For a .002 interference, you'd probably want to expand it .004 to make sure you got it on. You only get one chance.
steel fas a coefficient of expansion of 6.7 x 10 to the -6, or .0000067 in/in deg f

So for initial bore of .666 in circumference is .666 x pi = 2.0923 (that size is the devil to machine)
For expanded di of .670, circumference is .670 x pi = 2.1049

delta cir =2.1049 - 2.0923 = 0.0126

for expansion, delta L /L = coef of expansion x delta temp

0.0126 / 2.0923 = .0000067 x delta T

or delta T = 900 degree farenhiet over room temperature

Should be doable with a propane torch.

Cooling the lead screw wouldn't do much, you might drop it from say 70 to freezer temp say 20 f or delta t of 50 It would help but would only shrink 0.00022 in dia.


Check the math but it should work.

This was a smaller bore so less expansion but only a 1 thou interference, I was afraid of breaking the gear when it cooled. At blue temperature on the gear it literally fell on to the shaft.