Yes, I agree with you. Just be patient with your learning curve; it is steep initially. For example, I don’t even try to use 2D elements with ccx because they are always expanded as solid elements with constrained dofs and knots. So, I try to avoid this by expanding the model from the beginning.
Good luck learning ccx!
2 Likes
This is a common issue in ccx: There are no true rotational degrees of freedom. Therefore, your rotational nodes, which I assume are 3228 & 3230, only take DOFs 1, 2, and 3 for UR1, UR2, and UR3… Your 3227 & 3229 are the same for your U1, U2, and U3 displacements.
thank you for your courage!
So, from hat I understand it is easier to simulate a 3D case with asinle element in the thickness than a 2d plane stress problem?
In that case I may try this before trying to make work the 2d plane stress
Thanks @jbr
Sorry, I did not understand which is the common issue in ccx.
I block displacements in my 3227 and 3229 right, and rotations in 3228 and 3230.
The (x,y) coordinates of nodes 3227 and 3228 and 3229/3230 are the same
It seems that applying the torque is a bit unstable, in abaqus you can fix it using *CONTACT CONTROLS, stabilize but in CCX you need to tweak the model to make it work. So far I was unsuccessful. More elements in the contact region may help as in the convergence table the number of contact spring changes wildly:
Have a look to this thread on stiff contact problem: Snap-fit contact snagging problem
In ccx, there are technically no dofs 4,5,6 since all elements are expanded to 3D solid elements.
Yes, that is what I would do. Expand your 2D PS model to 3D and apply proper boundary conditions in the appropriate faces.
Hi,
Is Moment a follower load?
I mean, once the moment is distributed and translated as equivalent forces, ¿do they update direction as the gear rotates?
I’m obtaining weird results when the problem is solved by Moment driven.
EDITED: I have figured out a way to test if moment updates when rotating and it does. My problem must be something different sorry.
The plate is simply supported in the cube with a contact. It struggles at some points, but a constant moment is able to keep both in contact and bending is not changed.
EDIT: I tend to define as few master surfaces as possible. In this case it is not appropriate. Defining a contact pair in each space solved my problem.
(Constant moment on the Gear REF and Constant displacement on the RACK REF. Moment follows and contact pressure is automatically adjusted by ccx to preserve the required moment. Nonlinear geom quasiestatic.)
1 Like
Helo @Disla.
Just to see if I properly understood your suggestion. You are proposing me that a better ressult may be obtained if instead of creating a single contact pair between all the contact surface of the pinion and the rack, I generate one contact pair for each teeth in the pinion and rack??
Yes. The force needs to be perpindicular to the gear flank in contact. so I would say it is a follower load. Since programming this I think is more comples, that i swhay I always simulated by applying a torque (moment).
I do not understand what you said here:
Right, that made a difference in my simplified model. I’m not saying you can’t solve it with just one master and one slave but I couldn’t find the way to progress with that approach.
Action will induce reactions in perpendicular direction to the surface (generating contact pressure) and parallell direction to the surface (generating longitudinal and transverse shear).
By displacement driven I mean actions are imposed displacements.
By moment driven I mean actions are imposed Moments.
Convergence is typically more difficult in this last case.
My aproach is a mixture. First moment to engage with fixed rack and later I move the rack so the gear must follow to keep the system in balance of forces.
In the full model the contact opens, gear doesn’t follow rack anymore but keeps rotating ?¿?¿ and finally gear stops ?¿?¿?. Strangely *NODE PRINT for the ROT still shows there is moment in the gear?¿?¿?. Check Vid. Something is very wrong here. Looks like there is a breack somewhere in the gear.
I’m still working on it.
1 Like