I did a stress analysis with the mesh turned on of a steel plate with holes in it. I have created local meshes for each surface such as the small holes have a small mesh, the large hole has a medium sized mesh, and the top face has a wider mesh. My issue is in determining how many elements are created throughout the depth of the plate. I have a client that needs to know how the software is creating the 3d mesh elements. I can only display the surfaces and cannot see a full 3D mesh to see what the inner mesh looks like.
My work around is to stack 7 thin plates to make up the thickness of the single plate so I can manually create a 3D stress field, but it still does not answer the question for the future.
You can know for certain that it is at least one layer of elements thick. We pass the surface mesh off to our 3rd party volume mesher and they determine the volume mesh.
Even if you split the short "thickness" faces, to force the surface mesh to be more than one element thick, again the 3rd party mesher produces the volume mesh and we cannot guarantee how many "layers" will be created.
I should probably include more information why we don't necessarily need many layers thick of elements. It comes down to the element type used, and how the solver processes the element.
If using brick elements (not available in Inventor Stress Analysis) and the order of polynomial is 1 (linear displacement field) you may need more than one layer of elements since using just one layer will be overly 'stiff'. The displacement results will be under-estimated, which will cause the strain, subsequent stress to also be under-estimated. We can achieve better results using 'linear elements' by using more layers of elements through the thickness.
In Inventor Stress Analysis, we use tetrahedral elements with midside nodes. We also employ higher order polynomial elements. For parts we automatically do 2 p-refinements (cubic displacement field) and for assemblies we do one p-refinement to go from linear to quadratic displacement field.
In many cases, this can achieve comparable results to using brick elements multiple layers thick.
More information about adaptive h and p refinement (convergence and mesh refinements) can be found here: http://usa.autodesk.com/adsk/servlet/item?siteID=1
If the geometry is very thin, such as sheet metal parts or pressure vessels, using shell elements are preferable over the solid elements. Beginning in R2013, we introduced shell elements.
Please let us know if you have additional questions, comments or suggestions.
Best regards, -Hugh
I am currendly analyzing a turret shaft with shell slements. I was wondering if it is possible to use 8 noded quadratic shell elements in stead of 4 noded linear shell elements.
I hope you could help,
best regards, Rick
oo i forgot to say that i am working with inventor proffesional 2013. I was also wondering if i could use Quad is stead of tria's. When analyzing in inventor professional the software automaticaly creates tria's and with the mesh stetting i could not change the trias to a quad element.
For shell elements, we only have the DKT (Discrete Kirchoff Triangle) elements available. Not plate or membrane elements that are quads.
The DKT elements are six DOF, to allow both translation and rotation for both plane and bending stress. They have 3 corner nodes, and midside nodes. These assume a linear stress distribution through the thickness.
I've logged your request for the 8 node quad shell elements as 1494059 for consideration in a future release.
Thanks and Best regards, -Hugh
Thanks for your reaction,
I have another question regarding stress analysis.
See attachment for a concept of this stress analyze.
In this analyze I was wondering if I could make a graph of the stresses or the forces that occur on different angels of the shaft. In this concept I have used solid element just to illustrate, but most of the time I’m using shell elements.
I also read on a forum that the data of the FINS, FSAT, FMSH files could not be used and only Inventor AIP is able to use the stress Analysis result files. Is this also the case with inventor professional 2013?
So the question is if it is possible to make a graph of the stresses/forces that occur on the shaft or use data results and create a graph in Excel.
It would be wonderful if you could help me out,
Log into access your profile, ask and answer questions, share ideas and more. Haven't signed up yet? Register