Piping Losses in an air Vacuum System

[Briland33]

I operate a boat manufacturer and we recently setup a vacuum pump system to vacuum infuse our parts. My goal was to pull parts down to a consistent -28inHg. I bought a GVS 16A pump with 11CFM, 1.2HP, and attached it to a ~80gal tank and then after the tank is ~115ft of 25mm aluminum piping to the nozzle where we are pulling vacuum for the part. Now the issue is at the tank we will have -29.5inHg, great, but at the nozzle -24inHg, not great. I think we have a slight leak because with everything sealed on the part end (we confirmed) the pump should be able to eventually pull it down right? Unless I am underestimating how much power is required once we get into medium and high vac to continue and the pressure drop due to the lines is that significant.

I have someone diagnosing a leak now but if its not a leak I can up the pump from 1.2 hp to say 3 hp cheaper than I can have the lines removed and taken down.

 

[1503-44]

If the pressure difference is due to flow through 115 ft, of 1" pipe it would be a flow rate of around 20 ft3/min.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[EdStainless]

115' is way too far.
You should have a tank (and pump) much closer to the work location.
You also need larger piping.
We never went more than 4-5' with 2" piping.
You should be able to blank off (seal) your working end and shut off your pump and have the vacuum level hold for a few hours before you can measure any (say 0.5mm) change.
Does your system pump down to -29.5 inHg when you have it just sitting?
If your system isn't this tight then you may just want pumps and tanks near each work location.
Vacuum pumps don't suck air. The push it out and then everything left diffuses to be uniform again.
It isn't a fast or efficient process.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[georgeverghese]

I get a leak rate of about 30scfm, which is about 3x more than your current 16A vac pump can handle. So choose the 40A(2HP) or 60A pump(3HP) preferably if you cannot plug the leak or prefer not to make this line shorter.

 

[katmar]

I don't believe that upgrading the vacuum pump alone will help at all. Unfortunately my estimate of the leakage is different from the estimates by 1503-44 and georgeverghese. My calculations show that with a terminal pressure of -24"Hg for the pipe and -29.5"Hg in the tank you would get 11.5 scfm. This agrees well with the current pump capacity but it would also be at the sonic limit and installing a larger pump would not allow you to exceed the sonic limitation.

In case you or anyone else wants to check my flow estimate here are the values that I used
Gas assumed to be air at 20C and -24"Hg
Density = 0.233 kg/m3 (0.0146 lb/ft3)
Viscosity = 0.018 cP
Pressure drop = 5.5"Hg (29.5-24.0)
Inlet pressure = -24.0"Hg Gauge = 5.8"Hg abs
Pipe roughness = 0.0015 mm = 0.00006"
Pipe ID = 27.8 mm = 1.097" (1" Sch 10)
Flow = 23.5 kg/h = 0.87 lb/min = 11.5 scfm

It seems that the tank is not helping much as a store of vacuum. If you isolate or turn off the pump the pressure in the tank will come up to -28"Hg in a matter of seconds with my (or the others') estimate of the flow rate.

Increasing the pipe size to 1.5" would give you about 35 scfm and a 2" line would give around 65 scfm, but you would of course need a bigger pump to achieve these flow rates.

Hopefully the problem is a leak into the system and you can fix it. However we do not know what flow rate you require in operation but if you sized the pump correctly and there is minimal leakage you may be OK.


Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

I don't really get what you're trying to do with this "part"? What sort of gas flow does it leak out under vacuum?
How big is the chamber at the other end?

How are you joining the lengths of aluminium tubing - Why Aluminium?
Have you got any other joints or valves or instrument connections anywhere?
Can you pressure test the system with say 1 bar air and do a bubble test? Might not show up the leak as vacuum stresses the seals differently, but might show up a leak in a pipe or connection.

As katmar points out the density and hence velocity of your air is such that you will be at or close to max velocity. as the air expands further at the tank end it starts going faster and faster. Density at -29.5 ( that's an impressive number) is a tenth of that at -24".

Bigger pump may well not make any significant difference without some changes to pipe size or connections.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

I didn't see the MINUS 29. If air enters at +29.92, Yes, that's sonic.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[EdStainless]

Are you sucking anything other than air?
Any moisture, solvents, or other stuff?
If so then you have other problems to address also.
If there are any solvents in the system they will dissolve in your pump oil.
Their low vapor pressure will limit the vacuum.
And the thinned oil will destroy your pump.

Go to 2" line, get a bigger pump, fix all leaks.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Briland33]

The purpose of the vacuum system is for making vacuum infused boats.

EdStainless: The system dropped 4inHg when i shut all the valves over a few hours so that sounds like a leak however the installer came and pressurized the system and sprayed soapy water on the connectors and determined there isn’t a leak. Hmm not sure about that.

The system has a ridiculous amount of 90’s and its only 1” diameter. I think this contractor effed up by installing 1” piping. It’s aluminum tubing that’s connected with plastic quick connect fittings. My argument to him is that no matter what pump I get the piping will choke the flow and he needs to remove it and reinstall 2” with less 90’s.

Does that increase in diameter sound adequate for having a less than 1inHg drop over the run of piping?

 

[LittleInch]

" plastic quick connect fittings. "

No need to look any further.

Unless specifically designed for vacuum, they will almost certainly hold under positive pressure, but leak under negative pressure.

Without knowing what your flow is very difficult to say, but at that level of vacuum, density is very low so any vapour or air in the system will likely overwhelm your piping.

Unless you weld your pipes together you will never have a good system. IMHO.


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[katmar]

Some quick connects are designed such that internal pressure helps to seal the joints, but this would not happen if there was a vacuum inside the pipe. So a soapy water pressure test does not really prove that there are no leaks under vacuum.

LittleInch made a very important observation on how the density of air decreases very rapidly as you approach full vacuum. If you want to achieve -28"Hg at the end of the line and you want the pressure drop to be only 1"Hg then to make full use of your pump's capacity of 11 scfm the pipe ID would have to be at least 3.4" (86 mm). A 2" line would cope with less than 1.5 scfm at these pressures.

A note to 1503-44: In a recent thread I said that I had never seen a gas line run in the laminar regime. The 3.4" line running as specified above would have a Reynolds Number of 100 and would definitely be laminar because of the very low density. You live and learn.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[georgeverghese]

Am afraid this (-) inHg units got me into a tailspin. Agreed, flow is into sonic at the end of this line near the pump and flow cannot be any more than 5scfm with what you have currently. Pressure at pump is 1kpa abs. Its choking in the line somewhere near the pump. So the leak rate may be more if you "unchoked" the line and used 1.5inch or 2inch tubing.
How about using Parker or Swagelok compression connectors instead of these quick connects ? Parker has these in polypropylene also for 1inch.

 

[pierreick]

Hi,
Consider this resource to perform calculations.
Check your piping, the losses are huge considering the pressure losses. From the document 10% losses is acceptable.
Harvey, the mass flow rate is much higher, even using delta P/P =10% and rugosity = 0.04 mm, about 300 kg/h

Pierre

 

[katmar]

Pierre, thanks for the useful reference. Note that a pressure of -24"Hg is equal to 200 mbar absolute. Using the 10% rule of thumb would allow a pressure drop of 20 mbar or 0.6"Hg. There is absolutely no way to get 300 kg/h of air through a 1" line with this pressure and pressure drop.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[pierreick]

Hi Harvey,
Agree with your comment about flow rate. My calculation about physical properties matches yours(density).
Edit: error in my calculation, with the equation in the document I got 9.43 kg/h.
With your data, rugosity of 0.0015 mm and Delta P/P =10% I got # 40 kg/h; If I use the isothermal compressible model with your set of data I got 18 kg/h
Pierre

 

[LittleInch]

If when this is running you get -24" one end and -29.5 at the other, presumably running flat out, but then seal it off and it only goes to ??-20" over " a few hours" that doesn't sound right?

what is happening at the nozzle end? I think whatever you're connecting this to is giving off gases or vapour of some sort.

Got any more details / photos? "Vacuum infuse our parts" - What does that mean in reality? Infuse what?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Compositepro]

You cannot "store vacuum" in a tank. Tanks store pressure. A tank may help on initial application of vacuum to a vacuum bag when there are many leaks to seal and a large volume of air under the bag. But once the pressure rises in the tank it is just an additional volume that needs to be pumped down. You would be far better-off using 4" solvent-welded plastic pipe as a vacuum manifold throughout your plant. It would have far higher conductance and have more storage volume than your tank.

Are you infusing with epoxy resin or polyester resin that contains styrene?

 

[georgeverghese]

There is a big change in density from inlet to vac pump ( 0.08kg/m3 to 0.01kg/m3) so compressible flow equations apply.

Here are Parker connectors for vac service

https://www.parker.com/content/dam/...flo/Veriflo---pdf-files/25000260_VacuSeal.pdf