ERROR 6538 - Sliding Contact with interference (negative clearance)
I am running a model in which the contact has interference. as part of the installation process one part is pushed along the second one, therefore there is a frictional contact between them. This is the error message I get:
*** ERROR # 6538 ***
Finite sliding (TRACK=FINITE) CONTACT 1 is not compatible with CLEARANCE
in LGDISP analysis. CONTPRM,LSLDCLR,YES can be set with caution to bypass the
restriction.
I am working with HW2019 and the option LSLDCLR is not available. is there an alternative way to solve this issue?
Any insights?
Answers
-
Your model really needs large displacements? If not, you could switch off LGDISP.
Even though HM doesnt support it, you can add it using unsupported cards:
0 -
Thanks Adriano,
LSLDCLR did not avoid the issue. I need large displacmentes, the relative motion between parts is significant, it is part of the installation.
I keep getting the error 6538 as soon as I specify S2S or N2S for the contact discretization, when I leave it by default no error message shows.
Leaving it by default is however not the solution , since the contact elements of the part do not get update despite the relative motion.
The Initial contact elements are the only ones working during the simulation.
Any ideas? Also, where can I find more infor about error 6538?
Kind regards,
0 -
Are you sure you've activate it correctly?
Please try running the attached model. I've added a 5mm interference, with FINITE Sliding and LGDISP, rotating 1rad the pin.
<?xml version="1.0" encoding="UTF-8"?>
0 -
Also as an alternative, you can specify a GPAD in contact property (PCONT).
This will have the same behavior.
<?xml version="1.0" encoding="UTF-8"?>
1 -
Thank you Adriano,
Your answers were really helpful, How would you make the model slide axially?
When the contact is defined via PCONT, the padding paramenter GPAD is set to 5, why so? was it required to ease convergency? It seems to be incompatible with the negative clearance (press fit) Please find attached two modified models, in which I force the cylinder to slide axially. Dummy8 will not run due to the negative clearance.
Additionally I observed something awkward and I would like to have your opinion, after downloading the exampled I requested the output contact information, CONTF, and all of a sudden the error 6538 showed. I did download the .fem again, requested the contact output info and this time I did get results. What could be causing the error 6538 come out? Is this random?
0 -
the 5 stand for 5mm of clearance in my example.
If you use Clearance, it should be -5.0mm.
Using GPAD, 5.0mm.
For sliding axially i would add a SPC in axial direction with a predefined enforced displacement value.
0 -
Thank you Adriano,
Both models work well, however the contact pattern is not quite uniform only few nodes get paired actually (contact status). Intuitively due to the press fit the surfaces are in contact. Have you seen something like this before? How ccould I prove my model?
I have read the following about the control card GAPPRM, HMGAPST:
For correct visualization of their open/closed status in HyperMesh, this file needs to be imported before running the gap status command file, <my_job>.HM.gapstat.cmf.
Could you give some more detail about how to properly visualize contact status?
Best regards
0 -
Contact Status is visualized by CONTF output in OptiStruct, and it will show in HyperView as contact status.
If you use clearance, this should be more or less uniform
But you could also use some smoothening (SMOOTH in contact) to get a smoother contact distribution, if that's what you want.
0 -
Thanks Adriano,
What would be the pros and cons of using contact smoothing? Are there ay special precautions when using contact smoothing?
0 -
Hi again, depending on which of member of the contact pair the smoothing is applied, different normal and tangent forces are obtained, why so?
0
Altair Employee
