Jump to content

antonio.dortona95

Members
  • Content Count

    32
  • Joined

  • Last visited

About antonio.dortona95

  • Rank
    Advanced Member

Profile Information

  • Are you University user?
    Yes

Recent Profile Visitors

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

  1. Hi, I'm trying to build a model in order to simulate a shear test on a composite panel with honeycomb core. I've been trying with non-linear static load-step but I get a non-convergent results and the error you can read in the .out file attached here. One possible solution I've tried, could be tuning the NCUT parameter in the NLADAPT collector, but this couldn't solve the problem and I get the same error. I'm also not sure about the type of analysis I've chosen and how I have modeled the physics of the problem, do you think a non-linear transient or explicit simulation could be more suitable for this kind of problem? I attach here my model and the .out file of the run, I'll really appreciate your help! Thank you. solver_deck_s1_e.out Model_1.hm
  2. Hi, I'm trying to build a model in order to simulate a shear test on a composite panel with honeycomb core. I've been trying with non-linear static load-step but I get a non-convergent results and the error you can read in the title. One possible solution I've tried, could be tuning the NCUT parameter in the NLADAPT collector, but this couldn't solve the problem and I get the same error. I'm also not sure about the type of analysis I've chosen, do you think a non-linear transient or explicit simulation could be more suitable for this kind of problem? I attach here my model and the .out file of the run, I'll really appreciate your help! Thank you in advance. Model_1.hm solver_deck_s1_e.out
  3. So isn't there a way to verify with vectors on the elements, the x-axis (for example) direction after the user assignment? Thank you Simon you are the best!!
  4. Thank you for you very detailed and precise explanation, it will be very useful! Can I ask you something that could be a little OT with respect to the topic's title? Is there a way to orient a MAT9 ORT solid elements? I have tried with Aerospace--> Material Orientation and also with the classic 2D--> Composites --> Material Orientation as possible solutions as I've red here, but they actually don't work. To solve the problem I have changed the engineering constant's arrangement, in the material card definition (E1, E2, E3, G12, ecc..), in order to be congruent with global reference frame in the model. Is there a smart way to do this? I'm using the last version of Hypermesh 2019
  5. Thank you for your help, I'm trying to adjust my model. I need to ask a clarification about the point where you talk about the penetration between the plate and the supports: is the contact based on the shell mesh or do I have to take in account the ticked elements, to avoid intersection? I'll attach a picture to better explain what I'm meaning. In a previous model I've used slide contact between supports and plys, but the result has been (as you said) a sliding in Z direction, so I changed it into stick contact. I'll try to use friction to solve the problem.
  6. Hi, I need help with this model. I would like to perform a non-linear analysis on a composite specimen. It's a sandwich panel with honeycomb core, modeled as mat9 ort with two plies of carbon fiber per side. The main issue is the contact between the punch and the panel, the same as the one with supports. I'm not able to model them correctly. I've thought to use the contact slide but the results does not have sense. I have compared this with real test and the reaction force on the punch is more than 12 times bigger than reality. Looking at the results plot, It's clear that contacts behavior is not acceptable. Can you please help me? I attach here the .hm file. Thank you!
  7. Hi, I'm trying to perform a size optimization of a composite structure. When I launch the analysis i get this error. I've red something about a problem with the memory but free size optimizations are ok for the run. So I can't understand the problem, in particolar the word sayng "igply" seem to underline an issue related to plys. Thank you for your help.
  8. Thank you @Hyperman, I tried to remesh with QI optimize with the minimum size criteria (0.6), I checked the quality index and I used cleanup tools too but the same problem occurs, but if I decrease the time step of /DT/NODA it works but the time of simulation is too high. Is there another solution? braccetto_radioss.hm
  9. Hi, I'm trying to simulate a compression test to evaluate the buckling load, I use a CLOAD card to impose my load but the A-Arm fails also if I apply 1 N as soon as the simulation starts. I think that the problem is the mesh but there aren't problems if I try to do this simulation with Optistruct. Could someone help me? braccetto_radioss.hm
  10. Thanks again Ivan, I already read this tutorial, so I followed it. I used /DT/BRICK/CST with the same parameters of /DT/NODA/CST but the same error occurs
  11. Hi, Sorry for my question but I'm searching for advice. I have some errors with "run killed", energy and mass. I have also errors with "WARNING ZERO OR NEGATIVE VOLUME" and these elements belongs to the honeycomb. I think that these errors appear because I didn't defined a failure strain but I prefer to not define this paramater. Am I right? How could I solve these problems? Thanks for attention Best regards IA_5.0.hm IA_5.0_0000.out IA_5.0_0001.out
  12. Hi guys, Sorry if I ask another question. I don't understand some parameters in the card M28_HONEYCOMB. Is the strain the absissa of this curve? The last point of the curve is the yield or the rupture? Or it is only a function stress-strain and the epsilon values are our input to describe the rupture of the elements? Fscale is an addend to the function? Because I saw in the help that its dimension is Pascal so I thought that it isn't a factor. I'm asking all this questions because I'm thinking how I can simulate a buckling of the cell changing the yield function if my suppositions are right. Thanks a lot for the attention Best regards
×
×
  • Create New...