I have a few questions concerning variables (time dependent) and the formula editor.
Is it true that only the sticking factor (and therefore the pumpspeed) and the outgassing can be a self created variable via the time parameters window, instead of the temperature or sojourn time(binding energy and frequency) for example? Or is that a problem because of the "catch 22" issue you've mentioned before?
Second question, is it possible to use your own created variables in the formula editor? This will be very useful for me because it will allow me to see what happens at different times as the formula editor shows the solutions per step. (something I assumed as the solutions turn blue?)
If not, could ''time moments''[sec] be a variable in the formula editor? Then I will be able to insert the same formula I used for the time parameters and find the solutions per moment, insead of inserting self created variables directly into the formula editor.
Temperature would be straightforward to implement, but time-dependent mode was designed for processes up to a few seconds. Whereas the above 3 quantities are quickly changing (gas puffs, pumps starting, valves opening/closing), we couldn't imagine a situation where temperature (or sojourn time) changes abruptly. ("Catch 22" doesn't apply in this case as I guess that simulation results, like pressure, don't really heat/cool a surface.) What would be your usage scenario?
You're right, any number in blue (formula result, facet hits in lower right corner, physical quantity in facet details window) means that it's a time-dependent result belonging to the currently displayed step. Formula editor can't evaluate user parameters, although it would be a nice idea to add. FYI, user-defined parameters are linearly interpolated between defined time-value pairs, as plotted in the parameter editor, and lin-log interpolation is planned later.
It's a good idea to add "current time" and "parameter at this moment" as recognized formulas to the formula editor, I'll add it to our to-do list. As usual, I don't dare to give you a promised deadline but we'll do it soon.
This is part of the ET Pathfinder(interferometer). Prototype for the Einstein Telescope (third generation gravitational wave detector).The main focus will be the mirror surface and the surrounding shields.
There is a strategy to avoid ice buildup on the mirrors. The inner He-cooled thermal shield (blue) will remain at temperatures of 333 K while the outer liquid Nitrogen shield (green) will be cooled under 100 K while creating the UHV. It’s expected that the remaining water molecules will be trapped on the LN2-cooled shield and freeze instead of sticking to the mirror surface. Because the sojourn time is shorter at high temperatures. In the end the mirror temperature will be 10K with the least amount of water molecules left. We are using Molflow+ to simulate if this actually happens. If we don’t heat up the system it’ll take years to get rid of the molecules. If we can heat the full system to 338 degrees (The electronics prohibit us to heat above 338) we can deplete the monolayer of water to <0.1 percent coverage in about 2 weeks of time.
In this situation a lot of things will vary over time. Like temperature, sojourn, sticking, outgassing... ect. For some of the variables. I will have to use the iterative simulation like you said in the other forum post.
Do you think it will be possible to use the time dependent mode in Molflow+ for longer periods of time instead of its intended use (seconds)?
"Do you think it will be possible to use the time dependent mode in Molflow+ for longer periods of time instead of its intended use (seconds)?"
Hello Vera: yes, it is indeed possible to simulate reliably hours. I've been running since several days a simulation for 1 h, the pumpdown of a 1m3 cube initially filled with gas and with time-dependent desorption... just to validate the code against the analytical formula (initial exponential followed by a 1/t decrease, with t being the time). So, no problem to do hours... provided you have enough computing power. I'm running my simulations on a 16-core server, which we rent from our IT group.
On this subject I'd say that there is a growing interest from the external to have the possibility to have time-dependent variables all over Molflow+, not only the 3 mentioned by Marton.... in fact it would suit also some new needs we have here to simulate new vacuum systems for advanced projects.
We'll have to clearly address this internally, among us, to see when the code change could start... I understand from previous discussions with Marton that it is not a trivial software update, but needs to be done...
Possibility of running long simulations is good news! The computing power will not be a problem.
Do you have any experience running iterative simulations yourself using Molflow+ or know of someone or a company that has done it before? So multiple variables can be changed over time.
Just wondering if it is an achievable way to get realistic results.
I have done only tests, not practical simulations, sorry.
As discussed with Marton many times already, I think what is necessary, until the code is changed so that it can do it, is a "quick" way to change/update the value of selected facet properties.
A python script or something like that, capable of reading the files in Molflow+ format (gdml?) is what is needed. Unfortunately I'm not sure it can be done in a general way, so that it works for each and every case... it should be the particular user which writes it.
Marton had done a nice study case for the saturation front of NEG-coating in the MAX-IV crotch area, he presented it at an IPAC conference.
