Você está na página 1de 16

1

AER 504: Aerodynamics Crash Course in Vector Calculus


Vector Calculus is used extensively in advanced fluid dynamics and Aerodynamics. This summary gives a brief overview of some of the topics in this field of mathematics that applies directly to Aerodynamics. The summary begins with a review of partial differentiation and its applications along with vector notation and basic vector operations. Topics related to vector differentiation and multiple integration and applications are then presented. All figures are taken from reference 1. Review: Vector Notation Points in three dimensional Cartesian space are defined by three coordinates defined in terms of lengths along the Cartesian axis x, y, z. A location of a point is defined in terms of distance from the origin, P(x,y,z). The distance from the origin to a point P is given by, = || = 2 + 2 + 2

A vector in three dimensional space is defined by scalars applied to unit vectors which are aligned with the positive coordinate axis. A unit vector is one with a magnitude or length of 1. A vector defining point P is written as (, , ) = + +

+ + = || 2 + 2 + 2 The Cartesian components of the unit vector form of P can be define as the direction Cosines of the vector, =

The scalars x, y, z are the lengths from the origin to the point P. The bold variables i, j, k refer to the reference vectors, shown at left, aligned with the Cartesian axis, x, y, z respectively. These reference vectors are unit vectors. All bold variables commonly refer to vectors. The vector P has a length or magnitude defined by OP above and points in the direction defined by a line joining the origin O and point P. The vector P can be converted to a unit vector,

The angle x refers to the angle between the line OP and the positive Cartesian x axis. = + +

2 + 2 + 2

2 + 2 + 2

2 + 2 + 2

Review: Vector Addition, Subtraction If vectors A and B are defined as A = ax i + ay j + az k and B = bx i + by j + bz k, then when the vectors are added and subtracted, the operation is carried out on each individual component, + = ( + ) + + + ( + ) = ( ) + + ( )

Geometrically, the addition and subtraction of A and B creates entirely new directions. Multiple vectors can be added or subtracted in an easy fashion. Note that the vector A-B will have the same magnitude as vector B-A, but will be in completely the opposite direction.

Vector Differentiation: Vector differentiation is conducted in a similar fashion to the differentiation of single variable functions. By example, consider a point particle moving in time, where time is defined by the variable t. At every point in time the location of the particle is defined by the Cartesian coordinates, x, y, z. These coordinates can be viewed as functions of a single parameter, t, = (), = (), = ()

This defines the position of the particle at any time. The position vector r, can be used to locate the particle, = () = () + () + ()

The velocity of the particle can be found in a manner similar to single variable calculus, by taking the limit of the difference between positions at time t+t and t. = ( + ) () = =

This yields,

By considering that the derivative of the position vector was formed by taking the limit of the difference of r at two points in time, r, the derivative vector v, is tangent to the path traced out by the particle. In essence the velocity vector points in the direction that the particle is moving at one instance in time, t. The magnitude of v, gives the speed of the particle. = || = dy 2 dz 2 dx 2 + + dt dt dt

= + +

Scalar Products: The scalar product of two vectors a and b is defined as:

were is the angle between a and b. In effect, this product represents the magnitude of a multiplied by the projection of b onto a. Calculation of the scalar product for the two vectors is done via, = + +

= ||||

Since cos() =cos(-), the roles of a and b can be reversed, a b = b a. Vector Products: The vector product of a and b is defined as: | | = ||||

were is the angle between a and b. In effect, this product represents the area of a parallelogram formed from the vectors a and b as shown. The product yields a vector in a direction normal to both a and b. Calculation of the vector product for the two vectors is done via, =

= + ( ) +

The derivative of the cross product is,

The Gradient of a Scalar Function: It is assumed that the reader is familiar with partial derivatives (eg. f/x) which represent rates of change of some function, f, in the direction x when f is a function of more than one variable, f=f(x,y,z,.). The differential, df, is defined as an estimate of the change in a function of more than one variable, = + + + + +

d d ( ) = + dt dt

The gradient of a scalar function f(x,y,z) is denoted as f, which is a vector, and is formed: =

Assume now that the function represents a surface in 3D space. If the function f is now written in the form: f(x,y) = zo, where zo is a constant, then this form provides the level curves of the surface. As an example, if the function represents the surface of a mountain, then the form f(x,y) = zo gives the curve that is the cross-section of the mountain at a constant height of zo. In other words, this form gives the contour of the mountain at a specific height. By definition, the value of f along a level curve will be fixed. Thus the differential, df, will be zero along these level curves. A vector tangent to a level curve is, = + or = +

Then, on a level curve,

Since neither f nor dr are identically zero, then the angle between these two vectors must always be 90 degrees; they must be perpendicular. Since dr is always tangent to the curve, then the direction of the gradient f must always be at right angles to it, or always pointing in the direction of the largest change in f.
f dr f f constant f

+ = = |||| cos = 0

Sample Calculation of a Scalar Field Gradient The gradient of a scalar field appears frequently in the analysis of fluid systems. The driving force behind all fluid motion is ultimately the change in pressure inside a moving fluid. An example of the calculation of a gradient, and some interpretation is given below. If the pressure in a fluid is given by,

P( x, y ) = a ( x 2) 2 b( y 2) 2 + 20

Find, a) the gradient of P b) the location of the largest pressure c) the gradient at (2,0), (1,1), and (0,2) d) the equation of a contour of constant pressure, P = 5. a) The gradient is

P P P i+ j+ k = 2a ( x 2)i 2b( y 2) j P = x y z
b) The location of the largest pressure is at the local maximum,

2a ( x 2) = 0
Which is at the point x = 2, y = 2.

2b( y 2) = 0

c) The gradient of pressure at the points (2,0), (1,1), and (0,2), P(2,0) = 2a (2 2)i 2b(0 2) j = 0i + 4bj P (1,1) = 2a (1 2)i 2b(1 2) j = 2ai + 2bj P(0,2) = 2a (0 2)i 2b(2 2) j = 4ai + 0 j All three of these vectors point upslope approximately towards the point of highest gradient at (2,2). It can be shown (math texts) that the gradient of a scalar field provides the local magnitude and direction of the largest increase of the scalar. It should be emphasized that since the scalar field is continuous, the gradient is also; meaning its magnitude and direction will vary continuously.

d) The equation of a contour of constant pressure is essentially the equation of a curve corresponding to a value of P equal to some fixed value. In this case we wish P = 5.

5 = a ( x 2) 2 b( y 2) 2 + 20
15 = a ( x 2) 2 + b( y 2) 2 If a = b, then we would obtain the equation of a circle; meaning the level contours of pressure would be circular curves centered at the point (2,2). If a is not equal to b, then the curves would be elliptical. The vectors computed in part c) would all be normal to the local contours of constant pressure show below.

The Directional Derivative: Given that the gradient of a scalar points in the direction of maximum increase of the scalar function, we can use this property do determine the rate of change of the scalar in any direction, s. An estimate of the change in a function f, with a change in an arbitrary direction s is given by, ( + , + ) (, ) = 2 + 2 = +

By the mean value theorem,

and since,

Taking the limit as s goes to 0,

The components cos() and sin() can be interpreted as the components of a unit vector pointing in the desired s direction, = cos() + sin()

cos() + sin() = lim = cos() + sin() 0

= cos()

= sin ()

Therefore the derivative of any scalar function in an arbitrary direction s, can be computed from,

where n is a unit vector pointing in the direction s.

= = cos() + sin()

Line Integrals: A number of applications within Aerodynamics require the calculation of the total influence of a component of a vector field along a given path. For example the work done by an arbitrary force vector F along a path C going from point Po to P1. To calculate the work first recall that a vector tangent to the curve C is, = + +

Therefore, the work done by the force moving from point Po to P1 is

If the force vector is given in the form:

The work calculation reduces to,


1

= 1 (, , ) + 2 (, , ) + 3 (, , ) = + 2 +

This form may not be easily used since the force components F1, F2, F3 must be evaluated along the given curve in terms of the appropriate variable. For example, the integration of F1 with respect to x will require the variables y, z in F1 to be substituted for relations only in x. An easier approach would be to write the force vector, the curve, and dr in terms of a single parameter t. Then, =
1 1

If the position vector is,

Then,

Example: If F = xyi + (y2 + 1)j, find the work done in moving from Po(0,0,0) to P1(1,1,0) when, i. the curve is a straight line ii. the curve is along the x axis to (1,0,0), then along x = 1 to P1 iii. the curve is y2 = x i. The curve is the line y = x. The most convenient method is to change variables to a single parameter. In this case the easiest choice: x = t, y = t. Therefore, t = 0 referes to point (0,0) and t = 1 refers to point (1,1). = 2 + ( 2 + 1) = +
1

1 = + 2 + 3 1

() = () + () + ()

( 2 + ( 2 + 1)) ( + ) = (2 2 + 1) = 53
1 0

ii.

Here, the curve is made of two straight line components. Referring to point (1,0) as point R, the first line is from (0,0) to R, then from R to (1,1). The integral is broken into = +
1 1

On the first segment, y = 0, x is the parameter and ranges from 0 to 1, and F = j and dr = dx j. = = 0
0 1 1

On the next segment, x = 1, and dr = dy j, F = yi+(y2+1)j, with y going from 0 to 1. = ( + ( 2 + 1))


0 0 1 1

iii. On this segment, y2= x, and the parameterization is x = t2, y = t, were t ranges from 0 to 1. On the line, = 2 +
0 1

= ( 2 + 1) = 43 = (2 + )

Double Integrals: Double integrals operate in a similar way to their 1D counterparts where the area under a curve y = f(x), was easily computed by integration along the x coordinate. In a double integral, the integrand, z = f(x,y), when integrated over an area R in the x-y plane, gives the volume under the surface z = f(x,y). To be more precise, consider the figure below. The surface represented by, z = f(x,y), sits above an area R in the x-y plane with vertical walls from R to z . In the 1D case, the area under the curve was found by taking the limit as the spacings x where reduced to an infinitesimally small value. In the double integral case, the area of region R, is found by taking the limit as small areas A = xy, are reduced in size as the limit is taken. In this case, both x and y shrink simultaneously as the limit is taken. If the area R is divided along lines of x = xi where i = 1,2,3,.m, and y = yj where j = 1,2,3, n, an approximation to the volume under the curve over the element Aij = xiyj is , = ( , )

= ( + ( 2 + 1)) (2 + ) = (2 4 + 2 + 1) = 2615
1 0 1

= + ( 2 + 1)

10

The exact volume is found when the limit of the above expression is computed as m and n go to infinity. The variables and are the values of x and y on the range, xi-1 < i < xi and yj-1 < j < yj , and xi=xi-1-xi , yj=yj-1-yj.
m,n

( , )
=1 =1

The calculation of a double integral is best performed by considering a succession of single variable integrals. For instance, we could consider performing an integration parallel to the x axis, indicated by the i summation index, while holding the variable y fixed. (see figure at left)
n

( , )
=1 =1

The term in brackets is equivalent to computing, (with y held fixed),


= ()

This yields a sequence of areas under the curve f from y=a to y=b. The next step in the integration is to integrate this sequence of areas in the y direction,
= =

= ()

(, )

The order of integration could have been reversed. In other words we could have integrated with respect to y first, which would yield the same solution. It should be noted that if the value of the function is f(x,y) = 1, the double integral will yield the area of the region R.

= ()

= ()

(, )

11

Example of Double Integration Find the volume common to two cylinders, x2 + y2 = a2 and x2 + z2 = a2 The first cylinder is centred on the z axis while the other is centred on the y axis. The volume that represents the intersection of the two cylinders is desired. The volume is symmetric with respect to all Cartesian coordinate planes, x-y, x-z, y-z. The volume shown in the figure when multiplied by 8 will give the total volume. In this problem, the function that represents the top of the volume will be chosen as, = 2 2

The projected region R will lie in the x-y plane. So, the total volume is found from, = 8 2 2
= = 2 2

Considering the region R as shown, the top bound of the region is = 2 2 . The lower bound is y = 0. Therefore, = 8
= =0

= 8

= 8
=0

(2 2 ) = 8 2

=0

=0

2 2

2 2
2 2 0

3 163 = 3 0 3

12

Triple Integrals: Triple integrals are performed in a manner similar to double integrals. The only difference is that an extra layer of integration is added making the result applicable over a volume and not an area. The concept of using a curve as the bounds for the first integration in a double integral, is extended such that in a triple integral, the first bounds are surfaces. As an example, consider the calculation to find the total mass inside a volume where the density of the material varies in space, (x,y,z). Following the indexing developed under double integrals, the volume is divided along lines of x = xi where i = 1,2,3,.m, y = yj where j = 1,2,3, n, and z = zk where k = 1,2,3, p. Here, xi=xi-1-xi , yj=yj-1-yj, zk=zk-1-zk. An approximation to the mass in the volume is, , , = ( , , )

The variable k denotes a value of z such that zk-1 < k < zk. The exact mass is found when the edge lengths xi yj zk are reduced to zero, =
m,n,p

( , , )
=1 =1 =1

The iterative method used on the double integral is again used for the triple integral. Here, we will integrate with respect to z first, followed by y then x. The limits in z are now defined, not by curves as a result of projection onto the x-y plane, but by the equations of surfaces. The top limit is z = T(x,y), the bottom limit is, z = B(x,y). Once integration is done in z, the result is a double integral which is evaluated in a manner identical to the previous section. Therefore, the complete integral expression is, =
= =

It should be noted that if instead of using the density as the integrand, we used a value of 1, we would have obtained the volume of the region using the triple integration. Also, the order of integration of the Cartesian variables could be changed, depending on the easiest configuration. If the mass inside a volume is denoted as, =

= ()

= ()

= (,)

= (,)

(, , )

13

Then the coordinates of the centre of mass of the volume are, = =

Example calculation, triple integral, Find the volume enclosed by the two surfaces: z = x2 + 3y2 and z = 8 x2 y2. The two surfaces intersect on the elliptic cylinder, 2 + 3 2 = 8 2 2 2 + 2 2 = 4

or

The volume projects onto a region in the x-y plane bounded by x2+2y2=4. In this projected region, the limits of the ellipse are 2 in x, and 21/2 in y. Take the first integration to be in z, the second in y, then in x. =
=2 =2 =2 =2

4 2 8 4 2 = 2(8 2 ) 3 2 2 =2
=2

=(4 2 )/2 =(4 2 )/2 2 =(4 2 )/2

=(4 2 )/2

(8 2 2 4 2 )
3 2

= +

Surface Integrals: Some fields of study may require the evaluation of a function over a 3D surface. For example the force on the surface of a wing due to surface pressure can be evaluate provided the pressure is known p = p(x,y,z), and the wing surface shape S is defined by z = f(x,y). The pressure on the wing surface is then p = p(x,y,f(x,y)), which must be integrated over S. Using the nomenclature of the double integral, this surface force calculation is then, (, , ) = ( , , )Sij Here, ij is the value of f when evaluated at the points i, j, or ij = f(i, j). The parameter Sij is an element of surface area on the surface S. The integration can be completed if the surface
m,n =1 =1

42 =2 (4 2 )32 = 82 3 =2

14

z=f(x,y)

Sij

element Sij can be equated to a projected element on the x-y plane Aij, or some other plane. When this is done, we can use the double integral as previously defined to do the analysis. First, consider a surface in 3D space represented by the equation z = f(x,y). The differential of this surface can be written, = + = + + = 0

(i,j)
X

xi Aij
x=xi

x=xi-1

or

yj
y=yj-1 y=yj

This equation can be viewed as the product of the vector = +

with a vector on the surface

or

= + + = 0

which is given by

This relation gives the angle of inclination between the surface dS and its corresponding projection dA, so that dA = cos() dS. = 1 + 2 + 2
1 2

cos =

1 = 1 + 2 + 2 2 ||||

Since the vector dr is always along the surface, then the only way the equation above is true is if the vector n is normal to the surface. Consider a small element on the surface dS, which has a projection onto the x-y plane called dA. On the element dS, the angle between n and the vertical vector k is ,

So long as the surface S is simple, this relation allows the surface integral to be evaluated as we had done for the double integral. A simple surface is one where a vertical line cuts through the surface at only one point. The surface integral for p above can be written,

= 1 + 2 + 2

1 2

15

In this equation the projected region onto the x-y plane is denoted as R. If the projection was onto the z-x plane instead (the surface defines as y = f(x,z), (, , ) = (, (, ), )(1 + 2 + 2 )
1 2

(, , ) = , , (, )1 + 2 + 2

1 2

Note that if the value of the integrand is 1, the surface integral gives the area of the surface. Example Surface Integral Find the area of the plane ax + by + cz +d =0, inside the cylinder x2 + y2 = r2. Using a surface integral with an integrand of 1, the surface is easily projected onto the x-y plane. Re-writing, = (, ) = ( + + )/ and = = 2 2 = = 1 + + 2 2 = 1 + +
1 2 1 2

Since the remaining integral is just the area of the projected region R, which is a circle of radius r, 2 2 1 [ + 2 + 2 ] 2 = The Curl of a Vector Field, The curl of a vector field V = ui + vj + wk, is defined as, = x y w v u w v u + = + y z z x x y z curl V = V,

If the vector V is the velocity of the fluid, the curl is twice the local rotation rate of the fluid.

16

The Divergence of the Vector Field, The divergence of a vector field V, is div V = V, = =

The divergence of a velocity field can be shown to be an expression of mass conservation. The Divergence theorem, The divergence theorem, which is also known as Gauss theorem, states that the divergence of a vector when integrated over a closed volume (V) is equal to the scalar product of the vector with the unit normal of the volume surface integrated over the volume surface (S), dV = Gauss theorem in the plane is an extension of the divergence theorem applied to a 2D surface A with a perimeter l, = The divergence theorem is used to simplify the control volume approach to conservation relation in fluid dynamics. The Gradient theorem, The gradient theorem, like the divergence theorem, allows simplification of the basic conservation relations of fluid dynamics. dV = However, in this case the theorem can be used to show that the pressure force action on a surface is equal to the gradient of the pressure integrated over the enclosed volume.

+ +

References 1. Greenspan, H.P., Benney, D.J., Calculus: An Introduction to Applied Mathematics, McGrawHill Co., New York, 1973.

Você também pode gostar