[fixed] mistakes in the Manual

Clearly defined bug reports and their fixes
uwestoehr
Posts: 38
Joined: 05 Aug 2022, 14:42
Antispam: Yes

[fixed] mistakes in the Manual

Post by uwestoehr »

Dear Elmer team, many thanks for your new manuals (the ones from August 2nd)! (I like the new structure of the models manual.)

Here are some mistakes I found:

1. ElmerModelsManual.pdf, page 15:
"Heat Equation String
If set to True, solve the heat equation."

This does not work. When I input this to the case.sif:

Equation 1
Heat Equation = String "False"
Active Solvers(4) = Integer 1 2 3 4
End

I get this error:
"ERROR:: CheckKeyword: Keyword: [heat equation] in section: [equation 1] is given wrong type: [string], should be of type: [logical]"

It works, if I write therefore

Equation 1
Heat Equation = Logical False
Active Solvers(4) = Integer 1 2 3 4
End

2. ElmerSolverManual.pdf, page 34:
"below unity is sometimes required to achive convergence of the nonlinear system."
___________________________________ ^

3. all manuals the keyword "Steady State" is inconsistent, sometimes it is "Steady State", sometimes "Steady state". For the case.sif the difference doesn't matter but for me it was confusing.

At least 1. should be fixed.

thanks and regards
Uwe
Last edited by uwestoehr on 13 Feb 2023, 13:28, edited 2 times in total.
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: mistakes in the Manual

Post by raback »

Hi

Just fixed 1). Thanx! The "Heat Solver = True" and "Active Solvers(1) = 1" are btw interchangeable.

The sif file of Elmer is not case sensitive. The only thing case sensitive may be some filenames. To keep the case one can give, for example:

Code: Select all

Output File = "BigCaseExample.vtu"  ! known keyword of type File
Filename = File "MyFile.dat"
The type "File" will keep the case while type "String" will not. Otherwise it would be difficult to refer to files names with large caps.

-Peter
uwestoehr
Posts: 38
Joined: 05 Aug 2022, 14:42
Antispam: Yes

Re: mistakes in the Manual

Post by uwestoehr »

raback wrote: 08 Aug 2022, 12:15 Just fixed 1). Thanx! The "Heat Solver = True" and "Active Solvers(1) = 1" are btw interchangeable.
Thanks!

If they are interchangeable, do they overwrite each other? I mean does this:

Equation 1
Active Solvers(4) = Integer 1 2 3 4
Heat Equation = Logical "False"
End

and Solver 1 is Heat means the heat solver in effect not executed?

I tired this but then I get this error:

Code: Select all

LoadInputFile: Reading base load of sif file
ERROR:: SectionContents: Problem reading logical keyword: heat equation: False
when I try this:

Equation 1
Active Solvers(4) = Integer 1 2 3 4
Heat Solver = Logical "False"
End

I get this error:

Code: Select all

LoadInputFile: Reading base load of sif file
CheckKeyword:  Unlisted keyword: [heat solver] in section: [equation 1]
ERROR:: SectionContents: Problem reading logical keyword: heat solver: False
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: mistakes in the Manual

Post by Rich_B »

Hello,

Item 2 has been fixed. Several typos have been corrected in the Solver manual and the Models manual. The latest pdf is dated 17 October 2022 and is located here:

https://www.nic.funet.fi/pub/sci/physics/elmer/doc/

Rich.
uwestoehr
Posts: 38
Joined: 05 Aug 2022, 14:42
Antispam: Yes

Re: mistakes in the Manual

Post by uwestoehr »

Many thanks for the fixes!

Meanwhile, I found a new one:

Page 141, section "22. Computation of Magnetic Fields in 2D"
There is written that

Code: Select all

Variable String Potential[Potential Re:1 Potential Im:1]
However, Elmer creates as default this:

Code: Select all

Variable String Potential[Potential Re:1 Potential:1]
There the sentence

Code: Select all

The associated number of degrees of freedom should always be two.
Is not clear. How is this set? There is no two in the case.sif created by Elmer.
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: mistakes in the Manual

Post by raback »

Hi

The initial way of defining fields in Elmer was like:

Code: Select all

Variable = Potential  
Variable Dofs = 2
or later

Code: Select all

Variable = -dofs 2 Potential
These will give you vector field with components "Potential 1" and "Potential 2". For convenience a syntax where you give the component names and their sizes was defined,

Code: Select all

Variable = Potential[Potential Re:1 Potential Im:1]
The size of this is 2 because 1+1=2 and it has component names "Potential Re" and "Potential Im" fitting for a harmonic simulation.

I cannot find the one missing "Im" anywhere.

-Peter
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: mistakes in the Manual

Post by Rich_B »

Hello,

If one starts with the tutorial 'InductionHeatingGUI', by creating a copy of the files folder, then delete all files except for the msh file.

Follow the tutorial, create a new project, select the msh file for geometry, and add magnetodynamics2d.xml.

Add a new equation, selecting the tab for 'MgDyn2DHarmonic', click on Active, and select all bodies, then 'Edit solver settings', 'Solver specific options', and the Variable is shown as 'Potential[Potential Re:1 Potential Im:1]'. Everything is fine so far.

Save Project, click on Sif, Generate and then edit the sif file. Here is what has been added by the sif generator:
Variable = Potential[Potential Re:1 Potential:1]
I think we found the missing 'Im'. I'll see about creating a pull request to fix the sif generator.

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

Re: mistakes in the Manual

Post by Rich_B »

one more interesting point, the same thing happens in ElmerGUI when using vectorhelmholtz.xml.
Procedure = "VectorHelmholtz" "VectorHelmholtzSolver"
Variable = E[E re:1 E:1]
Once again, the 'im' is missing.

After looking at the sif generator code, I'm not confident about fixing the issue. Anyone feels like taking a look, that would be great.

Rich.
uwestoehr
Posts: 38
Joined: 05 Aug 2022, 14:42
Antispam: Yes

Re: mistakes in the Manual

Post by uwestoehr »

Rich_B wrote: 04 Feb 2023, 20:27 Save Project, click on Sif, Generate and then edit the sif file. Here is what has been added by the sif generator:
Variable = Potential[Potential Re:1 Potential:1]
Yes, this is the bug I see.

-------------

Concerning the docs, I miss sometimes consistence to what ElmerGui does. For example sometimes there is no output of the Variable at all while the models manual state the variable should be given. (name can be chosen freely).

This is for example the case for 2D and 3D magnetodynamics. If I run Elmer Gui, I don't get a variable setting at all. The solver run succeeds, but I see a warning that no variable name was given and thus

Code: Select all

av[av re:1 av im:1]
is used.
If I specify

Code: Select all

Variable = String "av[av re:1 av im:1]"
for the WhitneyAVHarmonicSolver, the warning goes away.

So either there should not be a warning of the variable should be output by Elmer Gui as stated in the manual.
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: mistakes in the Manual

Post by Rich_B »

Hello Uwe,

Regarding the bug of the missing 'Im', PR #382 has been accepted and the bug is now fixed. Wait for tomorrow's binary builds or build from the latest source.

Please take a look and confirm that this works for you.

Thanks, Rich.
Post Reply