Escolar Documentos
Profissional Documentos
Cultura Documentos
1.1 Introduction
This section covers the following topics. Basic Concept Early Attempts of Prestressing Brief History Development of Building Materials
The prestressing of a structure is not the only instance of prestressing. The concept of prestressing existed before the applications in concrete. Two examples of prestressing before the development of prestressed concrete are provided.
Force-fitting of metal bands on wooden barrels The metal bands induce a state of initial hoop compression, to counteract the hoop tension caused by filling of liquid in the barrels.
Metal bands
Figure 1-1.1
Pre-tensioning the spokes in a bicycle wheel The pre-tension of a spoke in a bicycle wheel is applied to such an extent that there will always be a residual tension in the spoke.
Spokes
Figure 1-1.2
For concrete, internal stresses are induced (usually, by means of tensioned steel) for the following reasons.
The tensile strength of concrete is only about 8% to 14% of its compressive strength.
Cracks tend to develop at early stages of loading in flexural members such as beams and slabs.
To prevent such cracks, compressive force can be suitably applied in the perpendicular direction.
Prestressing enhances the bending, shear and torsional capacities of the flexural members.
In pipes and liquid storage tanks, the hoop tensile stresses can be effectively counteracted by circular prestressing.
Mild steel rods are stretched and concrete is poured around them. After hardening of concrete, the tension in the rods is released. The rods will try to regain their original length, but this is prevented by the surrounding concrete to which the steel is bonded. Thus, the concrete is now effectively in a state of pre-compression. It is capable of counteracting tensile stress, such as arising from the load shown in the following sketch.
Figure 1-1.4
But, the early attempts of prestressing were not completely successful. It was observed that the effect of prestress reduced with time. The load resisting capacities of the members were limited. Under sustained loads, the members were found to fail. This was due to the following reason.
increases with increase in time. This is known as creep strain. The reduction in length due to creep and shrinkage is also applicable to the embedded steel, resulting in significant loss in the tensile strain.
In the early applications, the strength of the mild steel and the strain during prestressing were less. The residual strain and hence, the residual prestress was only about 10% of the initial value. The following sketches explain the phenomena.
c) Beam after long-term losses of prestress Figure 1-1.5 Variation of length in a prestressed beam
The residual strain in steel = original tensile strain in steel compressive strains corresponding to short-term and long-term losses.
Original tensile strain in steel Compressive strain due to elastic shortening of beam (short-term loss in prestress) Compressive strain due to creep and shrinkage (long-term losses in prestress) Therefore, residual strain in steel The maximum original tensile strain in mild steel
= (L2 L1)/L1 = (L2 L3)/L1 = (L3 L4)/L1 = (L4 L1)/L1 = Allowable stress / elastic modulus = 140 MPa / 2105 MPa = 0.0007
The total loss in strain due to elastic shortening, creep and shrinkage was also close to 0.0007. Thus, the residual strain was negligible.
The solution to increase the residual strain and the effective prestress are as follows.
Adopt high strength steel with much higher original strain. This leads to the scope of high prestressing force.
1824 Aspdin, J., (England) Obtained a patent for the manufacture of Portland cement.
1857 Monier, J., (France) Introduced steel wires in concrete to make flower pots, pipes, arches and slabs.
The following events were significant in the development of prestressed concrete. 1886 Jackson, P. H., (USA) Introduced the concept of tightening steel tie rods in artificial stone and concrete arches.
Figure 1-1.6
1888 Doehring, C. E. W., (Germany) Manufactured concrete slabs and small beams with embedded tensioned steel.
1908 Stainer, C. R., (USA) Recognised losses due to shrinkage and creep, and suggested retightening the rods to recover lost prestress.
1923 Emperger, F., (Austria) Developed a method of winding and pre- tensioning high tensile steel wires around concrete pipes.
1924 Hewett, W. H., (USA) Introduced hoop-stressed horizontal reinforcement around walls of concrete tanks through the use of turnbuckles.
Thousands of liquid storage tanks and concrete pipes were built in the two decades to follow. 1925 Dill, R. H., (USA) Used high strength unbonded steel rods. The rods were tensioned and anchored after hardening of the concrete.
Figure 1-1.7
1926 Eugene Freyssinet (France) Used high tensile steel wires, with ultimate strength as high as 1725 MPa and yield stress over 1240 MPa. In 1939, he developed conical wedges for end anchorages for post-tensioning and developed double-acting jacks. He is often referred to as the Father of Prestressed concrete.
1940 Magnel, G., (Belgium) Developed an anchoring system for post-tensioning, using flat wedges.
During the Second World War, applications of prestressed and precast concrete increased rapidly. The names of a few persons involved in developing prestressed concrete are mentioned. Guyon, Y., (France) built numerous prestressed concrete Abeles, P. W., (England) introduced the
concept of partial prestressing. Leonhardt, F., (Germany), Mikhailor, V., (Russia) and Lin, T. Y., (USA) are famous in the field of prestressed concrete.
The International Federation for Prestressing (FIP), a professional organisation in Europe was established in 1952. The Precast/Prestressed Concrete Institute (PCI) was established in USA in 1954.
Prestressed concrete was started to be used in building frames, parking structures, stadiums, railway sleepers, transmission line poles and other types of structures and elements.
In India, the applications of prestressed concrete diversified over the years. The first prestressed concrete bridge was built in 1948 under the Assam Rail Link Project. Among bridges, the Pamban Road Bridge at Rameshwaram, Tamilnadu, remains a classic example of the use of prestressed concrete girders.
Figure 1-1.8
In reinforced concrete, concrete and steel are combined such that concrete resists compression and steel resists tension. This is a passive combination of the two
materials. In prestressed concrete high strength concrete and high strength steel are combined such that the full section is effective in resisting tension and compression. This is an active combination of the two materials. The following sketch shows the use of the different materials with the progress of time.
Active combination
Figure 1-1.9
Prestressed Concrete
1.2.1 Definitions
The terms commonly used in prestressed concrete are explained. The terms are placed in groups as per usage.
Forms of Prestressing Steel Wires Prestressing wire is a single unit made of steel. Strands Two, three or seven wires are wound to form a prestressing strand. Tendon A group of strands or wires are wound to form a prestressing tendon. Cable A group of tendons form a prestressing cable. Bars A tendon can be made up of a single steel bar. The diameter of a bar is much larger than that of a wire.
The different types of prestressing steel are further explained in Section 1.7, Prestressing Steel.
Nature of Concrete-Steel Interface Bonded tendon When there is adequate bond between the prestressing tendon and concrete, it is called a bonded tendon. Pre-tensioned and grouted post-tensioned tendons are bonded tendons.
Unbonded tendon When there is no bond between the prestressing tendon and concrete, it is called unbonded tendon. When grout is not applied after post-tensioning, the tendon is an unbonded tendon.
Stages of Loading The analysis of prestressed members can be different for the different stages of loading. The stages of loading are as follows.
1) Initial
: It can be subdivided into two stages. a) During tensioning of steel b) At transfer of prestress to concrete.
2) Intermediate
3) Final
: It can be subdivided into two stages. a) At service, during operation. b) At ultimate, during extreme events.
The following text broadly mentions the advantages of a prestressed concrete member with an equivalent RC member. For each effect, the benefits are listed.
Increase in durability.
Increase in shear capacity. Suitable for use in pressure vessels, liquid retaining structures. Improved performance (resilience) under dynamic and fatigue loading. 2) High span-to-depth ratios Larger spans possible with prestressing (bridges, buildings with large column-free spaces) Typical values of span-to-depth ratios in slabs are given below. Non-prestressed slab Prestressed slab 28:1 45:1
Reduction in self weight More aesthetic appeal due to slender sections More economical sections.
3) Suitable for precast construction The advantages of precast construction are as follows.
Rapid construction Better quality control Reduced maintenance Suitable for repetitive construction Multiple use of formwork
Reduction of formwork
T-section
Double T-section
Hollow core
Piles
L-section
Inverted T-section
I-girders
Figure 1-2.1
The use of high strength materials is costly. There is additional cost in auxiliary equipments. There is need for quality control and inspection.
Source of prestressing force This classification is based on the method by which the prestressing force is generated. There are four sources of prestressing force: Mechanical, hydraulic, electrical and chemical.
External or internal prestressing This classification is based on the location of the prestressing tendon with respect to the concrete section.
Pre-tensioning or post-tensioning This is the most important classification and is based on the sequence of casting the concrete and applying tension to the tendons.
Linear or circular prestressing This classification is based on the shape of the member prestressed.
Full, limited or partial prestressing Based on the amount of prestressing force, three types of prestressing are defined.
Uniaxial, biaxial or multi-axial prestressing As the names suggest, the classification is based on the directions of prestressing a member.
Source of Prestressing Force Hydraulic Prestressing This is the simplest type of prestressing, producing large prestressing forces. The
hydraulic jack used for the tensioning of tendons, comprises of calibrated pressure gauges which directly indicate the magnitude of force developed during the tensioning.
Mechanical Prestressing In this type of prestressing, the devices includes weights with or without lever transmission, geared transmission in conjunction with pulley blocks, screw jacks with or without gear drives and wire-winding machines. This type of prestressing is adopted for mass scale production.
Electrical Prestressing In this type of prestressing, the steel wires are electrically heated and anchored before placing concrete in the moulds. This type of prestressing is also known as thermoelectric prestressing.
External or Internal Prestressing External Prestressing When the prestressing is achieved by elements located outside the concrete, it is called external prestressing. The tendons can lie outside the member (for example in I-girders or walls) or inside the hollow space of a box girder. This technique is adopted in
bridges and strengthening of buildings. In the following figure, the box girder of a bridge is prestressed with tendons that lie outside the concrete.
Figure 1-2.2
Internal Prestressing When the prestressing is achieved by elements located inside the concrete member (commonly, by embedded tendons), it is called internal prestressing. Most of the applications of prestressing are internal prestressing. In the following figure, concrete will be cast around the ducts for placing the tendons.
Figure 1-2.3
Pre-tensioning or Post-tensioning Pre-tensioning The tension is applied to the tendons before casting of the concrete. The pre-
compression is transmitted from steel to concrete through bond over the transmission length near the ends. The following figure shows manufactured pre-tensioned electric poles.
Figure 1-2.4
Post-tensioning The tension is applied to the tendons (located in a duct) after hardening of the concrete. The pre-compression is transmitted from steel to concrete by the anchorage device (at the end blocks). The following figure shows a post-tensioned box girder of a bridge.
Figure 1-2.5
The details of pre-tensioning and post-tensioning are covered under Section 1.3, Pretensioning Systems and Devices, and Section 1.4, Post-tensioning Systems and Devices, respectively.
Linear or Circular Prestressing Linear Prestressing When the prestressed members are straight or flat, in the direction of prestressing, the prestressing is called linear prestressing. For example, prestressing of beams, piles, poles and slabs. The profile of the prestressing tendon may be curved. The following figure shows linearly prestressed railway sleepers.
Figure 1-2.6
Circular Prestressing When the prestressed members are curved, in the direction of prestressing, the prestressing is called circular prestressing. For example, circumferential prestressing of tanks, silos, pipes and similar structures. The following figure shows the containment structure for a nuclear reactor which is circularly prestressed.
Figure 1-2.7
Full, Limited or Partial Prestressing Full Prestressing When the level of prestressing is such that no tensile stress is allowed in concrete under service loads, it is called Full Prestressing (Type 1, as per IS:1343 - 1980).
Limited Prestressing When the level of prestressing is such that the tensile stress under service loads is within the cracking stress of concrete, it is called Limited Prestressing (Type 2).
Partial Prestressing When the level of prestressing is such that under tensile stresses due to service loads, the crack width is within the allowable limit, it is called Partial Prestressing (Type 3).
Uniaxial, Biaxial or Multiaxial Prestressing Uniaxial Prestressing When the prestressing tendons are parallel to one axis, it is called Uniaxial Prestressing. For example, longitudinal prestressing of beams.
Biaxial Prestressing When there are prestressing tendons parallel to two axes, it is called Biaxial Prestressing. The following figure shows the biaxial prestressing of slabs.
Multiaxial Prestressing When the prestressing tendons are parallel to more than two axes, it is called Multiaxial Prestressing. For example, prestressing of domes.
1.3.1 Introduction
Prestressing systems have developed over the years and various companies have patented their products. Detailed information of the systems is given in the product catalogues and brochures published by companies. There are general guidelines of prestressing in Section 12 of IS:1343 - 1980. The information given in this section is introductory in nature, with emphasis on the basic concepts of the systems.
The prestressing systems and devices are described for the two types of prestressing, pre-tensioning and post-tensioning, separately. This section covers pre-tensioning.
Section 1.4, Post-tensioning Systems and Devices, covers post-tensioning. In pretensioning, the tension is applied to the tendons before casting of the concrete. The stages of pre-tensioning are described next.
Once the concrete attains the desired strength for prestressing, the tendons are cut loose from the abutments.
The prestress is transferred to the concrete from the tendons, due to the bond between them. During the transfer of prestress, the member undergoes elastic shortening. If the tendons are located eccentrically, the member is likely to bend and deflect (camber). The various stages of the pre-tensioning operation are summarised as follows. 1) Anchoring of tendons against the end abutments 2) Placing of jacks 3) Applying tension to the tendons 4) Casting of concrete 5) Cutting of the tendons.
During the cutting of the tendons, the prestress is transferred to the concrete with elastic shortening and camber of the member.
A prestressing bed is required for the pre-tensioning operation. There is a waiting period in the prestressing bed, before the concrete attains sufficient strength.
There should be good bond between concrete and steel over the transmission length.
1.3.5 Devices
The essential devices for pre-tensioning are as follows.
Prestressing bed End abutments Shuttering / mould Jack Anchoring device Harping device (optional)
Prestressing Bed, End Abutments and Mould The following figure shows the devices.
End abutment Mould Jack Prestressing bed
Anchoring device
Figure1-3.2
An extension of the previous system is the Hoyer system. This system is generally used for mass production. The end abutments are kept sufficient distance apart, and several members are cast in a single line. The shuttering is provided at the sides and
between the members. This system is also called the Long Line Method. following figure is a schematic representation of the Hoyer system
The
A series of moulds
Prestressing bed
Figure 1-3.3 Schematic representation of Hoyer system
The end abutments have to be sufficiently stiff and have good foundations. This is usually an expensive proposition, particularly when large prestressing forces are required. The necessity of stiff and strong foundation can be bypassed by a simpler solution which can also be a cheaper option. It is possible to avoid transmitting the heavy loads to foundations, by adopting self-equilibrating systems. This is a common solution in load-testing. Typically, this is done by means of a tension frame. The
following figure shows the basic components of a tension frame. The jack and the specimen tend to push the end members. But the end members are kept in place by members under tension such as high strength steel rods.
Loading jack High strength steel rods
Test specimen
Plan or Elevation
Free bodies
Figure 1-3.4
A tension frame
The frame that is generally adopted in a pre-tensioning system is called a stress bench. The concrete mould is placed within the frame and the tendons are stretched and anchored on the booms of the frame. The following figures show the components of a stress bench.
Jack
Threaded rod
Plan
Figure 1-3.5 Stress bench Self straining frame
The following figure shows the free body diagram by replacing the jacks with the applied forces.
Load by jack
Plan
Tension in strands
Figure 1-3.6
The following figure shows the stress bench after casting of the concrete.
Elevation
Plan
Figure 1-3.7 The stress bench after casting concrete
Jacks The jacks are used to apply tension to the tendons. Hydraulic jacks are commonly used. These jacks work on oil pressure generated by a pump. The principle behind the design of jacks is Pascals law. The load applied by a jack is measured by the pressure reading from a gauge attached to the oil inflow or by a separate load cell. The following figure shows a double acting hydraulic jack with a load cell.
Figure 1-3.8
Anchoring Devices Anchoring devices are often made on the wedge and friction principle. In pre-tensioned members, the tendons are to be held in tension during the casting and hardening of concrete. Here simple and cheap quick-release grips are generally adopted. following figure provides some examples of anchoring devices. The
Figure 1-3.9
Harping Devices The tendons are frequently bent, except in cases of slabs-on-grade, poles, piles etc. The tendons are bent (harped) in between the supports with a shallow sag as shown below.
Harping point
Hold up device
The tendons are harped using special hold-down devices as shown in the following figure.
Figure 1-3.11
Meanwhile the coarse and fine aggregates are batched, mixed with cement, water and additives in a concrete mixer. The stress bench is moved beneath the concrete mixer. The concrete is poured through a hopper and the moulds are vibrated. After the
finishing of the surface, the stress bench is placed in a steam curing chamber for a few hours till the concrete attains a minimum strength.
The stress bench is taken out from the chamber and the strands are cut. The sleepers are removed from the moulds and stacked for curing in water. After the complete curing, the sleepers are ready for dispatching.
Extension of ram
Fine aggregate
Coarse aggregate
(e) Material storage
(n) Storage and dispatching of sleepers Figure 1-3.12 Manufacturing of pre-tensioned railway sleepers
1.4.1 Introduction
Prestressing systems have developed over the years and various companies have patented their products. Detailed information of the systems is given in the product catalogues and brochures published by companies. There are general guidelines of prestressing in Section 12 of IS 1343: 1980. The information given in this section is introductory in nature, with emphasis on the basic concepts of the systems.
The prestressing systems and devices are described for the two types of prestressing, pre-tensioning and post-tensioning, separately. This section covers post-tensioning. Section 1.3, Pre-tensioning Systems and Devices, covers pre-tensioning. In posttensioning, the tension is applied to the tendons after hardening of the concrete. The stages of post-tensioning are described next.
Unlike pre-tensioning, the tendons are pulled with the reaction acting against the hardened concrete.
If the ducts are filled with grout, then it is known as bonded post-tensioning. The grout is a neat cement paste or a sand-cement mortar containing suitable admixture. The grouting operation is discussed later in the section. discussed in Section 1.6, Concrete (Part-II). The properties of grout are
In unbonded post-tensioning, as the name suggests, the ducts are never grouted and the tendon is held in tension solely by the end anchorages. The following sketch shows a schematic representation of a grouted post-tensioned member. The profile of the duct depends on the support conditions. For a simply supported member, the duct has a sagging profile between the ends. For a continuous member, the duct sags in the span and hogs over the support.
Figure 1-4.1
Among the following figures, the first photograph shows the placement of ducts in a box girder of a simply supported bridge. The second photograph shows the end of the box girder after the post-tensioning of some tendons.
Figure 1-4.2
Figure 1-4.3
The various stages of the post-tensioning operation are summarised as follows. 1) Casting of concrete. 2) Placement of the tendons. 3) Placement of the anchorage block and jack. 4) Applying tension to the tendons. 5) Seating of the wedges. 6) Cutting of the tendons.
The stages are shown schematically in the following figures. After anchoring a tendon at one end, the tension is applied at the other end by a jack. The tensioning of tendons and pre-compression of concrete occur simultaneously. A system of self-equilibrating forces develops after the stretching of the tendons.
Duct
Casting bed
(a) Casting of concrete
Side view
Jack
(b) Tensioning of tendons
Anchor
(c) Anchoring the tendon at the stretching end Figure 1-4.4 Stages of post-tensioning (shown in elevation)
1.4.5 Devices
The essential devices for post-tensioning are as follows. 1) Casting bed 2) Mould/Shuttering 3) Ducts
Casting Bed, Mould and Ducts The following figure shows the devices.
Anchoring Devices In post-tensioned members the anchoring devices transfer the prestress to the concrete. The devices are based on the following principles of anchoring the tendons. 1) Wedge action 2) Direct bearing 3) Looping the wires
Wedge action The anchoring device based on wedge action consists of an anchorage block and wedges. The strands are held by frictional grip of the wedges in the anchorage block. Some examples of systems based on the wedge-action are Freyssinet, Gifford-Udall, Anderson and Magnel-Blaton anchorages. The following figures show some patented anchoring devices.
Figure 1-4.6
Figure 1-4.7
Anchoring devices
Figure 1-4.8
Direct bearing The rivet or bolt heads or button heads formed at the end of the wires directly bear against a block. The B.B.R.V post-tensioning system and the Prescon system are
based on this principle. The following figure shows the anchoring by direct bearing.
Figure 1-4.9
Looping the wires The Baur-Leonhardt system, Leoba system and also the Dwidag single-bar anchorage system, work on this principle where the wires are looped around the concrete. The
wires are looped to make a bulb. The following photo shows the anchorage by looping of the wires in a post-tensioned slab.
Figure 1-4.10
The anchoring devices are tested to calculate their strength. The following photo shows the testing of an anchorage block.
Figure 1-4.11
Sequence of Anchoring The following figures show the sequence of stressing and anchoring the strands. The photo of an anchoring device is also provided.
Figure 1-4.12
Sequence of anchoring
Figure 1-4.13
(Reference: VSL International Ltd) Jacks The working of a jack and measuring the load were discussed in Section 1.3, Pretensioning Systems and Devices. The following figure shows an extruded sketch of the anchoring devices.
Figure 1-4.14
Couplers The couplers are used to connect strands or bars. They are located at the junction of the members, for example at or near columns in post-tensioned slabs, on piers in posttensioned bridge decks.
The couplers are tested to transmit the full capacity of the strands or bars. A few types of couplers are shown.
Figure 1-4.15
Figure 1-4.16
Figure 1-4.17
Grouting Grouting can be defined as the filling of duct, with a material that provides an anticorrosive alkaline environment to the prestressing steel and also a strong bond between the tendon and the surrounding grout.
The major part of grout comprises of water and cement, with a water-to-cement ratio of about 0.5, together with some water-reducing admixtures, expansion agent and pozzolans. The properties of grout are discussed in Section 1.6, Concrete (Part-II). The following figure shows a grouting equipment, where the ingredients are mixed and the grout is pumped.
Figure 1-4.18
Grouting equipment
(c) Stretching and anchoring of tendons Figure 1-4.19 Manufacturing of a post-tensioned bridge I-girder
The following photos show the construction of post-tensioned box girders for a bridge (Courtesy: Cochin Port Trust). The first photo shows the fabricated steel reinforcement with the ducts for the tendons placed inside. The top flange will be constructed later. The second photo shows the formwork in the pre-casting yard. The formwork for the inner sides of the webs and the flanges is yet to be placed. In the third photo a girder is being post-tensioned after adequate curing. The next photo shows a crane on a barge that transports a girder to the bridge site. The completed bridge can be seen in the last photo.
(e) Completed bridge Figure 1-4.20 Manufacturing of post-tensioned bridge box girders
The following figure shows a petrographic section of concrete. coarse aggregates and the matrix surrounding them. hydrated cement and tiny voids.
Figure 1-5.1
Aggregate The coarse aggregate are granular materials obtained from rocks and crushed stones. They may be also obtained from synthetic material like slag, shale, fly ash and clay for use in light-weight concrete.
The sand obtained from river beds or quarries is used as fine aggregate. The fine aggregate along with the hydrated cement paste fill the space between the coarse aggregate.
The important properties of aggregate are as follows. 1) Shape and texture 2) Size gradation 3) Moisture content 4) Specific gravity 5) Unit weight 6) Durability and absence of deleterious materials.
The nominal maximum coarse aggregate size is limited by the lowest of the following quantities. 1) 1/4 times the minimum thickness of the member 2) Spacing between the tendons/strands minus 5 mm 3) 40 mm.
The deleterious substances that should be limited in aggregate are clay lumps, wood, coal, chert, silt, rock dust (material finer than 75 microns), organic material, unsound and friable particles.
Cement In present day concrete, cement is a mixture of lime stone and clay heated in a kiln to 1400 - 1600C. The types of cement permitted by IS:1343 - 1980 (Clause 4.1) for prestressed applications are the following. The information is revised as per IS:456 2000, Plain and Reinforced Concrete Code of Practice. 1) Ordinary Portland cement confirming to IS:269 - 1989, Ordinary Portland Cement, 33 Grade Specification. 2) Portland slag cement confirming to IS:455 - 1989, Portland Slag Cement Specification, but with not more than 50% slag content. 3) Rapid-hardening Portland cement confirming to IS:8041 - 1990, Rapid Hardening Portland Cement Specification.
Indian Institute of Technology Madras
Water The water should satisfy the requirements of Section 5.4 of IS:456 - 2000.
Water used for mixing and curing shall be clean and free from injurious amounts of oils, acids, alkalis, salts, sugar, organic materials or other substances that may be deleterious to concrete and steel.
Admixtures IS:1343 - 1980 allows to use admixtures that conform to IS:9103 - 1999, Concrete Admixtures Specification. The admixtures can be broadly divided into two types: chemical admixtures and mineral admixtures. The common chemical admixtures are as follows. 1) Air-entraining admixtures 2) Water reducing admixtures 3) Set retarding admixtures 4) Set accelerating admixtures 5) Water reducing and set retarding admixtures 6) Water reducing and set accelerating admixtures.
The common mineral admixtures are as follows. 1) Fly ash 2) Ground granulated blast-furnace slag 3) Silica fumes 4) Rice husk ash 5) Metakoline These are cementitious and pozzolanic materials.
3) Minimum shrinkage and creep by limiting the cement content. The following topics are discussed. 1) Strength of concrete 2) Stiffness of concrete 3) Durability of concrete 4) High performance concrete 5) Allowable stresses in concrete. Strength of Concrete The following sections describe the properties with reference to IS:1343 - 1980. The strength of concrete is required to calculate the strength of the members. For
prestressed concrete applications, high strength concrete is required for the following reasons. 1) To sustain the high stresses at anchorage regions. 2) To have higher resistance in compression, tension, shear and bond. 3) To have higher stiffness for reduced deflection. 4) To have reduced shrinkage cracks.
Compressive Strength The compressive strength of concrete is given in terms of the characteristic compressive strength of 150 mm size cubes tested at 28 days (fck). The characteristic strength is defined as the strength of the concrete below which not more than 5% of the test results are expected to fall. This concept assumes a normal distribution of the strengths of the samples of concrete.
The following sketch shows an idealised distribution of the values of compressive strength for a sizeable number of test cubes. The horizontal axis represents the values of compressive strength. The vertical axis represents the number of test samples for a particular compressive strength. This is also termed as frequency. The average of the values of compressive strength (mean strength) is represented as fcm. The characteristic strength (fck) is the value in the x-axis below which 5% of the total area under the curve falls. The value of fck is lower than fcm by 1.65, where is the standard deviation of the normal distribution.
Frequency 1.65
The sampling and strength test of concrete are as per Section 15 of IS:1343 - 1980. The grades of concrete are explained in Table 1 of the Code.
The minimum grades of concrete for prestressed applications are as follows. 30 MPa for post-tensioned members 40 MPa for pre-tensioned members.
Since at the time of publication of IS:1343 in 1980, the properties of higher strength concrete were not adequately documented, a limit was imposed on the maximum strength. It is expected that higher strength concrete may be used after proper testing.
The increase in strength with age as given in IS:1343 - 1980, is not observed in present day concrete that gains substantial strength in 28 days. Hence, the age factor given in Clause 5.2.1 should not be used. It has been removed from IS:456 - 2000.
Tensile Strength The tensile strength of concrete can be expressed as follows. 1) Flexural tensile strength: It is measured by testing beams under 2 point loading (also called 4 point loading including the reactions). 2) Splitting tensile strength: It is measured by testing cylinders under diametral compression.
3) Direct tensile strength: It is measured by testing rectangular specimens under direct tension.
In absence of test results, the Code recommends to use an estimate of the flexural tensile strength from the compressive strength by the following equation.
fcr = 0.7 fck
(1-5.1)
Here, fcr = flexural tensile strength in N/mm2 fck = characteristic compressive strength of cubes in N/mm2. Stiffness of Concrete The stiffness of concrete is required to estimate the deflection of members. The stiffness is given by the modulus of elasticity. For a non-linear stress (fc) versus strain (c) behaviour of concrete the modulus can be initial, tangential or secant modulus. IS:1343 - 1980 recommends a secant modulus at a stress level of about 0.3fck. The modulus is expressed in terms of the characteristic compressive strength and not the design compressive strength. The following figure shows the secant modulus in the compressive stress-strain curve for concrete.
fc
fc fck
Ec
Figure 1-5.3
The modulus of elasticity for short term loading (neglecting the effect of creep) is given by the following equation.
Ec = 5000 fck
(1-5.2)
Here, Ec = short-term static modulus of elasticity in N/mm2 fck = characteristic compressive strength of cubes in N/mm2.
Indian Institute of Technology Madras
Durability of Concrete The durability of concrete is of vital importance regarding the life cycle cost of a structure. The life cycle cost includes not only the initial cost of the materials and labour, but also the cost of maintenance and repair.
In recent years emphasis has been laid on the durability issues of concrete. This is reflected in the enhanced section on durability (Section 8) in IS:456 - 2000. It is
expected that the revised version of IS:1343 will also have similar importance on durability.
The durability of concrete is defined as its ability to resist weathering action, chemical attack, abrasion, or any other process of deterioration. problems in concrete are as follows. 1) Sulphate and other chemical attacks of concrete. 2) Alkali-aggregate reaction. 3) Freezing and thawing damage in cold regions. 4) Corrosion of steel bars or tendons. The common durability
The durability of concrete is intrinsically related to its water tightness or permeability. Hence, the concrete should have low permeability and there should be adequate cover to reinforcing bars. The selection of proper materials and good quality control are
The durability is addressed in IS:1343 - 1980 in Section 7. In Appendix A there are guidelines on durability. Table 9 specifies the maximum water-to-cement (w-c) ratio and the minimum cement content for different exposure conditions. The values for moderate exposure condition are reproduced below. Table 1-5.1 Maximum water-to-cement (w-c) ratio and the minimum cement content for moderate exposure conditions (IS:1343 - 1980). Min. cement content Max w-c ratio* : 300 kg per m3 of concrete : 0.50
Table 10 provides the values for the above quantities for concrete exposed to sulphate attack.
To limit the creep and shrinkage, IS:1343 - 1980 specifies a maximum cement content of 530 kg per m3 of concrete (Clause 8.1.1).
High Performance Concrete With the advancement of concrete technology, high performance concrete is getting popular in prestressed applications. The attributes of high performance concrete are as follows. 1) High strength 2) Minimum shrinkage and creep 3) High durability 4) Easy to cast 5) Cost effective. Traditionally high performance concrete implied high strength concrete with higher cement content and low water-to-cement ratio. But higher cement content leads to autogenous and plastic shrinkage cracking and thermal cracking. At present durability is also given importance along with strength.
Some special types of high performance concrete are as follows. 1) High strength concrete 2) High workability concrete 3) Self-compacting concrete 4) Reactive powder concrete 5) High volume fly ash concrete 6) Fibre reinforced concrete
In a post-tensioned member, the concrete next to the anchorage blocks (referred to as end block) is subjected to high stress concentration. The type of concrete at the end blocks may be different from that at the rest of the member. Fibre reinforced concrete is used to check the cracking due to the bursting forces.
The following photo shows that the end blocks were cast separately with high strength concrete.
Indian Institute of Technology Madras
Figure 1-5.4
Allowable Stresses in Concrete The allowable stresses are used to analyse and design members under service loads. IS:1343 - 1980 specifies the maximum allowable compressive stresses for different grades of concrete under different loading conditions in Section 22.8.
Allowable Compressive Stresses under Flexure The following sketch shows the variation of allowable compressive stresses for different grades of concrete at transfer. The cube strength at transfer is denoted as fci.
0.54fci
0.37fci
0.51fci
0.44fci
M30 Post-tension
Figure 1-5.5
M60
M40
M60 Pre-tension
The following sketch shows the variation of allowable compressive stresses for different grades of concrete at service loads.
0.41fck 0.34fck M 30
Zone I Zone II
0.35fck 0.27fck M 60
Figure 1-5.6
Here, Zone I represents the locations where the compressive stresses are not likely to increase. Zone II represents the locations where the compressive stresses are likely to increase, such as due to transient loads from vehicles in bridge decks.
Allowable Compressive Stresses under Direct Compression For direct compression, except in the parts immediately behind the anchorage, the maximum stress is equal to 0.8 times the maximum compressive stress under flexure.
Allowable Tensile Stresses under Flexure The prestressed members are classified into three different types based on the allowable tensile stresses. The amount of prestressing varies in the three types. The allowable tensile stresses for the three types of members are specified in Section 22.7. The values are reproduced below. Table 1-5.2 Type 1 No tensile stress 3 N/mm2. Type 2 This value can be increased to 4.5 N/mm2 for temporary loads. Allowable tensile stresses (IS:1343 - 1980)
The purpose of providing hypothetical values is to use the elastic analysis method for Type 3 members even after cracking of concrete.
Stress-strain Curves for Concrete Curve under uniaxial compression The stress versus strain behaviour of concrete under uniaxial compression is initially linear (stress is proportional to strain) and elastic (strain is recovered at unloading). With the generation of micro-cracks, the behaviour becomes nonlinear and inelastic. After the specimen reaches the peak stress, the resisting stress decreases with increase in strain. IS:1343 - 1980 recommends a parabolic characteristic stress-strain curve, proposed by Hognestad, for concrete under uniaxial compression (Figure 3 in the Code).
fc
fc fck
0
Figure 1-6.1
cu c
a) Concrete cube under compression, b) Design stress-strain curve for concrete under compression due to flexure
The equation for the design curve under compression due to flexure is as follows. For c 0
2 fck = fck 2 c - c 0 0
fc = fck Here, fc = compressive stress fck = characteristic compressive strength of cubes c = compressive strain 0 = strain corresponding to fck = 0.002 cu = ultimate compressive strain = 0.0035
For concrete under compression due to axial load, the ultimate strain is restricted to 0.002. From the characteristic curve, the design curve is defined by multiplying the stress with a size factor of 0.67 and dividing the stress by a material safety factor of m = 1.5. The design curve is used in the calculation of ultimate strength. The following sketch shows the two curves.
fc fck
0.447 fck Characteristic curve Design curve
0
Figure 1-6.2
cu
In the calculation of deflection at service loads, a linear stress-strain curve is assumed up to the allowable stress. This curve is given by the following equation. fc = Ecc (1-6.3)
Note that, the size factor and the material safety factor are not used in the elastic modulus Ec.
For high strength concrete (say M100 grade of concrete and above) under uniaxial compression, the ascending and descending branches are steep.
fc fck
Es
Figure 1-6.3
Eci
0 c
The equation proposed by Thorenfeldt, Tomaxzewicz and Jensen is appropriate for high strength concrete. n c 0 fc = fck nk c n - 1+ 0 The variables in the previous equation are as follows. fc c 0 k n = compressive stress = compressive strain = strain corresponding to fck = 1 for c 0 = 0.67 + (fck / 77.5) for c > 0. The value of k should be greater than 1. = Eci / (Eci Es) Eci = initial modulus Es = secant modulus at fck = fck / 0. The previous equation is applicable for both the ascending and descending branches of the curve. Also, the parameter k models the slope of the descending branch, which increases with the characteristic strength fck. To be precise, the value of 0 can be considered to vary with the compressive strength of concrete. fck = characteristic compressive strength of cubes in N/mm2
(1-6.4)
Curve under uniaxial tension The stress versus strain behaviour of concrete under uniaxial tension is linear elastic initially. Close to cracking nonlinear behaviour is observed.
fc
fc
c
(a) Figure 1-6.4 under tension In calculation of deflections of flexural members at service loads, the nonlinearity is neglected and a linear elastic behaviour fc = Ecc is assumed. In the analysis of ultimate strength, the tensile strength of concrete is usually neglected. (b)
Creep of Concrete Creep of concrete is defined as the increase in deformation with time under constant load. Due to the creep of concrete, the prestress in the tendon is reduced with time. Hence, the study of creep is important in prestressed concrete to calculate the loss in prestress. The creep occurs due to two causes. 1. Rearrangement of hydrated cement paste (especially the layered products) 2. Expulsion of water from voids under load If a concrete specimen is subjected to slow compressive loading, the stress versus strain curve is elongated along the strain axis as compared to the curve for fast loading. This can be explained in terms of creep. If the load is sustained at a level, the increase in strain due to creep will lead to a shift from the fast loading curve to the slow loading curve (Figure 1-6.5).
fc
Figure 1-6.5
Creep is quantified in terms of the strain that occurs in addition to the elastic strain due to the applied loads. If the applied loads are close to the service loads, the creep strain increases at a decreasing rate with time. The ultimate creep strain is found to be proportional to the elastic strain. The ratio of the ultimate creep strain to the elastic strain is called the creep coefficient . For stress in concrete less than about one-third of the characteristic strength, the ultimate creep strain is given as follows.
cr,ult = el
(1-6.5)
The variation of strain with time, under constant axial compressive stress, is represented in the following figure.
strain
cr, ult = ultimate creep strain el = elastic strain Time (linear scale)
Figure 1-6.6
If the load is removed, the elastic strain is immediately recovered. However the recovered elastic strain is less than the initial elastic strain, as the elastic modulus increases with age. There is reduction of strain due to creep recovery which is less than the creep strain. There is some residual strain which cannot be recovered (Figure 1-6.7).
strain
Figure 1-6.7
The creep strain depends on several factors. It increases with the increase in the following variables. 1) Cement content (cement paste to aggregate ratio) 2) Water-to-cement ratio 3) Air entrainment 4) Ambient temperature. The creep strain decreases with the increase in the following variables. 1) Age of concrete at the time of loading. 2) Relative humidity 3) Volume to surface area ratio. The creep strain also depends on the type of aggregate. IS:1343 - 1980 gives guidelines to estimate the ultimate creep strain in Section 5.2.5. It is a simplified estimate where only one factor has been considered. The factor is age of loading of the prestressed concrete structure. The creep coefficient is provided for three values of age of loading. Table 1-6.1 Creep coefficient for three values of age of loading Age of Loading 7 days 28 days 1 year Creep Coefficient 2.2 1.6 1.1
It can be observed that if the structure is loaded at 7 days, the creep coefficient is 2.2. This means that the creep strain is 2.2 times the elastic strain. Thus, the total strain is more than thrice the elastic strain. Hence, it is necessary to study the effect of creep in the loss of prestress and deflection of prestressed flexural members. prestress and higher deflection. Curing the concrete adequately and delaying the application of load provide long term benefits with regards to durability, loss of prestress and deflection. In special situations detailed calculations may be necessary to monitor creep strain with time. Specialised literature or international codes can provide guidelines for such calculations. Even if the structure is loaded at 28 days, the creep strain is substantial. This implies higher loss of
Shrinkage of Concrete Shrinkage of concrete is defined as the contraction due to loss of moisture. The study of shrinkage is also important in prestressed concrete to calculate the loss in prestress. The shrinkage occurs due to two causes. 1. Loss of water from voids 2. Reduction of volume during carbonation The following figure shows the variation of shrinkage strain with time. Here, t0 is the time at commencement of drying. The shrinkage strain increases at a decreasing rate with time. The ultimate shrinkage strain (sh) is estimated to calculate the loss in prestress.
Shrinkage strain
sh
t0
Figure 1-6.8
Like creep, shrinkage also depends on several factors. The shrinkage strain increases with the increase in the following variables. 1) Ambient temperature 2) Temperature gradient in the members 3) Water-to-cement ratio 4) Cement content. The shrinkage strain decreases with the increase in the following variables. 1) Age of concrete at commencement of drying 2) Relative humidity 3) Volume to surface area ratio. The shrinkage strain also depends on the type of aggregate. IS:1343 - 1980 gives guidelines to estimate the shrinkage strain in Section 5.2.4. It is a simplified estimate of the ultimate shrinkage strain (sh). For pre-tension sh = 0.0003 For post-tension
sh = 0.0002 log10 (t + 2 )
(1-6.6)
(1-6.7)
Here, t is the age at transfer in days. Note that for post-tension, t is the age at transfer in days which approximates the curing time. It can be observed that with increasing age at transfer, the shrinkage strain reduces. As mentioned before, curing the concrete adequately and delaying the application of load provide long term benefits with regards to durability and loss of prestress. In special situations detailed calculations may be necessary to monitor shrinkage strain with time. Specialised literature or international codes can provide guidelines for such calculations.
Notations Geometric Properties The commonly used geometric properties of a prestressed member are defined as follows. Ac = Area of concrete section = Net cross-sectional area of concrete excluding the area of prestressing steel. Ap A = Area of prestressing steel = Total cross-sectional area of the tendons. = Area of prestressed member = Gross cross-sectional area of prestressed member. = Ac + Ap At = Transformed area of prestressed member = Area of the member when steel is substituted by an equivalent area of concrete. = Ac + mAp = A + (m 1)Ap Here, m = the modular ratio = Ep/Ec Ec = short-term elastic modulus of concrete Ep = elastic modulus of steel. The following figure shows the commonly used areas of the prestressed members.
A
Figure 2-1.1
Ac
Ap
At
CGC = Centroid of concrete = Centroid of the gross section. The CGC may lie outside the concrete (Figure 2-1.2). CGS = Centroid of prestressing steel = Centroid of the tendons. The CGS may lie outside the tendons or the concrete (Figure 2-1.2). I It = Moment of inertia of prestressed member = Second moment of area of the gross section about the CGC. = Moment of inertia of transformed section = Second moment of area of the transformed section about the centroid of the transformed section. e = Eccentricity of CGS with respect to CGC = Vertical distance between CGC and CGS. If CGS lies below CGC, e will be considered positive and vice versa (Figure 2-1.2).
CGC CGS
CGC CGS
= Initial prestressing force = The force which is applied to the tendons by the jack.
P0
= Prestressing force after immediate losses = The reduced value of prestressing force after elastic shortening, anchorage slip and loss due to friction.
Pe
= Effective prestressing force after time-dependent losses = The final value of prestressing force after the occurrence of creep, shrinkage and relaxation.
2.1.1 Introduction
In prestressed concrete applications, the most important variable is the prestressing force. In the early days, it was observed that the prestressing force does not stay constant, but reduces with time. Even during prestressing of the tendons and the transfer of prestress to the concrete member, there is a drop of the prestressing force from the recorded value in the jack gauge. The various reductions of the prestressing force are termed as the losses in prestress. The losses are broadly classified into two groups, immediate and time-dependent. The immediate losses occur during prestressing of the tendons and the transfer of prestress to the concrete member. The time-dependent losses occur during the service life of the prestressed member. The losses due to elastic shortening of the member, friction at the tendon-concrete interface and slip of the anchorage are the immediate losses. The losses due to the shrinkage and creep of the concrete and relaxation of the steel are the time-dependent losses. The causes of the various losses in prestress are shown in the following chart. Losses
Immediate
Time dependent
Elastic shortening
Friction
Anchorage slip
Creep
Shrinkage
Relaxation
Figure 2-1.3
Post-tensioned Members If there is only one tendon, there is no loss because the applied prestress is recorded after the elastic shortening of the member. For more than one tendon, if the tendons are stretched sequentially, there is loss in a tendon during subsequent stretching of the other tendons. The elastic shortening loss is quantified by the drop in prestress (fp) in a tendon due to the change in strain in the tendon (p). It is assumed that the change in strain in the tendon is equal to the strain in concrete (c) at the level of the tendon due to the prestressing force. This assumption is called strain compatibility between concrete and steel. The strain in concrete at the level of the tendon is calculated from the stress in concrete (fc) at the same level due to the prestressing force. relationship is used to calculate the strain from the stress. The quantification of the losses is explained below. A linear elastic
fp = Ep p = Epc f = Ep c Ec fp = mfc
(2-1.1)
For simplicity, the loss in all the tendons can be calculated based on the stress in concrete at the level of CGS. This simplification cannot be used when tendons are stretched sequentially in a post-tensioned member. The calculation is illustrated for the following types of members separately. Pre-tensioned Axial Members Pre-tensioned Bending Members Post-tensioned Axial Members
Pre-tensioned Axial Members The following figure shows the changes in length and the prestressing force due to elastic shortening of a pre-tensioned axial member.
Pi
P0
Elastic shortening of a pre-tensioned axial member
Figure 2-1.4
The loss can be calculated as per Eqn. (2-1.1) by expressing the stress in concrete in terms of the prestressing force and area of the section as follows.
fp = mfc P = m 0 Ac P f p = m i At Pi m A
(2-1.2)
Note that the stress in concrete due to the prestressing force after immediate losses (P0/Ac) can be equated to the stress in the transformed section due to the initial prestress (Pi /At). This is derived below. Further, the transformed area At of the prestressed member can be approximated to the gross area A. The following figure shows that the strain in concrete due to elastic shortening (c) is the difference between the initial strain in steel (pi) and the residual strain in steel (p0).
Pi
p 0
P0
Strain variables in elastic shortening
Figure 2-1.5
The following equation relates the strain variables. c = pi - p0 The strains can be expressed in terms of the prestressing forces as follows. (2-1.3)
c =
pi = p0 =
P0 Ac Ec
Pi Ap E p P0 Ap E p
(2-1.4)
(2-1.5)
(2-1.6)
or
P0 Pi = Ac At
(2-1.7)
Thus, the stress in concrete due to the prestressing force after immediate losses (P0/Ac) can be equated to the stress in the transformed section due to the initial prestress (Pi /At).
The following problem illustrates the calculation of loss due to elastic shortening in an idealised pre-tensioned railway sleeper.
Example 2-1.1
A prestressed concrete sleeper produced by pre-tensioning method has a rectangular cross-section of 300mm 250 mm (b h). It is prestressed with 9 numbers of straight 7mm diameter wires at 0.8 times the ultimate strength of 1570 N/mm2. Estimate the percentage loss of stress due to elastic shortening of
concrete. Consider m = 6.
40 250
40 300
Solution
a) Approximate solution considering gross section The sectional properties are calculated as follows. Area of a single wire,
Aw = /4 72
y=
115.5
Prestressing force,
Pi A
Pi .e y I
Pi Pi .e y A I
Since the wires are distributed above and below the CGC, the losses are calculated for the top and bottom wires separately. Stress at level of top wires (y = yt = 125 40)
P P .e ( fc )t = - i + i y t
= -4.9 N/mm2
= -6.7 N/mm2
Loss of prestress in bottom wires = 6 6.7 (5 38.48) = 7734.48 N Total loss of prestress = 4525 + 7735 = 12259.73 N 12.3 kN Percentage loss = (12.3 / 435) 100% = 2.83% b) Accurate solution considering transformed section. Transformed area of top steel,
A1 = (6 1) 4 38.48
(fc)b = -
435 103 (435 103 9.3)84.8 76.73 103 4.02108 = -5.67 - 0.85 = -6.52 N/mm2
(fc)t = -
435 103 (435 103 9.3)85.2 + 76.73 103 4.02108 = -5.67 + 0.86 = -4.81N/mm2
Loss of prestress in bottom wires = 6 6.52 (5 38.48) = 7527 N Total loss = 4442 + 7527 = 11969 N 12 kN Percentage loss = (12 / 435) 100% = 2.75 %
Indian Institute of Technology Madras
It can be observed that the accurate and approximate solutions are close. Hence, the simpler calculations based on A and I is acceptable.
The following figure shows the changes in length and the prestressing force due to elastic shortening of a pre-tensioned bending member.
Pi
wsw (self-weight)
Figure 2-1.6
Due to the effect of self-weight, the stress in concrete varies along length (Figure 2-1.6). The loss can be calculated by Eqn. (2-1.1) with a suitable evaluation of the stress in concrete. To have a conservative estimate of the loss, the maximum stress at the level of CGS at the mid-span is considered.
M e Pi Pe.e - i + sw A I I
fc = -
(2-1.8)
Here, Msw is the moment at mid-span due to self-weight. Precise result using At and It in place of A and I, respectively, is not computationally warranted. expression, the eccentricity of the CGS (e) was assumed to be constant. For a large member, the calculation of the loss can be refined by evaluating the strain in concrete at the level of the CGS accurately from the definition of strain. This is demonstrated later for post-tensioned bending members. In the above
For more than one tendon, if the tendons are stretched sequentially, there is loss in a tendon during subsequent stretching of the other tendons. The loss in each tendon can
The loss in the first tendon is evaluated precisely and half of that value is used as an average loss for all the tendons.
f p = 1 f p 1 2 1 = mfc1 2 1 n P = m i,j 2 j=2 A
(2-1.9)
Here,
Pi,j = initial prestressing force in tendon j n = number of tendons
The calculation of loss for tendons stretched sequentially, is similar to post-tensioned axial members. For curved profiles, the eccentricity of the CGS and hence, the stress in concrete at the level of CGS vary along the length. An average stress in concrete can be considered. For a parabolic tendon, the average stress (fc,avg) is given by the following equation.
fc,avg = fc1 + Here, fc1 = stress in concrete at the end of the member fc2 = stress in concrete at the mid-span of the member. A more rigorous analysis of the loss can be done by evaluating the strain in concrete at the level of the CGS accurately from the definition of strain. This is demonstrated for a beam with two parabolic tendons post-tensioned sequentially. In Figure 2-1.7, Tendon B is stretched after Tendon A. The loss in Tendon A due to elastic shortening during tensioning of Tendon B is given as follows. 2 ( fc 2 - fc1 ) 3
(2-1.10)
f p = E p c = E p [c 1 + c 2 ]
Indian Institute of Technology Madras
(2-1.11)
Here, c is the strain at the level of Tendon A. The component of c due to pure compression is represented as c1. The component of c due to bending is represented as c2. The two components are calculated as follows.
c 1 = c 2 = PB AEc
(2-1.12)
Here, A PB Ec L = cross-sectional area of beam = prestressing force in Tendon B = modulus of concrete = length of beam from left end I L = moment of inertia of beam = change in length of beam
(2-1.13)
(2-1.14)
eA1, eA2 = eccentricities of Tendon A at 1 (end) and 2 (centre), respectively. eB1, eB2 = eccentricities of Tendon B at 1 and 2, respectively.
Substituting the expressions of the eccentricities in Eqn. (2-1.12), the second component of the strain is given as follows. PB 1 2 8 = eA1eB1 + ( eA1eB 2 + eA2eB1 ) + eA2eB 2 Ec I 5 15 15
(2-1.15)
2.2.1 Friction
The friction generated at the interface of concrete and steel during the stretching of a curved tendon in a post-tensioned member, leads to a drop in the prestress along the member from the stretching end. The loss due to friction does not occur in pretensioned members because there is no concrete during the stretching of the tendons. The friction is generated due to the curvature of the tendon and the vertical component of the prestressing force. The following figure shows a typical profile (laying pattern) of the tendon in a continuous beam.
Figure 2-2.1
(Reference: VSL International Ltd.) In addition to friction, the stretching has to overcome the wobble of the tendon. The wobble refers to the change in position of the tendon along the duct. The losses due to friction and wobble are grouped together under friction. The formulation of the loss due to friction is similar to the problem of belt friction. The sketch below (Figure 2-2.2) shows the forces acting on the tendon of infinitesimal length dx.
d
R
N N P + dP
P d/2 P + dP
P
Figure 2-2.2 In the above sketch,
Force triangle
dx
Force acting in a tendon of infinitesimal length
P = prestressing force at a distance x from the stretching end R = radius of curvature d = subtended angle. The derivation of the expression of P is based on a circular profile. Although a cable profile is parabolic based on the bending moment diagram, the error induced is insignificant. The friction is proportional to the following variables.
Coefficient of friction () between concrete and steel. The resultant of the vertical reaction from the concrete on the tendon (N) generated due to curvature.
(2-2.1)
The friction over the length dx is equal to N = Pd. Thus the friction (dP) depends on the following variables.
Coefficient of friction () Curvature of the tendon (d) The amount of prestressing force (P)
Rigidity of sheathing Diameter of sheathing Spacing of sheath supports Type of tendon Type of construction
For a tendon of length dx, the friction due to wobble is expressed as kPdx, where k is the wobble coefficient or coefficient for wave effect. Based on the equilibrium of forces in the tendon for the horizontal direction, the following equation can be written. P = P + dP + (Pd + kPdx) or, dP = (Pd + kPdx)
(2-2.2)
Thus, the total drop in prestress (dP) over length dx is equal to (Pd + kPdx). The above differential equation can be solved to express P in terms of x.
Px x dP = - d + k dx 0 P P 0
or, lnP P = - ( + kx )
Px
0
Px = - ( + kx ) P0 Px = P0e -( +kx )
(2-2.3)
P0 = the prestress at the stretching end after any loss due to elastic shortening. For small values of + kx, the above expression can be simplified by the Taylor series expansion. Px = P0 (1 kx)
(2-2.4)
Thus, for a tendon with single curvature, the variation of the prestressing force is linear with the distance from the stretching end. The following figure shows the variation of
Indian Institute of Technology Madras
prestressing force after stretching. The left side is the stretching end and the right side is the anchored end.
P0
Figure 2-2.3
Px
Variation of prestressing force after stretching
In the absence of test data, IS:1343 - 1980 provides guidelines for the values of and k.
Table 2-2.1
Type of interface For steel moving on smooth concrete For steel moving on steel fixed to duct For steel moving on lead
The value of k varies from 0.0015 to 0.0050 per meter length of the tendon depending on the type of tendon. The following problem illustrates the calculation of the loss due to friction in a post-tensioned beam.
Example 2-2.1
A post-tensioned beam 100 mm 300 mm (b h) spanning over 10 m is stressed by successive tensioning and anchoring of 3 cables A, B, and C respectively as shown in figure. Each cable has cross section area of 200 mm2 and has initial stress of 1200 MPa. If the cables are tensioned from one end, estimate the percentage loss in each cable due to friction at the anchored end. Assume = 0.35, k = 0.0015 / m.
Solution
Prestress in each tendon at stretching end = 1200 200 = 240 kN. To know the value of (L), the equation for a parabolic profile is required.
dy 4ym = (L - 2 x ) dx L2
L ym y
Here, ym = displacement of the CGS at the centre of the beam from the ends L x y = length of the beam = distance from the stretching end = displacement of the CGS at distance x from the ends.
(L)
An expression of (x) can be derived from the change in slope of the profile. The slope of the profile is given as follows.
dy 4ym = (L - 2 x ) dx L2 At x = 0, the slope dy/dx = 4ym/L. The change in slope (x) is proportional to x. The expression of (x) can be written in terms of x as (x) = .x, where, = 8ym/L2. The variation is shown in the following sketch.
8ym/L 4ym/L 0
L/2
The total subtended angle over the length L is 8ym/L. The prestressing force Px at a distance x is given by Px = P0e( + kx) = P0ex where, x = + kx For cable A, ym = 0.1 m. For cable B, ym = 0.05 m. For cable C, ym = 0.0 m. For all the cables, L = 10 m. Substituting the values of ym and L
0.0043x for cable A x = 0.0029x for cable B 0.0015x for cable C
The maximum loss for all the cables is at x = L = 10, the anchored end.
0.958 for cable A = 0.971 for cable B 0.985 for cable C
-L
240 kN
Variation of prestressing forces The loss due to friction can be considerable for long tendons in continuous beams with changes in curvature. The drop in the prestress is higher around the intermediate supports where the curvature is high. The remedy to reduce the loss is to apply the stretching force from both ends of the member in stages.
Anchorage System Freyssinet system 12 - 5mm strands 12 - 8mm strands Magnel system Dywidag system
Due to the setting of the anchorage block, as the tendon shortens, there is a reverse friction. Hence, the effect of anchorage slip is present up to a certain length (Figure 22.4). Beyond this setting length, the effect is absent. This length is denoted as lset.
P0 Px
Figure 2-2.4
Here, x = + kx denotes the total effect of friction and wobble. The plot of Px gives the force variation diagram. The initial part of the force variation diagram, up to length lset is influenced by the setting of the anchorage block. Let the drop in the prestressing force at the stretching end be P. The determination of P and lset are necessary to plot the force variation diagram including the effect of the setting of the anchorage block. Considering the drop in the prestressing force and the effect of reverse friction, the magnitude of the prestressing force at a distance x from the stretching end is given as follows.
Px' = ( P0 - P ) e 'x
(2-2.6)
Here, for reverse friction is analogous to for friction and wobble. At the end of the setting length (x = lset), Px = Px
Px P0
P
Px
Px lset
Figure 2-2.5
Substituting the expressions of Px and Px for x = lset Since it is difficult to measure separately, is taken equal to . The expression of P simplifies to the following. P0e -lset = ( P0 - P ) e 'lset P0e - ( + ' )lset = P0 - P P0 1- ( + ' ) l set = P0 - P ' P = P0 ( + ' ) l set = P0 l set 1+ P = 2P0lset The following equation relates lset with the anchorage slip s.
(2-2.7)
(2-2.8)
s = s =
Transposing the terms,
(2-2.9)
2 lset = s
s Ap E p P0
Therefore,
lset =
s Ap E p P0
(2-2.10)
The term P0 represents the loss of prestress per unit length due to friction. The force variation diagram is used when stretching is done from both the ends. The tendons are overstressed to counter the drop due to anchorage slip. The stretching from both the ends can be done simultaneously or in stages. The final force variation is more uniform than the first stretching. The following sketch explains the change in the force variation diagram due to stretching from both the ends in stages.
Figure 2-2.6
The force variation diagrams for the various stages are explained. a) The initial tension at the right end is high to compensate for the anchorage slip. It corresponds to about 0.8 fpk initial prestress. The force variation diagram (FVD) is linear. b) After the anchorage slip, the FVD drops near the right end till the length lset. c) The initial tension at the left end also corresponds to about 0.8 fpk initial prestress. The FVD is linear up to the centre line of the beam. d) After the anchorage slip, the FVD drops near the left end till the length lset. It is observed that after two stages, the variation of the prestressing force over the length of the beam is less than after the first stage.
Example 2-2.2
A four span continuous bridge girder is post-tensioned with a tendon consisting of twenty strands with fpk = 1860 MPa. Half of the girder is shown in the figure below. The symmetrical tendon is simultaneously stressed up to 75% fpk from both ends and then anchored. The tendon properties are Ap = 2800 mm2, Ep = 195,000 MPa, = 0.20, K = 0.0020/m. The anchorage slip s = 6 mm. Calculate a) The expected elongation of the tendon after stretching, b) The force variation diagrams along the tendon before and after anchorage.
0.76
0.6
0.76
CL 0.6
13.7
13.7
3.7
15.2
15.2
3.7
Solution
Initial force at stretching end 0.75fpk = 1395 MPa
P0 = 0.75fpk Ap
= 3906 kN The continuous tendon is analysed as segments of parabola. The segments are The
identified between the points of maximum eccentricity and inflection points. segments are as follows: 1-2, 2-3, 3-4, 4-5, 5-6, 6-7 and 7-8. CL
3 4 5 6 7 8
inflection points are those where the curvature of the tendon reverses. The different
The following properties of parabolas are used. For segment 1-2, the parabola in the sketch below is used.
The change in slope from the origin to the end of the parabola is same as the slope at the end of the tendon which is = 2e/L, where
e1
e2
L
For the two parabolic segments joined at the inflection point as shown in the sketch above, the slope at the inflection point = 2(e1 + e2)/L. Here,
e1, e2 = eccentricities of the CGS at the span and support respectively L = length of the span
The summation is for the segments from the stretching end up to the point in the segment under consideration. denotes the point). Hence, the value of ( + kx) at the end of each segment is calculated to evaluate the prestressing force at that point (Px, where x
0.111
0.163 0.039
0.144 0.036
0.149
0.185
0.783 3058
Px (kN) 3906
The force variation diagram before anchorage can be plotted with the above values of
Px.
Since the
stretching is done at both the ends simultaneously, the diagram is symmetric about the central line. a) The expected elongation of the tendon after stretching First the product of the average force and the length of each segment is summed up to the centre line. PavL = 1 1 [3906 + 3718] 13.7 + [3718 + 3500] 13.7 2 2 1 1 + [3500 + 3363] 3 + [3363 + 3246] 3.7 2 2 1 1 + [3246 + 3058] 15.2 + [3058 + 2883] 15.2 2 2 1 + [ 2883 + 2718] 3.7 2 = 227612.2 kN
= =
Pav L AP EP
lset = =
s AP EP P0
Thus the value of the prestressing force at each stretching end after anchorage slip is 3906 424 = 3482 kN. The force variation diagram for lset = 15.46 m is altered to show the drop due to anchorage slip. The force variation diagrams before and after anchorage are shown below. Note that the drop of force per unit length is more over the supports due to change in curvature over a small distance.
4000 3500 3000 2500 0 20 40 60 80 Distance from end (m) After anchorage Before anchorage
Creep of Concrete Shrinkage of Concrete Relaxation of Steel Total Time Dependent Losses
characteristic strength, the ultimate creep strain (cr,ult) is found to be proportional to the elastic strain (el). The ratio of the ultimate creep strain to the elastic strain is defined as the ultimate creep coefficient or simply creep coefficient . The ultimate creep strain is then given as follows.
cr,ult = el
(2-3.1)
IS:1343 - 1980 gives guidelines to estimate the ultimate creep strain in Section 5.2.5. It is a simplified estimate where only one factor has been considered. The factor is age of loading of the prestressed concrete structure. The creep coefficient is provided for three values of age of loading. Curing the concrete adequately and delaying the application of load provide long term benefits time. with regards to durability, loss of prestress and deflection. In special situations detailed calculations may be necessary to monitor creep strain with Specialised literature or international codes can provide guidelines for such calculations. The loss in prestress (fp ) due to creep is given as follows. fp = Ep cr, ult (2-3.2)
Here, Ep is the modulus of the prestressing steel. The following considerations are applicable for calculating the loss of prestress due to creep. 1) The creep is due to the sustained (permanently applied) loads only. Temporary loads are not considered in the calculation of creep. 2) Since the prestress may vary along the length of the member, an average value of the prestress can be considered. 3) The prestress changes due to creep and the creep is related to the instantaneous prestress. To consider this interaction, the calculation of creep can be iterated over small time steps.
calculate the drop (or loss) in prestress (fp), the recommendations of IS:1343 - 1980 can be followed in absence of test data.
Example 2-3.1
A concrete beam of dimension 100 mm 300 mm is post-tensioned with 5 straight wires of 7mm diameter. The average prestress after short-term losses is 0.7fpk = 1200 N/mm2 and the age of loading is given as 28 days. Given that Ep = 200 103 MPa, Ec = 35000 MPa, find out the losses of prestress due to creep, shrinkage and relaxation. Neglect the weight of the beam in the computation of the stresses.
100
300 50 CGS
Solution
Area of concrete A = 100 300 = 30000 mm2 Moment of inertia of beam section I = 100 3003 / 12 = 225 106 mm4 Area of prestressing wires Ap = 5 (/4) 72 = 192.42 mm2 Prestressing force after short-term losses P0 = Ap.fp0 = 192.4 1200 = 230880 N
Indian Institute of Technology Madras
Modular ratio
Stress in concrete at the level of CGS fc = P0 P0e e A I 230880 230880 =502 4 6 310 22510
= 7.69 2.56 = 10.25 N/mm2 Loss of prestress due to creep (fp)cr = Ep cr, ult = Ep el = Ep (fc/Ec) = m fc = 5.71 10.25 1.6 = 93.64 N / mm2 Here, = 1.6 for loading at 28 days, from Table 2c-1 (Clause 5.2.5.1, IS:1343 - 1980). Shrinkage strain from Clause 5.2.4.1, IS:1343 - 1980 sh = 0.0002 / log10(t + 2) = 0.0002 / log10 (28 + 2) = 1.354 10-4 Loss of prestress due to shrinkage (fp)sh = Epsh = 2 105 1.354 10-4 = 27.08 N/mm2
From Table 2c-2 (Table 4, IS:1343 - 1980) Loss of prestress due to relaxation (fp)rl = 70.0 N/mm2 Loss of prestressing force Therefore, Loss of prestressing force due to creep = 93.64 192.42 = 18018 N Loss of prestressing force due to shrinkage = 27.08 192.42 = 5211 N Loss of prestressing force due to relaxation = 70 192.42 = 13469 N Total long-term loss of prestressing force (neglecting the interaction of the losses and prestressing force) = 18018 + 5211 + 13469 = 36698 N Percentage loss of prestress = 36698 / 230880 100% = 15.9 % = fp Ap
Concrete Institute (PCI) committee and is called the General method (Reference: PCI Committee, Recommendations for Estimating Prestress Losses, PCI Journal, PCI, Vol. 20, No. 4, July-August 1975, pp. 43-75). In the PCI step-by-step procedure, a minimum of four time steps are considered in the service life of a prestressed member. The following table provides the definitions of the time steps (Table 2-3.3). Table 2-3.3 Step 1 2 Time steps in the step-by-step procedure End Age of prestressing 30 days after prestressing or when subjected to superimposed load 3 4 End of Step 2 End of Step 3 1 year of service End of service life Beginning Pre-tension: Anchorage of steel Post-tension: End of curing End of Step 1
The step-by-step procedure can be implemented by a computer program, where the number of time steps can be increased. There are also approximate methods to calculate lump sum estimates of the total loss. Since these estimates are not given in IS:1343 - 1980, they are not mentioned here.
Introduction Analysis at Transfer Analysis at Service Loads Analysis of Ultimate Strength Analysis of Behaviour
Notations Geometric Properties A prestressed axial member may also have non-prestressed reinforcement to carry the axial force. This type of members is called partially prestressed members. The
commonly used geometric properties of a prestressed member with non-prestressed reinforcement are defined as follows. A Ac As Ap At = gross cross-sectional area = area of concrete = area of non-prestressed reinforcement = area of prestressing tendons = transformed area of the section = Ac + (Es/ Ec) As + (Ep/ Ec) Ap The following figure shows the commonly used areas of a prestressed member with non-prestressed reinforcement.
Ac
As
Ap
A
Figure 3-1.1
At
3.1.1 Introduction
The study of members under axial load gives an insight of the behaviour of a prestressed member as compared to an equivalent non-prestressed reinforced concrete member. Prestressed members under axial loads only, are uncommon. Members such as hangers and ties are subjected to axial tension. Members such as piles may have bending moment along with axial compression or tension. In this section, no
eccentricity of the CGS with respect to CGC is considered. The definitions of CGS and CGC are provided in Section 2.1, Losses in Prestress (Part I). The following figure shows members under axial loads.
Hangers
Figure 3-1.2
Piles
Members under axial load
The analysis of members refers to the evaluation of the following. 1) Permissible prestress based on allowable stresses at transfer. 2) Stresses under service loads. These are compared with allowable stresses under service conditions. 3) Ultimate strength. This is compared with the demand under factored loads. 4) The entire axial load versus deformation behaviour. The stages for loading are explained in Section 1.2, Advantages and Types of Prestressing
(3-1.1)
Here, P0 = prestress at transfer after short-term losses. In presence of non-prestressed reinforcement, the stress in the concrete can be calculated as follows.
fc = -
(3-1.2)
The permissible prestress is determined based on fc to be within the allowable stress at transfer.
(3-1.3)
Here, P = external axial force (In the equation, + for tensile force and vice versa.) Pe = effective prestress. If there is non-prestressed reinforcement, Ac is to be substituted by (Ac + (Es/Ec) As) and At is to be calculated including As. The value of fc should be within the allowable stress under service conditions.
(3-1.4a)
(3-1.4b)
In the previous equations, fy = characteristic yield stress for non-prestressed reinforcement with mild steel bars = characteristic 0.2% proof stress for non-prestressed reinforcement with high yield strength deformed bars. fpk = characteristic tensile strength of prestressing tendons. The ultimate tensile strength should be greater than the demand due to factored loads. The ultimate compressive strength of a section (PuR) can be calculated in presence of moments by the use of interaction diagrams. For a member under compression with minimum eccentricity, the ultimate strength is given as follows. Here, the contribution of prestressing steel is neglected. PuR = 0.4 fckAc + 0.67 fy As (3-1.5)
Equilibrium Equation At any instant, the equilibrium is given by the following equation. P = Acfc + Asfs + Apfp Here, fc = stress in concrete fs = stress in non-prestressed reinforcement fp = stress in prestressed tendons P = axial force. (3-1.6)
Compatibility Equations For non-prestressed reinforcement s = c For prestressed tendons p = c + p Here, c s p = strain in concrete at the level of the steel = strain in non-prestressed reinforcement = strain in prestressed tendons (3-1.8) (3-1.7)
p = strain difference in prestressed tendons with adjacent concrete The strain difference (p) is the strain in the prestressed tendons when the concrete has zero strain (c = 0). This occurs when the strain due to the external tensile axial load balances the compressive strain due to prestress. At any load stage, p = pe ce Here, pe = strain in tendons due to Pe, the prestress at service ce = strain in concrete due to Pe. The strain difference is further explained in Section 3.4, Analysis of Member under Flexure (Part III). (3-1.9)
Constitutive Relationships The constitutive relationships can be expressed in the following forms based on the material stress-strain curves shown in Section 1.6, Concrete (Part II), and Section 1.7, Prestressing Steel.
Indian Institute of Technology Madras
For concrete under compression fc = F1 (c) For prestressing steel fp = F2 (p) For reinforcing steel fs = F3 (s) The stress versus strain curve for concrete is shown below. (3-1.12) The first and third (3-1.11) (3-1.10)
fc
c
Figure 3-1.3
The stress versus strain curve for prestressing steel is as shown below.
fp
p
Figure 3-1.4 Stress versus strain for prestressing steel
fs
Figure 3-1.5
The equilibrium and compatibility equations and the constitutive relationships can be solved to develop the axial force versus deformation curve. The deformation can be calculated as cL, where L is the length of the member. The following plot shows the axial force versus deformation curves for prestressed and non-prestressed sections. The two sections are equivalent in their ultimate tensile strengths.
Axial force Cracking Tensile strengths
Deformation
Figure 3-1.6
From the previous plot, the following can be inferred. 1) Prestressing increases the cracking load. 2) Prestressing shifts the curve from the origin.
For the prestressed member, there is a compressive deformation in absence
3) For a given tensile load, the deformation of the prestressed member is smaller.
Prestressing reduces deformation at service loads.
4) For a given compressive load, the deformation of the prestressed member is larger.
Prestressing is detrimental for the response under compression.
6) For a partially prestressed section with the same ultimate strength, the axial load versus deformation curve will lie in between the curves for prestressed and nonprestressed sections.
3.2.1 Introduction
Similar to members under axial load, the analysis of members under flexure refers to the evaluation of the following. 1) Permissible prestress based on allowable stresses at transfer. 2) Stresses under service loads. These are compared with allowable stresses under service conditions. 3) Ultimate strength. This is compared with the demand under factored loads. 4) The entire load versus deformation behaviour.
The analyses at transfer and under service loads are presented in this section. The analysis for the ultimate strength is presented separately in Section 3.4, Analysis of Member under Flexure (Part III). The evaluation of the load versus deformation behaviour is required in special type of analysis. This analysis will not be covered in this section.
Assumptions The analysis of members under flexure considers the following. 1) Plane sections remain plane till failure (known as Bernoullis hypothesis). 2) Perfect bond between concrete and prestressing steel for bonded tendons.
Principles of Mechanics The analysis involves three principles of mechanics. 1) Equilibrium of internal forces with the external loads. The compression in concrete (C) is equal to the tension in the tendon (T). The couple of C and T are equal to the moment due to external loads. 2) Compatibility of the strains in concrete and in steel for bonded tendons. The formulation also involves the first assumption of plane section remaining plane after bending. For unbonded tendons, the compatibility is in terms of deformation.
3) Constitutive relationships relating the stresses and the strains in the materials.
Variation of Internal Forces In reinforced concrete members under flexure, the values of compression in concrete (C) and tension in the steel (T) increase with increasing external load. The change in the lever arm (z) is not large. In prestressed concrete members under flexure, at transfer of prestress C is located close to T. The couple of C and T balance only the self weight. At service loads, C shifts up and the lever arm (z) gets large. The variation of C or T is not appreciable.
The following figure explains this difference schematically for a simply supported beam under uniform load.
w1 z1 C1 T1 w 2 > w1 C2 T2
w1 z1 C1 T1
w2 z2
w2 z2
Prestressed concrete C2 C1, z2 > z1
C2 T2
In the above figure, C1, T1 = compression and tension at transfer due to self weight C2, T2 = compression and tension under service loads w1 = self weight w2 = service loads z1 = lever arm at transfer z2 = lever arm under service loads.
For the reinforced concrete member C2 is substantially large than C1, but z2 is close to z1. For the prestressed concrete member C2 is close to C1, but z2 is substantially large than z1.
There are three approaches to analyse a prestressed member at transfer and under service loads. These approaches are based on the following concepts. a) Based on stress concept. b) Based on force concept. c) Based on load balancing concept. The following material explains the three concepts.
Based on Stress Concept In the approach based on stress concept, the stresses at the edges of the section under the internal forces in concrete are calculated. The stress concept is used to compare the calculated stresses with the allowable stresses. The following figure shows a simply supported beam under a uniformly distributed load (UDL) and prestressed with constant eccentricity (e) along its length.
CGC CGS
Figure 3-2.2
e
A simply supported beam under UDL
The following sketch shows the internal forces in concrete at a section and the corresponding stress profiles. The first stress profile is due to the compression P. The second profile is due to the eccentricity of the compression. The third profile is due to the moment. At transfer, the moment is due to self weight. At service the moment is due to service loads.
CGC
P
(stress due to P)
P/A
Pey/ I My/ I Resultant stress (stress (stress profile due to P.e) due to M)
The resultant stress at a distance y from the CGC is given by the principle of superposition as follows.
f =-
P Pey My A I I
(3-2.1)
For a curved tendon, P can be substituted by its horizontal component. But the effect of the refinement is negligible.
Based on Force Concept The approach based on force concept is analogous to the study of reinforced concrete. The tension in prestressing steel (T) and the resultant compression in concrete (C) are considered to balance the external loads. This approach is used to determine the
dimensions of a section and to check the service load capacity. Of course, the stresses in concrete calculated by this approach are same as those calculated based on stress concept. The stresses at the extreme edges are compared with the allowable stresses.
C =T
Indian Institute of Technology Madras
(3-2.2)
M = C.z M = C(ec + e)
The resultant stress in concrete at distance y from the CGC is given as follows.
f =C Cec y A I
(3-2.3)
(3-2.4)
Substituting C = P and Cec = M Pe, the expression of stress becomes same as that given by the stress concept. (3-2.5)
f =-
P Pey My A I I
Based on Load Balancing Concept The approach based on load balancing concept is used for a member with curved or harped tendons and in the analysis of indeterminate continuous beams. The moment, upward thrust and upward deflection (camber) due to the prestress in the tendons are calculated. The upward thrust balances part of the superimposed load.
The expressions for three profiles of tendons in simply supported beams are given.
e wup L
M
Bending moment diagram
Figure 3-2.5 Simply supported beam with parabolic tendon
The moment at the centre due to the uniform upward thrust (wup) is given by the following equation.
M=
Indian Institute of Technology Madras
wup L2 8
(3-2.6) The moment at the centre from the prestressing force is given as M = Pe. The expression of wup is calculated by equating the two expressions of M. The upward deflection () can be calculated from wup based on elastic analysis.
w up = = 8Pe L2 5wup L4 384EI
(3-2.7)
P Wup
M
Bending moment diagram
Figure 3-2.6 Simply supported beam with singly harped tendon
The moment at the centre due to the upward thrust (Wup) is given by the following equation. It is equated to the moment due to the eccentricity of the tendon. As before, the upward thrust and the deflection can be calculated.
M=
Wup L
(3-2.8)
P aL W
up
P Wup
M
Bending moment diagram
Figure 3-2.7
The moment at the centre due to the upward thrusts (Wup) is given by the following equation. It is equated to the moment due to the eccentricity of the tendon. As before, the upward thrust and the deflection can be calculated.
M = Wup aL = Pe Wup = = Pe aL a ( 3 - 4a 2 ) Wup L3 24EI
(3-2.9)
Example 3-2.1
A concrete beam prestressed with a parabolic tendon is shown in the figure. The prestressing force applied is 1620 kN. The uniformly distributed load includes the self weight. Compute the extreme fibre stress at the mid-span by applying the three concepts. Draw the stress distribution across the section at mid-span.
45 kN/m CGC 145 At end 7.3m 500 750 At mid-span
Solution
a) Stress concept A = 500 750 = 375,000 mm2 I = (500 7503) / 12 = 1.758 1010 mm4 Bending moment at mid-span, M = (45 7.32) / 8 = 299.7 kNm
Area of concrete,
Moment of inertia,
Bottom fibre stress P Pe M ( fc )b = - - y bot + y bot A I I 3 162010 1620 103 145 299.7106 =375 + 375 375103 1.7581010 1.7581010 = -4.32 - 5.01+ 6.39 = -2.9 N/mm2 b) Force concept
ec e P
Applied moment M
C P z
C
= 299.7 kN-m
Lever arm
Eccentricity of C
ec
c) Load balancing method wup = 8Pe / L2 = 8 1620 103 145 / 73002 = 35.3 kN/m wres
Residual load
(fc)res (fc)t
= 64.6 106 375 / 1.7581010 = 1.38 N/mm2 = P/A (fc)res = 4.32 1.38 = 5.7 N/mm2
(fc)b
2.9 N/mm2
Introduction The analysis of flexural members under service loads involves the calculation of the following quantities. a) Cracking moment. b) Location of kern points. c) Location of pressure line.
Based on the allowable tensile stress the prestress members are classified into three types as per IS:1343 - 1980. The types are explained in Section 1.2, Advantages and Types of Prestressing. For Type 1 (full prestressing) and Type 2 (limited prestressing) members, cracking is not allowed under service loads. Hence, it is imperative to check that the cracking moment is greater than the moment due to service loads. This is satisfied when the stress at the edge due to service loads is less than the modulus of rupture.
The modulus of rupture is the stress at the bottom edge of a simply supported beam corresponding to the cracking moment (Mcr). The modulus of rupture is a measure of the flexural tensile strength of concrete. It is measured by testing beams under 2 point
loading (also called 4 point loading including the reactions or middle third loading). The modulus of rupture (fcr) is expressed in terms of the characteristic compressive strength (fck) of concrete by the following equation (IS:456 - 2000). Here, fcr and fck are in N/mm2.
(3-3.1)
The following sketch shows the internal forces and the resultant stress profile at the instant of cracking.
CGC
Mcr Pe
Resultant stress profile
yb
fcr
The stress at the edge can be calculated based on the stress concept as follows. The cracking moment (Mcr) can be evaluated by transposing the terms.
(3-3.2)
The above equation expresses Mcr in terms of the section and material properties and prestressing variables.
edge is zero. The levels of the upper and lower kern points from CGC are denoted as kt and kb, respectively. Based on the stress concept, the stress at the bottom edge corresponding to C at the upper kern point, is equated to zero. The following sketch shows the location of C and the resultant stress profile.
CGC
yb
Location of resultant compression
Figure 3-3.2
kt
Resultant stress profile
The value of kt can be calculated by equating the stress at the bottom to zero as follows. C Ckt y b + =0 A I C Ck y or, - + t 2 b = 0 A Ar r2 or, kt = yb -
(3-3.3)
The above equation expresses the location of upper kern point in terms of the section properties. Here, r is the radius of gyration and yb is the distance of the bottom edge from CGC. Similar to the calculation of kt, the location of the bottom kern point can be calculated by equating the stress at the top edge to zero. The following sketch shows the location of C and the resultant stress profile.
yt
CGC
kb
Resultant stress profile
Figure 3-3.3
C Ck b y t + =0 A I C Ck y or, - + b 2 t = 0 A Ar r2 or, k b = yt -
(3-3.4)
The kern points can be used to determine the cracking moment (Mcr). The cracking moment is slightly greater than the moment causing zero stress at the bottom. C is located above kt to cause a tensile stress fcr at the bottom. The incremental moment is fcr I/yb. The following sketch shows the shift in C outside the kern to cause cracking and the corresponding stress profiles.
z
CGC
ec e
C
CGS
kt
= fcr fcr
Resultant stress profile
Figure 3-3.4
The cracking moment can be expressed as the product of the compression and the lever arm. The lever arm is the sum of the eccentricity of the CGS (e) and the eccentricity of the compression (ec). The later is the sum of kt and z, the shift of C outside the kern.
Mcr = C ( e + ec ) = C ( e + k t + z ) or, Mcr = C ( e + kt ) + fcr I yb
(3-3.5)
Substituting C = Pe, kt = r2/yb and r2 = I/A, the above equation becomes same as the previous expression of Mcr.
(3-3.6)
(3-3.7)
A positive value of ec implies that C acts above the CGC and vice-versa. If ec is negative and the numerical value is greater than kb (that is |ec| > kb), C lies below the lower kern point and tension is generated at the top of the member. If ec> kt, then C lies above the upper kern point and tension is generated at the bottom of the member.
The pressure line is calculated from the moment due to the self weight. The following sketch shows that the pressure line for a simply supported beam gets shifted from the CGS with increasing moment towards the centre of the span.
CL
kt kb
Pressure line at transfer
The pressure line is calculated from the moment due to the service loads. The following sketch shows that the pressure line for a simply supported beam gets further shifted from the CGS at the centre of the span with increased moment under service condition.
CL
kt kb
Pressure line under service loads
Limiting Zone
For fully prestressed members (Type 1), tension is not allowed under service conditions. If tension is also not allowed at transfer, C always lies within the kern zone. The limiting zone is defined as the zone for placing the CGS of the tendons such that C always lies within the kern zone.
For limited prestressed members (Type 2 and Type 3), tension is allowed at transfer and under service conditions. The limiting zone is defined as the zone for placing the CGS such that the tensile stresses in the extreme edges are within the allowable values.
The following figure shows the limiting zone (as the shaded region) for a simply supported beam subjected to uniformly distributed load.
C L CGC
Locus of emin
Locus of emax
Figure 3-3.4
The determination of limiting zone is given in Section 4.4, Design of Sections for Flexure (Part III).
Example 3-3.1
For the post-tensioned beam with a flanged section as shown, the profile of the CGS is parabolic, with no eccentricity at the ends. The live load moment due to service loads at mid-span (MLL) is 648 kNm. The prestress after transfer (P0) is 1600 kN. Assume 15% loss at service. Grade of concrete is M30.
CGC
18.0m
500 200 150
Values in mm.
200 250
Cross-section at mid-span
Evaluate the following quantities. a) Kern levels b) Cracking moment c) Location of pressure line at mid-span at transfer and at service. d) The stresses at the top and bottom fibres at transfer and at service.
Compare the stresses with the following allowable stresses at transfer and at service. For compression, fcc,all = 18.0 N/mm2 For tension, fct,all = 1.5 N/mm2.
Solution
Calculation of geometric properties
The section is divided into three rectangles for the computation of the geometric properties. The centroid of each rectangle is located from the soffit.
1 + 2 500 100 3
y
yt
CGC
900
yb
Values in mm.
Area of the section = A1 = 500 200 = 100,000 mm2 = A2 = 600 150 = 90,000 mm2 = A3 = 250 200 = 50,000 mm2
Area of 1
Area of 2
Area of 3
= A1 + A2 + A3 = 240,000 mm2
Therefore,
Moment of inertia of
I1 =
Moment of inertia of
I2 =
Moment of inertia of
I3 =
r2 =
kb = =
r2 yt
CGC
182.2 255.1
+ Kern zone
Modulus of rupture
Mcr =
3.83 2.5521010 0.851600103 2.5521010 = + 583.3 240103 583.3 + 0.81600 103 433.3 Nmm = 167.6 + 247.9 + 554.6 = 970.1 kNm
At transfer
z=
Since ec is negative, the pressure line is below CGC. Since the magnitude of ec is greater than kb, there is tension at the top.
+ Kern zone
x
Value in mm.
At service
z= MDL+LL C (233.3 + 648.0)103 = 0.851600 = 648.0 mm
Since ec is positive, the pressure line is above CGC. Since the magnitude of ec is greater than kt, there is tension at the bottom.
x
CGC
214.7 +
CGS
Value in mm.
f =M
P Pey My A I I
+ + =
CGC
P
P/A
Pey/ I
My/ I
= 0.84 N/mm2
Stress at the bottom fibre
P0ey b 1600103 433.3 583.3 =2.5521010 I = -15.85 N/mm2
MDL y b 233.3106 583.3 = I 2.5521010 = 5.33 N/mm2
= - 17.19 N/mm2
Calculation of stresses at service (P = Pe)
= -10.44 N/mm2
Stress at the bottom fibre
Pf ey b = - 0.85 15.85 I = - 13.47 N/mm2
MLL y b 648.0106 583.3 = 2.5521010 I = 14.81N/mm2
= 1.0 N/mm2
The allowable stresses are as follows. For compression, For tension, fc,comp = 18.0 N/mm2 fc,tens = 1.5 N/mm2.
Of course, superposition cannot be used to calculate the deflection under combined loads.
Variation of Stress in Prestressing Steel In non-prestressed reinforced concrete members, the tension and consequently the stress in steel increase almost proportionately with increasing moment till yielding. The lever arm between the resultant compression and tension remains almost constant. In prestressed concrete members, the tension and consequently the stress in prestressing steel increase slightly with increasing moment till cracking of concrete. The increase in moment changes the lever arm significantly. This is explained in Section 3.2, Analysis of Member under Flexure (Part I). increases rapidly with moment. The following sketch explains the variations of the stress in prestressing steel (fp) with increasing load. The variations are shown for bonded and unbonded tendons. After the prestress is transferred while the member is supported at the ends, the stress will tend to increase from the value after losses (fp0) due to the moment under self weight. Subsequently the stress will tend to drop due to the time dependent losses such as from creep, shrinkage and relaxation. The losses of prestress are covered in Section 2.3, Losses in Prestress (Part III). The effective prestress after time dependent losses is denoted as fpe. Due to the moment under service loads, the stress in the prestressing steel will slightly increase from fpe. The increase is more at the section of maximum moment in a bonded tendon as compared to the increase in average stress for an unbonded tendon. The stress in a bonded tendon is not uniform along the length. Usually the increase in stress is neglected in the calculations under service loads. If the loads are further increased, the stress increases slightly till cracking. After cracking, there is a jump of the stress in the prestressing steel. Beyond that, the stress increases rapidly with moment till the ultimate load. At ultimate, the stress is represented as fpu. Similar to the observation for pre-cracking, the average stress in an unbonded tendon is less than the stress at the section of maximum moment for a bonded tendon. After cracking, the stress in prestressing steel
fp fpu
Bonded
fp0 fpe
Losses
Unbonded
Load
Figure 3-4.1
The above sketch assumes that the section is failing in flexure. Other types of failure are not considered.
Conditions at Ultimate Limit State In the limit states method of analysis, the limit state of collapse (ultimate state) of a member under flexure is defined as the state when the extreme concrete compressive strain reaches a value of 0.0035. At ultimate, let the extreme concrete compressive strain be denoted as cu. Thus, cu = 0.0035. Depending on the amount of prestressing steel, a section can be under-reinforced or over-reinforced. For an under- reinforced section, the amount of prestressing steel is less and the steel yields before the extreme concrete strain reaches 0.0035. For an over-reinforced section, the amount of steel is high and the steel does not yield at ultimate. The transition situation is called a balanced condition. three situations are shown below. The strain profiles across the depths of prestressed flexural members (up to the depth of CGS) for the
cu
cu
cu
p
pu > pu,bal Under-reinforced
p
pu = pu,bal Balanced
p
pu < pu,bal Over-reinforced
pu,bal = strain in the steel for a balanced section. The strain difference (p) is the strain in the prestressed tendons when the adjacent concrete has zero strain (c = 0). The strain difference gets locked during the transfer of prestress. The value can be determined as follows. For pre-tensioned members, the strain difference gets locked when the tendons are cut. The strain difference at that instant is given as follows.
p = pi 0
Here, pi = strain in tendons just before transfer c = strain in concrete is zero.
(3-4.1a)
For post-tensioned members, the strain difference gets locked when the tendons are anchored. The strain difference at that instant is given as follows.
p = p0 c0
Here, p0 = strain in tendons due to P0, the prestress after transfer c0 = strain in concrete due to P0. In general at any load stage,
(3-4.1b)
p = pe ce
(3-4.1c)
Here, pe = strain in tendons due to Pe, the prestress at service ce = strain in concrete due to Pe. As mentioned under material properties, the prestressing steel does not have a definite yield point. The 0.2% proof stress is defined when the steel reaches an inelastic strain of 0.2%. Hence, unlike reinforced concrete, the transition from under-reinforced to overreinforced section is gradual and there is no definite balanced condition. IS:1343 - 1980 does not explicitly enforce an under-reinforced section. But the IRS Concrete Bridge Code requires that the strain in the outermost tendon should not be less than the following.
0.87fpk + 0.005 Ep
The above value can be considered to be the strain in the steel at balanced condition.
Assumptions for Analysis The analysis of members under flexure for ultimate strength considers the following. 1) Plane sections perpendicular to the axis of the member remain plane till the ultimate state. 2) Perfect bond is retained between concrete and prestressing steel for bonded tendons. 3) Tension in concrete is neglected. 4) The design stress versus strain curves of concrete and steel are considered. The methods of analysis will be presented for three types of sections. 1) Rectangular section: A rectangular section is easy to cast, but it is not an efficient section. 2) Flanged section: A precast flanged section, with flanges either at top or bottom needs costlier formwork. But the section is efficient in flexure. A flanged section can also be made of precast web and cast-in-place slab. 3) Partially prestressed section: A section in a member containing both prestressed and non-prestressed reinforcement.
cu=0.0035
0.447fck
d Ap
pu Strain
xu
0.42xu
Cu Tu
Force
fpu
Stress
Cross-section
Figure 3-4.3
The variables in the above figure are explained. b d Ap = breadth of the section = depth of the centroid of prestressing steel (CGS) = area of the prestressing steel = strain difference = depth of the neutral axis at ultimate = strain in prestressing steel at the level of CGS at ultimate = stress in prestressing steel at ultimate
p
xu pu fpu
The stress block in concrete is derived from the constitutive relationship for concrete. The relationship is explained in Section 1.6, Concrete (Part II). The compressive force in concrete can be calculated by integrating the stress block along the depth. The stress in the tendon is calculated from the constitutive relationship for prestressing steel. The relationship is explained in Section 1.7, Prestressing Steel. In the force diagram,
Cu = 0.36fck xu b Tu = Ap fpu
(3-4.2) (3-4.3)
The strengths of the materials are denoted by the following symbols. fck = characteristic compressive strength of concrete fpk = characteristic tensile strength of prestressing steel
Indian Institute of Technology Madras
For analysis of a prestressed section, three principles of mechanics are used. First, the equilibrium relates the external applied forces with the internal forces. Second, the compatibility condition relates the strain in the prestressing steel with the strain in concrete at the level of CGS. This also considers the first two assumptions given in the previous section. materials. Based on the above principles of mechanics, the following equations are derived. 1) Equations of equilibrium The first equation states that the resultant axial force is zero. This means that the compression and the tension in the force couple balance each other.
F = 0 Tu = Cu
Ap fpu = 0.36fck xu b
(3-4.4)
The second equation relates the ultimate moment capacity (MuR) with the internal couple in the force diagram.
MuR = Tu (d - 0.42 xu )
= Ap fpu ( d - 0.42 xu )
(3-4.5)
2) Equation of compatibility The depth of the neutral axis is related to the depth of CGS by the similarity of the triangles in the strain diagram.
xu 0.0035 = d 0.0035 + pu - p
3) Constitutive relationships a) Concrete The constitutive relationship for concrete is considered in the expression under compression.
(3-4.6)
Cu =
0.36fckxub. This is based on the area under the design stress-strain curve for concrete
b) Prestressing steel
fpu = F ( pu )
(3-4.7)
The function F(pu) represents the design stress-strain curve for prestressing steel under tension. The known variables in an analysis are: b, d, Ap, p, fck, fpk. The unknown quantities are: xu, MuR, pu, fpu. The objective of the analysis is to find out MuR, the ultimate moment capacity. The simultaneous equations 3-4.1 to 3-4.7 can be solved iteratively. This procedure of analysis is called the strain compatibility method. The steps are as follows. 1) Assume xu . 2) Calculate pu by rearranging the terms of Eqn. 3-4.6. 3) Calculate fpu from Eqn. 3-4.7. 4) Calculate Tu from Eqn. 3-4.3. 5) Calculate Cu from Eqn. 3-4.2. If Eqn. 3-4.4 (Tu = Cu) is not satisfied, change xu. If Tu < Cu decrease xu. If Tu > Cu increase xu. 6) Calculate MuR from Eqn. 3-4.5. The capacity MuR can be compared with the demand under ultimate loads. In the strain compatibility method, the difficult step is to calculate xu and fpu. IS:1343 1980 allows to calculate these variables approximately from Table 11, Appendix B, based on the amount of prestressing steel. The later is expressed as a prestressed reinforcement index p.
p =
Ap fpk bdfck
(3-4.8)
Table 11 is reproduced as Table 3-4.1 which is applicable for pre-tensioned and bonded post-tensioned beams. The values of fpu and xu are given as fpu/(0.87fpk) and xu/d, respectively.
Table 3-4.1
Values of xu and fpu for pre-tensioned and bonded post-tensioned rectangular beams (Table 11, IS:1343 - 1980)
fpu/(0.87fpk) p
Pre-tensioned 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.9 tensioned 0.025 0.05 0.10 0.15 0.20 0.25 0.30 0.40 1.0 1.0 1.0 1.0 0.95 0.90 0.85 0.75 0.054 0.109 0.217 0.326 0.435 0.542 0.655 0.783
xu/d
Bonded posttensioned 0.054 0.109 0.217 0.316 0.414 0.488 0.558 0.653
The values of fpu/(0.87fpk) and xu/d from Table 3-4.1 are plotted in Figures 3-4.4 and 34.5, respectively. It is observed that with increase in p, fpu reduces (beyond certain values of p) and xu increases. This is expected because with increase in the amount and strength of the steel, the stress in steel drops and the depth of the neutral axis increases to maintain equilibrium.
1.2 1
0.2
0.3
0.4
Pre-tensioned
Post-tensioned (bonded)
Figure 3-4.4
0.9
0.6
xu / d
0.3
p
Pre-tensioned
Post-tensioned (bonded)
Figure 3-4.5
Thus given the value of p for a section, the values of fpu and xu can be approximately calculated from the above tables.
Example 3-4.1
A prestressed concrete beam produced by pre-tensioning method has a rectangular cross-section of 100 mm 160 mm (b h). It is prestressed with 10 numbers of straight 2.5 mm diameter wires. Each wire is stressed up to a load of 6.8 kN. The design load versus strain curve for each wire is given in a tabular form. The grade of concrete is M 40. The value of p is 0.0073.
Estimate the ultimate flexural strength of the member by the strain compatibility method.
100
CGC 40
160
p
0.006 0.008 0.010 0.012 0.014
P (kN)
5.4 7.6 9.0 10.0 10.7
Solution
Strain difference
= 0.0073
The effective depth of the CGS (d ) is 120 mm. The strain compatibility method is shown in a tabular form. Here,
Pu = load in a single wire obtained from the table Tu = 10 Pu , for the ten wires.
xu
(mm)
xu/d
pup
(3-4.6)
pu
Pu
(kN) (Table)
Tu
(kN)
Cu
(kN) (3-4.2)
Checking (3-4.4)
60 65 63.5
Tu > Cu Tu < Cu Tu Cu
T-section
Double T-section
L-section
Inverted T-section
I-girder
Figure 3-5.1
T T
T T
T T
Box section
Figure 3-5.2
Composite beam-slab
T-section
The analysis of a flanged section for ultimate strength is different from a rectangular section when the flange is in compression. If the depth of the neutral axis from the edge under compression is greater than the depth of the flange, then the section is treated as a flanged section. In the following figure, the first strain profile shows that the depth of the neutral axis (xu) is greater than the depth of the flange (Df). The section is treated as a flanged section. The second strain profile shows that xu is less than Df. In this situation, the section can be treated as a rectangular section.
bf Df d Ap xu xu
Cross-section
bw
Figure 3-5.3
The effective width or breadth of the flange (bf) is determined from the span of the beam, breadth of the web (bw) and depth of the flange (Df) as per Clause 23.1.2, IS:456 - 2000. Analysis of a Flanged Section The following sketch shows the beam cross-section, strain profile, stress diagram and force couples at the ultimate state. The following conditions are considered. 1) xu > Df : This requires an analysis for a flanged section.
2) Df (3/7) xu: This ensures that the compressive stress is constant at 0.447fck along the depth of the flange.
Df
bf d Ap bw
0.0035
0.447fck
0.5Df 0.42xu
xu fpu
Cuw + Tuw
Force
Cuf Tuf
Cross-section
Figure 3-5.4
p pu Strain Stress
The variables in the above figure are explained. bf bw Df d Ap = breadth of the flange = breadth of the web = depth of the flange = depth of the centroid of prestressing steel (CGS) = area of the prestressing steel = strain difference in the prestressing steel when strain in concrete is zero = depth of the neutral axis at ultimate = strain in prestressing steel at the level of CGS at ultimate = stress in prestressing steel at ultimate
p
xu pu fpu
The strain difference (p) is further explained in Section 3.4, Analysis of Member under Flexure (Part III).
The first
component (Tuw) balances the compressive force carried by the web, including the portion of the flange above web (Cuw). Thus Tuw= Cuw. The second component (Tuf) balances the compressive force carried by the outstanding portion of the flange (Cuf). Thus Tuf = Cuf. The stress block in concrete is derived from the constitutive relationship for concrete. The relationship is explained in Section 1.6, Concrete (Part II). The compressive force in concrete can be calculated by integrating the stress block along the depth. The
stress in the tendon is calculated from the constitutive relationship for prestressing steel. The relationship is explained in Section 1.7, Prestressing Steel.
Cuw = 0.36fck xu bw Cuf = 0.447fck (bf - bw )Df Tuw = Apw fpu Tuf = Apf fpu
The strengths of the materials are denoted by the following symbols. Apf Apw fck fpk = part of Ap that balances compression in the outstanding flanges = part of Ap that balances compression in the web = characteristic compressive strength of concrete = characteristic tensile strength of prestressing steel
Based on the principles of mechanics (as explained under the Analysis of a Rectangular Section in Section 3.4, Analysis of Member Under Flexure (Part III)), the following equations are derived.
1) Equations of equilibrium The first equation states that the resultant axial force is zero. This means that the compression and the tension in the force couple balance each other.
F = 0 Tu = Cu Tuw +Tuf = Cuw + Cuf
(3-5.5)
The second equation relates the ultimate moment capacity (MuR) with the internal couple in the force diagram. MuR = Tuw ( d - 0.42 xu ) +Tuf ( d - 0.5Df ) = Apw fpu ( d - 0.42 xu ) + Apf fpu ( d - 0.5Df )
(3-5.6)
From Tuf = Cuf and Eqns. (3-5.2) and (3-5.4), Apf is given as follows. The calculation of Apw from Ap and Apf is also shown.
Apf =
0.447fck ( bf - bw ) Df fpu
(3-5.7)
Apw = Ap - Apf
(3-5.8)
2) Equation of compatibility The depth of the neutral axis is related to the depth of CGS by the similarity of the triangles in the strain diagram.
xu 0.0035 = d 0.0035 + pu - p
3) Constitutive relationships a) Concrete
(3-5.9)
The constitutive relationship for concrete is considered in the expressions of Cuw and Cuf. This is based on the area under the design stress-strain curve for concrete under compression.
b) Prestressing steel
(3-5.10) fpu = F ( pu ) The function F(pu) represents the design stress-strain curve for the type of prestressing
steel used. The known variables in an analysis are: bf, bw, Df, d, Ap, p, fck and fpk. The unknown quantities are: Apf, Apw, MuR, xu, pu and fpu. The objective of the analysis is to find out MuR , the ultimate moment capacity. The simultaneous equations 3-5.1 to 3-5.10 can be solved iteratively.
The steps of the strain compatibility method are as follows. 1) Assume xu = Df. 2) The calculations are similar to a rectangular section, with b = bf. 3) If Tu > Cu, increase xu. Treat the section as a flanged section. 4) Calculate pu from Eqn. (3-5.9). 5) Calculate fpu from Eqn. (3-5.10). 6) Calculate Apf and Apw from Eqn. (3-5.7) and Eqn. (3-5.8), respectively.
Indian Institute of Technology Madras
7) Calculate Cuw, Cuf, Tuw and Tuf from Eqns. (3-5.1) to (3-5.4). If Eqn. (3-5.5) (Tu = Cu) is not satisfied, iterate with a new value of xu, till convergence. 8) Calculate MuR from Eqn. (3-5.6). The capacity MuR can be compared with the demand under ultimate loads. In the strain compatibility method, the difficult step is to calculate xu and fpu. Similar to the rectangular section, an approximate analysis can be done based on Table 11 and
Table 12, Appendix B, IS:1343-1980. The tables are reproduced in Table 3-4.1 and
Table 3-4.2, respectively, in Section 3.4, Analysis of Member under Flexure (Part III). The values of xu and fpu are available in terms of a reinforcement index pw. A f pw = pw pk bw dfck
(3-5.11)
Note that the index is calculated based on Apw instead of Ap. The calculation of Apw is from Eqn. (3-5.8). But Apf depends on fpu, which is unknown. Hence, an iterative procedure is required.
The steps are as follows. 1) Assume fpu = 0.87fpk. 2) Calculate Apf and Apw from Eqn. (3-5.7) and Eqn. (3-5.8), respectively. 3) Calculate pw. 4) Calculate fpu from Table 11 or Table 12. Compare the calculated value of fpu with the assumed value. Repeat steps 1 to 4 till convergence. 5) Calculate MuR. If Df > (3/7) xu, the flange depth is larger than the depth of constant compressive stress. An equivalent depth of the flange is defined as follows.
yf = 0.15xu + 0.65Df
The equivalent depth yf is substituted for Df in the expression of MuR.
(3-5.12)
Example 3-5.1
A bonded post-tensioned concrete beam has a flanged cross-section as shown. It is prestressed with tendons of area 1750 mm2 and effective prestress of 1100 N/mm2. The tensile strength of the tendon is 1860 N/mm2. The grade of concrete is M60.
Estimate the ultimate flexural strength of the member by the approximate method of IS:1343 - 1980.
460 175 550 175 460 Values are in mm. Cross-section at mid-span 140
900
115
Solution
Effective depth
Assume xu = Df = 175 mm. Treat as a rectangular section, with b = bf = 460 mm. Reinforcement index
P = = AP fPk bdfck
Tu
Cu
Assume
fpu = 0.87fpk
= 1618 N/mm2
Apf = =
Reinforcement index
pw = =
From Table 11,
= 1489 N/mm2
= 1521N/mm2
Reinforcement index
pw = 756 1860 140 785 60 = 0.21
The value of wpw is same as after 2nd iteration. Hence, the values of fpu, Apf and Apw have converged.
Indian Institute of Technology Madras
Tuw (d - 0.42 xu ) = Apw fpu (d - 0.42 xu ) = 7561521(785 - 0.42337) = 739.9 kNm Tuf (d - 0.5Df ) = Apf fpu (d - 0.5Df ) = 9941521(785 - 0.5175) = 1054.5 kNm
The ultimate flexural strength is given as follows. MuR = 1054.5 + 739.9 = 1794.4 kNm
Analysis of Partially Prestressed Section Analysis of Unbonded Post-tensioned Beam Analysis of Behaviour
reinforcement is provided in addition to the prestressing steel. When this reinforcement is considered in the flexural capacity, the section is termed as a partially prestressed section.
The reasons for using a partially prestressed section are as follows. 1) The section is economical. 2) The cambering is less compared to an equivalent section without conventional reinforcement. 3) The ductility is more in a partially prestressed section. 4) Any reversal of moments (for example, due to earthquake) is not detrimental as compared to an equivalent section without conventional reinforcement.
Analysis A partially prestressed section can be either rectangular or flanged. A section can be doubly reinforced with reinforcement near the compression face.
Here, the equations for a doubly reinforced rectangular section are given. The following sketch shows the beam cross section, strain profile, stress diagram and force couples at the ultimate state.
b dp As Ap As d d
0.0035 s
0.447fck
xu fs
fs fpu
0.42xu
Cs Cc Tp Ts
Force
p
pu
Cross-section
Figure 3-6.1
Strain
Stress
The variables in the above figure are explained. b d d dp As As Ap = breadth of the section = depth of the centroid of the reinforcing steel (tension side) = depth of the centroid of the reinforcing steel (compression side) = depth of the centroid of prestressing steel (CGS) = area of the reinforcing steel (tension side) = area of the reinforcing steel (compression side) = area of the prestressing steel = strain difference in the prestressing steel when strain in concrete is zero = depth of the neutral axis at ultimate = strain in reinforcing steel (tension side) at ultimate = strain in reinforcing steel (compression side) at ultimate = strain in prestressing steel at the level of CGS at ultimate = stress in reinforcing steel (tension side) at ultimate = stress in reinforcing steel (compression side) at ultimate = stress in prestressing steel at ultimate.
p
xu s s pu fs fs fpu
The strain difference (p) is further explained in Section 3.4, Analysis of Member under Flexure (Part III).
The stress block in concrete is derived from the constitutive relationship for concrete. The relationship is explained in Section 1.6, Concrete (Part II). The compressive force in concrete can be calculated by integrating the stress block along the depth. The stress in the tendon is calculated from the constitutive relationship for prestressing steel. The relationship is explained in Section 1.7, Prestressing Steel.
Indian Institute of Technology Madras
The expressions of the forces are as follows. Cs = Asfs Cc = 0.36fckxub Tp = Apfpu Ts = Asfs The strengths of the materials are denoted by the following symbols. fck fpk fy = characteristic compressive strength of concrete = characteristic tensile strength of prestressing steel = characteristic yield stress of reinforcing steel
(3-6.1) (3-6.2) (3-6.3) (3-6.4)
Based on the principles of mechanics (as explained under the Analysis of a Rectangular Section in Section 3.4, Analysis of Member Under Flexure (Part III)), the following equations are derived.
1) Equations of equilibrium The first equation states that the resultant axial force is zero. This means that the compression and the tension in the force couple balance each other.
F = 0 Tu = Cu Tp +Ts = Cc + C' s
' ' Ap fpu + As fs = 0.36fck xu b + As fs
(3-6.5)
The second equation relates the ultimate moment capacity (MuR) with the internal couple in the force diagram. MuR = Ts ( d - d p ) + Cc ( d p - 0.42 xu ) + C' s ( d p - d' ) = As fs ( d - d p ) + 0.36fck xu b ( d p - 0.42 xu ) + A' s fs' ( d p - d' ) 2) Equations of compatibility For each layer of steel there is a compatibility equation. If there are distributed
(3-6.6)
Ap
reinforcing bars in several layers and the spacing between the layers is large, then the use of compatibility equation for each layer is more accurate than the use of one compatibility equation for the centroid of the layers. The following equations are
(3-6.7)
(3-6.8)
(3-6.9)
3) Constitutive relationships a) Concrete The constitutive relationship for concrete is considered in the expressions of Cc. This is based on the area under the design stress-strain curve for concrete under compression.
b) Prestressing steel
fpu = F1 ( pu )
c) Reinforcing steel fs = F2 ( s ) f = F3 (
' s ' s
(3-6.10)
(3-6.11) (3-6.12)
(3-6.13) (3-6.14)
The known variables in an analysis are: b, d, d, dp, As, As, Ap, p, fck, fy and fpk. The unknown quantities are: MuR, xu, s, s, pu, fs, fs and fpu. The objective of the analysis is to find out MuR, the ultimate moment capacity. The previous equations can be solved by the strain compatibility method as discussed for the fully prestressed rectangular section. 1) Assume xu. 2) Calculate pu from Eqn. 3-6.7. 3) Calculate fpu from Eqn. 3-6.10. 4) Calculate Tp from Eqn. 3-6.3. 5) Calculate s from Eqn. 3-6.8. 6) Calculate fs from Eqn. 3-6.11. 7) Calculate Ts from Eqn. 3-6.4. 8) Calculate s from Eqn. 3-6.9.
Indian Institute of Technology Madras
9) Calculate fs from Eqn. 3-6.12. 10) Calculate Cs from Eqn. 3-6.1. 11) Calculate Cc from Eqn. 3-6.2. If Eqn. 3-6.5 (Tu = Cu) is not satisfied, change xu. If Tu < Cu decrease xu. If Tu > Cu increase xu. 12) Calculate MuR from Eqn. 3-6.6. The capacity MuR can be compared with the demand under ultimate loads.
A sectional analysis is not possible. The analysis involves integrating the strain in concrete to calculate the deformation p = cp Here, p = deformation of the tendon cp = deformation of the concrete at the level of prestressing steel (CGS). The change in stress in steel (fp) at ultimate is determined from p. The stress in steel at ultimate is given by the sum of the effective prestress (fpe) and fp. fpu = fpe + fp
(3-6.16)
over
the
length
of
the
member.
The value of fpu is less than that for a bonded tendon. The ultimate moment is given by the following equation.
MuR = Ap fpu ( d - 0.42 xu )
(3-6.17)
based on the amount of prestressing steel. The later is expressed as the reinforcement index p = Apfpk / bdfck. Table 12 is reproduced as Table 3-6.1 which is applicable for
unbonded post-tensioned beams. The values of fpu and xu are given as fpu/fpe and Xu/d, respectively. The effective prestress (after the losses) in a tendon is represented as fpe. Values of xu and fpu for unbonded post-tensioned rectangular beams (Table 12, IS:1343 - 1980) fpu/fpe p 30 0.025 0.05 0.10 0.15 0.20 1.23 1.21 1.18 1.14 1.11 For values of L/d 20 1.34 1.32 1.26 1.20 1.16 10 1.45 1.45 1.45 1.36 1.27 30 0.10 0.16 0.30 0.44 0.56 xu/d For values of L/d 20 0.10 0.16 0.32 0.46 0.58 10 0.10 0.18 0.36 0.52 0.64
Table 3-6.1
The values of fpu/fpe and xu/d from Table 3-6.1 are plotted in Figures 3-6.2 and
3-6.3,
respectively. It is observed that with increase in p, fpu reduces and xu increases. This is expected because with increase in the amount and strength in the steel, the stress in steel drops and the depth of the neutral axis increases to maintain equilibrium.
1.5
l /d
fpu/fpe
30 1.25 20 10
p
Figure 3-6.2
0.6
l /d
0.4
30
20
10
xu / d
0.2 0
0.05
0.1
0.15
0.2
p
Figure 3-6.3
Thus given the value of p for a section, the values of fpu and xu can be approximately calculated from the above tables.
=
Here,
c + cp d
(3-6.18)
c = extreme concrete compressive strain cp = strain in concrete at the level of prestressing steel (CGS) d = depth of the CGS. The following sketch shows the curvature () in the strain profile.
d Ap
cp Cross-section
Figure 3-6.4
Strain profile
Definition of curvature
The analysis of behaviour involves the following three principles of mechanics. 1) Equilibrium of internal forces with the external loads at any point of the behaviour. There are two equilibrium equations. a) Force equilibrium equation b) Moment equilibrium equation. The internal forces in concrete and steel are evaluated based on the respective strains, cross-sectional areas and the constitutive relationships.
2) Compatibility of the strains in concrete and in steel for bonded tendons. This assumes a perfect bond between the two materials. For unbonded tendons, the compatibility is in terms of deformation.
3) Constitutive relationships relating the stresses and the strains in the materials. The relationships are developed based on the material properties.
The equilibrium and compatibility equations and the constitutive relationships can be solved to develop the moment versus curvature curve for a section.
The following plot shows the curves for a prestressed section and a non-prestressed section. The two sections are equivalent in their ultimate flexural strengths.
Figure 3-6.5
From the previous plot, the following can be inferred. 1) Prestressing increases the cracking load. This leads to the following benefits.
For the prestressed member, there is a negative curvature causing camber in absence of external moment.
4) For a given reverse moment, the curvature of the prestressed member is larger.
6) For a partially prestressed section with the same ultimate strength, the moment versus curvature curve will lie in between the curves for prestressed and non-prestressed sections.
Ductility
The ductility is a measure of energy absorption. For beams, the curvature ductility () is defined as
=
Here, u = curvature at ultimate y = curvature at yield.
u y
(3-6.19)
For prestressed beams, y can be defined corresponding to a plastic strain of 0.002 in the prestressing tendons. It has been observed that the ductility of prestressed beams is less than that in reinforced concrete beams.
In design of members for seismic forces, ductility is an important requirement. In addition, seismic forces lead to reversal of moments near the supports of beams in a moment resisting frame. Hence, prestressing of beams in a moment resisting frame is not recommended in seismic areas.
Experimental Investigation
The behaviour of a beam and its ultimate strength can be determined by testing prototype specimens. The tests can be conducted under static or dynamic loads. Testing also helps to check the performance of the anchorage units.
The following photo shows the set-up for testing a prototype bridge girder.
Figure 3-6.6
Introduction The design of prestressed concrete members can be done by the limit states method as given in Section 4 of IS:1343 - 1980. First, the force demand in a member under the design loads is determined from a structural analysis. A preliminary size of the member is assumed for analysis. Next, the member is designed to meet the demand. If necessary, another cycle of analysis and design is performed. The following material explains the calculation of the demand in a member under the design loads.
Characteristics Loads For dead loads, a characteristic load is defined as the value which has a 95% probability of not being exceeded during the life of the structure. This concept assumes a normal distribution of the values of a particular dead load. In the following figure, the
shaded area above the characteristic value represents 5% probability of exceedance of the load in the design life of the structure.
Values of load
Figure 4-1.1
For live load, wind load and earthquake load, a characteristic load is defined based on an extreme value distribution. For example, the characteristic wind load is defined as the value which has a 98% probability of not being exceeded during a year.
Frequency 2% probability of exceedence Characteristic value Annual maximum mean wind speed
Figure 4-1.2
The characteristics loads can be obtained from IS:875 - 1987 (Code of Practice for Design Loads for Buildings and Structures) and IS:1893 - 2002 (Criteria for Earthquake Resistant Design of Structures) as follows.
Table 4-1.1 Type of load Dead load Live (imposed) load Wind load Snow Load Earthquake load
Codes covering information of loads Code (DL) (LL) (WL) (SL) (EL) IS:875 - 1987 Part 1 IS:875 - 1987 Part 2 IS:875 - 1987 Part 3 IS:875 - 1987 Part 4 IS:1893 - 2002 Part 1
For special loads, there are some guidelines in IS: 875 - 1987, Part 5. In addition, specialised literature may be referred to for these loads. The special loads are listed below. Temperature Hydrostatic Soil pressure Fatigue Accidental load Impact and collision Explosions Fire For special situations, the loads are determined from testing of prototype specimens. Dynamic load tests, wind tunnel tests, shake table tests are some types of tests to determine the loads on a structure. Finite element analysis is used to determine the stresses due to concentrated forces and dynamic loads.
Load Factors and Load Combinations The load factors and the combinations of the various types of loads are given in Table 5 of IS:1343 - 1980. The following are the combinations for the ultimate condition. 1.5 (DL + LL) 1.2 (DL + LL WL) 1.2 (DL + LL EL) 1.5 (DL EL)
1.5 (DL WL) 0.9 DL 1.5 EL The load combinations for service conditions are as follows. DL + LL DL + 0.8 (LL EL) DL EL DL WL
Analysis of Structures Regarding analysis of structures, IS:1343 - 1980 recommends the same procedure as stated in IS:456 - 2000. A structure can be analysed by the linear elastic theory to calculate the internal forces in a member subjected to a particular type of load.
Design of Members There can be more than one way to design a member. In design, the number of unknown quantities is larger than the number of available equations. Hence, some quantities need to be assumed at the beginning. These quantities are subsequently checked. The member can be designed either for the service loads or, for the ultimate loads. The procedure given here is one of the possible procedures. The design is based on satisfying the allowable stresses under service loads and at transfer. Initially, a lumpsum estimate of the losses is considered under service loads. After the first round of design, detailed computations are done to check the conditions of allowable stresses. Precise values of the losses are computed at this stage. The section is then analysed for the ultimate capacity. The capacity should be greater than the demand under ultimate loads to satisfy the limit state of collapse.
Design of Prestressing Force First, a preliminary dimension of the member is selected based on the architectural requirement. The prestressing force at transfer (P0) should be such that the compressive stress in concrete is limited to the allowable value. should be within the allowable value. At service, the designed prestressing force (Pe) should be such that the tensile stress in concrete The amount of prestressing steel (Ap) is determined from the designed prestressing force based on the allowable stress in steel. At transfer, in absence of non-prestressed reinforcement, the stress in concrete (fc) is given as follows.
fc = P0 Ac
(4-1.1)
Here, Ac = net area of concrete P0 = prestress at transfer after short-term losses. In presence of non-prestressed reinforcement, the stress in the concrete (fc) can be calculated as follows.
fc = P0 Ac + (Es /Ec )As
(4-1.2)
Here, As = area of non-prestressed reinforcement Es = modulus of elasticity of steel Ec = modulus of elasticity of concrete. At service, the stress in concrete (fc) can be calculated as follows.
fc = Pe P Ac At
(4-1.3)
Here, At = transformed area of section P = external axial force Pe = effective prestress. The external axial force is considered positive if it is tension and negative if it is compression. In the above expression, non-prestressed reinforcement is not considered. If there is non-prestressed reinforcement, Ac is to be substituted by (Ac + (Es/Ec) As) and At is to be calculated including As. Analysis of Ultimate Strength The ultimate tensile strength of a section (PuR) is calculated as per Clause 22.3, IS:1343 - 1980. The ultimate strength should be greater than the demand due to factored loads. In absence of non-prestressed reinforcement, the ultimate tensile strength of a section (PuR) is given as follows.
PuR = 0.87fpk Ap
(4-1.4)
(4-1.5)
In the previous equations, fy = characteristic yield stress for non-prestressed reinforcement with mild steel bars = characteristic 0.2% proof stress for non-prestressed reinforcement with high yield strength deformed bars. fpk = characteristic tensile strength of prestressing tendons. The following example shows the design of a post-tensioned hanger for tension.
Example 4-1.1
Design a post-tensioned hanger to carry an axial tension of PDL = 300 kN (dead load including self-weight) and PLL = 130 kN. The dimension of the hanger is 250 250 mm2.
Design the section without considering non-prestressed reinforcement. Tension is not allowed under service loads.
The grade of concrete is M 35. The age at transfer is 28 days. Assume 15% long term losses in the prestress.
The following properties of the prestressing strands are available from tests. Type of prestressing tendon : 7 wire strand Nominal diameter Nominal area Tensile strength fpk Modulus of elasticity = 12.8 mm = 99.3 mm2 = 1860 N/mm2 = 195 kN/mm2.
Solution Preliminary calculations at transfer Ac A = 250 250 = 62,500 mm2 Allowable stress for M35 concrete under direct compression at transfer
fcc,all = 0.8 0.51fci = 0.8 0.5135 = 14.3 N/mm2
Pe P + Ac At
fct,all = 0 N/mm2
Pe = 0.85P0
Substituting the values
0= 0.85P0 P + A A
Prestress at transfer
Indian Institute of Technology Madras
P0 = 397.21488 N = 591kN
Final calculations at transfer Ac = 62,500 - 397 = 62103 mm2 Stress in concrete fc = =P0 Ac 591,000 62,103 OK
= - 9.5 N/mm2
fc < fcc,all
E p = 195 kN/mm2
Ec = 5,000 35 = 29,580 N/mm2
At = 62,103 +
Stress in concrete
fc = Pe P + Ac At
fc = -
= - 1.4 N/mm2
No tensile stress in concrete. OK. Final calculations for ultimate strength
250
250
Nominal non-prestressed reinforcement is provided for resisting thermal and shrinkage cracks.
Calculation of Moment Demand For simply supported prestressed beams, the maximum moment at the span is given by the beam theory. For continuous prestressed beams, the analysis can be done by moment distribution method. The moment coefficients in Table 12 of IS:456 - 2000 can be used under conditions of uniform cross-section of the beams, uniform loads and similar lengths of span. The design is done for the critical section. For a simply supported beam under uniform loads, the critical section is at the mid span. For a continuous beam, there are critical sections at the supports and at the spans. For design under service loads, the following quantities are known. MDL MLL = moment due to dead load (excluding self-weight) = moment due to live load.
The material properties are selected before the design. The following quantities are unknown. The member cross-section and its geometric properties, MSW Ap Pe e = moment due to self-weight, = amount of prestressing steel, = the effective prestress, = the eccentricity.
There are two stages of design. 1) Preliminary: In this stage the cross-section is defined and Pe and Ap are estimated.
2) Final: The values of e (at the critical section), Pe, Ap and the stresses in concrete at transfer and under service loads are calculated. The stresses are checked with the allowable values. The section is modified if required.
Here, h is in meters and M is in kNm. M is the total moment excluding self-weight. 3) Select the type of section. For a rectangular section, assume the breadth b = h/2. 4) Calculate the self-weight or, estimate the self-weight to be 10% to 20% of the load carried. 5) Calculate the total moment MT including self-weight. The moment due to self-weight is denoted as Msw. 6) Estimate the lever arm (z). z 0.65h, if Msw is large (Msw > 0.3MT). z 0.5h, if Msw is small. 7) Estimate the effective prestress (Pe) Pe = MT / z, if Msw is large. Pe = MI L / z, if Msw is small. If Msw is small, the design is governed by the moment due to imposed load (MI L = MT MSW). 8) Considering fpe = 0.7fpk , calculate area of prestressing steel Ap = Pe / fpe. 9) Check the area of the cross-section (A). The average stress in concrete at service C/A (= Pe /A) should not be too high as compared to 50% of the allowable compressive stress fcc,all . If it is so, increase the area of the section to A = Pe /(0.5fcc,all).
Msw C T +
CGC CGS
kt kb
ct C e cb fb
C/A = P0/A
Stress in concrete
= gross area of cross section = maximum compressive stress in concrete at bottom edge = total height of the section
kt, kb = distances of upper and lower kern points, respectively, from CGC ct, cb = distances of upper and lower edges, respectively, from CGC P0 = prestress at transfer after initial losses.
Here, the magnitude of C or T is equal to P0. The value of P0 can be estimated as follows. a) 90% of the initial applied prestress (Pi) for pre-tensioned members. b) Equal to Pi for post-tensioned members. The value of Pi can be estimated from the amount of prestressing steel determined in the preliminary design. Pi = Ap(0.8fpk) (4-2.3)
Here, the permissible prestress in the steel is 0.8fpk, where fpk is the characteristic tensile strength. 2) Recompute the effective prestress Pe and the area of prestressing steel Ap. With increasing load, C further moves up. Under the service loads, C should lie within the kern zone to avoid tensile stress at the bottom. The highest permissible location of
C due to total load is at the top kern point (at a height kt above CGC) to avoid tensile stress at the bottom. The following sketch explains the highest possible location of C due to the total moment (MT).
ft h C T
+
MT
CGC CGS
kt kb
ct e cb C/A = Pe/A
0 Stress in concrete
(4-2.4)
Considering fpe = 0.7fpk , the area of prestressing steel is recomputed as follows. Ap = Pe / fpe 3) Recompute eccentricity e First the value of P0 is updated. The eccentricity e is recomputed with the updated value of P0. If the variation of e from the previous value is large, another cycle of computation of the prestressing variables can be undertaken. 4) Check the compressive stresses in concrete. The maximum compressive stress in concrete should be limited to the allowable values. At transfer, the stress at the bottom should be limited to fcc,all , where fcc,all is the allowable compressive stress in concrete at transfer (available from Figure 8 of IS:1343 (4-2.5)
- 1980). At service, the stress at the top should be limited to fcc,all , where fcc,all is the allowable compressive stress in concrete under service loads (available from Figure 7 of IS:1343 - 1980). a) At Transfer The stress at the bottom can be calculated from the average stress P0/A.
P0 h A ct
fb = -
(4-2.6)
To satisfy |fb| fcc,all, the area of the section (A) is checked as follows.
A P0h fcc,all ct
(4-2.7)
If A is not adequate then the section has to be redesigned. b) At Service The stress at the top can be calculated from the average stress Pe/A.
P h f =- e t A cb
(4-2.8)
To satisfy |ft| fcc,all, the area of the section (A) is checked as follows.
A Pe h fcc,all cb
(4-2.9)
At transfer, the stress at the bottom is calculated using the following stress profile.
CGC
e MSW/P0 C fb
C/A =P0/A
Figure 4-2.3
M P0 e - sw P0 P fb = - 0 - A I
Substituting I = Ar2 and r2/cb = kt
cb
(4-2.10)
e - Msw P P0 fb = - 0 1+ A kt
(4-2.11)
To satisfy |fb| fcc,all, the area of the section (A) is checked as follows. e - Msw P P0 A 0 1+ fcc,all kt
(4-2.12)
Example 4-2.1
Design a simply supported Type 1 prestressed beam with MT = 435 kNm (including an estimated MSW = 55 kNm). The height of the beam is restricted to 920 mm. The prestress at transfer fp0 = 1035 N/mm2 and the prestress at service fpe = 860 N/mm2. Based on the grade of concrete, the allowable compressive stresses are 12.5 N/mm2 at transfer and 11.0 N/mm2 at service.
The properties of the prestressing strands are given below. Type of prestressing tendon : 7-wire strand Nominal diameter Nominal area = 12.8 mm = 99.3 mm2
Solution
A) Preliminary design The values of h and MSW are given. 1) Estimate lever arm z.
Msw 55 = MT 435 = 12.5 %
Since MSW < 0.3 MT , Use z = 0.5h = 0.5 920 = 460 mm 2) Estimate the effective prestress. Moment due to imposed loads
4) Estimate the area of the section to have average stress in concrete equal to 0.5 fcc,all .
A= =
Pe 0.5fcc,all
920
1 410 CGC 2
ct = 460
3 Values in mm.
Check area of the section
I = 2I1 + I2 1 1 = 2 390 1003 + (390 100) 4102 + 100 7203 12 12 = 1.6287 1010 mm4
Square of the radius of gyration
r2 = I A 1.6287 1010 = 150,000 = 108,580 mm2
e= =
Msw + kb P0
Since Pe is very close to the previous estimate of 826 kN, Ap, P0 and e remain same. The tendons are placed in two ducts. The outer diameter of each duct is 54 mm. Select (10) 7-wire strands with
Ap = 10 99.3 = 993.0 mm2
b) At service
A = Pe h fcc,all cb
The governing value of A is 158,976 mm2. The section needs to be revised. The width of the flange is increased to 435 mm. The area of the revised section is 159,000 mm2. Another set of calculations can be done to calculate the geometric properties precisely. Designed cross-section at mid-span
435 100 100 CGC CGS 100 290 920 (10) 7-wire strands with P0 = 994 kN
The steps for Type 1 members are explained in Section 4.2, Design of Sections for Flexure (Part I).
The following material provides the steps for sections with small self-weight moment. For sections with large self- weight moment, the eccentricity e may need to be determined based on the cover requirements.
1) Calculate eccentricity e to locate the centroid of the prestressing steel (CGS). Under the self-weight, C may lie outside the kern region. The lowest possible location of C due to self-weight is determined by the allowable tensile stress at the top.
The following sketch explains the extreme location of C due to self-weight moment (Msw) at transfer.
fct,all h C T
+
MSW
CGC CGS
kt kb e1 e 2
ct C cb fb
Stress in concrete
fb fct,all h
= maximum compressive stress in concrete at bottom edge = allowable tensile stress in concrete at top edge = total height of the section
kt, kb = distances of upper and lower kern points, respectively, from CGC ct, cb = distances of upper and lower edges, respectively, from CGC e1 e2 P0
= distance between the bottom kern point and the location of compression = distance by which the compression travels from CGS due to self weight = prestress at transfer after initial losses.
From the previous figure, the shift of C due to self-weight gives an expression of e2. It is evident that if C is further shifted upwards by a distance e1 to the bottom kern point, there will be no tensile stress at the top. e2 = Msw P0
(4-3.1)
The value of e1 is calculated from the expression of stress corresponding to the moment due to the shift in C by e1.
(4-3.2)
fct,all Ak b P0
e1 + e2 =
Msw + fct,all Ak b P0
(4-3.3)
e=
Msw + fct,all Ak b + kb P0
(4-3.4)
The above expression can be compared with the expression of Type 1 member e = (Msw / P0) + kb. Note that the eccentricity has increased for a Type 2 member due to the allowable tensile stress fct,all. 2) Recompute the effective prestress Pe and the area of prestressing steel Ap. Under the total load, C may lie outside the kern region. The highest permissible location of C due to total load is determined by the allowable tensile stress at the bottom.
The following sketch explains the highest possible location of C due to the total moment (MT).
ft MT h C T
+
CGC CGS
e3 kt kb
ct cb fct,all
Stress in concrete
Figure 4-3.2
From the previous figure, the expression of e3 is obtained by the tensile stress generated due to the shift of C beyond the upper kern point.
Pe e3cb = fct,all I f I e3 = ct,all Pe cb e3 = fct,all Akt Pe
(4-3.5)
The above expression can be compared with the expression of Type 1 member Pe = MT /(e + kt). Note that the prestressing force has decreased for a Type 2 member due to the allowable tensile stress fct,all. This will lead to a decrease in the area of prestressing steel (Ap). Considering fpe = 0.7fpk, Ap is recomputed as follows.
Ap = Pe/ fpe
3) Recompute eccentricity e
(4-3.7)
First the value of P0 is updated. The eccentricity e is recomputed with the updated value of P0. If the variation of e from the previous value is large, another cycle of computation of the prestressing variables can be undertaken.
4) Check the compressive stresses in concrete The maximum compressive stress in concrete should be limited to the allowable values.
At transfer, the stress at the bottom should be limited to fcc,all , where fcc,all is the allowable compressive stress in concrete at transfer (available from Figure 8 of IS:1343 - 1980). At service, the stress at the top should be limited to fcc,all , where fcc,all is the allowable compressive stress in concrete under service loads (available from Figure 7 of IS:1343 - 1980).
a) At Transfer The stress at the bottom can be calculated from the stress diagram.
C C ( kb + e1 ) cb fb = - A I kbcb Ce1cb C fb = - 1+ 2 A r I
(4-3.8)
(4-3.9)
To satisfy |fb| fcc,all, the area of the section (A) is checked as follows.
C h fct,all + cb fcc,all A ct ct P0 h fcc,all ct - fct,all cb A P0 h A fcc,all ct - fct,all cb
(4-3.10)
b) At Service The stress at the top can be calculated from the stress diagram. C C ( kt + e3 ) ct A I kt ct Ce3ct C ft = - 1+ 2 A r I ft = From fct,all = Ce3cb / I, substituting Ce3 / I = fct,all /cb
ft = ft = C ct fct,all ct 1+ A cb cb C h fct,all ct A cb cb
(4-3.11)
(4-3.12)
To satisfy |ft| fcc,all, the area of the section (A) is checked as follows.
(4-3.13)
The following table shows a comparison of equations for Type 1 and Type 2 members.
Table 4-3.1
Eccentricity
e=
Msw + kb P0 MT e + kt
e=
Effective prestress
Pe =
Minimum area based on stress at bottom at transfer Minimum area based on stress at top at service
A A
P0 h fcc,all ct
A A
Pe h fcc,all cb
The following example shows the design of a Type 2 prestressed member. The same section was designed as a Type 1 member in Section 4.2, Design of Sections for Flexure (Part I). The solutions of the two examples are compared at the end.
Example 4-3.1
Design a simply supported Type 2 prestressed beam with MT = 435 kNm (including an estimated MSW = 55 kNm). The height of the beam is restricted to 920 mm. The prestress at transfer fp0 = 1035 N/mm2 and the prestress at service fpe = 860 N/mm2. Based on the grade of concrete, the allowable compressive stresses are 12.5 N/mm2 at transfer and 11.0 N/mm2 at service. 2.1 N/mm2 at transfer and 1.6 N/mm2 at service. The allowable tensile stresses are
Type of prestressing tendon : 7-wire strand Nominal diameter Nominal area = 12.8 mm = 99.3 mm2
Solution
A) Preliminary design
4) Estimate the area of the section to have average stress in concrete equal to 0.5 fcc,all . A= = Pe 0.5fcc,all
The following trial section has the required depth and area.
Trial cross-section
920
The section is symmetric about the horizontal axis. Hence, the CGC lies at mid depth. The section is divided into three rectangles for the computation of the geometric properties.
1 410 CGC 2
ct = 460
3 Values in mm.
Check area of the section
C) Final design
1) Calculate eccentricity e
P0 = Ap fp 0 = 960 1035 = 993.6 kN
e1 + e2 =
= = 130mm
Indian Institute of Technology Madras
Since Pe is substantially lower than the previous estimate of 826 kN, Ap, P0 and e need to be recalculated.
Ap = = Pe fpe
3) Recompute eccentricity e
P0 = Ap fp 0 = 7271035 = 752.4kN
e=
Msw + fct,all Akb + kb P0 55 103 + 2.1 150,000 236 103 + 236 752.4
= = 172 + 236
= 408 400 mm
Assuming the outer diameter of duct equal to 54 mm 1 Clear cover for the duct = 460 - 400 - 54 2 = 33 mm
The clear cover at the bottom is greater than 30 mm (Clause 11.1.6.2, IS: 1343 - 1980), which is satisfactory. The side cover in the web is slightly less than 30 mm. The thickness of the web can be increased to satisfy the requirement.
Since the value of e has changed from 366 mm to 400 mm, prestressing variables are recomputed.
Pe = MT - fct,all Akt e + kt 435 103 = = 592.0 kN 1.65 150,000 236 103 400 + 236
Ap = 799.3 = 695.1mm2
The tendons can be placed in one duct. The outer diameter of the duct is 54 mm.
P0 = 695.11035 = 719.4kN
Since the maximum possible eccentricity is based on cover requirement, the value of e is not updated.
A =
At service
A =
Pe h fcc,all cb - fct,all ct
The solutions from the examples of Type 1 and Type 2 members are placed together in the next figure for comparison.
435 100 100 CGC CGS 100 290 920 CGC 400 CGS 100 (7) 7-wire strands with P0 = 719 kN Type 2 100 100 920 335
Figure 4-3.3
Indian Institute of Technology Madras
The following observations can be made. 1) In Type 2 section, the amount of prestressing steel and the prestressing force are less than those in a Type 1 section. The area of cross-section is less for Type 2 section. Type 2 section is relatively economical.
2) The eccentricity in Type 2 section is larger than in Type 1 section. For unit prestressing force, the prestressing is more effective in Type 2 section.
(a) More room for Remarks on the sections tendons in lower flange. (b) Better stability during erection. Fabrication Space for reinforcement Easy Easy Less than adequate Expensive form work Good Difficult Torsionally stiff and strong.
Adequate
Good 1) Very
inefficient. Very good, z 0.7h 2) Small ultimate moment capacity 1) Especially 1) Good for long span and Msw/MT large 2) Msw/MT
1) Light load and good for long short span. 2) Msw/MT large. LL DL.
Very good
1) Topping serves to tie all Application of composite sections Good for building construction sections together. 2) No form required for composite pour. The different types of sections can be compared by a measure of flexural efficiency . The flexural efficiency is defined in terms of the radius of gyration r as follows. Long span buildings and bridges. Bridges
= =
r2 ct cb r 2 ct + cb ct cb h
> 0.33.
allowable values. Note that the limiting zone is a restriction for the CGS. The individual tendons may lie outside the limiting zone. The following figure shows the limiting zone (as the shaded region) for a simply supported beam subjected to uniformly distributed load.
C L CGC
Locus of emin
Locus of emax
Figure 4-4.1 Limiting zone for a simply supported beam
The limiting zone is determined from the maximum or minimum eccentricities of the CGS along the beam corresponding to the extreme positions of
C.
The maximum eccentricity (emax) at any section corresponds to the lowest possible location of C at transfer, that generates allowable tensile stress at the top of the section. Also, the maximum compressive stress at the bottom should be within the allowable value. The minimum eccentricity (emin) at any section corresponds to the highest possible location of C at service, that generates allowable tensile stress at the bottom of the section. Also, the maximum compressive stress at the top should be within the allowable value. The following material gives the expressions of emax and emin for Type 1 and Type 2 sections. The zone between the loci of emax and emin is the limiting zone of the section for placing the CGS. The values of emax and emin can be determined by equating the stresses at the edges of concrete with the allowable values. Else, explicit expressions of
emax and emin can be used. Here, the expressions of emax and emin based on allowable
tensile stress are given.
The following sketch shows the stress profile in concrete when C is at the lowest permissible location due to self-weight moment (Msw) at transfer.
0
CGC CGS
kt kb
ct C T emax c b
Stress in concrete
Internal forces
fb
Figure 4-4.2
From the shift of C due to self-weight, the following expression can be derived.
emax - k b = or, emax = Msw P0 Msw + kb P0
(4-4.2)
Note that since MSW varies, emax varies along the length of the beam. Also, the stress at the bottom needs to be checked to satisfy the condition |fb| fcc,all. At Service The following sketch shows the stress profile in concrete when C is at the highest possible location due to the total moment (MT).
ft kt kb C emin T ct cb
0 Stress in concrete
CGC CGS
Internal forces
Figure 4-4.3
From the shift of C due to total moment, the following expression can be derived.
emin + kt = or,
Indian Institute of Technology Madras
MT Pe MT - kt Pe
emin =
(4-4.3) Note that since MT varies, emin varies along the length of the beam. Also, the stress at the top needs to be checked to satisfy the condition |ft| fcc,all. If for a particular section emin is negative, it implies that the CGS can be placed above CGC. This happens near the supports.
Type 2 Section At Transfer The following sketch shows the stress profile in concrete when C is at the lowest permissible location due to self-weight moment (Msw) at transfer.
Internal forces
Stress in concrete
fb
Figure 4-4.4
(4-4.4)
Note that emax for a Type 2 section is larger than that for a Type 1 section due to the term fct,all Akb in the numerator. The stress at the bottom needs to be checked to satisfy the condition |fb| fcc,all. At Service The following sketch shows the stress profile in concrete when C is at the highest possible location due to the total moment (MT).
ft e3 kt kb C ct emin c b fct,all
Stress in concrete
CGC CGS
Internal forces
Figure 4-4.5
emin + kt =
or,
emin =
Note that emin for a Type 2 section is smaller than that for a Type 1 section due to the term fct,all Akt in the numerator. The stress at the bottom needs be checked to satisfy the condition |ft| fcc,all. The zone between emax and emin is the limiting zone of the section for placing the CGS for a given loading condition. The values of emax and emin for several sections can be determined at regular intervals along the length of the beam to get their loci. Note that the limiting zone for a Type 2 member is larger than the limiting zone for a Type 1 member. The following table shows a comparison of equations for Type 1 and Type 2 members. Table 4-4.2 Comparison of equations for Type 1 and Type 2 members Type 1 Maximum eccentricity
e= Msw + kb P0
Type 2
e=
M sw + fct,all Ak b + kb P0
Minimum eccentricity
e=
MT - kt Pe
e=
MT - fct,all Akt - kt Pe
The following example shows the calculation of limiting zone based on equating the stresses at the edges of concrete with the allowable values.
Example 4-4.1
For the Type 2 post-tensioned beam with a flanged section as shown, the span is 18 m. For uniform loads, the profile of the CGS is parabolic. The live load
moment at mid-span (MLL) is 648 kNm. The prestress after transfer (P0) is 1600 kN. Assume 15% loss at service.
Evaluate the limiting zone of CGS, if the allowable stresses at transfer and at service are as follows. For compression, fcc,all = 18.0 N/mm2 For tension, fct,all = 1.5 N/mm2.
500 200 150
200 250
Values in mm.
Solution
A) Calculation of geometric properties The section is divided into three rectangles for the computation of the geometric properties. The centroid of each rectangle is located from the soffit.
1 + 2 500 100 3
y
ct
CGC
900
cb
Values in mm.
Area of the section Area of 1 Area of 2 Area of 3 A = A1 = 500 200 = 100,000 mm2 = A2 = 600 150 = 90,000 mm2 = A3 = 250 200 = 50,000 mm2 = A1 + A2 + A3 = 240,000 mm2 Distance of CGC from the soffit y= A1 900 + A2 500 + A3 100 A = 583.3 mm
Therefore,
I1 =
Moment of inertia of 2
I2 =
Moment of inertia of 3
I3 =
Moment of inertia of the section I = I1 + I2 + I3 = (1.036 + 0.336 +1.184)1010 = 2.5521010 mm4 Calculation of moment due to self weight.
MSW =
B) Determination of limiting zone The values of emax and emin are determined by equating the stresses at the edges of concrete with the allowable values. The expression of stress is given below.
f =M
P Pey My A I I
+ + =
CGC
P
P/A
Pey/ I
My/ I
Limiting position at mid-span For emax, consider the load stage at transfer. i) Calculate e based on fb = 18.0 N/mm2.
Solving, e =
Solving, e =
emax = 455.2 mm
For emin, consider the load stage at service. i) Calculate e based on ft = 18.0 N/mm2.
Solving, e =
Solving, e =
emin = 418.4 mm
Indian Institute of Technology Madras
Solving, e =
emax = 309.6 mm
At service For fb = 1.5 N/mm2
fb = - 6.67 - 0.0366e = 1.5 N/mm2 Solving e = - 223.0 mm
Since the values of e are negative the CGS lies above CGC. The position of CGS closer to the CGC is selected.
emin = - 223.0 mm
Similarly, the values of emax and emin can be determined at regular intervals along the span. The limiting zone is available by joining the points by straight lines.
Indian Institute of Technology Madras
CL
223.0 309.6
From the sketch of the limiting zone, it is evident that the tendons can be spread out at the ends. This is necessary to anchor the tendons and reduce the stress concentration at the ends.
With reduced P0 at transfer, emax is increased. Thus, the limiting zone for placing the CGS and the available zone for the shift in C under service loads are also increased.
Notations The variables used in this section are as follows. A ct cb e ft fb = area of cross section of member = distance of the top of the section from CGC = distance of the bottom of the section from CGC = eccentricity of CGS with respect to CGC = stress at the top of the section = stress at the bottom of the section.
fcc,all = allowable compressive stress in concrete fct,all = allowable tensile stress in concrete I kt kb MT P0 Pe r Zt Zb = moment of inertia of cross section of member = distance of top kern point from CGC = distance of bottom kern point from CGC
MSW = moment due too self weight = total moment = prestress at transfer after immediate losses = prestress at service after long term losses = radius of gyration, r2 = I/A = section modulus corresponding to top of the section = I/ct = section modulus corresponding to bottom of the section = I/cb = ratio of prestressing forces = Pe /P0
There are essentially four stress conditions to be checked. These conditions are as follows.
The above expressions are algebraic inequalities where the stresses ft and fb are positive if tensile and negative if compressive. The allowable tensile stress fct,all is assigned a positive value and the allowable compressive stress fcc,all is assigned a negative value. The allowable stresses are explained in the Section 1.5, Concrete (Part I). It is to be noted that the values of fcc,all at transfer and at service are different. They are calculated based on the strength of concrete at transfer and at service, respectively. Similarly, the values of fct,all at transfer and at service can be different. As per IS:1343 1980, the values of fct,all at transfer and service are of course same. The stresses ft and fb in the four inequalities are expressed in terms of the initial prestressing force P0, the eccentricity e at the critical section of the member, the section properties A, Zt, Zb, kt, kb and the load variables Msw and MT. After transposition, 1/P0 is expressed in terms of e by linear inequality relationships. For a selected section, these relationships are plotted in the 1/P0 versus e plane. The acceptable zone shows the possible combinations of 1/P0 and e that satisfy all the four inequality relationships. A combination of P0 and e can be readily calculated from the acceptable zone.
The method is explained in a general form. For Type 1, Type 2 and Type 3 members, the value of allowable tensile stress (fct,all) is properly substituted. For Type 1 members, fct,all = 0 N/mm2.
At Transfer
The following sketch shows the variation of stress in concrete after the transfer of prestress and due to the self weight.
ft MSW e P0 fb
Figure 4-5.1
ct cb
The stress at the top is calculated from P0, e, Msw as follows. P0 P0ect Msw ct + A I I P ect Msw = 0 -1+ 2 A r Zt = P0 e M -1+ - sw A k b Zt
(4-5.1)
ft = -
The inequality relationship satisfying the stress at the top is expressed in terms of 1/P0 and e as follows.
ft fct,all P0 e M -1+ - sw fct,all A k b Zt or, (-1+ e/k b ) 1 P0 Msw fct,all + Z A t
(4-5.2)
The following sketch shows the plot of inequality relationship. The straight line given by the above inequality is plotted in the 1/P0 versus e plane and the acceptable zone is shaded.
1 P0
Acceptable
kb
Figure 4-5.2
The following expression relates the stress at the bottom with the load and section variables. fb = P0 P0ecb Msw cb + A I I P ecb Msw =- 0 1+ 2 + A r Zb =P0 e 1+ A kt Msw + Z b
(4-5.3)
The inequality relationship satisfying the stress at the bottom is expressed as follows.
fb fcc,all P0 e 1+ A kt Msw + Z fcc,all b ( 1+ e/kt ) 1 or, P0 Msw -fcc,all + Z A b
(4-5.4)
Acceptable
- kt
Figure 4-5.3
At Service
The following sketch shows the variation of stress in concrete at service and due to the total moment.
ft MT e P 0 fb
Figure 4-5.4
ct cb
Here, Pe is expressed as P0, where is the ratio of effective prestress (Pe) and prestress at transfer (P0). The expression of the stress at the top is given below.
ft = P0 P0ect MT ct + A I I P ect MT = 0 -1+ 2 A r Zt = P0 e M -1+ - T A k b Zt
(4-5.5)
The inequality relationship satisfying the stress at the top is expressed as follows. ft fcc,all P0 e M -1+ - T fcc,all A k b Zt e -1+ kb 1 or, P0 MT fcc,all + Z A t
(4-5.6)
The following sketch shows the plot of inequality relationship. The straight line given by the above inequality is again plotted in the 1/P0 versus e plane and the acceptable zone is shaded.
1 P0
Acceptable
kb
Figure 4-5.5
The following expression relates the stress at the bottom with the load and section variables.
fb = P0 P0ecb MT cb + I I A P ecb MT =- 0 1+ 2 + A r Zb =P0 e 1+ A kt MT + Z b
(4-5.7)
(4-5.8)
Acceptable
- kt
Figure 4-5.6
Next, the four lines are plotted simultaneously. The common region is the acceptable zone.
(4-5.6)
1 P0
Acceptable zone
- kt
kb
e
Acceptable zone
Figure 4-5.7
A combination of a trial section, prestressing force (P0) and eccentricity (e) at the critical section, can be plotted in the form of the above graph. If the point lies within the acceptable zone, then the combination is valid.
Example 4-5.1
The section shown is designed as a Type 1 member with MT = 435 kNm (including an estimated MSW = 55 kNm). The height of the beam is restricted to 920 mm. The prestress at transfer fp0 = 1035 N/mm2 and the prestress at service fpe = 860 N/mm2. Based on the grade of concrete, the allowable compressive stresses are 12.5 N/mm2 at transfer and 11.0 N/mm2 at service.
Type of prestressing tendon : 7-wire strand Nominal diameter Nominal area = 12.8 mm = 99.3 mm2
For the section, find the acceptable zone by Magnels graphical method. Compare the designed values of eccentricity (e) and the inverse of prestressing force at transfer (1/P0) with the acceptable zone.
e = 290
100
Solution
A) Calculation of geometric properties
The section is symmetric about the horizontal axis. Hence, the CGC lies at mid depth. The section is divided into three rectangles for the computation of the geometric properties.
1 410 CGC 2
ct = 460
3 Values in mm.
Area of the section
Section moduli
Zb = Zt = I = 38,712,174 mm3 ct
Kern levels
k b = kt =
r2 = 243.5 mm ct
B) Calculation of the inequality relationships of Magnels graphical method Ratio of effective prestress and prestress at transfer = = Pe P0 fpe fp 0 = 860 1035
0.0015
1/P 0 (1/kN)
At Transfer fb fcc,all
( 1+ e/kt ) 1 P0 Msw -fcc,all + Z A b
1 1+ e 243.5 P0 55106 12.5 + 159,000 38,712,174 1 1+ e 2,213,397.9 243.5 The relationship is plotted in the following graph.
1/P 0 (1/kN)
0.0015 0.001 0.0005 0 -300 -200 -100 0 100 200 300 e (mm)
At Service
ft fcc,all
e -1+ kb 1 P0 MT fcc,all + Z A t
1 (-1+ e 243.5)0.83 P0 435106 -11.0 + 159,000 38,712,174 1 -1+ e 45,358.0 243.5
0.0015 1/P 0 (1/kN) 0.001 0.0005 0 -300 -200 -100 0 100 200 300 e (mm)
At Service
fb fct,all
e 1+ k 1 t P0 MT -fct,all + Z A b
The four relationships are plotted in the following graph. The acceptable zone is shown. The zone is zoomed in the next graph.
0.0015
1/P 0 (1/kN)
0.0012
1/P 0 (1/kN)
0.001
The calculated values of e and 1/P0 for the Type 1 section are as follows. e = 290 mm 1/P0 = 1/(994 kN) = 0.001kN1. The solution of the design is shown in the graphs. It lies in the acceptable zone.
Introduction The detailing of the prestressing tendons and the reinforcing bars is important to satisfy the assumptions in the analysis, proper placement of concrete and durability. After the design calculations, drawings are prepared for construction. These drawings are referred to as the design drawings which become a part of the construction documents. The steel fabricator may prepare another set of drawings which are called shop drawings. These drawings are similar to the design drawings but they contain additional information such as the bar designations and bar bending schedule. It is essential to show the detailing in the design drawings so that there is no ambiguity during construction. It is also necessary to check the details in the shop drawings. IS:1343 - 1980 specifies some minimum requirements. Here, these requirements are briefly mentioned. There are requirements for the non-prestressed reinforcement as per IS:456 - 2000 which are not covered here. The detailing requirements for shear and torsion are covered in Section 5.3, Design for Shear (Part II) and Section 5.6, Design for Torsion (Part II), respectively. Of course the detailing is best learned by preparing drawings for construction projects.
y = (
Here, L = span of the beam x = distance from one end
4ym ) (L - x ) L2
(4-6.1)
Y = vertical displacement of the tendon (from the level at the ends) at distance x
Ym = vertical displacement of the tendon (from the level at the ends) at the middle of the beam The following sketch shows the plot of the equation.
L ym y
Figure 4-6.1 Profile of a parabolic tendon
Note that an individual tendon may be displaced from the CGS. Hence, the tendon need not pass through the CGC at the ends. The figure below shows the parabolic profiles of the ducts for placing tendon in a simply supported bridge girder.
Figure 4-6.2
Tendon profiles in a simply supported bridge girder (Courtesy: Larsen & Toubro Rambll)
For continuous beams or slabs, parabolic profiles at the spans and at the supports are connected to get the continuous profile of a tendon. The following sketch shows the profile of the CGS in a continuous beam. The eccentricities of the CGS at the end span, first interior support and first interior span are represented as e1, e2 and e3 respectively.
e1
e2
e3
Figure 4-6.3
A parabolic segment connects a point of maximum eccentricity with a point of contraflexure. A point of contraflexure is the location where the curvature of the profile reverses. It is also known as the inflection point. For varying spans and loading, the segments on two sides of a point of maximum eccentricity, may not be symmetric. In the second sketch of the above figure, the different parabolas between the points of maximum eccentricity and the points of contraflexure are numbered. The convex segment over a support is required to avoid a kink in the tendon. The length of a convex segment is determined based on the minimum radius of curvature for the type of tendon. A parabolic segment satisfies two conditions. 1) It has zero slope at the point of maximum eccentricity. 2) At a point of contraflexure, the slopes of the parabolic segments on both sides should match. The equation of a parabolic segment is given below.
y = ym
x l
(4-6.2)
y l
Figure 4-6.4 In the previous equation, l = length of the parabolic segment = vertical displacement of the profile at distance x eccentricity. x = distance from the point of maximum eccentricity y
ym x
The origin is selected at the point of maximum eccentricity at a critical section. The equation satisfies the first boundary condition of zero slope at the point of maximum eccentricity. The length (l) is determined from the requirement of minimum radius of curvature at the support. should match. The following photo shows the profiles of the tendons in a continuous bridge girder. The displacement ym is determined from the boundary condition that at the point of contraflexure, the slopes of the segments on both sides
Figure 4-6.5
The profile is implemented by the use of hangers or cross bars or chairs of varying depth at regular intervals. In beams, the duct is supported by hangers from the top bars or by cross bars attached to the stirrups. The depth of the hanger or cross bar at a
Indian Institute of Technology Madras
location can be calculated from the equation of the profile. supported on chairs resting on the form work.
The CGS of the tendon shifts from the centre line of the duct after stretching. The following sketches show the shifts at the low and high points of the tendon. The shift in the CGS is available from the type of tendon used and can be accounted for in precise calculations.
CGS
CGS
Figure 4-6.6
(4-6.3)
As
Ap
Figure 4-6.7
The minimum reinforcement can be reduced to 0.15% A, if high yield strength deformed bars are used.
Minimum Longitudinal Reinforcement with Unbonded Tendon In a post-tensioned member when the ducts are not grouted, beyond the cracking load, the number of cracks is small and the crack width is large. To reduce the crack width, a minimum amount of non-prestressed reinforcement should be provided. Since the non-prestressed reinforcement is bonded to the concrete, there are several cracks with small crack width.
Unbonded tendon
Non-prestressed reinforcement
Figure 4-6.8 with unbonded tendon
Unbonded tendon
As per the code of the American Concrete Institute (ACI 318), the minimum amount of such reinforcement (As) is 0.4% At , where At is the area under tension between the
centroid of the section (CGC) and the tension edge. The above reinforcement is not intended to provide flexural strength.
CGC
At
Unbonded tendon
As
Figure 4-6.9 Cross-section of a beam showing longitudinal reinforcement and area under tension
Minimum Side Face Reinforcement When the depth of the web exceeds 500 mm, a minimum amount of longitudinal reinforcement should be placed at each face (side face) of the web to check thermal and shrinkage cracks. According to Section 18.6.3.3-b, the minimum amount of side face reinforcement (As,sf ) is given as follows. Minimum As,sf = 0.05% Aw Here, Aw = vertical area of the web. The maximum spacing of the bars is 200 mm. (4-6.4)
As,sf
As,sf
Figure 4-6.10
Clear cover
Figure 4-6.11 Cross-section of a beam showing cover
According to Section 11.1.6, the minimum cover requirements are as follows. For pre-tensioned members, minimum cover for tendons is 20 mm. For post-tensioned members, minimum cover for sheathing (duct) is 30 mm or size of the tendon. The minimum cover should be increased by 10 mm in aggressive environment.
Minimum Spacing Requirements A minimum clear spacing of the tendons or reinforcing bars is necessary for the flow of concrete during casting and for the bond between concrete and steel.
Clear spacing
Figure 4-6.12
According to Section 11.1.7, the minimum spacing requirements are as follows. For single wires in a pre-tensioned member, Clear spacing 3 wire diameter 1 maximum aggregate size.
Indian Institute of Technology Madras
For large bars or tendons, Clear spacing 40 mm maximum size of tendon / bar maximum aggregate size + 5 mm. For grouped tendons (maximum four tendons per group), the requirement is for the spacing between the groups of tendons.
Vertical spacing
Horizontal spacing
Figure 4-6.13
According to Section 11.1.8, for grouped tendons the spacing requirements are as follows. Horizontal spacing 40 mm maximum aggregate size + 5 mm Vertical spacing 50 mm.
Anchorage of Reinforcement In a partially prestressed section, where the non-prestressed reinforcement contributes to flexural strength, the development length of the bars needs to be checked at the critical section. The bars should be anchored at the supports by hooks to avoid anchorage failure. The following photo shows the fabrication of the reinforcement for a post-tensioned boxgirder of a bridge.
Figure 4-6.14
Fabrication of reinforcement
Stress in an Uncracked Beam Types of Cracks Components of Shear Resistance Modes of Failure Effect of Prestressing Force
Introduction The analysis of reinforced concrete and prestressed concrete members for shear is more difficult compared to the analyses for axial load or flexure. The analysis for axial load and flexure are based on the following principles of mechanics. 1) Equilibrium of internal and external forces 2) Compatibility of strains in concrete and steel 3) Constitutive relationships of materials. The conventional analysis for shear is based on equilibrium of forces by a simple equation. The compatibility of strains is not considered. The constitutive relationships (relating stress and strain) of the materials, concrete or steel, are not used. The strength of nature. Shear stresses generate in beams due to bending or twisting. The two types of shear stress are called flexural shear stress and torsional shear stress, respectively. In this section, the analysis for shear refers to flexural shear stress. The torsional shear stress is covered in Section 5.4, Analysis for Torsion. To understand flexural shear stress, the behaviour of a simply supported beam under uniformly distributed load, without prestressing, will be explained first. prestressing force will be subsequently introduced. following sequence. The effect of The presentation will be in the each material corresponds to the ultimate strength. The strength of concrete under shear although based on test results, is empirical in
1) Stresses in an uncracked (homogenous) beam. 2) Types of cracks generated due to the combination of flexure and shear. 3) Components of shear resistance and the modes of failure. 4) Effect of prestressing force.
b h
Shear force diagram Momentd iagram
Figure 5-1.1
1 2
Under a general loading, the shear force and the moment vary along the length. The normal stress and the shear stress vary along the length, as well as along the depth. The combination of the normal and shear stresses generate a two-dimensional stress field at a point. At any point in the beam, the state of two-dimensional stresses can be expressed in terms of the principal stresses. The Mohrs circle of stress is helpful to understand the state of stress. Before cracking, the stress carried by steel is negligible. When the principal tensile stress exceeds the cracking stress, the concrete cracks and there is redistribution of stresses between concrete and steel. For a point on the neutral axis (Element 1), the shear stress is maximum and the normal stress is zero. The principal tensile stress (1) is inclined at 45 to the neutral axis. The following figure shows the state of in-plane stresses.
v
1
2 2 = 45o
2 1 Mohrs circle
Principal stresses
At the level of neutral axis, the normal stress is zero and the shear stress is maximum. An element at that level is under pure shear. A state of pure shear can be conceived as a state of biaxial tensile-compressive stresses. These principle stresses are inclined at 45 with respect to the axis of the beam. It is necessary to study the principle stresses to understand the cracking of concrete. The Mohrs circle is a representation of the state of in-plane stresses on surfaces of various inclinations passing through a point. The horizontal and vertical axes represent the normal and shear stresses, respectively. For a state of pure shear, the centre of the Mohrs circle coincides with the origin of the axes. It is expected that the reader is familiar with these concepts from a course in strength of materials. Since the shear force is maximum near the supports, cracks due to shear occur near the supports. The cracks are formed around the neutral axis and perpendicular to the principal tensile stress (1). The cracks are thus inclined at 45 to the axis of the beam. The following sketch shows the inclination of the cracks forming at the neutral axis.
v
1
1 2
Figure 5-1.3
For a point near the bottom edge of the beam (Element 2), the normal stress is maximum and the shear stress is close to zero. The principal tensile stress (1) is almost parallel to the bottom edge. The angle of inclination of 1 with respect to the axis of the beam () is much smaller than 45. The following figure shows the state of inplane stresses.
Indian Institute of Technology Madras
v
2
2 2 < 45o 2
(f,v) 1
Principal stresses
Mohrs circle
Adjacent to the bottom edge (edge under tension), the tensile stress due to flexure is maximum and the shear stress is zero. The state of stress is nearly uniaxial tensile stress. The principal compressive stress is negligible. The Mohrs circle is shifted towards the axis of principal tensile stress. Since the moment is maximum at mid span, cracks due to flexure occur near mid span. The cracks are formed at the bottom edge and perpendicular to 1. Since 1 is parallel to the edge, the cracks are perpendicular to the edge.
v
2
Figure 5-1.5
The previous concepts can be used to develop the principal stress trajectories. The following figure shows the trajectories for a simply supported beam under a uniformly distributed load. The crack pattern can be predicted from these trajectories.
Figure 5-1.6
b) Growth of flexural cracks and formation of flexure shear and web shear cracks.
Flexural cracks
Vcz Va Vs VP
Vd
Figure 5-1.8
The notations in the previous figure are as follows. Vcz = Shear carried by uncracked concrete Va = Shear resistance due to aggregate interlock Vd = Shear resistance due to dowel action Vs = Shear carried by stirrups Vp = Vertical component of prestressing force in inclined tendons The magnitude and the relative value of each component change with increasing load.
Diagonal tension failure In this mode, an inclined crack propagates rapidly due to inadequate shear reinforcement.
Figure 5-1.9
Shear compression failure There is crushing of the concrete near the compression flange above the tip of the inclined crack.
Figure 5-1.10
Shear tension failure Due to inadequate anchorage of the longitudinal bars, the diagonal cracks propagate horizontally along the bars.
Figure 5-1.11
Web crushing failure The concrete in the web crushes due to inadequate web thickness.
Figure 5-1.12
Arch rib failure For deep beams, the web may buckle and subsequently crush. anchorage failure or failure of the bearing. There can be
Figure 5-1.13
The objective of design for shear is to avoid shear failure. The beam should fail in flexure at its ultimate flexural strength. Hence, each mode of failure is addressed in the design for shear. The design involves not only the design of the stirrups, but also limiting the average shear stress in concrete, providing adequate thickness of the web and adequate development length of the longitudinal bars.
With increase in the load beyond the cracking load, the cracks generate in a similar sequence. But, the inclinations of the flexure shear and web shear cracks are reduced depending on the amount of prestressing and the profile of the tendon. The effect of prestressing force is explained for a beam with a concentric effective prestressing force (Pe).
Pe
Pe
Figure 5-1.14
A simply supported beam under concentric prestress and uniformly distributed loads
For a point at the neutral axis (Element 1), there is normal stress due to the prestressing force (fpe). The principal tensile stress (1) is inclined to the neutral axis at an angle greater than 45. With the combination of shear stress, the principal compressive stress (2) is inclined to the neutral axis at an angle much smaller than 45. The following figure shows the state of in-plane stresses.
v fpe
1
fpe
(fpe,v) 2
2 > 90 1
Principal stresses
Mohrs circle
In the following figure, the formation of cracks for a prestressed beam with large spanto-depth ratio and uniformly distributed loading is shown. This figure can be compared with that for a reinforced concrete beam.
Pe
Figure 5-1.16 Formation of cracks in a prestressed beam
Pe
After cracking, in presence of prestressing force, the length and crack width of a diagonal crack are low. Thus, the aggregate interlock and zone of concrete under compression are larger as compared to a non-prestressed beam under the same load. Hence, the shear strength of concrete (Vc) increases in presence of prestressing force. This is accounted for in the expression of Vc.
Design of Stirrups The design is done for the critical section. The critical section is defined in Clause 22.6.2 of IS:456 - 2000. In general cases, the face of the support is considered as the critical section. When the reaction at the support introduces compression at the end of the beam, the critical section can be selected at a distance effective depth from the face of the support. The effective depth is selected as the greater of dp or ds. dp = depth of CGS from the extreme compression fiber ds = depth of centroid of non-prestressed steel. Since the CGS is at a higher location near the support, the effective depth will be equal to ds. To vary the spacing of stirrups along the span, other sections may be selected for design. Usually the following scheme is selected for beams under uniform load. 1) Close spacing for quarter of the span adjacent to the supports. 2) Wide spacing for half of the span at the middle.
For large beams, more variation of spacing may be selected. The following sketch shows the typical variation of spacing of stirrups. The span is represented by L.
L/4
Figure 5-2.1
L/2
L/4
Vc includes Vcz (contribution from uncracked concrete), Va (aggregate interlock) and Vd (dowel action). The value of Vc depends on whether the section is cracked due to flexure. Section 22.4 of IS:1343 - 1980 gives two expressions of Vc, one for cracked section and the other for uncracked section. Usually, the expression for the uncracked section will govern near the support. The expression for the cracked section will govern near the mid span. Of course, both the expressions need to be evaluated at a particular section. The lower value obtained from the two expressions is selected. For uncracked sections, Vc = Vco Vc = 0.67bD ft 2 + 0.8fcp ft Vco is the shear causing web shear cracking at CGC. In the above expression, b = breadth of the section = bw, breadth of the web for flanged sections
Indian Institute of Technology Madras
(5-2.2)
D = total depth of the section (h) ft = tensile strength of concrete = 0.24fck = Pe/A. The value of fcp is taken as positive (numeric value). Note that, a reduced effective prestress needs to be considered in the transmission length (explained in Section 7.1) region of a pre-tensioned beam. fcp = compressive stress in concrete at CGC due to the prestress
The previous equation can be derived based on the expression of the principal tensile stress (1) at CGC.
v fcp fcp
1 2
(fcp,v) 2 1
Principal stresses
Mohrs circle
Figure 5-2.2
The principal tensile stress is equated to the direct tensile strength of concrete (ft). 1 = == ft In the previous equation, I = gross moment of inertia Q = At y At = area of section above CGC
y = vertical distance of centroid of At from CGC.
fcp 2 fcp 2
+ +
2 fcp
4
2 fcp
+ v2
2
Vc 0Q + 4 Ib
(5-2.3)
At
+ CGC
Figure 5-2.3
Cross-section of a beam showing the variables for calculating shear stress in the web
0.67bD ft 2 + 0.8fcp ft
The term 0.67bD represents Ib/Q for the section. It is exact for a rectangular section and conservative for other sections. To be conservative, only 80% of the prestressing force is considered in the term 0.8fcp. For a flanged section, when the CGC is in the flange, the intersection of web and flange is considered to be the critical location. The expression of Vc0 is modified by substituting 0.8fcp with 0.8 (the stress in concrete at the level of the intersection of web and flange). In presence of inclined tendons or vertical prestress, the vertical component of the prestressing force (Vp) can be added to Vc0.
Vcr is the shear corresponding to flexure shear cracking. The term (1 0.55fpe /fpk)cbd is the additional shear that changes a flexural crack to a flexure shear crack. The notations in the previous equation are as follows. fpe = effective prestress in the tendon after all losses 0.6fpk fpk = characteristic strength of prestressing steel
CGS. The values are plotted in the next figure. b = breadth of the section = bw , breadth of the web for flanged sections d = distance from the extreme compression fibre to the centroid of the tendons at the section considered M0 = moment initiating a flexural crack Mu = moment due to ultimate loads at the design section Vu = shear due to ultimate loads at the design section.
1.2
c (N/mm2)
0.8 0.4 0 0 1
M30
2 A p /bd x 100
M40
Figure 5-2.4
The term (M0/Mu)Vu is the shear corresponding to the moment M0, that decompresses (nullifies the effect of prestress) the tension face and initiates a flexural crack. The expression of M0 is given below.
M0 = 0.8fpt I y
(5-2.7)
In the above expression, fpt = magnitude of the compressive stress in concrete at the level of CGS due to prestress only. An equal amount of tensile stress is required to decompress the concrete at the level of CGS. The corresponding moment is fptI / y. In the expression of M0, I = gross moment of inertia y = depth of the CGS from CGC. The factor 0.8 implies that M0 is estimated to be 80% of the moment that decompresses the concrete at the level of CGS. Since the concrete is cracked and the inclination of tendon is small away from the supports, any vertical component of the prestressing force is not added to Vcr.
Maximum Permissible Shear Stress
To check the crushing of concrete in shear compression failure, the shear stress is limited to a maximum value (c,max). The value of c,max depends on the grade of concrete and is given in Table 7 of IS:1343 - 1980.
Vu c,max bdt
(5-2.8)
In the previous expression, dt = greater of dp or ds dp = depth of CGS from the extreme compression fiber ds = depth of centroid of regular steel Vu = shear force at a section due to ultimate loads.
6 c, max (N/mm )
2
0 30 40 50
2
60
f ck (N/mm )
Figure 5-2.5
Asv V -V = u c sv 0.87fy dt
The notations in the above equation are explained. sv = spacing of the stirrups dt = greater of dp or ds dp = depth of CGS from the extreme compression fiber ds = depth of centroid of non-prestressed steel fy = yield stress of the stirrups
(5-3.1)
Design of Stirrups for Flanges For flanged sections, although the web carries the vertical shear stress, there is shear stress in the flanges due to the effect of shear lag. Horizontal reinforcement in the form of single leg or closed stirrups is provided in the flanges. the shear stress in the flange at the face of the web. The following figure shows
Figure 5-3.1
The horizontal reinforcement is calculated based on the shear force in the flange. The relevant quantities for the calculation based on an elastic analysis are as follows. 1) Shear flow (shear stress width) 2) Variation of shear stress in a flange (f) 3) Shear forces in flanges (Vf). 4) Ultimate vertical shear force (Vu) The following sketch shows the above quantities for an I-section (with flanges of constant widths).
f max
bf Df Vf Vf Vu Vf Vf
Shear flow
Figure 5-3.2
Shear forces
Shear flow and shear forces in an I-section
Vf =
Here, bf Df = breadth of the flange = depth of the flange
f max bf
2 2
Df
(5-3.2)
f max =
Here, Vu = ultimate vertical shear force I
Vu A1 y I Df
(5-3.3)
= moment of inertia of the section. = distance of centroid of half of the flange from the neutral axis at CGC.
A1
Figure 5-3.3
Cross-section of a beam showing the variables for calculating shear stress in the flange
The amount of horizontal reinforcement in the flange (Asvf) is calculated from Vf. Asvf = Vf 0.87fy (5-3.4)
The spacing of stirrups (sv) is restricted so that a diagonal crack is intercepted by at least one stirrup. This is explained by the following sketch.
dp
bw
ds h
Cross-section
Figure 5-3.4
Elevation
sv
As per Clause 22.4.3.2, the maximum spacing is 0.75dt or 4bw , whichever is smaller. When Vu is larger than 1.8Vc , the maximum spacing is 0.5dt. The variables are as follows. bw = breadth of web dt = greater of dp or ds dp = depth of CGS from the extreme compression fiber ds = depth of centroid of non-prestressed steel Vu = shear force at a section due to ultimate loads Vc = shear capacity of concrete.
Minimum Amount of Stirrups
A minimum amount of stirrups is necessary to restrict the growth of diagonal cracks and subsequent shear failure. For Vu < Vc, minimum amount of transverse reinforcement is provided based on the following equation. Asv 0.4 = bsv 0.87fy
(5-3.5)
b = breadth of the section = bw, breadth of the web for flanged sections. If Vu < 0.5Vc and the member is of minor importance, stirrups may not be provided. Another provision for minimum amount of stirrups (Asv,min) is given by Clause 18.6.3.2 for beams with thin webs. The minimum amount of stirrups is given in terms of Awh , the horizontal sectional area of the web in plan. The area is shown in the following sketch.
A Elevation
Awh
Section A - A
Figure 5-3.5
In presence of dynamic load, Asv,min = 0.3% Awh = 0.2% Awh , when h 4bw With high strength bars, Asv,min = 0.2% Awh = 0.15% Awh , when h 4bw In absence of dynamic load, when h > 4bw Asv,min = 0.1% Awh There is no specification for Asv,min when h 4bw .
Anchorage of Stirrups
The stirrups should be anchored to develop the yield stress in the vertical legs. 1) The stirrups should be bent close to the compression and tension surfaces, satisfying the minimum cover. 2) Each bend of the stirrups should be around a longitudinal bar. The diameter of the longitudinal bar should not be less than the diameter of stirrups. 3) The ends of the stirrups should be anchored by standard hooks. 4) There should not be any bend in a re-entrant corner. In a re-entrant corner, the stirrup under tension has the possibility to straighten, thus breaking the cover concrete. The following sketches explain the requirement of avoiding the bend of a stirrup at a reentrant corner.
Re-entrant corner
Incorrect detailing
Figure 5-3.6
Correct detailing
To check web crushing failure, The Indian Roads Congress Code IRC:18 - 2000 specifies a minimum thickness of the web for T-sections (Clause 9.3.1.1). The minimum thickness is 200 mm plus diameter of the duct hole.
After a member is designed for flexure, the self-weight is known. It is included as dead load. The grade of concrete is known from flexure design. The grade of steel for stirrups is selected before the design for shear. As per IS:1343 - 1980, the grade of steel is limited to Fe 415. The following quantities are unknown. Vc = shear carried by concrete Asv = total area of the legs of stirrups within a distance sv sv = spacing of stirrups. The steps for designing stirrups along the length of a beam are given below. 1) Calculate the shear demand Vu at the critical location. 2) Check (Vu / bdt) < c,max. If it is not satisfied, increase the depth or breadth of the section. Here, b is the breadth of the web (bw) and dt is larger of dp and ds. 3) Calculate the shear capacity of concrete Vc from the lower of Vco and Vcr. In presence of inclined tendons or vertical prestress, the vertical component of the prestressing force (Vp) can be added to Vc0. 4) Calculate the requirement of shear reinforcement through Asv / sv . Compare the value with the minimum requirement. 5) Calculate the maximum spacing and round it off to a multiple of 5 mm. 6) Calculate the size and number of legs of the stirrups based on the amount required, type of section and space to accommodate. Repeat the calculations for other locations of the beam, if the spacing of stirrups needs to be varied.
Example 5-3.1
Design the stirrups for the Type 1 prestressed beam with the following section (location of tendons shown at mid span).
435 100 100 CGC CGS 100 290 920 (10) 7-wire strands with Pe = 826 kN
The properties of the sections are as follows. A = 159,000 mm2, I = 1.7808 1010 mm4 Ap = 960 mm2 The grade of concrete is M 35 and the characteristic strength of the prestressing steel (fpk) is 1470 N/mm2. The effective prestress (fpe) is 860 N/mm2. The uniformly distributed load including self weight, is wT = 30.2 kN/m. The span of the beam (L) is 10.7 m. The width of the bearings is 400 mm. The clear cover to longitudinal reinforcement is 30 mm.
Solution
1) Calculate Vu at the face of the support (neglecting the effect of compression in concrete).
L Vu = 1.5 wT - x 2 10.7 = 1.530.2 - 0.2 2 = 233.3 kN
Indian Institute of Technology Madras
Here, x denotes half of the width of bearing. x = 200 mm. 2) Check (Vu / bdt) < c,max . Effective depth dt = total depth cover diameter of stirrups longitudinal bar. Assume the diameter of stirrups to be 8 mm. diameter of
1 dt = 920 - 30 - 8 - 12 2 = 876 mm
Vu 233.3 103 = bw dt 100 876 = 2.7 N/mm2
fcp =
Vco = 0.67bD ft 2 + 0.8fcp ft = 0.67 100 920 1.422 + 0.85.19 1.42 = 173.4 kN
The vertical component of the prestressing force can be found out from the equation of the parabolic tendon.
y =
4ym x (L - x ) L2
L y
The following is the expression of the slope of the parabolic tendon.
ta n = dy 4ym = (L - 2 x ) dx L2
ym
ym
y
At x = 0.2 m, y = 20 mm, dy/dx = 0.105 and = 6.0.
Vp = Pe sin = 826 0.104 = 86.0kN
Vp
Pe
Here,
fpt = -
M0 = 0.85.21
Therefore,
Vu < Vc .
4) Calculate Asv / sv . Provide minimum steel.
5) Calculate maximum spacing sv = 0.75 dt sv = 4bw = 0.75 876 = 656 mm = 4 100 = 400 mm
Select sv = 400 mm. 6) Calculate the size and number of legs of the stirrups Select fy = 250 N/mm2.
Asv = bw sv 0.4 0.87fy 0.4 0.87 250
Provided amount of stirrups is larger. OK. Provide same spacing of stirrups throughout the span. Design of stirrups for flange
1 A1 = bf Df 2 1 = 435 100 2 = 21750 mm2
y = 410 mm
f max =
= Vu A1 y I Df
Vf =
Asvf = =
Asvf = Df sv
Stresses in an Uncracked Beam Crack Pattern Under Pure Torsion Components of Resistance for Pure Torsion Modes of Failure Effect of Prestressing Force
Introduction The analysis of reinforced concrete and prestressed concrete members for torsion is more difficult compared to the analyses for axial load or flexure.
The analysis for axial load and flexure are based on the following principles of mechanics. 1) Equilibrium of internal and external forces 2) Compatibility of strains in concrete and steel 3) Constitutive relationships of materials.
The conventional analysis of reinforced concrete and prestressed concrete members for torsion is based on equilibrium of forces by simple equations. The compatibility of strains in concrete and steel reinforcement is not considered.
The strength of each material, concrete or steel, corresponds to the ultimate strength. The constitutive relationship of each material, relating stress and strain, is not used.
Torsion generated in a member can be classified into two types based on the necessity of analysis and design for torsion. 1) Equilibrium torsion: This is generated due to loading eccentric to the centroidal axis. For example, a) in a beam supporting cantilever slab or precast slab or floor joists on one side, b) in a (curved) bridge deck subjected to eccentric live load and c) in an electric pole subjected to loads from wires on one side.
The torsion demand is determined by equilibrium condition only. The member needs to be analysed and designed for torsion. The following figure shows the situations where eccentric loads are acting on the members.
+ CGC CGC
a) Bridge deck
Figure 5-4.1
b) L-beam
2) Compatibility torsion: This is generated by twisting, to maintain compatibility in deformation with the connected member. This type of torsion generates in a primary beam supporting secondary beams.
In compatibility torsion, the torsion demand is determined by both equilibrium and compatibility conditions. Else, the torsion can be neglected. This implies that primary beam need not be analysed and designed for torsion, if the secondary beams are designed as pin supported.
In this section, the emphasis is laid on equilibrium torsion. To understand the behaviour of a beam under torsion, the presentation will be in the following sequence. 1) Stresses in an uncracked (homogenous) rectangular beam without prestressing due to pure torsion (in absence of flexure), with constant torque along the span. 2) Crack pattern under pure torsion. 3) Components of resistance for pure torsion. 4) Modes of failure under combined torsion and flexure. 5) Effect of prestressing force. Although pure torsion is absent in structures, understanding the behaviour of a beam under pure torsion helps to analyse a beam under combined torsion, flexure and shear.
the longer side. Hence, the subsequent explanation will refer to the stress condition at the middle of the longer side.
max
Figure 5-4.2
At any point in the beam, the state of two-dimensional stresses can be expressed in terms of the principal stresses. The Mohrs circle of stress is helpful to understand the state of stress.
Before cracking, the stress carried by steel is negligible. When the principal tensile stress exceeds the cracking stress, the concrete cracks and there is redistribution of stresses between concrete and steel.
For a point at the middle of the longer side (Element 1), the torsional shear stress is maximum. The principal tensile stress (1) is inclined at 45 to the beam axis.
2 2 2 1 Mohrs circle
Since the torsion is maximum at middle of the longer side, cracks due to torsion occur around that location and perpendicular to 1.
1 2
Crack inclination
Figure 5-4.4 Inclination of crack at the side of a beam
In structures, a beam is not subjected to pure torsion. Along with torsion it is also subjected to flexure and shear. Hence, the stress condition and the crack pattern are more complicated than shown before.
where the shear flow is assumed to be constant. The portion of concrete inside the shear flow zone can be neglected in calculating the capacity.
The components in vertical and horizontal sections of a beam are shown below.
Figure 5-4.6
The components can be denoted as below. Tc = torsion resisted by concrete Ts = torsion resisted by the longitudinal and transverse reinforcing bars. The magnitude and the relative value of each component change with increasing torque.
Failure surface
Figure 5-4.7
For a beam of rectangular section, the plane of failure is further influenced by warping. Torsional warping is defined as the differential axial displacement of the points in a section perpendicular to the axis, due to torque.
For a reinforced concrete beam, the length increases after cracking and after yielding of the bars. For a beam subjected to flexure and torsion simultaneously, the modes of failure are explained by the Skew Bending Theory. The observed plane of failure is not perpendicular to the beam axis, but inclined at an angle. The curved plane of failure is idealised as a planar surface inclined to the axis of the beam. The skew bending theory explains that the flexural moment (Mu) and torsional moment (Tu) combine to generate a resultant moment inclined to the axis of the beam. This moment causes compression and tension in a planar surface inclined to the axis of the beam. The following figure shows the resultant moment due to flexural moment and torsion in a beam.
Mu
Figure 5-4.8
Resultant moment
Beam subjected to flexural moment and torsion
Tu
The modes of failure are explained based on the relative magnitudes of the flexural moment (Mu) and torsional moment (Tu) at ultimate. Three discrete modes of failure are defined from a range of failure. The idealised pattern of failure with the plane of failure and the resultant compression (Cu) and tension (Tu) are shown for each mode (Courtesy: Pillai, S. U., and Menon, D., Reinforced Concrete Design). 1) Modified bending failure (Mode 1): This occurs when the effect of Mu is larger than that of Tu.
Zone under Cu
Tu
Figure 5-4.9 Idealised pattern for Mode 1 failure
2) Lateral bending failure (Mode 2): This is observed in beams with thin webs when the effect of Mu and Tu are comparable.
Zone under Cu
Tu
Figure 5-4.10 Idealised pattern for Mode 2 failure 3) Negative bending failure (Mode 3): When the effect of Tu is large and the top steel is less, this mode of failure occurs.
Tu
Zone under Cu
Figure 5-4.11 Idealised pattern for Mode 3 failure
difference in the torque versus twist curves for a non-prestressed beam and a prestressed beam.
Non-prestressed beam
The effect of prestressing force is explained for a beam under pure torsion with a concentric prestressing force (Pe). The following figure shows such a beam.
Pe T
Figure 5-4.13 Beam subjected to pure torsion and prestressing force
For a point at the middle of the longer side (Element 1), there is normal stress due to the prestressing force (fpe). The principal tensile stress (1) is inclined to the neutral axis at an angle greater than 45.
fpe
(fpe,) 2
Principal stresses
> 45o
In the following figure, the formation of cracks for a prestressed beam under pure torsion is shown. This figure can be compared with that for a reinforced concrete beam.
Figure 5-4.15
In presence of prestressing force, the cracking is at a higher torque. After cracking, the crack width of a spiral crack is low. Thus, the aggregate interlock is larger as compared to a non-prestressed beam under the same torque. Hence, the torsional strength of
concrete (Tc) increases in presence of prestressing force. This is accounted for in the expression of Tc.
General Comments Limit State of Collapse for Torsion Design of Longitudinal Reinforcement
Tu tu L
Figure 5-6.1
The torsional moment near the support is given by the following expression.
Tu =
Here, L = clear span of the beam
tu L 2
(5-5.1)
For a straight beam with a point torque, the maximum torsional moment (Tu) is near the closer support. If the location of the point torque is variable, Tu is calculated for the location closest to a support. For a curved beam, Tu is calculated based on structural analysis.
The design is done for the critical section. The critical section is defined in Clause 41.2 of IS:456 - 2000. In general cases, the face of the support is considered as the critical section. When the reaction at the support introduces compression at the end of the beam, the critical section can be selected at a distance effective depth from the face of the support. To vary the amount of reinforcement along the span, other sections may be selected for design. Usually the following scheme is selected for the stirrup spacing in beams under uniformly distributed load. 1) Close spacing for quarter of the span adjacent to the supports. 2) Wide spacing for half of the span at the middle. For large beams, more variation of spacing may be selected. The following sketch shows the typical variation of spacing of stirrups. The span is represented by L.
L/4
Figure 5-6.2
L/2
L/4
First, an equivalent flexural moment Mt is calculated from Tu. Second, for the design of primary longitudinal reinforcement, including the prestressed tendon, the total equivalent ultimate moment (Me1) is calculated from the flexural moment (Mu) and Mt. Third, the design of longitudinal reinforcement for other faces based on equivalent ultimate moments Me2 and Me3 is necessary when the equivalent moment Mt is larger than Mu. The following sketch shows the equivalent ultimate moments for design.
Me1
Figure 5-6.3
Me2
Equivalent ultimate moments
Me3
The design for Me1 is similar to the design of a prestressed section for flexure. The design for Me2 is similar to the design of a prestressed concrete or reinforced concrete section. The design for Me3 is similar to the design of a reinforced concrete section. The design of stirrups including torsion is similar to the design of stirrups in absence of torsion.
bw
t
Solid rectangular
Figure 5-6.4
Hollow rectangular
Flanged
The variables are as follows. b = breadth of the section = bw for flanged section D = total depth of the section t = thickness of the section. The average prestress in a section at the level of CGC, is limited to 0.3fck.
Mode 1 Failure
Indian Institute of Technology Madras
Zone under Cu
M e1 Tu
Figure 5-6.5
The equivalent ultimate bending moment for Mode 1 failure (Me1) is given by the following equation.
(5-5.2)
M e1 = M u + M t
The equivalent bending moment for Tu is given as follows.
2D Mt = Tu 1+ b
In the previous expression, Mu = applied bending moment at ultimate. Mt = additional equivalent bending moment for torsion. Tu = applied torsion at ultimate.
(5-5.3)
Since, the torsion generates tension in the reinforcement irrespective of the sign, the sign of Mt is same as that of Mu.
Mode 2 Failure
Zone under Cu
Figure 5-6.6
Tu
Me 3
The equivalent ultimate transverse bending moment for Mode 2 failure (Me3) is given as follows.
1+ 2b x1 D Me 3 = Mt 1+ 2D 2e 1+ b
2
(5-5.4)
In the previous expression e = Tu/Vu, ratio of ultimate torsion and ultimate shear force at a section. x1 = smaller dimension of a closed stirrup. The larger dimension of a closed stirrup is represented as y1. the dimensions are shown in the following sketch.
y1
x1
Figure 5-6.7
The transverse bending moment Me3 is considered when the numerical value of Mu is less than Mt. Me3 acts about a vertical axis.
Mode 3 Failure
Tu
Zone under Cu
Figure 5-6.8
Me2
The equivalent ultimate bending moment for Mode 3 failure (Me2) is given by the following equation. Me2 = Mt Mu The expression of Mt is same as for Mode 1 failure, given
(5-5.5)
before.
Mode 3 failure is checked when the numerical value of Mu is less than that of Mt. Me2 acts in the opposite sense of that of Mu. The longitudinal reinforcement is designed for Me1 similar to the flexural reinforcement for a prestressed beam. The design of flexural reinforcement is covered in Section 4.2, Design of Sections for Flexure (Part I) and Section 4.3 Design of Sections for Flexure (Part II). When Me2 is considered, longitudinal reinforcement is designed similar to a prestressed concrete or reinforced concrete beam. When Me3 is considered, longitudinal reinforcement is designed similar to a reinforced concrete beam. For a singly reinforced rectangular section, the amount of longitudinal reinforcement (As) is solved from the following equation.
f A 0.87fy As d 1- y s = Mu fck bd
(5-5.6)
In the previous equation, d = effective depth of longitudinal reinforcement fy = characteristic yield stress of longitudinal reinforcement fck = characteristic compressive strength of concrete Mu = one of Me2 and Me3.
Capacity of Concrete under Pure Torsion The capacity of concrete is determined based on the plastic theory for torsion. The capacity is equal to the torque generating the first torsional crack (Tcr). The For a reinforced concrete beam, Tcr is estimated by equating the maximum torsional shear stress (max) caused by Tcr to the tensile strength of concrete (0.2fck). does not plastify as assumed in the plastic theory. The estimate of the cracking torque (Tcr) for a rectangular section is given below.
Tcr 0.2 fck b 2D b 1 2 3D b Tcr = 0.1 b 2D 1 fck 3D
estimated tensile strength is less than that under direct tension because the full section
(5-6.1)
For flanged sections, the section is treated as a compound section. A compound section is a summation of rectangular sections. The cracking torque is estimated as a summation of the capacities of the individual rectangular sections. Since the interaction between the rectangular sections is
neglected in the summation, the estimate of the cracking torque is a lower bound estimate. The following flanged section is shown as a compound section of five rectangles. For an individual rectangle, the short side is b and the long side is D.
1 3
Figure 5-6.1
The estimate of the cracking torque (Tcr) for a compound section is as follows.
b Tcr = 0.1 b 2D 1 fck 3D
(5-6.2)
Here, the summation is for the individual rectangles. For a prestressed concrete beam, the strength of concrete is multiplied by the factor p, which is a function of the average effective prestress (fcp).
p = 1+
12fcp fck
(5-6.3)
The value of fcp is taken as positive (numeric value). It can be observed that the strength increases with prestress. The cracking torque (Tcr) and the capacity of concrete to resist torsion (Tc) for a prestressed concrete beam are thus estimated as follows.
Tc = Tcr b Tc = 0.15b 2D 1 p fck 3D
(5-6.4)
In the previous expression, b = breadth of the individual rectangle D = depth of the individual rectangle.
In presence of flexural shear, the torsional capacity of concrete reduces. Similarly, in presence of torsion, the flexural shear capacity of concrete reduces. This is referred to as interaction of shear and torsion. The capacity of concrete under shear is explained in Section 5.2, Design for Shear (Part I). A linear interaction of the shear and torsion capacities of concrete is considered as shown in the following figure. In the horizontal axis, the shear demand is normalised with respect to the capacity of concrete under flexural shear. In the vertical axis, the torsional demand is normalised with respect to the capacity of concrete under pure torsion.
1.0
Tu/Tc
Vu/Vc
Figure 5-6.2
1.0
(5-6.5)
This is a linear interaction equation. In the previous expression, Tu = applied torsion at ultimate Vu = applied shear at ultimate Tc = capacity of concrete under pure torsion. Vc = capacity of concrete under flexural shear. Based on the interaction equation, the reduced capacity of concrete to resist torsion (Tc1) is given below.
e Tc1 = Tc e + ec Tu / 2
(5-6.6)
Tc1 is limited to Tu/2 to restrict concrete reaching its capacity. The parameter e is the ratio of torsion and shear demands at ultimate. The parameter ec is the ratio of the corresponding concrete capacities. e = Tu/ Vu ec = Tc/ Vc The reduced capacity of concrete to resist shear is given below.
Vc1 = Vc ec e + ec
(5-6.7) (5-6.8)
(5-6.9)
The transverse reinforcement is provided in the form of closed stirrups enclosing the corner longitudinal bars. The amount (Asv) is equal to the higher value determined from two expressions. The first expression is based on the Skew Bending Theory.
Asv =
The notations are as follows.
Mt sv 1.5b1d1fy
(5-6.10)
b1 = distance between the corner longitudinal bars along the short side d1 = distance between the corner longitudinal bars along the long side. Mt = additional bending moment from torsion. sv = spacing of the stirrups fy = characteristic yield stress of the stirrups. The dimensions b1 and d1 are shown in the following sketch.
d1
b1
Figure 5-6.3
Asv = Av + 2 AT
(5-6.11)
The first component Av corresponds to the flexural shear to be carried by the stirrups. The second component At corresponds to the torsional shear to be carried by the stirrups. The factor 2 considers that the torsional shear is additive to flexural shear in both the legs. The following sketch shows the addition of flexural and torsional shears for a hollow section.
Due to Vu
Figure 5-6.4
Due to Tu
The two shears are additive in the left web, whereas they are subtractive in the right web. Since, the stirrups have equal areas in the two legs, the torsional shear is considered additive to flexural shear in both the legs. In solid sections, the two shears are not additive throughout the web. The flexural shear is distributed, whereas the torsional shear is restricted in the shear flow zone. Thus for
solid sections, the expression of Asv is conservative. The following sketch shows the addition of flexural and torsional shears for a solid section.
Due to Vu
Figure 5-6.5
Due to Tu
If the breadth of the web is large, the two shears can be designed separately. The stirrups for flexural shear can be distributed throughout the interior of the web. For torsional shear, closed stirrups can be provided in the peripheral shear flow zone. The expressions of Av and At are derived from the truss analogy for the ultimate limit state. Av = AT =
(Vu -Vc1 ) sv
0.87fy d1
(5-6.12)
(Tu -Tc1 ) sv
0.87fy b1d1
(5-6.13)
The minimum amount of transverse reinforcement is same as that for shear in absence of torsion.
Asv 0.4 = bsv 0.87fy
(5-6.14)
When any side is larger than 450 mm, provide side face reinforcement (As, sf), as per the following. Minimum amount Maximum spacing As,sf,min = 0.1% bD smax = 300 mm or b, whichever is less.
This amount is sufficient to check thermal and shrinkage cracks. 2) The closed stirrups should be bent close to the tension and compression surfaces satisfying the minimum cover. The stirrups should be perpendicular to the axis of the beam. Closed stirrups should not be made of pairs of U-stirrups lapping one another. This is clarified in the following sketch.
Incorrect detailing
Figure 5-6.6
Correct detailing
3) The maximum spacing is (x1 + y1)/4 or 200 mm, whichever is smaller. Here x1 and y1 are the short and long dimensions of the stirrups respectively. 4) Proper anchorage of stirrups as mentioned under detailing requirements of shear reinforcement. It is recommended to bend the ends of a stirrup by 135 and have 10 times the diameter of the bar (db) as extension beyond the bend. The following sketch clarifies the detailing of end hooks.
135 10db
Figure 5-6.7
5) The stirrups should be continued till a distance h + bw beyond the point at which it is no longer required. Here, h is the overall depth and bw is the breadth of the web.
2a) Calculate Me1. 2b) Design Ap and As. The design procedure involves preliminary design and final design, which are explained in the Section 4.2, Design of Sections for Flexure (Part I) and Section 4.3, Design of Sections for Flexure (Part II) 3a) Calculate Me2 if Mu < Mt. 3b) Design As. The design procedure is similar for a reinforced concrete section. If Ap is provided, the design is similar to a prestressed concrete section. 4a) Calculate Me3 if Mu < Mt. 4b) Check the adequacy of transverse bending based on the corner bars. If inadequate, design side face reinforcement (As,sf). As,sf includes the corner bars. The design is similar to that for a reinforced concrete section.
5a) Calculate Tc, Eqn. (5-6.4). 5b) Calculate Vc from the lower of Vc0 and Vcr. 5c) Calculate e (if not calculated earlier) and ec. 5d) Calculate Tc1 and Vc1. Limit Tc1 to Tu/2. 6) Calculate Asv / sv from the greater of the values given by Eqns. (5-6.10),
(5-6.11), (5-6.12), and (5-6.13).
Compare the value with the minimum requirement Eqn. (5-6.14). 7) Calculate maximum spacing and round it off to a multiple of 5 mm. 8) Calculate the size of the stirrups based on the amount required. Repeat the calculations for other locations of the beam if the spacing of stirrups needs to be varied.
Example 5-6.1
Design a rectangular section to carry the following ultimate loads. Tu = 44.5 kNm Mu = 222.5 kNm (including an estimate of self-weight) Vu = 89.0 kN. The material properties are as follows. fck = 35 N/mm2 fy = 250 N/mm2 fpk = 1720 N/mm2 The prestressing is fpe = 1035 N/mm2.
Solution
1) Calculate Me1. Let D/b = 2 Mt = Tu 1+ 2D b
= 44.5 1+ 2 2 = 99.5 kNm M e1 = Mu + MT = 222.5 + 99.5 = 322.0 kNm 2) Select section. Design Ap and As. Select b = 250 mm D = 500 mm d = 450 mm.
Provide (2) 16 mm diameter corner bars. The flexural design results are as follows. As = 2 201 = 402 mm2. Required amount of prestressing steel with dp = d = 450 mm is Ap = 484 mm2. Provide 11 mm diameter strands with area = 70 mm2. Required number of strands = 484 / 70 = 6.8 7 Provided amount of prestressing steel Ap,prov = 7 70 = 490 mm2 3) Calculate Me2 . Since Mu > Mt , design for Me2 is not required. 4) Calculate Me3 . Since Mu > Mt , design for Me3 is not required. 5a) Calculate Tc .
fcp = =
Pe A fpe Ap
p = 1+ = 1+ = 1.55
Indian Institute of Technology Madras
fpt = -
M0 = 0.8fpt
I y 2.604109 200
f V Vcr = 1- 0.55 pe c bd + M0 u fpk Mu 0.46 250 450 89 = (1- 0.550.6) +123.43 3 10 222.5 = 84.0 kN
Asv Mt = sv 1.5b1d1fy 99.5 106 = 1.5 200 400 250 = 3.3 mm2 /mm
Indian Institute of Technology Madras
T -T AT = u c1 sv 0.87fy b1d1 (44.5 - 19.26)106 = 0.87 250 200 400 = 1.45 mm2 /mm
Estimated values x1 y1 = 250 46 = 204 mm = 500 78 = 422 mm. The other values of sv do not govern. 8) Calculate the size of the stirrups Select 2 legs of 12 mm diameter stirrups. Asv = 2 113 = 226 mm2
sv =
226 3.48 = 65 mm
The spacing can be increased by bundling the stirrup bars. Designed section
250 12 mm diameter stirrups @ 65 mm c/c 500 CGC 200 (4) corner bars 16 mm diameter
Dimensions in mm.
6.1.1 Introduction
The deflection of a flexural member is calculated to satisfy a limit state of serviceability. Since a prestressed concrete member is smaller in depth than an equivalent reinforced concrete member, the deflection of a prestressed concrete member tends to be larger. The total deflection is a resultant of the upward deflection due to prestressing force and downward deflection due to the gravity loads. Only the flexural deformation is considered and any shear deformation is neglected in the calculation of deflection. Shear deformation is included in members such as deep beams and wall type of structures. The deflection of a member is calculated at least for two cases. 1) Short term deflection at transfer 2) Long term deflection under service loads The short term deflection at transfer is due to the prestressing force (before long term losses) and self-weight. The effect of creep and shrinkage of concrete are not considered. The long term deflection under service loads is due to the effective prestressing force (after long term losses) and the gravity loads. The permanent components of the gravity loads are considered in the effect of creep. These components are dead load and sustained live load.
5 wL4 384 EI
P at centre
PL3 = 48EI
Cantilever beams
=
P
wL4 8EI
PL3 = 3EI
e wup L
wup = 8Pe L2
4 5 w up L P = 384 EI
(6-1.1)
P Wup
Wup = P = 4Pe L Wup L3 48EI
(6-1.2)
P aL
Wup
Pe aL a ( 3 - 4a 2 ) Wup L3 24EI
Wup
Wup = P =
(6-1.3)
st = -P 0 + SW
Here, P0 = magnitude of deflection due to P0 sw = deflection due to self-weight P0 = prestressing force before long term losses.
(6-1.4)
The long term deflection under service loads is difficult to calculate because the prestressing force and creep strain influence each other. Creep of concrete is defined as the increase in deformation with time under constant load. Due to the creep of
concrete, the prestress in the tendon is reduced with time. The ultimate creep strain is found to be proportional to the elastic strain. The ratio of the ultimate creep strain to the elastic strain is called the creep coefficient . The values of as per IS:1343 - 1980 are given in Section 1.6, Concrete (Part II). The following expression of the long term deflection under service loads (lt) is a simplified form, where an average prestressing force is considered to generate the creep strain. The effect of shrinkage on the prestressing force is neglected in the expression.
P 0 + Pe lt = -Pe - + ( DL + SL ) (1+ ) + LL 2
(6-1.5)
The notations in the previous equations are as follows. P0 = magnitude of deflection due to P0 Pe = magnitude of deflection due to Pe Pe = effective prestressing force after long term losses. DL = deflection due to dead load (including self-weight) SL = deflection due to sustained live load LL = deflection due to additional live load A more rigorous calculation of total deflection can be done using the incremental timestep method. It is a step-by-step procedure, where the change in prestressing force
due to creep and shrinkage strains is calculated at the end of each time step. The results at the end of each time step are used for the next time step. This procedure was suggested by the Precast / Prestressed Concrete Institute (PCI) committee and is also called the General method (Reference: PCI Committee, Recommendations for Estimating Prestress Losses, PCI Journal, PCI, Vol. 20, No. 4, July-August 1975, pp. 43-75). In the PCI step-by-step procedure, a minimum of four time steps are considered in the service life of a prestressed member. The following table provides the definitions of the time steps.
Table 6-1.1
Time steps in the step-by-step procedure End Age of prestressing 30 days after prestressing or when subjected to superimposed load 1 year of service End of service life
Step 1.
Beginning Pre-tension: Anchorage of steel Post-tension: End of curing End of Step 1 End of Step 2 End of Step 3
2. 3. 4.
The step-by-step procedure can be implemented in a computer program, where the number of time steps can be increased.
the structural element and the appearance of the finishes or partitions are not adversely affected. The limits of deflection are summarised next. 1) The total deflection due to all loads, including the effects of temperature, creep and shrinkage, should not exceed span / 250. 2) The deflection after erection of partitions or application of finishes, including the effects of temperature, creep and shrinkage, should not exceed span/350 or 20 mm, whichever is less. 3) If finishes are applied, total upward deflection due to prestressing force should not exceed span / 300.
These types of members are designed to be uncracked under service loads. The gross moment of inertia (Ig) can be used to calculate the deflections.
Type 3 Members
This type of members is expected to be cracked under service loads. Strictly, the gross moment of inertia (Ig) cannot be used in the calculations. IS:1343 - 1980, Clause 22.6.2, recommends the following. 1) When the permanent load is less than or equal to 25% of the live load, the gross moment of inertia can be used. 2) When the permanent load is greater than 25% of the live load, the span-toeffective depth (L/d) ratio should be limited to bypass the calculation of deflection. If the L/d ratio of a member exceeds the limit, the gross moment of inertia can still be used if the tensile stress under service loads is within the allowable value. simpler as compared to an effective moment of inertia. This recommendation is suggested because the calculation of gross moment of inertia is
Here, L is in metres. Deflection calculations are necessary for cantilevers with L > 10 m.
6.2.1 Introduction
The crack width of a flexural member is calculated to satisfy a limit state of serviceability. Among prestressed concrete members, there is cracking under service loads only for Type 3 members. Hence the calculation of crack width is relevant only for Type 3 members. The crack width is calculated for the cracks due to bending which occur at the bottom or top surfaces of a flexural member. The flexural cracks start from the tension face and propagate perpendicular to the axis of the member. This type of cracks is illustrated in Section 5.1, Analysis for Shear. If these cracks are wide, it leads to corrosion of the reinforcing bars and prestressed tendons. Also, the cracks tend to widen under sustained load or cyclic load. To limit the crack width, Type 3 members have regular reinforcing bars in the tension zone close to the surface, in addition to the prestressed tendons. The crack width of a flexural crack depends on the following quantities. 1) Amount of prestress 2) Tensile stress in the longitudinal bars 3) Thickness of the concrete cover 4) Diameter and spacing of longitudinal bars 5) Depth of member and location of neutral axis 6) Bond strength 7) Tensile strength of concrete.
Wcr =
(6-2.1)
The notations in the previous equation are as follows. acr = shortest distance from the selected level on the surface to a longitudinal bar Cmin = minimum clear cover to the longitudinal bar h x m = total depth of the member = depth of the neutral axis = average strain at the selected level.
The values of Cmin and h are obtained from the section of the member. The evaluation of the other variables is explained. Evaluation of acr The location of crack width calculation can be at the soffit or the sides of a beam. The value of acr depends on the selected level. The following sketch shows the values of acr at a bottom corner (A), at a point in the soffit (B) and at a point at the side (C).
Neutral axis C
acr3 acr2
B
nearest longitudinal bar
acr1
A
Figure 6-2.1
Usually the crack width is calculated at a point in the soffit, which is equidistant from two longitudinal bars. This point is the location of maximum estimated crack width. The following sketch shows the variables used in computing acr.
s db Cmin
Figure 6-2.2
acr
dc
Using geometry, the value of acr is obtained from the following equation.
s 2 db acr = + dc 2 2
Here, db = diameter of longitudinal bar dc = effective cover = Cmin + db/2 s = centre-to-centre spacing of longitudinal bars.
(6-2.2)
The values of db, dc and s are obtained from the section of the member. Evaluation of x and m The values of x and m are calculated based on a sectional analysis under service loads. The sectional analysis should consider the tension carried by the uncracked concrete in between two cracks. The stiffening of a member due to the tension carried by the concrete is called the tension stiffening effect. The value of m is considered to be an average value of the strain at the selected level over the span. The following sketch illustrates the cracking and the uncracked concrete in a flexural member.
Cracked section
Figure 6-2.3
Uncracked concrete
The analysis of a Type 3 member should be based on strain compatibility of concrete and prestressing steel. IS:456 - 2000 recommends two procedures for the sectional analysis considering the tension stiffening effect. 1) Rigorous procedure with explicit calculation of tension carried by the concrete. 2) Simplified procedure based on the conventional analysis of a cracked section, neglecting the tension carried by concrete. An approximate estimate of the tension carried by the concrete is subsequently introduced. Here, the simplified procedure is explained. For a rectangular zone under tension, the simplified procedure gives the following expression of m.
m = 1 -
b ( h - x )( a - x ) 3Es As ( d - x )
(6-2.3)
For a prestressed member, (EpAp + EsAs) is substituted in place of EsAs. The second term considers the tension carried by the concrete approximately by reducing the strain (1) obtained from the analysis of a cracked section. In the above expression, a = distance from the compression face to the level at which crack width is calculated = h, when the crack width is calculated at the soffit b d = width of the rectangular zone = effective depth of the longitudinal reinforcement
As = area of non-prestressed reinforcement Ap = area of prestressing steel. Es = modulus of elasticity of non-prestressed steel Ep = modulus of elasticity of prestressed steel
1 s
= strain at the selected level based on a cracked sectional analysis = s(a x)/(d x) = strain in the longitudinal reinforcement.
The depth of neutral axis (x) can be calculated by a trial and error procedure till the equilibrium equations are satisfied. The following sketch shows the beam cross section, strain profile, stress diagram and force couples under service loads. The contribution of non-prestressed reinforcement is also included.
b dp d
fc x
0.33x
C Tp Ts
Force
Ap As
dec p
fs
Stress
fp
Cross-section
Figure 6-2.4
Strain
The expressions of the forces are as follows. C = 0.5Eccxb Tp = ApEpp Ts = AsEss Based on the principles of mechanics, the following equations are derived. 1) Equations of equilibrium The first equation states that the resultant axial force is zero. This means that the compression and the tension in the force couple balance each other. (6-2.4) (6-2.5) (6-2.6)
F = 0 Tp +Ts = C Ap E p p + As Es s = 0.5Ec c xb
couple in the force diagram. (6-2.7) The second equation relates the moment under service loads (M) with the internal
M A = Ts ( d - d p ) + C ( d p - 0.33 x )
p
= As Es s ( d - d p ) + 0.5Ec c xb ( d p - 0.33 x )
Indian Institute of Technology Madras
(6-2.8)
The value of M should be equal to the moment due to service loads. 2) Equations of compatibility The depth of the neutral axis is related to the depth of CGS and the depth of nonprestressed reinforcement by the similarity of the triangles in the strain diagram.
c x = d p c + p - dec c x = d c + s
(6-2.9)
(6-2.10)
3) Constitutive relationships Linear elastic constitutive relationships are used in the earlier expressions of C, Ts and Tp. The known variables in the analysis are: b, d, Ap, As, dec, Ec, Ep, Es, M. The unknown quantities are: x, c, p, s. The steps for solving the above equations are given below. 1) Assume c 2) Assume x. 3) Calculate p and s from Eqn. (6-2.9) and Eqn. (6-2.10), respectively. 4) Calculate C, Tp and Ts from Eqns. (6-2.4), (6-2.5), (6-2.6), respectively. 5) If Eqn. (6-2.7) is not satisfied, change x. If Tp + Ts < C, decrease x. + Ts > C, increase x. 6) Calculate M from Eqn. (6-2.8). If the value differs from the given value, change c and repeat from Step 2. If Tp
The limits of crack width are as follows. Crack width 0.2 mm for moderate and mild environments 0.1 mm for severe environment. The types of environments are explained in Table 9, Appendix A of IS:1343 - 1980.
Figure 7-1.1
(Courtesy: The Concrete Products and Construction Company, COPCO, Chennai) For a pre-tensioned member without any anchorage at the ends, the prestress is transferred by the bond between the concrete and the tendons. There are three mechanisms in the bond. 1) Adhesion between concrete and steel 2) Mechanical bond at the concrete and steel interface
3) Friction in presence of transverse compression. The mechanical bond is the primary mechanism in the bond for indented wires, twisted strands and deformed bars. The surface deformation enhances the bond. Each of the type is illustrated below.
Pitch
Elliptical indentations
Twisted strand
Deformed bar
Figure 7-1.2
The prestress is transferred over a certain length from each end of a member which is called the transmission length or transfer length (Lt). The stress in the tendon is zero at the ends of the members. It increases over the transmission length to the effective prestress (fpe) under service loads and remains practically constant beyond it. The following figure shows the variation of prestress in the tendon.
fpe
Lt
Figure 7-1.3 Variation of prestress in tendon along transmission length
Hoyer Effect After stretching the tendon, the diameter reduces from the original value due to the Poissons effect. When the prestress is transferred after the hardening of concrete, the ends of the tendon sink in concrete. The prestress at the ends of the tendon is zero. The diameter of the tendon regains its original value towards the end over the transmission length. The change of diameter from the original value (at the end) to the reduced value (after the transmission length), creates a wedge effect in concrete. This helps in the transfer of prestress from the tendon to the concrete. This is known as the Hoyer effect. The following figure shows the sequence of the development of Hoyer effect.
Original diameter
b) Casting of concrete
Sinking of tendon
Original diameter
fp0
c) Transferring of prestress
Figure 7-1.4 Hoyer effect
Since there is no anchorage device, the tendon is free of stress at the end. The concrete should be of good quality and adequate compaction for proper transfer of prestress over the transmission length.
Transmission Length There are several factors that influence the transmission length. These are as follows. 1) Type of tendon wire, strand or bar 2) Size of tendon 3) Stress in tendon 4) Surface deformations of the tendon
Indian Institute of Technology Madras
Plain, indented, twisted or deformed 5) Strength of concrete at transfer 6) Pace of cutting of tendons Abrupt flame cutting or slow release of jack 7) Presence of confining reinforcement 8) Effect of creep 9) Compaction of concrete 10) Amount of concrete cover. The transmission length needs to be calculated to check the adequacy of prestress in the tendon over the length. A section with high moment should be outside the transmission length, so that the tendon attains at least the design effective prestress (fpe) at the section. The shear capacity at the transmission length region has to be based on a reduced effective prestress. IS:1343 - 1980 recommends values of transmission length in absence of test data. These values are applicable when the concrete is well compacted, its strength is not less than 35 N/mm2 at transfer and the tendons are released gradually. recommended values of transmission length are as follows. Values of transmission length Lt = 100 Lt = 65 Lt = 30 The
Table 7-1.1
For plain and intended wires For crimped wire For strands
Here, is the nominal diameter of the wire or strand. To avoid the transmission length in the clear span of a beam, IS:1343 - 1980 recommends the following. 1) To have an overhang of a simply supported member beyond the support by a distance of at least Lt.
Lt
Figure 7-1.5
2) If the ends have fixity, then the length of fixity should be at least Lt.
Lt
Figure 7-1.6 End of a member with fixity
Development Length The development length needs to be provided at the critical section, the location of maximum moment. The length is required to develop the ultimate flexural strength of the member. The development length is the minimum length over which the stress in tendon can increase from zero to the ultimate prestress (fpu). The development length is significant to achieve ultimate capacity. If the bonding of one or more strands does not extend to the end of the member (debonded strand), the sections for checking development of ultimate strength may not be limited to the location of maximum moment. The development length (Ld) is the sum of the transmission length (Lt) and the bond length (Lb). Ld = Lt + Lb (7-1.1)
The bond length is the minimum length over which, the stress in the tendon can increase from the effective prestress (fpe) to the ultimate prestress (fpu) at the critical location. The following figure shows the variation of prestress in the tendon over the length of a simply supported beam at ultimate capacity.
fpu fpe
Lt
Figure 7-1.7 Variation of prestress in tendon at ultimate
The calculation of the bond length is based on an average design bond stress (bd). A linear variation of the prestress in the tendon along the bond length is assumed. The following sketch shows a free body diagram of a tendon along the bond length.
Lb
Figure 7-1.8 Assumed variation of prestress in tendon along the bond length
The bond length depends on the following factors. 1) Surface condition of the tendon 2) Size of tendon 3) Stress in tendon
4) Depth of concrete below tendon From equilibrium of the forces in the above figure, the expression of the bond length is derived.
Lb =
( fpu - fpe )
4 bd
(7-1.2)
Here, is the nominal diameter of the tendon. The value of the design bond stress (bd) can be obtained from IS:456 - 2000, Clause 26.2.1.1. The table is reproduced below. Design bond stress for plain bars M30 1.5 M35 1.7 M40 and above 1.9
bd (N/mm2)
End Zone Reinforcement
The prestress and the Hoyer effect cause transverse tensile stress (t). This is largest during the transfer of prestress. The following sketch shows the theoretical variation of t.
Lt
Figure 7-1.9 Transverse stress in the end zone of a pre-tensioned beam
To restrict the splitting of concrete, transverse reinforcement (in addition to the reinforcement for shear) needs to be provided at each end of a member along the
transmission length. This reinforcement is known as end zone reinforcement. The generation of the transverse tensile stress can be explained by the free body diagram of the following zone below crack, for a beam with an eccentric tendon. Tension (T), compression (C) and shear (V) are generated due to the moment acting on the horizontal plane at the level of the crack. The internal forces along the horizontal plane are shown in (a) of the following figure. The variation of moment (due to the couple of the normal forces) at horizontal plane along the depth is shown in (b).
Lt
C V
Lt
Free body diagram of zone below crack
Figure 7-1.10
The end zone reinforcement is provided to carry the tension (T) which is generated due to the moment (M). The value of M is calculated for the horizontal plane at the level of CGC due to the compressive stress block from the normal stresses in a vertical plane above CGC. The minimum amount of end zone reinforcement (Ast) is given in terms of the moment (M) as follows.
Ast = 2.5M fs h
(7-1.3)
In the previous equation, h = total depth of the section M = moment at the horizontal plane at the level of CGC due to the compressive stress block above CGC fs = allowable stress in end zone reinforcement. The lever arm for the internal moment is h/2.5. The value of fs is selected based on a maximum strain. The end zone reinforcement should be provided in the form of closed stirrups enclosing all the tendons, to confine the concrete. The first stirrup should be placed as close as possible to the end face, satisfying the cover requirements. reinforcement can be distributed in the remaining Lt. References: 1) Krishnamurthy, D. A Method of Determining the Tensile Stresses in the End Zones of Pre-tensioned Beams, Indian Concrete Journal, Vol. 45, No. 7, July 1971, pp. 286297. 2) Krishnamurthy, D. Design of End Zone Reinforcement to Control Horizontal Cracking in Pre-tensioned Concrete Members at Transfer, Indian Concrete Journal, Vol. 47, No. 9, September 1973, pp. 346-349. About half the reinforcement can be provided within a length equal to Lt from the end. The rest of the
Example 7-1.1
Design the end zone reinforcement for the pre-tensioned beam shown in the following figure.
The sectional properties of the beam are as follows. A = 46,400 mm2 I = 8.47 108 mm4 Z = 4.23 105 mm3
There are 8 prestressing wires of 5 mm diameter. Ap = 8 19.6 = 157 mm2 The initial prestressing is as follows. fp0 = 1280 N/mm2. Limit the stress in end zone reinforcement (fs) to 140 N/mm2.
Cross-section at end
Solution
1) Determination of stress block above CGC Initial prestressing force P0 = Ap.fpo = 157 1280 N = 201 kN
ft = -
Stress at bottom
fb = -
80 CGC CGS
140
C1 C2 C3 y3 y2 y1
2) Determination of components of compression block C1 y1 C2 y2 C3 y3 = 1.29 200 60 = 7.74 kN = 140 + 60 = 160 mm = 1.29 140 80 = 7.22 kN = 140 = 93.3 mm = 4.3 140 80 = 24.08 kN = 140 = 46.7 mm
Indian Institute of Technology Madras
8.60
3) Determination of moment M = Ci.yi = C1.y1 + C2.y2 + C3.y3 = (7.74 160) + (7.22 93.3) + (24.08 46.7) = 3036.6 kN-mm 4) Determination of amount of end zone reinforcement
Ast = =
2.5M fs h
Post-tensioned Members
yp0
y0 = larger
Bearing plate
y0
Figure 7-2.1
corresponding dimension of the bearing plate is represented as yp0. For analysis, the
Indian Institute of Technology Madras
end zone is divided into a local zone and a general zone as shown in the following sketch.
Local zone
General zone
Bearing plate
Figure 7-2.2
y0
The local zone is the region behind the bearing plate and is subjected to high bearing stress and internal stresses. The behaviour of the local zone is influenced by the anchorage device and the additional confining spiral reinforcement. The general zone is the end zone region which is subjected to spalling of concrete. The zone is strengthened by end zone reinforcement. The variation of the transverse stress (t) at the CGC along the length of the end zone is shown in the next figure. The stress is compressive for a distance 0.1y0 from the end. Beyond that it is tensile. The tensile stress increases and then drops down to zero within a distance y0 from the end.
0.1y0
0.9y0
Fbst
Figure 7-2.3
The transverse tensile stress is known as splitting tensile stress. The resultant of the tensile stress in a transverse direction is known as the bursting force (Fbst). Compared
to pre-tensioned members, the transverse tensile stress in post-tensioned members is much higher. Besides the bursting force there is spalling forces in the general zone.
Spalling force
Figure 7-2.4
Bursting force
IS:1343 - 1980, Clause 18.6.2.2, provides an expression of the bursting force (Fbst) for an individual square end zone loaded by a symmetrically placed square bearing plate.
y Fbst = Pk 0.32 - 0.3 p 0 y0
(7-2.1)
Here, Pk = prestress in the tendon yp0 = length of a side of bearing plate y0 = transverse dimension of the end zone. The following sketch shows the variation of the bursting force with the parameter yp0 / y0. The parameter represents the fraction of the transverse dimension covered by the bearing plate.
Fbst Pk
0.32
0.02 1
Figure 7-2.5
y p0 y0
It can be observed that with the increase in size of the bearing plate the bursting force (Fbst) reduces. The following sketch explains the relative size of the bearing plate with respect to the end zone.
(1)
Figure 7-2.6
(2)
(3)
End views of end zones with varying size of the bearing plate
In the above end views of end zones, the bursting force (Fbst) will be largest for Case (1) and least for Case (3). For a rectangular end zone, Fbst is calculated from the previous equation for each principle direction. For a circular bearing plate, an equivalent square loaded area is considered in the calculation of Fbst. For more than one bearing plate, the end zone is divided into symmetrically loaded prisms. Each prism is analysed by the previous equation.
End Zone Reinforcement Transverse reinforcement is provided in each principle direction based on the value of Fbst. This reinforcement is called end zone reinforcement or anchorage zone reinforcement or bursting links. The reinforcement is distributed within a length from 0.1y0 to y0 from an end of the member. The amount of end zone reinforcement in each direction (Ast) can be calculated from the following equation. Ast = Fbst fs
(7-2.2)
The stress in the transverse reinforcement (fs) is limited to 0.87fy. When the cover is less than 50 mm, fs is limited to a value corresponding to a strain of 0.001. The end zone reinforcement is provided in several forms, some of which are proprietary
of the construction firms. The forms are closed stirrups, mats or links with loops. A few types of end zone reinforcement is shown in the following sketches.
Mat
Figure 7-2.7
Links
Types of end zone reinforcement
The local zone is further strengthened by confining the concrete with spiral reinforcement. The performance of the reinforcement is determined by testing end block specimens. The following photo shows the spiral reinforcement around the guide of the tendons.
Figure 7-2.8
(Reference: Dywidag Systems International) The end zone may be made of high strength concrete. The use of dispersed steel fibres in the concrete (fibre reinforced concrete) reduces the cracking due to the bursting force. Proper compaction of concrete is required at the end zone. Any honey-comb of the concrete leads to settlement of the anchorage device. If the concrete in the end zone is different from the rest of the member, then the end zone is cast separately.
Bearing Plate High bearing stress is generated in the local zone behind the bearing plate. The bearing stress (fbr) is calculated as follows.
fbr =
Here, Pk
Pk Apun
(7-2.3)
= prestress in the tendon with one bearing plate = Area of contact of bearing plate.
As per Clause 18.6.2.1, IS:1343 - 1980, the bearing stress in the local zone should be limited to the following allowable bearing stress (fbr,all).
Abr Apun
(7-2.4)
= Maximum transverse area of end block that is geometrically similar and concentric with punching area fci = cube strength at transfer.
The expression of allowable bearing stress takes advantage of the dispersion of the bearing stress in the concrete. The following sketch illustrates the dispersion of bearing stress in concrete.
Apun Abr
Apun
Abr
End view showing bearing plate
Figure 7-2.9
The performance of anchorage blocks and end zone reinforcement is critical during the post-tensioning operation. The performance can be evaluated by testing end block specimens under compression. The strength of an end block specimen should exceed the design strength of the prestressing tendons. The following photos show the manufacturing of an end block specimen.
Example 7-2.1
Design the bearing plate and the end zone reinforcement for the following bonded post-tensioned beam. The strength of concrete at transfer is 50 N/mm2.
A prestressing force of 1055 kN is applied by a single tendon. eccentricity of the tendon at the ends.
There is no
400 100
400
400
600
Solution
1) Let the bearing plate be 200 mm 300 mm. The bearing stress is calculated below.
fbr = = Pk Apun
fbr,all = 0.48fci
= 0.48 50 = 48 N/mm2
Limit fbr,all to 0.8 fci = 0.8 50 = 40 N/mm2. Bearing stress is less than fbr,all. Hence OK. 2) Calculate bursting force. In the vertical direction
179.3103 = 0.87 250 = 824.6 mm2 Provide Ast = 824.6 = 550 mm2 within 0.1 y0 = 60 mm and 0.5 y0 = 300 mm from the end. Select (6) 2 legged 8 mm diameter stirrups. Provide Ast = 824.6 = 275 mm2 within 0.5 y0 = 300 mm and y0 = 600 mm from the end. Select (5) 2 legged 6 mm diameter stirrups.
200
300
End view
Introduction Prestressed cantilever beams are present in buildings and bridges. Usually, the
cantilever is provided with a back span (anchor span) to reduce the torsion in the supporting member. In a building, the cantilever can be an extension of a continuous beam. In a bridge, the cantilever is a part of the balanced cantilever girder. The construction can be in-situ, where the concrete is cast in formwork that is temporarily supported. Else, a segmental construction can be adopted, wherein slices of the girder are placed and temporarily prestressed.
The following figure illustrates the cantilevers in buildings and bridges schematically.
Segment
Back span
Cantilever span
Cantilever spans
(b) Cantilevers in a bridge girder
Figure 8-1.2
The analysis of a section of a prestressed cantilever for flexural stresses is similar to that for a simply supported beam. The difference is that for gravity loads, the bending moment in cantilever is negative, that is compression is generated at the bottom. Thus, the CGS is placed above the CGC and the eccentricity e is negative.
The following aspects need to be considered in the analysis and design of a prestressed cantilever beam. 1) Certain portions of the back span are subjected to both positive and negative moments. Hence, there will be two design moments at service loads. 2) The beam may be subjected to partial loading and point loading. 3) The sequence of loading is important to design the prestressing force. 4) High values of moment and shear occur simultaneously near the support.
8.1.1 Analysis
The analysis of a cantilever beam with a back span, is illustrated to highlight the aspects stated earlier. The bending moment diagrams for the following load cases are shown schematically in the following figure. 1) Dead load (DL) 2) Live load (LL) only on the back span 3) Live load only on the cantilever span 4) Dead load and live load along full length.
wDL
wLL
wLL
wDL + wLL
The envelop moment diagrams are calculated from the analysis of each load case and their combinations. In the following envelop moment diagrams, Mmax and Mmin represent the highest and lowest values (algebraic values with sign) of the moments at a section, respectively. Note that certain portions of the beam are subjected to both positive and negative moments.
Mmax
Mmin
Figure 8-1.4
For moving point loads as in bridges, first the influence line diagram is drawn. The influence line diagram shows the variation of the moment or shear for a particular location in the girder, due to the variation of the position of a unit point load. The vehicle load is placed based on the influence line diagram to get the worst effect.
The limiting zone was explained for a simply supported beam in Section 4.4, Design of Sections for Flexure (Part III). Here the concept and the equations are first reviewed for a simply supported beam with positive moment.
For full prestressed members (Type 1), tension is not allowed under service conditions. If tension is also not allowed at transfer, the compression in concrete (C) always lies within the kern zone. The limiting zone is defined as the zone for placing the CGS of
the tendons such that C always lies within the kern zone.
compressive stresses at transfer and service should be within the allowable values. For limited prestressed members (Type 2 and Type 3), tension is allowed at transfer and under service conditions. The limiting zone is defined as the zone for placing the CGS such that the tensile stresses in the extreme edges are within the allowable values. Also, the maximum compressive stresses at transfer and service should be within the allowable values.
The limiting zone is determined from the maximum or minimum eccentricities of the CGS along the beam corresponding to the extreme positions of C. Note that, the
limiting zone is related with the CGS of the tendons. Individual tendons may lie outside the limiting zone.
For a simply supported beam, the maximum eccentricity (emax) at any section corresponds to the lowest possible location of C at transfer, that generates allowable tensile stress at the top of the section. The maximum compressive stress at the bottom should be within the allowable value. The minimum eccentricity (emin) at any section corresponds to the highest possible location of C at service, that generates allowable tensile stress at the bottom of the section. The maximum compressive stress at the top should be within the allowable value.
The values of emax and emin can be determined by equating the stresses at the edges of concrete with the allowable values. Else, explicit expressions of emax and emin can be developed. The following material gives the expressions of emax and emin for Type 1 and Type 2 sections. The values of emax and emin can be determined at regular intervals along the length of the beam. The zone between the loci of emax and emin is the limiting zone of the section for placing the CGS. The equations are summarised
Type 1 Section Based on the stress at the top at transfer, the following expression of emax is derived.
Msw P0 Msw + kb P0
(8-1.1)
Also, the stress at the bottom should be less than the allowable value at transfer. Based on the stress at the bottom at service, the following expression of emin is derived.
MT Pe MT - kt Pe
(8-1.2)
Also, the stress at the top should be less than the allowable value at service. If for a particular section emin is negative, it implies that the CGS can be placed above the CGC. This happens near the supports.
Type 2 Section The corresponding equations for a Type 2 section are given below.
(8-1.3)
In a simply supported beam, the external moments are always positive. The minimum moment is due to self weight. The maximum moment is under service loads. For
cantilever beams, the minimum external moment need not be at transfer, when the moment is due to self weight (MSW). Also, under service loads there are two moments Mmin and Mmax at a location, obtained from the envelop moment diagrams. The maximum and minimum eccentricities emax and emin at a particular location are first determined for service loads from Mmin and Mmax , respectively, at that location and the effective prestress Pe. Next, another set of emax and emin are calculated at transfer from MSW and the prestress P0. The final emax is the lower of the two values calculated at service and at transfer. Similarly, the final emin is the higher of the two values calculated at service and at transfer.
The expressions of emax and emin for the simply supported beam were developed for positive moments. For a cantilever, corresponding to a negative moment, the The expressions for a eccentricity implies that the CGS is located above CGC. cantilever are given below.
emax = emin =
At Transfer
Mmin + kb Pe Mmax - kt Pe
(8-1.5)
(8-1.6)
emax = emin =
Msw + kb P0 Msw - kt P0
(8-1.7)
(8-1.8)
emax = emin =
At Transfer
emax = emin =
(8-1.9)
(8-1.10)
(8-1.11)
(8-1.12)
kt
Limiting zone
kb
Figure 8-1.5 Limiting zone for a non-prismatic beam
For a prismatic beam with uniform cross section along the length, the tendon profile is similar to the moment diagram under uniform load. Thus for regions of negative
moment, the CGS is located above the CGC. Since there cannot be a sharp kink in the tendons and, the supports are not true point supports, the profile is shown curved at the right support.
For a beam with varying depth, the tendon profile can be adjusted (within emax and emin) to be straight for convenience of layout of the tendons.
Analysis Incorporation of Moment due to Reactions Pressure Line due to Prestressing Force
Introduction Beams are made continuous over the supports to increase structural integrity. A
continuous beam provides an alternate load path in the case of failure at a section. In regions with high seismic risk, continuous beams and frames are preferred in buildings and bridges. A continuous beam is a statically indeterminate structure.
The advantages of a continuous beam as compared to a simply supported beam are as follows. 1) For the same span and section, vertical load capacity is more. 2) Mid span deflection is less. 3) The depth at a section can be less than a simply supported beam for the same span. Else, for the same depth the span can be more than a simply supported beam. The continuous beam is economical in material. 4) There is redundancy in load path. Possibility of formation of hinges in case of an extreme event. 5) Requires less number of anchorages of tendons. 6) For bridges, the number of deck joints and bearings are reduced. Reduced maintenance
There are of course several disadvantages of a continuous beam as compared to a simply supported beam. 1) Difficult analysis and design procedures. 2) Difficulties in construction, especially for precast members. 3) Increased frictional loss due to changes of curvature in the tendon profile. 4) Increased shortening of beam, leading to lateral force on the supporting columns. 5) Secondary stresses develop due to time dependent effects like creep and shrinkage, settlement of support and variation of temperature.
6) The concurrence of maximum moment and shear near the supports needs proper detailing of reinforcement. 7) Reversal of moments due to seismic force requires proper analysis and design.
End span
Intermediate span
End span
Intermediate span
(b) Continuous beam in a bridge Figure 8-2.1 Continuous beams in buildings and bridges
8.2.1 Analysis
The analysis of continuous beams is based on elastic theory. This is covered in text books of structural analysis. important. 1) Certain portions of a span are subjected to both positive and negative moments. These moments are obtained from the envelop moment diagram. 2) The beam may be subjected to partial loading and point loading. The envelop moment diagrams are developed from pattern loading. The pattern loading refers to the placement of live loads in patches only at the locations with positive or negative values of the influence line diagram for a moment at a particular location. 3) For continuous beams, prestressing generates reactions at the supports. These reactions cause additional moments along the length of a beam. For prestressed beams the following aspects are
The analysis of a continuous beam is illustrated to highlight the aspects stated earlier. The bending moment diagrams for the following load cases are shown schematically in the following figures. 1) Dead load (DL) 2) Live load (LL) on every span 3) Live load on a single span.
wDL
wLL
wLL
For moving point loads as in bridges, first the influence line diagram is drawn. The influence line diagram shows the variation of the moment or shear for a particular location in the girder, due to the variation of the position of a unit point load. The vehicle load is placed based on the influence line diagram to get the worst effect. An influence line diagram is obtained by the Mller-Breslau Principle. This is covered in text books of structural analysis.
IS:456 - 2000, Clause 22.4.1, recommends the placement of live load as follows. 1) LL in all the spans. 2) LL in adjacent spans of a support for the support moment. The effect of LL in the alternate spans beyond is neglected. 3) LL in a span and in the alternate spans for the span moment.
The envelop moment diagrams are calculated from the analysis of each load case and their combinations. The analysis can be done by moment distribution method or by computer analysis.
In lieu of the analyses, the moment coefficients in Table 12 of IS:456 - 2000 can be used under conditions of uniform cross-section of the beams in the several spans, uniform loads and similar lengths of span.
The envelop moment diagrams provide the value of a moment due to the external loads. It is to be noted that the effect of prestressing force is not included in the envelop moment diagrams. The following figure shows typical envelop moment diagrams for a continuous beam.
Mmin Mmax
Figure 8-2.3 Envelop moment diagrams for DL + LL
In the above diagrams, Mmax and Mmin represent the highest and lowest values (algebraic values with sign) of the moments at a section, respectively. Note that certain portions of the beam are subjected to both positive and negative moments. The
moment from the envelop moment diagrams will be represented as the M0 diagram. This diagram does not depend on whether the beam is prestressed or not.
The moment diagram due to the eccentricity of the prestressing force and neglecting the intermediate support is denoted as the M1 diagram. This diagram is obtained as M1 = Pe, where, P is the prestressing force (P0 at transfer and Pe at service) and e is the eccentricity of the CGS with respect to CGC. Neglecting the variation of P along the length due to frictional losses, the value of M1 is proportional to e. Hence, the shape of the M1 diagram is similar to the cable profile.
e l
Profile of the CGS
Pe M1 diagram
3wl / 8 10wl / 8
3wl / 8
w
Simplified free body diagram
wl2/2 = 4Pe
+
5wl2/8 = 5Pe
=
Pe M2 diagram
Figure 8-2.4 Moment diagram due to prestressing force for a two-span beam
Next, the moment diagram due to the prestressing force and including the effect of the intermediate support is denoted as the M2 diagram. This is obtained by structural analysis of the continuous beam subjected to the upward thrust. Since the profile of the tendon is parabolic in each span, the upward thrust is uniform and is given as wup = w = 8Pe/l2. The downward thrust at the location of the central kink is not considered as it directly goes to the intermediate support. The hold down force at the intermediate support neglecting the downward thrust is 10wupl/8 = 10Pe/l. The downward forces at the ends are from the anchorages. The moment diagram due to wup alone (without the
Indian Institute of Technology Madras
support) is added to that due to the hold down force. The resultant M2 diagram is similar to the previous M1 diagram, but shifted linearly from an end support to the intermediate support.
For a general case, the resultant moment (M2) at a location due to the prestressing force can be written as follows. M2 = M1 + M1/ In the above equation, M1 = moment due to the eccentricity of the prestressing force neglecting the intermediate supports = Pee. M 1/ Pe e = moment due to the reactions at intermediate supports. = effective prestress = eccentricity of CGS with respect to CGC. (8-2.1)
M1 is the primary moment and M1/ is the secondary moment. The moment due to the external loads (M0) that is obtained from the envelop moment diagrams is added to M2 to get the resultant moment (M3) at a location. M3 = M2 + M0 M3 = M1 + M1/ + M0 (8-2.2)
The variation of M3 along the length of the beam (M3 diagram) can be calculated as follows. 1) The M0 diagram is available from the envelop moment diagram. 2) Plot M1 diagram which is similar to the profile of the CGS. The variation of Pe along the length due to friction may be neglected. 3) Plot the shear force (V) diagram corresponding to the M1 diagram from the relationship V = dM1/dx. 4) Plot the equivalent load (weq) diagram corresponding to the V diagram from the relationship weq = dV/dx. Note, over the supports weq can be downwards. Also, a singular moment needs to be included at an end when the eccentricity of the CGS is not zero at the end.
5) Calculate the values of M2 for the continuous beam (with the intermediate supports) subjected to weq using a method of elastic analysis (for example, moment distribution or computer analysis). Plot the M2 diagram. 6) The M3 diagram can be calculated by adding the values of M2 and M0 diagrams along the length of the beam.
The following figures explain the steps of developing the M2 diagram for a given profile of the CGS and a value of Pe.
The important characteristics of the diagrams are as follows. 1) A positive eccentricity of the CGS creates a negative moment (M1) and an upward thrust. 2) The M2 diagram has a similar shape to the M1 diagram, which is again similar to the profile of the CGS. This is because the moment generated due to the reactions (M1/) is linear between the supports.
The calculation of pressure line from the M2 diagram is based on the following expression. length. ec= M2/Pe Here, ec = distance of the pressure line from the CGC at a location. A positive value of ec corresponds to a hogging value of M2 and implies that the pressure line is beneath the CGC. (8-2.3) The pressure line can be plotted for the different values of M2 along the
The following sketch shows the pressure line for a given profile of the CGS.
Pressure line
ec
CGC
The important characteristics of the pressure line are as follows. 1) The shift of the pressure line from the profile of the CGS is a linear transformation. It is because M2 diagram has a similar shape to the profile of the CGS.
Indian Institute of Technology Madras
The pressure line will have the same intrinsic shape as the profile of the CGS. 2) Since M2 is proportional to the prestressing force, the eccentricity of the pressure line (ec) remains constant even when the prestressing force drops from the initial value P0 to the effective value Pe. The location of the pressure line for a given profile of the CGS is fixed, irrespective of the drop in the prestressing force.
Example 8-2.1
The profile of the CGS for a post-tensioned beam is shown in the sketch. Plot the pressure line due to a prestressing force Pe = 1112 kN.
0.06
0.24
CGC
9m
6m
7.5 m
7.5 m
Solution
1) Plot M1 diagram The values of M1 are calculated from M1 = Pee. e (m) 0.06 0.24 0.12 0.27 M1 (kN m) 66.7 266.9 133.4 300.2
0.06
0.24
0.12
0.27
66.7
266.9
300.2
M1 diagram (kN m)
2) Plot V diagram For AD,
V=
For DB,
V= dM1 dx 133.4 - (-266.9) = 6 = 66.7 kN
For BC, to find dM1/dx, an approximate parabolic equation for the M1 diagram can be used.
M1 = -
L Pee M1 x
At B,
V= dM1 dx x=0 4P e =- e L 4 (133.4 + 300.2) =15 = -115.6 kN
V = - 107.0 kN
The difference of V between C and B is given from the change in slope of the M1 diagram. V|C - V|B = 0.176 1112 = 195.7 kN
266.9
300.2
M1 diagram (kN m)
22.2 66.7 89.0
Since B is a reaction point, the downward load at B need not be considered. Distributed load within B and C
w BC =
22.2
66.7
89.0
66.7 kN m
13.0 kN/m
244 244
Bal 194.7 CO Bal Total 66.7 97 122 38.5 38.5 327.5 327.5
In the previous table, Bal CO DF = Balanced = Carry Over moment = Distribution Factor
The moment at the spans can be determined from statics. But this is not necessary as will be evident later.
66.7 kN m
13.0 kN/m
M2 diagram (kN m)
5) Calculate values of ec at support. The values of ec are calculated from ec= M2/Pe.
The deviations of the pressure line from the CGS at the spans can be calculated by linear interpolation.
0.136 0.06
0.184
Profile of CGS
Principle of Linear Transformation Concordant Tendon Profile Tendon Profiles Partially Continuous Beams Analysis for Ultimate Strength Moment Redistribution
Introduction Before the discussion on the tendon profile (profile of the CGS), the following concepts are introduced. 1) Principle of linear transformation 2) Concordant tendon profile.
Linear transformation cannot involve the movement of the CGS at the ends of a beam or at the support of a cantilever.
Theorem In a continuous beam, a profile of the CGS can be linearly transformed without changing the position of the resultant pressure line. This theorem can be proved based on the requirement that the curvature of the profile of the CGS remains constant under linear transformation. The following sketch explains that the pressure line remains constant for linearly transformed profiles of the CGS.
Figure 8-3.3
Concordant profile
The advantage of a concordant cable profile is that the calculations become simpler.
1) There is no secondary moment in the spans due to the prestress. diagram coincides with the M1 diagram.
The M2
2) The pressure line due to the prestress coincides with the cable profile. The shift of the pressure line due to external loads can be measured from the profile directly. A concordant profile can be developed from the moment diagram due to external loads for a certain load combination using the following theorem.
Theorem Every real moment diagram for a continuous beam on non-settling supports produced by any combination of external loads, whether transverse loads or moments, plotted to any scale, is one location for a concordant tendon in that beam. The theorem can be proved based on the condition of no deflection at the supports due to external loads. Also, for a concordant profile since there is no reaction at any support, there is no possibility of deflections at the supports. Thus, it is easy to obtain a concordant profile from the moment diagram of the external loads for a certain load combination, drawn to a certain scale. The following figure shows the steps of the development of concordant profile from the moment diagram.
CGC
Discussion The computation of the concordant profile helps in the layout of the tendon profile. The tendon profile need not be designed to be a concordant profile. It should be such that the stresses in concrete at transfer and at service are within the allowable values. If a concordant profile is selected then the calculations become simpler.
4) Considering fpe = 0.7fpk , calculate area of prestressing steel Ap = Pe / fpe. 5) Check the area of the cross-section (A) based on A = Pe /(0.5fcc,all). 6) Calculate the kern distances kb and kt, and the maximum and minimum eccentricities (emax and emin) along the length. The zone between emax and emin along the length of the beam is the limiting zone. The equations of emax and emin are same as that for a simply supported beam. The value of P0 can be estimated from Pi as follows. a. 90% of the initial applied prestress (Pi) for pre-tensioned members.
b. Equal to Pi for post-tensioned members. The value of Pi can be estimated as follows. Pi = Ap (0.8fpk) Ap = Pe / 0.7fpk 7) Select a trial profile of the CGS within the limiting zone. the CGS. Calculate the shift in the pressure line due to external loads. For a Type 1 member, if the final pressure line lies within the kern zone, then the solution is acceptable. If final pressure line lies outside the kern zone, try another profile. For Type 2 and Type 3 members, if the final pressure line lies within a zone such that the stresses at the edges are within the allowable values, then the solution is acceptable. If final pressure line lies outside the zone, try another profile. 8) Linearly transform the profile of the CGS to satisfy the cover requirements and the convenience of prestressing. For a prismatic beam with uniform cross section along the length, the tendon profile can be selected similar to the moment diagram under uniform load. Since there cannot be a sharp kink in the tendons and the supports are not true point supports, the profile needs to be curved at an intermediate support. For a beam with varying depth, the tendon profile can be adjusted (within the limiting zone) to be relatively straight for convenience of layout of the tendons and reduction of losses due to friction. The tendons can be of segments of single curvature to reduce frictional losses. The following sketches show the profiles of the CGS for common continuous beams. (8-3.3) (8-3.4) If the profile is a
concordant profile, the pressure line due to prestress coincides with the profile of
c) Combination of a) and b)
introduced by additional prestressing tendons or coupling the existing tendons. Continuity can also be introduced in a composite construction, where non prestressed continuity reinforcement is introduced in the cast-in-place topping slab. A few examples are given in the following sketches. Other innovative schemes are also used.
Additional tendon
Coupler
Jack
xu M + 0.5 d 100
(8-3.5)
Here, xu = depth of neutral axis d = the effective depth M = the percentage reduction in moment.
Example 8-3.1
The prestressed concrete beam shown in the figure, is fixed at the left end and roller supported at the right. It is post-tensioned with a single tendon with a parabolic profile, with indicated eccentricities.
a). Locate the pressure line due to application of a prestress force of 1068 kN. b). Find the primary, secondary and total moments due to prestressing force at the face of the fixed support. c). What is the magnitude and direction of the reaction produced at the roller by prestressing force? d). What minor adjustment can be made in the tendon profile to produce a concordant profile?
250
150
300 300
6m
6m
Solution
a) Locate the pressure line. 1) Plot M1 diagram. The values of M1 are calculated from M1 = Pee. M1 (kN m) 267.0 160.2 0.0
250
150
267.0 160.2
M1 diagram (kN m)
2) Plot V diagram. The M1 diagram is made up of two parabolic segments.
267.0 160.2
M1 diagram (kN m)
For each segment,
M1 = 4Pex (L - x ) L2
V=
dM1 dx 4Pe = - 2 ( L - 2x ) L
L Pe M1
V
x=0
=-
4Pe
x =L
4Pe
4160.2 12 = 53.4 kN
V diagram (kN)
3) Plot equivalent load (weq) diagram.
w eq = dV dx 53.4 +142.4 = 12 = 16.3 kN/m
16.3 kN/m
FEM
16.3122 12 =195.8
195.8
195.8
FEM = Fixed End Moment. The moment at the span can be determined from statics. But this is not necessary as will be evident later.
16.3 kN/m
293.7
M2 diagram (kN m)
5) Calculate values of ec at support. The values of ec are calculated from ec= M2/Pe. M2 (kN m) 293.7 ec (m) 0.275
The deviations of the pressure line from the CGS at the span can be calculated by linear interpolation.
b) Calculation of primary, secondary and total moments. M1 M2 = 267.0 kN m = 293.7 kN m M1/ = M2 M1 = 293.7 267.0 = 26.7 kN m c) Calculation of reaction.
R1 = w eq L
primary total
secondary
R1 weq M2 R2
R1
2 16.3 12 = 2 = 97.6 kN
R2
R2 =
M2 L 293.7 = 12 = 24.5 kN
R1 R2 = 73.1 kN Resultant reaction at roller is downwards. d) The tendon can be shifted to coincide with the pressure line to get a concordant profile.
Values in mm.
Introduction Analysis of Composite Sections Design of Composite Sections Analysis for Horizontal Shear Transfer
9.1.1 Introduction
A composite section in context of prestressed concrete members refers to a section with a precast member and cast-in-place (CIP) concrete. There can be several types of innovative composite sections. A few types are sketched below.
T T T T T T
Box section
Figure 9-1.1
Composite beam-slab
T-section
The following photos show the reinforcement for the slab of a box girder bridge deck with precast webs and bottom flange. The slab of the top flange is cast on a stay-in formwork. The reinforcement of the slab is required for the transverse bending of the slab. The reinforcement at the top of the web is required for the horizontal shear
transfer.
Figure 9-1.2 Box girder bridge deck with precast webs and bottom flange and CIP slab; Top: Aerial view, Bottom: Close-up view (Courtesy: Jos Turmo)
The advantages of composite construction are as follows. 1) Savings in form work 2) Fast-track construction 3) Easy to connect the members and achieve continuity.
The prestressing of composite sections can be done in stages. The precast member can be first pre-tensioned or post-tensioned at the casting site. After the cast-in-place (cast-in-situ) concrete achieves strength, the section is further post-tensioned. The grades of concrete for the precast member and the cast-in-place portion may be different. section. In such a case, a transformed section is used to analyse the composite
The following diagrams are for a composite section with precast web and cast-in-place flange. The web is prestressed before the flange is cast. At transfer and after casting of the flange (before the section behaves like a composite section), the following are the stress profiles for the precast web.
Section
Figure 9-1.3
Here, P0 = Prestress at transfer after short term losses Pe = Effective prestress during casting of flange after long term losses MSW = Moment due to self weight of the precast web MCIP = Moment due to weight of the CIP flange. At transfer, the loads acting on the precast web are P0 and MSW. By the time the flange is cast, the prestress reduces to Pe due to long term losses. In addition to Pe and MSW,
Indian Institute of Technology Madras
the web also carries MCIP. The width of the flange is calculated based on the concept of effective flange width as per Clause 23.1.2, IS:456 - 2000.
At service (after the section behaves like a composite section) the following are the stress profiles for the full depth of the composite section.
Unpropped Propped
Section Pe + MSW
MLL
Figure 9-1.4
Here, MLL is the moment due to live load. If the precast web is unpropped during casting of the flange, the section does not behave like a composite section to carry the prestress and self weight. Hence, the stress profile due to Pe + MSW + MCIP is
terminated at the top of the precast web. If the precast web is propped during casting and hardening of the flange, the section behaves like a composite section to carry the prestress and self weight after the props are removed. The stress profile is extended up to the top of the flange. When the member is placed in service, the full section carries MLL. From the analyses at transfer and under service loads, the stresses at the extreme fibres of the section for the various stages of loading are evaluated. These stresses are compared with the respective allowable stresses.
f =-
P0 P0ec MSW c A I I
(9-1.1)
P P ec (M + MCIP )c f = - e e SW A I I
(9-1.2)
f =-
(9-1.3a)
f =Here,
(9-1.3b)
A = area of the precast web c = distance of edge from CGC of precast web c/ = distance of edge from CGC of composite section e = eccentricity of CGS I = moment of inertia of the precast web I / = moment of inertia of the composite section.
From the analysis for ultimate strength, the ultimate moment capacity is calculated. This is compared with the demand under factored loads. The analysis at ultimate is simplified by the following assumptions. 1) The small strain discontinuity at the interface of the precast and CIP portions is ignored. 2) The stress discontinuity at the interface is also ignored. 3) If the CIP portion is of low grade concrete, the weaker CIP concrete is used for calculating the stress block. The strain and stress diagrams and the force couples at ultimate are shown below.
p
pu Stress
Tuw
Tuf
Force
Figure 9-1.5
bf bw Df d Ap
p
= breadth of the flange = breadth of the web = depth of the flange = depth of the centroid of prestressing steel (CGS) = area of the prestressing steel = strain difference for the prestressing steel = depth of the neutral axis at ultimate = strain in prestressing steel at the level of CGS at ultimate = stress in prestressing steel at ultimate = characteristic compressive strength of the weaker concrete = resultant compression in the web (includes portion of flange above precast web)
= resultant compression in the outstanding portion of flange = portion of tension in steel balancing Cuw. = portion of tension balancing Cuf.
Apf Apw
= part of Ap that balances compression in the outstanding flanges = part of Ap that balances compression in the web
The equilibrium equations are given below. These equations are explained in Section 3.5, Analysis of Members under Flexure (Part IV). The ultimate moment capacity (MuR) is calculated from the second equation.
F = 0 (A
pw
(9-1.8)
(9-1.9)
serviceability, such as deflection and crack width (for Type 3 members only). Before the calculation of the initial prestressing force (P0) and the eccentricity of the CGS (e) at the critical section, the type of composite section and the stages of prestressing need to be decided. Subsequently, a trial and error procedure is adopted for the design.
The following steps explain the design of a composite section with precast web and cast-in-place flange. The precast web is prestressed before the casting of the flange. The member is considered to be Type 1 member. Step 1. Compute e. With a trial section of the web, the CGS can be located at the maximum eccentricity (emax). The maximum eccentricity is calculated based on zero stress at the top of the precast web. This gives an economical solution. The following stress profile is used to determine emax.
CGC
ct e
Stress profile
fb
Figure 9-1.6
emax = kb +
Here, CGC = Centroid of the precast web
Msw P0
kb = Distance of the bottom kern of the precast web from CGC Msw = Moment due to self weight of the precast web. P0 = A trial prestressing force at transfer.
Step 2. Compute equivalent moment for the precast web. A moment acting on the composite section is transformed to an equivalent moment for the precast web. This is done to compute the stresses in the precast web in terms of the properties of the precast web itself and not of the composite section. For a moment Mc which acts after the section behaves like a composite section, the stresses in the extreme fibres of the precast web are determined from the following stress profile.
ct
CGC
ct cb bw
Composite section
ft
fb
Stress profile
Figure 9-1.7
ft =
fb =
Mc ct' I'
Mc cb' I'
ct = Distance of the top of the precast web from the CGC ct = Distance of the top of the composite section from the CGC. cb = Distance of the bottom of the precast web (or composite section)
from the CGC
The following quantities are defined as the ratios of the properties of the precast web and composite section.
I mt = I' I mb = I'
ct ct' cb cb'
Then the stresses in the extreme fibres of the precast web can be expressed in terms of
fb =
(9-1.11)
Here,
A = Area of the precast web kb = Distance of the bottom kern of the precast web from CGC kt = Distance of the top kern of the precast web from CGC
The quantities mt Mc and mb Mc are the equivalent moments. Thus, the stresses in the precast web due to Mc are expressed in terms of the properties of the precast web itself. Step 3. Compute Pe Let MP be the moment acting on the precast web prior to the section behaving like a composite section. After Mc is applied on the composite section, the total moment for the precast web is MP + mbMc.
The stress at the bottom for Type 1 member due to service loads is zero.
Therefore,
or,
(MP + mbMc )
e + kt
(9-1.12)
Note that the prestressing force is acting only on the precast web and hence, e is the eccentricity of the CGS from the CGC of the precast web. Step 4. Estimate P0 as follows. a) 90% of the initial applied prestress (Pi) for pre-tensioned members. b) Equal to Pi for post-tensioned members. The value of Pi is estimated as follows.
Pi = Ap(0.8fpk) Ap = Pe / 0.7fpk
(9-1.13) (9-1.14)
Revise e, the location of CGS, as given in Step 1 based on the new value of P0.
emax = kb +
Msw P0
(9-1.15)
fb = -
(9-1.16)
The stress fb should be limited to fcc,all, where fcc,all is the allowable compressive stress in concrete at transfer (available from Figure 8 of IS:1343 - 1980).
At service,
ft = -
Pe Pee (MP + mt Mc ) + A Ak b Ak b
(9-1.17)
The stress ft should be limited to fcc,all , where fcc,all is the allowable compressive stress in concrete under service loads (available from Figure 7 of IS:1343-1980). If the stress conditions are not satisfied, increase A.
The stress ft/ should be limited to fcc,all in concrete under service loads.
Mc ct" (9-1.18) I' , where fcc,all is the allowable compressive stress ft=
The required shear friction reinforcement (per metre span) is calculated as follows.
Asv = 1000bv h 0.87fy
(9-1.19)
The minimum requirements of shear friction reinforcement and spacing are similar to that for shear reinforcement in the web.
In the previous equation, Asv = area of shear friction reinforcement in mm2/m bv = width of the interface of precast and CIP portions
The shear reinforcement in the web can be extended and anchored in the CIP portion to act as shear friction reinforcement, as shown below.
bv
Intentionally roughened
Figure 9-1.8
Example 9-1.1
The mid-span section of a composite beam is shown in the figure. The precast web 300 mm 920 mm (depth) is post-tensioned with an initial force (P0) of 2450 kN. The effective prestress (Pe) is estimated as 2150 kN. Moment due to the self weight of the precast web (MSW) is 270 kNm at mid-span. After the web is erected in place, the top slab of 150 mm 920 mm (width) is casted (unpropped) producing a moment (MCIP) of 135 kNm. moment (MLL) of 720 kNm. Compute stresses in the section at various stages. After the slab
concrete has hardened, the composite section is to carry a maximum live load
Solution
1) Calculation of geometric properties.
638 460
CGC/ CGC
2) Calculation of stresses in web at transfer f =P0 P0ec MSW c A I I 3 245010 2450103 260 460 270106 460 = 2.76105 1.951010 1.951010 = -0.22 N/mm2 = -17.54 N/mm2
3) Calculation of stresses in web after long term losses Pe Peec MSW c A I I 3 215010 2150103 260 460 270106 460 = 2.76105 1.951010 1.951010 = -0.97 N/mm2 = -14.61N/mm2
f =-
4) Calculation of stresses in web after casting of flange f =Pe Peec (MSW + MCIP )c A I I 3 215010 2150103 260 460 (270 +135)106 460 = 2.76105 1.951010 1.951010 = -4.16 N/mm2 = -11.42 N/mm2
5) Calculation of stresses in the composite section at service Stress due to MLL At top fibre
MLLct// ft = I/ 750103 432 =4.621010 = -7.01N/mm2
/
At bottom fibre
fb =
At top fibre of precast web, the stress due to MLL is calculated from proportionality of triangles. ft = 7.01 282 432 = -4.57 N/mm2
Total stress in precast web At top fibre ft = -4.16 - 4.57 At bottom fibre fb = -11.42 +10.36
2
= -8.73 N/mm
= -1.06 N/mm2
Total stress in CIP slab The total stress is due to MLL only. At top fibre At bottom fibre ft / = -7.01N/mm2
f = - 4.57 N/mm
/ b
Stress profiles
7.01 0.22 0.97 4.16 4.57 7.01 4.57 8.73
17.54 At transfer
14.61
10.36
1.06 At service
After losses
Due to MLL
9.2.1 Introduction
Slabs are an important structural component where prestressing is applied. With
increase in the demand for fast track, economical and efficient construction, prestressed slabs are becoming popular. The slabs are presented in two groups: one-way slabs and two-way slabs. The two-way slabs are presented in details in Sections 9.3 and 9.4. Rectangular slabs can be divided into the two groups based on the support conditions and length-to-breadth ratios. The one-way slabs are identified as follows. 1) When a rectangular slab is supported only on two opposite edges, it is a one-way slab spanning in the direction perpendicular to the edges. Precast planks fall in this group. 2) When a rectangular slab is supported on all the four edges and the length-tobreadth (L / B) ratio is equal to or greater than two, the slab is considered to be a one-way slab. shorter edge. The following sketches show the plans of the two cases of one-way slabs. The spanning direction in each case is shown by the double headed arrow. The slab spans predominantly in the direction parallel to the
L
(a)
L
(b) Plans of one-way slabs
(a) Supported on two opposite edges (b) Supported on all edges (L/B > 2) Figure 9-2.1
Indian Institute of Technology Madras
A slab in a framed building can be a one-way slab depending upon its length-to-breadth ratio. A one-way slab is designed for the spanning direction only. For the transverse direction, a minimum amount of reinforcement is provided. A hollow core slab is also an example of a one-way slab. A ribbed floor (slab with joists) made of precast double tee sections, is analysed as a flanged section for one-way bending. Other types of rectangular slabs and non-rectangular slabs are considered to be twoway slabs. If a rectangular slab is supported on all the four sides and the length-tobreadth ratio is less than two, then it is a two-way slab. If a slab is supported on three edges or two adjacent edges, then also it is a two-way slab. A slab in a framed building can be a two-way slab depending upon its length-to-breadth ratio. A two-way slab is designed for both the orthogonal directions. A slab is prestressed for the following benefits. 1) Increased span-to-depth ratio Typical values of span-to-depth ratios in slabs are given below. Non-prestressed slab Prestressed slab 2) Reduction in self-weight 3) Section remains uncracked under service loads Increased durability 4) Quick release of formwork Fast construction 5) Reduction in fabrication of reinforcement 6) More flexibility in accommodating late design changes. Precast planks are usually pre-tensioned. Cast-in-situ slabs are post-tensioned. Posttensioned slabs are becoming popular in office and commercial buildings and parking structures, where large column-free spaces are desirable. The maximum length of a post-tensioned slab is limited to 30 to 40 m to minimise the losses due to elastic shortening and friction. Slabs can be composite for the benefits of reduction in form work, cost and time of construction and quality control. A precast plank can be prestressed and placed in the final location. A topping slab is overlaid on the precast plank. The grades of concrete in
Indian Institute of Technology Madras
28:1 45:1
the two portions can be different. The following sketches show the sections of some one-way slabs.
Here, the moment due to imposed loads is given as MI L = MT Msw. 7) Considering fpe = 0.7 fpk , calculate area of prestressing steel Ap = Pe / fpe .
8) Check the area of cross section (A) A = 1000 mm h mm. The average stress C/A should not be too high as compared to 50% fcc,all . Final Design The final design involves the checking of the stresses in concrete at transfer and under service loads with respect to the allowable stresses. The allowable stresses depend on the type of slab (Type 1, Type 2 or Type 3). Here, the steps of final design are explained for Type 1 slabs only. For Type 1 slabs, no tensile stress is allowed at transfer or under service loads. For small moment due to self-weight (Msw 0.3 MT), the steps are as follows. 1) Calculate eccentricity (e) to locate the centroid of the prestressing steel (CGS). The lowest permissible location of the compression (C) due to self-weight is at the bottom kern point (at a depth kb below CGC) to avoid tensile stress at the top. The design procedure based on the extreme location of C gives an economical section. For this location of C, the following equation can be derived.
Msw + kb P0
e=
(9-2.1)
The magnitude of C or T is equal to P0, the prestress at transfer after initial losses. The value of P0 can be estimated as follows. a) P0 = 0.9 Pi b) P0 = Pi for pre-tensioned slab for post-tensioned slab
Here, Pi is the initial applied prestress. Pi = (0.8fpk) Ap The permissible prestress in the tendon is 0.8fpk. 2) Re-compute the effective prestress Pe and the area of prestressing steel Ap. For the extreme top location of C under service load, the shift of C due to the total moment gives an expression of Pe. (9-2.2)
Pe =
MT e+ kT
(9-2.3)
For solid rectangular slab, kb = kt = h / 6. Considering fpe = 0.7fpk , the area of prestressing steel is recomputed as follows. Ap = Pe / fpe (9-2.4)
The number of tendons and their spacing is determined based on Ap. The value of P0 is updated. 3) Re-compute e with the updated values of Ap and P0. If the variation of e from the previous value is large, another cycle of computation of the prestressing variables can be undertaken. For large Msw if e violates the cover requirements, e is determined based on cover. 4) Check the compressive stresses in concrete For the limiting no tension design at transfer, the stress at the bottom (fb) is given as follows.
fb = P0 h 2P =- 0 A ct A
(9-2.5)
The stress should be less than fcc,all , where fcc,all is the allowable compressive stress in concrete at transfer. The condition to satisfy can thus be written as | fb | fcc,all. For the limiting no tension design at service, the stress at the top (ft) is given as follows.
2P P h f =- e =- e t A cb A
(9-2.6)
The stress should be less than fcc,all , where fcc,all is the allowable compressive stress in concrete at service. The condition to satisfy can be written as | ft | fcc,all. For Type 2 and Type 3 slabs, the tensile stress should be restricted to the allowable values. For a continuous slab, a suitable profile of the tendons is selected similar to that
Indian Institute of Technology Madras
in continuous beams. The design of continuous beams is covered in Sections 8.2 and 8.3. When the value of e is fixed (in either pre-tension or post-tension operations), the design steps are simpler. If the tendons are placed at the CGC (e = 0), then the uniform compressive stress due to prestress counteracts the tensile stress due to service loads. To have zero stress at the bottom under service conditions, the value of Pe can be directly calculated from the following equation.
Pe MT = A Zb M or, Pe = A T Zb
(9-2.7)
Zb is the section modulus. The above expression is same as Pe = MT / kt, with e = 0. The stresses at transfer can be checked with an estimate of P0 from Pe. 5) Checking for shear capacity The shear is analogous to that generates in a beam due to flexure. The calculations can be for unit width of the slab. The critical section for checking the shear capacity is at a distance effective depth d from the face of the beam, across the entire width of the slab. The critical section is transverse to the spanning direction. The shear demand (Vu) in the critical section generates from the gravity loads in the tributary area.
For adequate shear capacity, VuR Vu where, VuR = Vc, the shear capacity of uncracked concrete of unit width of slab. The expression of Vc is given in Section 5.2, Design for Shear (Part I). If this is not satisfied, it is preferred to increase the depth of the slab to avoid shear reinforcement.
6) Provide transverse reinforcement based on temperature and shrinkage. As per IS:456 - 2000, Clause 26.5.2.1, the minimum amount of transverse reinforcement (Ast,min in mm2) for unit width of slab is given as follows. Ast,min = 0.15% 1000 h = 0.12% 1000 h for Fe 250 grade of steel for Fe 415 grade of steel.
Usually the transverse reinforcement is provided by non-prestressed reinforcement. The minimum reinforcement is sufficient for the transverse moment due to Poissons effect and small point loads. For a heavy point load, transverse reinforcement needs to be computed explicitly. The following example shows the design of a simply supported precast prestressed composite slab. Reference: Santhakumar, A.R., Partially Precast Composite PSC Slab, Published by Building Technology Centre, Anna University, Chennai.
Example 9-2.1
Design a simply supported precast prestressed (Type 1) composite slab for the following data. Width of the slab Clear span Effective span (L) Thickness of the precast plank Thickness of the cast-in-situ topping slab Grade of concrete in precast plank Grade of concrete in topping slab = 0.3 m = 2.9 m = 3.1 m = 50 mm = 50 mm : M60 : M15
The pre-tensioned tendons are located at mid depth of the precast slab. During the casting of the topping, planks are not propped.
Solution
1) Calculation of moments. Load per unit area Weight of precast plank Weight of topping slab Weight of floor finish Live load Total = 1.25 kN/m2 = 1.25 kN/m2 = 1.50 kN/m2 = 2.00 kN/m2 = 6.00 kN/m2
Total moment (MT) along the width of the slab is given as follows.
wBL2 6 0.3 3.12 = 8 8 = 2.16 kNm
The individual moments are calculated based on the proportionality of the loads. MSW = moment due to self weight of precast plank = 2.16 (1.25 / 6.00) = 0.45 kNm Mtop = moment due to weight of topping slab = 2.16 (1.25 / 6.00) = 0.45 kNm Mfin = moment due to weight of floor finish = 2.16 (1.50 / 6.00) = 0.54 kNm MLL = moment due to live load = 2.16 (2.00 / 6.00) = 0.72 kNm. 2) Calculation of geometric properties. Precast section
50mm 300mm
I1 = 1 300 503 12 = 3,125,000 mm4
c b = ct =
50 2 = 25 mm
Section moduli
Zb = Zt = 3,125,000 25 = 125,000 mm3
Composite section Since the grades of concrete are different for the precast- prestressed (PP) and cast-insitu (CIS) portions, an equivalent (transformed) area is calculated. The CIS portion is assigned a reduced width based on the equivalent area factor (modular ratio). Equivalent area factor = Modulus of CIS / Modulus of PP = (Grade of CIS / Grade of PP) =15/60 = 0.5
Equivalent section
300mm
Location of CGC from bottom Atop Abot A2 = 50 150 = 50 300 = 7,500 mm2 = 15,000 mm2
y=
CGC
58.3 mm 41.7 mm
Ibot =
Distance to the extreme fibres yb = 41.7 mm yt = 58.3 mm Section moduli Zb = 17.19 106 /41.7 = 412,527 mm3 Zt = 17.19 106 /58.3 = 294,703 mm3 3) Calculation of prestress The tendons are located at the mid depth of the precast plank. Hence, e = 0 for the precast plank. The value of Pe is calculated directly from the following stress profiles.
+ Section
Pe
MSW + Mtop
Stress profiles
Mfin + MLL
To avoid tensile stress at the bottom under service conditions, the resultant stress is equated to zero.
Pe MSW + Mtop Mfin + MLL + + =0 A1 Z1b Z2b M + Mtop Mfin + MLL Pe = A1 SW + Z1b Z2 b
or,
In the above expression, the first term inside the bracket corresponds to the precast section. The moments due to self weight and topping slab are resisted by the precast section alone. The second term inside the bracket corresponds to the equivalent section. section.
0.45 + 0.45 0.54 + 0.72 + 106 P = A 412,527 125,000 = 50300 (7.2 + 3.0) = 153,816 N
Indian Institute of Technology Madras
The
moments due to weight of the floor finish and live load are resisted by the equivalent
Assuming around 20% loss, The prestress at transfer (P0) = 1.2 153,816 = 184,579 N. Wires of diameter = 7 mm and ultimate strength (fpk) = 1500 MPa are selected for prestressing. Area of one wire (Ap) = 38.48 mm2. The maximum allowable tension in one wire = 0.8 fpk Ap = 0.8 1500 38.48 = 46,176 N. No. wires required = 184,579 / 46,176 = 3.99 4. Required pull in each wire = 184,579 / 4 = 46,145 N. Total prestressing force (P0) = 4 46,145 = 184,580 N. Effective prestressing force (Pe) = 0.8 184,580 = 147,664 N. 4) Checking of stresses in concrete a) At transfer The compressive strength at 7 days (fci ) = 0.7 fck = 0.7 60 = 42 MPa. Allowable compressive stress (fcc,all) = 0.44 fci
= 0.44 42 = 18.5 MPa. For Type 1 members, the allowable tensile stress (fct,all) is zero.
Stresses at the mid-span of the precast portion fc = P0/A1 MSW/Z1 = 12.3 (0.45 106 / 125,000) ft fb = 15.9 MPa = 8.7 MPa
8.7 15.9
|ft| fcc,all OK b) After casting of topping slab at 28 days Allowable compressive stress (fcc,all) = 0.44 fck = 0.44 60 = 26.4 MPa. The allowable tensile stress (fct,all) is zero. Stresses at the mid-span of the precast portion fc = P0/A1 (MSW + Mtop)/Z1 = 12.31 ((0.45 + 0.45) 106 / 125,000) ft fb = 19.5 MPa = 5.1 MPa
19.5
|ft| fcc,all OK
5.1
c) At service i) For the precast portion Allowable compressive stress (fcc,all) = 0.35 fck = 0.35 60 = 21 MPa. The allowable tensile stress (fct,all) is zero. Stresses at the mid-span of the composite section for unpropped construction fc fjunc = Pe /A1 (MSW + Mtop) / Z1 (Mfin + MLL) / Z2 = (147,664 / 15,000) ((0.45 + 0.45) 106 / 125,000) ((0.54 + 0.72) 106 / 2,063,625) = 17.6 MPa fb = (147,664 / 15,000) + ((0.45 + 0.45) 106 / 125,000) + ((0.54 + 0.72) 106 / 412,527) = 0.4 MPa 0 |fjung| fcc,all fb = fct,all OK OK
ii) For cast-in-situ portion Allowable compressive stress (fcc,all) = 0.35 fck = 0.35 15 = 5.2 MPa. Stresses at the mid-span of the composite section ft fjunc = (0.54 + 0.72) 106 / 294,703) = 4.3 MPa = (0.54 + 0.72) 106 / 2,063,625)
4.3
0.4 17.6
0.4
= 0.6 MPa ft fcc,all OK Note that the critical stress at the junction is in the precast portion. 5) Check for shear VuR = Vc = Vc0 = 0.67bh(ft2 + 0.8fcpft) = 0.67 300 50 (1.862 + 0.8 9.36 1.86) = 41.9 kN Vu = wuB L / 2 = 1.5 6 0.3 3.1 / 2 = 4.2 kN VuR Vu Therefore, the shear capacity is adequate. 6) Transverse reinforcement Using Fe 415 grade of steel, for 1m width Ast,min = 0.12% 1000 h = 0.0012 1000 100 = 120 mm2. Provide 8 mm diameter bars at 300 mm on centre. 7) Provide nominal reinforcement for shrinkage in the longitudinal direction of the topping slab. Using Fe 415 grade of steel, for 1m width Ast,min = 0.12% 1000 h
Reinforcement details
(4) 7 mm wires
Introduction Analysis and Design Features in Modeling and Analysis Distribution of Moments to Strips
9.3.1 Introduction
The slabs are presented in two groups: one-way slabs and two-way slabs. The oneway slabs are presented in Section 9.2. When a rectangular slab is supported on all the sides and the length-to-breadth ratio is less than two, it is considered to be a two-way slab. The slab spans in both the orthogonal directions. A circular slab is a two-way slab. In general, a slab which is not falling in the category of one-way slab, is considered to be a two-way slab. Rectangular two-way slabs can be divided into the following types. 1) Flat plates: These slabs do not have beams between the columns, drop panels or column capitals. Usually, there are spandrel beams at the edges. 2) Flat slabs: These slabs do not have beams but have drop panels or column capitals. 3) Two-way slabs with beams: There are beams between the columns. beams are wide and shallow, they are termed as band beams. For long span construction, there are ribs in both the spanning directions of the slab. This type of slabs is called waffle slabs. The slabs can be cast-in-situ (cast-in-place). Else, the slabs can be precast at ground level and lifted to the final height. The later type of slabs is called lift slabs. A slab in a framed building can be a two-way slab depending upon its length-to-breadth (L / B) ratio. Two-way slabs are also present as mat (raft) foundation. The following sketches show the plan of various cases of two-way slabs. The spanning directions in each case are shown by the double headed arrows. If the
a) Flat plate
b) Flat slab
d) Waffle slab
Plans of two-way slabs
The absence of beams in flat plates and flat slabs lead to the following advantages. 1) Formwork is simpler 2) Reduced obstruction to service conduits 3) More flexibility in interior layout and future refurbishment. Two-way slabs can be post-tensioned. The main advantage of prestressing a slab is the increased span-to-depth ratio. As per ACI 318-02 (Building Code Requirements for Structural Concrete, American Concrete Institute), the limits of span-to-depth ratios are as follows. For floors For roofs 42 48.
The values can be increased to 48 and 52, respectively, if the deflection, camber and vibration are not objectionable. The following photographs show post-tensioned flat plate and flat slab.
(b) Flat slab Figure 9-3.2 Post-tensioned two-way slabs (Courtesy: VSL India Pvt. Ltd.)
beams.
For two-way slabs with beams, if the beams are sufficiently stiff, then the
method (based on moment coefficients) given in Annex D, IS:456 2000, is applicable. The direct design method of analysing a two-way slab is not recommended for prestressed slabs. The equivalent frame method is recommended by ACI 318-02. It is given in Subsection 31.5, IS:456 - 2000. This method is briefly covered in this section for flat plates and flat slabs. The slab system is represented by a series of two dimensional equivalent frames for each spanning direction. An equivalent frame along a column line is a slice of the building bound by the centre-lines of the bays adjacent to the column line. The width of the equivalent frame is divided into a column strip and two middle strips. The column strip (CS) is the central half of the equivalent frame. Each middle strip (MS) consists of the remaining portions of two adjacent equivalent frames. The following figure shows the division in to strips along one direction. investigation is shown by the double headed arrow. The direction under
1 MS
2 CS
3 MS
l1
l2
Figure 9-3.3 In the above figure, l1 = span of the equivalent frame in a bay l2 = width of the equivalent frame. This is the tributary width for calculating the loads. The following figure shows a typical elevation of an equivalent frame. Equivalent frame along Column Line 2
Figure 9-3.4
The analysis is done for each typical equivalent frame. An equivalent frame is modelled by slab-beam members and equivalent columns. The equivalent frame is analysed for gravity load and lateral load (if required), by computer or simplified hand calculations. Next, the negative and positive moments at the critical sections of the slab-beam members are distributed along the transverse direction. moments per unit width of a slab. If the analysis is restricted to gravity loads, each floor of the equivalent frame can be analysed separately with the columns assumed to be fixed at their remote ends, as shown in the following figure. The pattern loading is applied to calculate the moments for the critical load cases. This is discussed later. This provides the design
Figure 9-3.5
The steps of analysis of a two-way slab are as follows. 1) Determine the factored negative (Mu) and positive moment (Mu+) demands at the critical sections in a slab-beam member from the analysis of an equivalent frame. The values of Mu are calculated at the faces of the columns. The values of Mu+
Mu Mu +
Figure 9-3.6 Typical moment diagram due to gravity loads
2) Distribute Mu to the CS and the MS. These components are represented as Mu,
CS
These
Mu Mu +
MShalf CS MShalf
Mu
MuCS
MuCS
Figure 9-3.7
3) If there is a beam in the column line in the spanning direction, distribute each of Mu,CS and Mu,+CS between the beam and rest of the CS.
Mu Mu+
MShalf CS MShalf
Mu
Mu
Mu
CS
MS
Mu+MS Mu+CS
MuMS MuCS
Beam width
Figure 9-3.8 Distribution of moments to beam, column strip and middle strips
Indian Institute of Technology Madras
4) Add the moments Mu,MS and Mu,+MS for the two portions of the MS (from adjacent equivalent frames). 5) Calculate the design moments per unit width of the CS and MS.
Design Once the design moments per unit width of the CS and MS are known, the steps of design for prestressing steel are same as that for one-way slab. The profile of the tendons is selected similar to that for continuous beams. than by tendon distribution. The flexural capacity of prestressed slab is controlled by total amount of prestressing steel and prestress rather But the tendon distribution effects the load balancing. Some examples of tendon distribution are shown.
Normal reinforcement
Figure 9-3.9
Maximum spacing of tendons or groups of tendons should be limited to 8h or 1.5 m, whichever is less. Here, h is the thickness of the slab. A minimum of two tendons shall be provided in each direction through the critical section for punching shear around a column. The critical section for punching shear is described in Section 9.4, Two-way Slabs (Part II). Grouping of tendons is permitted in band beams. A minimum amount of non-prestressed reinforcement is provided in each direction based on temperature and shrinkage requirement. given as follows. Ast,min = 0.15% 1000h for Fe 250 grade of steel = 0.12% 1000h for Fe 415 grade of steel. The ducts for placing the individual strands are oval shaped to maintain the eccentricity, reduce frictional losses and convenient placement of crossing ducts. The ducts are not commonly grouted as the use of unbonded tendon is not detrimental in buildings. The following photo shows the ducts for the prestressing tendons and the nonprestressed reinforcement in a two-way slab. As per IS:456 - 2000, Clause 26.5.2.1, the minimum amount of reinforcement (Ast,min in mm2) for unit width of slab is
Gross section versus cracked section For determining the stiffness of a slab-beam member, the gross section can be considered in place of the cracked section. This is to simplify the calculation of moment of inertia of the section.
Equivalent column The actual column needs to be replaced by an equivalent column to consider the flexibility of the transverse beam in the rotation of the slab. The portions of the slab in the MS rotate more than the portions in the CS because of the torsional deformation of the transverse beam. column. In the following figure, the size of the arrows qualitatively represents the rotation of the slab. Note that the rotation is higher away from the
c2
c1
Upper column
h
Transverse beam
Figure 9-3.11
l2
Lower column
The transverse beam need not be a visible beam, but a part of the slab in the transverse direction, bounded by the edges of the column or column capital. modelled transverse beam is taken as shown in the following sketches. In presence of beam or column capital or in absence of beam, the cross-section of the
The flexibility of the equivalent column is equal to the sum of the flexibilities of the actual column and the transverse beam.
1 1 1 = + K ec K c Kt
(9-3.1)
Kc,upper = flexural stiffness of the upper column Kc,lower = flexural stiffness of the lower column Kt = torsional stiffness of the transverse beam.
An approximate expression for the flexural stiffness of a column (Kc) is given below.
Kc = 4Ec Ic L - 2h
(9-3.2)
Here, Ec = modulus of concrete L = length of the column h = thickness of the slab Ic = moment of inertia of the column. An approximate expression for torsional stiffness of the transverse beam (Kt) is given below.
Kt = 9EcC c l 2 1- 2 l2
(9-3.3)
Here, C = equivalent polar moment of inertia of transverse beam c2 = dimension of column in the transverse direction l2 = width of equivalent frame. For a rectangular section, the expression of C is given below.
x x3 y C = 1- 0.63 y 3
(9-3.4)
Here, x and y are the smaller and larger dimensions of the transverse beam. The expression of C is a lower bound estimate, that is, the calculated value is always lower than the actual moment of inertia of the transverse beam. For a transverse beam of compound section, the value of C is the summation of the individual values of the component rectangles. The splitting of the compound section into component rectangles should be such, so as to maximise the value of C. For the following two cases of splitting, select the larger value of C.
Figure 9-3.13
If there is a beam in the column strip in the spanning direction, then Kt is replaced by Kt (Isb / Is). Here, Is = moment of inertia of slab without the projecting portion of the beam (shaded area in Sketch (a) of the following figure) Isb = moment of inertia of slab considering it as a T-section (shaded area in Sketch (b) of the following figure).
l2
(a)
Figure 9-3.14
(b)
Slabbeam members The variation of the flexural moment of inertia of a slab-beam member is considered as follows. The value of the moment of inertia (I) is constant (say equal to I1) in the prismatic portion, that is, in between the faces of the columns or column capitals or drop panels. It is also constant, with a different value (say equal to I2) in the region of a drop panel. The value varies in the region from the face of the column or column capital to the center line of the column. But it is approximated to a constant value equal to the following.
I2 c2 1- l 2
Here, I2 = moment of inertia at the face of the column or column capital c2 = dimension of column in the transverse direction l2 = width of equivalent frame. The following figure shows the variation of the moment of inertia of the slab beam member.
2
(9-3.5)
(a)
I2 c2 1- l 2
2
I1
I2
(b)
Figure 9-3.15
a) Elevation of equivalent frame, b) Variation of the moment of inertia of the slab-beam member
Arrangement of live load Since the factored live load (wu, LL) may not occur uniformly in all the spans in a floor, a distribution is considered to generate the maximum values of the negative (Mu) and positive moments (Mu+) at the critical sections. If the distribution of wu, LL is known, then the load is applied accordingly. If the distribution is not known, then a pattern loading is considered based on the value of wu, LL with respect to that of the factored dead load (wu, DL). Of course, the load case with wu, LL on all the spans should be also analysed. 1) For wu, LL wu, DL The possible variation in wu,LL in the different spans is neglected. uniformly on all the spans. wu,LL is applied
wu,LL.l2 wu,DL.l2
A B C D E
For maximum value of Mu+ in a span, wu,LL is applied on the span and the alternate spans. For example, if the maximum value of Mu+ in Span BC of the frame below is to be determined, then wu,LL is placed in Spans BC and DE. This distribution will also give the maximum value of Mu+ in Span DE. For maximum value of Mu near the
support, wu LL is applied on the adjacent spans only. For example, if the maximum value of Mu near Support B is to be determined, then wu,LL is placed in Spans AB and BC.
wu,LL.l2 A B C D E
wu,DL.l2
wu,LL.l2
wu,DL.l2
Distribution of live load for maximum Mu- near Support B Figure 9-3.17 Distribution of live load for wu LL > wu,DL
Critical section near a support The critical section is determined as follows. 1) At interior support At the face of support (column or column capital, if any), but not further than 0.175l1 from the center line of the column. 2) At exterior support At a distance from the face of column not greater than half the projection of the column capital (if any).
Distribution of Mu at interior support Mu,CS = 0.75 Mu Mu,MS = 0.25 Mu Here, Mu,CS = negative moment in the CS Mu,MS = total negative moment in the two MS at the sides. Distribution of Mu at exterior support If the width of the column or wall support is less than l2, Mu,CS = Mu Mu,MS = 0. distributed along the width l2. Distribution of Mu+ at mid span Mu,+CS = 0.60 Mu+ Mu,+MS = 0.40 Mu+ Here, Mu,+CS = positive moment in the CS Mu,+MS = total positive moment in the two MS at the sides. The total moments in MS (Mu,MS and Mu,+MS) are distributed to the two middle strips at the sides of the equivalent frame, proportional to their widths. The combined MS from two adjacent equivalent frames is designed for the sum of the moments assigned to its parts. (9-3.10) (9-3.11) (9-3.8) (9-3.9) (9-3.6) (9-3.7)
If the width of the column or wall support is greater than l2, then Mu is uniformly
Checking for Shear Capacity Spandrel Beams Anchorage Devices Additional Aspects
One-way shear The one-way shear is analogous to that generates in a beam due to flexure. This is checked in a two-way slab for each spanning direction separately. The critical section for checking the shear capacity is at a distance effective depth d from the face of the column, across the entire width of the frame. The critical section is transverse to the spanning direction. For gravity loads, the shear demand in the critical section generates from the loads in the tributary area shown in the next figure. For lateral loads, the shear demand is calculated from the analysis of the equivalent frame. In presence of a drop panel two critical sections need to be checked. The first section is at a distance d1 from the face of the column, where d1 is the effective depth of the drop panel. The second section is at a distance d2 from the face of the drop panel, where d2 is the effective depth of the slab.
Critical section
C L
Plan
l2 d
ln
Elevation
Critical sections
C L
Plan
l2
Elevation
d1
d2
ln
(b) Slab with drop panel Figure 9-4.1 Critical sections for one-way shear
The calculations for shear can be for unit width of the slab. The shear demand due to gravity loads per unit width is given as follows. Vu = wu (0.5ln d) Here, ln is the clear span along the spanning direction. The shear capacity per unit width is given as follows. VuR = Vc (9-4.2) (9-4.1)
Vc is the shear capacity of uncracked concrete of unit width of slab. The expression of Vc is given in Section 5.2, Design for Shear (Part I). For adequate shear capacity VuR Vu (9-4.3)
If this is not satisfied, it is preferred to increase the depth of the slab to avoid shear reinforcement along the width of the slab.
Two-way shear The two-way shear is specific to two-way slabs. If the capacity is inadequate, the slab may fail due to punching around a column. The punching occurs along a conical frustum, whose base is geometrically similar and concentric to the column crosssection. The following figure illustrates the punching shear failure.
Elevation
Figure 9-4.2
Isometric view
Punching shear failure
Two-way shear is checked for the two spanning directions simultaneously. The critical section for checking the shear capacity is geometrically similar to the column perimeter and is at a distance d / 2 from the face of the column. The depth of the critical section is equal to the average of the effective depths of the slab in the two directions. The sketches below show the critical section. The tributary area of the column is the area within the centre-lines of the spans minus the area within the critical section. It is shown shaded in the third sketch.
b1
Critical section
b2 c1
(a) Plana
c2
b2 d b1
(b) Isometric view r
C L Critical section C L C L
(c) Tributary area for shear due to gravity loads Figure 9-4.3 Critical section for two-way shear
C L
The lengths of the sides of the critical section along axes 1-1 and 2-2 are denoted as b1 and b2, respectively. b1 = c1 + d b2 = c2 + d Here, c1 = dimension of the column or column capital along axis 1-1 c2 = dimension of the column or column capital along axis 2-2. For a non-rectangular column, the critical section consists of the slab edges as per Figure 13, IS:456 - 2000. For edge and corner columns, the critical section consists of the slab edges as per Figure 14, IS:456 - 2000.
Indian Institute of Technology Madras
(9-4.4) (9-4.5)
V v = u + b0d
Here,
Muv
2-2
(9-4.6)
Vu = shear due to gravity loads from the tributary area Muv = fraction of moment transferred about an axis b0 = perimeter of the critical section = 2(b1 + b2). J = polar moment of inertia of the critical section about an axis
The second and third terms are due to transfer of moments from slab to column. The moment about an axis is due to the unbalanced gravity loads for the two sides of the column or due to lateral loads. It is transferred partly by the variation of shear stress in the critical section and the rest by flexure. The fraction transferred by the variation of shear stress about an axis is denoted as Muv. Muv|2-2 = Fraction of moment transferred about axis 2-2 Muv|1-1 = Fraction of moment transferred about axis 1-1 The forces and stresses acting at the critical section are shown below.
2 1 2 1
Vu
Muv|2-2
2 1 2 Due to Muv|2-2
Muv|1-1
Due to Vu
Due to Muv|1-1
The fraction of moment transferred by the variation of shear stress about an axis (Muv), is given in terms of the total moment transferred (Mu) as follows. Muv = (1- )Mu
(9-4.7)
The value of Mu due to unbalanced gravity loads is calculated by placing live load on one side of the column only. The value of Mu due to lateral loads is available from the analysis of the equivalent frame. The parameter is based on the aspect ratio of the critical section.
= 1 2 b1 1+ 3 b2
(9-4.8)
The polar moments of inertia of the critical section, about the axes are given as follows.
2 1 1 b2 3 3 J 1-1 = 2 b2d + db2 + b1d 12 2 12
(9-4.9)
(9-4.10)
v ks c
The shear stress capacity of concrete for a square column is given as follows.
(9-4.11)
c = 0.25 fck
(9-4.12)
Here, fck is the characteristic strength of the concrete in the slab. The effect of prestress is neglected. The factor ks accounts for the reduced shear capacity of non-square columns. ks = 0.5 + c
(9-4.13)
The value of ks should be less than 1.0. c is a parameter based on the aspect ratio of the column cross-section. It is the ratio of the short side to long side of the column or column capital. If v exceeds ksc, a drop panel or shear reinforcement needs to be provided at the slabto-column junction. The shear reinforcement can be in the form of stirrups or I section (shear head) or based on shear studs. The reinforcement based on shear studs reduces congestion for conduits and post-tensioning tendons. If v exceeds 1.5c, then the depth of the slab needs to be increased in the form of drop panels. The stirrups are designed based on the following equation. Asv is the area of the vertical legs of stirrups.
Asv =
( v - 0.5 c )
0.87fy
(9-4.14)
The stirrups are provided along the perimeter of the critical section. The first row of stirrups should be within a distance of 0.5d from the face of the column. They can be continued in outer rows (concentric and geometrically similar to the critical section) at an interval of 0.75d, till the section with shear stress v = 0.5 . The different types of reinforcement at the slab-to-column junction are shown in the following sketches. References: 1. Bureau of Indian Standards, Handbook on Concrete Reinforcement and Detailing (SP 34 : 1987)
0.5d
~ 2d
The following photo shows the ducts and reinforcement at the slab-column junction in a slab with a drop panel.
Figure 9-3.6
(Courtesy: VSL India Pvt. Ltd.) The residual moment transferred by flexure (Muf), is given in terms of the total moment transferred (Mu) as follows. Muf = Mu
(9-4.15)
Additional non-prestressed reinforcement is provided at the top of the slab over a width c2 + 3h (centred with respect to the column) to transfer Muf.
The following figure shows the distribution of the torsional loading on the spandrel beam.
Variation of torsion
Tu,max
l2
Figure 9-4.7
The maximum torsion (Tu,max) is given as follows. Here, Mu,e is the moment at the exterior support of the equivalent frame.
l2 - c2 Mu,e u,max = l2 2
(9-4.16)
The following photos show the anchorage devices, end zone reinforcement, spandrel beam before casting and stretching and anchoring of the tendons after casting of concrete in a slab.
Spandrel Beam
Bursting links
Bursting links
End-zone reinforcement and anchoring of tendons in a slab (Reference: VSL India Pvt. Ltd.)
Due to the restraint from monolithic columns or walls, the prestressing force in the slab is reduced. Hence, the stiff columns or walls should be located in such a manner that they offer least restraint. Alternatively, sliding joints can be introduced which are made ineffective after post-tensioning of the slab.
Calculation of deflection
The deflection of a two-way slab can be approximately calculated by the equivalent frame method. The deflection at a point is the summation of the deflections of the two orthogonal strips passing through the point. For an accurate evaluation, the following models can be adopted. a) Grillage model b) Finite element model.
Proportioning of drop panels and column capitals Section 31 of IS:456 - 2000 provide guidelines for proportioning drop panels and
column capitals. A minimum length and a minimum depth (beyond the depth of the slab) of a drop panel are specified. For column capitals it is preferred to have a conical flaring at a subtended angle of 90. The critical sections are shown in Figure 12 of the code.
9.5.1 Introduction
Prestressing is meaningful when the concrete in a member is in tension due to the external loads. Hence, for a member subjected to compression with minor bending, prestressing is not necessary. But, when a member is subjected to compression with substantial moment under high lateral loads, prestressing is applied to counteract the tensile stresses. Examples of such members are piles, towers and exterior columns of framed structures. As the seismic forces are reversible in nature, the prestressing of piles or columns is concentric with the cross-section. Some typical cross sections are shown below.
Figure 9-5.2
(Reference: Industrial Concrete Products Berhad) Precast prestressed piles have the following advantages. 1) Large bending and axial tension capacities. 2) Better quality control than bored and CIP piles. 3) Section remains uncracked under service loads Increased durability. 4) Use of high strength concrete gives reduced section. The piles are prestressed by pre-tensioning. The procedure of pre-tensioning is explained under Pretensioning Systems and Devices. Since a prestressed member is under self equilibrium, there is no buckling of the member due to internal prestressing with bonded tendons. In a deflected shape, there is no internal moment due to prestressing.
C T
P
b) Under internal prestressing
Figure 9-5.3
In the first free body sketch of the above figure, the external compression P causes an additional moment due to the deflection of the member. The value of the moment at mid-height is P. This is known as the member stability effect, which is one type of P- effect. If this deflection is not stable, then buckling of the member occurs. In the second free body sketch, there is no moment due to the deflection of the member and the prestressing force, since the compression in concrete (C) and the tension in the tendons (T) balance each other. When the additional moment due to deflection of the member is negligible, the member is termed as short member. The additional moment needs to be considered when the slenderness ratio (ratio of effective length and a lateral dimension) of the member is high. The member is termed as slender member. In the analysis of a slender member, the additional moment is calculated by an approximate expression or second order analysis. In this module only short members will be considered.
9.5.2 Analysis
Analysis at Transfer The stress in the section can be calculated as follows.
fc = P0 A
(9-5.1)
Here, A = Area of concrete P0 = prestress at transfer after short-term losses. In this equation, it is assumed that the prestressing force is concentric with the crosssection. For members under compression, a compressive stress is considered to be positive. The permissible prestress and the cross-section area are determined based on the stress to be within the allowable stress at transfer (fcc,all). Analysis at Service Loads The analysis is analogous to members under flexure. The stresses in the extreme fibres can be calculated as follows.
fc = Pe N Mc + A At It
(9-5.2)
In this equation, the external compression for a prestressed member is denoted as N and is concentric with the cross section. The eccentricity is considered in the external moment M. In the previous equation, A = area of concrete At = area of the transformed section c = distance of the extreme fibre from the centroid (CGC) It = moment of inertia of the transformed section Pe = effective prestress. The value of fc should be within the allowable stress under service conditions (fcc,all). Analysis at Ultimate When the average prestress in a member under axial compression and moment is less than 2.5 N/mm2, Clause 22.2, IS:1343 - 1980, recommends to analyse the member as a reinforced concrete member, neglecting the effect of prestress. For higher prestress, the analysis of strength is done by the interaction diagrams.
At the ultimate limit state, an interaction diagram relates the axial force capacity (NuR) and the moment capacity (MuR). It represents a failure envelop. Any combination of factored external loads Nu and Mu that fall within the interaction diagram is safe. A typical interaction diagram is shown below. The area shaded inside gives combinations of Mu and Nu that are safe.
NuR
Compression failure
1 eN
MuR
Figure 9-5.4 A typical interaction diagram for compression and bending
The radial line in the previous sketch represents the load path. Usually the external loads increase proportionally. At any load stage, M and N are related as follows. M = N eN (9-5.3)
Here, eN represents the eccentricity of N which generates the same moment M. The slope of the radial line represents the inverse of the eccentricity (1/eN). At ultimate, the values of M and N (Mu and Nu, respectively) correspond to the values on the interaction diagram. For high values of N as compared to M, that is when eN is small, the concrete in the compression fibre will crush before the steel on the other side yields in tension. This is called the compression failure. For high values of M as compared to N, that is when eN is large, the concrete will crush after the steel yields in tension. This is called the tension failure. The transition of these two cases is referred to as the balanced failure, when the crushing of concrete and yielding of steel occur simultaneously. For a prestressed compression member, since the prestressing steel does not have a definite yield point, there is no explicit balanced failure.
D
CGC +
B Ap1
Figure 9-5.5 The notations used are as follows.
d1
d2
Ap2
B = dimension of section transverse to bending D = dimension of section in the direction of bending Ap1 = area of prestressing tendons at the tension face Ap2 = area of prestressing tendons at the compression face d1, d2 = distances of centres of Ap1 and Ap2, respectively, from the centroid of the section (CGC). The strain compatibility equation is necessary to relate the strain in a prestressing tendon with that of the adjacent concrete. Due to a concentric prestress, the concrete at a section undergoes a uniform compressive strain. With time, the strain increases due to the effects of creep and shrinkage. At service, after the long term losses, let the strain be ce. Also, let the strain in the prestressing steel due to effective prestress be pe.
ce pe
Figure 9-5.6
pe
The strain compatibility equation for the prestressed tendons is given below.
p = c + p where, p = pe - ce
(9-5.4)
The strain difference of the strain in a prestressing tendon with that of the adjacent concrete is denoted as p. The design stress-strain curve for concrete under compression is used. This curve is described in Section 1.6, Concrete (Part II). The design stress-strain curve for the prestressed tendon under tension is expressed as fp = F(p). The calculation of NuR and MuR for typical cases of eN or xu are illustrated. The typical cases are as follows. 1) Pure compression (eN = 0, xu = ) 2) Full section under varying compression (0.05D < eN eN xu = D , xu D) 3) Part of section under tension (eN xu = D < eN , xu < D) 4) Pure bending (eN = , xu = xu,min) The above cases are illustrated in the following sketches.
e=0
e
Elevation
xu
Case 1 Case 2
Strain diagram
e
Elevation
xu
Case 3
Figure 9-5.7
Strain diagram
xu
Case 4
Typical cases of eccentricity and depth of neutral axis
In addition to the above cases, the case of pure axial tension is also calculated. The straight line between the points of pure bending and pure axial tension provides the interaction between the tensile force capacity and the moment capacity.
MuR
NuR
Figure 9-5.8
1. Pure compression (eN = 0, xu = ). The following sketches represent the strain and stress profiles across the section and the force diagram.
0.002 p1 p2 Strain profile
fp1
0.447 fck
fp2
Stress profile
Tu1
Figure 9-5.9
Cu
Tu2
Force diagram
Cu Tu1
(9-5.5) (9-5.6)
The steel is in the elastic range. The total area of prestressing steel is Ap = Ap1 + Ap2. The area of the gross-section Ag = BD. The moment and axial force capacities are as follows. MuR = 0 NuR = Cu Tu1 Tu2 = 0.447fck (Ag Ap) Ap Ep (pe 0.002 ce)
(9-5.8) (9-5.7)
In design, for simplification the interaction diagram is not used for eccentricities eN 0.05D. To approximate the effect of the corresponding moment, the axial force capacity is reduced by 10%.
NuR = 0.4fck (Ag Ap) 0.9Ap Ep (pe 0.002 + ce) (9-5.9)
2. Full section under varying compression (0.05D < eN eN xu = D , xu D) The following sketches represent the strain and stress profiles across the section and the force diagram.
xu = kD
0.002 p1 p2 3 / 7 D Strain profile
f p1 g Asector Tu1
Figure 9-5.10
fp2
0.447 fck Stress profile Force diagram
Cu Tu2
The limiting case for full section under compression corresponds to xu = D, when the neutral axis lies at the left edge of the section. The strain diagram pivots about a value of 0.002 at 3/7D from the extreme compression face. To calculate Cu, first the
reduction of the stress at the edge with lower compression (g) is evaluated. Based on the second order parabolic curve for concrete under compression, the expression of g is as follows.
4 D 7 g = 0.447fck 3 kD - 7 D 4 = 0.447fck 7k - 3
2
(9-5.10)
The area of the complementary sector of the stress block is given as follows.
1 4 Asector = g D 3 7 4 = gD 21
(9-5.11)
Asector x/ g
+ 4/7D
Figure 9-5.11
Distance of centroid from apex (x /) = (3/4)(4/7)D = 3/7 D The forces are as follows.
Cu = [0.447fck D - Asector ] B 4 = 0.447fck D - gD B 21 4 4 2 = 0.447fck BD 1- 21 7k - 3
(9-5.12)
(9-5.13)
Tu 2 = Ap 2fp 2 = Ap 2E p p 2 = Ap 2E p ( c 2 + p ) D -d x u 2 2 + = Ap 2E p -0.002 p 3D x u 7
(9-5.14)
The strains in the concrete at the level of the prestressing steels c1 and c2 are determined from the similarity of triangles of the following strain profile.
c1
xu (D/2 + d1) xu (3 / 7 D)
0.002
0.002
xu (3 / 7 D) xu ( D / 2 d 2 )
Figure 9-5.12
c2
The moment and axial force capacities are as follows. NuR = Cu Tu1 Tu2 MuR = Mc + Mp
(9-5.15) (9-5.16)
The expressions of Mc and Mp about the centroid are given below. Anticlockwise moments are considered positive. The lever arms of the tensile forces are shown in the following sketch.
(9-5.17)
(9-5.18)
d1 Tu1
Figure 9-5.13
d2
+
Cu Tu2
3. Part of section under tension (eN xu = D < , eN , xu < D) The following sketches represent the strain and stress profiles across the section and the force diagram.
p 1
xu p2
0.0035 Strain profile
fp1
fp2
0.447fck Stress profile Force diagram
Tu1
Figure 9-5.14
Cu
Tu2
The forces are as follows. The compression is the resultant of the stress block whose expression can be derived similar to a reinforced concrete section. Cu = 0.36fck xu B Tu1 = Ap1 fp1 = Ap1 F (p1) = Ap1 F (c1+ p) Tu2 = Ap2 fp2 = Ap2 Ep p2 = Ap2 Ep (c2+ p) The strains c1 and c2 are calculated from the similarity of triangles of the following strain diagram.
D + d1 - xu 2
c 1
0.0035 xu
(9-5.19)
c 2 0.0035 =D xu xu - - d2 2
(9-5.20)
c1
xu c2
0.0035
The moment and axial force capacities are as follows. NuR = Cu Tu1 Tu2 MuR = Mc + Mp The expressions of Mc and Mp about the centroid are as follows. Mc = 0.36fck xu B [ (D / 2) 0.42 xu ] Mp = Tu1d1 Tu2d2
(9-5.23) (9-5.24) (9-5.21) (9-5.22)
The lever arms of the forces are shown in the following sketch. The location of Cu is similar to that of a reinforced concrete section.
d1 Tu1
+
d2 Cu Tu2
(D / 2) 0.42 xu
Figure 9-5.16
4. Pure bending (eN = , xu = xu,min) The value of xu is determined by trial and error from the condition that the sum of the forces is zero. Cu Tu1 Tu2 = 0 or, 0.36fck xu B Ap1 fp1 Ap2 fp2 = 0
(9-5.25)
or,
xu =
Ap1fp1 + Ap 2E p p 2 0.36fck B
(9-5.26)
The strains p1 and p2 are calculated from the strain compatibility equations. The strain p2 is within the elastic range, whereas p1 may be outside the elastic range. The stresses fp1 and fp2 are calculated accordingly from the stress versus strain relationship of prestressing steel. The steps for solving xu are as follows. 1) Assume xu = 0.15 D (say). 2) Determine p1 and p2 from strain compatibility. 3) Determine fp1 and fp2 from stress versus strain relationship. 4) Calculate xu from Eqn. (9-5.26). 5) Compare xu with the assumed value. Iterate till convergence. The moment and axial force capacities are as follows. NuR = 0 MuR = Mc + Mp The expressions of Mc and Mp are same as the previous case. 5. Axial tension The moment and axial force capacities are as follows. The cracked concrete is neglected in calculating the axial force capacity. NuR = 0.87fpk Ap MuR = 0 The above sets of NuR and MuR are joined to get the interaction diagram.
(9-5.29) (9-5.30) (9-5.27) (9-5.28)
Example 9-5.1
Calculate the design interaction diagram for the member given below. member is prestressed using 8 strands of 10 mm diameter. stress relieved with the following properties. Tensile strength (fpk) Total area of strands = 1715 N/mm2. = 8 51.6 The
= 413.0 mm2 Effective prestress (fpe) Modulus (Ep) Strain under fpe (pe) Grade of concrete Strain under fpe (ce) = 1034 N/mm2 = 200 kN/mm2 = 0.0042.
= M40 = 0.0005.
300
300
50
200
50
Solution
Calculation of geometric properties and strain compatibility relationship. Ag Ap1 = Ap2 d1 = d2 p = 300 300 = 90,000 mm2 = 4 51.6 = 206 mm2 = 100 mm = 0.0042 + 0.0005 = 0.0047
p = c + 0.0047
1. Pure compression (eN = 0, xu = ) MuR Cu = 0 kNm = 0.447fck (Ag Ap) = 0.447 40 (90,000 413) = 1601.8 kN Tu1 = Tu2 = Ap1 Ep ( 0.002 + p) = 206.4 200 (0.0047 0.002) = 111.5 kN
Indian Institute of Technology Madras
NuR
With 10% reduction, to bypass the use of interaction diagram for eccentricities eN 0.05D NuR = 1204.9 kN
Select xu
= 400 mm = (4 / 3) 300 mm
=4/3
2
4 g = 0.447 fck 7k - 3
150
c1
c2
Tu 2 = Ap 2E p ( c 2 + p ) 350 = 206.4 200 -0.002 + 0.0047 271.4 = 87.5 kN NuR = Cu Tu1 Tu2 = 1486.9 148.4 87.5 = 1251.0 kN Limit NuR to 1240.9 kN to bypass the use of interaction diagram for eccentricities eN 0.05D. Mc = 10 gD 2B 147 10 = 7.133002 300 147 = 13.1 kNm
Tu2
= 73.6 kN
MuR
= 42.5 kNm.
3. Part of section under tension (eN xu = D < eN , xu < D) Select xu Cu = 200 mm. = 0.36fck xu B = 0.36 40 200 300 = 864.0 kN
c 1 =
Strain corresponding to elastic limit py = 0.87 0.8fck / Ep = 0.87 1715 / 200 103 = 0.0059. p1 < py
fp1
Tu1
200
c2
150
0.0035
f p2
Tu2
NuR
Mc
Mp
MuR
4. Pure bending (eN = , xu = xu,min) NuR = 0.0 kN Try xu = 100 mm. Cu = 0.36fck xu B = 0.36 40 100 300 = 432.0 kN c 1 = 0.0035 150 100 = 0.0052
c1
From stress-strain curve f p1 = 0.87fpk = 1492 N/mm2 Tu1 = Ap1 fp1 = 206.4 1492 = 308.0 kN
100
c2
50
0.0035
c 2 = -
p 2 = - 0.0017 + 0.0047 = 0.0029 f p2 = Ep p2 = 200 103 0.0029 = 580 N/mm2 Tu2 = Ap2 fp2 = 206.4 580 = 120.0 kN Tu1 + Tu2 = 428.0 kN This is close enough to Cu = 432.0 kN. Hence, the trial value of xu is satisfactory.
Indian Institute of Technology Madras
Mc
= 0.36fck xu B [ (D / 2) 0.42xu ] = 0.36 40 100 300 (150 0.42 100) = 46.6 kNm
Mp
MuR
5. Axial tension MuR NuR = 0.0 kNm = 0.87fpk Ap = 0.87 1715 413.0 = 616.2 kN The above sets of NuR and MuR are joined to get the following interaction diagram. The limit on axial force capacity to consider the effect of eccentricity less than 0.05D, is not shown.
1500 1000
N uR (kN)
Along with the interaction curve for the prestressed concrete (PC) section, the interaction curves for two reinforced concrete (RC) sections are plotted. The section denoted as RC 1 has the same moment capacity at zero axial force. The section denoted as RC 2 has the same axial force capacity at zero moment. The gross section of RC 1 is same as that of PC, but the section of RC 2 is smaller.
2500 2000 1500
NuR (kN)
PC RC 1
RC 2
Figure 9-5.17
0 0
NuR (kN)
20
40
60
80
M uR (kNm)
PC Strength
PC Cracking
RC 1 Strength
RC 1 Cracking
Figure 9-5.18
Comparing the curves for PC and RC 2, it is observed that if the moment demand is small, then a smaller reinforced concrete section is adequate to carry the axial force. Of course with increasing moment, the flexural capacity of the prestressed concrete section is higher.
Indian Institute of Technology Madras
sections with same flexural capacities, the axial load capacity of a prestressed concrete section is less. However if there is tension, the cracking load combination is higher for PC as compared to RC 1. Thus, prestressing is beneficial for strength when there is occurrence of: a) Large moment in addition to compression b) Moment along with tension. Such situations arise in piles or columns subjected to seismic forces. In presence of tension, prestressing is beneficial at service loads due to reduced cracking. prestressed reinforcement may be used for supplemental capacity. Non-
Introduction General Analysis and Design Prestressed Concrete Pipes Liquid Storage Tanks Ring Beams Conclusion
9.6.1 Introduction
When the prestressed members are curved, in the direction of prestressing, the prestressing is called circular prestressing. For example, circumferential prestressing in pipes, tanks, silos, containment structures and similar structures is a type of circular prestressing. In these structures, there can be prestressing in the longitudinal direction (parallel to axis) as well. Circular prestressing is also applied in domes and shells. The circumferential prestressing resists the hoop tension generated due to the internal pressure. The prestressing is done by wires or tendons placed spirally, or over sectors of the circumference of the member. The wires or tendons lay outside the concrete core. Hence, the centre of the prestressing steel (CGS) is outside the core concrete section. The hoop compression generated is considered to be uniform across the thickness of a thin shell. Hence, the pressure line (or C-line) lies at the centre of the core concrete section (CGC). The following sketch shows the internal forces under service conditions. The analysis is done for a slice of unit length along the longitudinal direction (parallel to axis).
CGS CGC
p Pe C C Pe T R
a) Due to prestress
Figure 9-6.1
To reduce the loss of prestress due to friction, the prestressing can be done over sectors of the circumference. Buttresses are used for the anchorage of the tendons. The following sketch shows the buttresses along the circumference.
Buttress
Figure 9-6.2
Analysis at Transfer The compressive stress can be calculated from the compression C. From equilibrium, C = P0, where P0 is the prestress at transfer after short-term losses. The compressive stress (fc) is given as follows
fc = -
P0 A
(9-6.1)
Here,
fc = Here,
Pe pR + A At
(9-6.2)
Design The internal pressure p and the radius R are given variables. It is assumed that the prestressing steel alone carries the hoop tension due to internal pressure, that is
Pe =
Apfpe = pR.
The steps of design are as follows. 1) Calculate the area of the prestressing steel from the equation Ap = pR / fpe. 2) Calculate the prestress at transfer from an estimate of the permissible initial stress fp0 and using the equation
P0 = Ap fp0.
3) Calculate the thickness of concrete shell from the following equation.
(9-6.3)
A = P0 / fcc,all
(9-6.4)
Here, fcc,all is the allowable compressive stress at transfer. 4) Calculate the resultant stress fc at the service conditions using Eqn. (9-6.2). The value of fc should be within fcc,all, the allowable stress at service conditions.
Figure 9-6.3 Stacked prestressed pipes (Courtesy: The Indian Hume Pipe Co. Ltd., Mumbai) The analysis and design of prestressed concrete pipes consider the stresses due to the different actions. A horizontal layout of the pipe is considered to illustrate them.
Analysis The stresses in the longitudinal direction are due to the following actions. 1. Longitudinal prestressing 2. Circumferential prestressing 3. Self weight 4. Transport and handling 5. Weight of fluid 6. Weight of soil above (fl1) (fl2) (fl3) (fl4) (fl5) (fl6)
fl 1 = Here,
Pe Ac1
(9-6.5)
(9-6.6)
Self weight If the pipe is not continuously supported, then a varying longitudinal stress generates due to the moment due to self weight (Msw).
fl 3 = Msw Zl
(9-6.7)
Here,
Transport and handling A varying longitudinal stress generates due to the moment during transport and handling (Mth).
fl 4 = Mth Zl
(9-6.8)
Weight of fluid Similar to self weight, the moment due to weight of the fluid inside (Mf) generates varying longitudinal stress.
fl 5 = Mf Zl
(9-6.9)
Weight of soil above The weight of soil above for buried pipes is modelled as an equivalent distributed load. The expression of stress (fl6) is similar to that for the weight of fluid. The longitudinal stresses are combined based on the following diagram.
Section of pipe
Figure 9-6.4
fl 1
fl 2
fl 3 + fl 4 + fl 5 + fl 6
The stresses in the circumferential direction are due to the following actions. 1. Circumferential prestressing 2. Self weight 3. Weight of fluid 4. Weight of soil above 5. Live load 6. Internal pressure (fh1) (fh2) (fh3) (fh4) (fh5) (fh6)
fh 1 = =-
Ps Ac 2 Ps 1 tc
(9-6.10)
Here,
Ps = tensile force in spiral wire in unit length of pipe Ac2 = area for longitudinal section of unit length tc = thickness of the core.
Actions 2. to 5.
For each of these actions, first the vertical load per unit length (W) is calculated. Moment (M) and thrust (T) develop across the thickness owing to distortion of the section due to W, as shown in the following sketch.
T
Figure 9-6.5
M
Forces due to vertical load
(9-6.11)
The expressions of M and T due to W are as follows. M = CM W R T = CT W Here, CM = moment coefficient CT = thrust coefficient W = vertical load per unit length R A = mean radius of pipe = area of longitudinal section for unit length of pipe = (1/6)t2 1000 mm3/m t = total thickness of core and coat
(9-6.12) (9-6.13)
Values of CM and CT are tabulated in IS:784 - 2001. The internal pressure is as follows.
fh 6 = pR At
(9-6.14)
fh 1 fh 2 + fh 3 + fh 4 + fh 5 fh 6
Figure 9-6.6
The construction of the tanks is in the following sequence. First, the concrete core is cast and cured. The surface is prepared by sand or hydro blasting. Next, the circumferential prestressing is applied by strand wrapping machine. applied to provide a coat of concrete over the prestressing strands. Shotcrete is
provides guidelines for the analysis and design of liquid storage tanks. sections of the code are titled as follows. Part 1: General Requirement Part 2: Reinforced Concrete Structures
The four
Analysis
The analysis of liquid storage tanks can be done by IS:3370 - 1967, Part 4, or by the finite element method. The Code provides coefficients for bending moment, shear and hoop tension (for cylindrical tanks), which were developed from the theory of plates and shells. In Part 4, both rectangular and cylindrical tanks are covered. Since circular prestressing is applicable to cylindrical tanks, only this type of tank is covered in this module. The following types of boundary conditions are considered in the analysis of the cylindrical wall. a) For base: fixed or hinged b) For top: free or hinged or framed. The applicability of each boundary condition is explained next.
For base
Fixed: When the wall is built continuous with its footing, then the base can be considered to be fixed as the first approximation. Hinged: If the sub grade is susceptible to settlement, then a hinged base is a conservative assumption. Since the actual rotational restraint from the footing is somewhere in between fixed and hinged, a hinged base can be assumed. The base can be made sliding with appropriate polyvinyl chloride (PVC) water-stops for liquid tightness.
For top
Free: The top of the wall is considered free when there is no restraint in expansion. Hinged: When the top is connected to the roof slab by dowels for shear transfer, the boundary condition can be considered to be hinged.
Framed: When the top of the wall and the roof slab are made continuous with moment transfer, the top is considered to be framed. The hydrostatic pressure on the wall increases linearly from the top to the bottom of the liquid of maximum possible depth. If the vapour pressure in the free board is negligible, then the pressure at the top is zero. Else, it is added to the pressure of the liquid throughout the depth. The forces generated in the tank due to circumferential prestress are opposite in nature to that due to hydrostatic pressure. underground, then the earth pressure needs to be considered. The hoop tension in the wall, generated due to a triangular hydrostatic pressure is given as follows. T = CT w H Ri The bending moment in the vertical direction is given as follows. M = CM w H3 The shear at the base is given by the following expression. V = CV w H2 In the previous equations, the notations used are as follows. CT = coefficient for hoop tension CM = coefficient for bending moment CV = coefficient for shear w = unit weight of liquid H = height of the liquid Ri = inner radius of the wall. The values of the coefficients are tabulated in IS:3370 - 1967, Part 4, for various values of H2/Dt, at different depths of the liquid. D and t represent the inner diameter and the thickness of the wall, respectively. The typical variations of CT and CM with depth, for two sets of boundary conditions are illustrated.
(9-6.17) (9-6.16) (9-6.15)
CT
(a) Fixed base free top
CM
CT
(b) Hinged base free top
Figure 9-6.8
CM
The roof can be made of a dome supported at the edges on the cylindrical wall. Else, the roof can be a flat slab supported on columns along with the edges. IS:3370 - 1967,
Part 4, provides coefficients for the analysis of the floor and roof slabs.
Design IS:3370 - 1967, Part 3, provides design requirements for prestressed tanks. A few of
them are mentioned. 1) The computed stress in the concrete and steel, during transfer, handling and construction, and under working loads, should be within the permissible values as specified in IS:1343 - 1980. 2) The liquid retaining face should be checked against cracking with a load factor of 1.2.
CL/WL 1.2
Here,
(9-6.18)
Values of limiting tensile strength of concrete for estimating the cracking load are specified in the Code. 3) The ultimate load at failure should not be less than twice the working load. 4) When the tank is full, there should be compression in the concrete at all points of at least 0.7 N/mm2. When the tank is empty, there should not be tensile stress greater than 1.0 N/mm2. Thus, the tank should be analysed both for the full and empty conditions. 5) There should be provisions to allow for elastic distortion of the structure during prestressing. Any restraint that may lead to the reduction of the prestressing force, should be considered.
Detailing Requirements IS:3370 - 1967, Part 3, also provides detailing requirements. The cover requirement is
as follows. The minimum cover to the prestressing wires should be 35 mm on the liquid face. For faces away from the liquid, the cover requirements are as per IS:1343 - 1980. Other requirements from IS:1343 - 1980 are also applicable.
Dome
Ring beam
Cylindrical wall
Raft foundation
Figure 9-6.9
Figure 9-6.11
Containment Structure, Kaiga Atomic Power Plant, Karnataka (Reference: Larsen & Toubro)
Analysis
The analysis of a ring beam is based on a load symmetric about the vertical axis. Since the dome is not supposed to carry any moment at the edge, the resultant reaction at the ring beam is tangential. The following figure shows the forces at the base of dome.
H V Rd sin Rd
Figure 9-6.12
Let the total vertical load from the dome be W. The vertical reaction per unit length (V) is given as follows.
V=
W 2Rd sin
(9-6.19)
Here, Rd = radius of the dome = half of the angle subtended by the dome. The horizontal thrust (H) is calculated from the condition of the reaction to be tangential. The value per unit length is given as follows.
H = V cot W cot = 2Rd sin
(9-6.20)
The thrust is resisted by the effective prestressing force (Pe) in the ring beam. Pe can be estimated from the equilibrium of half of the ring beam as shown in the following sketch.
H Pe Rd sin
Forces in the ring beam
Pe = H Rd sin = W cot 2
(9-6.21)
9.6.6 Conclusion
Prestressing of concrete is observed in other types of structural elements, such as bridge decks, shells and folded plates, offshore concrete gravity structures hydraulic structures, pavements and raft foundations. The analysis of special structures is based on advanced theory of structural analysis or the finite element method. After the analysis, the design of such structures follows the basic principles of prestressed concrete design. It is expected that in future, further innovations in structural form, prestressing systems and construction technology will promote the application of prestressed concrete.
Figure 9-6.14
Figure 9-6.15
Folded plate
Figure 9-6.17