Simulating beam deflection with free ends

stusona

Howdy folks. I'm playing around with the new simulation tools and started out with a simple beam deflection example. I got a "fixed end" solution up and running in minutes, but then spent a considerable amount of time trying to get it to solve a "free ends" model.

I was hoping I could just leave that top plate unconstrained and just allow OnShape to use contacts to figure out the "hinge" around which the material would bend, but no luck.

I tried a bunch of different configurations of mates too. I feel like a pin and slot at each end makes the most sense, but it still wouldn't solve.

Here's the document: https://cad.onshape.com/documents/d2440289f8117f7ffc496a11/w/3a480f7e650f42f3f0cf5455/e/710ba9a85c170463c8b0a1a8

Thanks in advance!

Here's the fixed end example:

S1mon

Pin and slot is not supported for simulation (this is a little buried in the help page).

• The following Mates are not translated, and therefore, not supported by simulation: Tangent Mate, Pin Slot Mate, and Parallel Mate. Relations (Gear relation, Rack and pinion relation, Screw relation, and Linear relation) are also not supported.

I was able to do what I think you're looking for by adding in two fake joint parts that have a revolute mate (to Part 1) and a planar mate (to Part 2).

It's unclear to me from the deformation view whether the simulation is correctly conserving the length of the plate (and sliding on the planar mates) or if it's getting "glued" at the planar mates. I set the connectivity method to "mates" to avoid any weird assumptions from "mates and touching faces".

Kevin_Cowles

Hello! First off, when you leave the top plate un-mated, it will still be treated as bonded to the edge faces by the simulation, since you are working with the Mates and touching faces Method for Simulation. See the bottom half of the Sim Panel to visualize what contacts are being honored in your simulation. This Tech Tip covers those features in detail.

In terms of achieving this "free ends" model, in general you need a statically determinate assembly for Onshape to solve. For instance, using the "Mates" method with the top plate unmated, you have rigid body motion and cannot converge. Additionally, Pin-Slot mates are not currently supported by simulation.

The approach I would take here is to put a revolute mate on the edge that is the "hinge". Note though, that that mate applies to all faces of the part in contact, or very close to being in contact. So if you truly wanted the outside edge to lift up, you'd need to use a simulation connection (see section below this) to specify that you only want that revolute mate connection to apply just around the edge.

Kevin_Cowles

Ah beat me to it S1mon, the dummy parts will work as well

S1mon

@Kevin_Cowles

I tried a single revolute mate on one side (without the helper parts) and it won't converge. If I add two revolute mates, it solves, but it glues the ends down and stretches the middle. This isn't what I think @stusona is asking for.

S1mon

I did a similar test with one revolute mate and one planar mate, and the deformation seems to be the same incorrect results.

I also tried adding the "simulation connection" selections to the mates. It didn't seem to help. 

I would really like to see tangent and pin/slot mates supported. We spent a lot of time figuring out 1) that they aren't supported 2) how to work around that.

stusona

Thanks all. Didn't see those notes about mates not supported! Wish that was more explicit in the simulation environment.

Yea I tried configurations like the single hinged revolute and had the fastening issue Simon is showing. And I tried with/without checking the "simulation connection" checkbox in the relevant mates as well. Though I didn't explore the "show connections" very well. That's helpful!

Got what I'm looking for: free end deflection! The max deflection formula predicted 12.35 mm, so 10.1 mm is pretty reasonable.

Can't wait to play around with the tool in the future.

stusona

Agreed. Also: how do you "save" results? If I click away into the part studio then come back I need to wait for it to solve again.

S1mon

stusona said:

Agreed. Also: how do you "save" results? If I click away into the part studio then come back I need to wait for it to solve again.

Exactly. We tried using Publications, which we thought might save the cached results, but nope. It's pretty frustrating. We basically end up taking a screen  shot and putting it in a Google Slide. 

I need to add this as an improvement request.

Kevin_Cowles

@S1mon Sorry that it's hard to find in the Help Docs! In the Learning Center Course that was just released, its much more upfront that they aren't yet supported. The workaround you've set up is a good one, and I'll see if we can add more workflows of that sort to our documentation. Definitely submit an IR for the addition Pin-Slot and Tangent, I can also pass the feedback along myself.

And I see that I mis-evaluated the original ask, of allowing the top plate to slide on the "hinged" contact. That cannot be achieved with the simulation connection, you will need a second mate/body. I was trying to help allow the edges of the plate to lift up and not be fused to the top face, but your solution is much more appropriate here.