Setting up coupled SR / photodesorption simulations

Wednesday, July 9, 2014 - 11:52

|

Synrad+ and Molflow+ were meant to be use together as a package. The latest beta versions allow you to do a SR simulation in Synrad+, then import the absorbed flux map to Molflow+, convert it to a desorption map, and do a standard vacuum simulation, this time having the walls as outgassing source. Follow the steps below for the whole procedure.

Part 1: Synrad+ simulation

In this example, we will use a short sample tube which receives SR light from an upstream dipole.

Image

As usual, you have to import or create the vacuum chamber and set up the magnetic regions which will generate the SR photons. You should watch out for two things:

  • Run the simulation Fluxwise (more statistics will be available on low energy photons)
    Image
  • Include low energy photons (4eV and up), as while their heat load is negligible, we're interested in their photodesorption effect
    Image

That should be enough for the simulation to run. As a next step, we need Synrad+ to count accurately the amount and the exact location of the absorbed SR flux (responsible for photodesorption). For that, we'll use textures.

You can determine manually which facets should count the absorbed photons. To be sure, you can use Selection->Select Abs>0 so every facet with at least one absorbed photon will be textured. At large geometries, however, you can do it more selectively: on the bottom right facet hit list, clicking the Flux column's header will sort the facets by the amount of flux absorbed:

Image

This allows you to select only the facets which get a high flux, and omit those that get several orders of magnitude lower flux.

Once it is done, using the Mesh... button, add a texture to these facets with Count absorption counting mode.

Image

This will show you where the flux is absorbed in your system:

Image

We can see on the above logarithmic color scale that the majority of the photons will absorb on the incidence surface and the pipe end, and the rest will be scattered.

That's all you have to do in SynRad+. Save the simulation and open MolFlow+.

Part 2 - Importing Synrad+ data and conversion to photodesorption

For a typical material, the molecule yield (that is, the number of molecules desorbed per incident photon) will depend on the absorbed photon dose: as the material's surface is cleaning up, less and less photons will desorb. The exact numbers must be measured, below is an example chart for some measurements at CERN with electroplated Copper.

Image

For reference, these plots are digitized, and can be dowloaded from here (also shipped with Molflow):

Desorption yields for common materials

Note: If you look at the content of these files, you will see two differences compared to the chart: the numbers on the X scale have been divided by 70 to convert between linear dose (photons/m) and surface dose (photons/cm2, used by Synrad+) (this factor of 70 is valid for a typical tube-geometry desorption measurement). The Y values show the yield in molecules/photon.

To set up a vacuum simulation with SR data, proceed the following way:

  • Open the simulation you just saved in Synrad+ (yes, Molflow+ can open .syn7z files). You will see the vacuum chamber without SR data (so magnetic regions, material data, etc. will be removed, and only the pure geometry will remain).
  • Import desorption data. For this, use the File->Import Desorption File->SYN file option
    Image
  • In the upcoming dialog, you have to specify the file the flux data is stored in (as textures).

    In our case, as we're working with the SYN file directly, you can press the Use current file button. If you have saved the geometry as a GEO file, you'll have to select the original SYN file with the flux data with the Load... button. You could even use a different SYN file as source, but beware: facets are referenced by their numbers, so make sure that the geometry is the same. Important: when you load a file, all facets in your geometry which have flux data in the source file are selected. Later, when you will press the Import for selected facets button, the desorption data will be imported for the selected facets only. With this method, you can import different photodesorption yield files for different materials.
  • Next, you have to specify the conditioning time. The flux (photons/second) stored in the SYN file will be multiplied by this value, resulting in a dose. Therefore, if you're interested in your system's vacuum state after 100h of conditioning, you would enter 360000 here.
  • Then you have to load the file containing the photodesorption yield. Here, you have to select one specific gas, do the simulation for it, and - if needed - add the partial pressures at the end. (You can do this as UHV systems are linear.)
  • Finish the import process by pressing the Import for selected facets button. For those facets where desorption data was succesfully imported, you will see the Use desorption FILE option, selected by default, in the desorption menu:
    Image
  • You might want to double check the order of magnitude of the gas desorption in Global Settings:
    Image
  • Imported data is available in the advanced facet details pane as well (bottom):
    2019-02-27 16_03_52-Molflow+ 2.7.5 (Feb 21 2019) [onepoint3_full.syn7z].png
  • Set gas mass in Global Settings, set up pumps, and you're ready to simulate:

Image

On the screenshot above, you can see that desorption, represented by blue dots, happens mostly on the locations where SR flux was absorbed. You can do a more quantitative analysis by adding textures with count desorption mode to the walls:

Image

Again, on this logarithmic color scale you can see that the highest desorption corresponds to the highest absorbed flux.

  • Final note 1: imported photodesorption data isn't saved with the old GEO file format, use the new XML type.
  • Final note 2: to be clear, in Molflow+, imported desorption maps are separate from textures. A facet could have an outgassing map but not have a texture, or vica versa. Also, the outgassing map and the texture resolutions don't have to be the same.