Introducing springs to a point/face as a boundary condition

Numerical methods and mathematical models of Elmer
kevinarden
Posts: 484
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Introducing springs to a point/face as a boundary condition

Post by kevinarden » 26 Mar 2020, 12:54

Both Target Boundaries and Target Nodes for Shell Solver and Spring 4 behaved the same way, the solution did not converge.
4point2.zip
(8.13 KiB) Downloaded 3 times
Kevin

kevinarden
Posts: 484
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Introducing springs to a point/face as a boundary condition

Post by kevinarden » 26 Mar 2020, 12:59

It solved with Direct, but the displacements are 5.5E10 when they should be about 0.02.

mika
Posts: 131
Joined: 15 Sep 2009, 07:44

Re: Introducing springs to a point/face as a boundary condition

Post by mika » 26 Mar 2020, 23:29

Thanks for testing,

The variable of shell solver is by default specified as

Variable = Deflection[U:3 DNU:3]

Since the sif file now specifies Dirichlet BCs in terms of the component name U, one shouldn't define a different variable name in the solver section 1. That is, the line "Variable = Displacement" should now be removed from the first solver section. Thus, give also in the solver section 2 the value of the keyword Displacement Variable Name as "Displacement Variable Name = Deflection". Then the normal deflection given by the solver is about 0.01 for this mesh. I didn't check whether it changes by refining the mesh.

-- Mika

kevinarden
Posts: 484
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Introducing springs to a point/face as a boundary condition

Post by kevinarden » 26 Mar 2020, 23:40

Perfect, Thank You.

Kevin

MoTN
Posts: 8
Joined: 17 Mar 2020, 04:25
Antispam: Yes
Location: Canada

Re: Introducing springs to a point/face as a boundary condition

Post by MoTN » 27 Mar 2020, 03:57

Wow! I am so impressed by all the discussions and the progress you have made.
I am also very excited to run Mika's elasticity_with_springs and Kevin's "shell" examples.
Mika, could you please explain how to add the new function library "SpringAssembly" to ELMER installed on a Windows OS?

Thank you both,
Mo

kevinarden
Posts: 484
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Introducing springs to a point/face as a boundary condition

Post by kevinarden » 27 Mar 2020, 12:03

There was a windows install update on 3/26 that should include it

http://www.nic.funet.fi/pub/sci/physics ... dows-test/

mika
Posts: 131
Joined: 15 Sep 2009, 07:44

Re: Introducing springs to a point/face as a boundary condition

Post by mika » 27 Mar 2020, 20:00

kevinarden wrote:
26 Mar 2020, 12:18
Does the target boundaries also work so if there were many nodes each node would not have to be listed?
After a fresh update of the shell solver it should also be possible to add spring BCs by using the keyword "Target Boundaries", for example

Boundary Condition 4
Target Boundaries(2) = 5 6
Spring 4 = 1.E3
End

In this connection the boundary elements can be points or one-dimensional. This should even work in nonlinear analysis if the springs however are linear. So this works without the SpringAssembly utility, which is still needed if the places of the springs are listed using "Target Nodes".

-- Mika

MoTN
Posts: 8
Joined: 17 Mar 2020, 04:25
Antispam: Yes
Location: Canada

Re: Introducing springs to a point/face as a boundary condition

Post by MoTN » 29 Mar 2020, 01:37

kevinarden wrote:
27 Mar 2020, 12:03
There was a windows install update on 3/26 that should include it

http://www.nic.funet.fi/pub/sci/physics ... dows-test/
Thanks, Kevin.

MoTN
Posts: 8
Joined: 17 Mar 2020, 04:25
Antispam: Yes
Location: Canada

Re: Introducing springs to a point/face as a boundary condition

Post by MoTN » 29 Mar 2020, 01:51

mika wrote:
27 Mar 2020, 20:00
kevinarden wrote:
26 Mar 2020, 12:18
Does the target boundaries also work so if there were many nodes each node would not have to be listed?
After a fresh update of the shell solver it should also be possible to add spring BCs by using the keyword "Target Boundaries", for example

Boundary Condition 4
Target Boundaries(2) = 5 6
Spring 4 = 1.E3
End

In this connection the boundary elements can be points or one-dimensional. This should even work in nonlinear analysis if the springs however are linear. So this works without the SpringAssembly utility, which is still needed if the places of the springs are listed using "Target Nodes".

-- Mika
I was wondering if I can now model this problem using the new shell solver?

In this problem, I'd like to perform frequency analysis.
The plate is of an idealized light and stiff material. The stiffness comes from the translational and the torsional springs at point 2 which allow movement in the vertical (z) translation and pitch (y) rotation directions only (thus, the constrained degrees of freedom at point 2 are 1,2,4 and 6).

Boundary Condition 1
Target Nodes(1) = 2
Spring 3 = Real 1.0
Spring 5 = Real 1.0
End

Also, I need to establish mass properties using concentrated (point) masses at each of the grid points and a concentrated inertia at grid point 2. I was wondering if it is doable in the Shell solver?
aa.jpg
aa.jpg (43.39 KiB) Viewed 27 times

Many thanks

kevinarden
Posts: 484
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Introducing springs to a point/face as a boundary condition

Post by kevinarden » 29 Mar 2020, 13:10

This does everything except for the lumped masses. There are single node elements type 101, but I do not know how to assign them a mass value.
eigenvalue.zip
(2.14 KiB) Downloaded 2 times
Kevin

Post Reply