Using Solidworks Simulation (need some advise)

[Andrew07]

Hello guys.

I'm trying to run a good simulation on a braking assembly that consists of a rotor hat, bobbins (8 of them) and a rotor. Typical motorbike set up, and common in fsae too. Ive figured out how to apply the appropriate loads and fixtures and have run two simulations. One with solidworks having the component contact under Connections set to "bonded" and another with it set to "no penetration." Of course I'm wondering if the latter is appropriate (I don't think the former is). In any case, should I be setting up the "connections" portion of this simulation in a different manner?

[Andrew07]

Hello guys.

I'm trying to run a good simulation on a braking assembly that consists of a rotor hat, bobbins (8 of them) and a rotor. Typical motorbike set up, and common in fsae too. Ive figured out how to apply the appropriate loads and fixtures and have run two simulations. One with solidworks having the component contact under Connections set to "bonded" and another with it set to "no penetration." Of course I'm wondering if the latter is appropriate (I don't think the former is). In any case, should I be setting up the "connections" portion of this simulation in a different manner?

[thewoundedsoldier]

Why not FEA the different components individually? The results will be much more powerful.

Using a bonded connection would not show you any useful results because the entire bobbin geometry will stay fixed within the rotor or hat. No penetration is the way to go for an overhead view, but I would suggest looking at each part individually.

[Andrew07]

I could look at the parts individually but I don't know how the load is distributed across them. With normal bolts you can say they all contribute equally to shear (from the brake pad force, for instance) and all react to the moment created by that force. Then add the stresses together and look at the most stressed bolt. But with these bobbins I assume the ones towards the rear of the car at a given time instant cannot react to the primary (applied) shear force. Just the moment. Then there's the issue of how the other bobbins on the front side are distributing the loads. I figure Id have to make so many assumptions I might as well have solidworks look at the whole assembly during the analysis.

[BMH]

The bolts should actually be pretty easy to find the loading and stresses for. If you know your tractive/braking moments, then you should be able to find the stresses, factor of safety, etc. for the bolts. Check out a machine design book and look in the Bolts/fasterners section.

[Pennyman]

I remember attempting this several years ago.

Using the bonded connection will put the bobbins (I call them rotor buttons) in tension as well as compression, which isn't correct.

You need to (at the least) apply split lines to your tophat and your rotor buttons to isolate the contact points to the areas where load will actually be applied. This will get you a little closer to the way it's actually being loaded.

Using the "no penetration" contact restraint instead of bonded does make things more difficult, and I remember having trouble myself. I think I ended up analyzing the parts separately as well, and cross referencing that with the assembly analysis to see if I was doing it completely wrong. If I think of anything more useful I'll definitely post it.

[Andrew07]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The bolts should actually be pretty easy to find the loading and stresses for. If you know your tractive/braking moments, then you should be able to find the stresses, factor of safety, etc. for the bolts. Check out a machine design book and look in the Bolts/fasterners section. </div></BLOCKQUOTE>

Yes that works for when the bolts are holding together 2 peices that are not in plane. It would work for a non-floating rotor for sure. In my case the buttons rotor and hat are all in plane. The significance being that the buttons "behind" the force application will not be part of the reaction to the primary force, just the resulting moment. Ill put it in other words just in case that didn't come across right: I suspect all the buttons to react to the moment created by the pad force and distance from the centroid, but only half the buttons to react to the primary force.

There's a lot of options under connections to play around with. There is a TA for one of my classes that is pretty good with running these kinds of studies. I'll see what he thinks. Hopefully he can offer some good insight.

[thewoundedsoldier]

Remote load should solve your problem. They have an option for "direct transfer" that will apply the moment and force in the same hoorah.

There's another option called "rigid connection" that might work if you tinker around with it. Direct transfer will probably be the best way to get both load components though.

[Andrew07]

Thanks, Ill make sure to try that shortly. In the meantime I have seen that the following options had stresses concentrated in all the places one would expect:
component contact: no penetration
large displacement option enabled
improve accuracy for large displacement enabled

With that, everything looks right. For example: On the rotorside of the button, the left part of the button is reacting force but not the right. On the hat side, the right side of the button is reacting force but not the left. On the opposite side of the rotor this statement would be flipped. The buttons near the pad are experiencing much more load. I'm very pleased with how the results look, as they agree with one's intuition.

[Adambomb]

Sounds like you've got it worked out. Definitely don't want tension force on the bobbins. I was going to recommend using a bearing load, that should also work well for that. Or perhaps it's the same thing, I haven't played with Solidworks '10 yet.

As for whether or not you want to do it as an assembly or individual parts, one thing to consider is how rigid the rotors are. I've played with some that have been lightened so much that they have so little torsional rigidity (about the wheel axis) that most of the bobbin force goes to the one or two bobbins closest to the pad. Now you're talking a statically indeterminate problem with many degrees of freedom...definitely they kind of thing I'd want to do in an FEA assembly!