I am a hobbyist and am hoping someone can help with a question on tube bending. 
This is a Harbor Freight bender and I'm making up dies for the radius of bend I need. Although it bends fine, I'm finding that when bending (steel) at room temp, The radius I end up with is larger than the die (due to spring back effects). The tubing is 1/2 x 1", .062" wall and I'm bending in the 'easy' direction.
The die OD is 5" which I was hoping would give me a 2.5" ID radius on the 90 degree bend. Instead, I'm getting around 3" radius.
My question is, how much smaller does the die diameter need to be so that I'll end up with 2.5" ID radius on the tube? I've looked around on-line and can't find any free calculators that address this and I'm not going to invest in a fancy calculator. Anyone have a rule of thumb here that might give me some idea what diameter to machine the die to? I made some smaller radius bends where I heated the tube up to orange hot before bending and did not have this problem. I tried to do that with the larger bend radius but could not heat the tube uniformly enough over the length and bad bends resulted.

This is a Harbor Freight bender and I'm making up dies for the radius of bend I need. Although it bends fine, I'm finding that when bending (steel) at room temp, The radius I end up with is larger than the die (due to spring back effects). The tubing is 1/2 x 1", .062" wall and I'm bending in the 'easy' direction.
The die OD is 5" which I was hoping would give me a 2.5" ID radius on the 90 degree bend. Instead, I'm getting around 3" radius.
My question is, how much smaller does the die diameter need to be so that I'll end up with 2.5" ID radius on the tube? I've looked around on-line and can't find any free calculators that address this and I'm not going to invest in a fancy calculator. Anyone have a rule of thumb here that might give me some idea what diameter to machine the die to? I made some smaller radius bends where I heated the tube up to orange hot before bending and did not have this problem. I tried to do that with the larger bend radius but could not heat the tube uniformly enough over the length and bad bends resulted.