Take The Off-The-Shelf Steam Car Water Pump Challenge!

Hey y'all,

I have taken the Stanley water pump valve/port/plunger-dia/stroke dimensions, graciously supplied by accomplished steam veterans, y'all know who you are, and I thank you most sincerely, and have worked them into a blueprinted feed-water pump design which looks very buildable and workable.

Despite the extensive design/blueprinting work, I am still slightly disturbed by the criticism, from a few quarters, to the effect that people who design a custom shop-built feed-water pump for a steam car are time-wasting idiots, and doomed to fail, because off-the-shelf steam car boiler feedwater pumps are easy to find, and work perfectly. Several people have referred to Harry Schoell's unsuccessful attempts at designing a successful custom high-pressure boiler feed water pump, and have suggested that any similar efforts on my part will meet with similar failure.

I have researched this, and have found that off-the-shelf commercial high-pressure feed-water pumps -- generally designed for pressure washers -- tend to be incapable of handling particulates, traces of oil in feedwater, feedwater at 150 degrees F, etc.. Several manufacturers and retailers of such pumps have explicitly advised against using their products to feed a high pressure boiler in a condensing steam power system. They tend to be designed for 100F or less, 60+psi cold/clean feedwater from city water mains via a hose bibb, and will not suck water out of a tank at atmospheric pressure, let alone at 150F.

Your challenge, if you choose to accept it, for steam "experts" who think that designing/building a custom feedwater pump for a steam car is a stupid waste of time, when equivalent off-the-shelf pumps are easily available, is to show me an off the shelf feedwater pump that can flow enough water for a 750 lbs/hr boiler at 500 psi & 1100 rpm, without any problems with traces of particulates or oil, and with zero pressure on the feedwater inlet [IE, pulls water from 1 ATM tank] - and without costing like $1000-$2000.

Cue "Mission Impossible" music.

I have researched this, and found nothing, but maybe I missed something?

To win this challenge, you have to supply the name/manufacturer/model-number of the suitable pump, plus price and URL of manufacturer or retailer from which I can order this pump on-line via point-and-click. "Call SmezCo Pumps Inc at 1-555-555-5555 and ask for Ned Murbly in the Extra-Special Custom Orders Department" equals fail.

I am looking for the easily/inexpensively available off-the-shelf standard production pumps which are supposedly out there somewhere. Costly/custom/special-order stuff, no. If it's costly/special-order or modification-needed, then I'd rather build my own custom design.

Extra credit for anyone who knows an off-the-shelf source for ~150 psi engine-driven gasoline pumps delivering 4gph at 1000rpm.

Peter



Edited 1 time(s). Last edit at 06/18/2018 03:35AM by Peter Brow.

"Several manufacturers and retailers of such pumps have explicitly advised against using their products to feed a high pressure boiler in a condensing steam power system. They tend to be designed for 100F or less, 60+psi cold/clean feedwater from city water mains via a hose bibb, and will not suck water out of a tank at atmospheric pressure, let alone at 150F. "

They are just covering themselves liability wise. If they said you could do it, and you did it, and for some reason it failed to pump, and the boiler went dry and failed, they could be liable. Remember, these are simple diaphragm or piston pumps built using jellybean parts. They'll have off the shelf buna orings, that won't mind 150f. There aren't going to be any special plastics or seals in these pumps that can handle 100f, but fail with 150f. The oil MIGHT pose a problem if they used plastics that are incompatible with oil, but that would really only be a possibility with the seals, which will be standard sizes and can be replaced with seals of a material that can handle water and oil. As far as debris goes, you should have a filter upstream of the pump anyways! Priming may be an issue, slightly exacerbated by the high speed these pumps run at, but is going to be something to deal with for most styles of high pressure pump. You would want a low pressure in tank pump feeding the high pressure pump to really eliminate that issue.

I wouldn't ask the manufacturer of a pressure washer pump if it can be used for boiler feed, they'll say no to cover themselves. You have to look at the materials and construction of the pump and decide for yourself if a pump built like this could handle boiler feed water.

However, there is a middle ground between custom machined pump and a pressure washer pump. You can buy an off the shelf gear motor and hook it up to an off the shelf industrial air cylinder. You might have to use a hydraulic cylinder for your pressures though. This is what I did for my circulation pump. Industrial air cylinders (and maybe hydraulic? I never really looked into them) can be bought in stainless steel, and are usually rated for 190f or higher. Then you just plumb some check valves to it, and fab up a quick scotch yoke to connect it to the motor (I 3d printed the parts to connect the cylinder to the motor) and you're in business. The machining requirements for something like this are very loose.

A boiler feed pump is very simple to make and can be made in just a drill press. Just a ram in a housing with external off the shelf check valves, capable of many thousands PSI. Engine driven is most efficient. I doubt you're going to find an inexpensive off the shelf pump adequate for feedwater. Most likely, the pumps you do find have built in check valves that are often very difficult to remove for cleaning and are proprietary to the pump. Have a look at the Hypro pumps, great pumps when they are working, if the checks foul, it takes a special tool available from Hypro to remove them for cleaning. And an adequately sized Hypro is about 350 dollars.

Keep in mind anything involving distilling various qualities of water to steam or simply the water to create steam will always have a certain amount of minerals and impurities to deal with, unless distilled pure water is used, which isn't practical. Pumps need to be robust and easily serviced, and engine driven for highest efficiency.

This is how simple they can be and still be capable of thousands of pound per square inch:



-Ron

Hi Zimirkin,

Makers and retailers of pumps may indeed be concerned about liability when they recommend against using their products for oddball applications like light steam power systems; that thought has occurred to me too. I have heard of a number of people successfully using off the shelf high-pressure reciprocating pumps in modern/project and even antique steam cars, but have not found information on specific makes or models of pumps,at least no makes/models which are currently available. Yet I have seen frequent recommendations, over the years, to "just use an off the shelf pump". Which ones? No answer. It's frustrating. Maybe there are some off the shelf pumps which work so well that their users keep the make/model info a secret -- keeping the good stuff to themselves? Well, designing and building my own fuel and feed-water pumps is lots of fun, and I am learning more in the process.

Hi Ron,

Right, pumps are really not that hard to design or build, once you know some proven design parameters/features and follow them. Thanks for your encouragement. Actually I have completed blueprinting a feed water pump, and mostly finished the blueprints for a fuel pump. Mainly I was looking for info on off the shelf pumps as a backup/"Plan B" in case I run into problems with these designs. But as near as I can tell, serious problems look unlikely. Maybe some slight mods will be needed after initial tests. Only building and testing will tell for sure of course.

Your pump drawing is similar to my design in some ways; both are built-up mostly from standard plumbing fittings with minimal [and simple] machining. I even designed it for easily-adjustable valve travel [for tuning experiments] and easy valve and seat removal and replacement. I worked out a sealed oil-bath case, with splash lube, to keep airborne dust/grit off of the plungers and out of stuffing boxes, and to continuously lube the plungers and their drive mechanism. A rocking shaft enters the case through an oil seal, and levers integral with the shaft move the pump plunger via long links, one on each side of the plunger, so that the plunger is not "tilted" or thrust sideways as in Stanley "H" pumps. The long links, and their short stroke, means that thrust is applied in almost exactly a straight line to the plunger. Everything is fitted with oversized "Oilite" sintered bronze bearings and journals cut from precision shaft stock, located with thrust bearings and shaft collars.

Years ago there was a discussion about O-ring sealed pump plungers, I think on either this Forum or John Woodson's forum. Some people said O-rings are only good for static or barely/occasionally sliding/rotating seals, but not for continuous motion. OK for a hand pump, but not for a power driven pump, which needs stuffing boxes. My hand-lever hydro-test pump has o-ring plunger seals. Others said that the O-rings are inexpensive and easy to replace, and have lasted acceptably long in power-driven experimental/home-built pumps. I am keeping an open mind on the subject. I seem to recall that some off-the-shelf pumps use "Option C: None Of The Above": cup seals on the inboard ends of the plungers. My current design uses packed stuffing boxes where plungers enter the valve bodies, with 6 packing rings. There are O-ring seals [nitrile] on the adjustable stop rods for the valve balls, however.

Somebody, I forget who, might have been you, recently suggested locating the fuel pump below the fuel tank for easy gravity-priming. For that and a number of other reasons, I am designing separate fuel and water pumps, instead of putting all the pumps in one unit the way Stanley and some other traditional steam cars did. Otherwise the design is based closely on the fundamentals which Stanley used, with modifications in an effort to correct, or at least minimize, various problems which have been reported with Stanley and other steam car pumps. Looking forward to seeing how well [and if! ] it works. At first I plan to fit hand levers for testing the pumps and other powerplant parts.

Peter

"Years ago there was a discussion about O-ring sealed pump plungers, I think on either this Forum or John Woodson's forum. Some people said O-rings are only good for static or barely/occasionally sliding/rotating seals, but not for continuous motion. OK for a hand pump, but not for a power driven pump, which needs stuffing boxes. My hand-lever hydro-test pump has o-ring plunger seals. Others said that the O-rings are inexpensive and easy to replace, and have lasted acceptably long in power-driven experimental/home-built pumps. I am keeping an open mind on the subject. I seem to recall that some off-the-shelf pumps use "Option C: None Of The Above": cup seals on the inboard ends of the plungers. My current design uses packed stuffing boxes where plungers enter the valve bodies, with 6 packing rings. There are O-ring seals [nitrile] on the adjustable stop rods for the valve balls, however. "
O-rings are in use in constantly moving applications everywhere in industry and they last plenty long enough. Plenty of industrial air cylinders use orings for piston or piston rod seals. The better ones use cup seals though, which are much better for this application. Cup seals may be a bit harder to find than orings, but they should still be very cheap. I've noticed that air cylinders are more likely to use cup seals, and hydraulic cylinders more likely to use orings. I'm not a cylinder designer, so I don't know why that would be. I do know that air cylinders that use orings also have a plastic wear ring, so the orings aren't handling the side loads. The orings definitely do wear out faster than the cup seals, but they aren't bad.

As far as specific pressure washers, there's so much selection out there. This one on amazon is electric, goes to 1500 psi, and is approved for hot water. You're going to want to filter the incoming water for any high pressure pump. All these pressure washer pumps are going to be extremely similar in construction and materials. The more expensive ones might have more easily replaceable seals or less plastic in the drive, but they're all gonna be piston or diaphragm pumps. If it was me, I would buy 2 harbor freight electric pressure washers, and use them in parallel for redundancy. However, my personal preference is to drive auxiliaries off of electricity. If you wanted to drive these off the engine, you can just buy the pump part itself and hook it up.

edit: some of those pumps are only rated for water temps of 110f. This is most likely for cavitation reasons due to the speed they run at. They should be fine at higher temperature if driven more slowly.



Edited 1 time(s). Last edit at 06/19/2018 07:32AM by zimirken.

CAT pumps have been successfully used in several steam projects.
Peter Barret's car started with a home made multi plunger pump. It was replaced with a commercial plunger pump, not sure if it was CAT.

Download several CAT repair manuals (some are better with details than others), but many usable ideas / tips, and ready made repair parts (ceramic plungers fer instance). CAT also has high temp seal kits (hot car wash).

[www.autowashonline.com]



Edited 1 time(s). Last edit at 06/19/2018 08:46AM by Scott Finegan.

The issues with cup seals is they only work in one direction. A pump needs to of course create a head pressure, but it also needs to create a vacuum to overcome the incoming check valve on the suction side. A cup would only push well and would leak on the vacuum stroke and introduce air. The main feedpump on the Tugboat is just a simple ram with two small O-rings of high durometer and installed in cut grooves of proper dimension (should only be 10 - 15 % compression on the ring, and there needs to be enough side clearance so the ring has room to be compressed into an ellipse, and the bore needs to be a mirror finish) and the hold up perfectly well, I replaced them once in eight years, and only did because I had the engine out mounting a flywheel and installing a generator.

Temperature is not much of a consideration as feed water from the tank is usually at ambient temp, unless it's being pulled from a hotwell in a condensing system for instance, or for Forced circulation.

Driving the feed pump mechanically and directly from the engine is by far the most efficient method requiring the least amount of horsepower. As Jim Crank used to say "Energy systems in series, I don't think so" Have a look under the hood of any car, all pumps, generators etc that require a significant amount of power for operation are driven directly off of the engine.

-Ron



Edited 1 time(s). Last edit at 06/19/2018 08:48AM by IronChief.

Thanks guys!

Yep, I've been around this block a few times before; Amazon, eBay, Northern Tools, Harbor Freight, and a number of other retailers -- I have browsed many, many websites in recent years -- have extensive selection and at first glance it looks like a piece of cake to just pick one, buy, and install, problem solved.

But there are lots of unknowns here. One of these websites had a customer question/feedback page for each of the dozens of pumps it listed, and for a couple of the pumps somebody asked about pulling water out of a tank to feed a boiler. The reply was that the pump would not develop suction on the inlet, and would only move water if fed with pressurized water from a city water system [usually something like 60-70 psig] via a hose. Does this only apply to that one particular model of pump? Or are all these pumps with hose bibb inlet fittings the same way? Can any particular pump be modified to suck water from an unpressurized tank, perhaps with optional parts from the manufacturer, or with home-built/modified custom valves? Do the listings for most of these pumps not mention the issue because almost nobody tries to feed them from an un-pressurized water tank?

Then there are head-scratchers like the note on the CAT pump website that their Model 310 pump [~$500] has a minimum pressure of 100psi. Does this mean that the delivery check valves are spring-loaded and won't open until the pump cylinder develops 100psi? Extra hp wasted even when the feedwater bypass is full-open? Or does that mean it needs to be fed 100psi water from a hose in order for the inlet checks to open? In some of these pumps, the valves need their springs to be properly located relative to the valve seat, and fitting lighter springs to cut inlet cavitation/vapor lock with hot water, or to avoid horsepower waste when running with bypass valve open, might not properly locate the valves.

I am going to cross this bridge if/when I get to it. Hopefully the custom pumps I have blueprinted will do the job acceptably and I won't have to spend hours on the phone or email trying to get answers to these and numerous other strange questions from busy manufacturers or retailers.



Edited 1 time(s). Last edit at 06/20/2018 08:35PM by Peter Brow.

Peter
Most of these three cylinder pumps as well as the Hypro Pump 3GPM 500 PSI 1800RPM pumps are available with different pistons and valves.
I have used two cyl Hypro 500 PSI as well as three cylinder Teel 1200 psi pumps .
My emergency low water pump on my 35 foot Boat was a three cyl Teel 1200 PSI pump driven by a ½ HP 12V DC motor. I ordered the pump with SS pistons and corrosion resistance valves.
Rolly

Peter,

The feed pump I used on the lawn tractor project was a 2,700 psi pressure washer replacement pump head I bought on Amazon for about 60 bucks. It was designed to be used with a 5 hp gas engine and fits a 7/8” keyed shaft. I coupled it with a strong 24vdc electric motor and it worked great as a feed pump. The RPMs have to be at about 3,500 or higher....that is crucial or it won’t work. The water on the suction side was gravity fed with a 3/4” id hose from a water tank elevated only about a foot above the pump inlet. The lower inlet pressure caused the pump to take about 10 seconds to build up full outlet pressure, but it worked great once it did. 2,700 psi will put water in your boiler very quickly! If you can live with the delay, it will work for you. Another solution is to put a small 12vdc transfer pump (Harbor Freight carries them) in-line at the inlet and that will give you the 50+psi to make it work as originally designed.

Jamison



Edited 1 time(s). Last edit at 06/21/2018 05:54PM by Arch-Tone.

Thanks Rolly & Jamison,

I am compiling a mental checklist in case I end up having to use an off the shelf pump as "Plan B". Now adding, check for available plunger/piston & valve options, and select from those, and check for minimum inlet pressure and any minimum RPM & flow rate, and/or delay in building up pressure. Also possible need for separate "booster" pump, up-line, to pre-pressurize feedwater. Long conference call with the pump factory salesman or engineer -- if they have the time. Lots of time & talk to buy/sell one pump.

These pumps need rotary drive too; tricky to do with a direct-drive style steam car engine. Best for that might be a right-angle drive on engine or axle, and a rotating back-to-front shaft with U-joints at each end; which has been done before. With that one, the alternator could be on the chassis instead of on axle. I don't like the right-angle drive part, though. Electric pump drive works, but that's a heckuva lot of watts and energy losses, as Ron & Jim Crank have mentioned. Also, I'm trying to avoid building a car that runs fine on the highway but drains the battery in extended stop/go/slow driving. Though I could just treat it as a "plug-in hybrid"; plug in to charge up batteries which run various auxiliaries, and the car goes hundreds of miles on a charge. Partly electric powered, partly steam powered. Hey, there's production cars now which run partly on gasoline and partly on remotely-generated electricity.

As a digression, the auxiliary-running batteries could be recharged via regenerative braking under some conditions. Would that make up for various electrical or other energy losses, would any savings be worth the extra equipment cost, so many questions there. A lot of highly-touted tech costs you $5000 extra to save $4000 worth of fuel over the life of the car. Regenerative braking does have the "cool factor", though.

Sometimes off-the-shelf components lead to a "design cascade effect", To use component A, you need to use B, and B requires C, then the engine and drive train have to be redesigned for this and that, and so on. Sometimes it ends up a "for want of a nail" story, sometimes it brings to mind the old adage "a camel is a horse designed by a committee". The committee in this case being the manufacturers of the various & sundry "not-entirely-suitable-for-steam-cars" off the shelf components. I have gone pretty far down the rabbit hole this way with some of my previous ideas for steam cars.

I have heard of Stanleys which were retrofitted with modern off the shelf water pumps, but maybe those were only used for racing or highway running.

There are three reasons auto manufacturers are going to electric auxiliaries. First, it allows you to move all these pumps and things out of the middle of the engine bay. Second, it reduces parasitic friction losses from spinning all these things even when they aren't doing work. Third, they are working to get rid of idling when stopped, but still need to drive the air conditioner and stuff.

Regenerative braking only makes sense when you already have a large electronically controlled electric motor(s) connected to the wheels, like in hybrid or electric cars. Braking dissipates as much energy as accelerating, and you would need an electric motor of equivalent power output to capture it. Plus some real sophisticated power electronics.

Also speaking of car systems. Many cars have dual fuel pumps. A low pressure pump in the tank feeds the high pressure in line fuel pump.

The deciding factor on if a given pressure washer pump can suck from a tank or not is the cracking pressure of the inlet valve. it must be low enough to operate under atmospheric pressure.

I run a Hypro on the tug and it simply pulls from the lake i.e. 0 psi or 14 psi atmospheric if one wants to be really technical about it

It requires quite a bit more energy to accelerate a vehicle to a given speed than it does to stop it. This is displayed well in heavy trucks, in stop and go driving the brakes aren't even needed, if the vehicle needs to stop, going slow, simply push the clutch in. The resistance to the road and mechanical friction will bring it to a stop. Nuclear comparison Cube law and all that.

Ram pumps like we are discussing here are extremely powerful - infinitely. Hundreds of PSI is being discussed as some sort of maximum, maximum pressure on a ram pump is the limitation of input force and structural integrity of the pump itself. To give one an idea on the amount of pressure than can be had, Water jet cutting machines use a multi piston ram pump at high speed and operate in the 50-70,000 psi range which requires many horsepower to generate, around 15 - 20 horsepower if memory serves me correct.

-Ron

IronChief Wrote:
-------------------------------------------------------
> I run a Hypro on the tug and it simply pulls from
> the lake i.e. 0 psi or 14 psi atmospheric if one
> wants to be really technical about it
>
> It requires quite a bit more energy to accelerate
> a vehicle to a given speed than it does to stop
> it. This is displayed well in heavy trucks, in
> stop and go driving the brakes aren't even needed,
> if the vehicle needs to stop, going slow, simply
> push the clutch in. The resistance to the road and
> mechanical friction will bring it to a stop.
> Nuclear comparison Cube law and all that.


Stopping returns exactly as much energy as was put in by acceleration. It's just super super easy to use friction to convert all that energy into heat. Using the friction of the drivetrain to stop creates just as much heat as using the brakes. In fact, that sounds like a very inefficient drivetrain.

Hopefully my custom pump design will work out well, and I won't have to deal with some pump manufacturer salesman or engineer telling me that it is impossible to have a plunger pump pull water from a 1atm tank and push it into a 500psi boiler at varying [0-1000rpm] pump speeds/delivery rates, because nobody builds pumps like that nowadays. At that point, knowing that Stanley built such pumps at least as early as 1905, my "Plan B" would turn into "Plan 9 From Outer Space", and I'd probably deliver a tirade something like this:

[www.youtube.com]



Edited 4 time(s). Last edit at 06/24/2018 04:23AM by Peter Brow.

One comment I neglected to post here earlier: engine-driven fuel and water pumps in a direct-drive steam vehicle deliver more fuel/water at higher vehicle/engine speeds and less at lower speeds. Thus they perform part of the boiler control function. Of course, fuel/water flow are not directly proportional to vehicle/engine speed for many reasons, so other controls are necessary. But engine-driven fuel/water pumps make the other controls simpler. That simplification more than compensates for difficulties in design/fabrication of engine-driven/proportional-flow fuel & water pumps. Current commercial pumps tend to be single-optimum-flow-rate designs, keyed to IC-driven or electrically-driven "full-or-zero"/ON/OFF rpm/pressure/flow-rate applications. In some ways, these "modern, therefore by definition better" pumps are actually more primitive than old-school steam car plunger pumps, with their unsprung check valves and "continuously/infinitely-variable" operating cycles.

Un-sprung check valves have a variety of other engineering issues/trade-offs/downsides, of course.

Gas-car company sales mottos/hype to the contrary, "perfect is the enemy of good enough".

My fuel and water pump blueprints are now finished; lotsa materials/parts to order; then machining, assembly, and test processes. Some parts/materials & assembly are already finished. I think the results will be good, but only building & testing will tell.

Peter



Edited 1 time(s). Last edit at 06/29/2018 05:32AM by Peter Brow.

It should be remembered that, unless the car has a transmission, an engine driven water pump should be sized to at least double the normal capacity. It's quite possible for consumption to outpace pump output when running in a long cutoff mode at low speed There are pictures of Stanleys parked alongside a mountain road with one rear wheel jacked up while the engine runs to refill the boiler. Jim Crank told me that driving a Doble in parades, or stop and go traffic, could be a tense experience....the engine needed long cutoff just to run smoothly but the low rpm weren't delivering satisfactory water flow. Jim says Doble came to the same conclusion as French and Staley, that it was necessary to decouple the pump from the engine. Doble used a steam powered feed pump whereas French and Staley employed an auxiliary engine, allowing them to connect the fuel pump and blower as well. I was looking at Karl's Citroen conversion at the last mini-meet, and could see he anticipated the same difficulty. He'd put a centrifugally actuated set of variable sheaves (like a snowmobile transmission) between the engine and the feed pump. This way the pump would run proportionately faster as the car slowed down.

Ken

Hi Ken,

Absolutely; excess water pump capacity is very important in steam cars. My current water pump design is the same displacement/mile as Stanley used in their later 4000-5000 lb big touring cars and sedans, but I'll be using it in a light roadster which should end up weighing well below 2000 lbs, probably around 1500 lbs.. So that equals or exceeds the double-capacity rule. I ran the numbers on how much steam (water) the engine could gobble per revolution under the worst conditions -- effective full-stroke admission due to ultralow rpm & engine valve leakage, plus high pressure [or low pressure with lots of ring leakage] and some other possible steam-hog factors -- and it looks like the pumps will more than keep up with water [and fuel] demand even then.

Driving the pumps by something other than the engine does have the potential advantage of reducing horsepower loss at the pump when the feedwater bypass is partly open. Another approach is to give a variable drive ratio between pump speed and main engine speed, while still driving the pump from the engine, the way that Karl did. One idea I looked at is fitting a Stephenson-link type drive to vary the pump stroke. Just move the curved/slotted link up or down for longer or shorter pump stroke. Making that automatic gets kind of complicated, though.

Peter.

To throw another ingredient in to the stew, so to speak. I was listening to a talk about model Locomotives from the President of Pennsylvania Live Steamers, 5 and 7 inch? gauge and the subject of feedwater was covered. Strangely, it appears that they do not use a bypass valve or even a variable stroke pump to control the feedwater. Rather they simply have a needle valve in between water tank (tender) and the feedpump that they adjust to the load and can be adjusted on the fly. They claim that it is easier to control and the performance increase is stark by administering feedwater this way and omitting the power requirements of returning unneeded feedwater to the tender in bypass.

Feedpumps always have a vacuum on the suction side regardless of the pressure on the output, worst case (valve shut) max vacuum is going to be 29.9 inches, not harmful in any way providing the seals on the piston are good, which they should be.

-Ron

Hi Ron,

This is indeed fascinating; I deleted from my previous post a speculation about throttling down the feedwater pump inlet to induce some cavitation and reduce the feedwater flow. I omitted it to avoid the usual criticism of my wacky speculations. But, interestingly, some folks actually already do this with good results. Yes, use good seals. Bad plunger/piston seals with a throttled feedwater pump inlet equals air-bound pumps, and/or, in my splash-lube pump, pump-box oil sucked straight into the feedwater/boiler. But pulling a hair more vacuum on the pump inlet stroke at low feedwater flows would mean much less feedwater-pump horsepower waste than the usual steam car bypass-valve feedwater control.

On the other hand -- there's always another side -- some years ago I did a mathematical analysis and found that with carefully-sized feedwater pumps, feedwater horsepower-reduction savings from reduced bypass losses would mean small overall savings, at least in a steam automobile.

On the third hand, it's simple/inexpensive equipment and continuously-variable control, a common combination in classic steam cars,

Definitely food for thought.

Peter

Two coins more to this discussion. There is one more way to control feed flow from the pump. It is counter piston which has movable stop above it letting this counterpiston move trom zero movement to full stroke of working piston. This counterpiston may be loaded by light spring to not cause any vacuum on the suction stroke of the pump. When You need full delivery of water You rest this stop so that counterpiston would not move. And when need zero output You let it follow working piston so that they shift this portion ot water between them withing the cylinder. This stop can be easily rest in any intermediate point right on the "fly". Such idea, not mine.

Serge

And a spring on the counter piston can be used to adjust the maximum pump pressure, regulating the system pressure.

Lohring Miller

After working with feedpumps for a good while, I don't see how a "spring loaded counter piston" would work on a steam generator. I'm assuming this would be like a spring loaded head on the pump cylinder.

Feedpump pressure varies wildly, especially when an economizer is used. Pressure in the economizer can be greater than the boiler pressure.

I have gauges on the feedpumps on the tug and they easily bounce to 500 psi feeding the boiler that is at ~70 psi. A simple elbow fitting in the line can increase the pump head pressure significantly and most feedpump systems require a some sort of shock arrestor to prevent water hammer. On my Locomobile with engine driven 1:1 pump, I have a feedwater heater in the muffler which is a coil of tubing up and around and back down to the bottom, it does two things, is uses the exhaust steam as a first stage feedwater heater and it also traps air at it's highest point providing an air cushion for the piping and doing away with hydraulic shock/water hammer. I ran it without and when taking feedwater, it sounded like someone was tapping the economizer with a hammer. Another remedy is to slow the pump speed down and undoubtedly why Stanley and other drove the large ram pumps from the rear axle which was much slower RPM than the engine, but problematic at low speed.

The point I'm making is the pump pressure varies wildly with no correlation to the water level in the boiler. As boiler pressure has no correlation to the water level other than completely full or completely empty.

A note on the gauges, a 1000 psi gauge perhaps, will get destroyed in a short order measuring between 0 psi and 500 psi 400 times per minute. The remedy was a 10,000 psi gauge with the needle barely fluctuating, it's only real purpose is to show that the feedwater is flowing. I also have 500 psi relief valves on the two engine driven pumps in case the valve is accidentally left closed at the boiler to prevent destroying the pumps.

-Ron

Hi Ron

In my proposal spring does not regulate anything. It only serves to help to return counter piston to it's bottom point on the suction movement of the working plunger. The movable stop regulates volume of the delivering portion of water. When working plunger begins to press water in the chamber it can not create pressure because counterpiston moves back until rests onto it's (adjustible) rest and only then begins create pressure and delivery water into the boiler regardless whatewer pressure in it. Adjusting the length of this retreat of counterpiston You could change flow rate of the pump at constant RPM accordingly level or temperature sensor.
Serge

Russ,

Keep in mind that the ram pump operates between some vacuum less than 0 psi and whatever the maximum head pressure is, be it 5,6,700 etc. psi and it does that many times per second. The plunger design and associated spring would have to survive that sort of abuse.

-Ron

This is an old toolmakers rule or commandment or adage: Avoid Springs, all types.

The reasoning for that is, most of the time they are applied in to a design, the engineering specs for the particular spring is unknown or ignored. They have a spec for loading, operating temperature, range of compression or extension, that must be observed for the proper service and service life.

Hi Ron,

Did you see any reports of problems with the "throttled-intake feed-water pumps"? No problem with the intake stroke, as long as the piston or plunger seals are good, but I wonder about the delivery stroke. Let's say it is throttled to where at the end of intake stroke, the pump cylinder is half solid water and half sub-zero-psig water vapor. On the return/delivery stroke, half of the stroke is nearly no resistance, but then the plunger or piston hits solid water at a fair rate of speed. I wonder if the pressure in the pump would spike momentarily, with big impact loads on the mechanism and pump pressure vessel, while the inertia of the liquid and delivery check valve is overcome. Sounds like it might run rough and/or noisy. Maybe not an issue with "live steamers", but for automobile-sized pumps & powerplants?

Didn't the early Stanley cylinder lubricator pumps use a spring-loaded "counter-piston" like Serge describes? I seem to recall with adjustable spring tension to adjust the plunger stroke and thereby the cylinder oil delivery rate. Later, in the condensing era, they used an impact element whose position was adjustable, to vary the cylinder oil pump stroke, then an off-the-shelf "box lubricator" where a ratchet pulled the cylinder oil plunger back and released it, to be driven on delivery stroke by a spring. I am currently considering the early-condensing-era-Stanley style lubricator pump, which is fully adjustable and springless.

Peter

Quote: "the pump cylinder is half solid water and half sub-zero-psig water vapor."

The only way vapor could happen is if the water is at high temperature and the vacuum caused it to boil. Feedwater in a non condensing system is usually somewhere near ambient or less - a long way from boiling. It would simply be water at vacuum which occurs in every feedpump supply side, they cannot work without it. Or if it was pressure fed, something entirely different.

Think about it like this: If the valve to the pump were completely closed, the piston would pull maximum vacuum at the bottom, on the upstroke, much of he travel would be cancelling the vacuum, the last bit of stroke at the top would equalize with exhaust head pressure through the check valve. Nothing in - nothing out.

Quote: " I wonder if the pressure in the pump would spike momentarily, with big impact loads on the mechanism and pump pressure vessel, while the inertia of the liquid and delivery check valve is overcome. Sounds like it might run rough and/or noisy. Maybe not an issue with "live steamers", but for automobile-sized pumps & powerplants? "

That is a good description of the problems with all ram feedpumps at high speed and pressure. No way to fix it other than to install some sort of shock arrestor. Water will not compress, forcing it or displacing it at high speed through narrow passages and elbows - restrictions against head pressure is going to cause a spike in pressure on every stroke. High RPM? Many times per second.

Lubrication and feedwater are two totally different services, where feedpumps operate at much higher volumes and pressures. If the Stanley's used this spring loaded counter piston, they answered this question by using it for the low pressure low volume lubrication pump and not using it for the feedwater pump. It would be short lived and noisy.

Here is that video, he talks about it at the 37:00 minute mark. And again at 46:50

[www.youtube.com]

-Ron



Edited 2 time(s). Last edit at 07/02/2018 06:34AM by IronChief.

Hi Ron,

Even if the unfilled volume in the pump cylinder at end of intake stroke were a vacuum with zero vapor -- no time for a vapor physics discussion -- throttling the intake to a pump, to control flow at a given speed, would give part of the cylinder containing water and part of it not containing water, at the end of the intake stroke. So on the delivery stroke, there would be very low resistance for part of the stroke and then bam, the plunger or piston hits incompressible water. Especially if that happened around the middle of the stroke, where plunger speed is highest, I wonder about the results. For a car-sized pump, it might get noisy or vibratey -- is that a word? -- and might require seriously beefing up all sorts of things, relative to a full-stroke-filling pump.

But, yes, the pressure would equalize and it would flow water, if built strong enough for the loads. Not built strong enough, of course, it would still flow water, but only for one stroke and all over the pump enclosure and surrounding areas, along with broken pump parts.

This would not be an issue if the inlet control were simply on/off. Off, there would only be vacuum inside the pump; on, it would run with cylinder completely filling. Even then, though, yes, the pressure is inherently "spikey", especially in a double-headed plunger pump with lots of inherent flow restriction. Some triplex pumps are said to deliver a much smoother flow. One Stanley guy reported smoothing out the spikes by simply using a longish flexible delivery line, but then somebody else said that was a Code-flunker. Years ago somebody mentioned an off-the-shelf accumulator [Code-OK] to smooth out the feedwater pressure and eliminate the "water hammer" in Stanleys, but now I don't remember who, or when, or any details.

I agree about Stanley using the counter-piston only for the cylinder oil pump, and not for feed pumps, providing the answer. On top of that, they only used it for oil pumps for a few years, then changed to other types of oil pumps. I don't know which years they used it, but I have seen a number of illustrations and explanations in the old literature. Apparently the principle worked, at least for small low-pressure pumps, but other oil pumps must have worked better, at least in the Stanley Brothers' experience. I should double-check that though. I _think_ it was a counter-piston design, with the counter-piston providing oil feed rate adjustment. But maybe I am mis-remembering it, and the spring-loaded piston only acted as a delivery valve? That tech section in "Floyd Clymer's Steam Car Scrapbook" has an illustration and explanation.

Thanks for the link to the video; I will check it out later. I am supposed to be blueprinting that cylinder oil pump. So far I'm not having much success getting started on it.

Peter

I'm not advocating that this would work and a bypass is no longer necessary. The optimum method of adding feedwater is "little and often", the old railroad firemen's motto. Ideally feedwater should be administered at the same rate the boiler is using it, that is the most efficient (Serpolet, White etc). So this valve to the pump could be throttled for continual adjusted flow, I think it would be easier to control than trying to meter with a bypass.

Water will boil at lower temperatures, but it doesn't happen instantly, and keep in mind, too much feedwater is rarely an issue and long spans of not adding feedwater do not happen while running, on purpose anyway, water consumption is high on any people carrying vehicle. At Greenfield Village they have a small rail loop that they run two or three small Locomotives on, the loop is maybe a bit over a mile long, they use 50 gallons of feedwater on every lap. Stopping, blowing whistles, running steam pumps. Most road vehicles will use about 1 gallon per mile.

Of course we can imagine all sorts of things, but until a throttle to the pump were actually tried, it's just speculation.

-Ron

Hi Ron,

It would be great if it worked, and if the "water throttle" could be controlled to feed just the right amount of feedwater, without excess water jetting out a bypass valve at full feed pressure. The throttle could save a decent amount of horsepower relative to the old-style bypass control, especially with pumps with lots of extra capacity. Meaning lower fuel and water consumption, longer vehicle range, and faster vehicle acceleration & speed, with the same engine. Right, we won't know till it's tried in a car. Looking at the feedwater pump I just blueprinted, it's pretty beefy and just might handle any extra loads from the cavitation. I plan to try standard bypass control first, with an "E-Z-Bild" feedwater automatic I blueprinted several years ago, but I will keep the throttled-pump idea in mind for future experiments. Thanks for the information on this.

Peter

All files from this thread

File Name	File Size		Posted by	Date
test pump.jpg	129.4 KB	open | download	IronChief	06/18/2018	Read message