Applying a moment couple to ends of a cylinder 1

[Techsan123]

I'm trying to twist a cylinder by applying a moment to each end in opposite directions. The error message I recieve tells me that there are Inactive DOF for which BC are specified. I tried not using any BC's just to see what it would tell me and I still get the same error. I was able to get the twist by just applying a bunch of concentrated forces but now I would like to incorporate a moment because the concentrated forces deform the ends too much.

 

[gurmeet2003]

This problem has been addressed before. Create a rerence point and then a kinematic coupling between this point and the surface. Apply the moment on the reference point. It will be transmitted to the surface.

Gurmeet

 

[Techsan123]

ok, I had tried this but I applied the kinematic coupling to the cylinder base faces and then applied opposite moments at each reference point but had no luck. I guess you are saying apply it to the lengthwise surface? Also, what about the BC's? How can apply the BC's without causing unwanted distortion? I'm trying to get the cylinder to twist and possibly kink in the middle.

 

[Techsan123]

I applied the kinematic coupling to the lengthwise surface of the cylinder and applied the moment to the reference point. I went ahead and put U1,U2,U3,UR1,UR2,UR3 all equal to zero on the opposite side of the moment. I tried various moment loads varying from .02 to .2 but I keep getting error stating too many attempts for this increment. The results show that the cylinder moves very slightly. Any suggestions on how I can fix this. I am aiming for the cylinder to twist somewhere along the length.

 

[gurmeet2003]

Techsan,
The cylinder contains two circular surfaces and one cylindrical. Fix the edge of circular face on one side. Define a reference point on the cylinder axis but outside the cyliner towards the free surface. Connect this reference point to the nearby circular face with a kinematic coupling. The cylindrical face is free. Apply the moment on the reference point along the axis. I am assuming that you want to study the cylinder's twisting under torsion. Also I am assuming that your material definition is linear.

With above constraints you will get artificially high stresses at the fixed bc. Ignore it. The stresses in the middle of cylinder should be realistic.

Gurmeet

 

[Techsan123]

Gurmeet2003,
First of all let me say thanks for taking the time to help me out with this problem. Your solution does work but it does not allow the cylinder to twist until it buckles which is what I need. It will abort before buckling even when I adjust the step time increments. Is there another way? I have also tried applying the moments directly to nodes along the circular base surface but this is where I would get the error "Nodes have inactive DOF for specified BC".

I have been able to accomplish this by applying concentrated forces as a couple directly to the one end of the circular base and also by modeling the cylinder with a rectangular base at the end and applying a couple to the rectangular base but I need to incorporate the moment feature.

Thanks again

 

[gurmeet2003]

Techsan123,

I do not understand the engineering problem you are trying to solve. Why do you expect the cylinder to buckle? Are you talking about torsional buckling? Please explain.

Thanks,

Gurmeet

 

[Techsan123]

Gurmeet,
Yes. Think of a rubber hose with one end being fixed, as you begin to twsit the hose from the other end eventually it will begin to deform, kink, or twist (whatever you want to call it) somewhere along the long axis. This is what I am trying to accomplish with the moment feature. This occurs easier with a short hose of course.