hexagon logo
Register Sign in
  • State Not Answered
  • Replies 4 replies
  • Subscribers 11 subscribers
  • Views 344 views
  • Users 0 members are here
Options
Related

Static equilibrium between rigid/modal simulation

sylvain.grosroyat
Hello,
 
The system i simulate is a propelled vehicule that is composed of two parts : first is a rigid part, the second is a flex part.
 
Until now, i started the simulations with a sim/statics command, in modal representation => the result is the nodes of the flex part move to the equilibrium positions before the transient simulation.
Good idea (or not ?...) : i use a fixed joint, only during the sim/statics command, to link my vehicule with the ground. Without the joint, the static analysis is not successful. Note that the 2 parts are linked by bushings.
 
But now, due to controls consideration, i need to do few seconds of simulation before. As there is no interest to simulate this part of the simulation in modal mode, i tried to :
  1. realize the simulation in rigid mode during few seconds
  2. then activate modal representation of the flex part
  3. realize the sim/statics command in order to obtain equilibrium of the flex part,
  4. realize the rest of the simulation
But step 3 is never successfull, even playing with activate/deactivate commands of the fixed joint and bushings.
 
Do you have any advice please ?
 
Thanks for your help.
  • 0
    It is always a good idea to get static equilibrium before you start the dynamic portion. But not always possible.
    Using a fixed joint might be overkill. In many cases a "stake" between the flexible part and the ground is enough. This stake can be implemented using a in-line jprim that prevents the x and y motion, but allows for all rotations and the motion along z (assuming that z is your vertical axis). This is how most vehicle model are implemented. Sometimes you would also add a perpendicular jprim to look the yaw motion.
     
    In the process you describe above, after the first few seconds of simulations, do you move the j-marker (on ground) to coincide with the i-marker (on the moving/flex part) before activating the fixed joint?
  • 0
    Thank you for your response.
     
    Yes, I created a i-marker attached to the ground, relative to (positions&orientations = {0,0,0} ) the j-marker i use for the joint. As the rigid body and the flex body are linked by bushings before separation, i tried as well with the j-marker attached on the rigid body, and with the j-marker attached on the flex body.
     
    Regarding the joint, as it's a space propelled vehicule, i tried fixed joint, jprim at point and spherical joint (even if it might be equal to the jprim at point).
     
    I wonder if it is really relevant to realize such static step as the vehicule speed is quite high...
    I tried all day but, so far, it only works (3 iterations) when the sim/statics command is at the beginning of the script.
     
  • 0
    I don't think you got the important part of what I was trying to convey.
    I assume that the vehicle moves in the initial dynamic simulation?
    Then when you activate the fixed joint, that will pull it back to the initial position and literally tear the system apart.
    So your script MUST look like this (assuming your fixed joint has id=13, the i_marker has id 42 and the j_marker has id 43.
    deactivate/joint,id=13
    simulate..... !initial dynamics
    ! do whatever you need to do to activate the modal representation
    marker/43,qp=0,0,0,reuler=0,0,0,RM=42 ! Move marker 43 (on ground) to coincide with marker 42 (on flex body)
    activate/joint,id=13
    sim/stat
    simulate.........
     
     
  • 0
    And of course, there can't be any external forces acting on the system (trying to propel it) when doing the static.
    The static will also set all velocities to zero, so if it was moving with a high velocity before the static, it will have stopped after the static.
© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement