Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Numerical methods and mathematical models of Elmer
Adev
Posts: 30
Joined: 02 Oct 2019, 22:18
Antispam: Yes
Location: Florida, USA

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by Adev »

I have read and worked through several of the related posts on this forum, and in particular, those by Z Losonc, CWeing, Raback and MZenker.
The first question I have is how to scale my model into metric units. It seems that there is a function "Scale" described in the ELMER Solver Manual (p63).
Q1: Do I merely include a line in the sif file stating: Coordinate Scaling = Real 0.0254 ?
The larger question is on how to set up the analysis using a sweep of SPLs (in dB) at the Inlet and how to extract the SPLs at the Outlet, starting with the Pipe-only model. From comments by raback, there is a way to use Savescalars to solve my cases without need to use Paraview. However, I do not know even where to start doing that, despite having read the ELMER Models Manual. So let me ask my other questions step by step:
Q2: Which solvers do I need to Append" using the File - Definitions menu in the GUI, and would I need more than 1 instance of any of those?
Adev
Posts: 30
Joined: 02 Oct 2019, 22:18
Antispam: Yes
Location: Florida, USA

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by Adev »

While waiting for responses, I blindly followed/attempted to mimic the sif file presented by CWeng in viewtopic.php?f=4&t=5001#p17893
The result was an error about rho0 not defined or something like that … I am vague because when I reloaded the project file, absolutely all the work of the past 4 hours was gone, wasted … no materials, equations, boundary conditions etc …. nothing! Grrrrr!!!!! What is wrong with this software!: I know, the answer will come back that its the user, not the software., despite the fact that I regularly saved the project after every significant step, like entering an equation, including one more time before running it. I know the software can do this and that I am doing/not doing something … please help.
Adev
Posts: 30
Joined: 02 Oct 2019, 22:18
Antispam: Yes
Location: Florida, USA

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by Adev »

Here is my sif file that failed
FNP_pipe_only case.sif
(4.61 KiB) Downloaded 302 times
kevinarden
Posts: 2316
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by kevinarden »

I don't think sound speed can equal c0 nor Density rho0

Sound speed = c0
Heat Capacity = 1005.0
Density = rho0

I think it is

Sound speed = $ c0
Heat Capacity = 1005.0
Density = $ rho0
Adev
Posts: 30
Joined: 02 Oct 2019, 22:18
Antispam: Yes
Location: Florida, USA

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by Adev »

Hi Kevin, I really appreciate your feedback. It will take me a while to recover and regenerate the problem using the GUI (which I do want to do)., but I will attempt your comments. In the meantime, here is the sif file from one of the resident experts, CWeng posted on the forum topic referenced above. Which is what I copied and attempted to replicate. I use his notation.
For the moment, I am enjoying a whiskey to sooth my frustrations following the loss of several hours of work (that a seasoned pro would probably do in 10 minutes).
areaExpansion - CWeng.sif
(4.21 KiB) Downloaded 253 times
kevinarden
Posts: 2316
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by kevinarden »

yes in Cweng's file

Sound speed = $c0
Heat Capacity = 1005.0
Density = $rho0

in your file

Sound speed = c0
Heat Capacity = 1005.0
Density = rho0

You do not have the $
Adev
Posts: 30
Joined: 02 Oct 2019, 22:18
Antispam: Yes
Location: Florida, USA

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by Adev »

Aha, now I see it, thank you very much Kevin!
I'll have to re-do it from scratch, use the $ sign this time, and post the outcome.
Adev
Posts: 30
Joined: 02 Oct 2019, 22:18
Antispam: Yes
Location: Florida, USA

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by Adev »

Referring to the CWeng sif file, lines 120 to 125 are not available in the GUI … any explanation? From the ELMER Manual, under "Model -Setup" there should be choices available for the coordinate system that include Cartesian 1D, Cartesian 2D, Cartesian 3D, Polar 2D, Polar 3D, Cylindric, Cylindric Symmetric and Axi Symmetric (copied and pasted from the manual). However, the GUI has only 3 choices available: Cartesian; Axi symmetric; and Cylindric symmetric. A note in line 122 of the CWeng sif file implies that variables and Operators 8 to 8 are only available for 3D models. How come my model is 3D and I cannot explicitly set the Coordinate System to "Cartesian 3D", only to "Cartesian"? Does this have anything to do with the missing Variables and Operators 6 to 8?
kevinarden
Posts: 2316
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by kevinarden »

Elmer has many features, solvers, options, not available in the GUI. The GUI is a small subset of the full ElmerSolver code. ElmerGUI does, on every input panel, allow 'free text input' . Free text input allows you to enter options into that section that the GUI is not yet supporting. Also you can edit the sif file before running ElmerSolver. In general, ElmerGUI will only get you so far. Eventually you will have to understand and edit the sif file to fully appreciate the ElmerSolver.

As far as cartesion 1d, 2d, 3d it only impacts the input data of the nodes. in the Cweng sif file the analysis is set to 2D in the solver equation when Variable is set to dofs 2.

cartesion sets the input and output to a cartesian system, whether it is 1d, 2d, or 3d depends on the geometry, nodes coordinates, elements used, and the variable set in the solver section.

Solver 1
Equation = Helmholtz Equation
Procedure = "HelmholtzSolve" "HelmholtzSolver"
Variable = -dofs 2 Pressure Wave
raback
Site Admin
Posts: 4832
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Acoustics – Basic Postprocessing With ElmerSolver & Paraview

Post by raback »

Hi

Yes, ElmerGUI supports a subset of ElmerSolver.

Concerning the coordinate system. If given "cartesian" the dimension of the mesh is used to define whether 2D or 3D. The other coordinate systems depend somewhat on the module. Most support axially/cylindically symmetric. Unfortunately Polar is not often supported.

-Peter
Post Reply