Integrate mass transfer coefficient calc

[sashahafner]

Idea is to call Kga_onda function from tfmod function.

Some issues to sort out first:

• Can we use only normalized flow are input (not Qstd etc.)?
• ssa = specific surface area (m2 area / m3 total volume)?

[afeilberg]

No problem at all to use normalized flow. Yes, ssa is the specific surface area of the filter material in m2/m3 total filter volume

[sashahafner]

I've made these and some other changes in 0f35f5e75e4f91df7f5c5f4c7c628fe5e2e27793 . See demo 5. We can discuss these.

[afeilberg]

@sashahafner: Not sure how to access the value of Kga estimated by Onda correlation, in order to compare with my own calculation. Must be something really simple...

[sashahafner]

@afeilberg I just added a demo subdirectory for this: demos/08... You can edit the inputs in demo.py and run the whole script and find the result in Kga_output.txt, or else run the py file contents interactively and check the Kga value in the console:

I was just thinking that this function should be "vectorized" so any of the inputs could have more than one value. I have to think/learn how to do this in Python.

[sashahafner]

I also added it to the output from tfmod() so if you run the 07... demo lines in demo.py interactively you can see the value calculated by tfmod():

I should have the demos export this and probably some other information to a file.

[sashahafner]

Note that I had to move the Kga function to the same script as the model function. See here https://github.com/sashahafner/pystupid/issues/3

[afeilberg]

OK

I found the reason for the high mass transfer coefficients. In the Onda correlations, densities are divided by viscosity in units of kg/(m*sec). Therefore, densities should be provided in kg/m3 to give the correct units of kga and kla. I don't think we are using densities except in the Onda correlations...?

[sashahafner]

I think that unit is right. But that is what is already used in demo 08:

# Set input values in function call
Kga = Kga_onda(pH = 7, temp = 20, henry = (0.1, 2000.), pKa = 7., pres = 1., ssa = 1084., v_g = 0.03, v_l = 0.003, por_g = 0.28, dens_l = 1000.)

https://github.com/AU-BCE-EE/tric-fil-mod/blob/ef4771cc96eb74f6c78d536973d62c8ab7fcdecb/demos/08_Kga_onda/demo.py

And defined in model:

   # dens_l = solution (liquid) density (kg/m3)

Line 132 here https://github.com/AU-BCE-EE/tric-fil-mod/blob/d8c7431904e495c902e12eda9ce224276a536754/mod_funcs.py

[sashahafner]

Oh! But now I see dens_l isn't actually used for viscosity conversion in the Onda function. Instead it has rho_l hard-wired to 1E6 on line 133 here https://github.com/AU-BCE-EE/tric-fil-mod/blob/d8c7431904e495c902e12eda9ce224276a536754/mod_funcs.py.

I'll fix it.

[sashahafner]

Fixed in 543a5325fa59536b445f5054db1f00fdfae511b4 but have not tested.

Note that air density is also in g/m3. We probably need to change that too right?

[sashahafner]

Do we want gas density set to air only? Here is an function that could possibly be converted from the biogas package:

> biogas:::gasDens('20O2:80N2')
[1] 0.00128613
> biogas:::gasDens('0.01O2:0.99N2')
[1] 0.001252195

(All g/mL, for common biogas units.) From here https://github.com/sashahafner/biogas/blob/dev/R/gasDens.R

It is for 0C, so would need temperature and pressure correction as in stdVol() https://github.com/sashahafner/biogas/blob/dev/R/stdVol.R

Almost certainly not worth the effort, but fun to think about.

[afeilberg]

Fixed in 543a532 but have not tested.

Note that air density is also in g/m3. We probably need to change that too right?

Yes, both should be in kg/m3

[afeilberg]

Oh! But now I see dens_l isn't actually used for viscosity conversion in the Onda function. Instead it has rho_l hard-wired to 1E6 on line 133 here https://github.com/AU-BCE-EE/tric-fil-mod/blob/d8c7431904e495c902e12eda9ce224276a536754/mod_funcs.py.

I'll fix it.

Yes, we have used both dens_l and rho_l, the latter for the onda calculation. I guess we can just use one of them...

[afeilberg]

Do we want gas density set to air only? Here is an function that could possibly be converted from the biogas package:

> biogas:::gasDens('20O2:80N2')
[1] 0.00128613
> biogas:::gasDens('0.01O2:0.99N2')
[1] 0.001252195

(All g/mL, for common biogas units.) From here https://github.com/sashahafner/biogas/blob/dev/R/gasDens.R

It is for 0C, so would need temperature and pressure correction as in stdVol() https://github.com/sashahafner/biogas/blob/dev/R/stdVol.R

Almost certainly not worth the effort, but fun to think about.

As a starting point we are focusing on air matrices, but you never know; there could also be other applications for biogas or N2. But it is not urgent to make flexible with respect to gas matrix.

[sashahafner]

I've tested updated function in demo 08 and get

Kga(onda) = 9.33e-02 sec-1

Does that look correct now @afeilberg?

63f1b78f0c7cbea57d391e862efe2581ff7c5e68

[sashahafner]

Oh! But now I see dens_l isn't actually used for viscosity conversion in the Onda function. Instead it has rho_l hard-wired to 1E6 on line 133 here https://github.com/AU-BCE-EE/tric-fil-mod/blob/d8c7431904e495c902e12eda9ce224276a536754/mod_funcs.py. I'll fix it.

Yes, we have used both dens_l and rho_l, the latter for the onda calculation. I guess we can just use one of them...

I switched everything to dens_l because it was the variable name used in the other functions (beyond Onda stuff, it is just used for converting units on Henry's law constant I think). But let me know if you want to switch dens to rho throughout. Would be fine with me.

[afeilberg]

@sashahafner Now we get the same with 4 significant digits. Found a few issues in my spreadsheet such as using the same gas constant and the same enthalpy for the Henrys law coefficient. This is great!

[sashahafner]

Great! I'll close this issue for now then.