Change of variables examples

Example 1

Compute the double integral

$\displaystyle \iint_{D}^{}$g(x, y)dA    

where g(x, y) = x2 + y2 and D is disk of radius 7 centered at origin.

\includegraphics[width=2.in]{disk.eps}

Solution: Since computing this integral in rectangular coordinates is too difficult, we change to polar coordinates. (We chose polar coordinates since the disk is easily described in polar coordinates.) Let

T(r,$ \theta$) = (r cos$ \theta$, r sin$ \theta$).    

be the change of variables function from polar coordinates into rectangular coordinates.

Then, by the change of variables formula, the integral becomes

$\displaystyle \iint_{D}^{}$g(x, y)dA = $\displaystyle \iint_{{D^{\textstyle *}}}^{}$g(T(r,$ \theta$))$\displaystyle \left\vert\vphantom{ \det D{\mathbf{T}}(r,{\textstyle \theta})}\right.$det DT(r,$ \theta$)$\displaystyle \left.\vphantom{ \det D{\mathbf{T}}(r,{\textstyle \theta})}\right\vert$dr d$ \theta$,    

where D* is the region in the r$ \theta$-plane that is mapped by T onto the disk D. In polar coordinates, the disk is described by 0$ \le$r$ \le$7 and 0$ \le$$ \theta$$ \le$2$ \pi$, so this is the region D*.

We calculate

g(T(r,$ \theta$)) = g(r cos$ \theta$, r sin$ \theta$) = r2cos2$ \theta$ + r2sin2$ \theta$ = r2    
$\displaystyle {\frac{{\partial T_1}}{{\partial r}}}$(r,$ \theta$) = cos$ \theta$    
$\displaystyle {\frac{{\partial T_2}}{{\partial r}}}$(r,$ \theta$) = sin$ \theta$    
$\displaystyle {\frac{{\partial T_1}}{{\partial \theta}}}$(r,$ \theta$) = - r sin$ \theta$    
$\displaystyle {\frac{{\partial T_2}}{{\partial \theta}}}$(r,$ \theta$) = r cos$ \theta$    
DT(r,$ \theta$) = $\displaystyle \left\vert\vphantom{ \begin{array}{cc} \cos\theta & \sin\theta -r\sin \theta & r \cos\theta \end{array} }\right.$$\displaystyle \begin{array}{cc} \cos\theta & \sin\theta -r\sin \theta & r \cos\theta \end{array}$$\displaystyle \left.\vphantom{ \begin{array}{cc} \cos\theta & \sin\theta -r\sin \theta & r \cos\theta \end{array} }\right\vert$    
det DT(r,$ \theta$) = r cos2$ \theta$ - (- r)sin2$ \theta$ = r(cos2$ \theta$ + sin2$ \theta$) = r    

Since r$ \ge$ 0 in D*, we know that $ \left\vert\vphantom{ \det
D{\mathbf{T}}(r,{\textstyle \theta})}\right.$det DT(r,$ \theta$)$ \left.\vphantom{ \det
D{\mathbf{T}}(r,{\textstyle \theta})}\right\vert$ = | r| = r. Therefore the integral is

$\displaystyle \iint_{D}^{}$g(x, y)dA = $\displaystyle \int_{0}^{{2\pi}}$$\displaystyle \int_{0}^{7}$r2rdr d$ \theta$    
  = $\displaystyle \int_{2}^{{2\pi}}$$\displaystyle {\frac{{r^4}}{{4}}}$$\displaystyle \bigg\vert _{0}^{7}$d$ \theta$    
  = $\displaystyle \int_{2}^{{2\pi}}$$\displaystyle {\frac{{7^4}}{{4}}}$d$ \theta$ = $\displaystyle {\frac{{\pi 7^4}}{{2}}}$    

Note: In case you have trouble believing the formula that dA becomes $ \left\vert\vphantom{ \det
D{\mathbf{T}}(r,{\textstyle \theta})}\right.$det DT(r,$ \theta$)$ \left.\vphantom{ \det
D{\mathbf{T}}(r,{\textstyle \theta})}\right\vert$dr, d$ \theta$, we could motivate the result that, for polar coordinates, dA becomes r dr d$ \theta$ as follows.

Chop up the disk into grid corresponding to a grid size in polar coordinates of $ \Delta$r and $ \Delta$$ \theta$, just as we did for the intro reading. A single "curvy rectangle" with a corner (r,$ \theta$) is pictured below.

\includegraphics[width=2in]{polargrid.eps}

Let $ \Delta$A be area of the "curvy rectangle". You can see that the area depends on the radius r, since its width is approximately r$ \Delta$$ \theta$ and its height is approximately $ \Delta$r. Hence the area $ \Delta$A is approximately r$ \Delta$r$ \Delta$$ \theta$. I hope that now you can believe that dA becomes r dr d$ \theta$ when changing variables to polar coordinates.

Example 2

Evaluate

$\displaystyle \iint_{D}^{}$(x2 - y2) dx dy    

where D is the region pictured below.
\includegraphics[width=2.5in]{changevarex.eps}

Solution:

The region can be described simply if we change to coordinates u and v where

u = y - x    
v = xy. (1)

With this change of variables, our new region of integration D* is 0$ \le$u$ \le$1, 1$ \le$v$ \le$2.

Our change of variables as expressed in equation (1) give u and v in terms of x and y. In our change of variables formula, we need to have x and y expressed in terms of u and v using some function (x, y) = T(u, v). So one way to solve this problem is to solve equation (1) for x and y to determine the function T. We could then compute

$\displaystyle {\frac{{\partial (x,y)}}{{\partial (u,v)}}}$,    

which is what we need for the change of variables formula.

For this example, we will use another method that skips this step. We will leave u and v expressed in terms of x and y (as in equation (1)) and compute $\displaystyle {\frac{{\partial (u,v)}}{{\partial (x,y)}}}$ instead of $\displaystyle {\frac{{\partial (x,y)}}{{\partial (u,v)}}}$. Since we know that

$\displaystyle {\frac{{\partial (x,y)}}{{\partial (u,v)}}}$ = $\displaystyle {\frac{{1}}{{\displaystyle \frac{\partial (u,v)}{\partial (x,y)}}}}$,    

we can easily get the correct factor for out change of variables formula.

To compute $\displaystyle {\frac{{\partial (u,v)}}{{\partial (x,y)}}}$, we view u and v as functions of x and y: s(x, y) and t(x, y). The partial derivatives of u and v with respect to x and y are

$\displaystyle {\frac{{\partial u}}{{\partial x}}}$ = - 1,        $\displaystyle {\frac{{\partial u}}{{\partial y}}}$ = 1,        $\displaystyle {\frac{{\partial v}}{{\partial x}}}$ = y,        and        $\displaystyle {\frac{{\partial v}}{{\partial y}}}$ = x.    

We compute

$\displaystyle {\frac{{\partial (u,v)}}{{\partial (x,y)}}}$ = det$\displaystyle \left(\vphantom{ \begin{array}{cc} \displaystyle \frac{\partial u...
...{\partial x} & \displaystyle \frac{\partial v}{\partial y} \end{array} }\right.$$\displaystyle \begin{array}{cc} \displaystyle \frac{\partial u}{\partial x} & \...
...artial v}{\partial x} & \displaystyle \frac{\partial v}{\partial y} \end{array}$$\displaystyle \left.\vphantom{ \begin{array}{cc} \displaystyle \frac{\partial u...
...{\partial x} & \displaystyle \frac{\partial v}{\partial y} \end{array} }\right)$    
  = $\displaystyle {\frac{{\partial u}}{{\partial x}}}$$\displaystyle {\frac{{\partial v}}{{\partial y}}}$ - $\displaystyle {\frac{{\partial u}}{{\partial y}}}$$\displaystyle {\frac{{\partial v}}{{\partial x}}}$    
  = - x - y,    

so that

$\displaystyle {\frac{{\partial (x,y)}}{{\partial (u,v)}}}$ = $\displaystyle {\frac{{1}}{{-x -y}}}$.    

The factor we need for our change of variables formula is then

$\displaystyle \left\vert\vphantom{\frac{\partial (x,y)}{\partial (u,v)}}\right.$$\displaystyle {\frac{{\partial (x,y)}}{{\partial (u,v)}}}$$\displaystyle \left.\vphantom{\frac{\partial (x,y)}{\partial (u,v)}}\right\vert$ = $\displaystyle {\frac{{1}}{{x +y}}}$    

since x and y are positive in our domain.

Once we change variables, we will need to integrate over the region D* the expression

(x2 - y2)$\displaystyle {\frac{{1}}{{x +y}}}$.    

Since we can factor x2 - y2 = (x - y)(x + y) the above expression simplifies to x - y.

Since the integral over D* is an integral in u and v, we need to write x - y in terms of u and v. Glancing at equation (1), we see that x - y = - u, so we simply need to integrate - u over the region D*.

Our integral is

$\displaystyle \int_{D}^{}$(x2 - y2)dx dy = $\displaystyle \iint_{{D^{\textstyle *}}}^{}$ - u du dv    
  = - $\displaystyle \int_{1}^{2}$$\displaystyle \int_{0}^{1}$u du dv    
  = - $\displaystyle \int_{1}^{2}$$\displaystyle {\frac{{u^2}}{{2}}}$$\displaystyle \bigg\vert _{0}^{1}$dv    
  = - $\displaystyle \int_{1}^{2}$$\displaystyle {\frac{{1}}{{2}}}$dv = - $\displaystyle {\frac{{1}}{{2}}}$    

where x and y are functions of u and v.

Example 3

Evaluate

$\displaystyle \iint_{{D_c}}^{}$e-x2-y2dA    

where Dc is the region in the first quadrant of the xy-plane where x2 + y2 < c2.
\includegraphics[width=1.3in]{changevarex2.eps}

Solution:

Change to polar coordinates. Region is sector: 0$ \le$$ \theta$$ \le$$ \pi$/2 and 0$ \le$r$ \le$c.

$\displaystyle \iint_{{D_c}}^{}$e-x2-y2dA = $\displaystyle \int_{0}^{{\pi/2}}$$\displaystyle \int_{0}^{c}$e-r2r dr d$ \theta$    
  = $\displaystyle \int_{0}^{{\pi/2}}$ - $\displaystyle {\frac{{1}}{{2}}}$e-r2$\displaystyle \bigg\vert _{0}^{c}$d$ \theta$    
  = $\displaystyle {\frac{{1}}{{2}}}$$\displaystyle \left(\vphantom{1-e^{-c^2}}\right.$1 - e-c2$\displaystyle \left.\vphantom{1-e^{-c^2}}\right)$$\displaystyle \int_{0}^{{2/\pi}}$d$ \theta$    
  = $\displaystyle {\frac{{\pi}}{{4}}}$$\displaystyle \left(\vphantom{1-e^{-c^2}}\right.$1 - e-c2$\displaystyle \left.\vphantom{1-e^{-c^2}}\right)$    



Duane Nykamp
nykamp@math.umn.edu
2005-10-27