Page 1 of 1

Diagnosing pressure rise in bell jar

Posted: Tue May 11, 2021 6:39 pm
by Arun Luthra
I thought I would post this pressure curve here in case there was some obvious suspects that people could identify.

I have a new vacuum chamber setup, it is a steel baseplate with a glass bell jar on top. The gasket is 1/8" thick orange silicone rubber. Currently the plumbing is all KF16 and KF25 piping (and metal bellows hose), connected with 9 KF clamps.

2-stage rotary vane pump -> Valve -> pirani gauge on tee -> chamber.

This is what the pressure vs time looks like when I close the valve:
pressure-log.png
pressure-linear.png
The fact that the linear-scale chart shows some curvature suggests to me that at least some of the pressure rise is due to outgassing, and the outgassing material is slowly getting depleted over time.

If there are any obvious improvements I could make that jump out at you, let me know. For example, I have a foreline trap that I can put between the valve and pump. Would this help a lot?

2-stage rotary vane pump -> foreline trap -> Valve -> pirani gauge on tee -> chamber.

Re: Diagnosing pressure rise in bell jar

Posted: Tue May 11, 2021 7:44 pm
by Mark Rowley
Is the pressure rise while the pump is running or is this something occurring after it's turned off?

Any new system takes awhile to stabilize and remove the moisture. However, looking at the list of material you posted, I'd bet your problem is with the orange rubber gasket. Bell jars and rubber gaskets are famous for being leaky. The best bell jar arrangement has no gasket at all. Just a flat ground glass mating surface and silicone vacuum grease.

Mark Rowley

Re: Diagnosing pressure rise in bell jar

Posted: Tue May 11, 2021 10:38 pm
by Richard Hull
I am sure Mark nailed it. Outgassing of components in the freshly sealed system as opposed to leaks. The very lightest coating of the gasket with vacuum grease might help. However all of this assumes you are sealed well all the way down the line from the valve to the chamber.

On great test is to warm the steel base plate while pumping....(DO NOT over heat or the gasket may get involved) If the pressure rises while pumping, the plate has occluded gases. Give the system time and keep running the tests after pump down. Taking the bell jar on and off will ruin all your good deeds obtained from long pump downs. So, seal the system for real use and go from there.

Richard Hull

Re: Diagnosing pressure rise in bell jar

Posted: Wed May 12, 2021 12:52 am
by Arun Luthra
Just for fun I fit the data to a leakage + outgassing model. I made the simple assumption that outgassing is proportional to the remaining mass of some volatile substance. It fits the data well. Leakage rate is a constant (should be valid for low pressure).

The outgassing decreases by 1 e-folding every 483 seconds.

I'm probably going to get some vacuum grease...

Richard, your point about keeping it sealed is well taken. I got a similar curve in 3 separate pump-downs which means the H2O/etc keeps resetting (letting in air and removing the jar each time).
pressuremodel.png

Re: Diagnosing pressure rise in bell jar

Posted: Wed May 12, 2021 1:16 am
by Arun Luthra
Mark Rowley wrote: Tue May 11, 2021 7:44 pm Is the pressure rise while the pump is running or is this something occurring after it's turned off?

Any new system takes awhile to stabilize and remove the moisture. However, looking at the list of material you posted, I'd bet your problem is with the orange rubber gasket. Bell jars and rubber gaskets are famous for being leaky. The best bell jar arrangement has no gasket at all. Just a flat ground glass mating surface and silicone vacuum grease.

Mark Rowley
The pressure rise is while the valve is closed and the pump is off.

Re: Diagnosing pressure rise in bell jar

Posted: Wed May 12, 2021 2:20 am
by Mark Rowley
Looks like this was discussed pretty thoroughly here:
viewtopic.php?f=6&t=13086&p=85013#p85013

A bell jar on top of a 1/8" o-ring sounds really problematic. Maybe a picture would help make it more clear. Still though, refer to the above thread. A wide ground glass mating surface with silicone grease directly in contact with the base plate is a sure win. No troublesome o-rings required.

In the off chance you don’t know, practically every operational fusor requires continual engagement of the pumping system. Granted, a demo fusor could do its thing for awhile with no active pumping but it will eventually leak in air as well as outgas from a hot grid. You’ll notice changes in plasma formation as well as the slow demise of “star mode”.

Mark Rowley

Re: Diagnosing pressure rise in bell jar

Posted: Wed May 12, 2021 11:40 am
by Arun Luthra
I went with an 1/8" thick o-ring because a thicker one would have had greater lateral forces, and if there was no o-ring it would require more extreme flatness requirements, and more grease. I have read bad things about vacuum grease getting all over the system.

Re: Diagnosing pressure rise in bell jar

Posted: Wed May 12, 2021 11:41 am
by sinan olcun
I have a similar set-up and I agree that the silicone gasket is probably the issue, I was able to get down to and hold vacuum at around 45 microns once I replaced the silicone gasket with a Viton one that I baked according to the instructions in the link below. Though having two ground surfaces is probably the best bet, I had no way of getting my metal plate ground so Viton it is. You can get Viton gaskets from Mcmaster-carr or any specialty providers.

https://www.duniway.com/images/_pg/vito ... orings.pdf

Re: Diagnosing pressure rise in bell jar

Posted: Wed May 12, 2021 7:47 pm
by Arun Luthra
sinan olcun wrote: Wed May 12, 2021 11:41 am I have a similar set-up and I agree that the silicone gasket is probably the issue, I was able to get down to and hold vacuum at around 45 microns once I replaced the silicone gasket with a Viton one that I baked according to the instructions in the link below. Though having two ground surfaces is probably the best bet, I had no way of getting my metal plate ground so Viton it is. You can get Viton gaskets from Mcmaster-carr or any specialty providers.

https://www.duniway.com/images/_pg/vito ... orings.pdf
Did it hold 45 microns with the pump turned off?

Re: Diagnosing pressure rise in bell jar

Posted: Thu May 13, 2021 4:56 am
by John Futter
Arun /Sinan
Arun please do not repeat with block repeats --We are all pretty clever here and can read what the thread is saying--a few years ago this was costing extra money with another provider not now, but we ask that people do not block repeat to keep things succinct.

Sinan
I see this is your first post
Your first post should have been in the "please introduce Yourself" forum as the rules you signed up to dictate-- please amend!!!

Re: Diagnosing pressure rise in bell jar

Posted: Sun May 30, 2021 1:14 pm
by Arun Luthra
Here is a longer run that is a lot more promising, in the sense that it had a good fit to a two-outgassing-component model with a fast component and a slow component. The fit suggests that leakage is low.
double-outgassing.png
Not sure if this is enough information to diagnose, but does anyone think some H2O microfilm is the main source of the outgassing? I've heard that chamber bakeouts are a common method for fixing this?

I added a foreline trap. It does not seem to have electrical connections. The flow conductance on the trap seems really low (pumpdown is a lot slower). It has been sitting in open air for probably 2 years or more, maybe soaking up water vapor, so, I imagine it would need a multi-hour pumpdown to help dry it out. I might have to ditch the trap.

At the end of the day, I'm not doing a fusor experiment, my operating pressure is more like 1 torr. So I may not need to figure out all of these issues.

Re: Diagnosing pressure rise in bell jar

Posted: Sun May 30, 2021 1:20 pm
by Dennis P Brown
So it looks like about one micron per three minutes - that is classical outgassing and not an issue. One the system is clean (been under vaccuum and not exposed to moisture) it will hold a lot longer. An old trap has a lot of contaminates and it too will outgas so, again, not to be an issue.

Some traps use an active agent that can be removed for replacement or separate cleaning. If yours can be removed, a quick heat in an oven might help - not too hot (100 - 150 C). Then place in (still hot) and pump it down for awhile.

Re: Diagnosing pressure rise in bell jar

Posted: Sun May 30, 2021 9:39 pm
by Arun Luthra
It is actually about 500 microns over the first 2000 seconds (even faster initially), so 15 microns per minute. With high vacuum grease on the gasket it is about the same.

Re: Diagnosing pressure rise in bell jar

Posted: Mon May 31, 2021 7:35 am
by Dennis P Brown
Ever since I learned calculus, my simple math skills are terrible ... ;) ( I saw the top line as brown but now realize its two other colors that appear brown on my screen.)

Fifteen microns per minute is a bit fast for simple out-gassing unless you have a huge surface volume, a very small, real leak or a very dirty system.

For leaks like this I generally break down all my KF's (if you use those), clean the 0-rings (a lint free cloth; maybe a hint of Isopropyl alcohol), bake anything that is all metal, and reassemble after an over night wait under vacuum. I check the pump oil (should look very clean even after being run for a long time), and also purge my vacuum pump (air inlet valve on the pump.) In extreme cases, clean my oil trap (for the pump - again, an alcohol rinse, allow to dry, then a bake out followed immediately by a pump out).)

When I pump the system down, I close the pump's cut off valve even if there is a oil trap (I have the valve above the vapor trap, of course). Then I test again. If issues, I spray every connection with a little alcohol - this can often close a very tiny leak or possibly, allow a short improvement in the system in which case you found a possible problem area.

If you have the equipment to get down to under 10^-4 torr, and an ion gauge, the alcohol test is ideal.