Jump to content

Simon Križnik

Members
  • Content Count

    1090
  • Joined

  • Last visited

  • Days Won

    99

Everything posted by Simon Križnik

  1. I do not have any SID at the moment, but a dummy model can be found in:
  2. Hi, side impact pole is preferably modeled by a rigid wall: -analysis>rigid walls>geom>shape=cylinder pick the appropriately offset node (edit node coordinates) and specify the corresponding diameter. Please refer to tutorial RD-E: 1600 Dummy Positioning in Radioss help.
  3. while this procedure does the job it will delete all shells, including shell elements that might be desired in the analysis.
  4. @Guerric RADIOSS Starter does not check if the kinematic conditions are truly incompatible. You have constrained the independent node of the rigid in all but Y traslational DOF and also imposed velocity in this direction so those kinematic conditions are compatible. But there are examples where they are not incompatible (see attached documentation from Help) I also revised the boundary conditions (the referred node for liberty and velocity was not the same) and contact definition (slave component was not specified, also I activated some flags) and moved the plate closer to ball by 27 units so the contact happens earlier. Incompatible Kinematic Conditions.pdf Model.hm
  5. Hi @yugang, yes, the error is due to your mesh method (solid map using shell elements). Shell elements used to construct 3D mesh with solid map should be deleted before simulation: 1. Isolate the components with solid mesh 2. mask 3D elements 3. delete shell elements used for construction It is a good practice to create constructional shell mesh in a different component.
  6. Hi Guerric, there are some modeling errors, listed from the starter .out file: shell property is defined for brick elements, which is incompatible. all components containing FE entities should have card image set to Part. There were also some shell elements which were used for solid mesh creation but should be deleted before the simulation. Always check the .out file for warnings and errors and use Tools>Model Checker>RadiossBlock for debugging purposes. Model.hm Model_0001.rad Model_0000.rad
  7. Hi Inge, the forces are deleted together with RBE2, because they are no longer referenced by any node after deletion. In order to keep the forces create temporary nodes (shift+F2) on independent nodes (where force is applied).
  8. Please check the attached document from Radioss help, section: "What is the meaning of: WARNING ID 94 and WARNING ID 477" If the issue persists please share the model. Contact Interfaces.pdf
  9. Hi @pohan, In the penalty based contact, the interaction between the bodies is handled with spring elements that model the stiffness of the contact. Deep penetrations increase spring stiffness which leads to timestep drop. Try increasing the Gapmin (set Inacti=6 and Fpenmax=0.8 to avoid initial penetrations) which will allow the contact to work sooner and prevent deep penetrations Or you could try making the interface stiffer, Istf=3 which uses the maximum stiffness of the slave and master. If timestep is low from the beginning, make sure there are no initial penetrations or intersections: -in Hypercrash: Quality>Check all Solver Contact Interfaces -in Hypermesh: Tool>penetration check The material parameters should be realistic as they can influence contact stiffnes (depending on Istf flag) and special treatment may be necessary if contact is between soft and hard part.
  10. These graphs are expected: -the model without EPS_max defined (Default = 1030) will absorb more energy, because it will continue to absorb energy even at large strains, which is less realistic. -the model with EPS_max defined will absorb less energy, because it will only absorb energy up to the plastic strain at failure defined by this parameter, after which the ruptured elements will not contribute stiffness. Also, if both models have the same boundary conditions the initial kinetic energy of the system should be the same. However, in your examples, the initial KE is different (3.3E+7 vs. 8.5E+6).
  11. Hi @misa_roll, the toughness (ability to absorb mechanical energy up to the point of failure) is less with failure plastic strain (EPS_max) defined because the area under the stress-strain curve is reduced.
  12. You can try with Optistruct- just import the .fem file and run either NLSTAT, NLGEOM, EXPDYN or IMPLDYN analysis.
  13. Hi, I have made 3 simple models to show the different ways to simulate preloaded spring hinge in Radioss using Type 8, 13 and Kjoint2 springs. torsion_pin_type8.hm torsion_pin_kjoint2.hm torsion_pin_type13.hm
  14. @Prakash Pagadala can you explain which factors affect the master node timestep of he RBODY?
  15. It has been a while since I last used the Inspire, so I am not up to date on its functionality. There was Clearance Parameter for Contacts introduced in 2018.1 version, which allows you to define when a contact is considered in cases where there is a slight separation between parts in contact. Also check contact validity. If the initial gap is too big it is a contact non-linearity, but the Inspire is based on linear solver, so it might not be suitable for such problems.
  16. Hi @fcolomb, nodes constrained by the rigids will not decrease the timestep during simulation. You can check this simply by deleting the rigids and comparing the timestep in the outfile. It is not a good idea to change stiffness or mass of the elements constrained by rigids, because this will change the physics of the problem. The stiffness is especially important for contacts as both master and slave stiffness can be taken into account for contact stiffness, depending on the Istf flag.
  17. Hi Yannis, you may find useful: Fllwe.fem
  18. Hi Andy, the procedure to extract contact forces in HV: -instead of the contour, plot the vector contact forces -in the Query panel activate the node ID, node coordinates and Vector (Contact Force) categories than pick the nodes of interest. Export the table in .csv format. To import the contact forces in Optistruct: -analysis>forces>linear interpolation (choose the .csv file) This procedure can be found on pages 32 and 34:
  19. Hi, Different initial design starting points (ply thickness or initial material fraction) could potentially result in different optimum solutions. Please refer to the attached document from Optistruct User guide. I would also recommend the free eBook: Practical Aspects of Structural Optimization (A Study Guide) Global Search Option.pdf
  20. Hi Omer Faruk GUMUS, This query was already treated: Pin hinge with torsions spring can be modeled with type 13 spring in combination with a preloading function (torque vs. angle) defined in the appropriate rotational DOF. The same can probably be done with type 8 spring.
  21. Hi Amasker, unless the collision is at slow speed (minimal kinetic compared to internal energy), this problem should be solved using an explicit time integration scheme to account for the effect of mass (inertia). Optimization is possible via the equivalent static load method, which is basically transferring the nonlinear problem into a series of linear problems. The problem can be solved in Optistruct transient (EXPDYN) or Radioss optimization (refer to tutorial RD-E: 5101 Thickness Optimization for B-Pillar). Attached are various documents that may be helpful. NL_Opti_Jun7.pdf 충돌_최적화_해석_2017_v01.pdf 충돌_최적화_해석_REV04.pdf expdyn_bump_opti.hm
  22. Hi, check the car test_0000.out file for warnings and errors. In this case the error is: There is no variable defined for output request. The default output requests are defined by entering DEF in the Data: Var table.
  23. Hi Rahul Prakash, the Free eBook: Practical Aspects of Finite Element Simulation (a Study Guide) is highly recommended. Please refer to chapter 2D Meshing With HyperMesh for guidelines. The following strategy is recommended if element quality is the most important: 2D>automesh (F12)>QI optimize>edite criteria to define element quality criteria 2D>qualityindex where element quality can be checked visually and edited using cleanup tools one at a time. The failing elements can be saved to later be retrieved in elem cleanup or automesh again 2D>elem cleanup can be used to cleanup a group of elements
  24. The connection between the support brackets and monocoque can be modeled more accurately by bolt connectors.
  25. Hi jarjun, I suggest google searching for relevant research papers and thesis. You can find simplified material properties on page 29. The tensile and shear stress functions can be found in https://www.researchgate.net/publication/249505709_Characterization_of_Aluminum_Honeycomb_Material_Failure_in_Large_Deformation_Compression_Shear_and_Tearing
×
×
  • Create New...