Strange results – Plate vs Shell

Strange results – Plate vs Shell

I'm developing a glass project and recently changed my version of Mecway and ended up doing some tests and noticed a big difference in the results

To illustrate the situation, I made an analytical calculation of a glass plate subjected to wind pressure of 2x2.50 with 2Kpa with 12mm

Voltage - 21.18Mpa
deformation 18.21 mm

When I solve this exercise with Mecway, I only consider plates with the thickness described in the materials

Voltage - 24.04Mpa
deformation 18.14 mm

which seems coherent to me, however when I take this same plate and extrude it to the thickness of the project and divide it into 4 stacks the results are very different

Voltage - 14.43Mpa
deformation 4.05 mm

Very far from what would be correct, could anyone give me a clue as to what might be happening?

Comments

  • The mesh is too coarse. Refine x2 leads to 19 MPa and 10 mm deflection.

    Generally, you shouldn't use the linear elements, especially in bending because they're terrible like this. However, the next version of Mecway will have incompatible mode hex8 elements which are much better (24.0 MPa and 18.1 mm for your mesh).

  • Hi Victor

    Thanks for this,

    I'm looking forward to testing the next version, these elements are really useful

    I imagined that when extruding the shell it would have hex8 elements, now I was confused what kind of elements are they?

    sorry for my lack of attention to this.
  • Is there any way to convert the elements with any calculix card?
  • They are hex8 elements but using the standard formulation from textbooks which is over-stiff in bending so it requires a fine mesh through the thickness so that each individual element hardly bends much.

    Yes, you can use the incompatible type with CCX. In the CCX branch of the outline tree, add a custom element type, select Hex8 and enter C3D8I as the CCX element type.

  • edited July 23
    Hi Victor,

    What do you mean by "incompatible mode hex8 elements" and later "incompatible type with CCX". The word incompatible is what is new to me here.

    Since you say that "They are hex8 elements but using the standard formulation from textbooks", does that relate to the Mecway solver, or CCX, or both? If it's just the Mecway solver, couldn't you just update the formulation, rather than add an "incompatible type" element. Sorry, if that's a dumb question. This has me a bit confused.

    Thanks,

    Anthony
  • Yea it's a confusing word, also called bubble functions. I don't fully understand what it means but the "I" in C3D8I stands for incompatible too. So CCX has both types.

    I agree with the principle of just updating the formulation instead of adding an option, but I'm not confident there isn't some disadvantage to it. So at first it'll be an option that's on by default, then in subsequent versions, if nobody seems to need the standard formulation, I intend to remove the option and add warning if you open an old file that uses standard formulation.
  • thanks victor,

    that gave me something to google and it makes more sense now
  • thanks victor
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!