Latest FEATool Version

Is FEATool 11.1 build 19.07.210 the latest available for trial/eval?
-Randal

randress wrote

Is FEATool 11.1 build 19.07.210 the latest available for trial/eval?
-Randal

The latest official FEATool Multiphysics software builds are always available from:

https://github.com/precise-simulation/featool-multiphysics

which is the same as using the downloads link from the featool.com website, Sourceforge, or the Mathworks file exchange. There is no difference between the trial and licensed software versions. When you purchase a full license you will be issued a license token which is used to unlock the unavailable features.

If you are interested you can also try the v1.12 alpha build:

http://forum.featool.com/FEATool-Multiphysics-v1-12-Alpha-td132.html

Precise Simulation wrote

The latest official FEATool Multiphysics software builds are always available from:

https://github.com/precise-simulation/featool-multiphysics

Good to know... but I'm afraid I am going to need a bit more direction to install it :-( .  I could not find a web or windows installer.  Or do I simply replace directories/files that are already installed for 1.11?

-Randal

The current releases and assets are bundled and available for download by clicking on "releases".

Installation depends on wether you use the stand-alone version or toolbox under Matlab and proceeds as per the installation instructions.

randress wrote

Precise Simulation wrote

The latest official FEATool Multiphysics software builds are always available from:

https://github.com/precise-simulation/featool-multiphysics

Good to know... but I'm afraid I am going to need a bit more direction to install it :-( .  I could not find a web or windows installer.  Or do I simply replace directories/files that are already installed for 1.11?

-Randal

I see 2 releases.  1.11.1  for the Matlab Toolbox and 1.11 for standalone. Are there instructions for installing 1.11.1 (or the 1.12beta) for a standalone version?
Thanks,
Randal

Due to both complete lack of interest (sales) and support from the Mathworks (who makes and sells Matlab) the stand alone version most likely will not be updated, but possibly eventually replaced with a version based on a different runtime.

randress wrote

I see 2 releases.  1.11.1  for the Matlab Toolbox and 1.11 for standalone. Are there instructions for installing 1.11.1 (or the 1.12beta) for a standalone version?
Thanks,
Randal

Precise Simulation wrote

Due to both complete lack of interest (sales) and support from the Mathworks (who makes and sells Matlab) the stand alone version most likely will not be updated, but possibly eventually replaced with a Python/Web/Julia/Octave version.

That is indeed sad to hear :-(.   As I am finding that FEATool seems to be able to support my use case (with probable challenges with regard to efficiency for larger real-world problems), I am disappointed to hear of the uncertainty of future standalone builds.  

I have a few questions about how I may still make this work for me.  

I see one of two paths: 1) Purchase Matlab and some yet unknown set of add-on Toolboxes so that I can make use of FEATool.     Can you tell me which of the add-ons I would need to run FEATool with Matlab efficiently? Here is a list from their site:  https://www.mathworks.com/store/link/products/home/ML

or

2)  I am a retired S/W engineer.  When the Matlab FEATool Toolbox is updated by Precise Simulation, I build the standalone myself from available source and libraries available at gethub. What will I need to setup a build environment so that I may build the standalone version? I assume only minor code/script tweaks would be necessary as you have an established build process that I could just execute.  Would I need to purchase anything from Mathworks? (if so, I'd likely be better off just to go the Matlab route.)

Thank-you for your continued help in evaluating FEATool and my personal acquisition options.
-Randal

randress wrote

I am disappointed to hear of the uncertainty of future standalone builds.  

The plan for the future is not to discontinue standalone builds, but to eventually replace the dependency on the Mathworks products (Matlab). However, this is not a quick fix and will a lot of work and time.

randress wrote

Can you tell me which of the add-ons I would need to run FEATool with Matlab efficiently?

FEATool does not require or make use of any Matlab toolboxes or add-ons beyond the base Matlab runtime.

randress wrote

When the Matlab FEATool Toolbox is updated by Precise Simulation, I build the standalone myself from available source and libraries available at gethub. What will I need to setup a build environment so that I may build the standalone version? I assume only minor code/script tweaks would be necessary as you have an established build process that I could just execute.  Would I need to purchase anything from Mathworks? (if so, I'd likely be better off just to go the Matlab route.)

If you already have Matlab (and the Matlab compiler required to make stand-alone builds) there would really be no reason or advantage to making such a build (it would only be useful if you want to distribute software to users that don't have Matlab). Also the raw sources necessary to make such a build are for obvious reasons not provided for free or publicly (nor redistribution rights of FEATool).

Precise Simulation wrote

The plan for the future is not to discontinue standalone builds, but to eventually replace the dependency on the Mathworks products (Matlab). However, this is not a quick fix and will a lot of work and time.

Ah.... thanks for the clarification.

Precise Simulation wrote

FEATool does not require or make use of any Matlab toolboxes or add-ons beyond the base Matlab runtime.

Good!

Precise Simulation wrote

If you already have Matlab (and the Matlab compiler required to make stand-alone builds) there would really be no reason or advantage to making such a build

Yeah, I was not thinking very clearly on that one :-)

Precise Simulation wrote

 (it would only be useful if you want to distribute software to users that don't have Matlab). Also the raw sources necessary to make such a build are for obvious reasons not provided for free or publicly (nor redistribution rights of FEATool).

Certainly not interested redistribution  .... so it looks like, for the near future, the only option is to get Matlab ($150) and FeaTool ($245). This certainly beats the many times that cost for ElecNet or COMSOL (two other products that I have recently tried).   Based on my limited experience with them and FEATool, it does look like they perform much more efficiently on my current hardware (i5 with 8Gb RAM).  But I have access to a beefier machine that should help.

Again, thanks for helping me figure out a path to success with FEATool.
-Randal

randall wrote

.... so it looks like, for the near future, the only option is to get Matlab ($150) and FeaTool ($245). This certainly beats the many times that cost for ElecNet or COMSOL (two other products that I have recently tried).   Based on my limited experience with them and FEATool, it does look like they perform much more efficiently on my current hardware (i5 with 8Gb RAM).  But I have access to a beefier machine that should help.

Depending on your use case, static conductive media DC simulations, you might be ok with the current stand-alone v1.11 distribution. If you purchase a license it will be valid for all distributions, so you can decide later if you think you need Matlab for using v1.12 or later.

Precise Simulation wrote

Depenging on your use case, static conductive media DC simulations, you might be ok with the current stand-alone v1.11 distribution. If you purchase a license it will be valid for all distributions, so you can decide later if you think you need Matlab for using v1.12 or later.

True. However, as I understand from my limited reading at featool.com/doc and from my initial use of the standalone FEATool, that being able to export/import both grid and results to/from Matlab will likely be necessary. An example is your answer to my question under the topic "Viewing Mesh/Grid Internal to Subdomains". I would need Matlab so that I could use the workaround: "plotgrid( grid, 'selcells', selcells(grid,'y>0.5') )".
-Randal