Jump to content

Felix Radisch

  • Content Count

  • Joined

  • Last visited

1 Follower

About Felix Radisch

  • Rank
    Advanced User

Profile Information

  • Are you University user?

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Hi Tjansari, Inspire 2018.1 is using a tetra mesh to form Lattice. Regards, Felix
  2. Hi Gojosema, Welcome to the Forum! For structural analysis and optimization we use Altair Optistruct in Inspire. In case you are going for a solver comparison, you might also check Altair Hyperworks to ensure you are not limited by Inspire which is just using very few options/capabilities of the core Optistruct. Regards, Felix
  3. Hi Yosef, What exactly do you mean with "still does not work"? Can you describe more in detail? Thanks! Felix
  4. Dear Lala, Please be aware this is a user support forum where noone has to answer your questions. Additionally, please immediately stick to the common rules of communication. Kind Regards, Felix Radisch
  5. Hi Lala, Have you tried to turn type of the optimization to "max. stiffness"? It seems to me, your min. safety factor isn't achievable with your BCs and loads. Regards, Felix
  6. Inspire is not allowing to set the max. amount of Iterations or any other values regarding the optimization process itself. Optistruct for sure is allowing this through Hypermesh - if there is a need for you to do it - please refere to HyperWorks user forum. Again - if there is no convergency within 80 Iterations, there is propably something wrong with the setup (maybe missing supports, large deformations...). -Felix
  7. Hi Jerome, Welcome to the forum! I don't think your result will be better if we could increase the amount of iterations within Inspire. If your Optimization takes more than 50-60 Iterations, there is usually something wrong with your setup - or - at least we can tweak it to force a convergence. If you like to share your model wide, please upload it here - or - PM, so I can send you an upload link. Regards, Felix
  8. Hi Rahul, Well - I think the issue is quite clear isn't? You are trying to run an optimization, which can't be achieved with the Boundary conditions you applied. You could play around with them to find out which it is - and change it. If I would have to guess: Your min. elementsize is too large or there is no solution possible with the direction of your draw-direction. -Felix
  9. Hi Guys, I'd like to share "Tips and Tricks" regarding Lattice Design and the NonDesign Space. First to mention: Please keep Inspire updated - we introduced Lattice in 2017.3 but the last few months, we put a lot of effort to increase performance and stability! ATM (April 2018), 2018.1 is best to work with! If you are unhappy with the "resolution" of your nonDesign spaces after a lattice optimization - you can manually force the nondesignspace (per part) to use smaller elementsize inside the model configuration: Comparison: Before: After: Furthermore, you can create a smoothed lattice before you export it as an STL: Regards, Felix
  10. Just got the results for my optimization I started - it seems like your boundary conditions (stamping) isn't suitable for your designspace - or you have to decrease the "min. feature thickness". Felix
  11. Hi Caleb, Glad to hear this was helpful! If you are changing your PC, you have to request another license through our website: https://solidthinking.com/Buy.aspx?category=Buy&item=Students Educators You can do mesh refinement with 2018.1, but it will not be helpful for the optimization - but the analysis: Click on this tiny triangle next to "Setup": Then choose "Mesh Control": You can use it to increase or decrease the elementsize by selecting surfaces of your model. But again - this isn't helpful for an optimization. Regards, Felix
  12. Hi Rahul, That's an impressive model! A few things I would suggest: 1. Delete all parts inside the model Browser which are not needed 2. Run the optimization for "max. stiffness" first - to find out what is an achievable SF with the "min. feature" thickness you are using 3. Take this info back into account for the "min. mass" optimization Regards, Felix
  13. Hi Rahul, Welcome to the sT Forum! Feel free to share your *.stmod with us, so we can have a look whats wrong there. Regards, Felix
  14. Hi Caleb, I guess you are runing "out of core" with your optimization: You can easily check if you are running "in core" or "out of core", by taking a look at the topoOpt.out file, which is located in your scratch directory (by default in the user directory). Inside you will find information regarding the recent Iteration (easy to see if your computation is still runing). Furthermore the info regarding your Memory usage: Within the Iteration 0 you will get the infomation if it is running "in core" or not: In case it isn't running incore, you might decrease the amount of elements by increasing the "Minimum Thickness constraint" inside the "Run Optimization" window: Out of Core Simulations are working usually, but they are way slower as the RAM isn't large enouth to fit the computation. Instead, data is shifted between your HDD, RAM and the Processor. As a result 90% of the time, no computation is going on - but data is read and written to your HDD. You might also should think about upgrading your RAM - 8GB RAM is really not much for optimizations. To put things into perspective: I'm working with 32GB of RAM. Highend Customers, usuallly got 128GB and even more. Regards, Felix
  15. Hi Virgile, Thanks for your question! "Faster" or "More Accurate" changes the mesh between Tetra 1th order (Linear Tetra 4) and Tetra 2th order (Parabolic Tetra 10): You always want to run an Optimization in 1th Order (faster), because it is way quicker. Results of 1th order Tetra analysis and optimizations are stiffer (self stiffening effect), but more conservative regarding the Factor of Safety. Second order is more accurate, but it will take much more ressources for the computation - so you might want to avoid this exept for the first analysis and the final analysis. The difference in the accuracy is ~5% to ~10% (depending on models). The difference in computation time can be between 1h and 10h for optimizations. Usually, no high-end workstation should run second order optimizations. The difference of results doesn't matter in nearly all usecases. A reanalysis based on an Optimization will use the same mesh - so if you select "Faster" for the optimization, it will be the same during reanalysis. If you select "more accurate", it will be 2th order for both optimization and reanalysis. Let us know if this was helpful! Regards, Felix
  • Create New...