hello everyone,
new to the forum and new to solidworks simulation.
I'm trying to determine if a gate assembly will be able to handle a certain amount of pound force. this is what I've done so far. all my knowledge on solidworks is from online searches and forums such as this one, so bare with me.
first I simplified the part by removing all the fillets. Then I looked for gaps and edited the individual parts to close the gaps (do I need to set Contact settings here, all I did was edit the dimensions). From what I've read, I needed to save the assembly as a part, and I did that. Now, I have about 58 solid bodies in the new part file. I then used the solidworks Features/combine to combine most parts, and reduced it down to 6 solid bodies. this is where I'm stuck. I looked at the part and found the inside door is not making contact with the outside frame. I guess I will have to fix that.
so my question is, is it possible to apply a load on to an assembly? or do I have to save it as a part?
and
is my approach correct? or how would a professional solidworks user go about doing FEA on a gate.
also,
will I need to set my Contact settings on the inside door and the frame as well as the hinges and other parts that are capable of moving? I think after combining the 58 solid bodies into 1 would make it a bonded contact right, and it will act as one solid piece.
hope I make sense...
thanks!
David
SolidworksSimulation
We do assembly FEAs all the time. If you have the simulation package and not SWX then it is fairly easy.
Usually fully contacting parts come in as bonded (not always). If you have some moving/non-contacting parts then you'll have to set up contact sets based on your desired result.
Set up your boundary conditions, loads and gravity then mesh & run, easy-peasy.
Wow. That sounds pretty easy. I'll look into that when I get off work.
I hope im understanding this corrrectly, from your comment, I should go back and use my assembly file and not my newly created part file with combined parts. Then assign contact sets, apply load, and gravity. Mesh and run.
I spent all Saturday trying to get this to mesh and run but with the thought that it had to be one solid body.
I'll post a followup of what Im able to do later.
Thanks!
Mr. Riddell,
One probelm i ran into is this....How do I do split line in assembly mode to project the load? I was told it was not possible that's why I went to save it as a part and edited in there...
Is there another way to apply/project the load to a certain area of the gate (middle area)?
David
That is also relatively easy. Just sketch on the surface where the load/force will be applied, go to Features, select the Curves drop-down and pick Split Line.
If you only have SolidWorks Express then you'll only be able to run FEA on a single part.
I have access to sw simulation premium.
This is what I tried to do last Saturday before saving the assembly as a part.
I selected the front plane. Drew a circle for testing. But when in assemble mode. I do not have the option for curve/split line. Maybe I am doing something wrong.
I am able to do split line in assembly when I go to edit feature on a particular part of the assembly. But this will only give me the option of selecting the faces on that part only and not the area of interest on the gate.
Any idea what I'm doing wrong?
"Any idea what I'm doing wrong?"
I love easy questions...
Instead of selecting a plane for sketching, select the part and edit it. You can use one of the part planes for sketching or a flat face. Then select Curves, etc. When you update the assembly and save it you will have a surface to add a BC to.
Just to clarify. I also did a quick search on google. Boundary conditions and contact set are the same thing right?
No, not at all. BCs are things like restraints, forces, pressure, etc. Contact sets are how surfaces interact - bonded; they stay in contact and transfer displacements, etc.: non-penetrating; surfaces may separate. With non-pen contact sets you can add friction between the surfaces.
I get it now. Almost off work now. Then im head to campus to tackle this fea problem again. Thanks for the replies. I'll post a follow up later of what I did and problems that I ran into.
Riddell,
I ran into a problem. what i did was i clicked the front face of the part i am wanting to sketch on. right-clicked. edit sketch. i drew a simple rectangle to test the load projection. when i try to do a split line I get the message "Please select from the solid body you are editing". so it is asking me to only select the faces on the part that i am editing. i'm looking to project this force over a certain area which would include multiple parts.. i'm doing all this in assembly mode. take a look at the screen shot below. I selected the top bar's face to do the edit sketch but looking to project the load on the two bars for example...
Hmmm, I don't know how you might edit two parts at once like that. You can move sketches from one sketch plane to another but that won't help in this case. When you do sketch on the second part you get your alignment for the new sketch from the split line from the first if you don't want to bother dimensioning it.
So maybe I do have to save as part?
Earlier when I did that , I was able to split line the area of the projection onto the multiple parts. I think I must've gotten confused when combining parts because I was trying to make the whole inside door and frame as one solid body. I forgot to do contact set for the door edges and the frame (there is gaps). That is why it was giving me the last 6 solid bodies after I combined it down from 58..... I'll try to set contact as no penetration and node to surface later on the door surfaces and door frame. Maybe that might work..
What do you think?
If you've got gaps, especially if it is "one" part, I don't think you will get transfer of displacement and won't see transfer of forces/stress.
Best practice is to eliminate gaps, slivers, etc.
I think that is why I setup contact set with no penetration and node to surface between those gaps. The gap is where the door would touch the frame when the load is applied..
Hope im understanding contacts set correctly. ..
I think you are getting it correct. Remember though, the part will need to move before contact is made, then it will transfer force, etc. Sometimes it is better to mate the surfaces and set up a non-pen contact set because if a part has to move too far you'll fall into the non-linear FEA realm and you won't get expected results.
I'll give it a try after work. I'll update later. Thanks Riddell
So I was able to combine the solid bodies down to 4 parts. Those extra parts I can exclude from the analysis because they are just covers and a hinge with a pin which I can replace with a fixture...that should leave me with 1 whole body for the entire gate. I was able to mesh it but haven't gotten the time to run yet. I tried first to let the bodies that are not initially touching have contact settings with surface to surface but the time to run took too long. 40 min and 20% done. So I terminated that run and did more simplification...which was what I described above.
I also tried to do split line in assembly mode again. This time. I added a plain. Sketch an area for the force to be projected. My plan was to go in and edit each part with the same area project from the plain I just created. The problem now is, because the horizontal bars that are hit with the projection are the same part but used 6 times, if I edit one, all 6 will have the same split line. .. any way to work around this? Or would I have to bring in 6 bars of the same part but with different file name to let the program know that it is 6 different parts? I guess an example would be a picture.jpg. but I can have picture1.jpg, picture2.jpg. picture3.jpg, etc...but they are all the same picture still... I hope I make sense
I'll run my single body part later when I get home.. hopefully it works..
If you want split surfaces on different locations (only) for each part then you are stuck with part1, part2, etc. If it is no big deal, you can have multiple split surfaces on a part used in many areas. The only downside to this that I have noticed is if the boundary causes a point to show high stress that isn't really there.
I didn't build this model and I'm just trying to learn to do fea on it.
I guess I'm a little confused. Let's say I want to project a circle. In assembly mode, I edit the top bar which is a part that is used 6 times down a column. If I created a split on that bar, the split at the same location would then be automatically be applied to the other 5 bars because it's the same part... And that is not what I want. I want it to show just one full circle projection after I get done splitting the other bars using an outline from an inserted plain....
I think I'm just confused..
From what you are describing you have instances of the identical part which may (or may not) be oriented/mated the same way every time - it could even be flipped end-to-end. Each surface of the part has a 'name' and if you split the one, all the others will automatically show up. This can be a (small) problem if you have an assembly with properly mated parts, however, it probably won't impact your FEA, you'll just see some added geometry.
If you truly want only one part with a split circle you could copy the part into a new file, substitute the new part in for the one you need the surface on and then do your split operation in the assembly.
In practice, I will many times, set up a part with the surfaces I need for constraints/loads and add it to the assembly.
Yes, that's it.... Instances. It looks like I will have to save the part as a new part and put it in each time. I'll give it a try later if my 1 body part fails to run.
Is there any advantages/disadvantages in doing FEA in part mode vs assembly?
IMHO, assemblies allow much more flexibility. I can substitute parts, reconfigure (i.e. rotate) subassemblies, change materials, etc.
On the other hand, a single part can't move if properly restrained.
When I ran the simulation. It says large displacement was found. Click yes to proceed with flag for large displacements no to continue with small displacement...
What does this mean. Currently searching Google for answers..
Either something isn't 'nailed down' or the gaps you mentioned are too big. Sometimes you can set SW to ignore initial gaps but it is much better practice to make sure the model is robust and has no gaps, slivers, etc.
I have my fixtures in place and looks correct. The way I fixed the gaps were by using the dimensioning tool (measuring tape) to take measurements of the gap and then dimensioing the part or extruding by adding the measured distance....some parts I did it using the "up to surface" option.
Did I need to do something extra besides just making the gaps closed? Set up a contact set or coincident for those new touching surfaces?
When I ran it earlier. I went ahead and clicked no to use with small displacement. I was able to get the von mises plot and displacement plot. It looks right but I'm not sure of the results because Im not sure of what I did by clicking no.
Material yield strength is 34000 psi. And tensile is 45000 psi btw. And I'm loading 34000 lbf on the center area of the gate....So I'm expecting a failure with that much force.... I think
Occasionally SW doesn't automatically find surfaces for Global Bonding. I've had to take seemingly perfect geometry and do the auto search for contacting surfaces and set those up using Contact Sets. That *might* be the problem you are seeing.
If the values you got agree with your 'back of the envelope' calculations then they are probably OK. I trust nothing looked odd and there were no parts flying off into the sunset.
How do you do auto search for contacting surfaces?
And yes, the part look it deformed the way it should. No parts were flying into space..
I just remembered. There are no gaps now that I have combined the whole gate into 1 solid body. But there is a support bar that runs vertically in the center of the horizontal bars that wouldnt combine so I set up contact set as bonded for that.....
Is SW giving me the large displacement/small displacement option because of the amount of load being applied is causing the model to fail or displace by a number that is too big?
That is certainly one possibility, what deflection are you expecting to see? Is it within the realm of linear analysis?
"Failure" would show up as stresses outside the linear analysis domain.
You can use the option under contacts sets to search for connections within the entire assembly. However, if you have made this all one part then it will not find any and you'll have to search manually and fill the gaps. (Another 'pro' to assembly over part analysis.)
The max deflection I got is about 540mm. So about 1.75 ft. I think that sounds about right for 34000 lbf. I havent showed the results to my teammates yet. So we'll discuss that later.
I don't recall the size of your structure but it would appear as if you are in the non-linear territory. Whole different ball game.
I'm not sure of the size at the moment. But of the top of my head, I think it was 8 ft wide. 6 ft tall....
What I'm trying to do is, determine if the gate can support the 34000 lbf load. And it look's like it cant , unless I'm interpreting something wrong here... if it fails, redesign the gate to support it.
If you are deflecting 10% of your largest dimension then you are way into non-linear territory.
Yeah. My material yields at 35,000 psi and the max von mises is at a pin with 1.029x10^6 psi.. the door frame bars are at 343,200 psi...
.
Ha-ha, it blew apart!
Yeah. Hahaha
When I simplified the part down, I took out a rod that the door frame uses as a rotating piece to open and close the door. Imagine a square tube with a rod inside..the square tube being the door frame.. I think if I put that rod back in, deflection should go down. I'll have to find out. I took it out to minimize mesh and run time by taking out contact settings. Surface to surface contact set takes a long time..
If it still fails.. I'll have to reconsider a different material and then a redesign.
Hi james. Been awhile since I've been on this forum. Been running different simulations trying to get my part to pass. Unfortunately, the load is too much.
Now I am wanting to know how much force magnitude will cause the material or gate to yield.
Is there an option in solidworks simulation to find this automatically, or will I have to keep running simulations at different load to find its yielding magnitude?
Since linear FEA really only uses E & nu in the calculation all you need to do is apply the correct material first then, based on the max stress shown (watch for singularities or really odd shaped elements), you can choose the particular strength of material you need and you can refine from there if you have different modulus or Poisson's ratios. Don't forget to apply your necessary safety factor to the material strength.
Hi james,
I have the material of interest already but I want to know at what load will the material start to yield. The 34,000 lbf is definitely too high for the material. That is with the factor of safety.
Is there a way to find this magnitude on this material?
Or is the only option to keep running simulations with incremental load between each run?
I do not know, but I don't believe there is a way in Sim to do what you want. I know with columns you usually apply a unit load and the factor that drops out is the critical buckling load.
If you have the material set up accurately with the YS then your FOS should be based on that - but that's not what you are asking.
I think you will have to re-run and zero in on your goal.
Ok. Thank you for your time james. You've been a great help!