Deciding LENGTH variable in NekRS .udf file

[nancysemwal]

@aprilnovak

In the pincell_multiphysics tutorial, the height of the pincell has been taken as 50 cms.

However, in the fluid.udf file, since we want to set temperature to a linear variation from 0 at the inlet to 1 at the outlet, we do

#define LENGTH 42.35158086795113

void UDF_Setup(nrs_t *nrs)
{
  // Other initial settings such as velocity, pressure

    nrs->cds->S[n + 0 * nrs->cds->fieldOffset[0]] = mesh->z[n] / LENGTH; // temperature
  }
}

I want to understand why the constant LENGTH has been assigned the value 42.35158086795113 and not 50, which is the actual length of the pincell?

Are these two even related? Has some kind of normalization been performed to obtain this value or we're talking about the active fuel length and hence the value is slightly less than 50 cms?

I agree this question is not directly related to Cardinal and more of a NekRS simulation setup detail, but I think the team can help me out here, or suggest better resources to ask the question.

[aprilnovak]

Hi @nancysemwal, great question! NekRS is solving this problem in non-dimensional form. This means that the actual units of the mesh used by NekRS (the .re2 file) are dimensionless (whereas the MOOSE heat conduction problem is set up in SI units).

You can find documentation about the non-dimensional solution approach here and in more detail how the non-dimensionalization happens.

[nancysemwal]

Thank you for your input April. I could understand and derive the value of LENGTH, albeit with some more help from some generative AIs 😉