Hi, I'd like some advice regarding applying initial velocities to the LSTC rigid body models ( https://lsdyna.ansys.com/lstc-dummy-models-overview/).
Usually, I'd just use *INITIAL_VELOCITY_GENERATION and assign an initial velocity to a group of deformable parts in a part set. However, in this model, there are many rigid bodies, discrete elements and joints. How would you go about adding an initial velocity to the whole dummy?
I'd like to do a simple test where I add an initial velocity and have the dummy impact a stationary object, like a block or a cylinder. I know the alternative would be give the block an initial velocity, but I'm curios to see the body kinematics after the collision.
Thanks!
Best thing for what you want is to define a node set for all the dummy nodes, and use *Initial_velocity_set and reference that node set. You can also define a vector that is perpendicular to your impact surface once you have the dummy in position, and reference that vector id in the *Initial_velocity card
Thanks for the suggestion @negativevolume. I've created a node set of all dummy parts using the "entity creation" toolbox. And assigned an initial velocity to the node set. I added option IRIGID = -2 to override any and all rigid body initial velocities.
However, when I run the simulation I get issues with the nodes connected to discrete elements in the model. I've tried excluding them from the initial velocity by defining node set #2 and excluing it from the initial velocity using the NSIDEX field. But that did not solve the issue.
It seems like such a trivial thing to do, but I guess it's never "that" simple with LS Dyna. Anything obvious that you see wrong? I've added the .k file with the impactor and initial velocity as an attachement.
Yeah this can be tricky. Did you try the other “irigid” options (0, 1, 2)?
I usually don’t have this problem using 0 or 2. But if you keep having problems, you can use irigid=-1 and apply initial_velocity_rigid to each rigid part, which is a pain.
Also, LSTC HIII model should have a manual that explains how you should be able to prescribe initial velocity with their model.
Yeah, I'm not yet desperate enough to apply *initial_velocity_rigid to each part, haha.
I tried a different version of the dummy, (LSTC 50th% Rigid-FE H-III Standing Dummy (Beta, End-June 2010)) and managed to get it to run with *INITIAL_VELOCITY_GENERATION applied to a part set (which includes all the dummy parts) and IRIGID = 0 for a start. It seemed to work, but there were some weird oscillations in the dummy torso (see below).
Also, the initial velocity of 10m/s was not applied to the whole dummy. The velocity of a node at torso appears to be only around 3.5m/s.
However, when I changed IRIGID = -2, it seems to have resolved both the weird oscillations and the initial velocity issue. Although it's not perfect, it's pretty close to 10m/s. The kinetic energy also checks out with a hand calculation. Good enough for now!
Thanks for your suggestions @negativevolume!