Endeavour to model basic Hypoplastic Sand using mfront: A newbie question

[s1966b]

Dear all,
Greetings from India...
This is the first time I am writing here. I am a new learner of mfront and my query might be naive for many of you. However, I am stuck up with the following:
I have tried to develop an mfront file for hypoplastic sand for ultimately using in my code-aster simulation. I managed to develop a small file, using Implicit DSL, which I had tried to compile using mfront-4.1.0 and 4.2.0 available in Salome-Meca-2024 container. However, after using different ways to handle a Stensor data D and performing a double dot product with another Stensor4 quantity L, I am facing primarily some type error. I am attaching my mfront script for your kind inspection.

My error message is about the declaration of the Stensor D which I declared within the Integrator block as
const Stensor D = de/dt;
The error message I am getting is
hypoplasticSandImplicit04.mfront: In member function ‘bool tfel::material::HypoplasticSand<tfel::material::ModellingHypothesis::TRIDIMENSIONAL, NumericType, false>::computeFdF(bool)’: hypoplasticSandImplicit04.mfront:105:1: error: expected primary-expression before ‘const’

Would someone of you kindly guide me?

Thanks and Best regards

Sukumar
hypoplasticSandImplicit04.txt

[thelfer]

Hi @s1966b,
You are facing the worst MFront's error message as I (dis)-like to call it. This is very annoying.
Basically, D is an MFront's variable and you can't just declare a new variable with the same name.
So,

const Stensor e_rate = de/dt;
```
shall work

[s1966b]

Dear @thelfer
Thank you so much for your prompt reply and correcting me not to use inbuilt variable names.

However, after changing the statement to const Stensor e_rate = de/dt; I am getting the same error message.

For your information, I have also installed the latest version from your repository and same error message persists in this case also.
The error message is
In file included from asterHypoplasticSand.cxx:11: hypoplasticSandImplicit04.mfront: In member function ‘bool tfel::material::HypoplasticSand<tfel::material::ModellingHypothesis::TRIDIMENSIONAL, NumericType, false>::computeFdF(bool)’: hypoplasticSandImplicit04.mfront:105:1: error: expected primary-expression before ‘const’ compilation terminated due to -Wfatal-errors. make: *** [Makefile.mfront:37: asterHypoplasticSand.o] Error 1 callMake: can't build target 'all' Libraries building went wrong. This may be due to an error in your implementation. In rarer cases, this may be due to a incompatibility with a previous run of MFront and you may want to remove to the 'include' and 'src' directories. You may also invoke MFront with the --verbose=debugflag for more (hopefully helpful) information. Command was: make -C src -f Makefile.mfront all -s
I think, there is some more fundamental error I have committed elsewhere. It would be great if you could kindly examine my file and advise me.

Thanks again and Best regards.
Sukumar

[thelfer]

@s1966b Please find attach an almost compiling version (contract does not exist). However, to be honest, I don't get your implementation. I left some comments.

hypoplasticSandImplicit04.zip

[s1966b]

Dear @thelfer

I am really grateful for your time to read my script, for your guidance in form of comments and your kind gesture to modify my script (if I can call .mfront file as a script)
Now, the main confusion and resulting conceptual errors I have made is arising out of my novice idea about variable names and due to my ignorance about reserved keywords or variable names in mfront.
In Geotechnical Engg, void ratio of soil is usually denoted by 'e', whereas here 'e' denotes strain tensor. And I mismatched both these variable nomenclature systems. And that has created all sorts of errors in compilation, I tend to believe now !!
I am extremely sorry for this silly misunderstanding, not to be expected from me!! I will have to correct my file completely.

By the way, with your notation of 'dedt' I am getting the error that there is no match for operator | for the operand types as follows:
hypoplasticSandImplicit04.mfront:120:9: error: no match for ‘operator|’ (operand types are ‘const Stensor4’ {aka ‘const tfel::math::st2tost2<3, double>’} and ‘tfel::math::BinaryOperationHandler<const tfel::math::stensor<3, double>&, tfel::math::Expr<tfel::math::stensor<3, double>, tfel::math::ObjectScalarOperation<tfel::math::Expr<tfel::math::stensor<3, double>, tfel::math::BinaryOperation<tfel::math::stensor<3, double>&, tfel::math::stensor<3, double>&, tfel::math::OpMinus> >&&, double&, tfel::math::OpDiv> >&&, tfel::math::OpMinus>’ {aka ‘tfel::math::Expr<tfel::math::stensor<3, double>, tfel::math::BinaryOperation<const tfel::math::stensor<3, double>&, tfel::math::Expr<tfel::math::stensor<3, double>, tfel::math::ObjectScalarOperation<tfel::math::Expr<tfel::math::stensor<3, double>, tfel::math::BinaryOperation<tfel::math::stensor<3, double>&, tfel::math::stensor<3, double>&, tfel::math::OpMinus> >&&, double&, tfel::math::OpDiv> >&&, tfel::math::OpMinus> >’})
I am sure, there are many more errors to be corrected, which I will try to do.

Now, one important thing to learn is where do I get the list of all inbuilt/reserved variable names for the various DSL types? I saw the lists for keywords in MFront Gallery page, which I need to study thoroughly. Do they also include the inbuilt variable names?

Another important issue is how to keep void ratio as one of the state variables. Hypoplastic model essentially is driven by the current void ratio and its "distance" from the critical void ratio. I think implementation of Cam-Clay model will give me some idea about that.

I will come back to you once I make a little progress to ask you more meaningful questions.

Thank you sooo very much for your kind help and your kind gesture to prepare a new file for me. Also kindly guide me to learn about the list of reserved variable names.

With warm regards

Sukumar

[thelfer]

You must use * to apply a fourth order tensor to a second order tensor.

[s1966b]

Dear @thelfer

Thank you once again for your timely guidance..
The file has compiled now !!

I will need to test if it represents anything meaningful at this stage..

Best regards..
Sukumar