...
Meeting Notes
Tom + Kevin
Fixing the bar and running chassis
The contact with the chassis being bonded could be a large source of error
Matt - can lead to lower stresses; but not drastic changes
Matt - (MS12) frictional and sliding contact; one node closes to the collision object near the centre to prevent it from sliding
prevents it from flying into space; there can be other options.
Shalin - (MS12) fixed points on the chassis and ran the bar into the chassis
Matt - not what he remembered
Shalin - have you tried simulating with no masses
previously done back in may – no large
simulation of no mass is good
as you add mass, you should watch how it trends
1 human
2 human → exponential increase can indicate something is wrong
Mesh convergence study-
Matt - not surprised that it didn’t compile
it is finnicky in how you can get results that are proper
don’t think it’s the mesh size is the thing hurting you and also not surprised of the result being seen
Shalin - mesh convergence may not be the main issue right now
5 mm element change would not give an order of magnitude difference in result
just set a size that runs relatively quick and get your geometry/results to work with that
Connor - reinforcements. Have we tried local refinements to the mesh? Are there areas that should have a greater concentration of elements (like around the gussets for example) ?
Shalin - x elements in gussets but beams are your global setting
refine locally at gussets will get better capture at those locations
Matt- understanding the effect of mass addition into simulation would be more insightful than accuracy checks
adds the most confidence to the simulation
previous investigation
Shalin - Model in L-shaped extrusion and provide mass
set material density and volume of the L-shaped object === 80 kg
Start with worst case → try same load with no mass; add masses slowly;
5G should never be anywhere near yielding
Setup more investigation preparation for bar running into the chassis
Shalin - need to make really good suggestion regarding where you fix the chassis
Matt - how would you account for the mass of bb box, passengers
Matt - distribute the entire mass of the chassis onto the entire thing
allocate the additional masses and distribute to the entire chassis
Tom - time crunch
run it again - same constraints - run no mass
if mass is not the concern, start looking at how it is constrained can be a factor
eliminating major factors and narrowing the potential factors
Idea: remove the complexities and analyze it on a simple level
suspect that point masses could be sources of the problem
test it unloaded
test it masses distributed across the chassis
simplifying your model and still missing by a wide margin, there is something potentially wrong from the fundamentals (how you sim it)
Summary of Action Items
As per shalin and matt:
try simulating just the chassis for the worst case orientation with no masses
then keep adding masses to check for exponential increase (unreasonable)
Shalin’s suggestion of L-shaped passenger (geometry+density =80kg)
distribute the additional masses of the entire vehicle across the entire chassis and see how that runs
make presentation and send to advisors for offline review (see unclosed items)
take a look at kimberly’s studies on making the bar go into the chassis with the chassis stationary
also take a look at doing what MSXII did for having the contact between the collision object and the chassis be frictionless
frictional and sliding contact; one node closes to the collision object near the centre to prevent it from sliding
Unclosed Items
Fixing the chassis and running the bar into the chassis
friction contacts
^^ weren’t able to conclude or no ground basis of justification
reached to that point – didn’t reach out to advisors for review
that is the time to bring in the advisors for review