If you mean stress concentration for torsion/torque, that is very localized to the corner of the keyway, so it should not matter how many or where, though obviously one could remove too much shaft. Consider the case of splines which are used to equalize the load over more of the shaft.

If there is enough clearance the two keys will share the torque and decrease the stress each sees. At worst case only one key will be carrying load.

Assume only one keyway carries 100% of the load. The load split on a double keyway will be nothing like 50/50.

I agree with rothers if the keyways have slight clearance fit then the load will be taken by one key at any one time, so assuming one key takes the load is conservative and any load sharing will be a bonus.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

Double key, no but I assume you can just double the shear area for a close enough calculation.

I used these guys calcs recently. https://www.engineersedge.com/calculators/shaft_ke...

Thanks for sharing your points of view. My doubt points to what 3DDave said, I am not concerned about the load that the 2 keys can carry but rather the reduction in the area of the shaft.

If i add a third keyway the "active" area on the shaft will be reduced and i do not know how it impacts on the new allowable torque. At the very begining of the keyway there is neither key nor coupling just grooves/keyways on the shaft and i supone this diminishes the torque capability.

Hi maugust

Try this site for a double keyway

https://www.engineersedge.com/calculators/shaft_wi...

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

Is the torque reversing or fluctuating each rotation? Probably not as much possibility of torsional fatigue as there is for bending fatigue. Where are the maximum bending and maximum torque locations? You probably need to evaluate the shaft based on the maximum bending moment and the corresponding torque at that location, and then evaluate the shaft based on the maximum torque and corresponding bending at that location.

I like the methodology of ASME B106.1 for shaft evaluation - this method assume more constant torque and fluctuating bending, and thus uses yield strength in shear and fatigue strength in bending for shaft diameter determination.


Also, if you cannot find the stress concentration factors, you might consider evaluating the shaft at the keyway by just looking at the diameter of the shaft between the keys. The geometric difference in moment of inertia is probably greater that the stress concentration factor that you would apply.

yep what dvd said