In assembly mode (or part) is there a way to make Solidworks apply force in a specified direction until the selected component in the assembly (or the part itself) reaches its yield limit?
Yield stress is typically in the linear portion of a stress-strain curve (by definition). You should be able to find yield in two iterations.
I followed your idea and realized that this method will yield very close but not exact results. For instance....
My first simulation had a force of 700 lbs generating 55374.3 psi of stress for a material (T6-6061) with yield strength of 39885.4 psi.
Doing simple math I calculated that the force to generate the yield stress of the material is about 504.201 lbs.
I reran the simulation with 500lbs of force and still ended up slightly above my yield value, which should not have been the case following a linear stress-strain curve.
I still believe that this is either a wild shortcoming of Solidworks or a wild lack of knowledge on my part.
If you have large deflection turned off, then I would expect to get an answer that scales perfectly. If you have large deflection turned on, then it gets trickier and I would not expect to be able to scale exactly. As the part deflects, things are no longer perfectly linear.
One other possible source of error is meshing. I might try running 5 iterations of h-adaptive and see if it converges on the sy if it is just a single body/part. You can check the option to coarsen the mesh in low stress volumes to speed up the solver.
Sooooo important to get convergence on refined mesh. Coarse mesh will give results that are too low.
You appear to be doing the calculation correctly. It could be the significant digits after three, are numerical round-off in the solver (certainly if a different solver), or the maximum stress value position has moved such as when contact is involved, or the mesh is slightly different (it should be an exact copy you could generate by "duplicating" the study).
Generally speaking, I tend to stick with no more than three significant figures. During my days as a lab test engineer, I typically observed repeatability within about 5 to 10% - depending on the humidity, temperature, test specimen, geometry variations, material lot, strain gage batch number, load cell used, time of day, even the day of the week, etc. As a result of that experience, for the FEA I would report 39,885.4 psi yield strength as 39.9 ksi, and any FEA stress values within about 10% of that is a match. I have noticed rather extensive "precision" in some of the material property values in the database. However, I find it extremely difficult to believe measurements to six significant figures are realistic.
Hi Bulent: Roland is correct. Assuming linearity to the yield strength. Usually, I apply 1 (one) pound (or some other multiple of 10) and look for highest stress, then just scale the force required to get to the yield stress. Example: Published material yield strength is 50 ksi. Apply 1000 lbs and observe max stress is 20 ksi. So the load to cause yield, scales as (50/20)*1000 = 2,500 lbs. Just for kicks, apply that load to the study and you should see 50 ksi in the model. That's two iterations, as Roland indicated.
I suspect also that you could use the built-in "event-based" simulation functionality, by applying a sensor at the point of interest, and parameterizing the load to have the software iterate and "stop" at yield. I have not tried that, however, but it appears useful for nonlinear material behavior.
Anthony Botting wrote: I suspect also that you could use the built-in "event-based" simulation functionality, by applying a sensor at the point of interest, and parameterizing the load to have the software iterate and "stop" at yield. I have not tried that, however, but it appears useful for nonlinear material behavior.
Anthony Botting wrote:
I'd like to learn how to do this as I believe it is the correct way to set yield limits in simulations. Do you happen to know a good tutorial or guide for sensors used for purposes like mine?
Retrieving data ...