A Pythagorean triad (or "Pythagorean triple") consists of three positive integers (A, B, C) such that:
A^2 + B^2 = C^2
For a Pythagorean triad (A, B, C), if you construct a triangle of sidelengths A, B and C, it will be a right triangle, and C will be its hypotenuse.
The 3-4-5 triangle is a well-known right triangle with integer-length sides, and also forms the smallest possible Pythagorean triad (3, 4, 5).
--
You could list the set of perfect squares {1, 4, 9, 16, 25, 36, ... } and then add pairs at random until you find a pair that adds up to a perfect square. But this would be tedious.
But, if we know some triad for (A, B, C) such that A^2 + B^2 = C^2, what if we multiply each side of the equation by some value n^2 ? Then we'd get:
n^2 A^2 + n^2 B^2 = n^2 C^2
(nA)^2 + (nB)^2 = (nC)^2
Now we have a new Pythagorean triad: (nA, nB, nC).
So, actually, this is the same as if we had simply multiplied A, B and C by n, (not necessarily n^2)--which means we have found a method of finding new triads using any value of n:
--
Starting with (3, 4, 5), try multiplying each value by some n ... say, 2.
This results in the triad (6, 8, 10). But is this a Pythagorean triad? Let's test it:
6^2 + 8^2
= 36 + 64
= 100
= 10^2
So we know that (6, 8, 10) is also a Pythagorean triad. This works for any value n. You can generate an unlimited number of Pythagorean triads simply by choosing different positive integer values for n, so this is probably sufficient for your purposes.
However, you can't find ALL Pythagorean triads this way.
For instance, consider (7, 24, 25)
7^2 + 24^2
= 49 + 576
= 625
= 25^2
... so it is a Pythagorean triple. But 7 is not a multiple of 3, so we could never find this particular triple using the above method.