Fusion Reactor Progress- Kuba and Noah

[Kuba Anglin]

Hello,

I wanted to post an update on my brother's and my progress on our reactor. Nearly all the parts have arrived and we have completed building the reactor. All flanges are connected using Viton gaskets and Apiezon type M vacuum grease. We spent a good hour making sure every screw was tight and every flange was flush with its connecting flange. Lowering the 50+ pound 8" gate valve with its bottom adapter onto the diffusion pump inside the wooden structure proved troublesome, but we managed to do it. We recently acquired a 50L lecture bottle of deuterium with a regulator as well as some capillary tubing and some more fittings. We also acquired a -40KV 5mA Spellman power supply from eBay and with some help from Cliff at Spellman, were able to modify it for operation at 10mA.

Specifications:

Mechanical pump: 12 CFM, 15 micron, 1 HP
Diffusion pump: Varian VHS-4
Gate valve: 8" CF manual
Vacuum chamber: 6" CF 6-way cross with 6.25" spherical body
Pressure gauge: Hastings DV-6M
TC contoller: DV-6 model CVT-16
HV feedthough: -45KV 3A 4.5" CF
Grid: Tungsten, 1.25" sphere
Power supply: Spellman PTV -40KV 5mA modified for 10mA operation
Neutron detection: Bubble detector (not yet acquired), silver or indium activation
Deuterium: 50L 99.999% purity

Here are some pictures:







Yesterday, we tried pumping the chamber down with just the mechanical pump. The dial on the TC gauge controller didn't budge after 20 minutes of pumping. I replaced it with a cheap dial gauge and it bottomed out within 30 seconds. I hooked up a variac-controlled, rectified 12KV 60mA NST to the feedthrough. Here is the result:



I used a piece of copper tubing to secure the grid to the feedthrough. It will be covered with a ceramic tube that is likely arriving tomorrow. The plasma clung to the grid, not forming the the plasma ball we all know and love. I am assuming that this is because the pressure is too high. I am unable to get an accurate pressure measurement at the moment, but am convinced that the pressure is the problem. I do not feel comfortable turning on the diffusion pump without verifying that the pressure is in the correct operating range. I suspect there is a leak in the vacuum system even though the bolts are very tight.

-Kuba

[Scott Moroch]

Kuba,

Really nice progress! I have a few suggestions. First, I would start by attaching a thermocouple gauge directly to your mechanical pump. Although the pump is rated for 15 microns, you will want to test it with your own equipment to get an idea of the pressure it is capable of achieving. Keep in mind that your chamber and diffusion pump are both a large volume and therefore I would have your chamber pump down for longer than 20 minutes. However, it is concerning that the dial did not move on the TC guage even after 20 minutes. I do not think that your conflats are leaking, assuming you tightened them properly, however if you do have a leak it might be on any NPT threads. Make sure you are placing the teflon thread on properly and I would check the manifold on the mechanical pump for a potential leak. Your chamber and gaskets could be outgassing which is why I would run your pump for longer than 20 minutes. By placing the TC gauge directly on the pump, this will verify that it is working properly.


I would also check the fitting between your diffusion pump and gate valve for a potential leak. Did you clean your chamber and components prior to assembly?

I hope you figure everything out soon.

Scott Moroch

[Kuba Anglin]

Thanks Scott,

You were right about the manifold on the mechanical pump. I removed the brass tee and installed only steel components with Teflon tape and vacuum grease, which greatly improved the vacuum when putting only the gauge on the mechanical pump. I don't have the correct fitting in order to attach the TC gauge to the mechanical pump, but I do have another electronic gauge that read 1 micron (wow!) when attached to the mechanical pump. I used this gauge.

However, when I connected the chamber back to the pump, I got the the same results as before. I decided to see if I could isolate the problem so I shut the gate valve. The mechanical pump could pull around 40 microns (likely more if I waited longer) after one minute of pumping down the diffusion pump body. This means that the problem exists somewhere above the gate valve. Now I need to figure out which connection is leaking. I have suspicions that the connection between the gate valve and the top adapter that connects to the main chamber is the problem. I will need to verify this somehow.

Thanks for your help,
Kuba

[Jerry Biehler]

If you are running with that diff pump you may not have a big enough roughing pump. That takes a pretty good sized pump to keep it backed, more than an Air Conditioner pump at least.

[JoeBallantyne]

Kuba -

You indicated that you used Viton gaskets for all the flanges, but it looks like those flanges are actually Conflat. Which take a copper gasket between knife edges machined into the flanges. Did you use copper gaskets with the conflat flanges or not? Or are those flanges actually not Conflat at all? They sure look like they are.

If they are, at a minimum, you should use the copper gaskets on flanges you don't expect to have to open frequently if ever. Likely at least 3 or 4 of your flanges could be sealed with a copper gasket and then left alone. Ideally all of your conflat flanges should use copper gaskets.

When tightening your flanges what tightening sequence did you follow on the bolts? Did you tighten opposite pairs, or just go around the circumference sequentially. Ideally you do opposite pairs and in an order that puts each sequential pair as close to exactly half way between the last pair of installed bolts as possible. (Tighten, top and bottom, then left and right, then 45 degree pair (top and bottom) on each side of vertical, etc.) Also, you don't want to crank each pair fully tight right away. I am sure there is documentation you can find on the web on the correct sequence and method to follow for tightening conflat flanges.

Also, that is a nice viewport, and you might want to put a sacrificial piece of glass in front of it on the inside, so that stray electron beams don't heat and destroy that likely expensive viewport. (They can heat and destroy the inner cheap piece of glass instead.)

Great progress, and a very nice looking build and chamber.

Hopefully you will find and fix your leaks easily and quickly.

Joe.

[JoeBallantyne]

After taking a closer look at your pictures, it looks to me like you could use copper gaskets on all of the large diameter flanges that are above the top of the wooden platform.

Since both the high voltage feedthrough and what looks like it may be the gas feedthrough are actually connected to the larger flange.

So you don't ever have to remove any of those large flanges, you can just take off the smaller ones that connect to them.

Normally conflats have a slight space between them even when tightened, and when you look down the crack, you can see the copper gasket in the middle.

Yours have no space between them at all. So either you REALLY cranked those bolts (which is bad), or there is no copper gasket.

You should get a torque wrench so you can accurately control how hard those bolts are torqued.

Joe.

[Kuba Anglin]

Joe,

Thanks for your reply. About the mechanical pump and diffusion pump, the mech pump is rated at 12 CFM. My diffusion pump documentation states 7-10 CFM. As long as this is accurate, I should be okay. The Viton gaskets I purchased were molded for conflat flanges. I removed the HV feedthrough and saw that the Viton gasket was damaged. This is likely because I tightened them incorrectly, like you explained. Yesterday I purchased copper gaskets for every connection. I will disassemble all the components and replace the damaged gaskets with copper ones. While I did tighten in the "star formation," in my haste, I did not gradually tighten each bolt. I have read up on bolt tightening documentation and have ordered a torque wrench. This is definitely a learning experience for me and I am so grateful to have this forum to help me.

Thank you,
Kuba

[JoeBallantyne]

While you have it the chamber taken apart, if I were you I would put in a piece of sacrificial glass, and maybe also some stainless, copper, or aluminum screen in front of that viewport on the inside.

It will make the view of the inside slightly less appealing, but I'm sure your parents will be happier if they don't have to fund a replacement viewport. Which is a distinct possibility if you don't protect it.

Doug Coulter discussed doing this in the past, on either this forum, or his own.

Note that the copper conflat gaskets are generally use once, and then replace every time you open them up. With very careful technique some people can reuse them, but you have to keep them positioned essentially exactly the same way, so that everything goes back together exactly like it came apart, and you have to minimize the initial torque the first time you use it, and then slightly increase it each time you reuse it.

Standard practice is to use a new copper gasket every time you open a connection, but that can get pretty expensive real quick for things you are opening a lot.

It should be fine to use a Viton gasket instead for the things you will be opening a lot. (Like probably the feedthrough. Since you will likely be making changes to the cathode.)

You should also be careful about the connection between the chamber and the gate valve, since likely the chamber side is conflat, but the gate valve is not.

Joe.

[John Futter]

This is the second time i have replied to this post

you can use viton or even ordinary buna rubber gaskets with conflat
it only gets a big deal below 1x 10^-8 millibar a place where no amateur fusor has gone
You can buy formed Viton gaskets for CF and they cost a lot more then a copper, but they are reusable and after one added opening you are on the right side of the cost ledger.

We use the viton gaskets @ work while proving new equip to save Time and money. ( PS on big cf flalnges you only have to put in every second or third bolt)
Joe please do not post misinformation

[Jerry Biehler]

Sorry, there is no way in hell that pump is rated 12 cfm. This is what a 12 cfm pump looks like: http://www.idealvac.com/product.asp?pid=1927

In reality you are probably getting 2 or 3 cfm out of that pump at ultimate vacuum range, if that.

[Rich Feldman]

In my unpracticed opinion, Kuba's backing pump ought to be OK even if it's only 3 cfm.

Diffusion pumps can't read backing pump nameplates. All they care about is pressure at discharge port. Actual pressure at that place depends on the actual gas load and backing pump speed, but not on diffusion pump size. Larger diff pumps have higher expected flows, hence higher recommended backing pump speeds.

With that said, a larger rotary pump gives you more margin for hose resistance, deuterium flow, etc.
Suppose a rotary pump and foreline hose deliver S = 3 CFM (85000 cc per minute) pumping speed at diffusion pump outlet.
If the deuterium flow + leaks amount to Q = 85 sccm, then P = Q/S = 1/1000 atm = 760 microns (at best).
If leaks are negligible and the deuterium flow is 8.5 sccm, then P = 76 microns (at best).
When in doubt, monitor the foreline pressure. At the diffusion pump end!

[Kuba Anglin]

Here is a link to the mechanical pump that I purchased. Compared to other 12 CFM rated pumps, it did seem a little suspicious, but I trusted the claim and bought it anyway. If achieving the optimum CFM as stated by the diffusion pump documentation is not actually necessary, that would be great.

-Kuba

[Richard Hull]

Jerry is correct. The backing pump would need to be rather massive as the volume of the diff pump probably rivals that of the chamber!! Wow!

For a stock amateur system, the diff pump's volume typically is about one third or less of the actual chamber volume to be pumped. This is not a law, but more of a norm. Giant diff pumps require giant backing pumps. I had posted years ago that a friend of mine was so proud that he obtained a 12" throat, water cooled diff pump for only $25.00 at a local junk yard. He had a Welch 1400 fore pump! (1.1 CFM) What was wrong with that picture?

I think he gave it to his wife as a planter pot after no one in our local group would accept it as a gift at one of our meetings........

Richard Hull

[Kuba Anglin]

I would like to verify the actual CFM rating of the pump I own currently. Is there a cheap and simple way to do this?

Thanks,
Kuba

[Rich Feldman]

.... Is there a cheap and simple way to ... verify the actual CFM rating of the pump I own currently?
Look at the picture in the "12 CFM" ebay listing.

cfm.PNG (164.68 KiB) Viewed 5280 times

Does that match the product they shipped to you?
Obviously one actual rating is 6 CFM, given by whoever had the nameplate printed and attached.
Another rating, more dubious, is that given by the ebay seller.
A nameplate with different values for 50 and 60 Hz power would earn a bit of respectability.

Is the pump ischarge port at the end of the handle? You could measure the rate at which it inflates an 11" latex balloon (0.5 cf), or a trash bag, etc.
Or run a hose into a large water bottle, inverted in a trough or bucket of water, and measure the displacement rate (in or out).

h2o.PNG (51.46 KiB) Viewed 5280 times

Can some expert please explain why more volume in a vacuum chamber, or diffusion pump,
seems in practice to need a higher capacity rotary pump?
Aside from pump-down time, which ought be proportional to chamber volume divided by pumping speed.
Thanks!

[Kuba Anglin]

Thanks for the suggestions Richard,

The pump I purchased indicates 12 CFM on the nameplate, not 6 CFM. I think it said somewhere in the description of the product that the image may not be correct.

Thanks,
Kuba

[Jerry Biehler]

I think CFM is calculated by measuring the mechanical displacement of the rotor. You would have to take the pump apart to do that.

But I can guarantee you it is just impossible for that pump to be 12 or even 6 CFM, it is just physically too small. I have a 6 cfm pump and the pump assy inside the oil reservoir is bigger than the oil reservoir on your pump. Also the hose fitting on the pump are way too small, they should be KF-25's on the pump itself to get the flow necessary.

Yes, a small pump can run a large diff pump in ideal situations. But this is just too small and the hose restrictions are too great. Any gas load in the chamber and used pump will easily overwhelm the little pump and push the foreline pressure up past it's operational range.

[JoeBallantyne]

Regardless of the capacity of your mechanical pump, you first need to get the leaks out of your chamber so that you can pump the chamber down to the point at which you can consider turning on the diffusion pump. Until you can get down to close to the pressure limits of the mechanical pump, and can maintain that pressure in the chamber with the mechanical pump valved off, you have leaks you may need to resolve.

Note that you should be able to use the gate valve to control how much gas loading is put on the mechanical pump.

Pump down everything to mechanical pump limits with the gate valve open, or partially open. Close the gate valve. Fire up the diffusion pump, and wait for the pressure between the diffusion pump and mechanical pump to again approach the lower limits of the mechanical pumps pressure, and then open the gate valve just a crack. That should limit the gas loading placed on the mechanical pump. Wait until the pressure between diffusion pump and mechanical pump drops down again, and then open the gate valve slightly more.

You can control the gas loading on the diffusion pump and the mechanical pump to some degree by using the gate valve appropriately.

When you are running the fusor, you will want to use the gate valve to help control the pressure anyway. I doubt you will have the gate valve anything close to full open when you are adding deuterium to the chamber and trying to do fusion.

[JoeBallantyne]

One other thing, is that it appears you have not setup the cooling loop on the diffusion pump. Which you will need to do, as that is a water cooled pump. Diffusion pumps have to be cooled on the outside in order to work properly. You will need to figure out how to cool that pump before you try turning it on. Otherwise it simply won't work properly.

Some people just flow water through the pump, and then down the drain. But it doesn't look like there are any drains in the room where you have that setup. The other alternative is to setup something you can use to cool the water, and make it a closed loop system.

[JoeBallantyne]

In addition it looks like you may have the bottom of that diffusion pump actually sitting on some wooden blocks. Which is a fire hazard, as the bottom of those pumps get very hot. (They basically have a heating element like those on old electric stove tops in them.) You should either have empty space under the pump, or put something under it that is not flammable and is a good insulator. Most diffusion pumps hang suspended from the chamber above them which is then attached to the support structure.

[Richard Hull]

Joe gives good advice. A diff pump that large will get very hot at the base. Check out how others, in the past, have their diff pumps mounted in images du jour.

Richard Hull

[Kuba Anglin]

Thanks to everyone who replied to this thread and offered their advice. I will modify the reactor accordingly and post an update soon.

Thanks,
Kuba

[Andrew Robinson]

I had posted years ago that a friend of mine was so proud that he obtained a 12" throat, water cooled diff pump for only $25.00 at a local junk yard. He had a Welch 1400 fore pump! (1.1 CFM) What was wrong with that picture? I think he gave it to his wife as a planter pot after no one in our local group would accept it as a gift at one of our meetings.

Kinda sounds like Alex

[Richard Hull]

Kinda' was Alex.......He tends to think bigger is better......This alone, qualifies him for placement at the highest levels of well funded , government fusion research.

Those who don't know this fine and long time member of the Richmond HEAS group are missing a great guy who is always doing things with his hands in an onward and ocassionally even upward manner, which is more than most of our members ever dream of.

In any group you have the "joiners", the "hangers-on" and the "doers". Oddly, the doers are always in the minority, but it takes all kinds to make a universe.

Richard Hull

[Andrew Robinson]

Haha, yeah I figured it was him. Off topic, but I should be heading your guys way at some point this Spring/Summer. Alex and I are finally tackling that new rotary gap for him to get his magnifier back up and running. Still working out some of the details, but hope to be getting some machining done for it in the near future.

And PS, Richard is right... Alex... Hahaha... GREAT GUY! Hilarious!