How to model a coupling between two shafts revolving

Hi, I am using Workbench 16.0 and I want to model a coupling between two shafts (A & B). Shaft A is attached with a revolute joint in another body and shaft B is supported with a Body - Ground Revolute Joint. 
A Moment and a rotational velocity are applied to Shaft A and through a coupling, they pass to Shaft B. The coupling allows for a max radial and axial displacement and an angle displacement. So I want to find a way to get results for these displacements in between the two shafts.
So I believe that I have to let free all DOFs between the shafts except the Rotational one around their axis (Z-axis). I tried the General Joints but when it comes to RDOFs it allows only for one or all of them to be free.

So if I use the General Joint and let free only one DOF (for example X) I won't get the right results since there should be rotation around the X & Y-axis. If on the other hand, I let all rotational DOFs free the Moment and Rotational Velocity will not pass to Shaft B. 

These two shafts are a part of a bigger assembly that is subjected to other forces and so it is expected to have deformations on its own. The purpose is to get the relative motions (X , Y , Z , ROTX , ROTY) of one shaft to another while letting the Moment and rotational velocity pass from A  to B. 


  • peteroznewmanpeteroznewman Member
    edited January 31

    Are you using Static Structural?  If so, Shaft B should not be free to rotate about Z.  You are applying a moment to shaft A which is transferred through the coupling to shaft B.  The moment must reacted to by a fixed rotational Z DOF on shaft B.

    You used an Inertial Load called Rotational Velocity which applies a radial acceleration load to the mass in the model from the axis that is proportional to the radius and the rotational velocity squared. It does not rotate a body.  All bodies in this two shaft and coupling model can have a Rotational Velocity load, but if the shaft speed is low and the radius is small, the stress that this will create will be insignificant compared with the moment load.

  • KTzanKTzan Member
    edited January 30

    Peteroznewman thank you for the answer but I am not exactly sure what you mean so I will be more specific this time and I will also upload the model. 

    A Hydraulic Motor (A) gives motion to a Pump (B) through Shaft C that belongs to a Torque Meter. Two couplings are connecting shafts A-C  and C-B. Connected to shafts A & B are the masses (each with a Moment of inertia around Z-axis) of the rotating parts of the motor/pump. The motor and pump are hanged on vertical brackets with bolts and thus the Moment and Weight are transmitted to the brackets, which are deformed as a result and that deformation leads to the relative displacement of the shafts. 

    This model will be used for static structural, modal and dynamic analysis. 


    1)  "The moment must reacted to by a fixed rotational Z DOF on shaft A." 
     I don't know how to fix just one Rotation DOF.  Can you please guide me? There is no joint that performs that task. 

    2)  "Are you using Static Structural?  If so, Shaft B should not be free to rotate about Z.  You are applying a moment to shaft B which is transferred through the coupling to shaft A."  
    Since the shafts are rotating in the real world, why they should not in the model? And what kind of support do you suggest that I use?
    Will it be the same for Modal and Dynamic analyses?

    (The mesh is of no concern)

  • peteroznewmanpeteroznewman Member
    edited January 31

    Your file shows a structure at the +z end called YK that has a platform under the center shaft and another structure at the -z end called ANTL.  Which end is the pump and which is the motor?

    Why is there a revolute joint from the platform up to the center shaft? Is that for the torque meter? Please provide more information about the torque meter.

  • KTzanKTzan Member
    edited January 31

    YK is the motor and only its face is visible, The rest is represented by masses A, B 
    Mass A is the mass of its case attached to its face
    Mass B is the mass of the rotating parts attached to the shaft 

    ANTL is the pump and the masses go likely .

    The shaft in the middle is of the torque meter. Only the weight of the torque meter is known and it is substituted by the force (200N) on the platform.
    I believe that the remote point which is the reference geometry of the revolute joint is supposed to be coupled with the face of the platform in order for it to follow its displacement, but i am not sure.

  • peteroznewmanpeteroznewman Member
    edited January 31

    Here is a Static Structural result for the 3000 N.m torque in the system.

    As you can see, most of the deformation is in the shaft, that gets wound up by the torque. If I plot the deformation without the shaft, you can see how small the deformation of the bracket is.

    To get this model to run, I changed the two General Joints to Fixed Joints, I suppressed the Revolute between the platform and the center shaft, since the fixed joints were now there to support the center shaft. I suppressed all the moments.  I created a Joint load on the motor side of 3000 N.m on the Revolute, while the Joint load on the pump side was Rotation and that was set to 0.  That means it was effectively a fixed joint, but with the ability to monitor a reaction torque.

  • KTzanKTzan Member
    edited February 7

    Sorry for  the late answer. 
    Eventually i restricted the ROTZ Dof  between the shafts with a Constraint equation and also used a General Joint between them with All DOFs free  as shown in the picture below



    I would like to ask one more thing. Is there any conflict between the General joint and the Restrained Equation that might lead to wrong results?

    Thank you

  • peteroznewmanpeteroznewman Member
    edited February 7

    A General Joint with all DOF free provides no constraints. You should get the same result after you delete the joint from the model.

  • KTzanKTzan Member
    edited February 7

    Thank you Peter. I've read a lot of your answers to many subjects and posts. You've been very helpful every time.

Sign In or Register to comment.