[Tobermory]

I have a scenario where I want to bathe a target piece in thermal radiation from a distant source, without having to explicitly model the distant source.

I can do this with the fvDOM model by setting a fixedValue boundary condition for G and one of the ray directions on one side the domain, having orientated the target to be normal to the ray direction. This works fine (see attached snapshot for a simple plate), but is a bit messy.

Earlier, I had been looking for a way to simply add a source term to the chosen intensityRay equation ... but I couldn't for the life of me work out how to do it! Does anyone know how I can add a source term to say ILambda_15_0?

I am working in OF v13. Any thoughts, comments or observations would be welcomed.

[dlahaye]

Apologies for not being of much help.

I wonder, however,

1/ why adding a source to a single ray? What motivates this approach? What are you trying to do (apologies for being lost in translation)

2/ to what extend does it help you to drill through the code and figure out how for instance the emission term is added to all rays together? Possibly I can assist here (likely not as you are at it for a longer time)

3/ to what extend does it help you to home-brew an implementation from scratch (using tools like FENICS or Gridap) and port it to OpenFoam is a second stage only?

Good luck. D.

[Tobermory]

Thanks for the reply Domenico. I am just trying to set up a case where a target piece is irradiated at a given level (e.g. 100kW/m2) from a distant hot source ... without having to mesh the distant heat source itself (saves a lot of gridcells!). In this scenario, the incident radiation is approximately unidirectional, and so I only need to enter the radiation in a single ray.

I can do it perfectly with boundary conditions - as per my figure in my previous post, and the results for a simple shape validate extremely well with an analytical mode.

However, it struck me whilst setting up the boundary conditions that it should be possible to do the same with a source term in the ILambda equations ... but I just couldn't work out how to add a generic source to an equation in OF13 ... without cloning the source code and writing my own version.

If you're interested, the intensity ray equation is formulated and solved in radiativeIntensityRay::correct():

         fvScalarMatrix IiEq
         (
             fvm::div(Ji, ILambda_[lambdaI], "div(Ji,Ii_h)")
           + fvm::Sp(k*omega_, ILambda_[lambdaI])
         ==
             1.0/constant::mathematical::pi*omega_
            *(
                 // Remove aDisp from k
                 (k - absorptionEmission_.aDisp(lambdaI))
                *blackBody_.bLambda(lambdaI)
  
               + absorptionEmission_.E(lambdaI)/4
             )
         );

... this is the equation that I'd like to add a source term to, using something in fvOptions for example.

Don't waste too much time on this Domenico - the original post was just in case someone said "that's easy - you just do ... blah blah blah". As I said, I have a perfectly good work-around.

[dlahaye]

The code that you share clearly has no fvOptions() construct that allows to add terms.

Does a coded scalar transport equation for myIiq to be solved ?

[Tobermory]

Yes, you're right - it was staring me in the face all the time. Digging deeper, I realise that in OF13 the source terms are defined in constant/fvModels ... and looking at
incompressibleFluid::momentumPredictor() or kEpsilon.C, I can see that there should indeed be a term like:

+ fvModels.source(blah blah)

in the fvm equation, to allow for source term inclusion. There isn't, so I can't. Anyway, specifying the source at the boundary is probably more appropriate anyway. Thanks again.