Jump to content

derMaster1

Members
  • Content Count

    17
  • Joined

  • Last visited

About derMaster1

  • Rank
    Beginner

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. I built up a simple model with nearly same Problem to share a screenshot. In my real model is no geometry, it's all FE-based.
  2. This time it's not possible to share a screenshot, sorry. Think of a C-Beam made out of Shell elements. But there are some elements missing in the middle of the length of the beam. That results in two beams which are not connected. I have to extend/Close thegap to get one beam again. So I want to fill it with Shell elements that connect one edge with the other.
  3. Hi, I need your advice again. I have two beams with a C crosssection. They are modeled by 2D-elements. The beams are co-linear and I want to connect the Ends of both beams (Frames) to get one beam as there is a gap between them at the Moment. A possibility would be the usage of RBE2, but I think closing the Gap between them with Shell elements is the best solution. The mesh size differs a little bit. Do you have some ideas how i can connect/Close the gap? Thanks a lot
  4. Hey tinh, thanks a lot - that soulution helped me! Also the other function in that Panel (self intersecting surfs) brought me a good result (the same I guess). As result I got 4 Surfaces (instead of 2), but that should be no probem. Thanks for the fast help.
  5. The geometry is not connected. I am sure, that this is the problem. How can I connect them at the junction?
  6. Hi at all, I have a simple problem: two planar surfaces intersect each other. From the top it Looks like a cross, so two crossing lines have been extruded to get these two surfaces. Hyperworks does not see that there is an edge due to intersection. So when I mesh the surfaces, I get two independet meshes. What I want to get is one mesh, so nodes have to be exactly at the intersection-line/edge and are shared of both surfaces. Do you have any hints for me? Thanks in advance.
  7. Hi, I want to use the relative displacement (magnitude) between node-pairs as a constraint. The values/bounds of the constraints are all the same. For that, I see at least two possible ways: 1) Create a dresp1 for each DOF (1,2,3) and each Node with the displacement. Let's assume, I have 20 pairs --> 40 Nodes --> 120 dresp1-Cards Also I write an equation: f(a,b,c,d,e,f,)= sqrt( (d-a)² + (e-b)² + (f-c)² ) and create a dresp2 for every pair (-->20 dresp2) Last step is to create 20 design-constrain-cards. 2) I'm not sure if this is really suitable: Create 3 dresp1-cards (one for every translational DOF) for all nodes on one side (each is one node of a node-pair), and the same for the second nodes of the pairs. I get 6 dresp1-Cards, each includes 20 nodes. The equation is the same. One dresp2 should be enough for the whole problem. And also only one constraint. But the question is HOW or IF optistruct understands, which nodes belong together. Is it done by the correct / same order for each of the 6 dresp1-Cards? Is there also a more simple way for the same result? One side-question: what system (x,y,z) is the dresp1 for displacements using? Global? May I change that in Hyperview? Thanks for help...
  8. I got a running Model now, but I didn't use dresp3 but dresp2 with dequation. With dreport I get the right values now in my outputfile. But i can't use these values for constraints / objectives. Optistruct unfortunately understands, that an equation with two stresses as input always results in a stress, so my resp2 is a stress-response (even though it is a "formula"-response..). Due to this fact i get error 1800, which tells me to use the stress-constraints in the desvar-panel instead... Any thoughts? What i want to do is to reduce the maximum stress (=objective). That means I only need the greatest stress as ONE response and not as many responses as myy number of elements... Any help on this point?
  9. Hello, as suggested in another thread in this Forum, i tried the usage of dresp3-Card. What i want to do: - let excel calculate the tresca-stress (Input: max principle, min principle, output: stress (=difference between both input variables). - get the tresca stress back and use it for an "minmax-" or "min-objective" in optimisation What i wrote until today: LOADLIB DRESP3 GR01 tresca.xlsx $solid DRESP1 1 MAX_PR1 STRESS PSOLID SMAP DRESP1 1 MAX_PR3 STRESS PSOLID SMIP DRESP3 20 TRESCA GR01 MYFUNC 1 2 + DRESP1 1 2 + CELLIN C2 C3 + CELOUT D1 I also use a small excel-file, which calculates the difference beween "cell C2" and "cell C3" in "cell D1". Questions: - would it work like that ? (I can't test it at the moment) - What do I get with the DRESP1 / DRESP3 ? When using DRESP1 with "ELEM" and enter an element-ID, it would give ONE value to the excel-file, which seems to be pretty clear. But when I use PSOLID as the PTYPE, there might be hundrets or thousands of values (here: principal stresses). At that point I do not understand my own programmed code.. Maybe someone can help? I read the Tutorial OS-4095 and also a lot of manual pages. Thanks a lot
  10. Yes, it is possible. I don't know if i got you right, but you might do the following: - create two (or more) loadsteps - create the things you need for optimization - when choosing a response, take the "weighted version" - e.g. weighted displacement or weighted freq.resp.. Then you have to input the values for the weight of each single loadstep. Did that answered your question?
  11. Hello again, looks like i can change the task a little bit. But again it seems like I need help. Now, the task is: minimize mass/volume with a global tresca stress constraint. Is it possible in hyperworks in a more simple way then the problem above? Would be great if i could get some input or help. Thanks again in advance
  12. Thanks again for your answer. I guess that it's to complex for me to write a function code. Especially due to the rare information about programming stress functions and using hypermath in general for responses I will have to resign this "little project". :-(
  13. Unfortunately I have not been successful. Are there any tutorials how to deal with dequations? I never did these things before. How to get the single components of the whole stress tensor into an equation? Hopefully someone can help :-)
  14. so it's not a real limitation but only less comfortable when postprocessing? Thanks for your fast help!
  15. Hi together, I am using V13 student edition at home. When doing an opimization (e.g. topology) I can only take a look at the optimization results (e.g. density, thickness, shape, ..) but NOT click to the next page to see the analysis of the result (e.g. static displacement after x iterations, stress after x iterations). It's written that it's page 1 of 1. When using Hyperview at my University I can take a look at other pages. Then I have more pages "loaded". Is the reason, that I use a student edition? Any ways to access the other results? Thanks in advance!
×
×
  • Create New...