General question regarding modelling in Elmer

General discussion about Elmer
Sayan
Posts: 112
Joined: 06 Dec 2020, 12:44
Antispam: Yes

General question regarding modelling in Elmer

Post by Sayan »

Hiiii,
I have two queries regarding modelling in Elmer,
i) Is it possible to assign different modulus of elasticity to some fraction of length of a beam? e.g If I want to assign middle 1/3rd length of beam a different modulus of elasticity than other part of the beam, is it possible?
ii) Whether loading can be assigned using some time history graph? i.e if I want to apply some past earthquake time history to a structure, is it possible to assign this as input?
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by kevinarden »

Yes on both.
1. You could make the middle 1/3 of the beam a different body and therefore it can be assigned its own material.
2. You can use variable time to make a force a function of time, then use a table to provide the function.

Force 1 = Variable Time
Real
0.0 0.0
.01 10.0
.02 20.0
.03 100.0
.04 50.0
.05 0.0
1.00 0.0
End
Sayan
Posts: 112
Joined: 06 Dec 2020, 12:44
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by Sayan »

Ok....got it..But using Freecad when I am exporting it to Elmer I have seen always it takes one body, though there may be several boundary present in the model. How do I then make different body for single beam?
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by kevinarden »

Not familiar with FreeCad, in solome you would partition the solid and group the volume elements in each partition to different groups, then they become separate bodies. Same in gmsh, you can use physical groups of elements to control bodies.
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by kevinarden »

I went to freecad and created 3 volumes, then used compound to make them one model. I exported the compound shape to gmesh and they stayed 3 volumes, Now I can use the physical group volume command to make each one a body.
volumes.png
(450.6 KiB) Not downloaded yet
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: General question regarding modelling in Elmer

Post by Rich_B »

There is an old non-gui tutorial demonstrating variable Young's Modulus, and it can be found here:

Code: Select all

 elmerfem-manuals/tutorials-CL-files/FluidStructureBeam
The actual tutorial text is not present in the current manuals, but it can be found in the 2009 pdf files here:

https://www.nic.funet.fi/pub/sci/physic ... orials.pdf

and look for tutorial 16, 'Fluid flow around an elastic beam'.

Rich.
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: General question regarding modelling in Elmer

Post by Rich_B »

Here's an excerpt from the tutorial:
In this problem, external Fortran functions are used to express a parabolic velocity inflow profile and to
define a variable stiffness for the deforming mesh. The stiffness of the mesh is controlled by the Youngs
modulus. In the case of deforming mesh, the absolute value of the parameter is not important but its changes
over the domain are influential. The following function defines the mesh to be stiffer near the corner of the
beam where the mesh is deformed most. The idea is to avoid too irregularly shaped elements in that area.

Code: Select all

FUNCTION Youngs( Model, n, x ) RESULT( s )
USE Types
IMPLICIT NONE
TYPE(Model_t) :: Model
INTEGER :: n
REAL(KIND=dp) :: x,s,xx,yy
xx = Model % Nodes % x(n)
yy = Model % Nodes % y(n)
s = 1.0d0 / SQRT( (xx-11.0)**2 + (yy-4.9)**2 )
END FUNCTION Youngs
Sayan
Posts: 112
Joined: 06 Dec 2020, 12:44
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by Sayan »

Hiii,
Thanks Kevin for the effort.....From the turorial of FreeCad last day I somehow managed to model 3 bodies and mesh it in gmsh using transfinite mesh definition and 3D optimize mesh... I analysed it...It executed well..But when I look into the result I find that only the last volume is responding to axial load..other two volumes are having zero deformation...I even skipped the initial condition for two bodies..But got same result..somehow the boundaries between two volume is assumed as fixed in Elmer I guess, though I have defined boundary only for the first volume which is fixed...How can I get rid of this??
Attachments
case.sif
(2.64 KiB) Downloaded 144 times
damageAx.msh
(45.94 KiB) Downloaded 144 times
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by kevinarden »

Looks like coincident nodes so the 3 bodies are not connected. You need to do a coherence command before you mesh to eliminate.
Sayan
Posts: 112
Joined: 06 Dec 2020, 12:44
Antispam: Yes

Re: General question regarding modelling in Elmer

Post by Sayan »

Hiii,
I used coherence command after adding physical group.then I mesh it...but same result...When using connect command it's showing only one volume in Gmsh...
Attachments
beam21.msh
(45.94 KiB) Downloaded 138 times
Post Reply