Hi Every one, I have a question for those who can help me. I read that simscale uses openfoam and code_aster under the hood and both of them are GPL license open source software. According to GPL license any software uses GPL software must also be GPL which means that simscale should provide the source. So does this mean that any one can get the source code of simscale?
1 Like
Hi @Mkhairy
Great question! It’s true that SimScale is using OpenFOAM and Code_Aster which are licensed under the GPL. It’s also not an easy one - software licenses are sometimes written in a form which makes them quite hard to read and understand. The question you’re asking is coming up from time to time, so I think it’s useful to share some more information.
The main point is that the event which triggers the GPL is distribution. Anyone who distributes software that contains GPL licensed code is forced to publish the source code alongside.
Allowing people to use GPL software (such as OpenFOAM) over the network is however not considered as distribution of the software and therefore doesn’t fall under the “viral” terms of the GPL.
(As a little sidebar: there is a different version of the GPL, called the AGPL, which also applies when an application can be accessed via the network. This license does not apply to OpenFOAM and Code_Aster, though.)
The difference between distribution of software (which is where the GPL does apply) and allowing people to use software via the network (where the GPL does not apply) is the reason why SimScale is not required to publish the source code.
I hope that I was able to answer your question. Feel free to ask if something isn’t clear.
Johannes
6 Likes
Hello @jprobst , my name is Mateus Alves Farah and I am a student of naval engineering from Brazil, I would like to ask something really important about this issue.
I am working on a project that I have to use OpenFOAM to validate a case of prediction of resistance in a boat, for my term paper which have to be delivered in the next month. I founded SIMSCALE in march and started to try to generate my mesh here, and after some time I could get it , but when I tried to do that in OpenFOAM (I am using version 6 and 7), I didn’t have succes.
So I sent a e-mail to the support of SIMSCALE and found that the version for mesh generate is 1606+ and for simulations is the 2.3.1, both with some improvements/differences from the source/base code of OF. So my question is : is there some way that I could run the mesh in some OF version with similar results founded here in SIMSCALE? If not, could I export the mesh results from here to OF and make a reference in my term paper saying that I used SIMSCALE because the results here were better but that it’s based on OF?
I am asking it because the mean goal of my term paper is to use a opensource code that allow the students of my university to use that, confirming from my validation case that this tool (OF in this case) is as good as the private programs of CFD that exist. This would be a good thing because the high price of these devices (even some students license) are impracticable sometimes.
Hope to see your reply.
Regards!
Hi Mateus
first of all, I’m happy to hear that you’ve chosen SimScale for your work 
Let me see if I got this right: the idea is to run some simulations with SimScale and then reproduce the results with a standard OpenFOAM?
First of all, it is possible to download the simulation results in OpenFOAM format and this will give you access to the mesh in OpenFOAM format which you can also use offline. The result will also contain all input files and boundary conditions. Regarding the version, I think the meshes should be compatible between the versions. So it should be possible to run a SimScale FOAM mesh in all other versions, too.
Now there are some extensions and modifications which we’ve made. For example there are some special boundary conditions, and some other utilities (functionObjects, automatic relaxation etc.) which aren’t publicly accessible. I would assume though, that it is possible to arrive at practically the same results with an offline OpenFOAM. There could be some small differences in the convergence behavior, and the input files will need some small (but easy) adaptations to run correctly. But I don’t see a big problem there.
I would suggest you just give it a try and report back if it worked, or not.
Johannes
Sorry if I didn’t make my self clear, but the real problem now is the mesh. Here is my mesh in SIMSCALE https://www.simscale.com/workbench/?pid=4733360224625830605.
I have been working parallel with OF version 6 and 7 together with SIMSCALE, and the results with the same geometry and parameters used in Hex PArametric Dominant generator (SIMSCALE) and SnappyHexMesh (OpenFOAM) are the exactly the same.
So the question is: can I do somethings in OF to achieve the same results ? Remembring that it seems that version 1606+ of OF was used here in SIMSCALES, with some modifications (as I saaw in some threads here). If it wouldn’t be possible (something that I really expect to be wrong because of the time spent in that), is there any problem to reference the SIMSCALE platafform in my term paper as a opensource tool (I am using the free access for example) so that the students can make their simulation in prediction the resistance of a boat and so on.
It’s because the goal of the work is to show another way to evaluate the resistance dispite the methods available for the first steps of a naval project, so it would be in a preliminar step, for example (refine of the project until achieve the chosen ship).
My concern is to know exactly what I could do because it’s important to define right now and I don’t have much time to start all over again the process of mesh generation.
Thanks for the attention!
Hello @jprobst there any update about I told in the last? Sorry to send another message , btu it’s just way I am worried about that subject.
Regards!
Hi @mateusfarah
let me try to understand the actual concern here. Are you asking if you may quote SimScale in the thesis? You certainly can reference it, however keep in mind that SimScale is not open source. It is free to be used for academic purposes (which seems what you’re doing) and so it’s fine. Please keep in mind that “open source” and “free to use” is not the same thing.
I’m actually happy to see someone complete his paper with SimScale. Once the work is done I’d be happy to read it 
But we cannot provide the source code to our modifications of OpenFOAM (in case this is what you’re asking).
I am not sure if I got your question right. Please let me know.
Johannes
1 Like
Thank you for the reply, and the work will have SIMSCALE and OF uses, because I came to the forum because I loved that everybody and due to the cloud system with 16 processors available, which helped me a lot (my notebook just have 4 cores), so if I could use it I would be glad to put in the reference.
The problem is that in the first time, in a conversation with my professor, the term paper were done using JUST OF, but I had some problems doing the mesh there and some memory problem due to my limitated computer. Since I came to SIMSCALE I resolved these issued, but this days I tried to do the SAME mesh in OF and it didn’t work.
So, following your answer, could you please tell me something that could help?
-
What versions of OF is used in the Parametric Hex dominant generator? Which is similar to the SnappyHexMesh in OF, which I am using.
-
What is the version of OF used in the simulation process? I am using version 6 and 7 for over a year but I would like to represent the results from here as similar as I can (because I know that were made modification by you here in SIMSCAL, what is great).
-
Is there any possibility that someone in the fĂłrum that knows OF very well to help me doing that: get the mesh as similar as possible from here in SIMSCALE to OF, but without exporting, I want to do there in OF because of the goal of the term paper as I told in the beginning of this message.
So, just to compile the informations: I am using SIMSCALE to hurry up somethings and get help with the users ( and try to help others too), and I will reference in my term paper certainly, but I really have to use OF because there will be others things to do in the work, and for that I will need an open source tool.
And I have invested SO MUCH time doing the mesh (because of the geometry particular, sharp edges and so on), that had to be with parametric (because of the use in SHM in OF), and I really don’t have much time to do all the things again (because of the dead line of the term paper itself), that’s way I am asking for help.
Thank you for the attention, I really appreciate that.
Regards!
So we are using snappy from OpenFOAM 16.06+ and for the simulations it is OpenFOAM 2.3.1.
Sharp edges is something we have improved on our side for snappy so I am not sure if you’ll get the same results with the standard snappy. If the idea is to reproduce SimScale results with an offline OpenFOAM, your mileage may vary. It all depends on how much influence our modifications have on this particular model.
Have you tried downloading our SimScale simulation result and running the data (i.e. the STL file, the snappyHexMeshDict etc.) through snappy again? One thing we’re doing a little different is that we run the layer creation separately. First, we run the “normal” meshing without layers and then we do the layers separately on top. There’s a setting in the snappy dict which you need to modify to run both at the same time.
Sorry for the late reply.
Thank you for the answer about the versions, and I wanted to tell you that I tried to export the mesh and it worked (I had some help from two users from SIMSCALE).
I noticed that the two steps are done separately, and it’s one thing that I do two, just put first the castellated and snap step as"true" while addlayers is “false”, and then turn off the two first steps and turn on addlayers. But the whole problem is that the “addlayers” step is done here in a different way from that exactly in OF 1606+, so I can’t get the same (as you told), and if someone that knows about OF could help me it would be awesome.
Thank you and hope to see the reply!
Hi @mateusfarah, that’s correct. This discussion is getting technical and way off track (this thread is about licensing). I’ll PM you.
Sorry for that, do you prefer that I transfer the messages from here to the thread about layer generation?