Fix rock density

[jdkio]

@diaza pointed out that the rock density of the ND cavern for 2x2 is wrong. The ND-LAr cavern will be very similar. So this updates the rock density to match. See this conversation.

Does this affect the rock muon rate?

The difference in density of the rock isn’t quite as important as it might look initially. This is because while at the higher density there’s more interactions in the rock, there’s also more attenuation of the events due to ranging out. They don’t exactly cancel out 100% but the effect isn’t 2.82/2.33=1.21 - R. Hatcher

What's the source of the difference?

iirc, 2.82 is the dry density of the crushed rock, and 2.33 accounts for the porous wet rock. Some percentage of the rock is filled with water which decreases the avg density - me

New Parameters:

<D value="2.33" unit="g/cm3"/>
<fraction n="0.0147547" ref="Hydrogen"/>
<fraction n="0.0114328" ref="Carbon"  />
<fraction n="0.5637993" ref="Oxygen"  />
<fraction n="0.0431255" ref="Calcium" />
<fraction n="0.0028548" ref="Sodium"  />
<fraction n="0.0946920" ref="Aluminum"/>
<fraction n="0.0179492" ref="Iron"    />
<fraction n="0.2420140" ref="Silicon" />
<fraction n="0.0093778" ref="Potassium"/>

[ast0815]

The original also had sulfur in it. Is this supposed to be completely gone now?

Also, why did you give this the invalid label? That one signals to me that the PR is invalid and should not be merged.

[jdkio]

I removed the invalid label. I interpreted it as the numbers were invalid

As for the sulfur, I'm not sure. These are the numbers I was given. I think if you add up the relative proportions of the other elements, you don't get exactly the same numbers as previous as well judging by Al for example. That requires extra investigation. I asked on the original thread

[ast0815]

Any news about this? Are we sure the new composition is "correct" and should merge it?

[jdkio]

Sounds like it's complicated. My interpretation is that 2.84 g/cm^3 is still too high, so perhaps the simplest solution is to revert the changes to the composition, and only change the density. Maybe we can find someone we a more definitive answer. Here's Matt's response:

DUNE ND rock may be similar to NOvA ND rock, but be careful, because there are different layers of rock. I don't remember the depth of the DUNE ND. If it is not the same, the rock will be different. See MINOS-doc-2777 (attached: numinear.pdf).

But for most purposes, the main concern is DO NOT LEAVE OUT THE HYDROGEN. :-)

I'm attaching my talk from 2018 on NOvA ND rock composition. I later updated these estimates a bit, with the computation done in novasoft:Geometry/gdml/compositions/ndrock.py . Here's the relevant part:

# This defines ND rock as it is relevant for beam events.
# It does not attempt to simulate geological layers which may be
# relevant for cosmics.
#
# There is a fair amount of guesswork and heavy use of Wikipedia here,
# but the composition should be reasonable, and a lot closer to reality
# than "standard rock".  See also nova-doc-25794.

compositions = {
  'SiO4':  { 'Si' : 1, 'O': 4 },
  'AlO4':  { 'Al' : 1, 'O': 4 },
  'SiO2':  { 'Si' : 1, 'O': 2 }, # Quartz
  'Al2O3': { 'Al' : 2, 'O': 3 },
  'K2O':   { 'K'  : 2, 'O': 1 },
  'Na2O':  { 'Na' : 2, 'O': 1 },
  'CaO' :  { 'Ca' : 1, 'O': 1 },
  'CaCO3': { 'Ca' : 1, 'O': 3, 'C': 1 },
  'FeO':   { 'Fe' : 1, 'O': 1 },
  'Fe2O3': { 'Fe' : 2, 'O': 3 },
  'MgO':   { 'Mg' : 1, 'O': 1 },
  'TiO2':  { 'Ti' : 1, 'O': 2 },
  'P2O5':  { 'P'  : 2, 'O': 5 },
  'MnO':   { 'Mn' : 1, 'O': 1 },

  'kaolinite': { 'Al': 2, 'Si': 2, 'O': 5+4, 'H': 4 },

  # Omitted to avoid adding an element: 'Mg': 1,
  'montmorillonite': { 'Na': 0.33/2, 'Ca': 0.33/2, 'Al': 1,
                       'Si': 4, 'O': 10 + 2 + 2, 'H': 2 + 4 },

  'illite': { 'K': 0.5, 'Al': 0.67 + 2,  'Fe': 0.67,
              'Si': 2, 'O': 10+2+1, 'H': 2+2 }, # Omitted 'Mg': 0.67,

  'water': { 'H'  : 2, 'O': 1 },

  # Just elements, for checking old composition + water
  'oxygen':   { 'O'  : 1 },
  'silicon':  { 'Si' : 1 },
  'calcium':  { 'Ca' : 1 },
  'sodium':   { 'Na' : 1 },
  'iron':     { 'Fe' : 1 },
  'aluminum': { 'Al' : 1 },
  'potassium':{ 'K'  : 1 },
}

# From MINOS-doc-2777

matrix_density = 2.84
water_density = 1.00

# Measurement for "Lower portion (Shalely) of
# Scales formation (Maquoketa)"
void_volume_fraction = 0.22
rock_volume_fraction = 1 - void_volume_fraction

# We have been draining water out of this rock for 20 years
# This is a guess at the remaning water content.
# We had hoped to find better data in the sump logs, but weren't able to.
water_volume_fraction = void_volume_fraction * 0.5

air_volume_fraction = void_volume_fraction - water_volume_fraction

water_mass_fraction = water_volume_fraction * water_density /(
water_volume_fraction*water_density + rock_volume_fraction*matrix_density)

# Mass of air in the voids is neglected
rock_mass_fraction = 1 - water_mass_fraction

density = matrix_density * rock_volume_fraction + water_density * water_volume_fraction;

print('<D value="{0:.2f}" unit="g/cm3"/>'.format(density))

fractions = {
  'water': water_mass_fraction,

  # Here's the old composition, for checking how much difference the water
  # alone makes
  #'oxygen':    0.437,
  #'silicon':   0.257,
  #'sodium':    0.222,
  #'iron':      0.020,
  #'aluminum':  0.049,
  #'potassium': 0.015,

  # Some not-totally-wild guesses using Wikipedia "Shale"

  # Clay minerals
  'montmorillonite': 0.3 * rock_mass_fraction,
  'kaolinite':       0.2 * rock_mass_fraction,
  'illite':          0.2 * rock_mass_fraction,

  # Quartz and calcite.  Quartz is mentioned more often.
  'SiO2':  0.2 * rock_mass_fraction,
  'CaCO3': 0.1 * rock_mass_fraction,
}

numinear.pdf
novageorocktalk.pdf

[diaza]

I confirmed with Matt my understanding of his calculation:

Let me see if I understand how you got your number:
From the table below, you take the Matrix Density of the Lower portion (Shalely) and calculated what the density would be if 50% of the water drained away and left behind air. You then use Wikipedia to form some composition of shale. Does that sound right?

where the "table" is Table 2 from this pdf linked earlier.

My current guesstimate from looking at the above pdf and in
numi0494.pdf is that the ND (~200 ft underground) will lie in the Upper portion of the Maquoketa Group, composed of dolomite CaMg(CO_3)_2.

Looking at the table I, would calculate that, assuming all the water remains, the inputs should be
density: 2.37 g/cm^3
water mass fraction: 0.07
dolomite mass fraction: 0.93

That's all assuming the water content remains the same. Matt made an assumption that 50% of the water in the rock around the MINOS hall had been drained. Not sure what would be the correct assumption with the ND.

[ast0815]

Do we have some civil construction engineer or geologist we could ask about the water drainage? Seems like the kind of thing that Fermilab should have someone with knowledge about it.