Implicit Dynamic Analysis (Drop Test) ANSYS Workbench 1

[Transient1]

Happy Holidays to all,

Does any one have any experience with setting up a transient dynamics simulation in workbench to simulate a drop? I have begun reading the manual. My questions are basically thus:

(1) Do I need to make the impact surface?

(2) If I do, what contact elements would be best to connect the dropped object and the floor? Frictional?

(3) Is it best to start with the object directly on the floor surface and apply a zero displacement and initial velocity condition?

(4) Would it be correct to apply a displacement versus time input (to simulate the acceleration profile) to the impact surface? If the object is a box, would I then constrain the top corners? That doesn't seem right to me.



Thanks for any help or guidance you can provide.

 

[Stringmaker]

Never done implicit transient analysis in WB but:

1) Yes you will need to make the contact surface. Make sure your object is initially not in contact.

2) In an implicit transient analysis you're probably only making your life more difficult by adding friction if you don't really need it. I would use surface to surface, or node to surface contact elements.

3) You should only have to fix the target surface and apply an initial velocity.

4) Initial velocity is the way to go here. Constraints would only be valid up until the point contact occurs.

 

[Transient1]

Thanks Stringmaker,

I was able to get a model of the impact of a rod on a rigid body surface. I accomplished it by specifying friction (or frictionless) contact elements betweent the flat end of the rod and the impact surface. The initial velocity condition in workbench, caused problems. I specified the initial velocity by starting at 0 displacement on the top face of the rod at time 0 and ramping it to a a higher displacement at a small time increment (1E-4). The parts started in contact and I had to icnrease the stiffness of the contact to reduce/eliminate penetration.

 

[ansysfreak]

Dear Transient

I did some droptests in WB and never had any problems with the initial velocity, it works fine. What is your displacement before impact? and element size on impact? Timestep size?

Regards
Garry

 

[Transient1]

Garry,

I specifiy zero displacement before impact. The rod is in contact with the rigid body impact surface. The surface is constrained on the bottom face in all DOF using a remote displacement. I solved using 2 steps, the first substep is in .001 second increments up to .005 and then the next substep is in ~0.01 increments up to 0.1 seconds. The problem solves, but the results are screwy, there are no stress waves and the values stay constant for all time steps.

The rod is 3 inches long, .337 inches in diameter. It is meshed with solid186 elements with I'd estimate an average size of 0.042 inches square and a total of 28,000 nodes.

The contact surface is meshed with .072 inch quads.

Frictional contact is specified using pure penalty method and a stiffness factor of 4.

Regards,

Chris

 

[davidmandis]

Hey guys I am currently working on a impact simulation problem in ANSYS WB. The problem is that I am using an educational version which does not have the LS DYNA solver, so I wanted to ask if you guys using a LS DYNA solver? If not where did you learn to carry out drop tests in ANSYS WB?

Thanks,
David

 

[Transient1]

David,

I believe forum rules are to start a new topic, when asking a new question. See ANSYS workbench help for flexible dynamics simulation to start. You probably will need to run 2-D simulations due to mesh size limitations in ANSYS ED.


Regards,

Chris

 

[ansysfreak]

Hi Chris

I was just wondering, are there results on time of inpacked? maybe due to time stepping these results are not available?
Maybe a stupid question but you never know.

Garry

 

[Transient1]

Garry,

Time of impact would be immediately after the simulation starts for the case of initial velocity and zero displacement. The rod is in contact with the rigid surface to start. I may experiment with the time step when I get a chance, but I would think that there would be some variation in stresses or displacement across the all my substeps. There is only an initial displacement and stress spike and then it remains constant.

Chris