The color of the sphere (in the case of a beam) is yellow instead of magenta, which indicates that the ends of your two tubes are not connected to another element, beam or other.
Moreover, if I see the diameter of the spheres, they are too small to make any connection. The diameter of the hypothetical spheres defining the bonds must be increased. (see in the connection group)
Could you make another image where the whole tree would be expanded so that I can see the parameters (or make a pack and go of your ASM so that I can look)?
A question independent of the cause of the message. Why do you put both global contact and also contact between sets when you actually only have one set. From what I understood (if I'm not sucked) is that there is a big difference in the way you calculate if you have a single ASM (which is your case) and the calculation of several subassemblies in a global ASM. In the second case, which is not your example , it would be appropriate to put contacts between ensembles.
[Off-topic On] Why use a mixed mode instead of treating everything in volume mode. Personally, I would only do mixed mode on large boilerwork where there are real beams with upper sections 200 mm x 200 mm with sheet metal. Typically a cyclone with a mix of sheet metal and metal structure (or metalwork boilermaking). The mixed mode is only interesting to avoid having too many elements and therefore multiplying the number of nodes and DDLs for the calculation. In your case the gain on the speed of calculation will be AMHA not significant.
I suggest you go all in volume for such small pieces, it will be easier: unless you have a particular interest later on to stay in mixed mode. [HS /Off]
I will try to clarify my words with the details below:
- I'm on a multibody part (profile + sheet metal) not an ASM!
- The yellow sphere is a visual element that indicates a node not connected to another beam, so that the user thinks to create an additional contact. In my case, I did add a contact with the sheet metal at each end of the beams. (the sphere always remaining yellow)
- No point in my case to increase the diameter of the sphere, since no node needs to be merged with another.
- The global contact is generated automatically. but the overall contacts are necessary between elements of different types (here: shell and beam)
- I have made here a simplified piece to isolate and understand my modeling problem. My part being much more complex with mainly sheet metal, I don't even plan to go into volume (especially with material thicknesses of 3mm, I would need tiny meshes...)
I'm attaching an image with exactly the same model, the same conditions, but cylindrical beams (the spheres remain yellow!) But as soon as I switch to a square or rectangular profile, the model doesn't rotate anymore!
Your problem comes from the way you make your contacts between assemblies (you have to associate the ends of the beams with the external face of the sheet metal and this for each sheet). The handling is less simple in the case of mixed mode.
I'm attaching my 2019 version of my model.
As you are in Version 2020 I could not see that you had used shells and I did not pay attention to the fact that you were in multi-body while your attached file was a PRT.
I hope I have answered your question.
Kind regards
PS for the mesh I used 89mm for the max and 4.76 for the min
Indeed your solution works. But in the long run on my complete model, some sheet metal will not necessarily be at the end of the beam.
After exchange with the hotline:
- It's the "global contact" that is a problem, because all my "contacts between ensembles" are enough to fully define the model. Just by deleting it, it works.
- Another alternative: transform the "beams" into "shell" elements and add a single global contact to make the model work.
