Final Design and Molflow Simulations of a Small Scale Multipurpose High Vacuum Chamber

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
Post Reply
Michael Bretti
Posts: 177
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Final Design and Molflow Simulations of a Small Scale Multipurpose High Vacuum Chamber

Post by Michael Bretti » Mon Dec 17, 2018 1:13 am

This post will briefly cover the final design details as well as some of the simulation results for the Small Scale Multipurpose High Vacuum System V5 assembly. I have posted extensively about this system in previous postings. The final design of the chamber is as follows:

2,75in Conflat High Vacuum Chamber V5.jpg

Here is the chamber attached to the previously posted Integrated High Vacuum Test Stand:

Small Scale Multipurpose High Vacuum System V5 Assembly.jpg

The full engineering specs and details of the system can be found on the project page for it:

http://appliedionsystems.com/portfolio/ ... system-v5/

Like the previously post detailing the Integrated High Vacuum Test Stand, I also ran Molflow+ simulations for this chamber. The full simulation details, including model preparation, input parameters, results, discussions, and simulation files, can be found here:

http://appliedionsystems.com/portfolio/ ... m-chamber/

All of the results are on the above page, so I won't post everything here. Like my first Moflow simulation, I looked at three different cases for system preparation: unbaked and pumped 1 hour, unbaked and pumped >24 hours, and baked and pumped >24 hours. For brevity, I will share the results of the first simulation, looking at the simulation results for the system unbaked and pumped for 1 hour.

The first result looks at average molecular trajectories in the system:

Small Scale Multipurpose High Vacuum System V5 Final - Molecular Trajectories.jpg

Based on the simulation results, for a pumping speed at the inlet of the diffusion pump of 600 L/s, the resulting speed at the inlet of the 5-way 2.75" conflat cross chamber is around 62 L/s for the first case, dropping down to about 38 L/s for the remaining simulation cases. These numbers are higher than I initially calculated, which means I can handle a much greater ultimate gas load than originally anticipated.

The main results focus on the pressure profile through the system. The following results shows the pressure gradient through the entire assembly for the first simulation case:

Small Scale Multipurpose High Vacuum System V5 Final - Unbaked Pumped 1hr.jpg
Small Scale Multipurpose High Vacuum System V5 Final - Pressure Mapping Scale Unbaked Pumped 1hr.jpg

By applying a transparent measurement facet through the model, a pressure plot can be generated. The x-axis consists of arbitrary units, where Molfow breaks up the measured plane into 100 equal length segments. The y-axis is in mbar:

Small Scale Multipurpose High Vacuum System V5 Final - Unbaked Pumped 1hr Facet.jpg

Log scale plot:

Small Scale Multipurpose High Vacuum System V5 Final - Unbaked Pumped 1hr Pressure Plot Log.jpg

Linear scale plot:

Small Scale Multipurpose High Vacuum System V5 Final - Unbaked Pumped 1hr Pressure Plot Linear.jpg

An interesting thing to note from the graphs is around 10 on the x-axis. This sudden increase in pressure is due to the flow impedance seen at the baffle. At around point 65 on the x-axis, we see a slight bump in the pressure. This is due to the pressure rise seen in the gate valve, where I estimated an increase in outgassing load due to the seals used for the valve. The pressure then levels off in the 5-way cross to around 2.8 x 10^-5 mbar. This shows that assuming no leaks and backstreaming in the system, it is reasonable to expect pressures in the chamber in the lower 10^-5 mbar region for a system that has just been assembled with minimal pumping and no baking. With baking and extended pumping it is expected to reach an ultimate pressure of 1.2 x 10^-6 mbar, which meets the criteria for this system.

One additional thing I was also interested in looking at was the difference in resulting pressure between the 5-way cross main chamber, and the 4-way cross vacuum instrumentation ports. Since high vacuum gauge will be located on one of the arms of the lower 4-way cross, due to the decreased conductance, and hence decreased pumping speed the further up towards the main chamber (away from the pump), the more of a difference there will be between the readings at the gauge and the actual chamber pressure. By plotting measurement facets from the 5-way cross chamber and 4-way cross instrumentation ports, I can now reasonably calibrate my gauge response to the difference between the two. The results are shown below. The green trace represents the 5-way cross chamber, and the blue trace represents the 4-way cross instrumentation input:

Small Scale Multipurpose High Vacuum System V5 Final - Unbaked Pumped 1hr Pressure Plot Comparison.jpg

As can be seen from these results, it is reasonable to expect that the pressures read at the instrumentation ports will be slightly lower than at the chamber, which can be factored in during system runs. In addition, with these plots, it is also possible to estimate the pressure at any part of the system during pumpdown based on the measurement at the gauge. For example, this means I would be able to estimate the pressure seen at the inlet of the diffusion pump, and use this to gauge pump performance.

Post Reply