Possibly Leaky Turbopump

For Short Term Learning Discussions ONLY. This area is for CURSORY questions and connecting with other users ONLY. ALL technical contributions need to be made in the appropriate forums and NOT HERE. All posts are temporary and will be deleted within weeks or months. You should have already search the extensive FAQs in each of the forums before posting here as your question may already be answered.
Post Reply
User avatar
Maxwell_Epstein
Posts: 79
Joined: Mon Jul 19, 2021 11:08 pm
Real name: Maxwell Epstein
Location: Columbia University
Contact:

Possibly Leaky Turbopump

Post by Maxwell_Epstein »

Hello all,

It has been a while since I last posted and quite a bit has happened! I was finally able to buy a turbopump and controller for a very reasonable price (still very high for me and my resources, but very good in general). A few weeks ago, I began testing the turbopump before hooking it up to my chamber and have encountered a pretty substantial leak that I am having trouble sealing.

The setup is as follows:
  • My roughing pump (a Precision D-25) is attached to a kf-25 foreline containing a vent/up-to-air valve and a manual right angle valve
  • On the other end of the valve, the Pfeiffer TPU-170 turbopump (overkill, I know, but all I could find) is connected by KF-25 connection
  • The 6" CF flange at the top of the turbopump is reduced to 2.75" CF which is adapted to KF-25 which attaches to my thermocouple
In operation, I activate the roughing pump until the system pumps down to 300 microns. If I at this point, isolate the turbopump and deactivate the roughing pump, the turbopump section will leak very quickly back to near atmosphere. There is no audible leak. If I instead activate the turbopump at 300 microns, the system will very quickly (and noiselessly) pump down to around 20 microns but will not go below this point. The foreline does not leak without the turbopump attached so it seems very likely that it is the turbopump itself that is leaky. I have replaced all gaskets and tried resealing all connections multiple times. All knife-edges seem to be in good condition.

I am not sure how to proceed here and would appreciate any advice or direction. I bought the turbopump from someone I trust with evidence that it could pump down to 10^-7 torr so I have no reason to believe that I was sold a defective unit. Could I be operating it incorrectly? Could it have been damaged in shipping? If it is damaged, what are my options for repair?

Thank you,
Max E.
Attachments
The setup with the leak. I later removed the valve on top of the pump and attached the thermocouple directly to the top of the turbo.
The setup with the leak. I later removed the valve on top of the pump and attached the thermocouple directly to the top of the turbo.
PXL_20220904_044224063.jpg (28.63 KiB) Viewed 5492 times
See my needlessly verbose fusor blog here: https://highschool-fusioneer.medium.com/!
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Possibly Leaky Turbopump

Post by Richard Hull »

You have a leak, obviously, but I would look at the turbo as not a point of leak. Yes to leaks at the turbo input and output....No to the pump itself.
does the pump have a small up to air port? This is usually covered or blanked off with a KF-16 cover and clamp.

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Maxwell_Epstein
Posts: 79
Joined: Mon Jul 19, 2021 11:08 pm
Real name: Maxwell Epstein
Location: Columbia University
Contact:

Re: Possibly Leaky Turbopump

Post by Maxwell_Epstein »

The pump does have an up to air port that I have blanked off with a KF-16 blank and clamp. I will try removing, cleaning, and resealing all connections and see if that fixes the leak problem. Thank you!
See my needlessly verbose fusor blog here: https://highschool-fusioneer.medium.com/!
User avatar
Liam David
Posts: 509
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Re: Possibly Leaky Turbopump

Post by Liam David »

I second what Richard said. My first suspicion is the NPT threads on the thermocouple, unless that is what you used to verify that the foreline doesn't leak. Be aware that something as small as a hair across one of your gaskets can create a sizable leak. Could you upload a larger picture? Hard to see what's going on in 293 x 220 pixels.
Chuck Sherwood
Posts: 36
Joined: Sat Jun 05, 2021 3:57 pm
Real name: Chuck Sherwood
Location: Naperville, IL

Re: Possibly Leaky Turbopump

Post by Chuck Sherwood »

I have a Pfeiffer TPU-170. It is a great pump IMO.
It has a vent port on the side that needs to be blanked off.
I tried to reuse the copper CF gasket on the top and it did not seal well. I ordered a copper gasket from McMaster which was not the best source (but fast) but it seems to OK after torqueing down all the bolts three times.
It looks like you have two CF gaskets and I would suspect that you leak is there.

If things are sealed that pump will get into the 5s in less than a minute and into the 6s in less than 10 minutes. This is where a cold cathode ion gauge is a very useful indicator. Activation at high pressure will not hurt it much and it quickly shows the rapid pump down.

One trick I read about and have used is to seal threaded connections with Apiezon Wax "W". I still think a cold cathode gauge is a better indicator for this test.

chuck
User avatar
Maxwell_Epstein
Posts: 79
Joined: Mon Jul 19, 2021 11:08 pm
Real name: Maxwell Epstein
Location: Columbia University
Contact:

Re: Possibly Leaky Turbopump

Post by Maxwell_Epstein »

Thank you all for your advice and suggestions!

I'm not sure why the picture was so low quality when I uploaded it, I have attached a higher resolution copy below. I can take better pictures of the resealing process once I get home to confirm that I'm not doing anything wrong there.

I did use that thermocouple to verify that the foreline doesn't leak and the NPT threads were okay there so I do not think that is the source of this leak (although it certainly might be leaky at higher vacuums). I plan to buy a more specialized gauge for high vacuum when my budget allows or a good deal appears either here or on eBay.

I do not have any Apiezon wax or grease, but I do have some Dow Corning high vacuum grease that I have used in very small amounts to seal viton gaskets. I know that Apiezon is higher quality and better for specific tasks, but would the Dow Corning grease that I have work for this purpose?

I have a few spare copper 6" CF gaskets from LDS Vacuum. I will unseal everything and replace and reseal with fresh gaskets. I am using a viton gasket for the connection between the 6" CF to 2.75" CF zero length reducer and the 2.75" CF to KF-25 adapter, and I will replace that gasket with a fresh one as well.

I am very relieved that nobody thinks the pump itself was damaged in shipping. If resealing everything fixes my problem, I will move on to attaching my chamber, upgrading the power supply, and eventually fusion!

Thanks again!
Attachments
This should be higher resolution
This should be higher resolution
See my needlessly verbose fusor blog here: https://highschool-fusioneer.medium.com/!
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Possibly Leaky Turbopump

Post by Richard Hull »

The key thing is to get just the foreline down well below 50 microns. I usually wait until my foreline is at 10 microns before I open the valve to the turbo. I will also second Liam's comment on the 1/8 NPT seal on the TC gauge sensor. Most all of my past frustrations in even testing a pump by itself at the head of the mechanical pump have involved leaks at the TC gauge sensor in the KF-25 adapter.

Most properly attended to KF and CF professional vacuum hookups will not leak to so great a degree. It is threaded and push on things like hose and TC gauge stuff that leaks.

Is that cross really needed?? It is 4 needless points of possible leaks.

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Maxwell_Epstein
Posts: 79
Joined: Mon Jul 19, 2021 11:08 pm
Real name: Maxwell Epstein
Location: Columbia University
Contact:

Re: Possibly Leaky Turbopump

Post by Maxwell_Epstein »

In this current configuration, the cross is superfluous and I will get rid of it. I was using it as an attachment point for my thermocouple tube, but now that the TC will go elsewhere I can get rid of it and put the up to air valve on the T instead.

Unfortunately, I only have one TC tube at the moment so I cannot measure pressure both on the foreline and at the opening of the turbopump in a single test. This is one of the main reasons that I want to get another gauge soon. I will, however, make sure that my roughing pump is properly conditioned to get my foreline to 50 microns pretty easily (I think its overdue for an oil change anyway).

Thanks for the advice!
See my needlessly verbose fusor blog here: https://highschool-fusioneer.medium.com/!
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Possibly Leaky Turbopump

Post by Richard Hull »

I have a log sheet on my forepump. I last changed the oil in 2018. Now, it is to be noted that the hose to the foreline was put on in 2004 running to a KF25 valve which hasn't had its KF25 to the turbo pump removed since 2017. I also vent my pump to the great outdoors and it has a ball valve right at the exhaust that remains closed all the time the pump in not running. This isolates the pump inlet and outlet forever when not on.

If you are in constant test mode and sucking lines and chambers filled with air regularly in numerous ad hoc lash-ups over weeks or months, this can foul oil in short order. Only change oil when you just can't get to the low micron numbers. Once in a system and the pump is isolated by inlet and outlet valves, oil changes might go for years.

The foreline valve and the fusor chamber valve also isolate the turbo in my system. Once stable and sealed well in such an arrangement, pump down to fusion levels goes very rapidly. In my case, over the last weekend, as listed in the HEAS image post and report, I vacuumed out the entire system, with the turbo up to full speed, the fusor V was submicron and ready for fusion in about 5 minutes.

If a system is to be even semi-permanent, this is the way to go. The vacuum system must fall into the category of a no-worry tool, and not a constantly fidgeted with appendage to doing fusion.

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Chuck Sherwood
Posts: 36
Joined: Sat Jun 05, 2021 3:57 pm
Real name: Chuck Sherwood
Location: Naperville, IL

Re: Possibly Leaky Turbopump

Post by Chuck Sherwood »

You state that if I shut things off, the system comes back up to ATM quickly. I don't see a valve between the roughing pump and the turbo. Without a valve to isolate the turbo it will air up quickly through the mechanical pump. BTW, this is not recommended as it is a good way to suck mechanical pump oil into your turbo.

You state that it goes to 300 and then down to 20 when you start the turbo. How trustworthy is your gauge?? Has it ever been "zeroed" at known high vacuum levels? Thermocouple gauges are not very accurate to start with and a contaminated sensor will likely never indicate a low level. What is the speed of the turbo here?? If its close to max, then this might be just a bad sensor or a now zeroed meter.

That turbo can tolerate a foreline pressure of 100 mTorr. I so start my turbo when the system pressure goes below 100. This is almost instantly with a Welch 1376 backing pump. This helps keep mechanical oil vapor out of the turbo.

chuck
Chuck Sherwood
Posts: 36
Joined: Sat Jun 05, 2021 3:57 pm
Real name: Chuck Sherwood
Location: Naperville, IL

Re: Possibly Leaky Turbopump

Post by Chuck Sherwood »

I did a little test today. I grabbed an old Duniway TC readout I bought long ago in a lot. I don't think I ever powered it up until now. I mated it with used tube removed from something long ago as well.

I put the 531 tube into a compression fitting and mounted on a system with a XacTorr 0-1Torr Capacitance Manometer.

I was amazed to see the 531 match the manometer very well down to about 150 mTorr. At that point, I turned on the Pfeiffer turbo and the pressure rapidly went into the 4s so the 531 TC should be reading close to zero, but it never went below 30 micros.

I adjusted the zero pot to zero under high vacuum. Now it reads poorly at higher pressures. For what it is worth, and it proves nothing, this scenario resembles Maxwells scenario.
IMG_2750.jpg
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Possibly Leaky Turbopump

Post by Richard Hull »

Occasionally I will come across old TC gauge tubes that are from old gear or just stored poorly and may even have some rust on the exterior. My first effort is to test them under vacuum metered by a capacitive manometer. If they are very far off the mark, I take a syringe and load them up with acetone of MEK. I shake them to move the liquid over the thermocouple. I then let this set for a short period and then shake again and dump the liquid out onto a clean white tissue. If the liquid is discolored, I repeat the process until it goes clear.

This process has corrected a number of "off base" reading, old, TC gauge tubes. In many cases this is just a colloidal coating of oil or other hydrocarbon on the thermocouple, fouling it from heat distribution as designed. It is very important to have a chart with the tube data and verify the heater current is correct and that the metering presents the correct load impedance as stated by the manufacturer. All of this is worth the effort as these tubes cost $60 to $100 each. An old tube can be as good as a new one with proper testing.

The 531 is typically AC operated at one of the highest heater currents of all tubes (~ 160 ma!)

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Maxwell_Epstein
Posts: 79
Joined: Mon Jul 19, 2021 11:08 pm
Real name: Maxwell Epstein
Location: Columbia University
Contact:

Re: Possibly Leaky Turbopump

Post by Maxwell_Epstein »

Thank you, all, for the valuable advice! I have finally had some time to work on the fusor and here is what I have found:

It is difficult to see in the picture that I posted, but my original turbopump test setup included a right angle valve between the KF-25 cross and the KF-25 tee. I used this valve to isolate the foreline and turbopump from the roughing pump so that I wouldn't suck in air through the oil and potentially foul parts of my system.

As per everyone's suggestions, I re-sealed all the gaskets and connections everywhere in the system and saw notable improvement! At this point, the roughing pump can pull the foreline (now just the cross with no extra KF connections) and turbopump down to 100 microns pretty quickly. There is still a small leak in one of the turbopump connections (I think the 6" CF connection), but it is not large enough to interfere with any sort of operation. I have also definitively found that there are very minor leaks in my barbed hose connections from the roughing pump to the foreline and in the threaded TC to KF connection. Putting some vacuum grease on these points lessened these leaks considerably. As of now, these leaks all together are small enough that they do not interfere with any function. My vacuum system is now a "no worry tool," as Richard put it, for my needs right now. Once I get a capacative manometer gauge and progress deeper into high vacuum territory, I will have a list of things to worry about, but for my demo system right now, this is adequate.

As far as my TC gauge and tube. I managed to nearly zero the gauge, but this was with the turbopump running full speed and given how quickly the needle on the gauge dropped, I do not doubt that this is just the TC inaccuracy some of you pointed out. This turbopump experiment has certainly prompted me to look into investing in higher quality gauges sooner.

I have encountered a problem, though, with how I am connecting the turbopump to my main chamber. Because I was apprehensive about connecting the turbopump directly to the chamber, I bought a length of flexhose like this one: https://www.ldsvacuumshopper.com/2flmeho18frl2.html (but for much cheaper on eBay) to connect them. Unfortunately, when I connected the turbo to the chamber in this way and pulled a vacuum with the roughing pump, the hose would contract nearly two inches and pull the turbopump closer to the chamber. When I vented the chamber back to atmosphere at the end of the test, the hose would lengthen back to its normal size. This is obviously not permissible. Is my solution, then, to remove the flexhose entirely? I would need to use a straight length of pipe to attach the two, is there a better way that would avoid adding extra connections?


Thank you again,
Max E.
Attachments
How I initially attached the turbopump to the chamber. The length of flexhose would contract nearly two inches when the roughing pump pulled a vacuum :(
How I initially attached the turbopump to the chamber. The length of flexhose would contract nearly two inches when the roughing pump pulled a vacuum :(
Nearly zeroed out vacuum gauge!!
Nearly zeroed out vacuum gauge!!
I attached my TC tube to a valve on top of the turbopump for testing
I attached my TC tube to a valve on top of the turbopump for testing
The foreline attached to the roughing pump and the turbopump. I will attach the TC tube in the place of the blank on the KF cross to monitor foreline pressure eventually.
The foreline attached to the roughing pump and the turbopump. I will attach the TC tube in the place of the blank on the KF cross to monitor foreline pressure eventually.
See my needlessly verbose fusor blog here: https://highschool-fusioneer.medium.com/!
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Possibly Leaky Turbopump

Post by Richard Hull »

I would never run a turbo in a lash up where the turbo is just sitting on a table. It should be bolted to the table or in a rigid fixed assembly. A turbo is like a gyroscope that can't have its rotating mass, (blade system), disturbed. At that speed, the blade system has what is termed "rigidity in space". This is great in a gyro as its high speed rotating mass can tumble over 3 axes in its pivoting ring system as its case is moved all over the place. For the tight fitting blade group in the turbo, it can be a bomb ready to explode destroying the blade system should the turbo body suffer a shock or rapid motion while running.

All the rest of the stuff can just hang around loose provided it doesn't give a hard jerk to the bolted turbo. You have been lucky. You are now warned. save your expensive turbo! Mount it firmly. Even if mounted firmly to a table or cage, never move the table or cage while the turbo is spinning.

Most turbos have threaded holes for mounting at the base. Make use of them.

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Maxwell_Epstein
Posts: 79
Joined: Mon Jul 19, 2021 11:08 pm
Real name: Maxwell Epstein
Location: Columbia University
Contact:

Re: Possibly Leaky Turbopump

Post by Maxwell_Epstein »

Thank you for the warning! I had thought I could get away with leaving it unbolted on the table so I'm really glad to have been told otherwise before my turbopump exploded rather than after. I will certainly set up a very stable base that I'll bolt everything to.

Using this base, will I still be able to use the flex hose to connect my pump to the chamber? I can imagine that the flex hose trying to contract while the turbo and chamber are both fixed in place will lead to torque being applied to various connections in the system. The one I'm particularly worried about is the horizontal torque applied to the vertical bolts attaching the valve to the top of the turbopump. Is this a non-issue?

Thanks!
See my needlessly verbose fusor blog here: https://highschool-fusioneer.medium.com/!
Chuck Sherwood
Posts: 36
Joined: Sat Jun 05, 2021 3:57 pm
Real name: Chuck Sherwood
Location: Naperville, IL

Re: Possibly Leaky Turbopump

Post by Chuck Sherwood »

It is easy to calculate the force compressing the tubing and the pull it will put on the end points. For a 1.5 inch tube that is about 35 pounds if the end are rigidly attached. Richard's point was that you really don't need to rigidly fix the fusor if you don't want to. But it is absolutely necessary to secure the turbo. Personally, I an not fond of flexible tubes at high vacuum levels because they have a ton of surface area for gas molecules to stick to and the stress they put on the attachment points.

That said, there are a fair number of things to attach to the fusor chamber so it needs to be supported in a reasonable way. Why not build a platform that securely supports the turbo and allows some flexibility in adjusting the chamber position so you can modify the plumbing between the two as you develop your throttle valve to control the chamber pressure. If you make a stand to hold the chamber on the same horizonal plain as the turbo exit pipe you can use solid pipes between the two and add valves or whatever is needed as you please.

If you look at the thread "setting up turbo pumps" you will see how I mounted my Pfeiffer 170. I used four studs in place of the mounting bolts on the top flange. This allowed me to hang the turbo for some brackets made from steel bar stock bought at Home Depot. I used a stand from "Rockler Woodworking" intended as a router table. You can use the supplied metal feet or add wheels with locks.
I used a similar approach to mount this diffusion pump. This time the mounting brackets are attached to the corner brackets. Once again the angle iron is available at Home Depot and this is just a matter of drilling holes in the stand to bolt things too. Eventually, I plan on replacing this DP with the Seiko Seiki turbo show in another thread.
IMG_2757.jpg
This setup has been extremely useful and flexible. The KF flanges allow for quick and easy changes. I can set a bell jar on top of the table or erect a column of KF50 crosses etc to support the current project. The wood shelf sets on top and holds gauges or power sources.

Another point that you probably need to research because I don't have a clear answer. Can the mechanical pump be mounted on the platform like I did with the chamber pump shown above. I believe that the vibration from the pump is very bad for the turbo. So I intend to leave the backing pump on the concrete floor and connected via a rubber hose.

Food for thought.
chuck
User avatar
Dennis P Brown
Posts: 3147
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Possibly Leaky Turbopump

Post by Dennis P Brown »

If you want to mount the roughing pump onto your system's base, use vibration damping rubber post mounts. That should reduce vibration issues.
Post Reply

Return to “New User Chat Area”