## General Mechanical

#### Pull out Simulation in Transient Structural

• boodoo.a.aa
Subscriber

Hi everyone,

I am trying to run a simulation of vertical pull out of a plant root out of a soil. The soil in this case is represented by the Mohr-Coulomb model. There is a frictional contact between the root and the soil. The simulation runs successfully when i apply a displacement condition to simulate the pull out but if I apply a force to pull out the root, then the solution does not converge.

• peteroznewman
Subscriber

Use the Displacement constraint.  You can Probe to get the Reaction Force for the Displacement.

Why are you solving in Transient Structural? You can solve the Displacement constraint problem in Static Structural.

Please reply with an image of the Force vs Time plot for the Displacement constraint.

If you insist on using a Force, use Transient Structural, and add a huge point mass at the top of the plant root to slow down the acceleration when the root lets go of the soil. You can subtract the effect of the point mass on the force by subtracting m*a where m is the point mass and a is the acceleration of the root.  But remove the  gravity load if you had it there in the first place.

• boodoo.a.aa
Subscriber

Thank you for your reply peteroznewman. Please see attached the force vs time plot for displacement constraint. I was using transient structural to (hopefully) do a FSI coupling with Fluent later on. By chance, do you know of a way in which I can maybe stop the force after a certain allowable displacement ie at the moment right before the root gets pulled out of the soil? I really appreciate the advice!

• peteroznewman
Subscriber

Notice that the force to "break free" the root is 7.46 N and the force after that is less, but you can slowly measure those lower forces because the displacement has a linear ramp over time.

If you apply a Force to the root instead of a Displacement in Static Structural, not much happens up to 7.46 N then there is no equiliburium after that point because the root broke free and is unconstrained, that is why you can only solve it in Transient Structural.

If you solve the model once, you know when a displacement threshold was crossed, so you can manually set the end time for a second solve to the value you looked up on the first solve.  Sorry, I don't know how to do it programatically.

• boodoo.a.aa
Subscriber

Thank you so much for all your advice peteroznewman. I will try everything you suggested.