Unable to mesh the geometry

  • 282 Views
  • Last Post 01 September 2018
KumarVaibhav posted this 29 August 2018

I have a geometry of a helmet with ribs. I applied the midsurface property in DM and then moved to the mesher.However i am unable to create the mesh Can some one tell me what is preventing to be meshed.I am on 16 academic version.Pls find the geometry file.Peter could you please guide me on this.

Order By: Standard | Newest | Votes
kkanade posted this 29 August 2018

Hello, 

Please check geometry in DM using Tools - Analysis Tools - Fault Detection. Usually if geometry has any errors, it may fail in meshing. Also you can right click on error message in meshing and check the problematic geometry and then take corrective action in DM or in meshing itself. In meshing you can change size or can use virtual topology with problematic geometry. 

Regards, 

Keyur

  • Liked by
  • KumarVaibhav
peteroznewman posted this 31 August 2018

Here is the mesh from the attached file above.

Kumar, please attach a wbpz archive file and I will take a look.

Regards,

Peter

 

  • Liked by
  • KumarVaibhav
peteroznewman posted this 31 August 2018

Kumar,

If you refine the mesh (make smaller elements), do those numbers remain stable or do they move around? You need to complete the mesh refinement study to know that the numbers accurately represent the response of the geometry.

Regards,

Peter

  • Liked by
  • KumarVaibhav
rwoolhou posted this 30 August 2018

Looking at the image have you put the block in the correct location? It looks like whatever you're hitting the helmet with is already partly inside it? 

The attached image only shows the the failed mesh, and the helmet still looks to have a thickness. Please post images of the geometry, but also work through how it's built to see if you can find the problems. Also note, it's quite possible for the geometry to be valid even if you've defined it incorrectly: valid doesn't mean it's correct. 

KumarVaibhav posted this 30 August 2018

pls take a look at the file attached.Do i need to define contacts in drop test when the block falls on top surface?Please look

 

rwoolhou posted this 30 August 2018

Not sure, but I suspect you'll need to do something so the two bodies know about each other. 

How are you getting on working through the geometry to make it all work? 

KumarVaibhav posted this 31 August 2018

Peter i was able to generate the mesh with applying clean bodies operation and changing the mesh settings.But i have a doubt.actually i made two helmets one with the reinforcement ribs as circular cross section and one with rectangular cross section.i checked both for a drop test and i got deformation identical in both cases.i got 0.37 for rectangular ribs and .35 for circular ribs using ABS as material.which is better to use a rib with rectangular cross section or a circular one.i think rectangular section have good section modulus but in ansys d results were differed.pls guide more on this

KumarVaibhav posted this 01 September 2018

could you further eleborate a little more.Actually when i change the mesh setting from curvature to fixed to proximity and various settings i saw the mesh was unable to generate most of the times so i resorted to proximity only.what step do i need to do?Kindly guide me

 

Do u have any idea based on physics which cross section has least chances to deform circular or rectangular so that i can validate my ansys results if its right or not?

peteroznewman posted this 01 September 2018

If the circle is inside the rectangle, then there is less material, so a physics argument is that it should take less force to deform the circle than the rectangle.

If you get two successful meshes on the same geometry, and one has an edge length of 1.5 mm and the other has an edge length of 1.0 mm, if the deformation remains at 0.35, then that is evidence of a mesh-independent result.  If you repeat for the other geometry and that value doesn't change for two successful meshes, then you can conclude that there is a difference between the geometries.

Ideally, you would also create a mesh with an element edge length of 0.75 mm and a coarse mesh of element edge length of 2.25 mm. Then you would have four points to plot the result vs. edge length and see if the value has converged on the exact value. See this discussion.

KumarVaibhav posted this 01 September 2018

ok got your point.will do iterations then

but tell me about the physics argument.What do u mean when you say the circle is inside the rectangle?

do u mean to say the skeleton ribs with(represented inside the helmet as crossed line body kind of) rectangular cross section is less deformable to the drop test i am interested in than over the ribs having circular cross section?i wanna validate it with ansys so pls elaborate more on the physics argument and how to cross check with ansys,it would be great learning for me

peteroznewman posted this 01 September 2018

Here is the thickness of the helmet shell with a rectangular or circular rib on it. The rectangle has a higher area moment of inertia than the circle profile of rib.

kkanade posted this 31 August 2018

Hi, 

Please use insert image instead of attach so that we can see those images. 

 

Regards,

Keyur

Close