A silly question but I cannot anyting in the doc on this subject. Does SOL401 use a fixed time/load step based on StepDuration/Nb of increments? Assuming that the analysis does not cut back it seems that the step is the same even if convergence rate is good. I am sure software like abaqus will increase (if requested) the load if it can.
If you are using NX Nastran 9 or 10, then it's not surprising that after cutting back the time step, the time step doesn't grow. Work has been done in NX Nastran 11 to address some aspects of adaptive time stepping. I suspect you aren't beta testing NX Nastran 11 though? This is an area that will get more attention in future versions.
Simulation Product Management
Simulation and Test Solutions
Siemens Industry Sector
Siemens Product Lifecycle Management Software Inc.
doing the work in NX10. I am not worried about the cutting back. Did a test with 300 increments and all the way (up to 93.3%) SOL401 kept the same time step (0.033) - no cut back. At no point it tried to increase the time step if convergence between successive time steps was good. If my memory serves me right ABAQUS, tries doubling the time step (by default) if the convergence rate is "good" (event after cut back). This can speed up the analysis (somewhat). There is not such feature/option in SoL401? The time/load step is fixed unless cut back is needed!
Is there any (step-by-step) example available (for NX10) for a multi-step analsysis? I cannot find anything in the doc! I am trying to “move” something along (an axis) in 4 steps (each step squeezing the components a bit more) and cannot do this over the 4 steps wanted. Step 1 works but as soon as I am trying to apply (add) a force in step 2 the analysis crashes. There is nothing too fancy about the analysis at the moment;
Step 1: move part # 1 by 1.4mm (x-axis) – the key component is pushed by part #1 (via contact) moving and is squeezed.
Step 2: apply an (axial) force F to part # 1 (x-axis) – the key component is pushed further by part #1.
Step 3: move part # 1 by a further .3mm (x-axis) – the key component is pushed further by part #1.
Step 4: move part # 1 by a further 0.1mm (x-axis) – the key component is pushed further by part #1
Steps are a cumulative. Overall disp is therefore 1.4+?+0.3+0.1. One doesn’t know by how much part #1 is moving when applying the force F
(frictionless) Contacts are defined and are working fine. Once this is working then one will need to add plasticity
Here a weird one for the SOL401 experts out there
carried out 2 tests
Waht is going on with SOL401 ?
There is an example of a thermal mechanical coupled solution using NX 10 Multiphysics on Learning Advantage. I searched for “multiphysics” and found the workshop below. The NX 10 online help has a section on multiphysics as well. I admit we could use more examples as they are the best learning tools. The example here focuses on the thermal mechanical coupling capabilities of NX Multiphysics.
In NX 10, the solver will cut back the time step, but not increase the time step beyond the user specified steps.
The description of your model sounds like a press fit (i.e. squeezed) that is performed by pushing a body that is in contact with the pin of the press fit:
The 4 steps you defined would correlate to 4 steps/subcases in SOL 401. Steps 3 and 4 probably could be combined into a single step. I can start from your model to triage the specific questions or create my own and send it to you as an example.
Thanks for that. Will have a look at the Learning Advantage again. haven't looked there for a long time but it looks like myaccount give me access to very little (nothing!!)
the idea of step 3 & 4 are part of the specified assmely process.
Do Step 3 if a specific check is not met
Do Step 4 is after step 3 the specific check is still not met
Therefore I would like to keep steps 3 & 4 seperate even tough I could combine them
Happy to send a copy of my test model off line if you want