Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Current images of fusor efforts, components, etc. Try to continuously update from your name, a current photo using edit function. Title post with your name once only. Change image and text as needed. See first posting for details.
Michael Bretti
Posts: 170
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Michael Bretti » Thu May 09, 2019 3:40 am

After a year of intensive research, design, simulation, and building, I have finally run the Integrated High Vacuum Test Stand at high vacuum levels with all subsystems for the first time last week. Over the course of the year I have posted extensive walkthrough details regarding the design process, CAD, thermal modeling, and vacuum modeling of my system. All of the engineering details and specs are on my website, and I have been posting all build progress and details elsewhere on social media, but just wanted to give a brief update here of these efforts. The updated complete CAD model, and actual build can be seen below. I currently have the Micro Propulsion Testing Chamber mounted, and have already started running propulsion tests with advanced micro pulsed plasma thrusters:

Micro Propulsion Test Chamber Full System Final Assembly.jpg
Completed Integrated test Stand with Micro Propulsion Testing Chamber Resize Smaller.jpg

Not only has the system performed beyond expectation, but it has performed exactly to engineering spec, and shockingly close to my initial calculations and simulations. A brief summary for reference for the simulated and calculated numbers vs. the actual measured numbers:

COOLING SYSTEM: Simulated steady-state average baffle temp with 15C ambient and cooling with full diffusion pump load was 21C. Actual measured steady-state baffle temperature with 15-17C ambient conditions and peltier chiller module activated was 22C.

HIGH VACUUM SYSTEM: Hand calculated final pressure for ideal outgassing numbers for very short pumping times (~1 hr) was 3.8 x 10^-6 Torr. Simulated pressure with Molflow+ for ideal outgassing numbers for very short pumping times was 3 x 10^-6 Torr. Actual ultimate measured pressure was 6 x 10^-6 Torr after 3 hours for the first pumpdown. Outgassing rates were higher than ideal since all fresh materials and new oil was used, in addition to several more viton o-rings used than initial calculations and simulation.

ROUGHING SYSTEM: Estimated maximum foreline pressure goal was set at 3.0 x 10^-2 Torr under full diffusion pump load during pumpdown. Actual measured foreline pressure was stable at 2.8 x 10^-2 Torr.

Below is the full pumpdown curve from the very first run. I have since collected more pumpdown data, and pumping performance has noticeably improved with chamber pumping and conditioning.

Run 05-01-2019 - Micro Propulsion Testing Chamber Pumpdown.jpg

From cold start now, I can hit 10^-5 torr in less than an hour, and with a bit of chamber conditioning, reach the 10^-6 torr level in about an hour. Pumping speed appears to be very high as expected, and is probably on par with my simulated number of about 400 L/s at the chamber inlet, and can handle large outgassing and test loads with ease. The closed-loop peltier chilled cooling system performs excellently with the thermal load from the diffusion pump, and keeps the water/glycol coolant below my average target temperature of 25C, with stable values around 22C for my extended pumping times.

One important thing that I am doing for every test is logging and saving all of the data for every test run and pumpdown. Now that I can record system temperature, environmental temperature and humidity, and pumpdown curves, I can compare collected data from each run to qualify my system performance based on operational and environmental factors. This can be a very powerful tool for benchmarking and troubleshooting down the road, and allows me to understand every operational parameter and variation in my system to a very low level.

The entire design, from roughing, to cooling, to high vacuum has been successfully verified to spec, and is in extremely close agreement with my calculations and simulations. Pumpdown is very easy and fast, and has performed excellently so far. First time turning the system on, and it just worked, no major leaks for the levels I am operating at, no adjustments were needed. This has been a massively involved and challenging endeavor, built completely from scratch from the ground up, and has been the most challenging engineering project I have worked on yet. For anyone looking to build a high vacuum system, I would strongly encourage you to take note of this example of what can be accomplished with extensive planning, research, and design work put into your vacuum systems. Take your time, plan things out, do thorough research and design, and you will avoid significant frustration, and can have a system that works as desired.

On a final note, here is a shot of one of my pulsed plasma thusters in operation. I have had some major technical issues running the test, but was able to capture successful ignition and firing of the thruster. This marks the start of my work on advanced open-source electric propulsion, and may be one of the first, if not only independent home-built high vacuum research and testing efforts in this area at the maker level:

First Ignition Test - 05-07-2019.jpg

User avatar
Richard Hull
Moderator
Posts: 11246
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Richard Hull » Thu May 09, 2019 7:45 am

Great work and a fine report on your system and the pump down info. and report. Newbies might take note of what a real vacuum system looks like and should act like, if assembled well. Good luck on the propulsion tests.

I assume you will have a way of measuring the minuscule thrust, and have planned to obtain some mass flow data in future to measure the impulse?

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: 170
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Michael Bretti » Thu May 09, 2019 2:09 pm

Richard,

Thank you for your compliments. The thrusters I will be focusing on for now are solid Teflon fuel pulsed plasma thrusters. No gas inlet or flow rates to measure. The fuel is provided through ablation of the Teflon fuel first through an ignition arc, which raises the localized pressure between the main discharge electrodes to allow for the main capacitor bank to discharge. Ignition voltage is on the order of several thousand volts, main bank is usually around 1kV. Since this is a very small thruster aimed at PocketQube-class satellites, I will be working with small bank capacities (around 1uF range).

For measuring things like specific impulse and impulse bit for these thrusters, the common practice is to carefully weigh the Teflon fuel before and after a set number of shots, and calculate impulse from the average mass loss per shot. I don't have a high-precision laboratory scale as they are very expensive, but I will ask around people I know in the community to see if I could get some help on precision weight measurements.

In terms of measuring thrust, this will be challenging since small PPTs operate on the level of tens of micronewtons. One reference paper I have come across utilized a small pendulum with Kapton flap. When the thruster fires, the pendulum moves, and by knowing the mass and distance traveled, thrust can be back calculated. These tests will be further down the road. My first goal is to achieve repeatable ignition. Once this is achieved, I will start on lifetime testing to see how many shots the thruster can fire for a particular set of ignition and main bank voltages. This will be in stages - 10k shots, 50k shots, 100k shots, etc. Normally these operate at roughly 1 Hz, so 10k shots is about 3 hours of testing. However, I should be able to run it at a bit higher rep rate to reduce testing time. I will have other instrumentation set up for measurements as well going forward, such as Rogowski coils for capacitor discharge currents, and a Faraday cup for discharge exhaust measurements.

User avatar
Richard Hull
Moderator
Posts: 11246
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Richard Hull » Thu May 09, 2019 4:11 pm

When I was working with water arc experiments. I had to measure currents in kiloamps and spent a full kilobuck to get a custom, Pearson pulsed current transformer calibrated to 1 volt per kiloamp.

It is interesting that the teflon thruster is its own fuel. cool idea.

Way back in the beginning of this fusion effort there was much discussion about kiloampere pulsed fusors using hydrogen thyratrons, but nothing was ever done in the effort.

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: 170
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Michael Bretti » Thu May 09, 2019 5:23 pm

Yeah, the Teflon PPTs are pretty interesting in their operation, and are very promising for miniaturization from the standpoint of no pressurized gas, liquid metal sources, or any difficult fuel-handling needs, as well as very low power consumption requirements. Very robust and relatively low cost compared to other EP out there. They are also some of the oldest electric propulsion out there, and was the first ever successfully flown in space on the Zond II.

I will actually be making my own rogowski coils for another project I am working on, the pulsed accelerator, and using my smaller calibrated Pearson coils to check the volts/kA of the coils I make.

It's really a shame there isn't more effort on pulsed fusors. One of my areas of expertise is in pulsed power, and I have been actively encouraging other enthusiasts in the area to try exploring pulsed fusors. I do not have a fusor build currently planned in the foreseeable future, but if I was building a fusor, I would immediately go pulsed. So much untapped potential. For the pulsed accelerator I am working on, I am already halfway done with a quite large pulse power driver that consists of two stages of pulse compression to get the very high voltages and currents (hundreds of kV at tens of kA) I am after. I am actually using a hydrogen thyratron now for the trigger pulser of the thruster igniter, and will also be using it to trigger the accelerator as well down the road. It would already be ready as is now to run a fusor in pulsed mode at modest voltages and currents (10s of kV at several amps), but as is with the thruster and accelerator project I am just way too spread thin on resources to take on another major build.

Rex Allers
Posts: 305
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Rex Allers » Fri May 10, 2019 6:31 am

Michael, very professional parts in your chamber. Must have a few dollars invested in all those components.

I did notice that your roughing pump seems to be a Yellow Jacket SuperEvac normally used in HVAC. As Richard has commented elsewhere, not a bad pump. No need for fancy brands if it works.

I have a small question. Apologies if you covered this in earlier posts. You seem to have two big, thick aluminum squares between the diff pump and the chambers. I'm wondering why such big pieces of metal here. Does it have something to do with your thermal control?
Rex Allers

Patrick Lindecker
Posts: 32
Joined: Mon Apr 09, 2018 1:47 pm
Real name: Patrick Lindecker

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Patrick Lindecker » Fri May 10, 2019 9:37 pm

Hello Michael,

>It's really a shame there isn't more effort on pulsed fusors.

Speaking of pulse, do you know what are, at the moment, the approximative minimum limits of time during which a pulse can be controlled (microsec? nanosec?), and precisely:
* in case of a magnetic field pulse, what is the minimum time during which a current can be controlled (for example to pass from 20 mT to 500 mT),
* in case of an electric field pulse, the minimum time during which a voltage can be controlled (for example to pass from 1 KV to 20 KV).

Web links are welcome. Thanks.

Patrick Lindecker

Michael Bretti
Posts: 170
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Michael Bretti » Sat May 11, 2019 12:48 am

Rex,

The project did cost a good amount, however no more total than any other well-designed fusor system (few thousand dollars total). I ended up spending a significant amount of time waiting and searching for the best deals I could find for parts, and designing my system around what was available. I ended up getting several big ticket items, such as the diffusion pump, and half of the propulsion chamber, along with several feedthroughs, for free.

I ended up going with the 6 CFM Yellow Jacket based on Richard's early recommendations and experience in the beginning, weighing out all of my options and the best cost going forward. I am very happy with its performance so far. While it can't get to single digit millitorr levels like a scientific grade foreline pump, it is more than enough to back my diffusion pump, with plenty of overhead, and reaches sufficient vacuum levels for what I need. After a few pumpdowns and system conditioning, it can get the whole system down to the upper 20 millitorr range in a few minutes. For those looking to just use a refrigeration pump instead of a scientific pump due to availability and cost, I would definitely recommend this one.

The thick aluminum plates between the diffusion pump and chamber serves multiple functions. They are firstly adapter plates to allow me to go from the diffusion pump, to the baffle, to the chamber. They are 1" thick to allow for plenty of thread engagement for mounting the pump and chamber, especially for the larger bolts needed to hold the pump on the bottom. The baffle in between the two plates is compressed with the large outer bolts surrounding it, and I wanted the plates to be thick enough so there wouldn't be deflection. The plates also serve as mounts to the 8020 table, and I wanted them to be thick enough to support all the weight of everything mounted to it. By combining multiple functionality into these parts, I could save cost as well as space. While it wasn't originally part of my plan, the large thermal mass of the plates also helps for thermal dissipation from heat rising up from the diffusion pump.


Patrick,

In terms of minimum limits of time for pulse control, it highly depends on the system and application. It is really application specific, and very dependent on the end parameters you want to achieve and the technology you are using. For example, for electric field pulses, I have studied systems that can create pulses of hundreds of kV in only a couple of hundred of picoseconds. These are GW-class, ultrafast pulsers used for intense UWB microwave generation. These require state of the art high pressure ultrafast spark gap switching. Theoretical lower limits for spark gap switching is on the range of about 10 ps (never practically achieved though), and can handle the highest voltage and currents that can be generated (used anywhere from kW to TW systems at MV, MA levels). For semiconductor switches, there are obviously technologies that can switch much faster than this, but on the level of volts, and nowhere remotely near the peak power levels seen in pulsed power technology (with the exception of specialized pulsed power semiconductors such as DSRDs). Tube based switches such as thyratrons switch in the microsecond range, and can handle tens of kV and thousands of amps for the biggest tubes. Then there are some in-between technologies such as pseudo-spark gap switches, magnetic switches (not for magnetic fields, but voltage pulse compression), and photoconductive semiconductor switches that cover a wide range of switching parameters. Massively important factors include type of load, rise time, pulse width, pulse repetition rate, voltage, current, etc. These factors will determine the technology required, and is again very application dependent. For 1 to 20kV, ultrafast switching is relatively trivial, and the range can be anywhere from hundreds of picoseconds to microseconds for standard pulsed power technologies.

For magnetic fields, it again highly depends on your setup. Things like inductance plays a massive role in speed when dealing with high speed, high power systems. There are certainly quite fast-changing pulsed magnetic field systems, but these are not your average electromagnets, nor will they reach the speeds for pulsed electric field systems. Typically pulsed magnetics are used in ultra-high magnetic field applications, and are on the range of milliseconds. Explosive magnetics and other non-destructive magnets are used to generate fields in the tens to hundreds of Tesla range for brief periods. However, for such incredibly low magnetic fields less than 1 Tesla like you are interested in, you could probably switch much faster, in the microsecond range, but generally pulsed magnets are not seen for such low field strengths.

In terms of applications specifically for pulsed fusors, you would not want to operate the fusor in the super-fast pulse regime. From the few papers I have seen on pulsed fusors, pulse-widths are on the order of tens to hundreds of microseconds, which is incredibly slow in terms of pulsed power. These systems mainly utilize a simple PFN switched by a hydrogen thyratron. Voltage can easily be stepped up to over 100kV without any difficulty with a small pulse transformer, and can deliver hundreds of amps or more. Far higher levels than any DC x-ray transformer based supply, and could be built very easily, low cost, and quite compact, and deliver orders of magnitude higher average neutron yields for similar or less input power. For reference, the system I am building for my pulsed accelerator will generate a pulse of up to 300kV, 30kA in about 20-30 nanoseconds (assuming everything works as designed.) I do not believe such a fast pulse would be desirable for optimally driving a fusor.

Patrick Lindecker
Posts: 32
Joined: Mon Apr 09, 2018 1:47 pm
Real name: Patrick Lindecker

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Patrick Lindecker » Sat May 11, 2019 5:36 pm

Hello Michael,

Thanks for your complete answer. It's not really for a Fusor but I study (by simulation) the following problem: when you inject electrons (to get a negative potential with the electrons cloud) in a set of magnetic mirrors (to avoid the "cusps" of the "biconic cusps" devices), if the electrons are confined with a static magnetic field (or a static electric field), there is noway to avoid the electrons to come back to their initial location and collide the injector, as you can see below (the total energy being conserved in a supposed isentropic evolution).
Sans_titre.jpg
Sans_titre.jpg (23.74 KiB) Viewed 651 times
So one solution is to increase the field (magnetic in this case), as for an adiabatic compression. At the moment, I found that the pulse would need to be of 30 ns long and might make pass the field from 100 to 500 mT.

But 30 ns is below the limit that you propose, i.e 1 microsec for a magnetic pulse, so I would have to see other solution ☺

Congratulations for your device (even if I wonder how it works exactly).

Patrick Lindecker

Michael Bretti
Posts: 170
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: Successful Completion and Operation of the Integrated High Vacuum Test Stand and Micro Propulsion Testing Chamber

Post by Michael Bretti » Sat May 11, 2019 6:10 pm

Patrick,

Certainly a very interesting problem. In terms of the lower limit for pulse widths for magnetic fields, I would not take the number I gave as a lower limit by any means, I just gave an estimate based off things I knew directly off the top of my head for more conventional high power pulsed systems I have studied, however it could certainly be possible to do what you need at your time scales. For such low magnetic field strengths, I think it could certainly be feasible in the 10s of ns range. Pulsed magnetics is not really an area I have covered as extensively as other pulsed technologies. Certainly though there are devices, for example, magnetically insulated transmission lines, that rely on self-induced magnetic fields to work, and operate on the nanosecond time scale. I would certainly recommend to look deeper into nanosecond pulsed magnetic fields, I am sure such areas of research exist. I will go through my resources and see if I can find anything specifically for your time scales and field strengths.

Post Reply