[question] How do I determine the sputtering and reflection yield from "output, incident, stopped = compound_bca_list_1D_py" code output in test_rustbca.py?

[HexSygon]

I believe I understand the inputs and outputs but not how to calculate sputter and reflection yields from the output.

Also, what is the compound reflection code used for/is this how you determine reflection coefficients for ions on compound targets?

I want to find the sputtering and reflection coefficients of D and C ions on a half W - half C target at various energies.

[drobnyjt]

@HexSygon the way to do it is to count the number of sputtered or reflected atoms and then divide by the number of computational ions / number of samples. In the examples folder, test_morse.py includes the way I usually do this:

...

if np.size(reflected) > 0:
  num_reflected = np.shape(reflected)[0]
  energy_reflected = np.sum(reflected[:, 2])
else:
  num_reflected = 0
  energy_reflected = 0.

return num_reflected/num_samples, energy_reflected/energy/num_samples

In this code snippet, the number of reflected atoms is found using np.shape as the number of rows in the reflected output file. Same procedure for the sputtered atoms - count the number of rows and divide by the number of computational ions.

[HexSygon]

Thanks Jon,

I'll give this a go and let you know if I come across any issues.

Best,
Alec

[drobnyjt]

@HexSygon You're welcome - I just realized I only responded to half of your comment; sorry about that.

Yes, the compound version is for homogeneous targets composed of more than one element. So for example H reflecting from CW.

[HexSygon]

Hey Jon, I hope you've been well. I have a few more questions for you if you don't mind.

1.) At the end of the compound material code input it states "[[target['n']/1030, target['n']/1030], [target['n']/10**30, 0.0]], [50.0, 1e6]" for a 50 Angstrom layer of W-H on top of the tungsten target

I find it interesting that the W-H layer is listed second, but it's thickness is listed first. Can you please confirm that this is corrected stated in test_rustbca.py?

2.) Is there a way to specify a mixed ion concentration? Like 95% deuterium and 5% carbon for example?

3.) Is there a way to specify a non 1:1 mixed target ratio? Like 75% carbon and 25% tungsten in the top mixed layer for example?

[drobnyjt]

Hi @HexSygon - of course, no problem. One thing to note - I am currently rewriting the documentation of the Python functions and should be finished this week. Hopefully the new documentation will be more understandable. Edit: this has been completed.

Question 1): I can see how this could be confusing. The material species are defined in the order [W, H] as per the Z2 argument (and the other material species arguments), [tungsten["Z"], 1.0]. This means the densities argument will be in the order [[nW1, nH1], [nW2, nH2]], for layers 1 and 2 - to state it explicitly, each nested list is its own layer. The following code snippet might help illustrate why this ordering was chosen instead of the transpose:

python3
>>> import numpy as np
>>> densities = np.array([['nW1', 'nH1'], ['nW2', 'nH2']])
>>> print(densities )
[['nW1' 'nH1'] # first layer
 ['nW2' 'nH2']] # second layer
>>> densities[0]
['nW1', 'nH1'] # first index of densities is first layer

Looking at the densities of the first layer, nW1 and nH1 are both set to the default tungsten number density, so that represents a layer with a 1:1 W:H ratio assuming zero swelling of the tungsten with increasing hydrogen concentration.

Moving on to the second layer, with densities [nW2, nH2], nW2 is the default tungsten number density but nH2 is zero, so the second layer is pure tungsten with zero hydrogen.

The layer thicknesses argument is [t1, t2], where t1 is the thickness of the first layer with number densities [nW1, nH1] and t2 is the thickness of the second layer with number densities [nW2, nH2] . t1 is 50, so that represents a 50 angstrom layer of W-H; t2 is 1e6, so that represents a 1e6 angstrom layer of pure W. Hopefully that explanation makes sense - keeping track of the different species and layer indices can be tricky, especially when the number of layers and species is the same. A similar issue exists for the interaction_potential argument in the standalone code, but like in this case the information is just two-dimensional so there's not really a better way to represent it.

Question 2): This is possible in the standalone code, but not in the python functions. The best thing to do would be to call the function twice, once for carbon ions and once for deuterium ions, and then when you process the results you can weight them appropriately. Edit: actually, I have to correct myself - any of the Python functions that accept a list of incident ion properties can do this. For example, if you wanted to run 2 incident ions, one of D and one of C, your arguments would look like Z = [deuterium["Z"], carbon["Z"]], m = [deuterium["m"], carbon["m"]], etc. To do some prescribed ratio, you could build two lists for each incident ion argument and concatenate them like this:

Z1 = [carbon["Z"]]*95
Z2 = [deuterium["Z"]]*5
Z = Z1 + Z2

Question 3): Yep - the densities argument is just the number density of each species in each layer. Ideally you would look up the compound in question and calculate the number density of each species in the layer from the mass density and the stoichiometry - in test_rustbca.py I chose to assume that the interstitial gas atoms did not increase the target volume just for the sake of simplicity, but this is not always a realistic assumption.

[jguterl]

suggesting to turn those issues into a FAQ in the wiki.

[drobnyjt]

@jguterl that's a great idea; I've added a new sidebar, an FAQ page, and this specific question to it. I'll spend some time going through closed issues with the [question] tag and add any others that have popped up as well.

[HexSygon]

This FAQ page sounds helpful and is a good middle point

…

On Wed, Apr 16, 2025 at 12:28 PM Jon Drobny ***@***.***> wrote: @jguterl <https://github.com/jguterl> that's a great idea; I've added a new sidebar, an FAQ page, and this specific question to it. I'll spend some time going through closed issues with the [question] tag and add any others that have popped up as well. — Reply to this email directly, view it on GitHub <#262 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BKCBKD3TJSQAWW2ZSLLLNGL2Z2AMLAVCNFSM6AAAAABXK4MVMKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQMJQGEYDOOBWHE> . You are receiving this because you were mentioned.Message ID: ***@***.***> *drobnyjt* left a comment (lcpp-org/RustBCA#262) <#262 (comment)> @jguterl <https://github.com/jguterl> that's a great idea; I've added a new sidebar, an FAQ page, and this specific question to it. I'll spend some time going through closed issues with the [question] tag and add any others that have popped up as well. — Reply to this email directly, view it on GitHub <#262 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BKCBKD3TJSQAWW2ZSLLLNGL2Z2AMLAVCNFSM6AAAAABXK4MVMKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQMJQGEYDOOBWHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>