Self - Regulating Feed Pump

Some years back, I had an idea for a feed pump that could control its own operation and, hopefully, be responsive enough to work with a monotube boiler.

I took yesterday evening to sketch up a slightly more complete drawing of the idea, plus a variant. The hydrolock version is on the left. The engine turns a drive eccentric which in turn, causes a lever to rock on a spring-loaded pivot. This spring is under a heavy pre-compression and therefore is effectively rigid at moderate pressures. The upper cylinder is termed the 'control pump/. It takes suction from the boiler at whatever point you wish to maintain the water level and it returns this flow back to the boiler by way of a needle valve. The theory is that, if the water level is low, steam will be pumped through the cylinder and will easily pass through the needle valve due to its lower density and viscosity. On the other hand, if the level is high, you will be pumping water, which will not readily pass through the needle. This will effectively cause the pump to hydro lock and will, in turn, cause the lever to compress the spring rather than move the piston. A second piston, on the bottom of the crosshead, is your feed water pump and is therefore regulated by the state of the upper control pump.


The version on the right is designed if you wanted to use something like a CAT pump to feed the boiler. The control pump has a relief valve which prevents the hydrolock from becoming so serious that it breaks the pump. A spring-loaded piston, or Bourdon tube is positioned between the control pump and the relief valve; this piston opens a recirculation valve that diverts flow from the feed pump discharge back to the feed water tank.

It just occurs to me that a third variant is possible. It would resemble the second variant but have its own feed pump cylinder, or cylinders, driven by the same shaft that powers the drive eccentric. These cylinders would be driven via a curved pivoting link and the spring-loaded piston would pivot the link so as to increase or decrease pump stroke(s).

I always wondered if this concept would be responsive enough to regulate a monotube.

Regards,

Ken

Hi Ken,
I hope to beet Rolly with this comment. I'm not following your logic to use a feed pump for a boiler on a mono-tube that doesn't necessarily feed water based on it's water level.

I think I'm missing something. Can you fill in the details?

As far as using it in a boiler with water level control, your concept is sound. However, my guess is that it will be a nightmare to get it working with all the check valves. One of the most troubling issues with steam engines is to get your check valves to work properly and consistently.

Kind regards,
Rick

Hi Rick

The logic is simple. Monotube boilers are unstable because rising water levels increase the amount of boiler tube surface area dedicated to generating steam while, at the same time, reducing the area dedicated to superheating that steam. This means we now have lots and lots of wet steam. Low level gives us a smaller area capable of producing steam but a lot of area providing superheat -- now we have less steam but it is extremely hot. This is in contrast to a boiler with a drum, in which the generator and superheater surface areas are fixed and, therefore, always operate in a regular and predictable relationship to one another. In these devices you can change the amount of generator surface to get the correct evaporation rate and the amount of superheater surface to control the output temperature. In a monotube, you are trying to control a boiler that is redesigning itself on a second by second basis.

Therefore, the more closely we can regulate monotube water level, the more predictable the boiler becomes. Just remember, after fooling around with monotubes for decades, Abner Doble stated in print that he was never really satisfied with any of his control systems. His controls got quite complicated, but they never directly controlled the water level; instead, they measured its effects indirectly via temperature in different segments of the monotube and this was subject to ongoing error due to thermal time lags and hysteresis.

Regards,

Ken

First lets get something straight. I do not consider a Monotube a boiler it has no circulation no drum or water level, it’s just a piece of pipe.
One way to control the temperature is to control constant input and output of the feed water as well as the BTU input.
Doesn’t work for a car stopping and changing speeds.

Doble did use a bank of pumps, if my memory serves me I think he used fore. I can’t remember if they were turned on as needed. I’m not going to reread Walton’s book to find out. All the good technical stuff was left out of Jims book that he had in his manuscript.

Today a variable speed pump might do the job with a controller looking at temperature along a given section of pipe. This would only work with a design limit to the maximum and minimum output. Temperature output limited to engine oil rating.

I think Manual mechanical controls would be a waste of time with today’s technology.
Solid state controls are so much faster.

At my age I am not interested in any of it.

Rolly

Hi Rolly,

I guess I have to classify a monotube as a boiler simply because I have seen engineering articles about 190 years old that refer to them by that name -- as Bulletin editor, I've given other people static about inventing their own terminology for existing hardware and, in fairness, I have to obey the same rules.

Computers aren't always an option. There's a group in Cuba looking at building simple little steam vehicles and they certainly don't have access to the kinds of sensors, microprocessors, or actuators that you would need. They have limited access to some machine tooling, but things really need to be kept simple since they aren't going to have much in the way of dies, jigs, and fixtures -- and it's hard to beat a monotube for simplicity. There might be something to be said for a standpipe to reduce boiler surging and you might justify a recirculation pump if you can make things simple. I think they are looking at something about like a Loco-surrey in capability, but they don't have the same manufacturing resources.

The same might apply to little hobby vehicles, I certainly wouldn't want a minibike that needed a battery, alternator and other such hardware -- I used to beat mine up so badly on back trails that just keeping the mechanicals in good tune was effort enough. Admittedly, things got better when dad and I junked the 5 hp Briggs & Stratton in favor of a 4 hp Kohler -- that Kohler even made the bike faster, making me question how people were rating horsepower in the mid to late 60s. Looking at one of the RJ Smith minibikes, I was struck by how unlikely it was for that control system to operate long without going out of adjustment, so something more brute force would be nice. I think this system is about as in-your-face brute force as you can get.

Regards,

Ken

So does your pump's boiler feed and return tap into the monotube at a point where you want to maintain the water level?

Lohring Miller

Ken
Theirs a lot written by so called qualified people (engineers). One needs to discern what is correct and what is not. I have worked reading specifications and bidding and building all kinds of construction projects. Bridges, power plants, buildings of all kinds. I made more money with change orders resulting from design errors, by the so called qualified people. Do your own work.

I do not consider I’m inventing new nomenclature. A boiler is a device that heats water above the boiling point designed for circulation. Look up the word circulate, it follows a coerce returning to the starting point. The boiler stores the energy of heated water and only release the energy as steam when the pressure is lowered.

A monotube has no circulation nor is it designed to store heated water. Mono (one) tube.
It’s not a boiler in any sense. I would call it a steam generator.

Rolly

Hi Lohring,

There are two pumps in the left-hand drawing. The upper one is the control pump which takes a suction at the desired boiler level and discharges back into the boiler. The lower pump is the actual feed pump; it draws water from the day tank, or condenser, and pumps it straight into the condenser. If the upper pump hydrolocks due to a rise in boiler water level, the directly attached lower pump cannot feed water to the boiler.

In the right-hand version, there is only a control pump. As in the first example, it draws from the boiler right at the desired water level and pumps back to the boiler. The pumping pressure is dependant upon the presence of either steam or water, and this controls a bypass valve connected to the actual feed pump (not shown).

Regards,

Ken

I understand that. I was just wondering about the actual plumbing in a monotube. It seems to me that the "boiler" in this case is a long tube. Therefore, the connection needs to be along the tube where you want the water steam interface.

Lohring Miller

Rollin White
On his patent calls his a steam generator. It’s other people writing about it that can’t rear and call it a boiler.
Most mono tube steam generators are coiled configuration of pipe. The water goes in at one end and the steam comes out the other end. White, Doble, and a few others used different systems to balance the amount of water to the burner to maintain the proper output. (Temperature and pressure) my experience has only been with the White, in two cars. In both cases the temperature and pressure cycled badly, and both burned out the two coils nearest to the burner.

Attached is a scale model of the mid 1907 version of the White.

Rolly



Edited 1 time(s). Last edit at 02/13/2021 01:44PM by Rolly.

To All,
I've been reading this book I bought from the SACA storeroom, "Experimental Flash Steam" by J.H. Benson and A.A. Rayman. It is a good book and I highly recommend it to folks considering taking the next step to high performance steam engines. It is only $30 with an additional $5 for shipping. It does talk about the experiences with flash steam as applied to model engineering and the model speed boat. Reference is made to a mono-tube steam generator that produces 163 lb/ft^2-hr. Something to be said about this when Doble's best is 26 lb/ft^2-hr. Note that the 163 is using 3/16" diameter brake line...food for thought.

Ken, after studying this book and reading the responses to your post, respectfully, I would say your device will not work in a mono-tube steam generator. I agree with Lohring that you would apply your pump regulating device serially into the mono-tube say after intended primary BTU input and before one might have super heat tubing. The intent is to separate the steam and keep the generating tube full. Also I would speculate that this pump would be properly sized to meet the lbs/hr required to keep the engine happy.

Please continue to present ideas for it forces us to think about what really is going on within these systems. This is in no way presented to discourage you, mean't to encourage you. In the flash steam book, the model boat actually wants to keep the feed water as water, not steam, all the way to the engine manifold or valve(s). The flash is supposed to happen there. This was Serpollet's intention also. This is intended to maximize the most energy of vaporization via the pressure drop. Note that when this occurs, the steam is super heated.

Your orifice will present a pressure drop and the feed water will always be water, not steam. This is the main reasoning for your system not to work at least from a theoretical stand point. If you don't agree, this is good. However, you'll have to build it to prove me wrong.

A far better solution is in the words of Ofeldt's Steam Automobile Specialties that again I bought from the SACA storeroom. I quote, "Upon trying to adapt the flash boiler to automobile purposes we found it was not as satisfactory as the water level boiler, for the resistance was not the same for more than five minutes at a time and it was impossible to keep an even steam pressure or take the hills without hand pumping. A water level boiler with a strong water pump does away with these annoyances, and enables one to take the longest and steepest hills without hand pumping."

I'm designing and going to build a mono-tube to a separator drum steam generator prototype. Also included is to incorporate natural circulation Ofeldt coils into this system. The water level will be used to control the feed pump. After steam separation, the exit from the drum will pass over the flame to introduce super heat. I'm in the design phase now and trying to figure out how to design it to make considering tube bending, forming and braze joint techniques. To build it is to say that it passes my theoretical test in my opinion of course.

Sincerely hope this helps,
Rick

Hi Rick,

The pressure drop at the needle valve is irrelevant, we don't care about the drop. What we are trying to achieve is the back pressure. The back pressure will be much higher when pumping water than steam, and it is this backpressure that regulates the feed pump.

Regards,

Ken

Hi,

Do a search on this forum for "waterman control system". I think you'll find what your looking for.

Keith T.

Hi Keith,

The underlying principle that water and steam have different physical properties is similar to the Waterman pump -- the actual system design has almost nothing in common.

Regards,

Ken

Keith,
Nice to go back and re-read the post from Jim Crank. Jim is definitely a favorite of mine in the Steam Engine world. He explains things clearly and succinctly…I can really understand what is happening. I have read his Doble book twice over.

Again, he confirms the use of a separator drum from a mono-tube and its use in feed water control. However, he envisions this application for a steam boat and not a steam automobile. Wondering what else he has in mind for steam automobile?

He also states that one must have the right-sized feed pump.

I purchased the plans for the Richard Smith Bicycle mono-tube system from the SACA store. Within his plans he shows a separator drum to collect the crud coming from the 1/8" mono-tube, black iron pipe. My belief that the real reason is to provide some buffer to the mono-tube so that the flow control valve can work. The separator tank is a controlled point in the mono-tube where the water will transition to steam. Also note that steam super heats when allowed to do this. This will heat any carry over and turn it to steam.

Last is Jay Carter (another hero of mine in Steam) stated to use an expansion volume just before the valve chest to allow for the volume of steam defined by the piston/cylinder. These words are taken from Chuk Williams during one of his talks at a SACA meet (hope I got it right). I believe that this volume (separator) will perform exactly like the Richard Smith mono-tube system description, the separator drum. The size of the separator volume is relatively applied the same in both cases.

Kind regards,
Rick

All files from this thread

File Name	File Size		Posted by	Date
SELF REGULATING FEED PUMPS.png	301.8 KB	open | download	frustrated	02/10/2021	Read message
Finished Model-3.jpg	527.9 KB	open | download	Rolly	02/13/2021	Read message
RHWhiteNomination-2_5.jpg	467.9 KB	open | download	Rolly	02/13/2021	Read message