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: 148
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 » Mon Jun 11, 2018 2:48 am

A few updates on the cooling system, control system, and overall direction of this project.

For the cooling system, the secondary tank capacity was increased a bit from the initial design. I had originally spec'd a 3"x10" pvc tank. However, since I was already using 4" pvc for the main tank and had some leftover, I decided to just increase the secondary tank capacity to a 4"x10" tank which takes up negligible extra space overall. Fill ports were also added to each of the tanks. Second, since I have purchased the main heat exchangers as sets of 2 for a great discount on eBay, I ended up with an extra leftover from my previous purchase. I will be adding this to the design to bring the primary cooling loop capacity up from 1200W to 1700W, which would allow me to run this cooling system now with either of my diffusion pumps. The contingency for this upgrade was already built in, so I figured I would add this now for minimal additional cost.

For the control system, as mentioned above, I will be going with an Arduino Mega microcontroller. Since I don't want to spend a massive amount of time creating my own fully custom GUI from scratch, I decided to look into several available solutions. The ones I initially researched and considered include the following:

1.) Azande - https://zeijlonsystems.se/products/azande/index.html
2.) MegunoLink - https://www.megunolink.com/?nabe=6011995085340672:1
3.) Instrumentino- http://www.chemie.unibas.ch/~hauser/ope ... index.html
4.) Makerplot- http://www.makerplot.com/

After a lot of looking into each of the options, researching their capabilities, evaluating my system requirements, and weighing their strengths and weaknesses, I have decided that I will most likely being going with MegunoLink. Azande and Instrumentino are both free, while MegunoLink and Makerplot go for about $40 each. I like both MegunoLink and Makerplot, and feel they both have a lot of powerful and robust capabilities to offer for monitoring, data logging, and control with the Arduino. However, based on my system, I feel that MegunoLink would be the best option for my needs. Now that I have the basic overview of the initial inputs and outputs for my control system, I will start coding up the controller, and probably download the free trial version first and experiment with it before I pull the trigger. The initial system inputs for monitoring and plotting will include the four thermocouple sensors for the cooling loop monitoring, the flow sensor, as well as a digital temp/humidity sensor for ambient monitoring, the roughing side thermocouple gauge, and the high-vacuum side wide range transducer. Outputs will need to control pwm for speed control of the three cooling pumps, as well as a 16-channel relay board for turning on and off the numerous heat exchanger cooling fans, water pumps, peltier modules, and interlocks for the diffusion pump main power, and other power and electronics needed for the system. Semi-automated to full automation of pumpdown is still expected, as well as real-time monitoring of all system and environmental parameters, in addition to data-logging system conditions each run and pumpdown to build up a log of run data to better control and understand the system.

In terms of the overall experimental direction for this build, I have decided on the first experiment and system that I will be concentrating on, and the main focus of my research going forward. This build will initially support a 300kV, 30MW peak-power nanosecond-pulsed electron gun for intense relativistic pulsed electron gun injector development, initially exploring scattering and transmission studies in open-air. The preliminary system cost analysis, gun design, Faraday cup, and extraction window are already in the initial design phase. This build will be used to support a much more high-power and ambitious system build in the future, and is a moderate stepping stone towards my ultimate goal. In addition to research into this rather unique and highly-specialized subset of the highest peak-power class of electron gun injectors in use for physics research (which usually reach in the many 10s of GW range for peak power), this build can also be used as the first starting point to support intense ion-beam development, which opens up the door for a lot of exciting, more high-energy experiments than normally done at the hobbyist level. However, electron beam work will be the primary focus, since I will not have to worry about the cost of gas handling subsystems for ion injectors.

Michael Bretti
Posts: 148
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 » Sun Jun 17, 2018 9:59 pm

Quick update on the cooling system build. The primary and secondary water tanks have been completed and assembled. Attached to the tanks are the thermocouple gauges, as well as fill ports at the top, and brass hose fittings for the cooling lines. A couple of additional ports were added as mentioned prior for future cooling capacity upgrades, which are currently capped off. No pvc glue was needed either for the pipe to end-cap fitting - it turns out that the standard 4" pvc to 4" clay pipe rubber adapters with hose clamps works perfectly between the pipe and cap interface, and provides a leak-tight seal that also allows me to completely disassemble the tanks if I ever need to get inside them in the future. Now that these are done, I can start mounting them to the MDF baseplate. Once these are on, then the rest of the cooling system mounting is pretty easy. The tanks have already gone through preliminary leak checking, and should be good to go.

Cooling Tank Pic 1.jpg

Cooling Tank Pic 3.jpg

Michael Bretti
Posts: 148
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 » Sat Jun 23, 2018 5:18 am

Another update on the cooling system. In my initial designs, I forgot a key parameter that has forced me to redesign the whole system from scratch. While mechanically and thermally the system was sound, chemically it was not. In my cooling loops, I had brass, copper, stainless, and aluminum. These metals in combination would induce galvanic corrosion over time. There are coolants that can be used for combination metals in loops, specifically inhibited glycol coolants. However, it also turns out that glycol, both ethylene and propylene, are not compatible with pvc, which my storage tanks are made from, and will eventually degrade pvc on contact over time.

However, it turns out that this initial issue, and subsequent redesign, has been the best thing to happen to the cooling system yet. Upon re-evaluating my options, and re-configuring the system, for only a little bit more in expenditures, I could radically improve and upgrade performance of the whole system in all aspects, and selected components that would make it much easier to build. The pvc cooling tanks were replaced with larger capacity hdpe tanks with integral mounting flanges. This increased the tanks capacity by 2.5x in a smaller amount of space. These tanks were purchased for $8, and are very robust. The primary and secondary loops, which share coolant, have only copper/brass components, and the third loop, for cooling the peltier hot side, will reuse all the aluminum cooling parts I had bought previously, and since it is an isolated loop, will not have corrosion issues. To replace the aluminum heat exchanger on the primary side from the initial design, I got a 7kW rated copper heat exchanger used for furnaces from ebay for only $30. The secondary loop, consisting of the peltier chiller block, will use an array of high quality solid copper heat-sinks scavenged from disposed servers, with coolant flowed between the fins. 8 peltiers will be used for this chiller block. Solid copper heat sinks with integral mounted fans also scavenged from servers will help cool the hot side aluminum blocks as well, three for each block. The aluminum cooling blocks will be doubled up on the hot-side as well, with an extra heat exchanger running in parallel. Two 1750 CFM performance radiator fans for cars were purchased from ebay for $25 each for cooling the main heat exchangers. In all, the heat capacity specs for each of the loops are as follows:

1.) Primary Loop (diff pump main heat exchanger) - 7kW
2.) Secondary Loop (peltier chiller block) - 300-400W
3.) Tertiary Loop (peliter hot-side cooling) - 1kW

I have most of the components needed to actually start building the cooling system, and will begin construction shortly. I will post pictures and further updates as I progress. This system should now be far more than enough to handle any heat load needed for any future experiments, and should last me indefinitely. Once this is complete, I will qualify it with my thermal imaging system, then proceed to writing the code for the automated control and monitoring system.

Michael Bretti
Posts: 148
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 » Sun Jul 15, 2018 8:12 pm

Quick update on the cooling system build. Below is a picture of the V2 system so far:
Cooling System Update - Front.jpg
Cooling System Update - Back.jpg
I currently have all of the hardware needed to complete the system, it is just a matter of finishing up some of the custom components. All brass hardware that was originally to be used has been replaced with much lower cost nylon components, which has more than wide enough temperature range as well as chemical compatibility for this build. In parallel with this build, I have also been working on rebuilding the thermocouple sensors I purchased for cheap on ebay, making them both waterproof and corrosion resistant with new 316 SS casings, and mounting them to the custom adapters needed for placing them in the system, as well as qualifying and calibrating their response. It turns out that the original sensors I bought were not actually SS, and a quick immersion test in water revealed rusting within hours, necessitating a rebuild of the sensors.

I am also about half way done with the upgraded peltier chiller module, which still requires custom machining some nylon manifold blocks for directing flow through the chiller, as well as a couple of other nylon manifolds needed for directing flow to the tertiary loop heat exchangers. Once the rest of these custom components are built and mounted, I can proceed to routing all of the cooling lines and bolting the assembly to the 8020 table. All three of these builds, (the thermocouple upgrade, the peltier chiller, and the completed cooling system) will have full and complete documentation available including cost analysis, components supply, specifications, CAD drawings, testing and calibration data, and full build walkthroughs. I will release links to the pages with this information as it is completed. I expect to finally complete the entire cooling system and have it qualified and tested with my thermal imaging camera within the next few weeks.

In parallel with these efforts I have been working on the pulsed power driver for the first intense e-beam system. The small multipurpose V4 vacuum system will implement a small, lower power beam to qualify intense beam transport through gases, as well as the first stage of the pulse power unit. Details of this build will follow after the cooling system is complete. For this system, I have obtained a large supply of large 28kV, 0.02uF pulse capacitors used for the primary stage of the driver. So far I have tested up to 2 capacitors in parallel into a dummy primary pulse transformer load, achieving a current of 1164A @ 23kV charging voltage, for a total peak pulse power of 26.77MW in about 1uS FWHM. I plan on using 8 parallel capacitors for the small system, and 16 for the large system I am working on. For the small gun, which will only be driven by the first slow stage of the pulse power driver, I expect input power into the gun in excess of 100MW, with gun efficiency of several 10s of percent. With additional pulse compression I expect to easily achieve much higher peak powers. Once the small gun has been qualified and running with the primary pulse power stage, then construction on the large gun will begin with the full pulse power unit.

Michael Bretti
Posts: 148
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 Aug 08, 2018 10:17 pm

A big update on the vacuum efforts so far. The cooling system is finally assembled, mounted to the test table, and all of the cooling lines routed. It has been a challenging and demanding design and build process over the past couple of months. Now that it is assembled, the next phases of writing the control code, wiring the electronics, and qualifying the system can begin.

The system should be able to support both of my diffusion pumps - my 850W EO4, and my 1450W NRC. My three major vacuum setups - the small-scale V4 system, the accelerator build, and the micropropulsion test chamber will all share the same infrastructure, and are designed with adapter plates that will allow me to rapidly switch between test setups with ease, reducing the overall cost of having multiple systems to one modular, inclusive test stand. I will post details of the other builds in other forum posts. I also have full documentation available for all of my builds, including engineering specs, datasheets, CAD drawings, CAD files, cost analysis, and build pictures. I have a lot of documentation left to organize for the full cooling system build, and will release the links when complete. For now here are the links to the peltier chiller module build and the thermocouple build:

http://appliedionsystems.com/portfolio/ ... er-module/

http://appliedionsystems.com/peltier-ch ... -pictures/

http://appliedionsystems.com/portfolio/ ... d-upgrade/

http://appliedionsystems.com/low-cost-t ... -pictures/

Here are a couple of pictures of the cooling system. Thermal imaging data to follow:

Completed Peltier Chiller Cooling System 1.jpg
Completed Peltier Chiller Cooling System 2.jpg

Michael Bretti
Posts: 148
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 » Tue Aug 14, 2018 2:52 am

Full technical specifications, CAD files, cost analysis, build pictures, and a detailed overview comparison article between the iterations of the cooling systems are now available for the peltier-based chiller for diffusion pump cooling system:

http://appliedionsystems.com/portfolio/ ... p-cooling/

http://appliedionsystems.com/closed-loo ... -pictures/

http://appliedionsystems.com/learning-f ... -redesign/

Michael Bretti
Posts: 148
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 » Mon Aug 27, 2018 12:50 am

New update on the cooling system build. I finally completed the first round of preliminary testing this week, and have analyzed the data and calculated all the major thermal parameters of the system. So far, it appears that it is performing grossly under spec - however, I have traced the problem back to the heat exchangers, which both use the same fan. It appears that the fans do not push nearly enough air through the heat exchangers, greatly reducing their ability to dissipate heat. I am in the process of replacing these with high static pressure fans that should fix this issue. Fortunately, since my system is so heavily instrumented (now 6 thermocouples, 3 flow sensors, and a general ambient temp/humidity sensor), debugging and analyzing system performance is quite easy. I also downloaded my free trial of MegunoLink to start seeing if it is a viable option for control and data logging. So far I really like the software, and was able to set up real time plotting and data logging in a only a couple minutes with my Arduino Mega, which was used for collecting the data for the test. I will be working on building the entire control system, which will be housed in a large computer casing mounted to the side of the 8020 test stand, and includes all of the power supplies, circuitry, and instrumentation cards for the high vacuum and cooling systems. Full details of the first round of testing and data analysis can be found here:

http://appliedionsystems.com/preliminar ... ng-system/

After the testing was complete, I also made the discovery that my small water cooled baffle that I was originally going to use for the EO4 diff pump is made from plated steel, not stainless steel, and had heavily rusted inside. As a result, I will not be using that baffle anymore, and will instead switch to using my larger 8" water cooled baffle, which I was originally saving for the NRC 0183 diff pump. This requires designing new adapter plates, which I will make so that both diff pumps can be used with the baffle, as well as all of my test chambers. While it does make the system even more modular as a whole now, it does add an extra expense that will delay pulling vacuum for a while, but I still have plenty of work to do in terms of documentation, CAD, testing, and coding the automated control system.

Michael Bretti
Posts: 148
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 » Thu Sep 20, 2018 7:21 pm

Another quick update on the project progress so far, and the scope going forward. The cooling system is about complete and almost ready for final testing, and is currently being integrated with the new control system I am working on. The control system hardware is being assembled, and is in the process of being wired together. Due to the extremely high peak power pulses I will be generating for one of my systems, I have decided to keep the control box separate from the main test stand, which relaxes EMP shielding requirements on the box. The project itself has evolved and expanded immensely since I first started, and has been rapidly gaining much more sophistication than I even previously first planned.

For the control system, I have decided to fully embrace MegunoLink (https://www.megunolink.com/) for developing the user interface and control architecture on the Arduino Mega. I have been working with it more and more in the past week and a half, and have come to really like the program. I keep finding new features to implement, and it has proven itself to be a very powerful and capable piece of software, and has really allowed my control system to explode in sophistication and features. Currently I have full data acquisition complete and running, which allows me to graph and save all the data for my cooling system sensors, as well as the roughing gauge and high vacuum gauge in the chambers. I am now working on implementing all of the controls for manual mode operation, which will allow for individual control and adjustment for every component in each of my subsystems (cooling pumps, fans, peltiers, vacuum pumps, etc.), and will be moving on to the automated safety interlocks, and a full automation mode, which will allow my system to run through the complete cooling and vacuum pumpdown cycles automatically with the single press of a button.

The original V4 design has also evolved, with the introduction of the new 8" baffle and adapter plates to replace the original smaller baffle that has rusted. Since I started working on a couple of new larger scale vacuum systems that are now higher priority, the small-scale V4 system is also currently on hold for the foreseeable future. However, the two large systems will still be using the entire infrastructure I have built up so far, including the 8020 test stand, updated cooling system, roughing system, diffusion pump, and control system.

As a result, I will probably start a new forum topic on updates going forward, since the project has rapidly evolved beyond the initial scope presented at the beginning of this post series. The new topic will continue from here, looking at the current status of the fully integrated cooling, vacuum, and control system that will be used to support all of the chambers that I am working on. I will also be introducing the two larger systems I am working on that will be used with this test stand as well, although the full details for the projects will be beyond the scope of this forum, and include a tremendous amount of engineering effort that would be inefficient to try and post here. The full details and technical documentation however will be available on my website as I progress forward with them.

Michael Bretti
Posts: 148
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 » Tue Oct 16, 2018 2:21 am

As of today, the Small Scale Multipurpose V4 system design is obsolete, and I have moved to the newest V5 version update. This update is to adapt the system to the newly designed high vacuum pumping assembly and utilize the highly modular design of my pumping station to its fullest. This update allows the small 2.75" chamber to adapt to the 6" conflat port that I will be using for my propulsion chamber and large beam chamber. While I still won't be running any experiments with the V5 system for a while based on funding and project priority, I am still moving forward with planning and designing all of the experiments and finalizing them for when I can begin testing down the road. The new V5 system will be used to support researching scaled versions of the large, high power beam system I am working on, which I will post about more as I progress.

In addition to the V5 design, I have completed new thermal modeling simulations for the new baffle and diffusion pumping assembly, which I will create a new post for. I will also be starting molecular flow simulations for high vacuum operation with some new software that I found and am planning to explore further.

Below are the CAD render files of the new V5 design, as well as the V5 system attached to the upgraded diffusion pump assembly.

2,75in Conflat High Vacuum Chamber V5.png
2,75in Conflat High Vacuum Chamber V5 and Pumping Assembly.png

Post Reply