# Why do herringbone/double-helical gears have a stable and unstable configuration?

I understand that double-helical gears are used over regular helical gears to balance axial forces in a shaft but, depending on the direction, it seems that these balanced forces can be either compressive or tensile.

A quick Google search shows that many suppliers recommend the compressive configuration over the tensile one when installing these gears, because it is more "stable." However, most of the websites I have looked at have similar wording when saying this, and I have not yet found any source that is credible (academic textbooks, well-know suppliers, etc.). This makes me suspect that this is either not a very important consideration, or it is not true. From what I understand, if the forces are compressive, they will correct any axial misalignment, but I would think that tensile configuration would also correct the misalignment. In other words, I can't really picture the "unstable" scenario.

Moreover, I do not understand how both the pinion and the gear could be both in a "stable" configuration, since any compressive forces in the gear would have to cause a tensile force in the pinion. If one direction is truly "stable," which of the two (pinion or gear) should be stabilized?

If anybody has any credible online sources or intuitive explanations on this stability problem, that would be a great help

• what gears and gearboxes have you looked at ? Google is good but real machines may give you a better insight.... Go to a scrap yard and get a couple of old transmissions : take them apart and work from there... Nov 24 '17 at 21:56
• I'm actually trying to analyze the transmission system of a mill. There's a way to reverse the double-helical gears such that they work in either the "stable" or "unstable" configuration. I'm not sure that we installed the gears correctly. 1 out of our 4 mills have a transmission in the "stable" configuration, and the other 3 are in the "unstable" configuration. I'm not sure if that actually is a problem, but I can't take our setup as a reference because it's what I'm trying to analyze. Nov 24 '17 at 22:10