Learning Forum

[dadidal96]

Greetings!

I'm attempting to simulate a simple model: a cylindrical specimen under gravity load supported by a rigid plate in LS-DYNA SMP double precision. The plate is fixed in all degrees of freedom. The SPG parameters are set as follows: KERNEL=2, ITB=3, MSC=1, PDAMP=-0.001, DX=DY=DZ=1.6. Despite its apparent simplicity, the model exhibits instability after about 0.06 seconds. The specimen begins to move without any reason (no changes made to the model). I've noticed unphysical oscillations in acceleration values - they start to grow to extremely high values (1e06 m/s^2 and more) and I don't know why. I tested different SPG parameters (different KERNEL, ITB, MSC, DX, DY, DZ values) but the issue remain unsolved. Also I tried to include some damping to model (PART_MASS, PART_STIFFNESS, FREQUENCY_RANGE) but nothing helped. The contact between SPG part (specimen) and FEM part (plate) is defined using DEFINE_SPG_TO_SURFACE_COUPLING card. Also I tried different contact thickness and stiffness inside this card, and make a little gap between SPG and FEM part. Nothing of these helped. Could anyone tell me why under just gravity loading the specimen start to move after some model time and accelerations are increasing? Looks like SPG method is very unstable at all.

You can download one of my input files with this link: https://drive.google.com/file/d/1xFyWjedxMHNY_ZBJ_B3iyiObG0ng53Bl/view?usp=sharing

[Ram Gopisetti]

Hi

Please know that SPG is stable in small time step problems and more tailor to mid to fast loading problems.

ANSYS/SYNOPSYS employees cannot share and download the files so I will leave that option to someone to check and comment.

However, I did not see any such issues based on same model on my side using the R14.1 and up for 1ms run with contact_automatic_node_to_surface with soft=2 

All my SPG section parameters are mapped to ITB=1 with 1.6 as dilation parameter. 

Cheers, Ram