PDA

View Full Version : need self priming pump ideas



Mcgyver
01-31-2010, 07:26 PM
any ideas on commonly available (read cheap) low volume, self priming pumps? a project of mine, still virtual, is a central coolant system and I'd like a centralized return pump. I thought of making one, piston style, but life would be easier if i could just buy (cheap like). I've decided to go overhead with the plumbing, so the challenge is potentially 8' of head

J Tiers
01-31-2010, 09:25 PM
any ideas on commonly available (read cheap) low volume, self priming pumps? a project of mine, still virtual, is a central coolant system and I'd like a centralized return pump. I thought of making one, piston style, but life would be easier if i could just buy (cheap like). I've decided to go overhead with the plumbing, so the challenge is potentially 8' of head

Probably less than that, as the descending pipe will restore some head, leaving the difference, plus frictional head. Unless you are going to put it in a sump area.

Dr Stan
01-31-2010, 09:49 PM
something like this? https://www.surpluscenter.com/item.asp?UID=3524013119492446&item=2-1284&catname=water

You should also be able to find a similar one at TSC, Rural King, or other farm supply store.

JCHannum
01-31-2010, 09:51 PM
Where do you want the pump to pump to and from? If it is a cental system with a return sump that will always have a liquid level, a self priming pump will not be a necessity.

JoeCB
01-31-2010, 10:57 PM
those garden pond pumps might do the trick. come in several GPM sizes. I had a scrap one apart and was surprised to see that all the internals are made of ceramic, very wear resistant. the drive was thru a magnetic coupling so that the motor was completely sealed from the pumped fluid. Unfortunatly I don't remember the maker of that particular pump.

Joe B

Mcgyver
01-31-2010, 11:27 PM
the idea is there there's a well or sump somewhere off each table drain.... with the table in its lowest position, the well is close to floor level. I want a central pump to drain these wells. Lines will go from each well to the ceiling and across to the tank. The plan is to either manually or via solenoids open only the lines at the active machine. The lines will initially be full of air and go from close to the floor to the ceiling, hence the 8 foot head.

Dr Stan
01-31-2010, 11:44 PM
Something to consider. When I worked for CMI we had a central coolant system that was full of bacteria and caused many staph infections. If you have one machine get nasty it is fairly easy to deal with, a central system however is much more problematic.

Mcgyver
01-31-2010, 11:49 PM
Stan, that's one of the objectives of single tank, put in the infrastructure to keep in clean....anyway, lots more Q's and points for the overall idea....for another thread, hopefully I can get the self priming pump thing solved in this one :)

the one you posted and Joes, are submersibles, so won't be self priming :(

JCHannum
02-01-2010, 12:09 AM
The submersibles are self priming as long as there is enough liquid to cover the inlet.You could also have pump close coupled to the outlet of the tank.

Install a check valve on the vertical lin directly after the pump discharge, and the line will remain full.

Mcgyver
02-01-2010, 12:20 AM
i can only see a submersible working if its sitting in the sump, which i want to avoid (one pump instead of 6). also i don't want coolant in the lines.....it needs to circulate through my bug killing stuff.

i'm no pump expert, but i thought self priming meant the pump can be dry and above the liquid...?

I can make one I think, but was hoping I'd missed some common home/farm/yard application that uses them so i could take advantage of consumer mass production pricing

Evan
02-01-2010, 12:59 AM
Sounds like a centrifugal sump pump would be the thing to use as a sump pump.

http://ixian.ca/pics7/sump.jpg

$59 at C-Tire.

Mcgyver
02-01-2010, 08:50 AM
Sounds like a centrifugal sump pump would be the thing to use as a sump pump.

http://ixian.ca/pics7/sump.jpg

$59 at C-Tire.


A submersible or sump pump wont work.

The pump isn't in the sump, its by the tank with lines going across the ceiling and down to a variety of sumps. only the lines to the active machines will be open, otherwise it would just suck air, but unless I'm missing something this precludes sump pumps or submersibles. The whole idea is there are going to be a lot of sumps and its 1) too costly to put a pump in each, and 2) a submersible or sump pump will stop pumping when the liquid is gone, leaving lots in the lines...I want the lines cleared so the coolant doesn't sit there and rot

no ideas on a self priming pump?

Luke55
02-01-2010, 09:47 AM
roller pump they are self priming and can remove the last drop. Expensive to buy but easy to built

JCHannum
02-01-2010, 09:56 AM
Possibly a vacuum system of sorts acting on a closed tank could satisfy your needs.

A type of pump that develops good suction and can be run dry is the diaphragm pump. Wilden and a couple of other manufacturers make small air operated pumps that often turn up on eBay at low prices. Several are of all plastic construction for chemical resistance. There are also some motor driven diaphragm pumps and shop fabrication is not out of the question.

Another thing to explore is a peristaltic pump, they can be home built, develop a good suction and can be run dry. The pumping member is subject to wear, but replacement is simple and inexpensive.

Duffy
02-01-2010, 09:59 AM
Second the roller, (peristaltic,) pump idea. They can run at any fairly low speed and the output is according to tubing size. But Mike, you will HAVE to build it, unless you steal it from a neighbor's dialysis unit! There were plans for one in a back issue of HSM. Tubing life is short, silicone-type preferred, but it is quick to change. The output that you need is probably only a couple of liters/min. Duffy

aboard_epsilon
02-01-2010, 10:09 AM
chemical dosing pumps / swiming pool chlorinators is are self priming with check valves ..

but dont pump an awfull lot .

a 5 ltr a min max
2 metres suction height

you dont say how much you want to pump .

i have a few of these

http://www.etatron.com/id450.htm

picked up from car boot sales and ebay ..they are very dependable ..very adjustable ..and perform without fault

all the best.markj

Mcgyver
02-01-2010, 11:25 AM
thanks Luke, a quick google shows they are similar in action to a vane pump.....when you say easy to build, can you point me a bit...any designer or builders notes? thanks

Mark & Jim, thanks for the ideas....I don't actually know how much i need to pump:o , I need to measure that, but is fairly low...ie even an aggresive flood on an open manual machine is not many gallons per minute

Duffy, i'm confused ..... a roller pump and a peristaltic pump are not the same thing are they?

roller http://www.google.com/patents?id=B-k_AAAAEBAJ&pg=PP2&source=gbs_selected_pages&cad=0_1#v=onepage&q=&f=false
peristaltic: http://en.wikipedia.org/wiki/Peristaltic_pump

aboard_epsilon
02-01-2010, 12:38 PM
the peristaltic pump is the same sort of pump they have on kidney dialysis machines ..

i don't know if these things have any head ..or suction.

all the best.markj

kendall
02-01-2010, 12:42 PM
A small automotive fuel pump? different styles and types available, from a captive piston type where the piston is driven by alternating magnetic fields (click in use until fluid fills them, then they quiet down) also the rotary vane type. The frame mount, and in tank mounts.

I've used the in-tank style for a coolant pump befor and it worked nicely, 5 gallon bucket, with a hardware cloth pre-filter, window screen after that, then a foam rubber final filter with the pump in a 'too small' hole cut in the foam so fluid -had- to come through the foam. draining the lines was managed by the simple method of using a loose fitting hose on the pump. I lost a percentage of pumping efficiency because of blow-by, but when pump was turned off the fluid in the lines siphoned back into the bucket.

Frame style pumps are cheap enough that you could afford to put one at each machine, in my case I had a lot of donor cars in and out of the shop so I'd snag them.

Ken.

RancherBill
02-01-2010, 02:15 PM
A submersible or sump pump wont work.

The pump isn't in the sump, its by the tank with lines going across the ceiling and down to a variety of sumps. only the lines to the active machines will be open, otherwise it would just suck air, but unless I'm missing something this precludes sump pumps or submersibles. The whole idea is there are going to be a lot of sumps and its 1) too costly to put a pump in each, and 2) a submersible or sump pump will stop pumping when the liquid is gone, leaving lots in the lines...I want the lines cleared so the coolant doesn't sit there and rot

no ideas on a self priming pump?

Maybe I am lost.

You want the lines / machines empty so that implies that it has drains back to a common sump. So, IMHO, you need a common sump, a distribution device that allows you control flow to individual machine and a drain device.

A drain device is a solenoid valve out of a dishwasher, for example. Get a Normally open one. Put an air line on your distribution center. When the system is OFF the valve is open and allow air into the system facilitating draining by breaking the coolant line vacuums. When you turn the pump ON the valve closes and allows the coolant to go where it supposed to go. Turn it OFF and the valve opens air comes in and it all goes back to the common sump.

Or maybe I didn't have enough coffee the AM.

camdigger
02-01-2010, 03:28 PM
thanks Luke, a quick google shows they are similar in action to a vane pump.....when you say easy to build, can you point me a bit...any designer or builders notes? thanks

Mark & Jim, thanks for the ideas....I don't actually know how much i need to pump:o , I need to measure that, but is fairly low...ie even an aggresive flood on an open manual machine is not many gallons per minute

Duffy, i'm confused ..... a roller pump and a peristaltic pump are not the same thing are they?

roller http://www.google.com/patents?id=B-k_AAAAEBAJ&pg=PP2&source=gbs_selected_pages&cad=0_1#v=onepage&q=&f=false
peristaltic: http://en.wikipedia.org/wiki/Peristaltic_pump

The short answer is no. Roller pumps allow the roller to be in contact with the fluid, Peristaic do not allow roller contact with the pumped fluid. They work by rolling a tube flat and count on the rebound of the tube (shape memory?) to create the suction.

What you're asking is a pretty tall order. Few pumps will create enough vacuum on the suction side to lift fluid 8'. To lift fluid out of a machine sump across the shop your pump will have to create a vacuum in the suction line sufficient to lift the fluid up. Once all the suction line is full, the lift/fall of the fluid almost cancel out if the vacuum at the top is low enough the liquid doesn't boil or cavitate. IIRC, that`s why most pumps are limited to 25`lift, - that`s where water boils at standard temps and limits the amount of vacuum that can be drawn.`

Even the so called self priming impellor pumps need some water in them to make a seal between the housing and the impellor. As a further caution, because they are reliant on having some liquid to run, the seals need fluid for lubrication too. Running an impellor pump completely dry can burn the seals out of it.

Impellor pumps are typically `low lift or `flooded suction`pumps. High lift pumps, in my experience are close head clearance plunger pumps of relatively large displacement.

aboard_epsilon
02-01-2010, 03:49 PM
i don't know ..never tried

i think you can increase the amount of suction on those pumps i linked to ,just by making new ball valves on them

the originals have weak Biro springs ..so build a couple of valves that have stronger springs ..and you may get your 8 foot.

all the best.markj

Evan
02-01-2010, 04:06 PM
There is a further problem that makes using a single pump impossible unless some additional gear is added to the system. If you have a pump that can create enough vacuum to lift the coolant and draw it all the way to the collection point the moment that one of the lines goes empty you lose your vacuum and cannot empty the rest of the open lines.

The only certain way I can think of to make it work is to have local check valves on the return lines at the machines along with forward check valves on the supply side of each machine's coolant system ahead of each machines coolant system. The return line to the main sump is also connected to the supply line at each machine with a check valve that permits flow from the return line to the supply line. Each machines sump tank is configured to collect used coolant only.

To supply coolant to the machines use a centrifugal pump of whatever type that will handle the flow requirement. When the centrifugal pump is turned on all the machines that have open delivery lines will then have coolant pressure created at the main sump. Each machine's coolant pump is arranged as a scavenge pump to deliver collected coolant into the return line from the local sump with a check valve as mentioned.

To empty the system a small (and cheap) poly air line is run to each machine and connected to the return line just after the scavenger pump check valve. Air is admitted at around 15 psi and blows down the return line and the supply line via the cross connected check valve.

jdunmyer
02-01-2010, 09:09 PM
I'm with Jim Hannum, use a diaphragm pump. I have a couple of air-powered diaphragm pumps that work great and will pump anything that will fit through the ports.

However, Evan is correct, you'll need to close off the individual lines as they empty, perhaps using a float valve. Unfortunately, that alone won't empty the line.

firbikrhd1
02-02-2010, 11:38 AM
Having been a firefighter for 28 years when you mentioned self priming pumps the first thing that came to mind was what we use on Fire Apparatus. The main pump on Fire Apparatus is centrifugal and requires the pump to be full before it will pump water. When we draft from a static water source the pump is empty so we use a "Primer" to get water from the static source (lake, canal, ocean, whatever) to fill the centrifugal pump. These primers are rotary vane pumps and will move both gases and liquids. For Fire Apparatus to pass certification tests the Primer Pump must lift water 10 feet through 20 feet of suction hose, so a rotary vane pump would most likely work for your lift distance.
You might be able to adapt a Primer from a Fire Apparatus or, adapt an air motor such as one used in an Impact wrench. I don't know what volume you need but I can tell you that the Primer Pumps in our Apparatus could lift from 10', fill a 6" hard suction hose and prime a 1500 GPM pump in about a minute. The rotary vane pumps are not particularly large and are powered by what looks like a starter motor for a car. There is a specified time requirement to pass Pumper Certification tests but I don't recall what it is. The old type required a tank which fed oil or water into the pump to lubricate the vanes until water reached them, the new ones have vanes made of a material that no longer requires this lubrication. In your application I assume your coolant has enough residual lubrication properties that it would be a non issue.

derekm
02-02-2010, 05:06 PM
Another issue it isnt just liquid. Its swarf fines you need to move as well.
My centralised coolant system drains to a common sump. the common delivery line opens into the high points of the drain system (via a restriction to ensure pressure), to ensure a high enough flow rate to keep the swarf fines on the move. When the pump is off the delivery line drains down.
it works with only 20 litres of liquid and it keeps the coolant aerated

firbikrhd1
02-02-2010, 11:26 PM
Derekm, your point is well taken with regard to "fines" and the works of a rotary vane pump, perhaps any pump for that matter if they aren't filtered out.

This may be over complicating the matter but, I propose no pump at all but rather a system where the coolant drains by gravity to a receptacle which, when full, activates a float switch/sensor which opens an air flow control valve pressurizing a lower receptacle with very low pressure air closing a check valve on the receptacle inlet line and forcing the coolant up to a reservoir where it is gravity fed back to where it is needed. A second float switch/sensor in the upper tank may be necessary to achieve proper timing between the two tanks. This would require an air source, float switch or sensor, regulator, an electric air control valve, check valve, manual shut off/flow control valve and some plumbing, but might do the job for not too much money. Perhaps a couple of pressure sprayers like those used for gardening could be utilized for reservoirs.

A centrifugal pump could be used in the lower reservoir instead of air pressure provided the pump is kept submerged at all times. The problem of particles wearing the pump might still be an issue.

This system is not unlike the lubrication system on some two cycle outboard motors with oil injection except that an electric pump in the main oil reservoir pumps oil to a secondary tank at a higher level where the oil is fed to the oil injection pump by gravity, the difference being that air pressure is used to move the coolant to the higher secondary tank instead of an electric pump. Level sensors and switches are used to time the process.

J Register
02-04-2010, 01:46 AM
One thing that might work would be to use a jet from a well pump. What I'm thinking is to submerge it in your main coolant reservoir to keep it primed, set it up to draw from the return lines, and have the output just shoot back into the reservoir. You would divert some of the output from the coolant pump to run the jet.

I've never tried this, and don't know how much flow and pressure it would take to actually empty the return lines, but thought it worth throwing out as a possibility.

Jim

Metalmelter
02-04-2010, 11:40 AM
any ideas on commonly available (read cheap) low volume, self priming pumps? a project of mine, still virtual, is a central coolant system and I'd like a centralized return pump. I thought of making one, piston style, but life would be easier if i could just buy (cheap like). I've decided to go overhead with the plumbing, so the challenge is potentially 8' of head


Self priming your gonna pay for since they are either a gear pump (expensive) or they are the flexible impeller style (neoprene and much cheaper but not so resistant as bronze gears).

I'd look on eBay for a while. That's where I found mine and the $25 or so bucks I spent for an older bronze geared pump was well worth the shipping price (flat rate priority mail ~$10). You sure you need self priming?? Re-design the system to take advantage of cheaper style pumps if you can ;)

Mcgyver
02-05-2010, 02:31 PM
thanks for all the ideas....Metalmelter, my design tree on coolant return, such that it is, came to a decsion point on pumps...centralized or not. Centralized is better, 1 pump vs 6, but it had be self priming for about an 8' distance (just above floor to ceiling) it had to be cheap to make or buy

what i'm hearing for lots of guys more knowledgeable on pumps than i is good luck.

One idea before i abandon centralization....

I started thinking how a shop vac works, it creates enough air flow that a certain amount of liquid gets carried along. So what about a vane pump (they can be used for liquid and air, right?) or small centrifugal blower? it might be possible with a vane pump to If I made a current sensing controller, the vane pump could be two speeds; high if its moving air, low if it actually gets saturated.

This might create new problems; noise, exhaust is laden with moisture, etc.

Plan B is abandoning central return and pumping from each machine. even the little giant pumps for fountains and what not start becoming more than I'd like to spend; maybe there's a cheaper way. Evan's idea of blowing out the lines had occurred to me; with a solenoid to cut off flow to the pump and a quick blast last of the coolant should be cleared....

and leads to firkiks idea; collect a bunch of coolant, close the input and send the rest home under air pressure....if i could identify a supply of cheap solenoid valves I'm liking this one!

Evan
02-05-2010, 02:51 PM
Found you a source of electric valves. Nice prices too. $18 for what you need isn't bad at all.

http://stcvalve.com/Process%20Valve.htm

madman
02-05-2010, 03:39 PM
I used to mess around with stinky coolant. We ended up (in a shop i worked art) with a central system that still needed LOTS of maintenance. For youre application perhaps a pump is NOT even needed. I suggest a stainless holding tank (old water heater tank is OK) Then run youre lines to machines (i would utilize Pex tubing cheap and leak proof) If you hooked up a small valve arrangement at each machine (from this point you would have another coolant hose mount valve arrangement Garden hose nFaucet for a cheap example)and you were to lightly pressurize the holding tank(water hearer tank) you could have a easy coolant flow system that would meet youre mcgyvery low dollar side LOL ,Just a idea take care Mike madman (Mike) OH PS spend a Little money on a stink free coolant to start with synthetic> I always liked Blaser from Switzerland a

camdigger
02-05-2010, 04:29 PM
Mike might be on to something. If you're going centralized, you could gravity feed a return run along the baseboards to a sump, then supply pressurized coolant to the machines. One pump and a throttling control valve at each machine. A self starting siphon in the machine sump might be the way to go. It does mean your centralized tank would have to be low profile or in a pit like the oil change pits in some garages.

Mcgyver
02-05-2010, 04:36 PM
Evan, thanks for the valve idea....Mike, the challenge isn't the flow to the machine, its getting it back to the holding tank.....on maintenance side, thats exactly why i want 1 tank. Plans is to build it UV light and a bubbler and maybe skimmer and maybe even a zapper (that idea needs work) to have a fighting chance....was it Rich who said here he was in his 7th year on coolant treated this way? the whole idea of centralization is to put these things to get long coolant life

Mcgyver
02-05-2010, 04:41 PM
Mike might be on to something. If you're going centralized, you could gravity feed a return run along the baseboards to a sump, then supply pressurized coolant to the machines. One pump and a throttling control valve at each machine. A self starting siphon in the machine sump might be the way to go. It does mean your centralized tank would have to be low profile or in a pit like the oil change pits in some garages.

goes against one the constraints....return line has to go to the ceiling; shop is too crowded to trench, lots of machines away from a wall etc. Delivery imo is easy in that its easier to push liquid than pull it....,a centrifugal pump and manifold with either manual or solenoid controlled valves with tubing to each machine...or maybe i'm missing what you're saying

Evan
02-05-2010, 08:49 PM
UV probably won't work if you use the regular water sterilization units. It won't penetrate far enough. It will though if you flow it down a stainless steel sheet as it returns. I would just add a bactericide, perhaps chlorophyll.

madman
02-05-2010, 10:39 PM
Use Gravity. flow from machine drains to a floor tank and then either pump that coolant (little giant pump or parts washer pump ) or you can also pressurize and force it back to the main high tank, you do not need a lot of pressure to do this. Good Luck mike

2ManyHobbies
02-06-2010, 12:14 AM
If you had an aisle or wall with machines on it, you could use a common sump and reduce from 6 pumps to a smaller number. It sounds like you need ~8 foot of head pressure (or lift) initially, but if your reservoir is lower than the machine sumps siphoning will do most of the work.

Here are the problems I see:

You will have to pull and maintain a vacuum on everything of at least 96" WG.
You will be lifting or pulling dirty fluids and possibly trash which will settle overhead and require regular flushing.
You will require some sort of fluid linkage back to your reservoir meaning:

You are spending some good money on a trash capable pump that can lift 96" WG.
You are spending more money on your supply pump so that it can overcome a 96" WG vacuum and still provide 8' head pressure plus friction plus any required force at the machine.
You are spending money on a 3rd pump to move from a central return sump back to your reservoir against 96" WG vacuum.



Can you situate your reservoir below the level of the sump on every machine? If so, just put together a quick disconnect return line and hook it to the active machine. Gravity will do the rest, and when you get done, you can either hang your return and let it drain, or clear it with the air compressor.

If you have extra capacity on your supply pump; build a venturi pump to clear the sump on each machine. It would probably be good to put a check valve on each pickup to prevent unexpected flooding though. At the end of the day, open all of your supply valves, plug the air compressor into the supply side, and wait till all the lines and sumps go clean. If you don't have the capacity in your supply pump, you could always substitute the air compressor.

Mcgyver
02-06-2010, 12:51 AM
UV probably won't work if you use the regular water sterilization units. It won't penetrate far enough. It will though if you flow it down a stainless steel sheet as it returns. I would just add a bactericide, perhaps chlorophyll.

I will probably go with an aquarium style that circulates liquid over the light, cheap and solves the problem that you mention; that just having a UV light pointed at the surface wont do much

madman
02-07-2010, 03:13 PM
Good luck with the coolant. Thought of using a ozone generator for our large system also copper helps to negate the smell .(I was told this new use for pennies)>>?????

Toolguy
02-07-2010, 04:17 PM
If I am understanding correctly, all of this is an effort to make it easier to deal with bad coolant in multiple machines. There is an easy solution. Most times the coolant starts going bad when it gets way lube mixed in. Often it's the sulfur in the way lube that causes problems. You can't keep the machine oiled without getting way lube in the coolant. My solution was to get Hangsterfer's coolant and compatible way lube. I have been using this combination continuously in the mills, surface grinder and cutoff bandsaw for about 10 years with no problems. I never have to change out the coolant, just add more as needed to keep the correct ratio of coolant to water. My shop never smells bad and coolant maintainence is a few minutes every month or two. Also, if this coolant is left on a machine overnight, it leaves an oily film that protects the machine from rust rather than making it rust. The main thing is it's very good for machining parts. Might want to check it out - all the complicated plumbing may not be needed.