Mr.Alb

Members

15

• Days Won

1

Mr.Alb had the most liked content!

• Rank
Beginner

• Website URL

Profile Information

• Gender
Male
• Interests
Research and CAE simulations
• Are you University user?
Yes

Recent Profile Visitors

374 profile views

1. FEM Theory Question: Creating a local model from a global model

Hello Crashphys, FEM is pretty general as field. In RADIOSS solver, you can performe a cut to consider only a part of your whole system for explicit analysis. (https://altairuniversity.com/28761-radioss-sub-modeling-cut-approach/) If you work with linear problems, maybe the field of most typical optimization problems, with OPTISTRUCT you have to consider how real components contraint the item you want to optimize. Unfortunately, I don't know how you can make a binding bond... In my opinion, when you performe an optimization the real limit is the submodel you're considering rather than the constraints you've imposed. Usually accounting all part as possible is a good choice to do a meaningful optimum. For instance, are you really sure that optimizing one panel at once, globally it results in an optimum structure? Loads pattern could be very different in my opinion... Let me know
2. Radioss - Quasi Static Analysis

It may depend on timestep. If it's too big, you could lose your slope in PLOAD. Anyway, to compare static and quasi-static i suggest you to try with less slope, because with a high one, analysis's closer to a dynamic than a static evaluation. Let me know!
3. i need to do the simulation in radioss with very high speed impact, the simulating overall time is 500ns, and elements size is also 0.01um, how much of the time step? i run my model with these so small values, but it doen't work because i give the overall

I suggest you to modify this title and rename it with some shorter one. In Radioss, there isn't a unit system panel, so you have to put consistent units in all your model (mesh, material, property, timestep, cards ecc). With timestep you subdivide your analysis as frames for a video, so you have to know how much resolution you need. Can you share your engine and model files?
4. Missing Element to import hypercrash

Hi Korosh, I checked your model. It seems that you have made some mistakes with elements. I fixed your model, but you can fix it yourself by updating elemet types (2D>elemt types>elems>All>update), then export your rad and import it again in hypercrash. I tried, it works. Let me know! new_0000.rad new_0001.rad
5. Info about the GRSM algorithm?

I have the same problem. Is not possible to know how the algorithm works in GRSM method? There are a lot of different algorithms, for instance: Line-Search Approach Trust Region Approach Simplex Method Newton’s Method Quasi-Newton Methods Conjugate Direction Methods Levenberg–Marquardt Methods Elimination Methods Lagrangian Methods Active Set Methods Penalty and Barrier Function Methods Sequential Quadratic Programming Mixed Integer Programming NLPQLP Many Thanks
6. Cut Approach - Sensibility

ok, let's suppose for instance that my changes were about thickness on the front rails for my vehicle. Thickness varies within few millimiters (about 3mm), but when It's thin, the car folds deeply against the wall, whereas if It's thick, the car folds a little and then has a rebound. With the cut approach I have imposed the displacement to the section force so that the car's dynamic is fixed, isn't it? In this way, I cant't appreciate differences due to the thickness changes, can I? (Total mass isn't very affected from thickness changes (less than 1%), but stiffness might be) Many thanks again for your reply
7. Cut Approach - Sensibility

Hi, I have an issue with cut approach (https://altairuniversity.com/learning-library/radioss-sub-modeling-cut-approach/). I'm modeling an half of full vehicle with the cut approach above. From this sub-model, with hyperstudy I modify a little bit the geometry and study how the crash changes dynamic. I wonder how dynamics could vary among runs if the section-force imposes original displacements? Many Thanks
8. Automatic creation .h3d file from cluster runs

Thank you so much, Andy. Thorough response!
9. Automatic creation .h3d file from cluster runs

Good morning, I have an issue when I run a RADIOSS .rad file on my local group's cluster. At the end of the analysis, It doesn't automatically extract the .h3d file, although It does so locally on my pc. I have all animations A0*, but I'd rather It extracts h3d automatically instead of doing manually with hvtrans. I use Radioss Block 120, Many Thanks
10. Quality Index Shortcut

ok sorry, just solved. From Preferences>keyboard settings just fill a free gap with "hm_pushpanel {qualityindex}".
11. Quality Index Shortcut

Hi everyone, I wonder whether it's possible to configurate as a shortcut the 2D--> quality index. I know that from Preferences>keyboard settings I can change shortcuts, but i need the scripts and I can't find that one in my folders. I use Hypermesh 2019.1, many thanks! Regards
12. Hyperstudy fails automating run execution

This solved my issue. Many thanks! Was there a bug in previous versions?
13. Hyperstudy fails automating run execution

It's a little tricky for me, but i could try as soon as possible if there aren't any other solutions. I will keep you informed
14. Hyperstudy fails automating run execution

No, I'm using full version 2017.2 of hyperstudy and I've indicated Radioss.bat (2017.1) as script, but also with the same version 2017.1 (of Hyperstudy and Radioss) I encountered this problem
15. Hyperstudy fails automating run execution

Hi, I'm using Hyperstudy 2017.2 and Radioss as a solver. I encounter an error when I perform the evaluate step in the DOE approach. It happens that during a run (whatever) the evaluate and extract steps can't finish succesfully, but there aren't errors in the messages box and in the .out file I can read that analysis is already finished with normal termination: in spite of this, the cursor keeps turning and seems to be waiting to finish evaluation. I realized that when it happens, there isn't the .h3d file in the run folder, even if run is completed and all animation A0# are created. I realized also that by finishing the "hvtrans" process from the task manager, or (is the same) Tclsh85t application, hyperstudy concludes the evaluate step, extracts results and goes to the next run. It's a random problem, but it always happens at least once. Thank you all for helping out.
×
×
• Create New...