[deal.II] solve on mesh imported through opencascade interface

I am trying to combine techniques from tutorial 54 and tutorial 38 - namely, I would like to solve a surface diffusion problem on a geometry that I bring in through the opencascade interface. I have created a geometry (using Trelis/Cubit) and have the mesh .ucd file, as well as the .iges file

Re: [deal.II] solve on mesh imported through opencascade interface

; order mappings on co-dimension one surfaces, defined through iges files. > This should works, but its highly untested… you may hit some high order > mapping bugs we are trying to address… I hope things will be resolved soon. > > Can you open an issue, so that I can refer back to it when th

[deal.II] Re: Ellipsoidal grid and boundary descriptor

Markus, I don't think this was ever added to dealii, do you still have an implementation of the ellipsoid handy? I'm trying to make an ellipsoid (just the surface) out of a ChartManifold and it's a bit of a pain, I sure wouldn't mind looking over someone's shoulder for some help here ;) Thank

Re: [deal.II] Re: Ellipsoidal grid and boundary descriptor

; > > Am Montag, 25. Juli 2016 21:42:39 UTC+2 schrieb thomas stephens: >> >> Markus, I don't think this was ever added to dealii, do you still have >> an implementation of the ellipsoid handy? I'm trying to make an ellipsoid >> (just the surface) out of

[deal.II] Re: Ellipsoidal grid and boundary descriptor

Markus, thank you On Thursday, July 28, 2016 at 1:13:09 AM UTC-4, Markus Klinsmann wrote: > > Hi Tom, I have attached the files. They were working with deal.ii 8.1, but > that is already quite some time ago. Hope it helps you, Markus > > > Am Montag, 25. Juli 2016 21:42:39 UT

[deal.II] obtaining an ellipsoid grid and manifold from GridGenerator::hyper_sphere() and a subclass of ChartManifold

I am trying to obtain the mesh for a codimension-1 ellipsoid and attach an ellipsoidal manifold to it in order to refine the mesh. My strategy is failin

Re: [deal.II] obtaining an ellipsoid grid and manifold from GridGenerator::hyper_sphere() and a subclass of ChartManifold

rounding points… in other words, the > average will not be in the middle of the four points (that should be the > north pole: a singularity of your mapping). It will lye on the same > latitude of the other four points. > > ChartManifold is doing exactly what you asked it to do, but in

[deal.II] does the latest dealii prepackaged image file for Mac OSX come with 'other software packages' such as p4est, PETSc, Trilinos, etc?

It's not clear to me from the github dealii wiki for MacOSX whether or not the prepackaged image file for OSX comes with all of the optional third party libraries that are listed as optional software

Re: [deal.II] weakly enforcing an identity - vector mean curvature as the surface laplacian of the identity on a codim-1 manifold

12:48:05 PM UTC-4, Wolfgang Bangerth wrote: > > On 08/17/2016 10:22 AM, thomas stephens wrote: > > *On to the question: *It is a fact that the vector mean curvature k_bar > > is equal to the surface laplacian of the identity function id_X on the > > manifold X, k_bar = laplac

Re: [deal.II] weakly enforcing an identity - vector mean curvature as the surface laplacian of the identity on a codim-1 manifold

6 at 12:24:23 PM UTC-4, Wolfgang Bangerth wrote: > > On 08/18/2016 09:39 AM, thomas stephens wrote: > > id_X looks like this: > > > > | > > template > > classIdentity:publicFunction > > { > > public: > > Identity():Function(){}

Re: [deal.II] weakly enforcing an identity - vector mean curvature as the surface laplacian of the identity on a codim-1 manifold

On Tuesday, August 23, 2016 at 12:51:53 PM UTC-4, Wolfgang Bangerth wrote: > > > Thomas, > > > I was able to solve for the vector mean curvature using the weak form of > > the identity k_bar = laplacian_X id_X, where X is a codimension 1 > > manifold without boundary. The image above is a plo

[deal.II] post processing a vector valued solution

I have a vector-valued solution, call it H, on a codimension-one geometry (described by a MappingQEulerian object) and I am very interested in obtaining H*nu, the dot product of H with the normal vector nu. I have already obtained H*H using a class derived from DataPostProcessorScalar, and th

[deal.II] Re: different shape_gradients for MappingQ and MappingQEulerian

nto something else. I suggest to myself to reread the Interplay of UpdateFlags, Mapping, and FiniteElement in FEValues: https://dealii.org/8.4.1/doxygen/deal.II/group__UpdateFlags.html, maybe this will all become clear one day ;) On Wednesday, October 19, 2016 at 7:19:47 PM UTC-4, thomas s

Re: [deal.II] post processing a vector valued solution

asses tests then I guess there's no problem, it just appears to be inconsistent. If this indeed needs a correction, then maybe we can walk through it together, I am interested in learning how to help. Tom On Wednesday, October 26, 2016 at 6:08:00 PM UTC-4, Wolfgang Bangerth wrote

[deal.II] Re: post processing a vector valued solution

Well, here's a pretty ugly attempt - One thing I want to point out is that my solution vector comes from a finite element built like this: (dim=2, spacedim = 3) FESystemfe; fe(FE_Q(fe_degree), spacedim) and the only DoFHandler I have lying around right now is dof_handler.distribute_do

Re: [deal.II] Re: post processing a vector valued solution

Your suggestions worked exactly as advertised. I've copy-pasta's my updated code below for completeness. > Thomas, > your code looks essentially correct, though it can be improved: > > > Now, for my attempt at this: > > > > | > > template(int spacedim> > > void data_dot_normal (Vector

[deal.II] mesh quality during deformations using MappingQEulerian

I'm working on a closed codimension-1 surface (think of an ellipsoid) and my goal is to iteratively evolve this surface to form tubulations and other localized significant deformations. My current strategy is to compute an euler_vector at each iteration and create a new mapping at each time ste

[deal.II] Re: mesh quality during deformations using MappingQEulerian

> > I can't say whether that's the same problem that is commonly reported in > the > literature, but this sort of mesh entanglement is the motivation for > people to > look at Arbitrary Lagrangian Eulerian (ALE) methods, instead of just > Eulerian > methods. > > Can you show a picture of yo

[deal.II] Re: mesh quality during deformations using MappingQEulerian

J-P, I should not have given Mesquite such short shrift in my reply yesterday. With regard to the MappingQEulerian approach, I would imagine Mesquite would take the mapped points as the mesh points (basically, what my figures show as the mesh), then smooth those out a bit, and return new verte

Re: [deal.II] Re: mesh quality during deformations using MappingQEulerian

> > > In the case that the mesh lives in a higher dimensional space, you also > have to enforce -- either as part of the problem formulation, or as a > postprocess step for the output of Mesquite -- that the new nodes still > need to lie on the geometry as described before. In other words, noda

[deal.II] Re: mesh quality during deformations using MappingQEulerian

> > With Mesquite I believe one would essentially redefine a triangulation in > whatever format it wants (I think you provide nodal coordinate and > connectivities), tell it which nodes remain fixed (i.e. provide some > constraints) and then simply specify some algorithm with which it must >

[deal.II] Re: mesh quality during deformations using MappingQEulerian

On Wednesday, November 16, 2016 at 4:40:05 PM UTC-5, Jean-Paul Pelteret wrote: > > Hi Tom, > > An alternative to implemented a complex mesh update strategy is to define > an initial mesh that, when mapped via your Euler vector, produces a higher > quality mesh. For example, you could remove > <

Re: [deal.II] Interfacing with Mesquite

Oded, I am working on an evolving codimension-1 surface, representing deformations through the MappingQEulerian object, and running into what I presume is a similar problem as you have - tangential mesh distortion. How has your experience with Mesquite turned out? Specifically, how has the su

[deal.II] Re: Interfacing with Mesquite

Oded, Thank you for your assessment. Mediocre documentation combined with a stalled development is certainly a recipe for frustration. I have recently obtained a license for cubit - I would like to ask you about that but it will have to wait until the middle of next week, I'm a little busy ri

[deal.II] Re: mesh quality during deformations using MappingQEulerian

J-P (and hopefully Oded is still reading this thread, too), I was looking into Mesquite over the weekend and noticed that the project seems to have stalled, and also that the doxygen documentation was a little weak. I also reached out to Oded (from this thread: https://groups.google.com/forum/

Re: [deal.II] Interfacing with Mesquite

> commercial users called Trelis. See: > > http://www.csimsoft.com/ > > > > I would definitely recommend the use of Cubit/Trelis. > > > > I will be happy to answer further questions, if you have any. > > > > Best, > > Oded > > >

Re: [deal.II] Re: mesh quality during deformations using MappingQEulerian

J-P, I have taken some good advice from Oded (it turns out we work one building over from each other!), and I started using Cubit to smooth my meshes. Cubit requires a license but it does seem to have some usage within the dealii community, so I hope whatever comes out of this is useful for somebo

Re: [deal.II] Re: mesh quality during deformations using MappingQEulerian

> Its great that you're getting somewhere with this! > > J-P, you (and the entire dealii community) are a paragon of encouragement, and I appreciate that very much. After I wrote you yesterday I was able to figure out how to 'play a journal file' from within an embedded python session within deali

Re: [deal.II] fe_function_values

Bernard, you probably want to take a look at Step 18 in the tutorial, specifically the move_mesh() function. In short, you will want to loop through the cells of your triangulation and call exactly cell->vertex(i)

Re: [deal.II] fe_function_values

Benard, (sorry for misspelling your name earlier) It's not clear to me whether or not you've solved your problem at this point. Timo's reply above is the precise reply to your original question: > I want to use fe_function_values to get values for the entire of degrees > and am using > fe_v