Você está na página 1de 7

Int. J. Mech. Sci. Vol. 29, No. 10/11.pp.

713-719, 1987

Printed in Great Britain.

0020 7403/87 $3.00+.00 ~31987 PergamonJournals Ltd.

THE

CONTACT PRESSURE DISTRIBUTION DURING COMPRESSION OF LEAD SPHERES S. P. TIMOTHY,* J. M. PEARSON t and I. M. HUTCH~NGS

PLASTIC

Department of Materials Science and Metallurgy, University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, U.K.
(Received 16 January 1987; and in rer,ised jbrm 25 June 1987)

Abstract--Pressure-sensitive film has been used to examine the form of the contact pressure distribution for lead spheres plastically compressed between hard steel platens. The mean contact pressure attained a maximum value of 0.6-0.7 times the Vickers hardness of the undeformed sphere when the sphere had been deformed to 0.25 ~ a/R ~, 0.30, where a is the contact radius and R is the radius of the undeformed sphere. The form of the pressure distribution changed markedly as the contact area increased in size, with an increasingly larger proportion of the load being supported towards the contact perimeter. The results are compared with theoretical pressure distributions proposed by Matthews [Acta Metall. 28, 311-318 (1980)] tor contact between spheres. Fair agreement is obtained if the sphere is assumed to detorm by power-law creep, which is thought to reflect the observed time-dependent plastic detormation of the lead.

I. I N T R O D U C T I O N

The mean pressure required to deform a metal sphere plastically between hard, smooth parallel platens has been shown to vary with the amount of deformation in a manner which is characteristic of the initial degree of work hardening in the undeformed sphere [1, 2]. This geometry is relevant to studies of metalworking processes such as forging, the deformation of asperities (often assumed spherical) during contact between solid bodies, and the evaluation of the strength properties of metal powders [2]. Previous work was restricted to measurements of overall loads and geometry changes, and is extended here by using a commercially available pressure-sensitive film to determine the form of the pressure distribution over the contact area.
2. E X P E R I M E N T A L METHOD

A recently developed pressure-sensitive polymeric film, Fuji Prescale Film (Dracard Ltd, Maidstone, Kent, U.K.) was used to measure the contact pressures. The film is used as a double layer of two films, designated A-film and C-film. The A-film is available in several types covering different ranges of pressure sensitivity. In the present work, 'high-pressure' Afilm was used, which registers pressures in the range 25-70 MPa. When sheets of A-film and C-film are placed in contact and subjected to normal pressure, a red colour develops on the initially white C-film, the density of which increases with the pressure. By suitable calibration, a correlation can be established between the colour density and the applied pressure. The thickness of the double layer of film is 200 ~m. The maximum pressure that could be measured with the film was low compared with the plastic flow stresses of most metals. However, the indentation hardness of lead lies within the range of measurements covered by the film, and so lead was used for these experiments. Lead spheres (0.03 ~o Cu), 50.8 mm in diameter, were produced by sand-casting in a suitable mould. The risers and runners were removed from the castings, and the surfaces of the spheres were ground to a smooth finish with fine emery paper. One sphere was sectioned along its midplane and no significant porosity was evident. Its mean Vickers hardness was 59 __+2 MPa. The mean Vickers hardness of the two spheres deformed by compression was found,

* Present address: Alcan International Ltd, Banbury Laboratories, Banbury, Oxfordshire OX16 7SP, U.K. t Present address: Central Electricity Generating Board, North West Region, Scientific Services Department, George Street, Macclesfield, Cheshire, U.K. 713

714

S.P. TIMOTHY et al.

however, to be 2o"'v/o lower; this variation in mean hardness between different spheres probably results from small differences in the as-cast microstructures. The spheres were compressed in a series of incremental steps between quenched and tempered tool steel platens which had been ground and polished to 1 #m diamond finish, in a servo-hydraulic testing machine. The loading configuration is illustrated in Fig. 1. Experiments were performed under nominally unlubricated conditions, and also with a thin layer of molybdenum disulphide grease between the sphere and the film. The load on the sphere was increased gradually over 2 min and then maintained constant for a further 2 rain, as recommended by the manufacturers of the pressure-sensitive film. The sphere was then unloaded and removed together with the film; the diameters of the flattened contact surfaces of the sphere were measured to within 0.5 mm, and the coloured image developed on the Cfilm was photographed on to colour transparency film for long-term preservation of the results, since the image may become degraded more than two hours alter loading. The contact surfaces of the unloaded spheres were found to be plane. The sphere was then reloaded to a slightly higher load than before against new sheets of pressure-sensitive film. The colour transparencies of the films and of a reference chart relating pressure to colour density, supplied by the film manufacturer, were digitized with a scanning densitometer. The distributions of pressure along the radii of the contact circles were then computed from the radial colour-density profiles, after these had been averaged circumferentially.
3. R E S U L T S

The mean contact pressure P (calculated from the applied load and the measured contact area of the deformed spheres) is plotted against aiR for lubricated and unlubricated spheres in Fig. 2, where a is the mean contact radius and R is the radius of the undeformed sphere. The spheres were observed to creep during the period of constant load maintained at the end of each deformation increment, so that the contact radius expanded by a small amount during this time. Vickers hardness measurements on the final flattened sphere surfaces revealed no systematic variation in hardness across the contact diameters. The mean Vickers hardness of these surfaces was found to be 48 _+ 2 MPa for the lubricated sphere (final a/R ratio = 1.12) and 50 + 1 MPa for the unlubricated sphere (final a/R ratio = 1.23). Images developed in the pressure-sensitive film at different stages of deformation of the unlubricated sphere, together with the corresponding pressure profiles, are shown in Figs 3 and 4. The forms of the pressure profiles for the lubricated spheres were quite similar to these. The maximum contact pressure Pmax registered by the film at each stage of compression is compared with the mean contact pressure P in Fig. 2. It was found that the total load acting over the contact area, as calculated by integrating the pressure profiles registered by the film, was always greater than the measured load for a/R ~, 0.8, whereas for aiR ~ 0.8 it became less (Fig. 5).

bad

V//////////////////////////A

A filmI~I[
2a

film

F//////////////////////////////~

t
FIG. 1. The spheres were compressed between hard plane parallel platens, with the pressure-sensitive film interposed at both contact points.

Contact pressure distribution of lead spheres

715

Lubricated
/ ~an

Pmax

60

8. z+O
~ ~ o ~

o 20

02

04

06 air

08

10

1-2

unlubricated :~ 60
/ m / ~ ~ m ' ~ - m \ Pmax

~m ' ' ~

m m

n ~ um'~,..,.~

"E 8 20

O f

~ 0 ~ " -""-~0 ~

Po

/ /o
l I I I I I I 1 I I I I

02

04

a/R

06

08

10

12

FIG. 2. Variation with dimensionless contact radius, a/R, of the mean contact pressure, P, calculated from the applied load and the measured value of a and of the maximum contact pressure, Pmax, registered by the pressure-sensitive film. (a) Lubricated contact; and (b) unlubricated.

4. D I S C U S S I O N

The variation of P with a/R (Fig. 2) was similar for both the lubricated and unlubricated spheres, increasing to a maximum mean contact pressure of ~ 33 MPa (0.6--0.7 times the Vickers hardness of the cast lead) at 0.25 ~ a/R ~, 0.30, and then diminishing with increasing a/R. This behaviour is close to that reported by Chaudhri et al. [1] for the compression of brass and phosphor bronze spheres which were capable of work-hardening by a small amount. Unlike those materials, however, the lead spheres used in the present work continued to deform under constant applied load. The low effective rate of work-hardening of the lead implied by Fig. 2 may be attributed to dynamic recovery (or possibly recrystallization) during the loading cycle of deformation, which would tend to counteract any strain hardening due to crystallographic slip in the plastic zone at the relatively high homologous temperature of these experiments (about 0.5 times the melting temperature); when the loading cycle is completed, deformation continues further under constant load by creep. The creep behaviour observed under these conditions is in agreement with earlier work by Mulhearn and Tabor [3] and Atkins et al. [-4], who found that the indentation hardness of lead at room temperature decreased with increasing time of loading. It should be noted from Figs 2 and 4 that the normal stresses at the interface were often high; both the mean and local contact pressures were in some places greater than the uniaxial yield stress of the lead (assumed to be approximately one third of its Vickers indentation hardness, i.e. 16--20 MPa).

716

S.P. TIMOTHYet al.

a)
0

b)

a/R - 0.13

0"2

c)

d)

0 70

0"93

e)

1"23

50 mm

FIG. 3. Typical colour density patterns registered by the pressure-sensitive film at various stages of compression. Similar behaviour was observed under lubricated and unlubricated conditions.

It can be seen from Figs 3 and 4 that the distribution o f the contact pressure registered by the pressure-sensitive film changed significantly as the a m o u n t o f compression was increased. At low a/R, the pressure is distributed relatively uniformly, but the region away from the centre o f the contact area supports an increasingly larger p r o p o r t i o n o f the applied load as the sphere is compressed further. The pressure on the axis first tends to zero at a/R ~. 0.7; this

Contact pressure distribution of lead spheres

717

~a)

aiR=0-13

:ir

radial distance

I a'R:070

FIG.4, The radial pressuredistributions (solidlines)derived from the densitymaps shown in Fig. 3, In each case the colour density pattern was digitized and translated into a two-dimensional pressure distribution. The pressure values plotted here represent the circumferential average of" that distribution for each valueof radius. The mean pressure over the wholecontact area, derived from the measured pressure distribution, is also shown in each case (broken line).

W" --1 W

e-o
"

02

04.

06

0"8 a/R

1.0

12

FIG. 5. Variation in the load as calculated from the pressure-sensitive film, W', with a/R for unlubricated contact; W is the measured load.

very low pressure region expands in size as d e f o r m a t i o n proceeds. The form o f the pressure d i s t r i b u t i o n does n o t depend strongly o n the extent of lubrication at the sphere-film interface; the insensitivity o f the m e a n contact pressure to the presence of a lubricant o n the contact surface (Fig. 2) has also been reported previously [1]. Agreement between the actual applied loads a n d those determined by integrating the pressure profiles derived from the film was, however, poor, due to a s s u m p t i o n s made a n d to several errors a n d uncertainties in the experimental techniques, the influence o f which was difficult to assess quantitatively. F o r example, it was assumed that the pressure d i s t r i b u t i o n
M~ 29:10/II-D

718

S.P. TIMOTHY et a/.

did not change significantly with increments of aiR during both the loading cycle and constant load periods of each test, the influence of shear tractions at the contact was unknown, estimation of pressures below 25 MPa was only approximate since the film calibration was uncertain in that range, fluctuations occurred in the background density or' the photographic film, pressure gradients undoubtedly existed over distances small compared with the resolution limit of the film, and the pressures close to the contact edge were likely to be underestimated since this region was exposed to the pressure for times less than the full exposure time required by the Prescale film. This last limitation to the experimental method may be important in explaining why the total load calculated by integrating the pressure profiles was less than the applied load tbr a/R ~, 0.8 (Fig. 5), and why Pmax (Fig. 21 decreased from its maximum value beyond the interval 0.6 ~ a/R ~, 0.8, since a large proportion of the applied load was supported very close to the contact perimeter (see Fig. 4). At low a/R, it may be that shear stresses generated at the sphere-film interface during expansion of the contact edge influence the response of the film; the calculated normal loads would theretbre always be larger than the true loads since their effect would not be compensated for in the integration. Despite these shortcomings, the technique of using a pressure-sensitive film does appear to provide useful qualitative information on the contact pressure distribution which would be difficult to obtain in other ways. The experimental results in Fig. 4 may be compared to the theoretical pressure distributions for contact between spheres proposed by Matthews [5] and discussed by Johnson [6], For materials which exhibit power-law work-hardening during uniaxial

work- hardening

n:2 n=5 / ,. n = t~ 1 0 ~ ~

n:oe

v (3_

02

O-Z+

0-6

018

1-0

r/a

power taw creep

m=~

m=5 m t-~ = 2

f
I I I I I I

m=l

0-2

04

r/a

0"6

0"8

10

FIG. 6. Variation in radial pressure distribution P(r/a) with r/a (Matthews, 1980) for (a) work.hardening [equation (1)] and (b) power-law creep [equation 12)].

Contact pressure distribution of lead spheres

719

deformation, the contact pressure P(r) is given by, P(r) - (2n + 1) (1 - rE/a 2) 1/2n/5, 2n (1)

where P is the mean contact pressure, a is the contact radius, r is the radial distance from the centre and n is the reciprocal of the work-hardening index. For materials which display power-law creep, P(r, t) - ( 2 m - 1) (1 - r 2 / a 2 ) 2m
- 1/2rav(t),

(2)

where m is the creep exponent relating uniaxial strain rate to stress, and t is the time of loading. The pressure distributions given by equations (l)and (2) are quite different, as shown in Fig. 6, depending strongly on the constitutive behaviour of the material. The experimentally determined pressure distributions resemble more closely the family of curves given by equation (2) [Fig. 6(b)], although the pressure rises to a maximum value near the contact perimeter rather than tending to infinity. The general correlation between the experimental results in Fig. 4 and the curves in Fig. 6(b), rather than those in Fig. 6(a), is consistent with the work described above, indicating that the lead is deforming at room temperature in a timedependent manner, although its actual behaviour may not be accurately described by powerlaw creep.
5. C O N C L U S I O N S

The form of the contact pressure distribution for lead spheres plastically compressed between hard plane parallel platens has been investigated by using pressure-sensitive film. It was found that the variation in both the mean contact pressure, and the form of the pressure distribution, with dimensionless contact radius aiR was relatively insensitive to the degree of lubrication at the sphere-film interface. The mean contact pressure reached a maximum value of ~ 0.6-0.7 times the Vickers hardness of the undeformed spheres at 0.25 ~ a i r ,~ 0.30. A region of low pressure was observed at the centre of the contact area which increased in size as the deformation progressed, and thus the proportion of the applied load supported towards the contact perimeter also increased. The general form of the experimentally determined pressure distributions is similar to that proposed by Matthews [5] for contact deformation of a sphere undergoing power-law creep.
REFERENCES 1. 2. 3. 4. 5. 6. M. M. CHAUDHRI, I. M. HUTCHINGS and P. L. MAKm, Phil. Mag. 49A, 493-503 (1984). M. M. CHAUDHRIand I. M. HurcslrqGS, J. Mater. Sci. Lett. 3, 79-82 (1984). T. O. MULHEARNand D. TAaOR, J. Inst. Metals 89, 7-12 (1960--1961). A. G. ATKINS, A. SILVERIOand D. TABOR, J. Inst. Metals 94, 369-378 (1966). J. R. MATTHEWS, Acta Metall. 28, 311-318 (1980). K. L. JOHNSON,Contact Mechanics, pp. 199-201. Cambridge University Press (1985).

Você também pode gostar