Transient gear modelling

[amarillogears]

Hi all,

I'm attempting to simulate the dynamic loading between two gears in mesh during one complete cycle. So far I've managed to get some successful simulations using a static structural analysis, however this is not appropriate for what I aim to achieve. As the gears come in/out of mesh there will be small deflections which the teeth will undergo, which as opposed to no deflection, will increase the stresses present. It seems a transient analysis is necessary to model the rotation and also calculate the stresses, although I'm not very familiar with this. My main difficulties are getting the contact right so the gears dont go through eachother and getting the rotation to work correctly, although I get a fountain of errors from every transient simulation I even try. Any help would be greatly, greatly appreciated!

Thanks,
Mitch.

 

[IanME]

My main difficulties are getting the contact right so the gears dont go through eachother and getting the rotation to work correctly

Look into the 'Adjust to Touch' contact setting.



Also, a quick 'google' for 'Transient Gear Analysis' resulted several links that could be of some use to you:

http://www.caeai.com/engineering-an...Contact/Transient+Dynamics+Analysis&caseID=10

http://www.scribd.com/doc/29351219/Transient-Stress-Analysis-on-Spur-Gear

Hope this helps,
-Ian

"All models are wrong, but some are usefull" - George E. P. Box

 

[amarillogears]

Hi Ian, thankyou for the reply.

I've read that article before and although I follow most of it, actually doing it is another story, I'm still quite new to Workbench.
The CAE Associates simulation seems perfect, although once again quite complex and all that seems to be provided on the site is that brief summary?

The 'adjust to touch' approach seems like it could help the contact problem, although I still havn't been able to get the gears moving in the Transient analysis.. Would applying a moment or setting a rotation velocity be more appropriate? I'm currently exploring using a rotational velocity, but when looking at the solution it seems not too much is happening.. Stresses and deformation are produced but the gears seem to remain stationary?

If it would help at all I can upload the "progress" i have made so far.

I'm sorry this seems so vague but everything I am trying just causes more problems and I am reaching my wit's end..

 

[IanME]

I'm currently exploring using a rotational velocity, but when looking at the solution it seems not too much is happening.. Stresses and deformation are produced but the gears seem to remain stationary?

I seem to remember (albeit vaguely) that rotational velocity in ansys (CMOMEGA in apdl) equates the stress and deflection fields from the velocity vectors created but does not simulate the movement in the environment. I'd definitely try using a moment load, although it really depends on your application (do you know the moment load?). The other thought that occurs from the quote above is that your model is over-constrained. How have you constrained the gears' motion in space?

I did a little more research and found this nice step-by-step gear analysis in workbench:

http://www.youtube.com/watch?v=J_o3hWQnEHM

. The analyst, Huei-Huang Lee, uses a moment load and also some pretty nifty contact settings between the two gears in initial contact (Frictionless, Aggressive Lagrange, Adjust to Touch). It seems this analysis might not be as in depth as yours, but it looks like a really great starting point.


I hope this is clear and helps you in some way.

Best of Luck,
-Ian

"All models are wrong, but some are usefull" - George E. P. Box

 

[amarillogears]

Hi Ian, thanks again, and I'm sorry for the delayed response.

I currently have the model's constraints as one fixed support and one frictionless support, both at the centre of the gears. The moment is then applied to the frictionless support gear. The value of the moment load itself isn't overly important.

Yes that video is what I have based my initial simulation off of. It was very useful and achieved good results, however I am still struggling to change from a Static Structural analysis. I've also found this;

http://www.youtube.com/watch?v=oaAnStqYq3M

which demonstrates motion in an Explicit Dynamics simulation. Although linear, it takes me a tiny step closer..

Nothing seems to want to work! And I do not have enough time to completely learn the ins and outs of each analysis. I am at the point now where if anyone can get this to work I would be happy to pay them to do it just to get it done.

Thanks Ian.