HI
I tried a test consisting of two plates connected by a bolt to get the forces on the bolt.
a) In the first case I applied a surface force using a rigid body on the upper plate, but the result was not consistent as the upper plate was unloaded. I don’t understand why. In the attached model I have deactivated this behavior.
b) To the same model I applied the forces on the upper part of the plate as surface forces (Fx= Fy = 50000 N)
On the bolt the recation forces are in balance with the external forces but an Fz also appears on the bolt (Fz= -15780.37 N) which does not seem to me to be balanced by the contact forces on the surfaces of the two plates, I don’t understand why.
Thank you for any clarifications you can provide me
Agazzotti
Ag_1.pmx (5.8 MB)
I can’t check your file now but if you share some screenshots of the model (to show how it’s assembled and loaded), I might be able to give you some tips. Sharing the .inp file would also be helpful.
I’m mostly wondering how you modeled the connection (shell plates and solid bolt?) and how you measure the forces.
All elements are modeled with solid elements
Force:N
Dimension: mm
Uploading: Ag_1_file.zip…
It seems that the upload failed, probably due to clicking Reply too soon (before uploading was completed).
I resend the files
Ag_1_file.zip (4.5 MB)
Thanks. Is the bolt split into two parts to be able to apply preload to it later ? Because it might be better to try with a single/compound part for the bolt, only having an internal surface in the middle to request a surface print with SF (instead of a node print with RF). You will need keywords for that though.
In addition to FEAnalyst suggestion, sum all contact pair forces component by component. I mean, look at all the contacts acting on a component and add them. They should be in balance with the external load.
NOTE: this problem is like a bottle opener.
EDITED: I have check the force balance and ccx gives the CF for the Slave surface not the Master as I assumed. I guess that’s why contact pressure is also given for the slave. Not an issue but that changes the sign criteria with respect what I said. (I was wrong sorry) .
Negative for the master and positive for the slave.
Thanks for your replies.
I ran further checks and got the following results.
If we consider the lower surface of the upper plate as the MASTER element we obtain a total contact force on the surface equal to Fz= 14082.58N which does not balance the force on the bolt Fz= 15857.60N (error = 11.2% not acceptable).
If we consider the lower surface of the plate as a SLAVE element we obtain a total force Fz= 15857.6N which perfectly balances the Fz of the bolt. I don’t understand why. To avoid incorrect evaluations, it would be better not to provide the results of the MASTER surfaces as they would appear to be incorrect.
Previously I had also verified that if the surface force is assigned through the rigid body the plate is not stressed and that therefore the rigid body command not only serves to distribute the loads but modifies the elastic behavior of the element. If this were the case, it would be appropriate to inform users of this aspect
Upper plate balance involve one external action and three contacts. TIE contact must be extracted separately with node print. Contact print do not recognize it as there is no contact pair.
I have recently update my previous post. CF extracted with the Contact print card provides the contact force at the Slave . Master CF= -Slave CF . MAster or Slave definition should not affect the overall force balance (Althought I experienced some issues with two masters defined on the same element in oposite faces).
Yep. It becomes Rigid. That should not affect the equilibrium equations ¿isn’t it?
No. The stresses change because the contact surface between the two plates (one rigid and the other not) are different; in the specific case the action Fz increases by approximately 9% considering a plate as a rigid body. For this reason I believe that users must be informed of this aspect in order to evaluate when to insert the rigid body or not.
I forgot… as already indicated the forces change considering the SLAVE or MASTER surface with a difference of approximately 11%. I don’t understand why.
This is what I get. It matches. I have centered the model to see if I can understand the moments (They are given about the origin)
EDITED: Bolt Top and Bottom sorted in their right position.
Section print don’t provide as accurate results. Maybe that’s the origin of your discrepancy. Marked in red colour.
It’s explained in the CalculiX User’s Manual but IMO Abaqus documentation does it better:
A rigid body is a collection of nodes, elements, or surfaces whose motion is governed by the motion of a single node, called the rigid body reference node. The relative positions of the nodes and elements that are part of the rigid body remain constant throughout a simulation. Therefore, the constituent elements do not deform but can undergo large rigid body motions.