Valve Testing

24 views
Skip to first unread message

Russell Philips

unread,
Dec 6, 2011, 10:47:35 AM12/6/11
to Open Source Steam

Here is a drawing of a prototype drum valve assembly.
This should be a reasonable build (with access to the right tools).

http://www.SunriseEnergy.org/images/drum_valve_17.pdf

Driven by a 3/8" I.D. steam line.
Spin the drum shaft with a cordless drill for an initial test.

Mark Norton

unread,
Dec 6, 2011, 11:02:57 AM12/6/11
to open-sou...@googlegroups.com
While the fabrication steps are clear, I'm having a bit of trouble seeing how it works.  Can you do an assembled section view?  Label steam inputs.  I assume the "port" in your current diagrams are the exit port into the cylinder.

- Mark

Ken Helmick

unread,
Dec 6, 2011, 11:59:38 AM12/6/11
to open-sou...@googlegroups.com

How do you control cutoff?

Russell Philips

unread,
Dec 6, 2011, 3:46:30 PM12/6/11
to Open Source Steam
Mark, Does this make the proto workings clearer?

http://www.SunriseEnergy.org/images/drum_valve_18.pdf


Ken, How bout a quick description of throttle and cut-off and
importance.

Ken Helmick

unread,
Dec 6, 2011, 4:59:33 PM12/6/11
to open-sou...@googlegroups.com
There are a few different ways to regulate steam engine power output.
 
1.  Install a throttle valve
2.  Actively control boiler pressure
3.  Adjust cutoff
 
Throttles, as the name implies, throttle the steam from boiler pressure down to a lower pressure which produces the amount of engine work you desire.  This is all well and good for accurate power control, the problem is that it takes significant heat input to generate the steam in the first place and when you throttle both the temperature and pressure decline without producing any useful work.  Actively controlling boiler pressure is a theoretical ideal but difficult to implement as hysteresis causes results to lag input significantly, you would need a boiler with almost no thermal mass to give immediate response.  The nicest way to control power output is to regulate the engine cutoff, the point in the stroke at which the admission valve closes.  
 
Cutoff is usually given in percent of stroke, so if the admission valve closes 1/4 of the way down the cylinder bore we refer to 25% cutoff, another way to think of it is as the reciprocal of the expansion ratio (assuming no clearance volume).  By closing the admission valve earlier, less steam is admitted to the cylinder, and less power is produced.  Since the steam hasn't been throttled (discounting the steam that undergoes wire drawing just as the valve is coming shut) we haven't suffered a loss there as well.  Ideally, you would like to size your engine such that your typical loads are done with the most economical cutoff and then be able to employ longer cutoff in those instances when greater power is needed.
 
Variable cutoff is produced by a variety of valve apparatus referred to as valve gears.  The most common valve gear is the Stephenson's link but there are many others including the Marshall, Hackworth, Fink, Gootch, Baker, Joy, Waelscherts and so on and so on.   Some valve gears use variably eccentric eccentrics, these are typically employed with speed regulating governors.
 
Cutoff control is not applicable under all conditions, so throttle valves are still included, but to whatever degree possible it is desirable to "work the steam expansively". 
 
Steam temperature and pressure both affect expansion ratios.  Obviously, we can only expand to some minimum feasible pressure below which friction, exhaust back pressure, pump loads and so on exceed the power we can extract from the steam  Since there is a minimum feasible pressure, it is apparent on inspection that increasing operating pressure improves overall efficiency because we can expand the steam further and extract a greater portion of the energy contained therein.  Temperature is also important, however, and its importance can not be overstated.  Assume two identical engines both running at 150 psi, one using saturated steam with no entrained moisture and the other running with 200 degrees of superheat.  As the piston moves down the cylinder, the pressure and temperature drops in both cylinders. The steam in the saturated cylinder contains no sensible heat, it is all latent heat, so before very long the steam begins to condense while the superheated cylinder gives up sensible heat and the superheat begins to drop.  Since condensate occupies much less volume than steam, we see an accelerated drop in pressure in the saturated steam cylinder, and this cylinder reaches the minimum practical operating pressure well before the steam in the superheated cylinder.  From this we can demonstrate that higher temperature steam can operate with shorter cutoff (higher expansion ratios) than cooler steam.  Unless the steam system is tightly controlled to maintain stable temperatures at all times, the ability to control cutoff as temperature varies can significantly improve efficiency.


-----Original Message-----
From: Russell Philips <russell...@hotmail.com>
To: Open Source Steam <open-sou...@googlegroups.com>

Mark Norton

unread,
Dec 6, 2011, 5:09:51 PM12/6/11
to open-sou...@googlegroups.com
While I'm aware that control over cut-off is a desired approach to controlling an engine, I must admit that I'm a bit vague on how the valve gear actually works.  Do you have any links or pointers to explanations of the valve gear types you mention?  Stephenson, Marshall, etc.  I'd like to understand this better.

- Mark

Mark Norton

unread,
Dec 6, 2011, 5:10:57 PM12/6/11
to open-sou...@googlegroups.com
Much clearer.  Thank you.

Ken Helmick

unread,
Dec 6, 2011, 6:37:43 PM12/6/11
to open-sou...@googlegroups.com
Here's a quick sketch I just threw togeher of a valve set suitable for mounting on air compressors or other SA engines
 
 
valves 2.png



 
-----Original Message-----
From: Mark Norton <markj...@gmail.com>
To: open-source-steam <open-sou...@googlegroups.com>
Sent: Tue, Dec 6, 2011 5:10 pm
Subject: Re: Valve Testing

valves 2.png

Ken Helmick

unread,
Dec 6, 2011, 8:28:24 PM12/6/11
to open-sou...@googlegroups.com
There are a ton of old books on the internet that cover valve gearing in any kind of detail you want from basic concepts to Zuener diagrams. 
 
 
Charlie Dockstader wrote a series of programs to allow you to simulate various valve gear on the computer:
 
 
and others...
 
 
 
 
For some good papers on superheat, condensation, valves: http://5at.co.uk/index.php/references/software.html
 
Regards,
 
Ken
-----Original Message-----
From: Mark Norton <markj...@gmail.com>

Mark Norton

unread,
Dec 7, 2011, 8:38:37 AM12/7/11
to open-sou...@googlegroups.com
Thanks for the links and the diagram, Ken.  Can I add your diagram to my web site?

- Mark

Mark Norton

unread,
Dec 7, 2011, 8:48:55 AM12/7/11
to open-sou...@googlegroups.com
Ah, never mind.  I see it on the site already.

- Mark

Russell Philips

unread,
Dec 7, 2011, 9:26:57 AM12/7/11
to Open Source Steam

Ken, The poppet valve is/was my number two go to. This design is well
done. It is the design, already formed that i was working on in my
head. I recognized the key aspects as the link painted: transverse
oriented poppet valves, one leading/ one lagging the steam flow
(pressure must seal both intake / exhaust), wobble linkage
activation. My missing form was the 90* conversion, the three point
rocker plate- well done! Simple and robust!

Does the valve guide seal steam pressure?
Does the valve guide require lubrication?
How do standard components hold up with steam/corrosion?
(i've heard they last 'long enough')
(plus, replacements are simple enough)
(stainless is always a go to)

Mark, I found this link helpful:
http://en.wikipedia.org/wiki/Stephenson_valve_gear

Ken Helmick

unread,
Dec 7, 2011, 9:38:33 AM12/7/11
to open-sou...@googlegroups.com
One valve gear you might want to consider is the swinging eccentric.  It's dead simple, especially if you don't worry about the operator directly controlling the valve gear and instead use some variant of Rite's Governor.
 

 
RITE'S INERTIA GOVERNOR1.png

Francis Rite patented and marketed his governor near 1900 when steam was on the way out, so it is less well known than the fly ball governor.  While not theoretically quite as accurate as the flyball, in practice is holds a much more steady speed.  It's also far more simple to build, the flyball itself is more complex and then must be connected to the engines valve gear while the simple Rite's IS the valve gear.
 
A weighted arm is mounted on the flywheel by a pivot and forced to swing on the pivot due to the action of the tension spring so that the line between the pivot and the swinging arm center of gravity is line AC.  The swinging eccentric operates the steam admission valve through the reach rod, since it is resting at the intersection of line AB and arc EE it is off center to the flywheel which revolves around the intersection of line AC and arc EE and being off center acts as a crank to push the reach rod.  As the engine speeds up, centrifugal force pulls the weighted arm radially outward from the pivot causing it to rotate and precess clockwise in an effor to place the center of gravity on the center of the flywheel rotation.  This precession is opposed by the tension spring causing centrifugal force and spring tension to come into equilibrium at some speed.  This precession also causes the center of the eccentric to move closer to the center of the flywheel, reducing the eccentricity and decreasing the distance the reach rod is moved, thereby shortening the period of time the admission valve is open.
 
One advantage of the inertia governor is that the weighted bar has a LOT of inertia and any small change in engine speed causes the bar to surge ahead or back, instantly changing cutoff to match demand.  This made inertia governors popular with people operating small electric generators, cream separators and other machinery needing close speed regulation.
 
Next up, I'll sketch up Marshall's valve gear, possibly the simplest radial valve gear available.

Ken Helmick

unread,
Dec 7, 2011, 1:26:45 PM12/7/11
to open-sou...@googlegroups.com
Russell:
 
Thanks!
 
The one really bad design flaw that I made was showing the two valves in one housing.  Should NEVER do that if avoidable, you want to prevent heat from from the admission to the exhaust through the valve bodies so as to preserve the enthalpy for use in the cylinder.
 
The valve stem doesn't really seal pressure, but it doesn't leak that much due to the tight clearance and long passage.  This might work out to simply be a case of benign neglect because the valve stem is only intermittently under pressure in the case of the admission valve and even then the pressure is not constant but declining immediately after cutoff.  Turns out blowby is a function of MEP and residence time, and with a shorter cutoff and decent rpm these both diminish.  Automotive IC engines use polymer seals to keep oil from running down into the head through the stems, these might be adequate to address blowby if needed.  If tighter sealing is warranted, I suspect a teflon O ring on the end of the guide held in place by a packing nut should do the trick.
 
Generally speaking, poppet valves don't need much in the way of lube, they have almost no side loading and can hardly twist in the guide bushing, so not a lot of chance to wear.  Of course, if the guide bushing is not aligned to the valve seat, all bets are off.  If a bit of cylinder lube is injected into the valve to lube the cylinder, the poppet should be more than adequately oiled.   Other engines in the past had a small hole drilled into the guide bushing and a small tube was connected to the oiler to supply a tiny flow directly to the stem.  I suspect the oil was also viscous enough to retard any blowby.
 
I don't expect significant problems with IC engine grade parts in a steam application, they are certainly designed for high impact loads at peak rpm and the IC engine runs extremely hot.  Poppets were considered the ideal steam engine valve by many industrial builders back in the day, but the metallurgy wasn't good enough to make for valves that didn't require frequently regrinding.  Today our metals are much, much better.  To be far, the valves they did use often required resurfacing as well.
 
Rust is one of those things that can largely be addressed by proper operation and maintenance.  The engine should have some vent valves that can be opened immediately after securing, if the engine is properly hot any initial condensate will flash to steam and exhaust out. Allow to dry for a few minutes and then close the valves to prevent any humidity from creeping back in.  You won't get rusting while in operation because there is no significant oxygen in the engine, it has all been displaced by the steam.  Note the last statement may not apply to boilers, when I was in the navy we used a deaerating feed tank to get air out of the steam system and others use hydrazine or other chemicals to do the same.  Water chemistry is what usually separates professionals from guys that own steam engiens, commerical powerplants and naval propulsion systems) run tens of thousands of hours without significant corrosion due to proper feed water chemistry.
 
Regards,
 
Ken


-----Original Message-----
From: Russell Philips <russell...@hotmail.com>
To: Open Source Steam <open-sou...@googlegroups.com>
Sent: Wed, Dec 7, 2011 9:27 am
Subject: Re: Valve Testing

Ken Helmick

unread,
Dec 7, 2011, 2:17:47 PM12/7/11
to open-sou...@googlegroups.com
Here is the Marshall valve gear, which is a variant of radial valve gear for those whom have an interest in that sort of thing.
 
The eccentric is mounted on the crankshaft or a parallel layshaft and turns with the engine.  The eccentric strap oscillates with the eccentric.  The eccentric strap follows a curved path imposed on it by the swing of link "B".  Link B is connected to the top of link A and free to swing from the top of A.  The control arm is controlled by the human operator as he desires, and it serves to swing link A around the stanchion to which it is connected, the stanchion being rigidly affixed to the engine.
 
When link A is in the vertical position, link B swings in a primarily horizontal plane due to the oscillations created by the eccentric strap and eccentric and the reach rod (which operates the steam admission valve) experiences only slight vertical movement.  Moving the control arm so as to position link A at an angle away from vertical causes link B to swing with less of a horizontal componentand more of a vertical, causing the reach rod (which is connected to the eccentric strap by the clevis) to travel further and open the steam admission valve more fully and for a greater length of time.
 
Unlike some valve gears, there are no particular proportions to Marshall valve gears and I have seen the positions of the reach rod clevis and link B reversed...and also seen link B and the clevis share the same pin.  Compared to a Stephenson's link, it is easy to build and a working version can be constructed with a hacksaw, drill press, drill set, a couple of reamers and some files.
Marshall valve gear1.png

Russell Philips

unread,
Dec 7, 2011, 3:09:42 PM12/7/11
to Open Source Steam
Ok, im getting it, (however a link to a picture would be great). The
Marshall gear (manual lever) sets the distance of the rod travel -
which affects the open and close time of the poppet valve. Great
stuff, and easy to implement. Testing of valve timing, proper cut-off,
and efficiency will be much easier with variable control! This
control should raise steam engine efficiencies above 10%, towards 20%.
I like it!

Ken Helmick

unread,
Dec 7, 2011, 4:44:17 PM12/7/11
to open-sou...@googlegroups.com
Hi Russell:
 
The link to my sketches is:
 
 

For more information on the Marshall gear:
 
 
 
 
 
 
 
 
 
 
 
OK, back when I first got really interested back in 96 I had to go to used book stores and part with hard earned cash, search old book in university and big city libraries and delve through tons and tons of microfiche at the Detroit patent office archives.  Now you just type a few words into Google and get more results than you can read....for free.  Just seems a bit unfair, somehow.
 
Ken
 
 
-----Original Message-----
From: Russell Philips <russell...@hotmail.com>
To: Open Source Steam <open-sou...@googlegroups.com>
Sent: Wed, Dec 7, 2011 3:09 pm
Subject: Re: Valve Testing

Mark Norton

unread,
Dec 7, 2011, 5:32:52 PM12/7/11
to open-sou...@googlegroups.com
It's a nice service that Google offers, though they do make money from it, too.

- Mark

Russell Philips

unread,
Dec 7, 2011, 8:44:13 PM12/7/11
to Open Source Steam
The complexity of the Marshal gearing can be simplified for our
purpose (non-reversing). As i considered upon the movements, the goal,
and the minimum movement - is simply to change the ratio of eccentric
distance -vs- valve distance.

We already have a complex 90* conversion device.
My focus went to this area.
We can simplify the eccentric strap to a simple eccentric rod (wobble
shaft).
Shifting complexity towards the 90* area.

I know this is ratio is adjustable. I have already thought of several
ways to make adjustments while it is still (which is good). However,
There are moving parts in constant motion. The trick is to make it
adjustable on the fly.

By changing the simple three point rocker plate to a more complex set
of arms, we can achieve adjustments over this ratio.

I believe one axis on an a threaded rod gives us the needed
articulation for dynamic ratio adjustment.
I'm still mulling this over- and need to draw it....

Russell Philips

unread,
Dec 8, 2011, 8:26:46 AM12/8/11
to Open Source Steam
The control thread changes the vertical/horizontal movement aspect
ratio.

http://www.SunriseEnergy.org/images/Rocker_Gear_20.pdf

Russell Philips

unread,
Dec 9, 2011, 6:34:29 PM12/9/11
to Open Source Steam
Ken is proposing a uniflow engine design. We have previously decided
that the re-compression issues puts bottom of cylinder ports and
uniflow out of scope. Ken mentioned somewhere that he had a 'poppet in
piston design'. I would assume this is a uniflow design. Perhaps this
would eliminate recompression and allow us to harness the advantages
of uniflow. As i considered upon the poppet in piston design....

I thought of a 'latching' poppet valve in piston concept.
This gave way to an 'inertia' poppet valve in piston concept.
Which actually requires an 'inverted inertia' activated poppet
(cantilevered mass).

LATCHING VALVE
When the poppet is extended (open/exhaust position) a spring tensioned
latch engages and locks the poppet open. This occurs at/near BDC. The
piston travels upward venting most/all expanded vapor. At TDC a
'piston button' touches top of cylinder and the trips a manual
release, allowing the poppet to spring close. Steam poppet opens re-
pressurizing the cylinder for expansion stroke. The steam expands
against the now closed poppet in piston, until the piston reaches BDC,
where the poppet is manually opened - engaging the spring tensioned
latch.

The thought occurred that the spring may not offset the poppet
valve's inertia forces at higher/extreme rpm.

I thought of the mass-vector timings...
Could the poppet naturally (inertia) open and close at the right time?
Unfortunately it is opposite of what we need for sympathetic inertia
help.
This is how the inertia concept originated.

The piston can mount a mass greater than the poppet on a cantilever to
induce sympathetic inertia help.

Actually, the mass need only be strong enough to allow the valve
spring to close at our target maximum rpm.
(This assumes that the spring is not strong enough to close it in the
first place!)

Perhaps the cantilevered piston poppet valve would open and close
naturally without need of manual (activation, latching, release)?

Not sure? My head hurts...

A uniflow without re-compression issues (by venting expanded vapors
through a poppet that stays open throughout the entire return stroke)
is EPIC!

Immediate upcoming issue: steam vents to gear box.... i'm working on
it :)

What do you all think?
Has this been done before?


Ken Helmick

unread,
Dec 9, 2011, 7:28:32 PM12/9/11
to open-sou...@googlegroups.com
Hi Russell,
 
Generally speaking, the thing to understand regarding compression and uniflow engines is how to calculate the necessary clearance volume.  The ideal solution is to recompress back to admission pressure, so you have to know about what your steam pressure will be, as well as the exhaust pressure, and then understand the basic formulas.
 
The Williams Brothers used a relief valve in the cylinder head that vented excess recompression back to the steam chest, so they never overcompressed, but my math does show a penalty....but less than if they had over compressed.  Others use an auxiliary clearance volume that is adjustable depending on exhaust pressure...this typically being found in condensing engines that routinely exhaust at subatmospheric pressures.
 
If most of these engines can make use of good condensing, then the demand for clearance volume to handle recompression drops, and efficiency goes up as clearance volume is a bad thing.
 
All that being said, there is a certain allure to having the ability to vent pressure off through the piston down to the uniflow port and delay compression onset...you do NOT want to give up compression altogether as that reduces efficiency (because you will not get a perfectly zero clearance volume and recompression reduces losses due to clearance).  But delaying it is certainly nice.  I know my own engine has that capability, and with projected operation at 1000 psi and 1000 degrees, it should pay off nicely.
 
The idea is not exactly original with me, see the links below for a sampling.  See what happens when you do your research?  :-)
 
I tend to think uniflow is favorable, but I know the engineer Tom refers to who favors counterflow, and I respect him highly.  I don't think our agreement is huge, I am sure that the thermodynamic advantages of uniflow drop as you apply more superheat although they don't disappear.  I can't argue that the counterflow engine has some mechanical points in its favor.
 
The engine I sketched up has a few features that I am not sure the group appreciates.  The piston rings are displaced quite a ways down the cylinder wall to keep them away from the hottest portion of the cylinder wall.  Likewise, it is hard to see but the upper part of the piston is slightly undersized and has small labyrinth grooves cut in.  This serves to retard steam flow but allows the upper end of the piston to remain out of contact with the wall, the piston being long enough that the lower end is more than adequate.
 
The piston wrist pin is displaced sideways, which has two related benefits.  The offset creates a small toggle action, making the residence time at TDC nonsymmetrical, in the first place this means the side wall pressure will be more advantageously distributed when the piston is near the top and under highest pressure.  The same nonsymmetrical events cause the dwell time at TDC to be exxagerated, which allows for shorter cutoff for the same valve operating speeds.  Lastly, the piston is built up of fairly simple to make parts and pulled together with a single Allen bolt, obviating the need for complex machining or casting of elaborate cores.
 
The poppet valve also acts as a relief valve, venting pressure back to the steam chest.  This can be useful on cold starting if someone forgot to drain condensate from the head.
 
The valve gear uses a single lay shaft chain driven at 1:1 speed to the crank.  Displacing the shaft axially causes the eccentric to shift sidewas and thereby increase or decrease cutoff...or facilitate reverse operation.  Worked just find for Muller.
 
The main crank bearings are just spherical housing ball bearing pillow blocks.
 
Anyhow, this was less of a solid proposal than an attempt to throw a lot of things out at once and give people something to work from.  Ever read Jefferson's draft of the Declaration?  Wasn't that good, the Continental Congress improved it immensely.  On the other hand, without his draft to spring from they would still be haggling over the Preamble and we'd all be adding the letter u to a lot of words unnecessarily, like colour.
 
Regards,
 
Ken

 
 
 
 
 
 
 

 
 
 
 
 
 


 
-----Original Message-----
From: Russell Philips <russell...@hotmail.com>
To: Open Source Steam <open-sou...@googlegroups.com>
Sent: Fri, Dec 9, 2011 6:34 pm
Subject: Re: Valve Testing

Reply all
Reply to author
Forward
0 new messages