BEAM

Hi,

I have a part attached , I just want to do an analysis of putting a load (just for example 600.0kg total on 2 surfaces on the beam.) and get FOS.

I have the 2 surfaces shown in red here, beam fixed at both ends.

The simulation doesn't seem to run, wondering if someone could take a look its probably very basic , but I don't use simulation much at all

Thanks

SolidworksSimulation
Answers 3
RD
RD 2021-04-09 Best Answer

Frank,

The way you have created solid body plates to apply force to then translate the force onto the beam body is not good FEA. If you want to be applying force loads along the beam you should instead create two Reference Points at the locations you wish the forces applied on the beam then apply forces to those points instead. The points will transfer the force directly into the beam body and remove the need to have the solid bodies which are, as you have it setup, not meshing.

comment imagecomment imagecomment imagecomment image
DS
DS 2021-04-09

Hey Frank,

I would second Ryan's answer, and to give a little more context let's think about the different ways we can mesh this beam:

1. Solid Tetrahedral Elements

2. Shell Elements

3. Beam Elements

Now, if your entire analysis is JUST this beam as you have it set up, then I would seriously consider using solid elements, as this is always the best route - but we employ shell and beam elements as a means of simplifying our problem and easing the load on our system resources (often quite necessary!).

When you have a weldment in the CAD space and start a Sim study, that weldment member will automatically be treated as a beam element type, unless you change it.

The main issues you have as the analysis is set up now are:

1. You're applying a load to two solid mesh components that then are transferred to a beam element type. In itself, this is not always bad, but in general that technique would be used when we are concerned about the "macro" picture of what's happening to an overall structure, not local interactions like we have here. The results won't be representative of what's really happening between the solid component faces and the beam, though if meshed properly we may get decent results w.r.t. overall deflection.

2. The two solid components mentioned above are extremely thin components, where your (solid) mesh settings indicate very very large element size. You'll never be able to cram such large elements into thin components like this.

There isn't any reason to actually have more than JUST the beam component present in this analysis. You have three options as I see it:

1. Run the analysis as a beam with the point loads as indicated above (most simplified setup, fastest solve time). Single component.

2. Mesh and run the beam as solid tet elements (most realistic, slowest solve time - you will need really really small elements here). Use a single component, and use "Split Line" to cut the FACES of the model to apply loads to.

3. Mesh and run as shell elements. Similar to number (2) you will use split line to "cut" the faces in order to apply loads, so we still have just a single component. Using this method we can really kick up the mesh density without increasing solve time dramatically.

I would start with method 3 above and follow up with a solid mesh (2) analysis as necessary (most likely necessary). There are actually a lot of variables and things to consider here. The first run as shown below isn't necessarily a final answer, but rather a starting point. As shown below, though, your FOS of course is less than 1 already, but this is why you will need to investigate further with more runs and solid elements especially to examine the areas of high stress concentrations we see here.

comment imagecomment image
FA
FA 2021-04-09

thanks Ryan