Find the bounds after changing the variables in a double integral

5 replies · 3K views · last reply
Leo Liu
Messages
353
Reaction score
156
Homework Statement
.
Relevant Equations
.
Homework problem 3D-4: evaluate the double integral of (2x − 3y)^2 (x + y)^2 dx dy over region R, the triangle bounded by the positive x-axis, negative y-axis and the line 2x − 3y = 4, using the substitution u = x + y, v = 2x − 3y.

The answer calculates the integral with ##du## before ##dv## as shown below.
Handwritten solution to a double integral change of variables: u = x + y, v = 2x - 3y, Jacobian -5, with the region R sketched in uv-coordinates between lines v=0, v=2u, v=-3u and v=4 and the integral evaluated in order du dv.

However I decided to compute it in the opposite order with different bounds. Here is my work:
According to the definitions, $$\begin{cases} u=x+y\\ v=2x-3y \end{cases}$$
First we need to convert the boundaries in xy into uv. The equation of y-axis is ##v=-3u## and the equation of x-axis is ##v=2u##. This two lines give the lower and upper limits of the inner integral dv. The bounds for u can be found by substituting ##x=0## and ##y=0## into ##2x-3y=4##. The lower bounds appears at the lower vertex of the triangle with the value of ##u=-4/3##. Similarly, the upper bounds is the right vertex at ##u=2##. After putting things together we get
$$\frac 1 5 \int _{-4/3}^{2}\int _{-3u}^{2u}v^2u^2\,dvdu$$
I'd like someone to check if the bounds in this integral are correct. Thanks in advance.
 
Hi,

If you did everything correctly, your integral would yield the same as what the alternative yields. I get a different result, though.

Draw some lines where ##u## is constant and check the bounds of ##v##

##\ ##
 
BvU said:
Hi,

If you did everything correctly, your integral would yield the same as what the alternative yields. I get a different result, though.

Draw some lines where ##u## is constant and check the bounds of ##v##
Thank you! I think I messed up the bounds for v.
Graph in the UV plane with axes from -4 to 4: a horizontal dashed blue line at v = 4, two lines through the origin (one descending steeply to the left, one ascending to the right), red dashed vertical segments from the v = 4 line down to those lines, and labels (-4/3, 4) and (2, 0).

The graph above shows that heights of the endpoints of the line segment spanning the height of the triangle at a fixed x value are ##v=4## and ##v=-3u## or ##v=2u##; the choice of the lower bound depends on x. Hence the double integral can be written as

$$\frac 1 5 \int _{-4/3}^{0}\int _{-3u}^{4}v^2u^2\,dvdu+\frac 1 5 \int _{0}^{2}\int _{2u}^{4}v^2u^2\,dvdu$$
Is this correct?
 
You can evaluate and check for yourself ...

##\ ##
 
BvU said:
You can evaluate and check for yourself ...

##\ ##
I think I got it right. I appreciate your help.
Mathematica notebook screenshot defining f = 1/5 v^2 u^2 with limits umin1 = -4/3, umax1 = 0, vmin1 = -3u, vmax1 = 4, umin2 = 0, umax2 = 2, vmin2 = 2u, vmax2 = 4, a double integral sum evaluated as 1792/243, and a second double integral over u from -v/3 to v/2 and v from 0 to 4 also giving 1792/243.
 
Looks ok. (didn't check it, but I see you have to do two integrals -- which is the reason the book solution prefers the other order)