Ace 8095 said:
I calculated the integrals and it was the correct answer, however I don't understand the concept of breaking an inequality into two integrals. Could you or someone else explain when this is necessary.
Sure, I'll try to explain. This problem is a little tricky.
Suppose we were told to perform a similar calculation, but this time in 2D, so just one variable. Let's say we had to solve the following:
Calculate the area under the graph y = 3x from 0 to 1, subject to the constraint 0 < y < 1. Draw this out on a piece of paper and you end up with something like the following, where the shaded green area is the area you need to compute.
We can calculate the area using basic geometry of a trapezoid, but we can also do it by treating it as 2 separate integrals. The first integral is from x=0 to x=1/3 of the line y=3x. The second integral is from x=1/3 to x=1 of the line y=1. If you correlate these two integrals to the graph you've drawn, you will notice that they cover the entire area of the shaded region. Since integral is just area represented, you know this must work.
So now how to apply this to your case? It's hard to draw on a sheet of paper since it's in 3D but the concept is the same. The first inequality tells you that 0 < y2 < y1. The other tells you that y2 < 1/3 and y1 < 1/2.
If you were to just integrate this normally (as you originally did), there would be two problems:
First, y2 would be bounded only by the number 1/3. In the 2D world, this is like taking the area of a rectangle. When you're doing multiple integrations like this, ALL possible combinations of the 2 ranges become valid. So if you have a double integral over the range [0,1/2] [0,1/3] then every possible pair of numbers in those 2 ranges are factored into the result. 0,0 is one possibility, so is 1/2,1/3, and 0,1/3, and 1/2,0, and everything else. What you need is a way to say "as I am integrating over the first range, limit the the second range to whatever number is currently the value of the function in the first range. It seems confusing in words but it's easier to visualize.
The second problem is that, if you were to just solve the first problem described above, you would still not be integrating over the correct range. If your only condition is that 0<y2<y1 then since y1 goes up to 1/2 you would be letting y2 run all the way up to 1/2 once you get near the end of the y1 interval. But we only want y2 going up to 1/3. The way to solve this is by breaking it into a separate interval. Just like in the 2D example I gave originally, there are 2 different conditions.
1) As long as y1 <= 1/3, then we can integrate directly over the range 0<y2<y1, there will never be a problem. All the inequalities will be satisfied. But as soon as y1 goes over 1/3, we'll have a problem because then so will y2. So we need to stop the first integral at y1=1/3
2) That means we still haven't integrated anything in the range y1=1/3 to y1=1/2. The integral that says that 0<y2<y1 we don't even need to look at for this part of the integral, because it's ALWAYS true, since the smallest y1 can even be is 1/3, and the max that y2 can be is 1/3. So in this case we want to integrate y2 from 0 to 1/3.
This leads to the 2 integrals I gave originally.