Calculate max force before breaking for a very simple object

vladimir_petrov190

How can I measure the maximum force before breaking for a very simple object (like a beam or pipe)?

I created the model in Onshape and used SimSolid (from the App Store) to simulate the load. As a result, I received the bending magnitude, but not the maximum force the object can withstand before failure.

MichaelPascoe

Onshape has simulation capabilities directly built into the assembly now! No need to use a 3rd party unless doing more complex simulations.

I'm a little rusty with my inner engineer, so someone step in please to correct me:

Onshape doesn't have a "Find breaking point" button, though this is a great idea for an improvement request. The shape of your part is going to greatly affect the outcome. But here are the general steps you can take to easily find that breaking point:

• Assign your Onshape material to the part and note the Ultimate Yield Stress (psi).
• Insert your parts into the assembly. Make sure one is fixed and the other has a force applied to it. Note: The mate connecting them needs the "Simulation connection" check box checked. (Update: Just found an easier way. If your parts aren't moving parts, just insert them as a rigid part studio then fix only one of the parts. The simulation now supports this approach and it is way faster for set up time.)
• Click the show simulation results.
• When using the simulations, switch it to Signed von Mises Stress (psi).
• Hovering over the simulated part will show you how much stress is being applied to your part. If that number is below your (UTS) Ultimate Yield Stress, your part shouldn't break.

Onshape Simulation Example

https://cad.onshape.com/documents/427b7835b7bd87f214d6e137/w/db846339ef05228c726baa3d/e/4ad8495…

GregBrown

There's a lot to unpack here and it's late but I wanted to add a couple of things today before coming back to it tomorrow!

There is a "safety factor" display in the results options. This is the material yield strength divided by the Von Mises stress calculated in the model. Meaning, anything below safety factor 1.0 is exceeding the yield strength of the material. ("Yield strength" is the term used for material properties, and is measured in same units as stress. The words do get interchanged)

Now the trickier part - "failure" can mean many different things and especially for different materials. Brittle materials fail differently than ductile ones, for example. Various failure criterion exist depending on the material: steel, aluminum, titanium, etc etc (and the type of loading but that's an even tricker topic for now.) The Von Mises criterion (also known as maximum distortion energy criterion) is one that is widely used (certainly is the default on many postprocessors) and applicable to ductile material failure. There is a LOT of documentation available on this and other failure models….

Final point for now, the UTS (Ultimate Tensile Strength) is not considered in the calculation and essentially has no meaning in any discussion here. It is a material property for sure, but in terms of a linear static analysis (like you'd be doing in Onshape), the response of the material is modeled as linear based on its Young's modulus and Poisson's ratio. The Yield Strength is just an (important) marker that implies "beyond this the material has yielded and any numbers higher that this are not physically relevant" (for a linear static analysis. If you are running a nonlinear analysis then that's another topic)

Long story short and extremely simplified:
Safety factor of 1.0 means you are at the yield strength of the material. Adjust the load to make sure you don't drop below 1.0

PS: Linear static analysis has a wonderful property that you can take advantage of - that is, the relationship between load and resulting stress is….. linear. So if a 1N load gives a max stress of 0.3 MPa, then the load to reach yield (of 276MPa for Al 6061-T6) would be 276/0.3 = 920N.

Oliver_Couch

The short answer is you can't. It's beyond the capabilities of the stress analysis built into Onshape.

What Onshape can tell you is confined to the linear elastic region of a material's stress/strain curve:

Onshape assumes that the stress-strain relationship is a nice straight line. But as you can see in the graph above, past the yield point the graph is no longer a straight line; any stress values onshapr calculates that are greater then yield stress (or strain) are unreliable, as the assumption of linearity is no longer true.

However, most of the time this is good enough. Most of the time you don't really care when a part actually breaks apart, because it will fail to do whatever job it is meant to be doing long before that. Or because it's behaviour becomes much less predictable past the yield point, and because metals become permanently bent (and more brittle due to cold working) after that, it's not a point you want your part to go past eg. Piston conrods go in the bin once they're bent. A bent pin will no longer go in its hole so it is no good.

Here's a great video on stress and strain:

https://www.youtube.com/watch?v=aQf6Q8t1FQE

The Wikipedia article is great too:

https://en.m.wikipedia.org/wiki/Stress%E2%80%93strain_curve

Oliver_Couch

https://forum.onshape.com/discussion/comment/106805#Comment_106805

"If that number is below your (UTS) Ultimate Yield Stress, your part shouldn't break." Use caution here, this is not correct. The built-in simulation is linear only, so the UTS - Ultimate Tensile Strength - isn't used or useful, except to let you know you're way past the bounds of what Onshape knows if you're near your material's UTS.

It would probably be handy for Onshape to throw a warning about this actually, for those who haven't had any formal training/education in FEA. Well actually I once had an Engineer say to me "Why would they give you the button if you shouldn't use it?" in reference to using a function that displayed gaussian stress across the whole element instead of the stress gradient as extrapolated to the nodes in order to 'reduce' the apparent stress magnitude when the stress gradient was very large - so even for those that have had some training a warning would be good.

To determine when a part will actually fail you need to run a non-linear analysis which uses a stress strain curve/lookup data to determine the material response beyond the yield point (unless your part is extremely brittle, in which case essentially the Yield Stress = the UTS).

MichaelPascoe

Thanks for the corrections! Yep it's been a while, I knew it wasn't as simple as I thought. This is a great forum post for others wandering about this same thing. Glad it's here.