Re: [deal.II] Step-77: line search fails

Hi Wolfgang, I was using version 9.2.0 and 9.3.0. Let me try the current version. Thanks, Andrew > On 14 Jun 2022, at 17:29, Wolfgang Bangerth wrote: > > On 6/14/22 09:59, Andrew McBride wrote: >> I’m trying to replicate the results reported in step-77. However when I run

[deal.II] Step-77: line search fails

Hello deal.ii, I’m trying to replicate the results reported in step-77. However when I run step-77 the line-search algorithm fails on the first refinement step with the error message below. I’ve tested different versions of deal and on different platforms and it’s not obvious to me what needs t

Re: [deal.II] Questions on Step-18

My view here is that the problem is quasi-static and hence time serves to order events. Hence a displacement increment can be viewed as equivalent to the velocity. For example, let’s fix the number of time-steps to 10. If the total time is 1s or 10s we should get the same results (assuming the

Re: [deal.II] Questions on Step-18

Hi both, Jean-Paul has addressed the second point nicely. On the first point, I think there is a 1/2 missing. The curl of the velocity gradient is the vorticity which is twice the angular velocity - hence I think you need a 1/2. Happy to be corrected on this. Best, Andrew > On 24 Jun 2021, at

Re: [deal.II] Creating a FESystem with bulk and surface FE spaces

Hi Manuel, I’m not sure if this helps, but some time ago we developed a code for surface elasticity in deal.ii. In this problem the surface of a bulk has its own energetic structure. One effectively constructs system matrices for volume cells and surface cells and these contribute to the same

Re: [deal.II] Polycrystal on deal ii

In the classical continuum approach to plasticity there is no grain boundary. The elastic anisotropy of the crystal is captured at the Gauss point level. So you would start by associating different materials properties (elastic / plastic) to Gauss points depending on their spatial location. Ha

Re: [deal.II] step-42 now available

Hi all Classical plasticity is underpinned by return mapping algorithms that operate at the level of the quadrature point. One has a global predictor for the displacement field where we assume frozen plastic flow. Then at each quadrature point you compute a trial strain, and hence a trial stre

Re: [deal.II] small strain (additive strain decomposition) elastoplastic code

Are you sure that the existing plasticity code (step-42) is behaving incorrectly? Can you demonstrate this via a simple example (consider only the mechanical problem)? I find this conclusion very surprising. Best Andrew > On 21 Nov 2019, at 18:02, Wolfgang Bangerth wrote: > > On 11/21/19 10:

Re: [deal.II] Are shell elements available in deal.II?

Hi I’ve hesitated to add this link as it predates many of the features in deal.II that would have made the implementation a lot more elegant and simple. Nevertheless have a look at https://open.uct.ac.za/handle/11427/5552 It’s a fully nonlinear shell t

Re: [deal.II] Rigid Body Interaction

Hi Jim In addition to the below you might have a look at https://doi.org/10.1007/s10237-017-0973-8 This builds on the the great contribution http://journals.ub.uni-heidelberg.de/index.php/ans/issue/view/1244

[deal.II] postdoctoral positions

finite element method. We are looking for people with experience in continuum mechanics / continuum physics and the use of open-source software packages (i.e. people like you). Feel free to contact me if you have any questions. All the best Andrew Andrew McBride Glasgow Computational Engin

Re: [deal.II] L2-norm gets decreased to certain value, but stays above threshold

I’m a bit confused by what you are asking. The entries in the residual vector corresponding to constrained dofs will converge to a non-zero number. Those corresponding to unconstrained dofs should converge to zero. As is stated in step-15 … we have to set to zero the elements of the residual ve

Re: [deal.II] L2-norm gets decreased to certain value, but stays above threshold

I’m referring specifically to a line similar to the following where all entries in the residual associated with constrained dofs are zeroed: for (unsigned int i=0; ihttp://www.dealii.org/developer/doxygen/deal.II/classDoFHandler.html#a54e7270f2ba6206604f794114b39a2aa>(); ++i) if (boundary_dofs

Re: [deal.II] L2-norm gets decreased to certain value, but stays above threshold

Are you ensuring that the entries in F(U) associated with the constrained dofs are set to zero? As shown in http://www.dealii.org/developer/doxygen/deal.II/step_15.html#MinimalSurfaceProblemcompute_residual

Re: [deal.II] Modifying step-44 - Null pressure case

Hi Lucas Not quite - have a look at the struct Errors in step 44 where: void reset() { norm = 1.0; u = 1.0; p = 1.0; J = 1.0; } void normalize(c

Re: [deal.II] Re: Modifying step-44 - Null pressure case

I think the problem arises from the fact that the error residual is essentially zero in the absence of loading. There is nothing wrong with the theory, it’s just that we assume loading will / should be applied. I think this is reasonable. Try a small value (positive or negative) for the pressur

Re: [deal.II] Re: Question about Step 44 shape function gradient computations

Dear Claire You might also be interested in the one-field version of step-44 in the code gallery https://dealii.org/developer/doxygen/deal.II/code_gallery_Quasi_static_Finite_strain_Compressible_Elasticity.html A > On 9 Nov 2016, at 18:24, Claire wrote: > > Dear Wolfgang and Jean-Paul, >

Re: [deal.II] Re: Learning Finite Element Method

The following course introduces FE and deal.II https://www.youtube.com/playlist?list=PLJhG_d-Sp_JHKVRhfTgDqbic_4MHpltXZ A > On 8 Nov 2016, at 12:41, Jean-Paul Pelteret wrote: > > Dear Divya, > > Its incredibly diffic

Re: [deal.II] Writing local_quadrature_point_history() in parallel (different sub_ranges)

Hi Chenchen Just so that I understand: are you currently solving the problem using the same approach (i.e. distributing the cells, solving the microscopic problem at each quadrature point of the cells). You have the solution and now you want to write it into a global history vector? When you