Faster Pumping of Chambers: Two Pumps?

One bigger pump is usually cheaper and easier to maintain seals and such than it would be for two.

Maybe you should review alternate technology for extremely low vacuum. There is a point where conventional "piston" or displacement pumps become challenged and items like diffusion pumps are then more effective.

Have a look at "Leybold" to get some key words, then search on those for more details.

How low are you trying to maintain? This will enable the experts to give you better advice.

If there are no leaks on the vacuum side, two equal pumps should evacuate almost twice as fast as a single pump. (If they share any suction piping, it won't be quite twice as fast, owing to additional pipe friction.) Additional pumps, piping, or other components increase the chance of leaks, which again would make evacuation not quite twice as fast.

A few years ago I've met this problem in Kazakhstan, at the JSC "YRYSTY_AECRW" shop. During VPI process vecuumization took more than 4 hours. We have to replace the pump and use more powerful. GE never undertakes the rash acts.

Boldov Andrey,

Leading Eng.

"RUSELPROM ELECTRIC DRIVE"

Add to that list trapped volumes like large surfaces in contact with no relief, screws in blind holes with no vent, these all add to the pumping time.

Is the longer time on a turbo pump or cryo pump? We turbo pump UHV chambers to 10^-6 in 1 hour after baking for a day at 350°C then to 10^-12 on a cryo pump in a few hours.

Yes, I agree totally.

I had some big trouble by a small screw with the threads cut or rolled with some cutting or rolling oil.

This trapped the oil in very small volumes beneath the surface and the residual channels through wich the vapors were conducted being exceedingly small.

So first a small surface evaporating getting bigger and bigger as the channels were nearly empty and the free surface of the oil inside these reservoirs multiplied.

No cleaning procedure could remove these outgassing sources.

Only specifying: "no fluids used during manufacturing" was successful.

RHABE

All things being equal, adding another pump will definately reduce your pumping time.

Sir, Yes is the answer but you would need a non return valve on each pump to avoid sucking back through the other pump, this would also alleviate the need to have two identical pumps, you would need a short piece of pipe between the NRV and the pump just as a reservoir.

I have to ask, why 5 hours to pump out? either the pump you're using is faulty or the wrong type for the job, In my experience if you are trying for high vacuum you need a combination of three pumps, 1 a high volume booster pump,(similar to a blower on a hotrod), 2 what is known as a Kinny pump (which is similar to a compressor in reverse) & 3rd a high vacuum pump (diffusion pump) which is an oil pump that has no moving parts & works with silicon oil which is heated & cooled in a specially shaped chamber. If this is an on going process I could tell you how to speed up the whole cycle but I would need to know more about your equipment also about the type of material you are trying to put under vacuum, ie if its wood then its very porous and would take longer than say steel, also with high vacuum cleanliness is of paramount importance,I need to know if you are using a chamber? & by the way if you are then 5 hours is really over the top.

Although having two pumps (with one set on spare parts) will be a definite time reducer, another option is a second (larger) pump in series with the first.

Larger because it will be pulling on the discharge of the first (like the second stage of an air compressor is smaller because the same mass flow of air has already been made volumetrically smaller from the lead compressor).

The combined suction should get to the desired vacuum faster.

As I think about this a little more- the discharge of the first pump will be essentially atmospheric, so the second pump could be the same size as the first, moving the same volume of air.

In either case- there would be no need for backflow protection since each pump is in series, not parallel.

Working in line or series as you put it would mean that the first pump would always be doing all the work & in the case where pump 1 starts to lose it's suction (as it wears) pump 2 would have to pull through p 1 which is wasted energy also there is the possibility of sucking oil out of p 1, pumping in parallel with NRVs means it doesn't matter if the pumps are sucking at different rates.

NB There is all ways need for backflow protection if there was a power outage & it's taking 5 hrs to evacuate the chamber then you've wasted however long its been pumping.

Bazzer

You're right.

Don't know anybody in their right mind would pump in series.

Parallel pumps don't need any additional NRV. The pump inlet valve is already one. If it isn't good enough to hold the load, then the pump is RS and shouldn't be there.

In my preceding example, the "geniuses" attempting to pull the vac down with a standard vac pump hadn't engaged brain, and the poor damned thing was there for 3 days, rattling away, eventually run out of oil.

I uncoupled it and got the big workshop vacuum cleaner, coupled up to the vac line and we had the vessel down to 'pressure' in less than 2 hours. Job done.

You've got to use a machine with a separate fan to cool the motor tho'. Light duty ones rely on the air flow of the suction line to cool the motor, and in our example there is damn all of that for quite some time.

Hope this helps.

OK, the idea behind NRVs is that the short piece of pipe itself acts as a chamber so that unequal pumping rates can smooth out the performance,ie,the pump can evacuate the small "chamber" until it is at a higher vacuum than the main chamber then equalise before starting again. We are all assuming (well most of us) that this is a high vacuum application but it could be a bagging system for panel assembly, without knowing what the machine is being used for we are in the dark.

Bazzer

" without knowing what the machine is being used for we are in the dark."

Yep!

Helpful if the OP gave us some clues as to where we're at, after all he DID ask for help.

Cheers,

Stu.

Iyer, Before I can advise you I need to know what is the size of your chamber. I just had to pump down a chamber : dia 3m x 12m for integrity and leak testing. Had to achieve 1400mm water column, static for 2 hours. Are you this big? Stu.

Our OP, after expressing his problem, has not come back to in reinforcing his existing context.

We all are trying to assume his existing context and trying to help him.

Well, even if we have not precisely predated his existing context, we all derive some befit out of such vague discussions. Every unique experience is knowledge for others; in this way going through others experience (though not relevant to the original scenario) is interesting and beneficial.

In the first place, OP has expressed, "Can I speed up the pumping of chambers with 2 pumps instead of one? This is for high vacuum applications that takes about 5 hours to get to the required vac levels."

In his set-up every equipment/component could be in their best reliable level and from day one it takes 5 hours to get the required vac, who knows? And his present requirement may be to improve the system's performance by reducing the overall time. If this were the case, our discussions towards finding faults on Vacuum effecting and canceling are irrelevant for his existing context.

Wish, OPs should continually participate by adding more relevant reinforcing data pertained to their particular context as other experienced experts to assist them solving their problems.

Sorry folks, I was following the discussions and wanted to interject many times. But I was trying ot get specific details of the chamber size and the existing pumps etc. from the customer. But couldnt get the precise info.

From what I know, these are large chambers for CVD/PVD deposition techniques. The pumping time of >5 hours occurs when the entire chamber is opened up for troubleshooting or PM etc.

It simply occured to me that intuitively adding addl pumps should improve the time of pumping. I started reading about pumps, pumping curves etc. etc and couldnt find any simple mathematical treatment on how much time would be saved.

Its one of those "If I were running this.. I would add an extra pump "... idea !!

Hence the question. I really appreciate various ideas (had me thinking more with all the series pumping etc.) and I do like this forum which helps keep the thinking caps active.

--Iyer

Iyer,

Like I said, get a bloody great vacuum cleaner and couple it into the vessel with decent diameter hose, and a stop valve, and you'll have it down in minutes. You'll be able to turn 'watching the manometer' into a spectator sport.

If perchance you need to pull more vacuum than the cleaner can manage, have the vac pump coupled in parallel, and start it up once the great majority of the air has been expelled.

Cheers,

Stu.

Just for information,when I was involved in the high vacuum plating industry we had Edwards vacuum plant which was 1800mm dia x 1500mm long & we were using it for atomising aluminium to coat small components, We could get two-three cycles per hour depending on the turnaround & that was to pull it down to 10⁹-10⁶ Torr but it was all kept extremely clean & free of oil & contaminates, the plant had a Roots blower,a Kinney mechanical pump & three diffuser pumps, based on the above dimensions maybe you can see if what you are trying is anything like this.THE END

Bazzer