Hello,
I am trying to model a threaded joint in Abaqus. I would like to study the impact of the initial acceleration on the sliding behaviour of the joint. I have taken the contacts into account with a classic Coulomb formulation. The first step is applying the preload with a bolt load. The second step is applying a transverse displacement. During this last step, the bottom part is fixed while the upper part is moving and I look at the tangential load.
My problem is the one you can see on the curves attached. I have a "strange" initial behaviour, some kind of discontinuity, at the beginning (2nd increment !) of the tangential load curve. I don't know if this is physical or not and if not, if there is a way to fix it...
Do you have any clue ?
Thanks a lot.
Rémi
Probably it is due to the mesh discretization which creates small peaks and crests. Try to include an initial phase of approach between the geometries.
Dear Sir
Very interesting pattern. To try to solve problems, maybe you should read the: https: //sites.google.com/site/mechanicalengineeringforlife/FEA/typesofcontactsinansys
where it says:
"Surfaces are allowed to slide by user defined coefficient of friction (Mu = user defined) and can open and close. This is real like having two surfaces sliding against each other. Sliding will occur after the threshold resistance value of shear stress depending on the Mu. ) 0.2 is considered as the ‘magic number’ in FEA world. Any Mu below this value would converge easily. That has to do with use of Unsymmetrical Eigen solver like Lanczos which is covered in another post."
Very good the Ariadne Sambani example, the last file.
Interestingly, if your model contains profiles spiral thread. In this case there will be slipping on the wing and will be a natural friction.
Sincerely Yours
Cardei Petru
Probably if the preload is too important and applied that a direct constant value, for several computing time incremens, for several points the shear stress value reached exacly the plastic limits Tau_max and consequently differences occur between mu*normal stress for the other points and Tau_max for the previous points. This phenomenon can explain osccilations of the tangential lods. To avoid this problem and to obtain good convergence solution, small time increments must be used togthether with a short time increase of the preload until the imposed value.
- Badges
- Science topic