[vtkusers] ExodusII problem or bug?? Please help

Tue May 17 09:03:33 EDT 2011

Thanks for the help.

1. I tried to figure out but it seems to be a little bit beyond me.
Given the directions I will try to fix the problem myself and maybe then I
will submit my first patch to vtk!

2. I can confirm that the node orders for quad9/bi-quadratic-quad are
identical.
So when quadratic quad is used instead of bi-quadratic quad, the missing
node is the last one (8).

3 -- 6 -- 2
|       |      |
7 -- 8 -- 5
|       |      |
0 -- 4 -- 1

However, one question is how does  vtkExodusIIReader
differe from vtkExodusReader.
If I open the file and use PrintSelf methods They produce slightly different
information:

vtkExodusIIReader:
*
Debug: Off
Modified Time: 34
Reference Count: 2
Registered Events: (none)
Executive: 0x1536d60
ErrorCode: Success
Information: 0x1533ed0
AbortExecute: Off
Progress: 1
Progress Text: (None)
FileName: sphere-6.exo
XMLFileName: (null)
DisplayType: 0
TimeStep: 0
TimeStepRange: [0, 0]
ExodusModelMetadata: OFF
PackExodusModelOntoOutput: ON
ExodusModel: 0
Metadata:
  Exoid: -1
  AppWordSize: 8
  DiskWordSize: 8
  ExodusVersion: 4.57
  ModelParameters:
    Title: cubit(/home/dansari/Desktop/sphere-6.exo): 05/13/2011: 16:08:57
    Dimension: 3
    Nodes: 26
    Edges: 0
    Faces: 0
    Elements: 6
    Edge Blocks: 0
    Face Blocks: 0
    Element Blocks: 1
    Node Sets: 0
    Edge Sets: 0
    Face Sets: 0
    Side Sets: 0
    Element Sets: 0
    Node Maps: 0
    Edge Maps: 0
    Face Maps: 0
    Element Maps: 0
  Time steps (0):
  TimeStep: 0
  HasModeShapes: 0
  ModeShapeTime: -1
  Blocks:
    Element block 1 "Unnamed block ID: 1 Type: QUAD9 Size: 6" (6)
        FileOffset: 1
        CachedConn: 0x153b380 (1)
        PointMap: 26 entries, ReversePointMap: 26 entries
        Type: QUAD9
        Bounds per entry, Node: 9 Edge: 0 Face: 0
        Attributes (0):
  Sets:
  Maps:
  Array Cache:
    Debug: Off
    Modified Time: 10
    Reference Count: 1
    Registered Events: (none)
    Capacity: 128 MiB
    Size: 0.00390625 MiB
    Cache: 0x1536d10 (4)
    LRU: 0x1536d40
  SqueezePoints: 1
  ApplyDisplacements: 1
  DisplacementMagnitude: 1
  GenerateObjectIdArray: 1
  GenerateFileIdArray: 0
  FileId: 0*

ExdosuReader:

*
Modified Time: 85
Reference Count: 2
Registered Events: (none)
Executive: 0x2310390
ErrorCode: Success
Information: 0x2307ed0
AbortExecute: Off
Progress: 1
Progress Text: (None)
GenerateBlockIdCellArray: On
GenerateGlobalElementIdArray: On
GenerateGlobalNodeIdArray: On
PackExodusModelOntoOutput: On
ApplyDisplacements: On
ExodusModelMetadata: Off
File Name: sphere-6.exo
XML File Name: (none)
Title: cubit(/home/dansari/Desktop/sphere-6.exo): 05/13/2011: 16:08:57
Dimensionality: 3
NumberOfUsedNodes: 20
NumberOfNodesInFile: 26
NumberOfUsedElements: 6
NumberOfElementsInFile: 6
NumberOfBlocks: 1
  6 elements in block 1
NumberOfTimeSteps: 0
NumberOfPointArrays: 0
NumberOfCellArrays: 0
NumberOfSideSets: 0
NumberOfNodeSets: 0
TimeStep: 0
TimeStepRange: 0 -1
DisplacementMagnitude: 1
DisplayType: 1
HasModeShapes: 0
*
The output from ExodusIIReader clearly suggests that my elements are 9 node
quads!

Please advise

Davood

On Fri, May 13, 2011 at 12:02 PM, David Thompson <dcthomp at sandia.gov> wrote:

> Hi Davood,
>
> The ExodusII reader was created before VTK had Soeren Gebbert's 9-node
> quadratic elements, so the 8-node quad was used as an approximation. It
> appears that vtkExodusIIReaderPrivate::DetermineVtkCellType needs to be
> changed to use the bi-quadratic quad, but there may be more depending on the
> order nodes are specified in by Exodus vs. VTK (see
> vtkExodusIIReader::GetCacheOrRead around line 2225 for what's done with
> hexahedra).
>
> If someone wants to create a patch, I'll be happy to review it in Gerrit.
>
>        David
>
>
> On May 12, 2011, at 9:17 , Davood Ansari wrote:
>
>  Hi all
>>
>> I am having some trouble with reading ExodusII files generated from Cubit
>> 12.
>> The elements are supposed to be (and based from what I see in the .exo
>> file are)
>> biqudratic quads that have 9 points. When I read the file using
>> vtkExodusIIReader,
>> however, I get quadratic quads with 8 sampling points.
>>
>> Any idea what is going one?
>>
>> Attached is the header for my .exo mesh file, as you see num_nod_per_el1 =
>> 9 ; but I still get 8 point quadratic quad in VTK??
>>
>>
>> netcdf sphere {
>> dimensions:
>>    len_string = 33 ;
>>    len_line = 81 ;
>>    four = 4 ;
>>    time_step = UNLIMITED ; // (0 currently)
>>    num_dim = 3 ;
>>    num_nodes = 386 ;
>>    num_elem = 96 ;
>>    num_el_blk = 1 ;
>>    num_qa_rec = 1 ;
>>    num_el_in_blk1 = 96 ;
>>    num_nod_per_el1 = 9 ;
>> variables:
>>    double time_whole(time_step) ;
>>    char qa_records(num_qa_rec, four, len_string) ;
>>    char coor_names(num_dim, len_string) ;
>>    char eb_names(num_el_blk, len_string) ;
>>    int elem_map(num_elem) ;
>>    int eb_status(num_el_blk) ;
>>    int eb_prop1(num_el_blk) ;
>>        eb_prop1:name = "ID" ;
>>    int connect1(num_el_in_blk1, num_nod_per_el1) ;
>>        connect1:elem_type = "QUAD9" ;
>>    double coord(num_dim, num_nodes) ;
>>
>> // global attributes:
>>        :api_version = 4.57f ;
>>        :version = 4.57f ;
>>        :floating_point_word_size = 8 ;
>>        :file_size = 0 ;
>>        :title = "cubit(/home/dansari/Desktop/sphere/sphere.exo):
>> 03/22/2011: 19:24:22" ;
>>
>> _______________________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Please keep messages on-topic and check the VTK FAQ at:
>> http://www.vtk.org/Wiki/VTK_FAQ
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtkusers
>>
>
>
>

-- 

    work as if you'd stay forever,
prepare as if you'd leave tomorrow
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20110517/8ef5850e/attachment.htm>