Bike frame, FEA, Creo 2.0

[mcadek]

Hello,

Im strugling with a bicycle frame FEA for a while now (please check the pictures below).

After getting over some basic problems like absolute/relative accuracy values to mesh the whole thing, I have no errors during simulation, but unfortunately and probably, I have no convergency either.Solver is stucked at Time Step 1 (so no load), but its 106309 elements and I have slow PC.

So I had to ask myself a question...

Is it even possible to do this in Creo Simulate? (I have seen this job done with Ansys)

The rear wheel should move for about 120 mm before the mechanism reaches equilibrium with spring stiffness I used. (120 mm only because it is a kids bike on 20" wheels)

I know that theres other way around this problem, which I was using the whole developing time and that is: mechanism + "use in structure", however I still want to compute the whole assembly.

I have some other ideas to avoid the movement of a rear axle and usage of idealized spring to get equivalent situation or load.

For example, have head tube fixed, suspension to known compression state with known force at the rear axle and fix pos.of both shock pins. But Im not quite sure if that would be the right way yet.

Any ideas?

Or, is it possible with spring deformation and wheel travel this big, to get some decent results?

I used contact only between bearings inner and outer rings and in some other parallel surfaces (see the picture below). I also used rigid links for the pins where the spring is attached to avoid stress peaks since the spring can be attached to points only. The rear axle is also defined as rigid as I dont care about stress values there.

THANKS

M.

This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.

[JonathanHodgson]

If you're after frame stress then I would build it with rigid shock absorbers.

To be honest, unless you have a good reason I would analyse just the frame component, applying calculated loads at all the relevant points (use Review Total Load to check that the overall load is correct / balanced). It's probably more efficient to calculate the loads externally and run a smaller, simpler FEA model.

[mcadek]

Thank you for your reply Jonathan

I am after all the stresses

I have calculated the parts alone many times using inertia relief during "development phase"

Its for my master thesis, basically to spare pages (I have few more load sets) but mainly for presentation

I have now applied symetry to the model, I also used shell elements and tweaked the mesh control to get fewer elements.

I also changed the job to the following: spring was replaced by rigid link, suspension was moved to the steady state position known for that particular load from dynamic simulation.

Analysis has now finished step 1, after 24 iterations and 48 minutes on older Core Duo 2

I have secured much better machine for friday at university, since it seems that the model is defined ok I will try that spring thing again,

I will keep thist topic posted

and I have first iteration with Load factor = 1

[JonathanHodgson]

Milos Cadek wrote:

I have calculated the parts alone many times using inertia relief during "development phase"

Milos,

Why have you been using inertia relief? It should be possible to constrain a single part such as these fully. As I understand it, inertia relief would be used primarily in a case where acceleration of the mass of the part itself is a significant part of the free body diagram. For your bike frame, I wouldn't expect the inertia of the parts to be significant compared to the reaction loads between the suspension and the weight / mass of the rider.

[unickque]

As far as I know you can use Inertia Relief always. All it does is balance out the forces on your model with some inertia spread over the whole model. That way you can for instance apply only loads and no constraints, and still get decent convergence, even if your loads are not 100% accurate.

The risk is that when you are not careful in balancing your forces, a lot of inertia can be added to your model which may be unrealistic. So use with caution, and only when needed.

[mcadek]

Patrick: correct

Jonathan: Inertia relief was used for simulation of mechanism links, ie the rockers. I wanted to use all three reaction forces from dynamic simulation combined, so inertia relief was the way to do that.

[unickque]

Did you mesh with shells or with solid elements?

Have you tried to model a rotating contact like this in a smaller model, to see if it works?

[mcadek]

Thanks for the reply

Im using both types, could that be a problem? Having mixed mesh?

Yes, I have tried that and the results seemed OK, I tried a lever on a pin with force on one side and spring on the other