CosmosWorks Distributed Loads

[TomFin]

Here's the situation I'd like to analyze. A torsion spring with one of the straight spring legs resting flush against a sheet metal flange, the other...well I dont think it has bearing on potential discussion. This spring leg should deliver a triangular distributed load from torque: Zero above the spring axis, Maximum and the extreme tip of the spring leg. I'd like to mimic this real-life loading scenario in CosmosWorks. Under the Loads Feature box resides an expandable window containing parameters for a Non-Uniform Load. I'd like to at least try experimenting with this function but it requires that a coordinate system is selected. I've tried picking lines,vertexes,origins, you name it. I've followed the idealistic help instructions without success.
Any suggestions would be greatly appreciated.

Thanks,
Tom

Failure is a prerequisite of successful design

 

[gfbotha]

As I understand, you are actually interested in modelling the spring (on its own).

I'm not sure whether you are right to assume a distributed load on that leg. It depends on the layout, the relative length/width of the sheet metal tab, any offset, etc. If you start (at no load) with the spring leg perfectly in line against the tab, then chances are good that as the leg starts bending under load, the tab edge will start reacting most of the load?

Anyhow, you might be able to do what you want to by using Geometry > Curve > Split Line. This might allow you to split the leg surface into planned portions using an initial sketch. Alternatively, edit or cut away a small/thin sliver from the spring leg 3D CAD geometry.

Regards
Gert

 

[TomFin]

Thanks Gert for the insight about the realistic sheet metal behavior. In my realistic scenario the leg of the spring rests flush against a sheet-metal flange incipient to the initial loading. Therefore it is that situation which I would like to analyze accurately. I understand that the edge farthest from the spring will deform first rendering another scenario, but I'd like to gather data about the initial loading of the flange.
The modelling of the parts is done and doesn't pose as a a problem. I am quite familiar with split-lines. Yes I could cut the flange into small slivers each with its own load to represent the distributed triangular load, but I'd rather try and take advantage of the COSMOS features rather than be reminded of integral calculus.


Failure is a prerequisite of successful design

 

[gfbotha]

Tom, I have appreciation for you trying to use the functionality that is supposed to be there. You mentioned exploring HELP, but did you see the tutorial on this functionality? (The storage tank with liquid pressure.) Or, let me rather say; we have such a tutorial in the case of Cosmos DesignStar... However, I did not yet had an opportunity to try out this new feature myself. Will try it out when I have some time.

 

[TomFin]

Thanks gfbotha, I just found out how to insert coordinate systems. Reading the Cosmos tutorial I've learned that the tutorial instructs on coordinate systems as well. I tried using a non-uniform PRESSURE distribution on my flange with success. But as soon as I change it to a FORCE, the study runs with an error message of "Force1 transfer failed." I then closed SW and re-opened the program and it worked, I did not get this message upon completion of inputing my parameters. Although no errors were pointed out to me by SW/CW the preview of the force vectors did not look like a distributed load. Instead it just had 2 vectors representing the maximum force, none transitioning to a zero force.

Failure is a prerequisite of successful design

 

[gfbotha]

OK, I spent some time trying out the non-uniform loading feature of Cosmos DesignStar. I analysed a sheetmetal part bent in a L-shape with the load applied normal to the face of the short leg. The load varied from zero to some max value. FEA results were verified against some manual calcs.

I did not encounter any errors, and the loading was depicted correctly with force vectors which varied (increased) in length. I successfully tried out Force loading, Normal Force loading as well as Pressure loading.

But, analysing in SI units (m) after having started with solid geometry created in mm, I initially had a hard time figuring out the correct Force/Pressure values and the matching equation coefficients… Maybe the most important clue is that in the case of force loading, the force value should be the TOTAL (per entity) and the coefficient could be 1 or 1000 (it does not matter!). With pressure loading, one of course has to use a specific coefficient in order to achieve the correct max pressure at a given distance.

Regards