Hello everyone,
I am not an expert in Molflow. To gain further familiarity with this software, I have a question. I have a vacuum chamber with a volume of 3 liters, and I want to evacuate it from atmospheric pressure (1000 mbar) to 1E-9 mbar using a turbomolecular pump and an ion pump. Is it possible to calculate the entire evacuation time using Molflow?
Hello Cristian: Molflow+ implements the Test Particle Monte Carlo (TPMC) method, and as such it doesn’t consider molecular collisions (other than for a new feature recently introduced which considers a STATIC background gas). This means that its range of validity, pressure wise, is only that for which the mean-free path (MFP) is at least a few times bigger than the “typical” dimension of your vacuum system, usually referred to as “molecular flow conditions”.
The typical dimension is, for a cylindrical tube, its diameter, for a cube its edge. For 3 l volume, the equivalent cube has an edge of slightly less than 15 cm, and a 150 mm ID cylindrical tube has an equivalent length of ~17 cm. So, you need to calculate the MFP for the gas you are simulating, multiply it by 2-3 (doesn’t matter much) and then see what pressure that corresponds to. For N2 (or air), the MFP corresponding to 15 cm is around 4.510^-4 mbar (see MFP calculator linked at the end of the message, and screenshot). So, pumping down from atmosphere to this pressure, let’s say 210^-4 you need to use the analytic formula P=Po*(1-exp(-S/Vt), where Po is 1013.25 mbar (atmospheric pressure), S the pumping speed in l/s, and V the volume (3 liters in your case).
Starting at that point, ~2 10^-4 after xx seconds (xx from the formula above), you can simulate the pump-down in molecular flow, via a time-dependent simulation. For that you need to look at the relative webinar and entry on our website, which explains how to do it.
Comments: neither the turbo pump (TP) nor the ion-pump (IP) can pump down from atmospheric pressure, so in the formula above you should use the pumping speed of your primary pumps (typically given by manufacturers in m3/h or per minute, sometimes in other fancy non-metric units, especially if you live in the USA). Once the TP is at full speed, its pumping speed will depend only on the temperature and mass of the gas, i.e. it has a flat pumping speed curve, while the IP has a variable pumping speed curve. Typical IP pumping speed refers to SATURATED pump at its peak pumping speed value (which is reached around few 10^-7 mbar or so, depending on the manufacturer and type of pump (diode, triode, StarCell, etc…)… so if you reallyu want a precise calculation you need to parametrize the IP pumping speed as a function of the pressure, which CAN’T be done (yet) in Molflow+. The pump down difference between constant pumping speed and variable pumping speed for the IP, if the TP is pumping at the same time, is not probably going to be large.
Remember to simulate in your geometry also any tubular connections, elbows, crosses, diameter restrictions, etc… that may be installed between the TP and IP and the chamber you are pumping down, because the EFFECTIVE pumping speed in the 3 liter chamber depends on the CONDUCTANCE between each pump and the vacuum chamber.
I conclude here, hope it clarified this issue.
Cheers.
MFP calculator: Mean Free Path Calculator (omnicalculator.com)
Dear Roberto,
Thank you for your informative answer!
I am not concerned about pumping down from atmospheric pressure to mid vacuum. Indeed, I would like to assess the evacuation time from mid vacuum to UHV, and investigate whether our piping and connections from the main chamber to the pumps are correct or not.
Your answers are very hepful to me, and I appreciate it.
Thank you for your time.
Best Regards
Let me know if you need more information. Thank you very much. Best regards.
Obs.: Correcting the image: in the first two chambers there are mechanical pumps and in place of the Vacuum Conductance use Molecular Conductance, it’s more correctly.
Cristian,
The fact these chamber are separated by simple orifices allows you to separate it into 5 coupled equations, which can further be simplified if the pump speed is larger than the flow through the downstream orifices, so the pressure will be independent of any downstream chambers. There are simple analytic expressions for flow through orifices through the whole pressure range. I would only use molfow on the last two chambers if you needed to know the spatial distribution of the gas. You would use the results of solving the equations for the other upstream chambers to use as the gas input for the first molecular chamber.
Alan
Hello Gabriel (and Alan)!
I ran through the paper, and if I understand correctly, it shows the analytic solution, citing the short tube conductance in molecular, viscous and atmospheric conditions (and proceeds to calculate the pressure using these forumalae), just like Alan suggested. If I understand correctly, you’re looking for the static solution, not the evacuation time.
Recreating this in Molflow doesn’t work because Molflow assumes molecular flow regardless of the input. We don’t have much experience going to higher pressures, we only know that Molflow and its physics isn’t valid.
What is the final goal? Benchmarking the paper’s results? Getting the evacuation time?
I’m afraid I can’t help too much. If you remember our previous discussion, I was suggesting to go to COMSOL for the high-pressure parts, here is the animation I made last October:
Maybe this is something you can use for the high-pressure parts.
