There are several issues that are raised here. For one there is the matter of units. Do we want to have a consistent set of units which would make densities kg/m^3? I know this looks kind of ugly but from a practical point of view it will lead to less errors.

Another issue is material naming. For elements and compounds I would use the OpenPMD naming convention #33. And I would get away from the G4_ prefix unless, for a particular material, there is a good reason not to.

Another issue is that having a single Matter element to represent foils, gas cells, etc. to my mind lumps too many different types of objects under one name. I would rather see separate elements. Especially since the parameter set needed for these different types will undoubtedly be different.

Yes. density in kg/m^3 is ugly but reasonable. Do you have a suggestion for the atomic mass?

Is there a list of OpenPMD materials? I could not find any reference on that.

As someone working on fragment separators, to me everything is just matter with composition and shape.
What element specific parameters come to your mind?

• You don't need atomic mass directly. EG: #3He is good enough.
• OpenPMD does not define materials yet. I propose adding to OpenPMD a materials list and then PALS can inherit from that. Much better to have everything is in one place.
• in terms of element parameters, there could be something like gas pressure that is needed. We should make a survey to see what people need.

Also: Vacuum windows was mentioned. This should definitely be in a category by itself.

@danielkallendorf

There are a few conflicts to be resolved after the latest merge in main. Could you resolve the conflicts and apply the suggestions from @DavidSagan's review? Thanks!

Comments from today: define ratios more precise (mass/number of atoms/...)
Allow for Compounds of Compounds or introduce an intermediate element

Updated Version

Matter Element

Materials which fully occupy the beamline, like targets, (stripper-)foils, vacuum windows, gas cells or degraders.
This element can cause energy-loss, angualar and energy straggling, as well as change of charge state or particle type.

Element parameter groups associated with this element kind are:
...

• MaterialP: Matter parameters.

MaterialP: Definition of materials

MaterialP:
- state:             # ["solid", "liquid", "gas"]
- material:          # [CompoundP,ElementP] definiton of the material in terms of compounds or elements.

The definition of materials is inspired by Geant4's material handling. One can either use the name, which is the symbol of an element ("Cu", "H", ...) or compound ("G4_STAINLESS-STEEL").
In the future all elements and materials in the G4 Manual will be supported, without the G4_-prefix for elements.

Alternatively custom elements can be defined by

ElementP:
- name:      # [string] The name of the material
- density:   # [kg/m^3] The density of the material
- Z:         # [int]    The atomic number
- A:         # [int]    The mass number (equals Z+N)
- N:         # [int]    The neutron number
- m:         # [u]      The atomic mass

and compounds by

CompoundP:
- name:       # [string] The name of the material
- density:    # [kg/m^3] The density of the material
- elements:   # List of MaterialP or CompoundP of which the compound consists
- ratio:      # List of ratios of the elements
- mass_ratio: # List of mass-ratios of the elements

ratio describes the ratio in terms of atom/molecule count, while mass_ratio describes it int terms of mass fractions.
If density is not given it will be inferred form the densities and mass fractions of the elements.

Examples

matter: "Cu" or matter: "G4_STAINLESS-STEEL" or define a new material.

Liquid deuterium can be defined with the ElementP parameters:

ElementP:
- name: "Deuterium(l)"
- density: 160
- Z: 1
- A: 1
- m: 2.01

and deuterated polyethylene (C2H4) can be defined based on the previous definition

CompoundP:
- name: "deutPE"
- density: 1050
- elements:
  - "C"
  - "Deuterium(l)"
- ratio:
  - 2
  - 4

This is great! I think I spot a possibility to simplify this a little by using 2 instead of 3 parameter groups, let's discuss tomorrow :)

My idea was to separate the material properties MaterialP (what is it made of) from the actual piece of matter MatterP with thickness (and in the future maybe geometry), because MaterialP has probably other applications, like defining the material of the aperture.

Ah, good point.

@danielkallendorf After thinking it over, to answer your question, I would advocate that things like atomic masses can be output parameters (#107) and one of the aims of the PALS C++ interface I am involved with will be to output this info with the expanded lattice.