January 2014 - cgma-dev - mailman3.cels.anl.gov

Re: [cgma-dev] Facet/Cholla engine
by David Thompson 18 Jan '14

18 Jan '14

Hi Jane, > I'll put this into our bug system and hopefully find chances to fix it either myself or by others. I've attached a version of the tetrahedron code that demonstrates split_surface and feature_angle do nothing that you can include in the bug report. > Sorry for the inconvenience. Thanks for looking into things. We are also looking at using MOAB to represent models composed of multiple faceted surfaces -- so it is not a really high priority, but it would help me with the CGM work I am doing even if we use MOAB for faceted models. David

1 0

Re: [cgma-dev] CGM Instance
by Patrick Shriwise 17 Jan '14

17 Jan '14

On 01/16/2014 11:47 AM, Paul Wilson wrote: > On 01/16/2014 11:30 AM, Patrick Shriwise wrote: >> On 01/15/2014 03:33 PM, David Thompson wrote: >>>> I'm currently writing tests for MOAB's Read_CGM. In this process >>>> I'm loading an ACIS file multiple times in different MOAB >>>> instances. This causes the file loads to 'stack' upon one another >>>> when I do this however. Any recommendations on how to clear the CGM >>>> data if I want to load a file more than once? >>> Have you tried >>> >>> GeometryQueryTool::instance()->delete_geometry(); >>> >>> or does that not do what you want? >>> >>> David >> Haven't tried that yet. Going to give it a shot today. Do you think >> the function is accessible from within a MOAB instance? >> > Patrick, > > This is available through the same (or similar) CGM interface that is > used to to call get_graphics() in ReadCGM. You'll need to set it up > in your tests probably. > > Paul > Just got this implemented. Works well. Thank you! Patrick -- Patrick C. Shriwise Research Assistant University of Wisconsin - Madison Engineering Research Building - Rm. 428 1500 Engineering Drive Madison, WI 53706 (608) 446-8173

1 0

Re: [cgma-dev] Facet/Cholla engine
by Jane Hu 17 Jan '14

17 Jan '14

Hmmm... I'll put this into our bug system and hopefully find chances to fix it either myself or by others. Sorry for the inconvenience. Jane On Fri, Jan 17, 2014 at 5:57 AM, David Thompson <david.thompson(a)kitware.com>wrote: > Hi Jane, > > > ... Other than sending facet group separately to make individual > surfaces, I understand it's the most accurate approach, can you try to give > an feature_angle and set split_surfaces=true to see if, first of all, it > gives 4 surfaces? > > I tried using feature_angle+split_surfaces but always get a single surface > back. > > Thanks, > David > > > On Wed, Jan 15, 2014 at 2:47 PM, David Thompson < > david.thompson(a)kitware.com> wrote: > > Hi Jane, > > > > > I debugged your program today, it seems that the code does have some > bug when trying to create individual surface and body through individual > build_facet_surface() calls, here you are trying to create 4 surfaces and > 4 bodies. > > > > I do need 4 surfaces, but just 1 body. I thought that by calling > make_facet_shell() with all 4 of the surfaces returned by > build_faceted_surface() that the surfaces might be stitched (or at least > included in the same shell and thus body). Is that wrong? > > > > > However, I am not familiar with the code enough to make any changes. > > > > Thanks for taking a look at it; it's good to know it's not all my fault. > :-) > > > > > I did checked that if you pass in all the facets, with empty point > list, it can create a single FacetBody and you can export it. > > > > > > I attached the changed code for you, don't know if it's sufficient for > your purpose. > > > > The problem is that the faceted model I am trying to import has multiple > model-faces (each with multiple triangles) on the same shell, which I would > like to model with CGM. Among other things, I need to attach attributes > (ToolData) to portions of a shell, not an entire shell at a time. Is there > another way to do that? > > > > Thanks, > > David > > > > > > > > -- > > Jane Hu > > > > Asst. Researcher > > Dept. of Engineering Physics > > UW @ Madison > > > > "And we know that for those who love God, that is, for those who are > called according to his purpose, all things are working together for good." > (Romans 8:28) > > -- Jane Hu Asst. Researcher Dept. of Engineering Physics UW @ Madison "And we know that for those who love God, that is, for those who are called according to his purpose, all things are working together for good." (Romans 8:28)

1 0

Re: [cgma-dev] Facet/Cholla engine
by David Thompson 17 Jan '14

17 Jan '14

Hi Jane, > ... Other than sending facet group separately to make individual surfaces, I understand it's the most accurate approach, can you try to give an feature_angle and set split_surfaces=true to see if, first of all, it gives 4 surfaces? I tried using feature_angle+split_surfaces but always get a single surface back. Thanks, David > On Wed, Jan 15, 2014 at 2:47 PM, David Thompson <david.thompson(a)kitware.com> wrote: > Hi Jane, > > > I debugged your program today, it seems that the code does have some bug when trying to create individual surface and body through individual build_facet_surface() calls, here you are trying to create 4 surfaces and 4 bodies. > > I do need 4 surfaces, but just 1 body. I thought that by calling make_facet_shell() with all 4 of the surfaces returned by build_faceted_surface() that the surfaces might be stitched (or at least included in the same shell and thus body). Is that wrong? > > > However, I am not familiar with the code enough to make any changes. > > Thanks for taking a look at it; it's good to know it's not all my fault. :-) > > > I did checked that if you pass in all the facets, with empty point list, it can create a single FacetBody and you can export it. > > > > I attached the changed code for you, don't know if it's sufficient for your purpose. > > The problem is that the faceted model I am trying to import has multiple model-faces (each with multiple triangles) on the same shell, which I would like to model with CGM. Among other things, I need to attach attributes (ToolData) to portions of a shell, not an entire shell at a time. Is there another way to do that? > > Thanks, > David > > > > -- > Jane Hu > > Asst. Researcher > Dept. of Engineering Physics > UW @ Madison > > "And we know that for those who love God, that is, for those who are called according to his purpose, all things are working together for good." (Romans 8:28)

1 0

Stretched grid implementation
by Alexander Vizard 16 Jan '14

16 Jan '14

Hello, I am looking tom implement the EBMesh software to work with a 3D cartesian grid that is stretched in the x and z directions. In terms of the code, my understanding is that I want to create a vtxMesh by reading i_arr, j_arr, and k_arr (the locations of the points, not the index values) from a file, rather than calculating them based on a total dimensional length and a number of gridpoints. Can you verify that to do this, these values must be changed in SCDMesh.hpp, EBMesher.cpp and EBMesher.hpp. However, a stretched grid will mean a variation of the grid spacing (currently, dx, dy, and dz, though not necessarily the same, are constants, rather than an array of values). Is the EBMesh software capable of calculating these dx values at each cell already, or would this also need to be added to the existing code structure to get the intended results? Thanks for your help, and I look forward to your reply. -- Alexander Vizard Research Assistant George Washington University avizard(a)gwmail.gwu.edu

1 0

Re: [cgma-dev] CGM Instance
by Paul Wilson 16 Jan '14

16 Jan '14

On 01/16/2014 11:30 AM, Patrick Shriwise wrote: > On 01/15/2014 03:33 PM, David Thompson wrote: >>> I'm currently writing tests for MOAB's Read_CGM. In this process I'm >>> loading an ACIS file multiple times in different MOAB instances. >>> This causes the file loads to 'stack' upon one another when I do >>> this however. Any recommendations on how to clear the CGM data if I >>> want to load a file more than once? >> Have you tried >> >> GeometryQueryTool::instance()->delete_geometry(); >> >> or does that not do what you want? >> >> David > Haven't tried that yet. Going to give it a shot today. Do you think > the function is accessible from within a MOAB instance? > Patrick, This is available through the same (or similar) CGM interface that is used to to call get_graphics() in ReadCGM. You'll need to set it up in your tests probably. Paul -- -- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -- Paul Wilson ~ UW-Madison ~ 608-263-0807 ~ cal: http://bit.ly/pphw-cal Professor, Engineering Physics. ~ http://cnerg.engr.wisc.edu Faculty Director, Advanced Computing Infrastructure

1 0

Re: [cgma-dev] CGM Instance
by Patrick Shriwise 16 Jan '14

16 Jan '14

On 01/15/2014 03:33 PM, David Thompson wrote: >> I'm currently writing tests for MOAB's Read_CGM. In this process I'm loading an ACIS file multiple times in different MOAB instances. This causes the file loads to 'stack' upon one another when I do this however. Any recommendations on how to clear the CGM data if I want to load a file more than once? > Have you tried > > GeometryQueryTool::instance()->delete_geometry(); > > or does that not do what you want? > > David Haven't tried that yet. Going to give it a shot today. Do you think the function is accessible from within a MOAB instance? -- Patrick C. Shriwise Research Assistant University of Wisconsin - Madison Engineering Research Building - Rm. 428 1500 Engineering Drive Madison, WI 53706 (608) 446-8173

1 0

Re: [cgma-dev] Facet/Cholla engine
by Jane Hu 16 Jan '14

16 Jan '14

Ok, my fault, I did notice that you just wanted to create 4 individual surfaces, but kept shell and body making out of the loop, so they should be coming one. They are correct logically and codewise. Just the bug stepped in when the code is doing the second surface, while querying for the shared edge( with the first surface created) info, it failed and returned a NULL pointer, thus resulted the crash. Other than sending facet group separately to make individual surfaces, I understand it's the most accurate approach, can you try to give an feature_angle and set split_surfaces=true to see if, first of all, it gives 4 surfaces? Jane On Wed, Jan 15, 2014 at 2:47 PM, David Thompson <david.thompson(a)kitware.com>wrote: > Hi Jane, > > > I debugged your program today, it seems that the code does have some bug > when trying to create individual surface and body through individual > build_facet_surface() calls, here you are trying to create 4 surfaces and > 4 bodies. > > I do need 4 surfaces, but just 1 body. I thought that by calling > make_facet_shell() with all 4 of the surfaces returned by > build_faceted_surface() that the surfaces might be stitched (or at least > included in the same shell and thus body). Is that wrong? > > > However, I am not familiar with the code enough to make any changes. > > Thanks for taking a look at it; it's good to know it's not all my fault. > :-) > > > I did checked that if you pass in all the facets, with empty point list, > it can create a single FacetBody and you can export it. > > > > I attached the changed code for you, don't know if it's sufficient for > your purpose. > > The problem is that the faceted model I am trying to import has multiple > model-faces (each with multiple triangles) on the same shell, which I would > like to model with CGM. Among other things, I need to attach attributes > (ToolData) to portions of a shell, not an entire shell at a time. Is there > another way to do that? > > Thanks, > David -- Jane Hu Asst. Researcher Dept. of Engineering Physics UW @ Madison "And we know that for those who love God, that is, for those who are called according to his purpose, all things are working together for good." (Romans 8:28)

1 0

Re: [cgma-dev] CGM Instance
by David Thompson 15 Jan '14

15 Jan '14

> I'm currently writing tests for MOAB's Read_CGM. In this process I'm loading an ACIS file multiple times in different MOAB instances. This causes the file loads to 'stack' upon one another when I do this however. Any recommendations on how to clear the CGM data if I want to load a file more than once? Have you tried GeometryQueryTool::instance()->delete_geometry(); or does that not do what you want? David

1 0

Re: [cgma-dev] CGM Instance
by Grindeanu, Iulian R. 15 Jan '14

15 Jan '14

Isn't this related to heavy use of "singleton" pattern by CGM ? I think that it is hard to get rid of some of the classes in cgm. so even if you have a new moab instance, some cgm classes might linger around. Although, I thought that a new moab instance would automatically mean a new cgm instance. Probably you can check in the debugger if you are really getting an old cgm class. Iulian ________________________________________ From: cgma-dev-bounces(a)mcs.anl.gov [cgma-dev-bounces(a)mcs.anl.gov] on behalf of Patrick Shriwise [shriwise(a)wisc.edu] Sent: Wednesday, January 15, 2014 3:28 PM To: cgma-dev(a)mcs.anl.gov Subject: [cgma-dev] CGM Instance Hi all, I'm currently writing tests for MOAB's Read_CGM. In this process I'm loading an ACIS file multiple times in different MOAB instances. This causes the file loads to 'stack' upon one another when I do this however. Any recommendations on how to clear the CGM data if I want to load a file more than once? Thank you, -- Patrick C. Shriwise Research Assistant University of Wisconsin - Madison Engineering Research Building - Rm. 428 1500 Engineering Drive Madison, WI 53706 (608) 446-8173

1 0