Nonzero latent heat melt without melting

Hello everyone,

I am trying to use a Compositing material model to include batch melting and latent heat of melting into a model with viscoplastic rheology.
It is a simple testing case that should represent decompression melting beneath and extending lithospheric plate. The viscosity looks fine and decompression melting occurs as well, but I am confused by the latent heat. In the visualization output, the latent heat is non-zero even if there is zero melt fraction. What is the source of this heat?

Thanks in advance for your help!
Petra

Here is the PRM file:

set Dimension                              = 2
set Use years in output instead of seconds = true
set Output directory                       = output-extension-freesurf-latentheat
set End time                               = 2e6 #years

set Nonlinear solver scheme                = single Advection, iterated defect correction Stokes
set Nonlinear solver tolerance             = 1e-5
set Max nonlinear iterations               = 100

subsection Geometry model
  set Model name = box
  subsection Box
    set X repetitions                      = 2
    set Y repetitions                      = 1
    set X extent                           = 100e3
    set Y extent                           = 50e3
  end
end

subsection Mesh deformation
  set Mesh deformation boundary indicators  = top: free surface
  subsection Free surface
    set Surface velocity projection = vertical
  end
  set Additional tangential mesh velocity boundary indicators = left, right
end

set Pressure normalization = no

subsection Gravity model
  set Model name = vertical
  subsection Vertical
    set Magnitude                           = 10.0
  end
end

subsection Initial temperature model
  set Model name = function
  subsection Function
    set Variable names      = x,y,t
    set Function constants  = Ttop=273, Tbot=1473, ysize=50e3
    set Function expression = Ttop+(Tbot-Ttop)*(ysize-y)/ysize
  end
end

subsection Boundary temperature model
 set Fixed temperature boundary indicators = top, bottom
 set List of model names = function
 subsection Function
   set Function constants = Ttop=273, Tbot=1473, ysize=50e3
   set Function expression = (y>ysize/2 ? Ttop : Tbot)
 end
end

subsection Boundary velocity model
  set Prescribed velocity boundary indicators = bottom y: function, left x: function, right x: function
  subsection Function
    set Variable names      = x,y
    set Function constants  = v=0.02, xsize=100e3, ysize=50e3
    set Function expression = (x<xsize/2 ? -v : v); (y<ysize/2 ? ysize/xsize*v : -ysize/xsize*v)
  end
end

subsection Compositional fields
  set Number of fields                        = 2
  set Names of fields                         = pyroxenite, weak_seed
  set Compositional field methods             = field, field
end

subsection Initial composition model
  set List of model names = function
  subsection Function
    set Variable names      = x,y
    set Function constants  = xsize=100e3, ypos=30e3, rad=5e3
    set Function expression = 0; if((x-xsize/2)^2+(y-ypos)^2<rad^2, 1, 0)
  end
end

subsection Boundary composition model
  set Fixed composition boundary indicators    = top, bottom, left, right
  set List of model names                      = initial composition
end

subsection Heating model
  set List of model names =  latent heat
end

subsection Material model
  set Model name =  compositing
  set Material averaging =  harmonic average only viscosity

  subsection Compositing
     set Viscosity = visco plastic
     set Density = visco plastic
     set Thermal expansion coefficient = visco plastic
     set Specific heat = visco plastic
     set Thermal conductivity = visco plastic
     set Compressibility = visco plastic
     set Entropy derivative pressure = latent heat melt
     set Entropy derivative temperature = latent heat melt
     set Reaction terms = latent heat melt
  end
 
  subsection Latent heat melt
    set A1 = 841.32
    set A2 = 1.3463e-7
    set A3 = -5.3158e-18
  end

  subsection Visco Plastic
    set Densities                                  = 2800 # kg/m3
    set Thermal diffusivities                      = 1e-6
    set Heat capacities                            = 1000
    set Thermal expansivities                      = 0

    set Viscous flow law                           = dislocation

   set Prefactors for dislocation creep          = 2.7e-18, 1, 1e-15
    set Activation energies for dislocation creep = 480e3
    set Activation volumes for dislocation creep  = 1.1e-5
    set Stress exponents for dislocation creep    = 3.5

    set Angles of internal friction                = 10.0, 1, 9.0 # degrees
    set Cohesions                                  = 10e6 # Pa
    
    set Minimum viscosity                          = 1e19 # Pa s
    set Maximum viscosity                          = 1e25 # Pa s
  end
end

subsection Mesh refinement
  set Initial global refinement                = 4
  set Initial adaptive refinement              = 2
  set Time steps between mesh refinement       = 1
  set Strategy                                 = composition, strain rate
  set Refinement criteria merge operation      = max
  set Refinement criteria scaling factors      = 1, 2
end

subsection Postprocess
  set List of postprocessors                   = visualization
  subsection Visualization
    set Time steps between graphical output    = 1
    set Output mesh velocity = true
    set List of output variables               = material properties, strain rate, heating, melt fraction
    subsection Material properties
      set List of material properties          = density, viscosity
    end
  end
end

Hi Petra,

Thanks for the post! I’ll admit I haven’t actually used the latent heat heating model before, but looking at the source code for the latent heat model, the actual heating terms (which is what is output for the heating visualization postprocessor), do not depend on the melt fraction. Instead, they only depend on temperature, densities, some derivatives, and the velocity.

I hope this helps!

Cheers,
Daniel

Hi Daniel,
I think that the entropy_derivative_temperature and entropy_derivative_pressure are set in the Latent heat melt material model and that they depend on the change of the melt fraction… Isn’t it true?
Petra

Hi Petra,

I think I know what the problem is.
You use the latent heat melt material model to compute latent heat of melting, and the melt fraction postprocessor to visualize the melt fraction. They both use the same melting model, but they do not know about each other’s parameters. So when you change the A1, A2, A3 parameters for latent heat melt, this does not change them in the postprocessor. So if you add another subsection

subsection Melt fraction 
    set A1 = 841.32
    set A2 = 1.3463e-7
    set A3 = -5.3158e-18
end

within Postprocess/Vislualization then both should hopefully be consistent.

I know this is confusing; and we’ll have to think about a way to make this more consistent. But for now, this should hopefully work.

-Juliane

Hi Juliane,
Thank a lot, it works!
Then, is it true that there is no direct way to plot the melt fraction computed by the Latent heat melt material model?
One more question (sorry for that): what does the original (wrong) visualized output of latent heat mean (see the figure)? It looks like a combination of two different sources of latent heat. (The white contour shows the extent of partial melting computed by the Melt fraction postprocessor.)

Thanks again!
Petra

Hi Petra,

Then, is it true that there is no direct way to plot the melt fraction computed by the Latent heat melt material model?

No, there’s no way to directly plot the melt fraction computed by the Latent heat melt material, and that is also because it does not directly compute a melt fraction. It only computes the derivative of the melt fraction with respect to temperature and pressure, which are the terms needed for latent heat generation.

What does the original (wrong) visualized output of latent heat mean (see the figure)? It looks like a combination of two different sources of latent heat.

That is very hard to tell without looking at all the other outputs (temperature, pressure, etc.), but I do have an idea. The Katz 2003 melting parametrization has a kink at the point where all clinopyroxene has melted. If you increase melting beyond that, the change in melt fraction with increasing pressure at first becomes smaller (because there is no clinopyroxene to melt anymore). Since the latent heat generation depends on the derivative of the melt fraction with respect to temperature, I would expect a discontinuity at that point. But that is just a guess.

Cheers,
Juliane

Yes, you are probably right! It is the first time I played with this melting model, so it surprised me.
Thanks again!
Cheers,
Petra