Rotating Shaft with spline end

UGMENTALCASE

Morning all,

I have attached a little something I'm looking at out of interest. I'm basically running bits and bobs of tests to get a rough understanding of things. So when you open this up, you'll see the unit. Unfortunately it's too big a file to attach with all constraints on. I have fixed it at the 4 holes, back face and front face of the diameter piece.

The forces I'm a little iffy on. I wan to try and figure out how much that end splined bit will take before it snaps off. What would be the best way to apply the forces to this?

Thanks in advance

Victor

Could you please also attach the shaft.step file? The geometry doesn't get included inside the .liml file until you mesh it. Liml files without solution data compress very well so zipping can help if it's too big.

Not having seen the geometry, it sounds like a cantilever with a lateral end load, so probably just apply a force to the end face to get a worst case bending moment. If it's not very slender then it might be more important to distribute the force closer to how it would be distributed in real life, such as spread over the surface of the spline.

UGMENTALCASE

Cheers

Victor

I see, thanks. It depends how it's being loaded. Are you interested in bending, shear, or torsion?

For a rough first approximation of shear strength, I would put a Y-direction force on the end face of the spline's shaft. But this could be wildly off depending on how much bending there is or isn't.

For torsion, you can make a moment from two forces in opposite directions. There are also more difficult ways to distribute a moment over a surface if really want to avoid modelling the part connected to the spline. See these other threads for that:

http://mecway.com/forum/discussion/comment/632/#Comment_632

http://mecway.com/forum/discussion/comment/281/#Comment_281

http://mecway.com/forum/discussion/91/reaction-summation-on-fixed-support

Don't constrain the face of the large disk nearest the spline unless it's actually constrained in real life because that may need to deform near the narrow shaft.

UGMENTALCASE

Thank you very much i'll have a read of those posts and see. That's what i started thinking about, modelling the other half of the parts. I could do, and try that could look into both ways.

Thanks again for the help :-)

UGMENTALCASE

Hi guys, back again!

So I'm working on the first link example above. So I see we fit this diameter piece on the end which acts like the rotation device (if you will), is that correct?

Could some one help me out as to why we have that as the same material as the main body? Is that so it 'acts as one' so to speak? When I run the analysis on this example we get a higher MPa stress at the ends of the cross section. Is it safe to say we ignore that as that piece is almost false, it's just put there to twist the main body?

cheers

UGMENTALCASE

attachment

Victor

Yea, the extra parts are just to transfer the load. If it's too stiff, it will artificially stiffen what it's attached to. There's no general correct way to make a distributed moment because it depends on how the load is really applied - if it's through another part which is very stiff, then the extra part should be stiff too.

And yes, ignore results next to the attachment.

UGMENTALCASE

Awesome! I think I'm getting the hang of what I'm seeing. Trying to blend that into a contact as well. Found a video on youtube where a part is rotated into the side of a fixed block. So when it contacts the force (from the rotation) keeps going and the part bends. I'm almost there with it but can't get the ccx contact constraint to work right.