<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=utf-8">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Reference: ParaView Version 4.3.1<br>
<br>
(Q1) Is it intended that ParaView's EnSight reader support
non-symmetric 2nd-order tensors?<br>
<br>
The attached EnSight formatted file set contains two test cases
defined on a single 4-node quadrilateral in 3D --<br>
<br>
<b>Test Case 1:</b> File fmaego.hist.1st6 designated as "tensor symm
per element"<br>
<br>
This is a symmetric tensor defined with 6 components as <br>
(Txx,Tyy,Tzz,Txy,Txz,Tyz) = (1,2,3,4,5,6).<br>
<br>
<br>
<b>Test Case 2:</b> File fmaego.hist.2st9 designated as "tensor asym
per element"<br>
<br>
This is a symmetric tensor defined with 9 components as <br>
(Txx,Txy,Txz,Tyx,Tyy,Tyz,Txz,Tzy,Tzz) = (1,4,5,4,2,6,5,6,3)<br>
<br>
(Test Case 2 follows the EnSight Gold format specification. It also
is the Tij component sequence specified in the VTK format for a
2nd-order tensor; the VTK format has no special case for a symmetric
tensor.)<br>
<br>
ParaView Response: With the attached file set, ParaView crashes
while reading the case file. An empty message window appears. When
the line for Test-2st9 is removed, from the case-file, PV reads the
file set without any trouble.<br>
<br>
(Just so you know, I am a rabid PV supporter, and greatly appreciate
the arbitrary n-vertex polyhedral finite element support provided by
PV's EnSight reader and displayed by PV.) <br>
<br>
Samuel W Key<br>
FMA Development, LLC<br>
1005 39th Ave NE<br>
Great Falls, Montana 59404<br>
USA <br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
</body>
</html>