[ansys wb 13] multibody static simulation

[tele5palla]

Hello everyone,
I am trying to have the clearest ideas on how to handle multibody in ansys workbench.
I have a very simple mechanism (a manovellism) designed with solidworks and I know the moment(constant) applied to the tree.

I have to analyze it in a given position (hand crank angle) and I don't understand if I have to use joints (body-ground and body-body) to assemble it or simply contacts.

the doubt comes from the fact that using joints software obviously creates contacts, but the movement I do not care.

I have difficulty understanding well what function joints have, or better if it makes sense to use them in a static analysis.

I hope I've explained,
Thank you.

 

[stef_design]

Hello everyone,
I am trying to have the clearest ideas on how to handle multibody in ansys workbench.
I have a very simple mechanism (a manovellism) designed with solidworks and I know the moment(constant) applied to the tree.

I have to analyze it in a given position (hand crank angle) and I don't understand if I have to use joints (body-ground and body-body) to assemble it or simply contacts.

the doubt comes from the fact that using joints software obviously creates contacts, but the movement I do not care.

I have difficulty understanding well what function joints have, or better if it makes sense to use them in a static analysis.

I hope I've explained,
Thank you.

What form are you using ansys?

 

[tele5palla]

Hello and thank you for your interest.
I am using ansys workbench13 and simulation is static structural(ansys mechanical)
anyway in a full immersion in the help I found my carry, and in a way I also understand why of your question.
the connections defined joint, i.e. those I used, are created with the elements mpc184. I realized that in a static analysis it makes little sense to use them, or rather it is necessary to use them as possible because they also create cinematic constraints as well as contacts.
I would use them in a static analysis of a hinge if, for example, I did not model the pin.

should be used instead in dynamic analysis, between rigid and flexible bodies to limit the gdl.

I have modeled all the components of my mechanism and I want to analyze it statically, I do not need to use joints, but I can very well use mpc formulations for contacts i.e. bonded and no-separation that among other things have a linear composure.
an application can be found in the customer portal in training-ansys mechanical- introduction to ansys mechanical 13.0-tutorial 3.1

What I wrote is in line with what you would have answered me?

hello and thanks again

 

[tele5palla]

I am reflecting that I could have asked the question differently, for example:
What is the difference between a "cilindrical support" and a "body to ground joint-cilindrycal"?

I hope I understand the answer...

 

[stefano.garbin]

pin on hole... I would avoid things like "bonded" opure "no separation" because you find yourself a super-winned and unrealistic system.
Bonded is to be avoided because the hinges do not transmit moment, while bonded does.
no separation is to avoid because the surfaces that exchange forces do it only in compression and not in traction: a loaded shaft only presses half the hole. with the "no seperation" all the hole surface works.

It wouldn't hurt to do a sketch on paper with force breakdown, also because it helps understanding, reduces the complexity of the system and prevents you from messing up your life with contacts.

 

[tele5palla]

Maybe you're right.
But I have to tell you that some other simulation with workbench I did it, messing my life a lot... just like you say!
but I still think it's a necessary step to learn how to use the software. Get your head down!
The mechanism I've been studying for months now... .despite its apparent simplicity, it is my first true project. I am also very busy because I am extinguished of what I am learning at this time of my life... including asnys
Anyway, I used bonded for hinges... I have no axial efforts and in any case I could use for problems of convergence of the frictionless supports on the unbounded parts.
I actually launched some analysis between yesterday and today just using this kind of contacts and it works well to me.
true what you say about the transmission of forces... but I can overlook it for the moment. Maybe after another help I'll improve this too.


Hi.
claudio

 

[stefano.garbin]

Anyway, I used bonded for hinges... I have no axial efforts and in any case I could use for problems of convergence of the frictionless supports on the unbounded parts.
I actually launched some analysis between yesterday and today just using this kind of contacts and it works well to me.

the eye that bonded transmits tangential forces and is not good for hinges!
on the fact that it converges well I agree, because more a system is bound and better converges... but converges to a wrong result!

never forget the "physics" of the problem to analyze, and you always have an idea (qualitative) of what you have to get!

 

[tele5palla]

I was wrong to write! obviously I meant no-separation

However the result is right... I've been studying for months! paper and pen of course. I would simply like to complete with a fem analysis!

 

[tele5palla]

pin on hole... I would avoid things like "bonded" opure "no separation" because you find yourself a super-winned and unrealistic system.
Bonded is to be avoided because the hinges do not transmit moment, while bonded does.
no separation is to avoid because the surfaces that exchange forces do it only in compression and not in traction: a loaded shaft only presses half the hole. with the "no seperation" all the hole surface works.

It wouldn't hurt to do a sketch on paper with force breakdown, also because it helps understanding, reduces the complexity of the system and prevents you from messing up your life with contacts.

Stefano, I know I didn't want to listen to your advice... but I am sorry

How would you simulate a zipper? would you use other types of contact?
Thank you.