Negative Jacobian

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • Vaishakhraju
    Member
    • Jun 2019
    • 36

    Negative Jacobian

    Hi,

    While running my simulation i am getting the error "Negative Jacobian was detected at element", What is the meaning of this error? How can we avoid this error to get a converged solution?


    Thanks in advance
    Vaishakh R
  • maas
    Lead Code Developer
    • Nov 2007
    • 3400

    #2
    Hi,

    This error means that at least one element in your mesh has turned inside out due to the large deformation. At this point the FEBio cannot continue since the solution has become nonphysical. There are many reasons why this could happen. I would suggest to take a look at Chapter 7 of the online manual. If you can't find the answer in there, you may send me your model and I'd be happy to take a closer look.

    Cheers,

    Steve
    Department of Bioengineering, University of Utah
    Scientific Computing and Imaging institute, University of Utah

    Comment

    • Vaishakhraju
      Member
      • Jun 2019
      • 36

      #3
      Hi Steve,

      Thanks for the reply, I am attaching the preview file that is having the convergence issue.

      I want to implement gait cycle in openKnee model, for that i provide rotation curve and load curve for 1 second. I have given rotation in initial step using curve editor with maximum extension angle of 22 degree. I want to give a load curve with 1000 N of maximum load, First i tried with 10 N in step-1 and the simulation run successfully, after i increased the load to 100 or 1000 N, error termination is coming. It would be very helpful if u can figure out the isssue and with a valuable feedback.


      Thanks
      Vaishakh
      Last edited by Vaishakhraju; 07-12-2019, 11:29 PM.

      Comment

      • maas
        Lead Code Developer
        • Nov 2007
        • 3400

        #4
        Hi Vaishakh,

        I think the problem is that you are applying the force too quickly. At the first iteration of the first time step, contact is not yet established, so there is nothing to stop the femur from moving down. Since the force is already big at this point, the femur moves quite a bit and inverts elements. This is causing your negative jacobians. I suggest that you ramp up the force more slowly, so that contact can be established and the femur "settle in", before the loading begins. You could setup a two-step analysis, where you do this initial slow loading in the first step, and then your desired loading pattern in the second step.

        Cheers,

        Steve
        Department of Bioengineering, University of Utah
        Scientific Computing and Imaging institute, University of Utah

        Comment

        • Vaishakhraju
          Member
          • Jun 2019
          • 36

          #5
          Thank you so much for your feedback,

          I will try to change the time step settings and load as per your suggestion.

          Thanks
          Vaishakh

          Comment

          • aliraza
            Junior Member
            • May 2020
            • 5

            #6
            Hi,

            I am facing the negative jacobian problem. First I ran the simulation with the following material

            "Elastic
            Solid Mixture
            Neo Hookean
            Fiber EXP Pow
            Fiber EXP Pow"
            It worked
            But as I changed the material from elastic to uncoulped it gave me negative jacobian error

            "Uncoupled
            Uncoupled Solid Mixture
            Mooney Rivlin
            Fiber EXP Pow uncoupled
            Fiber EXP Pow uncoupled"
            I also went through the FEbio user manual but I did not understand what exact problem I have. I changed the time step but it did not help me.
            any kind of help would be greatly appreciated.
            Here is the error I am getting while running in preview.


            1
            ************************************************** ***********************
            * ERROR *
            * *
            * Negative jacobian was detected. *
            ************************************************** ***********************


            ------- failed to converge at time : 0.1

            Retrying time step. Retry attempt 1 of max 10


            Thank you!
            Ali

            Comment

            • maas
              Lead Code Developer
              • Nov 2007
              • 3400

              #7
              Hi Ali,

              When you use an uncoupled solid mixture, you have to set the bulk modulus (parameter k) to a suitable value. First, ensure that this parameter is there and then try changing the value. You have to set it high enough so that the material behaves nearly-incompressible, but if you set it too high you could introduce ill-conditioning, which could lead to negative jacobians.
              If that's not the problem, check your boundary conditions and loads. Since the error is happening in the first time step you might be applying the load to quickly. Check if the loads and prescribed boundary conditions have a load curve associated.
              If none of this helps, feel free to send me the feb file and I will take a closer look.

              Best,

              Steve
              Department of Bioengineering, University of Utah
              Scientific Computing and Imaging institute, University of Utah

              Comment

              • aliraza
                Junior Member
                • May 2020
                • 5

                #8
                Thank you so much for your kind response. I followed the steps you mentioned but still it is not working. Is that possible can I send my .feb file to your email.

                Thank you!

                Ali

                Comment

                • maas
                  Lead Code Developer
                  • Nov 2007
                  • 3400

                  #9
                  Hi Ali,

                  Yes, you may send your file to steve.maas@utah.edu.

                  Best,

                  Steve
                  Department of Bioengineering, University of Utah
                  Scientific Computing and Imaging institute, University of Utah

                  Comment

                  • Shah
                    Junior Member
                    • Dec 2021
                    • 12

                    #10
                    Hi, I have taken Negative jacobian was detected. error. How can this error be solved in a soft tissue extension test model?
                    I have sent my .feb file to the steve.maas@utah.edu . I would highly appreciate it if you let me know how can I solve the error.
                    Best,
                    Shah

                    Comment

                    • pooyan.vahidi
                      Junior Member
                      • Jun 2023
                      • 9

                      #11
                      Dear Steve,
                      I hope you are doing well.
                      Many thanks in advance for your reply. when I am running my simulation the error " negative Jacobian was detected" will appear. I checked everything based on the Febio manual but unfortunately, I didn't find the solution for it. I will be happy and so appreciative if you could help me in this regard. I shared my file in the link below:

                      Pooyan.fs2

                      Many thanks for your consideration
                      Kind Regards
                      Pooyan​​

                      Comment

                      • maas
                        Lead Code Developer
                        • Nov 2007
                        • 3400

                        #12
                        Hi Pooyan,

                        In this model you assigned the neo-Hookean material, which appears to be for the soft tissues, to the tibia. Was that your intention? I assume you intended to assign a rigid body to the tibia as well?

                        Best,

                        Steve
                        Department of Bioengineering, University of Utah
                        Scientific Computing and Imaging institute, University of Utah

                        Comment

                        • pooyan.vahidi
                          Junior Member
                          • Jun 2023
                          • 9

                          #13
                          Dear Steve,
                          Thank you so much. I have sent an email to steve.maas@utah.edu to explain this issue—many thanks for your consideration.
                          Kind Regards
                          Pooyan​

                          Comment

                          • pooyan.vahidi
                            Junior Member
                            • Jun 2023
                            • 9

                            #14
                            Dear Steve,
                            I hope you are doing well.
                            I modified my model and the loading condition, but unfortunately, I am still getting the “ negative Jacobian error was detected”. Please see the shared model below:

                            PooyanModel.fs2

                            In rigid section, I applied rigid force, rigid fix, rigid, rigid moment, and rigid rotation. The problem is that when I put a value for the rigid rotation (0.785), the error “ negative Jacobian error was detected” will appear. To resolve this issue, I did the following steps:
                            • Model modification
                            • The rigid rotation was applied in different steps up to 4 (slowly ramped up)
                            • The tied contact between the graft and the hole
                            • Modifying the loading curve so that the load applies in step form.

                            I didn't apply the rigid rotation value (0.785) in the attached model, and it is working. I will be happy and so appreciative if you could help me in this regard—many thanks for your consideration.
                            Kind Regards​

                            Comment

                            • maas
                              Lead Code Developer
                              • Nov 2007
                              • 3400

                              #15
                              Hi Pooyan,

                              We communicated via email so I'm not sure if this is a new issue or if this issue was resolved. Can you please let me know?

                              Thanks,

                              Steve
                              Department of Bioengineering, University of Utah
                              Scientific Computing and Imaging institute, University of Utah

                              Comment

                              Working...
                              X
                              😀
                              😂
                              🥰
                              😘
                              🤢
                              😎
                              😞
                              😡
                              👍
                              👎