Elmer Discussion Forum

Hi,

I noticed some examples in which boundary conditions in stressSolver have been set using the keyword "displacement n load".

I could not find this keyword in the ElmerModelsManual. Could somebody explain its usage, please.

Apology if this information is already there in the documentation. Honestly, I tried to find it.

Thanks in advance,
Ajit

Hi,
Can you exactly mention which example is that ?
I guess that as i see some examples with So, it is something like For further information, i hope this can be useful for us.
viewtopic.php?f=4&t=2286&p=6884&hilit=d ... load#p6884
This post says that this notation is used for "Target Nodes" and "Target Coordinates" in BCs.

In summary, this "displacement n load" is not for mentioning "displacement and load" but it is to mention for the "target nodes" and "target coordinates" by using the number "n = 1,2,3" between displacement and load.
So, what i guess is that for a given "target coordinate" setting in BCs, displacement 1 load means the load component related to the displacement in x (more practically the normal) direction .
yours
annier

Hi,

Here is an example usage.

http://numsim.de/cantilever2d.html

" ...what i guess is that for a given "target coordinate" setting in BCs, displacement 1 load means the displacement in x direction for a given load. "......

May be you are right, but in the above example, displacement 1 load = Real -10e3. If "displacement 1 load" means displacement of the coordinate in the x direction then -10^3 is a pretty huge displacement constraint, is it not?

My guess is that this keyword is for specifying point loads.

Thanks,
Ajit

Hi Ajit,
Yes you are right.
i made a hasty decision on the basis of i=1,2,3 as they mean for the commonly used x,y,z directions.
Then, it(displacement 1 load) might be the component of the load related to displacement in the x (= 1) direction. However, '1' here does not explicitly mean the global x direction but it is something like normal direction to the object concerned. in the example you have mentioned, displacement 1 load means the load normal/perpendicular to the surface/line with the respective target coordinate point - and for these two normal loads, the corresponding two target coordinate points on the surface/line are (2.0, 0.0) and (2.0, 0.3) . The positive direction for load at (2.0,0.0) means "normally outward" and negative direction for load at (2.0,0.3) means "normally inward" to the surface".
Suppose, If additional vertical load were also put in the example at these two points, it would be represented as displacement 2 load for it being tangential to the surface.

and if load in both normal and tangential directions are shown for a given force.
then, in 2D,
and for a 3D geometry In Elmer, 1,2,3 mean x,y and z and are used x =1= normal to the object, y =2= tangential to the object and z =3= remaining direction for 3D geometry. or, now you can just use 1,2 and 3 directions. These x,y,z are localized to the region of analysis and are thus represented as 1,2,3.
If two surfaces for a same geometry are not coplanar, then the directions '1' for these two different surface are not same. If two lines for a same geometry are not collinear, then the directions '1' for these two different surface are not same. Can we say now that the coordinate system in Elmer are local? This is how i understood about the coordinate system in Elmer solver.

Note: I have made update in my earlier post for further reference.
yours
annier

Hi

The "Displacement i load" is indeed a nodal force in direction i. For example, in FSI you could compute "Flow Solution i Loads" from the solution and associate it as a nodal term on the r.h.s. when solving the structual equation with Or something like that. Of course you could apply the node just on one point if your BCs are applied just on one node.

-Peter

Hi Peter,
Thank you for helping us understanding about these concepts. The illustration you have given can be utilized as a coupling tool in FSI.
Yours
Annier

Hi Peter,

I am indeed using this functionality to couple a Emer CSD code with a CFD solver, where I obtain the nodal forces for structural boundary nodes from the CFD solver. I use the following boundary condition in Elmer sif file to impose the nodal loads obtained from CFD solver:

Boundary Condition 2
Name = 'Aerodynamic loading'
Target Boundaries(2) = 5 6
Displacement 1 load = Real 0.0
Displacement 2 load = Real 0.0
Displacement 3 load = Real Variable time
Real Procedure "AERO_PRESSURE" "GETFORCE"
END

However, I am having trouble assigning nodal forces to boundary nodes using 'Displacement i loads. Can you please clarify following concerns regarding its usage:

1) What does '3' means in 'Displacement 3 load'? Is it global Z direction or is it normal to the boundary that contains the corresponding boundary element?
2) is there any difference between 'Deflection 3 load' and 'Displacement 3 load'? Turns out there is s difference in the time-history I obtained using Deflection? Can you explain the difference?

I am attaching the concerned files (Sif file and UDF file) for your understanding.

Thanks,
Rimple

Hi

Component "3" of displacement load refers to the z-component of the force, or if using normal-tangential coordinate system to the 2nd tangent direction. In N-T system you typically would like to enforce the normal direction i.e. the 1st component.

For plate solver the deflection has also three components. The 1st is the normal displacement and the 2nd and 3rd the derivatives of the displacement in the two plane directions.

-Peter