Hi all
I have a problem modeling a connection in an assembly, I isolated the two parts for the example:
- I have a fixed part with a curved tongue, in which a crown with a curved groove (tongue within groove) is placed. On each of the 3 faces of the groove/tongue, the contact is set to No penetration, surface surface, no play or friction. Theoretically, this curved slide link gives a pivot link of center to the center of curvature of the groove, except that it doesn't work properly.
- Statics 1: To test the behavior of the two parts, I apply a force of 100N on the side of the crown to make it rotate, and I block the rotation on the other side with a spring with a stiffness spring of 20N/mm. So I should have a displacement of my crown of ~5mm and an axial force on my spring of ~100N, but this is not the case at all...
- Statics 2: By replacing the non-penetration contacts with springs, I get a consistent behavior
Do you have an idea? I'm a little desperate to be able to get the simulation module of Solidwors to work properly... Assembly + simulation are in PJ solidworks 2016 with the results.
Kind regards
2020-04-22_glissiere_courbe.zip
Hello
I have a bit of trouble understanding what you want to do because you don't use a coefficient of friction.
You have if I understand springs on the opposite side to the 100 N force.
This is equivalent to having a portion of the bearing on which we would like to measure the loss due to friction. but without defining the coefficient of friction (gulp!)
In addition, you want to do almost dynamic while you probably don't have the PRO version which is the only one that can do dynamic.
In any case, I don't believe that simulation is used to do what you want to do, especially under the conditions you describe. If you don't set distance limits to the normally rotating movement, SW should offer you a trip to the Parthenon.
I may not have understood anything, in which case I apologize. ;-)
Kind regards
Hello
Thank you for your answer, the assembly shared here is an example, it's a small piece of a more complete assembly, it has no other purpose than to check that my load descents are consistent so that my simulation is valid in my more complex assembly.
Precisely, there is no coefficient of friction and yet I find almost no force on my spring when it comes out, it's good that the input force was dissipated somewhere in the link when it shouldn't have, which makes the link unrepresentative.
Ditto if I remove the exit spring (which compensates for the input effort), SW doesn't offer me a trip to the Parthenon at all, the part moves slightly, proof that something is blocking the movement when it should be leaving it free.
The simulation, on the other hand, by replacing the contacts with very rigid springs, gives very consistent results:
- The input force of 100N is found on the output spring
- The spring rate is 20N/mm, as we put 100N as an input the part must move 5mm, which is the result obtained
I don't know if the zip of the assembly was transmitted on my 1st message, I'm putting it back here
2020-04-22_glissiere_courbe.zip
Hello @benjamin.perez
I don't understand your approach at all, nor why you use simulation for this. Because there is no load descent in your case, because it is like wanting to measure the loss due to friction in a ball bearing (since you do not put friction)
But hey, you have to have your reasons.
For Static 1 no debate since force at the input minus against force of the spring at the other end The travel limited to 5mm is since the two parts Crown_up and block are spaced 5mm apart and come to stop without penetration on the pyramid tips.
Note that your connections are redundant (excess) but have no effect on the result
Attached is an image of the slightly more complete assembly:
- The green piece is guided in a pivot to the grey part thanks to the two top and bottom crowns
- Efforts are applied to the green parts through a mechanism not shown here, the FEM calculation is used to size my mechanism, my grooves....
For my calculation to be valid, the groove modeling must behave correctly. The pink and blue pieces should slide on the grooves of the gray piece, but this is not the case at all, so I made a submodel to try to investigate what doesn't fit. Is it clearer?
Regarding Static 1:
- Inlet force 100N, if the modeling is correct the two parts should slide over each other without loss of effort (no friction) and compress the spring until the force on the spring opposes the inlet force.
- I should therefore read 5mm of displacement and 100N in the spring, but here I have 0.16mm of displacement and -3N in the spring. Removing the spring would give exactly the same thing, which is not normal.
- The 5mm spacing does not change anything since the contact is not declared in SW, there is no stop between the two pyramids. I'll have the exact same thing with 50mm of spacing.
porte_couronne.png
Hello
To solve your problem, you need to deal with it in the full model with the two standard SW tools, namely collision detection and clearance checking.
After inspecting the model by changing the current constraints which is the correct way to see the constraints in conflis.
To me you seem a little stubborn or stubborn but the simulation won't tell you anything about where it gets stuck because from the beginning you will have a message about the collision without it telling you more.
If you really don't get out, then give a link of your ASM on any drive in private message I can take a look. It must not be very complicated for a new eye ;-)
Kind regards
PS:
[HS On]
I cordially try to answer your request patiently about the simulation, but I can assure you if you allow me to offer you an opinion, well: as I see your model in the last image you will have buttresses obligatory. Thus, and even with a correct design, nothing will slide unless there are copious clearances and the green and grey part does not deform under any load, I am not even talking about the welded mechanic.
;-) ;-) ;-)
[HS /Off]
No problem, thank you for taking the time to take an interest in the problem. I'm sending you the ASM in mp, tell me if you can unravel the problem.
The guidance is much longer, I cut it for the simulation and to avoid very long calculation times with the contacts, the time it takes for the model to seem coherent. Indeed the guiding principle is a bit atypical, that's precisely why I would like to behave correctly in these places to see what is going on.
Hello Benjamin,
For me, your simulation has no problems. Static versions 1 & 2 give more or less the same values.
I think your error comes from the fact that you measure in "static 1" reaction forces (for connectors) when they are given as "reaction forces" (SolidWorks 2019 help). This works for "static 2" because the springs are connectors.
PS: we see 3.92 N of tangential force on the bearing, probably because the direction of application of the spring force changes and not that of the force on the face.
M.
In fact, I think that this effort modeling is not adapted to your problem. If you replace the two forces (on the side side of 100 N and the top side of 1 N) with a remote force according to the following parameters, you will have the result you were hoping for:
- With a pure torque component of 34 N.m (100 N to 340 mm; midpoint of the slide)
- With a "vertical" force component of 1 N
- (In the screenshots below, I added a tangential force of 10 N for the form)
This gives me a spring displacement of 4.72 mm, which corresponds well because it is located 360 mm from the axis of rotation (5 / 360 * 340).
M.
1 Like
Hello mgoroy,
Thank you for your answers,
I have the impression that we have different results on the same simulations...
For your 1st message, how do you get the table indicated?
- I have 0.16 in displacement resulting
- The detail of the contact/friction forces on the 3 surfaces of the groove give results that I can't interpret: +/- 5e3 on Z? (they compensate on both sides but too high intensity); the sum on X gives -147N while it is announced -86N in the summary table
- The summary table is correct, the projection of my 100N at the input gives 86 in X and 46 in Z, the result is close to 100N.
For the second simulation, ok I understand, but again we don't have the same results, I have exactly the same thing as on static1. In addition, the application of my efforts in my main assembly is close to the configuration in static1
2020-04-22_glissiere_courbe_2.zip
Hello Benjamin,
The table I give is simply a summary of the measurements made with SolidWorks and the projections calculations made with Python/Jupyter.
For displacement measurements, I use the displacement result probe tool on the node at the end of the diamond shape, where the spring is attached. For the force measurements, I had just looked at the contact force on the inside of the slide. I redid the measurement on all three sides this morning and it's still good.
I don't explain why you get a 0.16mm displacement. This morning I downloaded your files again and launched them directly before making any changes and I get the same results as yesterday.
mgauroy,
Which version of SW are you on? Is it possible to upload your assembly with the result file for static1 please?
Hello
I think I'm dreaming!
If the problem is as it seems to appear that this the problem is actually that the slide is not sliding somewhere. In this case I maintain that it is a collision detection in the model that will indicate where the mechanical problem is.
If I understood correctly, this would confirm that the simulation does not have to do anything to solve the problem.
Or it is an exercise in style to better understand how the simulation module works, in which case the statement and the search can be expressed differently.
Kind regards
I'm on SoldiWorks 2019. I attach my files as a PC.
2020-04-22_glissiere_courbe_mod.zip
mgauroy,
Thank you for your file,
very surprising, my results differ again when I perform the simulation by launching your file under SW 2019 (I find 0.16mm)... It's rather worrying!
