
Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Siavash Khajehhasani Sep 13, 2016 8:52 AM (in response to Will Kim)Can you upload your model with both simulations here?

Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Will Kim Sep 13, 2016 12:43 PM (in response to Siavash Khajehhasani)Hello Siavash,
I can not upload everything as this would violate disclosure agreement.
Instead, I have added more pictures showing load, restraint conditions in my design in my reply to James.
Your advice will be really appreciated!
Thanks,
Willk


Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Mike Pogue Sep 13, 2016 10:13 AM (in response to Will Kim)Will,
I'm not seeing where you applied the symmetry conditionthis is not automatic. You half model is putting a moment into the truss that is not in your full model. Your model is very small and should run very fast. I don't think you need to worry about symmetry. But if you wanted to, you would need to add back half of the cross members, then apply the symmetry condition to the geometry at the middle.

Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Will Kim Sep 13, 2016 12:43 PM (in response to Mike Pogue)Hello Mike,
When I said i used symmetry condition I meant that my design is just in symmetry with respect to the center of mass, not that I used symmetry fixture in simulation. Sorry about that.
I have added more pictures in my reply to James Riddell answer, perhaps would you be able to explain why there is moment developing if i use just one truss?
Thanks,
Willk

Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Mike Pogue Sep 13, 2016 4:19 PM (in response to Will Kim)The moment is trying to twist the one truss, where if you have two trusses, the twisting is much less., because in order to twist it you have to stretch and bend other members. When you have only half the structure, the twisting of the one truss is decoupled from other deformations and it's stiffness goes way down



Re: The result I got using symmetry condition is not the same as the one i got using entire model.
James Riddell Sep 13, 2016 10:23 AM (in response to Will Kim)If I was attacking this problem I'd move my 1/2 force back to the plane of the truss. In the picture it looks like you are applying a moment and a shear force. (Can't really tell where your restraints/symmetry are/is.)

Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Will Kim Sep 13, 2016 12:43 PM (in response to James Riddell)To aid your understanding in my design and simulation, I have added extra pictures showing location and type of restraints and force used for the model. The truss is not exactly symmetric with respect to the plane of truss so If I follow your advice and add 1/2 force back to the plane of truss, where would you be attaching remote load to?
Also, hopefully, with more pictures below, can you please explain why there is moment of shear force developing when only one truss is used?
Thanks,
this is one end. It has roller at the bottom and zero radial displacement in the face where bolt would be bearing loads.
This is the other end. the restraining condition is similar to the restraints i made above because this part will be bolted as well.
The pic above shows where remote mass will be attached too. In real design this is where idler will sit on.

Re: The result I got using symmetry condition is not the same as the one i got using entire model.
James Riddell Sep 13, 2016 1:03 PM (in response to Will Kim)If you apply your forces through the shear center of your crosssection then the truss side will act more in pure bending and should give a closer result to the previous complete model.
Just because you have roller/slider restraints doesn't mean you only have a perpendicular reaction force.
Since the remote loads applied are connected to the surfaces with rigid elements I suspect that they are not allowing the 'full' model to twist as in your 'half' model. However, since you are not showing any geometry between the two trusses I'm not clear on how SW would even be able to handle a symmetry constraint in the first place.



Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Mark Stowart Sep 13, 2016 6:39 PM (in response to Will Kim)Theoretically it should work, bus as you said it gives totally different results

Re: The result I got using symmetry condition is not the same as the one i got using entire model.
Bill McEachern Sep 13, 2016 4:23 PM (in response to Will Kim)you need to apply the symmetry constraint at the load point. You would need to add a remote displacement and put in the symmetry condition.