Homemade jet/venturi pump

[logoslarry]

I made a jet/venturi pump from a 2" T ,1" discharge and 1/2" air inlet through a 1/8 nozzle,air at 120psi. Suction is 2" pipe about 5' long. Problem is that I'm only getting about 3gpm. Would decreasing the diameter of the suction allow for better flow? Or would I do better increasing the diameter of the discharge? Would steam @ 100 psi give better rate?
Trying to give a hand to our sump pumps at work(poultry plant) which keep clogging up and burning out.Results in a flooded boiler room On the other hand the air lift I made works quite well.

 

[chicopee]

If you can get your hands on mechanical or chemical engineering handbook you'll see that steam is mostly used.
The trouble with air lift pumps is their inefficiency.
Have you researched diagphram pumps (mudsuckers in the construction vernicular)?

 

[zdas04]

Nothing wrong with an eductor/ejector for this service. They're used as bilge pumps all over the world.

One thing to keep in mind is that they are mass transfer devices, and the best I've ever seen done is the ratio of the suction mass flow rate to the power fluid mass flow rate is about 1.5:1, cheep units can be as bad as 2.5:1--your crappy 1/8" air nozzle is pretty cheep and is not letting much mass in so your suction is around 1/4 to 1/2 of the air mass flow rate. I'm surprised you're getting 3 gpm.

Since you are doing this on the cheep, I wouldn't go with steam on a bet. With properly designed convergent/divergent nozzles you can get steam up to around 1.5Mach and really move some fluid, but it is expensive.

Try an "eductor" instead of an "ejector". Do you have a fire-water system or a potable water system? First, change the 1-inch outlet to a full 2-inch (and connect it to something cheep like 2-inch fire hose). Then change the 1/2 power "nozzle" to about 1-inch (and if possible extend the 1-inch into the tee just past the branch, it doesn't have to be rigid), turn on the water and you should get suction flow about equal to 1/2 your power water flow.

I've seen a similar gadget used to suck 55 gallon barrels flat for the land fill.

David

 

[tr1ntx]

You can do as zdas04 said, or if you want even better, you can buy a properly manufactured eductor. There is some amount of technical know-how involved in the function of eductors, so one that is made "right" will greatly outperform a home-built one. It involves the converging/diverging aspects of the throat and fluid dynamics.

For your homemade one, if you don't have it that way already, do like zdas04 said and extend the air nozzle past the T-branch and into the exit leg. If the end is threaded and you can screw something onto it to decrease the annular ares, it should help. Also, without reviewing any notes, I'm thinking you should reduce the suction line to 1".

 

[logoslarry]

Hey guys and or gals,
I had thought of using steam but the air was more easily available. May do it yet if I can talk my boss around to putting in the connections. The original idea came from the PM Shopnotes 1921 pg 3491,probably find it on Google. Another thing I can try is high pressure water,available at about 1000 psi. Of course with air I know when I'm done pumping out.
I'm well aware of eductors used for bilge pumping, spent a few decades at sea. Found out the hard way to leave the hatch open when sucking out the bilge in an enclosed space. A new,properly designed one is not in the cards, hmmm maybe if I could get a hold of an old Pemberton.
As for air lifts, yes not particularly efficient, depends on free air and height of surface above air nozzle. Best I've heard of is around 65% efficiency. However they do work well enough to help.
Anyhoo, thanks for the comments and ideas.

 

[zdas04]

More pressure does not necessarily give you more pumping hp. The function maxes (for water with atmospheric suction stream and 50 ft of discharge hose) at around 35 psig power water. City potable water system works just fine.

David

 

[logoslarry]

Interesting. I was thinking that the pressure drop at the nozzle exit would be higher with higher pressure upstream of the nozzle. I guess I'll have to dig out some of my old texts. Here we go, O'Rourke, "General Engineering Handbook" 1940 pg 813. To paraphrase, ejector usually uses a small quantity of the working medium at high pressure, to pump a large quantity at low pressure. Efficiencies between 15 and 30%, capacities to 700gpm. Well I don't expect I'll get anywhere near that. I'd be satisfied if I could get another 10 or 20 gpm, and stunned if I got 100gpm.

 

[Compositepro]

The tricky part in design is matching the size and divergence of the exit nozzle with the divergence of the jet. If the exit nozzle is too large you get backflow around the jet. If too small you get back pressure from the restriction. A converging/diverging nozzle is more accommodating of changes it jet shape. The jet shape will change with the amount of fluid it entrains. Entrainment flow rate will depend on inlet and outlet restrictions as well as nozzle shape. I've used water hose nozzles to make some very effective eductors. Straight pipe will not make an efficient jet pump.

 

[logoslarry]

Hey Compositepro,
Thanks for the comments. I am aware of the design criteria for a proper venturi, unfortunately we don't have the tooling or capability to make one,nor the budget to buy. Wish we did, but it is a poultry processing plant after all. The MBAs running the place think they save money by skimping on the maintenance budget and materiel.
I hadn't thought of using water hose nozzles, makes sense though.

I do recall a fellow who emptied his pool with a length of pipe with a water hose attached to the inside of the submerged part. Took him most of the day and a case of beer, but it did work.

 

[zdas04]

Ever see a blue plastic water bed pump? Connect the power to a garden hose, exhaust to a hose out the window and suction shoved into the bed. Empty a queen-sized bed in about half an hour. Made out of plastic and cost $12. Trouble is finding a waterbed store today.

David

 

[logoslarry]

Hey zdas,
Yup, had one for years,some years back. Aaah, the memories, best dang bed I've ever had. Mine took somewhere around 1 to 2 hours to empty, probably because I had to run about 20 ft of 1/2" hose both ways to the bathroom. Size that eductor up 3 or 4 times would probably be perfect for the sewer.
A few years back I did use a homemade jet to clean out the last inch or so of oil out of a compressor crank case, gravity drain was a bit higher than it should have been. Worked pretty well, though I did have to drape a rag over the discharge bucket, oil came out more as a mist than a stream. Took most of the sludge out too.

 

[logoslarry]

Hey all,
Did some experimenting the other day. Changed suction diameter to 1 1/2 " using air no real change in flow. Hooked up to 1000psi water line emptied 5gl bucket in 5 seconds max, so 60 gpm. Quite satisfied with that. Boss won't let me try steam,dang it.
Here's a properly designed one that looks quite nice,

http://www.youtube.com/watch?v=HHa42cY0QZI&feature=related

 

[logoslarry]

Yeehaw. Used the jet pump the other night, sewer backing up. Cleared it in about ten minutes, could actually watch the level dropping.It's fun to make something that works.