Cancel
Showing results for 
Search instead for 
Did you mean: 

JACOBIAN CONFLICT

Experimenter
Experimenter

Hello,

 

I have two Jacobian plots and they have conflicting data from what I can tell.  The Jacobian reading from the mesh toolbox gives an all green while the element contour option gives bad elements almost equal to a jacobian of 1.  Which one is correct, and how do I fix the hex elements that are bad without messing up the entire mesh?

 

Thanks

 

2 REPLIES

Re: JACOBIAN CONFLICT

Solution Partner Phenom Solution Partner Phenom
Solution Partner Phenom

Hello!,

I am happy you are worried for Jacobian mesh quality, in general people do not check mesh quality at all, and is very important to produce quality mesh form the point of view of getting accurate results, more when we have in FEMAP powerful tools to check quality mesh, then let's use it, no excuse!!.

 

Well, first at all you must understand the diference between the various quality checks tools we have in FEMAP:

  • Quality Check in MESHING TOOLBOX: this is exclusively for 2-D surface elements, like Shell CQUAD4 or 2-D Solid plane stress, plane strain & axisymmetric elements, OK?. Then if you have a 3-D Solid CTETRA/CHEXA/CPENTA/CPYRAM elements this is not your command. Also, please note the name of the tool: SURFACE MESH QUALITY

mesh-quality-check-meshing-toolbox.png

 

  • VIEW > ADVANCED POST > CONTOUR MODEL DATA > ELEMENT QUALITY: well, this is a powerfull command, I use it a lot when meshing because I receive interactively direct input about the quality mesh I am doing, here you can control ALL the quality checks, not matter your mesh is 2-D o 3-D, you must choose the right quality check, then you must know what check (Aspect Ratio, Skew Angle, Alternate Taper, Jacobian, etc..) you want to use in each moment. For instance, if I run a 2-D SHELL mesh based in CQUAD4 elements I will check element quality for ALTERNATE TAPER & JACOBIAN as being the two most resttrictive checks, then Aspect Ratio, etc.. Please also note that new FEMAP 11.2 has added Element Quality Checks for NX NASTRAN solver!!. This is of critical importance: now we can see in advanced (before solving!!) if the NX NASTRAN solver will complaint or not with the created mesh, thanks FEMAP guys!!.

contour-model-data-element-quality-new.png

 

  • And finaly you can use the classical "TOOLS > CHECK > ELEMENT QUALITY" COMMAND", that in FEMAP V11.2 includes the specific mesh quality checks for NX NASTRAN solver, then you can see the quality of your mesh just in FEMAP by advanced, without to wait to solve your model with NX NASTRAN and read the GEOMCHECK in the *.F06 file, OK?.

check-element-quality-new.png

 

Regarding your question of how to solve your issue with Jacobian check "without meshing again", well, FEMAP don't have a magic wang, this is the typical problem when creating 3-D solid mesh using horrible WEDGE CPENTA five-sided elements (ie, extruding 2-D 3-nodes triangles to form 3-D six-nodes Wedge solid mesh), simply avoid to have triangles in your 2-D mesh, work the geometry to have regular regions to mesh with perfect QUAD-only elements and you will arrive to 3-D solid CHEXA quality mesh for sure, OK?.

 

Best regards,

Blas.

Blas Molero Hidalgo, Ingeniero Industrial, Director
IBERISA • 48004 BILBAO (SPAIN)
WEB: http://www.iberisa.com
Blog Femap-NX Nastran: http://iberisa.wordpress.com/

Re: JACOBIAN CONFLICT

Experimenter
Experimenter

The hex mesher has automatically created some wedge elements from degenerate hexes.  When is it ok to leave a wedge element alone?  I have attached my stress countour and selected a few wedges.  If the stress peaks are far enough away from the wedges is it ok to leave them?