General Mechanical

General Mechanical

Topics relate to Mechanical Enterprise, Motion, Additive Print and more

Reaction forces of moving SCARA robot

    • SC83


      I have to design a table for a SCARA robot which will be mounted on top of this table. I'm not allowed to bolt the table (consisting of 4 legs bosch frame and aluminium top plate) on the ground (drilling of holes in the ground is not allowed). I need to know if the table will stand still or shake/move under the acceleration forces of the robot. The table should also not tip over under the acceleration forces of the robot.

      I have a NX design of the table with top plate. In the top plate are four mounting holes for mounting the SCARA robot.
      The SCARA robot consists of three joints/arms - see attached - and translates/rotates the end effector in the XY-plane.
      I have the masses of the three arms, their lengths and their COG. I know the max. linear (radial) acceleration and rotation (theta) acceleration. I don't have detailed/CAD model of the SCARA robot.
      I also know the reaction torque of the motors which rotate the arms.

      How should I model this in ANSYS? I only need to check the worst case situation, i.e. stretched arm.

      The robot cannot move in the vertical (Z-direction). Thus only in the horizontal plane.

      I have a license for static structural and modal analysis in ANSYS. Any suggestions are welcome.

    • Sampat Kumar
      Ansys Employee

      I would recommend using the Rigid body dynamics to simulate this model. You can download the student version of the ansys and you can use this Rigid body dynamics module. you can download the Ansys free version from the attached link.
      you can find a number of relevant resources regarding the simulation of the robot in the rigid body dynamic on Google and youtube
      Rigid & Flexible Transient Dynamics Analysis in ANSYS Workbench - e-Learning

      You may follow the steps mentioned to simulate this model in Ansys Structural.
      1. Try to use a revolute join in the connection between the arms for free movement of the arms.
      2. Please try joint load as the input BCs for the joint to provide the rotation and moments as per your requirements.
      3. You may apply for frictionless support between the table legs and the ground surface. A frictionless reaction probe may help to calculate the reaction force between the table legs and the ground surface. If the direction of the reaction probe changes, then the table may topple.

      Please let me know if it worked or not. 



    • SC83

      Hi Sampat,

      Thank you for your reaction.

      I prefer to use Ansys static structural if possible. I have defined the mass of each arm with a point mass in its COG (see screenshot and blue dots in the top view) as I don't have a CAD model of the robot. Each mass is connected to the red colored surfaces (this is the interface between the robot and table) via a named selection. I want to simulatie the most extre condition, i.e. all arms in line rotating around the Z-axs. For this I have calculated the inertias of each arm as:

      • I_arm_1_Z = I_arm_1_COG + m1.r1^2
      • I_arm_2_Z = I_arm_1_COG + m2.r2^2
      • I_arm_3_Z = I_arm_3_COG + m3.r3^2

      I which r1, r2, r3 is the shortest disctance from each point mass to the Z-xis.

      Or should I define for arm_2 and arm_3 the distance r2 and r3 wrt to their own rotation point, i.e. the joint_theta_2 and joint_theta_3?

      When I support the table on its legs using frictionless support, I get the following error: An internal solution magnitude limit was exceeded. Please check your Environment for inappropriate load values or insufficient supports.  Please see the Troubleshooting section of the Help System for more information.

      The model, currently without any accelerations - thus only gravitational forces, solves when I use fixed supports. Acutaully, there is friction between the legs of the table and the concrete floor (XY-plane). How can I add this to the model?


    • Sampat Kumar
      Ansys Employee


      If you are using point mass and directly connected to the base of the robot then it may also be good for your purpose because you are only interested to calculate the topple of the table when the arm would be fully stretched. 

      You may apply a spring with very less stiffness between the legs and the ground surface to avoid this unconstrained behaviour.

      I would suggest using a transient structure to better capture the Dynamic behaviour of the robot arm and its impact on the legs of the table.





Viewing 3 reply threads
  • You must be logged in to reply to this topic.