High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

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.
Michael Bretti
Posts: 152
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Michael Bretti » Wed Jan 31, 2018 1:21 am

INTRODUCTION

PART I - OVERVIEW

For the past several months, I have been working heavily on the design of my high vacuum system. A good deal of my initial research has been going through and reading as many posts as possible from this forum, which has provided an extraordinary amount of knowledge from many years of cumulative experience of many members, slowly transitioning into a more intensive academic study of the subject. I have noticed that there does not appear to be a post yet diving into more rigorous and intensive engineering design, simulation, and calculation for high vacuum systems. This is certainly not needed for the average fusor builder, however I have found this field to be very fascinating and on the rather obscure and lesser practiced areas of engineering in general, and could prove helpful and useful for those looking to take a more rigorous approach to the design process of their system. Since this hobby largely involves long periods of time for waiting on finding the right components, sourcing parts from eBay and other various resources, as well as being a rather expensive hobby, I have decided to spend my free time while waiting for parts to complete all of the major design calculations and simulations that would apply to my system. This has proven to be a tedious and challenging endeavor, but incredibly rewarding in understanding the deeper principles of high vacuum systems and engineering. I have decided to focus purely on the high vacuum system itself for this next year while I await for more funds for my projects. This includes all of the required calculations, simulations, and designs related to the system, as well as building, conditioning, and working with the high vacuum system itself. The main goal for this year will be to establish a well prepared system capable of achieving the desired ultimate vacuum needed, as well as getting all peripheral support systems (power, control, instrumentation, data acquisition) properly running, and generating data on the vacuum system itself, including pump down curves, rate of rise curves, and other data to compare with how close the system actually behaves to my initial calculations. I have decided to share this long term design work here, which I will continue to update and post continuously throughout the year as progress is made. Because there is such a massive amount of information to be presented, I will break each informational post in this walk-through/walk-along in titled sections in the following order:

1.) INITIAL HIGH VACUUM SYSTEM CONCEPT DESIGN, REQUIREMENTS, AND MODELLING

2.) CALCULATION AND COMPARISON OF SYSTEM CONDUCTANCES AND EFFECTIVE SPEEDS FOR MOLECULAR FLOW FOR VARIOUS PROCESS GASES

3.) CALCULATION AND COMPARISON OF SYSTEM CONDUCTANCES AND EFFECTIVE SPEEDS FOR TRANSITIONAL FLOW FOR VARIOUS PROCESS GASES

4.) TOTAL GAS LOAD DUE TO OUTGASSING AND DETERMINATION OF ULTIMATE PRESSURE DURING PUMPDOWN

5.) PUMPDOWN TIMES FOR ROUGH AND HIGH VACUUM

6.) MAXIMUM GAS LOADS FOR PROCESS GASES FOR VARIOUS EXPERIMENTS

7.) THERMAL MODELLING OF THE HIGH VACUUM SYSTEM

8.) THERMAL MODELLING OF SYSTEM RUNNING A STANDARD FUSOR GRID

9.) ELECTRONICS, CONTROL, AND INSTRUMENTATION

10.) SYSTEM BUILD

11.) SYSTEM CONDITIONING, PUMPDOWN, AND BAKEOUT

12.) PUMPDOWN AND RATE OF RISE CURVES

To supplement the information I will also be providing PDFs of the calculations I have done for the various sections in the process. Due to the fact that the calculations are very long and tedious, I will not post the math here but refer to it from the PDFs. Since I have many calculations repeating the same thing for different gases, I may just only post a single example for each section to reduce document clutter. Pictures, models, and other data will be provided as I go along. While this is not meant to be a substitute in any way for proper research and reading of high vacuum engineering from academic literature, it is my hope that I may be able to provide useful information and help in these areas and add a contribution to this group that has already been such a major help to myself and others. A few additional notes to consider:

- This series of posts is meant as both documentation of my own efforts, as well as a walk-along for more intensive engineering design. I am by no means an expert in this field, and do not claim complete accuracy to data and math present. Inevitably, I will make mistakes along the way, as I am still also in the process of learning, so constructive criticism and correction is always welcome.

- The initial calculations presented are not meant to be hard set actual numbers. Since there are so many variables present in high vacuum systems, these calculations can only provide a very rough guide to what may be expected, and provide a point of initial design comparison and order of magnitude expectations. They are meant as rough estimates and a guide, not exact figures.

- Many material properties and constants used, such as outgassing rates, will vary between sources. Data I have presented here are general values acquired from a variety of sources, and will differ depending on a wide number of factors due to the complex nature of vacuum systems in general.

- This intensive of an approach to high vacuum system design is by no means necessary for any fusioneer to accomplish fusion. The math and modelling can be quite long and tedious, but it certainly do-able - the complexity comes from accounting for such a large amount of variables present in such designs. Most members here are looking to build simple fusors with whatever available resources, and a good, solid working fusor does not require any of these calculations. However, some calculations may be very useful to look at if you want to have a more fundamental grasp of what exactly is happening in the system and understanding how high vacuum systems work. Modelling, whether CAD, thermal, electrostatic, or other, can also provide a very powerful way of understanding your system, and there are so many free resources available now for various types of modelling it is much more accessible now to the determined hobbyist. Since I am not focusing on building a fusor, but various types of experimental setups, these calculations and approach has provided very useful for my own endeavors, as well as any general high vacuum system one may be working on.

Michael Bretti
Posts: 152
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Michael Bretti » Wed Jan 31, 2018 3:10 am

INTRODUCTION

PART II - BASIC FLOW OF CALCULATIONS

The above list, from Sections 2-6, deals with the bulk of calculations and examples from my system. These calculations are presented in the order which I followed in order to figure out the necessary key parameters of my system. I will not go over the basic fundamentals of high vacuum calculation - there are numerous texts and websites on this. However, I will show the progression of steps used to find each number.

Since I am planning to work with a wide range of test setups across a wide range of pressures, various gases need to be accounted for depending on the mode of operation. A quick breakdown includes the following:

1.) Pumpdown from atmosphere: 10^-2 Torr - 10^-7 Torr - molecular and transitional flows with water vapor loading
2.) Standard fusor operation without deuterium: 10^-2 Torr - transitional flow with air and water vapor loads
3.) Standard fusor operation with deuterium: 10^-2 Torr - transitional flow with deuterium
4.) Beam on Target and Beam Injected Fusor Systems for neutron production: <10^-4 Torr - molecular flow with deuterium
5.) Ion Beam systems with argon injection for non-neutron systems: <10^-4 Torr - molecular flow with argon
6.) Electron Gun Systems: <10^-6 Torr - molecular flow with water vapor loading
7.) Micro-thrusters: <10^-6 Torr - molecular flow with water vapor loading and argon
8.) Plasma Sources: <10^-3 Torr - molecular and transitional flows for deuterium, argon, and water vapor loading

As a quick summary, molecular flow governs the flow of gases in high vacuum systems at pressures of around 10^-4 Torr and lower. This is determined due to the fact that mean free path of molecules is large enough that molecules and residual gases interact with the walls of the system more than each other in the space between. Other factors can be used to calculate this region based on the system, which will be presented later. Transitional flow governs the flow of gases in a vacuum system in a region between molecular flow and roughing, usually between 10^-2 Torr and 10^-3 Torr. Due to the wide range of operating parameters, both regimes need to be calculated for my system, using the various gases present in the system. At vacuum levels from low vacuum to about 10^-7 Torr, the dominant gas load due to outgassing is water vapor. Since my systems will not be operating yet in the ultra-high vacuum regime at levels of 10^-8 and lower, which is dominated by the outgassing of hydrogen sorbed in the metals, calculation in this area is not immediate. Since deuterium is already hydrogen, and the molecular masses are almost identical between the two, any calculations used for deuterium would be reasonable for hydrogen in the ultra-high vacuum regime for estimates as well.

The fundamental and most important relationship in high vacuum systems is determined in the equation S=Q/P, where S is the speed in L/s, Q is the gas load in Torr*L/s, and P is the pressure in Torr. By finding the effective speed of the system, ultimate pressure and gas loads for various process can be derived. Speed is determined by a large range of factors, including conductance, which is also a critical factor in high vacuum systems. Conductance itself is dependent on several factors, such as geometry, gas used, and temperature. For all processes calculated in my systems, a temperature of 20C is assumed, which is the standard temperature used in literature for calculations and comparisons, usually using nitrogen or air.

My ultimate goal for these calculations can be boiled down into the following:

a.) Calculate the effective speed of my system for various gases based on processes I will use for molecular flow
b.) Calculate the effective speed of my system for various gases based on processes I will use for transitional flow (derived from molecular flow)
c.) Calculate the gas load due to outgassing (derived from materials and pumping conditions)
d.) Calculate the theoretical ultimate pressure of the system during pumpdown (derived from a and c)
e.) Calculate maximum allowable gas loads for various experiments (derived from a, c, and d)
f.) Calculate pumpdown times from atmosphere to ultimate vacuum (derived from all of the above)

Due to the amount of calculations involved, initially only conductance and pumping speeds were determined for my initial designs. Once a design was selected, the rest of the parameters could be calculated for that final design. Based on all of the above information, I would be able to know roughly how my system should behave for each test setup, assuming the chamber and system is well prepared, sealed, and functioning prior to the experiment. Again these are very rough estimates in ideal scenarios, but should be agreeable with basic expectations for vacuum systems.

Michael Bretti
Posts: 152
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Michael Bretti » Wed Jan 31, 2018 3:46 am

1.) INITIAL HIGH VACUUM SYSTEM CONCEPT DESIGN, REQUIREMENTS, AND MODELLING

The current final version of my system is the result of numerous design iterations, based on a wide variety of parameters, including: costs and available funding, availability of components, experimental goals, CAD modelling, and initial calculations involving system conductances and effective speeds for various gases. When I first decided to tackle this endeavor, I had originally purchased a quite massive chamber. However, over time I realized it would be much more beneficial and safer to start as small as possible, get a solid system running, and then gradually increment up to larger systems. Because of this, I eventually decided upon a topology that is built around 2.75” Conflat hardware, which is ubiquitous and relatively cheap on eBay. Despite its small size, 2.75" CF hardware can be plenty of room for many types of systems - the current 60MeV L-band electron beam LINAC at the research facility that I work travels through 2.75" CF pipeways, so a seemingly small diameter system can go a long way!

My initial goals from my larger system stayed the same however – I wanted to work on a wide variety of systems that interested me. These systems include: electron guns, ion guns, plasma sources, electric space propulsion, beam on target systems, and the fusor. I therefore needed to design a system capable of supporting all of these experiments, while being modular and low cost to build. This requires that I not only be able to operate over a wide range of pressures, but also utilize different gases, and have enough input ports for both adequate instrumentation as well as being able to accommodate the systems I wanted to test.

Prior to obtaining the hardware required for this build, I initially had acquired an excellent diffusion pump which would serve as the backbone to the high vacuum pumping of the system. Based on the above, I came up with an outline of key requirements needed for the system:

1.) Cost
2.) Small size, portable, and manageable to move
3.) Ability to operate from low to high vacuum, in the range of 10^-2 Torr to greater than 10^-7 Torr
4.) Modularity and expandability to support wide range of experiments
5.) Direct pumping line to maximize pump capability in a small system
6.) Input port for high vacuum pumping and roughing
7.) At least 2 input ports for a range of high vacuum gauges capable of reading from 10^-2 torr to greater than 10^-7 Torr
8.) At least 3 input ports for experimental setups
9.) Ability to support electron guns, ion guns, plasma sources, standard fusor, beam on target system, and potentially micro-thrusters, in addition to required feedback such as faraday cups, beam profiling, etc.
10.) At least one viewport for visual feedback
11.) Ability to both isolate and throttle the main chamber from the high vacuum pump

Before I started purchasing parts, I decided to model my system with CAD software. I currently use Fusion360, which I highly recommend to anyone – it is free, and has a massive range of capabilities and is incredibly powerful with a not-too-steep learning curve like some other CAD packages. In addition, it turns out that several online vendor such as Kurt J. Lesker and Ideal Vacuum provide a large selection of free CAD models for vacuum components.

Below is a rendering of V1 of my small-scale multipurpose system:

2.75in Conflat Multipurpose High Vacuum System V1.jpg

Initially, my first design, V1, incorporated a rather odd topology, utilizing a 90 degree manual valve from the diffusion pump to the pipeline. The valve is also useful as a throttle control for the diffusion pump depending on the process I am running. An adapter plate, made of 1" aluminum, would be needed to go from the 5" inlet of the diffusion pump to the 2.75" CF flange on the valve. The pipeline further branches off to a 4-way 2.75” CF cross and KF25 cross to support instrumentation. The valve on the right is for roughing the system. Above the CF cross is the main chamber, consisting of a 2.75” 5-Way CF cross. The initial design was largely dictated by what was available on eBay. 90 degree valves are very cheap and easy to come by, and utilizing a mix of KF25 hardware with CF hardware could reduce the cost considering KF hardware is relatively inexpensive and easy to find. The 5-way cross was chosen due to its ability to support 3 inputs, as well as a viewport, and connection for pumping, and can be found at much lower prices than a 6-way cross, which would be preferable for functionality. The 4-way CF cross allowed me to have connections to both high vacuum and roughing lines, which were initially separated for the system.

Although I could satisfy instrumentation and system input requirements, the setup seemed to be a bit awkward physically, and due to the pipeline and 90 degree bends in it, conductance and pumping speed would suffer. Since 2.75" CF hardware already would have low limits in pumping speed in molecular flow, I did not want to risk scrapping further speed. The lower the speed, the less gas I can support for certain systems that require operation at low vacuum. I scrapped the design in favor of further designs before I got to calculations for the system, as well as the availability of new parts on eBay to improve the design. After many more hours of searching on eBay, I was able to find some inline valves that made me redesign the system and come up with V2. This will be detailed in the next section, along with the associated calculations for V2.

User avatar
Dennis P Brown
Posts: 1544
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Dennis P Brown » Wed Jan 31, 2018 10:41 am

A minor point I'd like to mention. Do not use torr in calculations or for units at all in a technical approach. Use only SI units. One can convert later but all formal work must use SI. If you are writing for just us here, torr is ok but isn't appropriate for a rigorous approach.

Jerry Biehler
Posts: 831
Joined: Tue Nov 24, 2009 8:08 am
Real name:
Location: Beaverton, OR

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Jerry Biehler » Wed Jan 31, 2018 11:53 am

A lot of current papers and books still use torr so you can really use what you want. I just looked through about 10 research papers on my drive and all of them but one uses torr, the other uses Pa.

Michael Bretti
Posts: 152
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Michael Bretti » Wed Jan 31, 2018 1:27 pm

Thank you for your concern, however, as Jerry Biehler noted, Torr would be more than acceptable. Every source I have looked at so far for high vacuum engineering, both textbook as well as online, except for a few, refer to Torr for pressure in calculations. Another thing to consider is that most data presented on things like outgassing rates, which become crucial for later calculations, are almost always represented by the units Torr*L/s per cm^2, where a couple I have seen represented using mbar and inches. At this point I would find it very odd to represent any calculations either here or for my own personal work in the SI unit Pascal for vacuum, since most numbers and data are still referred to in Torr in this field. The equations are also independent of any particular unit, so if one wanted to use Pa they still could - however it seems largely inefficient and kind of a moot point when it would probably have to be converted back to Torr for data and comparison to literature anyway. Pressure measurement for high vacuum gauges are also represented and sold usually rated either in Torr or mbar, so when working with vacuum systems, calculating in Torr conceptually would make more sense since it is still the most commonly referred to unit still. All of the calculations I will present here are in Torr, and going forward Torr will be the default measurement for pressure for reference here.

User avatar
Richard Hull
Site Admin
Posts: 10946
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Richard Hull » Wed Jan 31, 2018 7:01 pm

I have given my thoughts in the past regarding pressure units solely for use by fusioneers using the fusor. We are not highly technical or the types to write papers for Nature. We are forced into two worlds in our quest for vacuum. One of technical vacuums and the other the scientific vacuum.

For our purposes and for the sake of most reading, we might encounter, the micron and the torr. These are all we need concern ourselves with.

All fusors run solely in a technical vacuum.....Above 1 micron. All fusion pressures can be expressed in microns. Why?.... The cheapest and most used gauges found surplus are TC, (thermocouple gauges). Virtually all of these are in microns. Thus, we work, do fusion and talk mostly in microns in a technical vacuum.

However, most of us strive with diffusion and turbo pumps to achieve some sort of scientific vacuum level in our chambers before introducing the fusion fuel, (deuterium). As this is the case, we might resort to the time honored Torr found in the bulk of classic literature on vacuum. Thus, we might speak of achieving 10e-5 or 10e-6 torr as a base pressure for those advanced enough to speak in scientific notation using torr-speak. This is proper when discussing deeper vacuums than fusor operational pressures.

The beauty of the micron is that when operating a fusor we can use whole numbers. (5, 12, 25 microns). We have an intimate grasp of these pressures due to the base level of most fore-pumps and inexpensive vacuum gauges which we deal with every day. We know that we must be well below 50 microns before starting our diffusion or Turbo pumps and that regardless of where they may take us in the "torr" range, we must add deuterium to at least 5 microns before attempting fusion.

Likewise, there is little sense in using .03 microns. If we are going to use fractional units, use torr as the whole point of using microns was to escape fractional units in fusion operating pressures, preferring whole number units.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Michael Bretti
Posts: 152
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Michael Bretti » Wed Jan 31, 2018 7:42 pm

Thank you for your input. For the sake of consistency in my calculations and notes, everything here that I post will be referred to in terms of Torr, for both low and high vacuum. Conversion to micron from Torr is trivial, and easily convertible between the two.

The goal here again is to describe an example of how one might approach more intensive calculation and design of a high vacuum system in general, and is not meant necessarily solely for fusors. However, all fusors require a good high vacuum system, so it is directly applicable to any fusor. The range for fusors is included in my current scope of requirements, but is not as much of a priority as establishing high vacuum working. For example, one important outcome specifically for fusors that can be gleaned from this could be how one would approach estimating the max gas flow rate of deuterium into the system given the vacuum chamber design and pumping parameters at a given vacuum level. This can be useful in qualifying a design before a lot of time and money is spent, and provides a good understanding of why things work the way they do in terms of vacuum systems.

Again, this is certainly not necessary for anyone working on a fusor, but since I have a lot of time before I build my system, and have an insatiable thirst for understanding the deeper principles of these technologies (not to mention I have found these pursuits not only highly educational and a great way to spend time while waiting for parts, but quite fun as well), I present my information here for those who might find it useful. I also like knowing that I can fully characterize my system and understand what is going on at all levels of operation, and have some idea of what to expect during my experiments.

I am also interested in seeing how close real operation is from theory for the systems I build. I won't be able to afford to run deuterium or other gas injected plasma systems for quite a while, however I will be able to run experiments on the vacuum itself and qualify the system at the vacuum engineering level. I think it could be very interesting to see how the system behaves and focus just on vacuum for now instead of the application used in the vacuum, though I probably would like to at least build an qualify some electron gun designs since they only require high vacuum and no gas input.

User avatar
Dennis P Brown
Posts: 1544
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Dennis P Brown » Wed Jan 31, 2018 10:13 pm

When you say
I am also interested in seeing how close real operation is from theory for the systems I build. ... however I will be able to run experiments on the vacuum itself and qualify the system at the vacuum engineering level.
You do realize the how and whether you clean these used components will affect issues for high vacuum a great deal and needs to be documented - one also must be careful about any contact with ones hands in assembly after cleaning - as such, gloves are critical so as not to cloud your results. Also, issues of bake out/temperature and the time used all matter and these are not easy to do calculations upon since the state of your stating systems (if not new) will be an unknown. Further, you fail to indicate that you will track room humidity so you can include that issue when the system is assembled or ever opened to the atmosphere - I've found that essential for any high vacuum work; even with purged systems.

Strangely, you have said nothing about the fore line system, and whether and how you will deal with back flow from that pump. Again, necessary for others to follow and benefit from your work.

When you say
greater than 10^-7 Torr
relative to its upper limits that will not be attainable unless you have a cooled trap for your DP. That adds further to issues of design and operation of your simple/low cost system relative to your calculations.

Michael Bretti
Posts: 152
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: High Vacuum Engineering Design, Analysis, and Build of a Small-Scale Multipurpose System

Post by Michael Bretti » Wed Jan 31, 2018 10:37 pm

Dennis P Brown,

I appreciate your comments regarding this information. Note however that I have only posted about the introduction and very first stages of the design process I started months ago - everything has evolved since then and been accounted for. Everything that you mentioned will be addressed as I go - preparation, cleaning, and baking of the system will not happen for several months, and will be well documented when the time comes to it. This has already all been weighed in and accounted for.

The foreline system is not mentioned yet because it is not necessary in order to perform the initial calculations at the high vacuum level. The foreline pump has already been properly sized to deal with the system. Although the diffusion pump is rather large, it is severely choked and the gas loads present should not be an issue for the roughing pump. Eventually, I would like to switch to a series diff pumped system, with one diffusion pump backing the main diffusion pump (backed by the foreline pump) to attain higher vacuum levels, among other modifications to allow for operation with Viton o-rings theoretically up to the 10^-9 torr level (yes, viton can be operated to this level with proper implementation). These will be addressed in following sections. Everything I listed out follows a logical progression for the approach I took. Again, this is an example of how one might go about it, and not the only answer. As stated above, these numbers are not hard numbers, but guidelines of what roughly to expect. I am well aware that the amount of variables in real life are far too many to account for accurately, but rough estimates can still be generalized.

The calculations, as already mentioned above in the introduction, also assume that the system is already well prepared, cleaned, baked, etc - they will not be indicative of initial pumpdown efforts, but rather steady state operation when a quality vacuum is achieved. They are calculated for a rather ideal case, but are weighted with correction factors for the worst case scenarios of these ideal cases.

For my upper limits, I do in fact have a cooled baffle. Under a well prepared system in my current configuration given all of the necessary parameters once steady state is achieved, I should in fact be able to achieve in the 10^-7 torr range.

Another reason for data logging the pump down and rate of rise curves will be precisely to identify patterns in pumpdown, including humidity, time pumped, baking, etc. Ultimately, I do have my system set up so that I will be able to seal off the main chamber and continuously bake and pump it with an ion pump, though this will not happen for quite some time.

Thank you again for your comments. I do encourage this to be an open discussion build along, and I know that this community is very scientifically critical in its approach, which is of benefit to everyone.

Post Reply