How do you perform an accurate simulation on a part in an assembly?

Hi all

Let me explain, I would like to know if after a first simulation of an assembly it was possible to isolate a part in order to carry out a more precise study, for example to identify the forces, displacements that apply to this part in the assembly (during the first simulation of the entire assembly) and then try to apply them to the part on its own in order to to represent its behavior in the assembly and this in order to make a very refined mesh of the part alone.

Do you think it's too complex to want to represent the behavior of an assembly by only forces, imposed displacements (and all the options offered by Solidworks simulation)?

Hello

Anything is possible, but you have to ask yourself why you want to do this.

Let me explain!
When you do a simulation on an isolated part, it most of the time comes down to simplifying the constraints, except when the movements are almost 100% predictable. Example: a link in the chain.

The second example is a simulation on an assembly whose operation is static. Typically a mechanically welded frame with different plates that support one or more loads. If you do a piece-by-piece analysis, you will have virtually no deviation from the simulation on the assembly.

Third example: your assembly has moving parts, typically an assembly with kinematics and displacements of parts and loads (and depending on the vibrations, only the assembly allows you to see the amplification by reasoning). In this case you may have big discrepancies depending on the version of solidworks simulation you have.

With the standard version you always work in static mode, while with the PRO version you work dynamically. The fact that it works in dynamic mode will mean that the number of price parameters in account for the different deformations that your model will undergo will be a hundred times greater than what you can design on a single part which is also tested in static.

Wanting to do a more detailed analysis on a solo part implies that either you are in a borderline condition on this part in the assembly. And that you want to solidify certain parts before re-testing it in an assembly context.

If, on the other hand, you are in normal safety conditions, you want to do some refining, it's bead threading and cutting with a lace scissor. :-)  :-)

We would have to be given indications on the part and the assembly to be more precise.

Kind regards

2 Likes

Thank you Zozo_mp!

So I was asking myself this question in general, in my work it mainly concerns static studies with mechanically welded assemblies. So I wouldn't have a gap between a piece-by-piece analysis and an analysis of an assembly? 

And what do you call normal security conditions? And what is this annoying thing about?

Thanks in advance!

Sincerely, 

Normally no deviation but we cannot give a universal rule.

In a mechanically welded frame it depends on where the load is (and the nature of the fixed or non-fixed load). If you have two mechanically welded frames that are bolted together and the load is placed on the top frame, remote loads must be placed on the bottom frame. Remote charging replaces the frame or ground that does not exist in the individual room.

Be careful not to be told what I didn't say. You would have to submit a case to us from time to time to show you the point.

1 Like

Very well, thank you once again Zozo_mp ;)

Yes indeed it all depends on the case and there I didn't have a specific example to show you But I won't hesitate next time! I still have a bunch of questions^^

Sincerely, 

 

Simulation also allows you to mix the meshes and to make meshes of different sizes depending on the parts: it is therefore possible on a mechanically welded tubular assembly to simulate everything in beam except for the local part that you really want to study where you will use fairly fine volume mesh.

This type of beam/volume or beam/shell/volume mesh is nevertheless less and less interesting given the power of today's PCs. A mesh size of 20mm on parts far from the critical area to end up with parts meshed in 1 mm where you want a correct result will perhaps be sufficient without penalizing the calculation time too much.

It all depends on the complexity of the assembly and the type of junction between the parts (if we manage the contacts between parts, the calculation times are much longer and the models have more difficulty converging than if we consider that all the parts are in rigid contact (in this case the calculation of the assembly comes down to the calculation of a monobloc part of the same volume))

Do not hesitate to take into account symmetries to lighten the model (many designs are symmetrical, calculating half of the model basically reduces the number of nodes, not two, and therefore the computation time)

1 Like