I've posted topics asking how to come up with the proper print temperature and read many posts as well. All that came back or what I read was "do it by trial and error", no one really gave any direction. The more I thought about it, the correct print temperature has to result in the highest bond strength.
So I developed this... The base is 25mm x 5mm, spindle is 10mm x 40mm.
NOTE: This might be obvious to some, but pay attention to the print quality as well as the strength. I just tried my test pieces with some filament from a2aprinter and the best bond was 245C, but the print quality was garbage. So I analysed the pieces (top surface) and picked the strongest piece with the best print quality. This roll has some air bubbles in it as well which would affect the bond strength and the print quality.
UPDATE: Green filament is from Repraper (Voxel Factory in Canada). There are 2 rows of 215c, 220c, 225c. The second row is with a FAN cover installed UNDER the headsinks/fans on the Replicator 1. These pieces were the strongest pieces I've ever made!! I could barely break them. However I'm having trouble printing the amazing marble run (http://www.thingiverse.com/thing:30166) with the FAN cover installed. The bottom of the tracks where the magnet is inserted curls up - the slide portion is fine. I'm going to try reducing my HBP heat from 115C to 105C. To all those that say the FAN cover "might make a difference", I can tell you that the bonding seems about 10x the strength. So much so I was about to get some pilers to break the temp calibration pieces!
** Make sure your platform is level, your filament diameter is correct as all these factors can contribute to bond strength.
Print with 10% and 1 extra shell (2 in total). Print at the temperature you think is close to the best. I used 4 settings 230C,235C,240C,245C. Make sure you label them (I labelled base & spindle).
As pointed out by theverant in the comments. Speed also affects the bonding of the plastic my test was done at 80mm/s. Higher/Lower will change the print temperature slightly, I'm not sure by how much I haven't experimented with that yet.
Once all your desired pieces are printed, grasp the base & spindle and break off the spindle, the result was shocking to me! I was expecting them all to look like 230C (as that's how my objects have been since I got the replicator).
The end result 235C was the best bond, it actually ripped off the 3 layers of infill from the base and produced a clearly visible hole! As you can see 245C was going downhill again in bond strength.
I believe I'll have to do this for Black, White and coloured plastics.
Just wanted to add...
The amount of force required to break the pieces changes as well (which is expected), so even if the end result is similar. Try and gauge the required force as you break the pieces. Unless you own a force gauge :)