Jump to content

QuentinB

Members
  • Content Count

    15
  • Joined

  • Last visited

Posts posted by QuentinB


  1. Hi, 

    I am trying to learn how to realize an impact test with RADIOSS. I folowed this tutorial from Altair University: https://www.youtube.com/watch?v=aDeNhQFvqjE&t=304s

     

     

    I folowed every steps of this video but my simulation seems not to converge... I stopped it after 30 minutes and here is the last iteration message: 

    568955877_Error99.9.PNG.4bfdd1db251e989a8a3f0c98c1ccb549.PNG

    You can see that after 1715s I had an error of 99.9%, and the remaining time was increasing again and again (so the response couldn't converge)

     

    Before running the test, I had no error in the model checker:

     

     1837989501_Errorcheck.PNG.e3054743cc8060e0fd7f9705a572b0b2.PNG

     

    And here is the picture of mu model:

     1531121560_Droptestpic.PNG.67c16f88ea4700a1af2095ae116f3c19.PNG

     

    I attached a file to this post which is the model i use for this simulation. 

     

    Thank you very much for the help. Quentin BERMOND 

     

     

    Impact test-tutorial.hm


  2. 17 hours ago, Pranav Hari said:

    Hi QuentinB

     

    Is there a specific reason why you are applying pressure? or else you can use this suggestion and try out

     

    1. Select all the nodes of the surface area in contact  (white color portion) and make it as rigid. You can use the calculate node and multiple node option inside the rigids panel. Input the mass of the block as force into the master node of the rigid. This will equally divide and apply the force exerted by the block to all the nodes, which will give reasonable result

     

    2. Also the contour image will show the deformation according to the scale factor you have provided. So check the actual deformation values of the deformed region and interpret the result; the deformation may be too low

     

     

     

    There was no particular reason why I used pressure so I tried your solution and it worked, the simulation gives me the expected results. I am not sure the magnitude is right but the type of response totally is. Thank you very much. 

    2118210394_Rigidsimulationbehavior.PNG.32f70f329cd69c1dc2d7e44fc601082e.PNG

     

     

     

     

    13 hours ago, tinh said:

    You made the model complicated.

    If concrete block is non deformable, it does not cause bending so I think just devide its mass by 2 and apply to each pillar but not the middle beam - only compression is considered.

     

    You're right, it would work for that case. As I just mentioned previously, Pranav solution also seems to work, I'll compare both results and use the best one for my thesis. Thank you very much for your answer. 


  3. 16 hours ago, Pranav Hari said:

    Hi QuentinB

     

     

    How did you apply the pressure load. Will it resemble a distributed load along the structure?

     

    Hi thank you very much for this reply, I used the "pressures" tool from the analysis panel. 

     

    The concrete block is 500 kg

    The "white area" which is the contact between the block and the structure is 34 950 mm2 

    g=9.81

     

    So by using the relation Pressure= Mass*g/white area 

    I find that pressure = 0.140 MPa

     

    This is the value of the magnitude I use for the pressure on the entire white area. Should I proceed differently ? 

     

    I attached some pictures to show how the pressure is applied on the elements.671570095_Loadzoom.PNG.c2e76b98fc9ef41ececff44e46c78813.PNG

     

    1458658810_Loadapplied.thumb.PNG.fec3b94b6003135b1aaefa58089b0a22.PNG

     

     

    15 hours ago, Jack Vincent said:

    Looks like a standard three point bending response. If you could include a scale that would be helpful.

     

    Could always try using a remote force where the top of the block is 

     

    Hello ! You're right, it is a standard response, but this is not what I expected because the concrete block cant't be deformed like that in its center area.

    I might be wrong but the constraint is supposed to be concentrated in the "pillars" relied to the ground. The following picture shows the case that happens ( case number 1 on the top) and the one I expect to model (case 2, on the bottom). In the second case, the bloc is not deformed, which is what happens in reality (we have some measurement that shows the concrete can't bend in that case. 

     

    You asked for a scale. The maximum displacement in the center is around 2 mm while I expect a value 10 times smaller. 

     

     

    721707969_Expectedresults.thumb.jpg.61900b2add5734d1e3479291cdb62c4e.jpg


  4. Hello, 

     

    I am a master degree student and i'm currently try to solve a finite element problem for a school project. 

     

    I want to run a a static analysis about a structure that is supporting concrete blocs. I've linked a picture of the problem to this message. 

    The concrete blocks contact with the structure is not continuous. I have linked a picture of my Hypermesh model with the contact highlighted in white. 

    I use a pressure reparted on this white area to model the actions of the concrete blocks on the structure. The concrete bloc is not deformable.

     

    The third picture coresspond to my simulation results. 

     

    I am not sure these results represent the reality because as I mentionned before, the concrete block is not deformable so the structure is not supposed to bend that much in the middle area. I suppose the software consider that these pressures are independent, so how should I proceed to make Hyperwork understand that all these pressures belong to a same solid in order to see the "pillars" of the structure being deformed instead of the middle area ? 

     

    For this type of loading (mass that has a complex contact with the structure) should I run a multi-body simulation instead ?

     

    I don't know if my questions are clear, feel free to ask me some more details if you don't understand it. 

     

    (i've linked in 4th picture the constraints in the structure)

    Concrete block holder with block.PNG

    Hypermesh model.PNG

    Concrete block results.PNG

    Constraints concrete block holder.PNG


  5. I did, but it was only about the theory...This is the first time I use a software like hypermesh (before I was using automeshing tools without thinking about the quality of my mesh). Actually, I practice a lot because i found out recently that finite element analysis is extremely interesting so I want to understand it as much as possible. My next goal will be the understanding of optimization with Optistruct. I already red the free E-books about Hypermesh and Optistruct but sometimes years of practice are way more  important than just reading theory. I asked those trivial questions here to have some experimented users feedback and to be sure it was the most efficient thing to do 

     

    Thank you for all those answers Tinh and Pooja S !


  6. Hi,

    I’m a beginner with altair Hypermesh and today I try to connect two parallel surfaces together. those midsurfaces are exctracted from the model shown below (the plates are considered to be welded together in reality).My goal is to mesh those surfaces and recreate the contact in yellow shown on the second picture. For the moment, due to the mid-surface extraction, I have a gap between the two planes. should I use RBE2/RBE3 elements to connect them ? My goal is to have a contact that behave as much as possible as if the two planes are welded. 

     

    Thank you very much for the help

    parallel planes connexion.PNG

    surface shared parallel.PNG

    midsurface parallel.PNG


  7. Thank you very much for this fast answer tinh

    I tried the first solution, my surfaces are in contact but i still have a free edge as shown on the pictures below (basically, I extended the vertical planes to the horizontal one). Is the contact existing even with this free edge, or I still have to use an other tool to "merge" those planes to be sure my nodes will be aligned ?

     

    surface edit panel used.PNG

    Surfaces not connected.PNG


  8. Hi,

    I’m a beginner with altair Hypermesh and today I try to connect two perpendicular surfaces together. My goal is to mesh those surfaces and be sure that the nodes of the horizontal plane are coincident to the ones on the vertical plane. For the moment, due to the mid-surface extraction, I have a gap between the two planes that's why the edge is a "free edge", from what i understand of the software. I assume that before starting the meshing part, i should connect the two planes together and have a "shared edge"... Is it the right way to deal with this situation ? 

     

    Thank you very much for the help

    system picture.PNG

    midsurfaces extracted.PNG

    edges to connect.PNG

×
×
  • Create New...