3
$\begingroup$

I'm designing an aluminium shaft hub for a 16mm keyed shaft with a maximum torque of 30Nm. However, stress analysis tells me that the keyway can suffer too much stress:

enter image description here

If my understanding is correct, keyways have little or no fillet, then how is stress relieved? In general, is driving an aluminium hub with a keyed shaft for ~30Nm a bad idea? There are some aluminium hubs with keyway on the market, but it's not clear if they're designed for 30Nm. Could I be missing something in my design?

UPDATE: I've added 0.5 fillets in the keyway, but the stresses are still very high:

enter image description here

Improve this question
$\endgroup$
16
  • $\begingroup$ The keyway geometry is dependent on what tooling you use to form it. You could add a second keyway 90 degrees away to reduce some of the stress, or use a tapered shaft and hub to lock on since your torque is relatively low. $\endgroup$
    – jko
    Jun 11, 2020 at 12:04
  • $\begingroup$ @jko The input to the hub is a planetary gearbox which I'm trying to keep intact. The keyway matches its output shaft. Like you said, 30Nm is rather low, but is it within expectation that so much stress will build up in the keyway? I'm just not sure if my analysis is correct. $\endgroup$
    – Kar
    Jun 11, 2020 at 12:10
  • $\begingroup$ Use a bigger key so spread the load more. $\endgroup$
    – Solar Mike
    Jun 11, 2020 at 12:21
  • $\begingroup$ @Kar the analysis looks accurate for the model you have made, but keep in mind your model has perfectly sharp edges. These will act as stress concentrators in FEA. In reality the keyway shaping tool will have a radius/edge break, even if it's too small to see. This will reduce the stress to that of areas close to the corner, so about 17 MPa based on your analysis. You should be fine. $\endgroup$
    – jko
    Jun 11, 2020 at 12:53
  • 1
    $\begingroup$ Keeping the solution simple for the the computer to solve and simple for you to set up is useless if the results that you get out are wrong. $\endgroup$
    – Jonathan R Swift
    Jun 17, 2020 at 14:32

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.