Cylinder and valve starting point

When we took our first trip to Georgia to look over 110, we opened the left side front cylinder head and found remaining counterbore of about .250 at the top and almost nothing left on the bottom. Later we got the drawings, and discovered that the original counterbore was only .250 to start with.......so, i think it's safe to presume we will want to bore out the cylinders and line them........but, how thick should we make the new liners?

Hugh and Chris Zahrt both mentioned that for the diameter of the drivers and boiler pressure, the spec bore diameter seems a bit large. It should also be mentioned that while the boiler is designed for 200PSI at a factor of safety of 5 with a lot of margin, the engine itself was set for an operating pressure of 180 PSI and our regulations require a factor of safety of only 4. So, we have plenty of room to consider using thicker liners than .250 to allow for future boring out, and can juggle the pressure as necessary to compensate.

Piston and valve ring assemblies are the usual American style, using two compression rings in separate grooves. We KNOW this is an inefficient design which encourages blowby, even before things wear. So, we should come up with a new multiple ring design and appropriate lubrication system for it.

Piston and valve rod packings can likewise be modernized for less leakage and longer life.

It might be feasable to also redesign the valve sleeves and valve gear for better distribution.

Just a couple of comments to get things started:

Is it really necessary for the locomotive to make its full 'original' power in the service for which it will be used? If not, then I'd bush the cylinders smaller (to allow the repeated boring out) and perhaps reduce the boiler pressure without reducing the degree of superheat [EDIT - know now that 110 is not superheated].

I would definitely design for the multiple-ring style of valve, although I do think the largest possible steam-chest volume, and hence thin liners, may be a better optimization than thick 'n boreable. I would be tempted to include the Meiningen-style (cushioned) type of Trofimov valve, which should be relatively easy to machine and then to keep maintained.

Was there ever any sense in two-piece bushings, where the 'inner' bushing was easier to R&R and perhaps made of a better grade of material? That's kind of an intermediate stage between full hard liners (expensive) and hard wear coating or plating.

Robert- one thing you'll see from the diagram I sent out with the group invitation is this engine is considerably "over-cylindered". The factor of adhesion if the engine were operated at 200 PSIG (the stamped working pressure) would be MUCH less than 4 to 1. Even at 180 PSIG, it's still well under 4 to 1. So, yea, it could stand to have the cylinders sleeved down a bit.

Despite having piston valve cylinders, the engine is saturated. I imagine it'd be beyond the group's resources to add superheating but who knows?

One last thing to consider in evaluating improvements in this area- the railway is strictly a low-speed operation; I'd guess 20 MPH is about the maximum they'd ever want to run, and lower speeds are typical. The current operation is a 4 mile out-and-back trip. There is a decent hill at either end of the run, so low speed pulling power is far more important than high speed horsepower.

Hugh

Until all the drawings are uploaded, I've uploaded the cylinder drawings here:

http://www.eliodommusic.com/cylinders%202.pdf

(This is my son's website- if you know anyone that wants to take guitar lessons, live or via Skype, please pass it along!)

The steam passages mostly appear to be pretty decent. The only obvious shortcoming I see is the passages from the valve to the cylinder; these appear to be what Dusty Durrant referred to as "Z" ports, but they're not as circuitous as some I've seen.

Hugh, please explain your numbers and what they mean and how you arrived at them. Some of the members may not know that calculation.

I think it would be best for now to not limit ourselves by worrying about the current operation at NHVRR with 0-4-0T 17. One thing we learned in moving 110 is that she's at the high end of very transportable on our usual truck without major disassembly. It's not inconceivable that 110 could be a roving ambassador to some extent, just as we send 17 to Roanoke for special events. I'd personally love to see 110 named as some thing like the official historic steam locomotive for the State of North Carolina and have a larger career making visits anywhere in the state - Spencer is without steam, so is Craggy Mountain, GSMRR and the guys working on the Newton depot project have access to a decent run of track.

Dave

Robert, thanks. I'm in agreement about multiple rings rather than the two ring pattern....anybody familiar with the FCAF tests on NORA and what that showed would agree. That story starts on page 91 of the Camden print of three technical papers by Porta, which I think is still in print. EDIT Nora was built with the usual American 2 ring plan and wasted about half the steam the boiler made blowing past them and out the stack without doing any work. She was in good condition at the time. 110 has the same sort of ring arrangement. the situation is much improved by replacing the two large rings with many narrow rings.

Anyhow, that leads into what our options might be - What would make Troffimov pattern the best choice? How better than reconfiguring our existing valves to hold more rings? Willoteaux?

And, I'm unclear on a few of the things mentioned in Red devil and in other places.... such as the definition of "diesel quality rings" which could mean any variety of things.....and about lubrication - Wardale says "tangent to the rings" or some such for supply - huh?

Dave

DaveLathrop57 wrote:Hugh, please explain your numbers and what they mean and how you arrived at them. Some of the members may not know that calculation.

A little background: the amount of starting pull a locomotive can achieve is limited by the coefficient of friction between the steel tires and steel rails which is generally taken to be 25%. In other words, if you have a load of 1000 lbs on a piece of steel that's sitting on top of another piece of steel, you can push on it without about 250 before it starts to slip. Or looking at it the other way a pair of steel wheels supporting 1000 pounds can exert about 250 pounds of tractive effort before they start to slip (clean, dry rails and wheels). For locomotive design, you try to keep the tractive effort equal to or less than one-quarter of the weight on the drivers, or expressed another way, you try to keep the weight on the drivers at least four times the tractive effort. This number is referred to as the Factor of Adhesion, and is normally 4 or greater.

Factor of Adhesion (FOA) = weight on drivers / tractive effort

The tractive effort for a 2-cylinder steam locomotive is calculated by the formula TE = (d (squared) x P x S x 0.85)) / D

This translates as Tractive Effort = ((diameter of cylinders squared) x (boiler pressure) x (cylinder stroke) x 0.85 / (diameter of driving wheels). The "0.85" is a coefficient for the loss of pressure through the throttle valve, piping, superheater elements (if present), valves, and steam passages.

In the case of #110, TE = (16) squared x 180 PSIG x 24 x 0.85 / 44 = 21,364 pounds

According to Vulcan's records, the 110 carries 78,000 pounds on the drivers, so

FOA = weight on drivers / TE
= 78,000 / 21,364
= 3.65

which is significantly less than 4. This would seem to indicate #110 would be a very slippery engine, as the cylinders can exert more tractive power than the drivers can put to the rails.

I certainly don't think this was a goof on the part of Vulcan's designers. I'd guess for an engine in this service having ample tractive effort under adverse conditions (low boiler pressure, worn rings, valve timing off, etc.) was considered pretty important.

So, to maintain a Factor of Adhesion of 4 :
weight on drivers / 4 = 78,000 / 4 = 19,500
= target tractive effort.

You can work backwards to determine the cylinder diameter which will give that tractive effort; it comes out to about 15.29 inches. So, you could install thicker liners in the cylinders and bore then to 15-1/4 inches or so, and you'd have the desired FOA of 4.

Hugh "Low Water" Odom

DaveLathrop57 wrote:Robert, thanks. I'm in agreement about multiple rings rather than the two ring pattern....anybody familiar with the FCAF tests on NORA and what that showed would agree. That story starts on page 91 of the Camden print of three technical papers by Porta, which I think is still in print.

Anyhow, that leads into what our options might be - What would make Troffimov pattern the best choice? How better than reconfiguring our existing valves to hold more rings? Willoteaux?

And, I'm unclear on a few of the things mentioned in Red devil and in other places.... such as the definition of "diesel quality rings" which could mean any variety of things.....and about lubrication - Wardale says "tangent to the rings" or some such for supply - huh?

Dave

Porta was not a fan of Troffimov valves (excess weight, no need for by-pass function if mid-gear drifting is used) nor Willoteaux (they give double the steam port opening for a given valve travel, but they're heavy and complex, and have circuitous steam passages which result in high pressure drop and high heat transfer). All his designs used light steel fabrications with LOTS (~10 or so at each end) of very narrow piston rings. Some include "exhaust diffusers" at the ends, which are sort of a 360 degree "turning vane" which makes the exhaust steam sweep against the inside of the valve body on the backside of the ring grooves to cool them.

I agree "diesel quality" seems to be a somewhat nebulous term, but based on what Wardale applied to the Red Devil, I'm pretty sure he meant "diesel quality" in terms of materials (like high chromium cast iron) and finish. It shouldn't be too hard to find out what current state-of-the-art production diesel engines use and strive for something close to that.

Regarding lubrication being applied "tangent to the rings", I'm pretty sure i understand this. You're trying to get the oil to spread on the inner face of the valve bushings (or cylinder liners). The lubrication holes would be drilled perpendicular to the valve bushing axis and tangential to the valve bushing bore. Imagine standing beside the locomotive by the cylinder casting- you drill a hole straight into the casting so that the hole just intersects the valve bushing bore at either the top or the bottom. The lubrication hole placement is determined based on the valve travel. The holes should be placed so that the oil always squirts in between the valve rings, so most of the oil stays on the area where the rings "rub" where it can do the most good. I think Porta envisioned that the oil would squirt into the space between the rings and be carried around the entire circumference of the valve, where it can "wet" all the rings and the entire rubbing surface of each valve bushing.

This method is particularly important in engines with high superheat. The standard U.S. practice of atomizing the oil into a flow of superheated steam results in the oil being needlessly "cooked" before it can provide any lubrication. By applying the lubrication directly to the rubbing surfaces, most of the oil winds up exactly where you want it, and the oil temperature remains much lower (especially if cooling of the rubbing surfaces is provided) so that much higher superheat temperatures can be used without requiring "super" lubricants.

If #110 remains saturated, I doubt this would be a major consideration, but I suspect the "tangential" lubrication would still allow you to use a lot less oil while still maintaining low wear rates, which might be a significant operational cost savings over time.

Hugh

I always thought of 'diesel quality rings' referring to the metallurgy and controlled gap, etc. that have been developed for large crosshed-diesel-engine technology. Some of this of course has to be modified for the wet environment (remember the abortive attempts at hard-chroming steam cylinders?) but in general the approach is a positive one.

I understand 'tangent to the rings' as meaning direct supply at right angles to the ring surface (e.g. through ports circumferentially spaced around the bore at the right place, or via pressure through small ports in the valve body. I think Wardale preferred lube through the mass of the cylinder block to positive-pressure lube through the valve body, but that consideration won't apply on a saturated engine. Of course, modifying only the new valve with oiling passages and ports is quite a bit cheaper, and doesn't affect the 'legacy' block or new bushings... you'd run the pressurized oil into the valves with hardline/flexlinesup along the radius rod and into a hollow valve rod. Might be similar arrangements for the piston, as reduction of piston mass is not as significant on 110...

... although it might be an interesting part of the exercise to see exactly how well-balanced this engine can be made, and experiment with some of the theories on building cheap in-situ balancing and wheel-turning arrangements for these locomotives... or for providing a portable balancing and profile-grinding service for steam railroads...

Seems to me that one big advantage of tangential lubrication on pistons and piston valves is that you no longer need viscous goo for tribology. I am not expecting oil carryover in the exhaust to be an issue on this engine, but it's possible EPA regs would come to cover it.

Can we PLEASE standardize on 'Trofimov' for the Romanized spelling, as that's what Meiningen used and they are the folks who gave us the modern configuration of these things?

The point of using these valves, WITH THE MEININGEN CUSHIONING PISTONS, is to make the engine more tractable to run. I see very little compromise necessary with the multiple rings; in fact, they make the movable heads more self-locating. Saturated = less tendency to coke or glaze the lubricant; these valves should run very well if sensible hard coating is done on the bushings where the valve slides on the rod. We can always provide a positive lock in the rod design to eliminate the bypass action if that turns out undesirable for some reason.

As I see it, the only reason to superheat this locomotive is to make it less prone to carryover if operated out on a main line somewhere. We do not care about minimizing water rate, really, and we don't particularly care about loss of efficiency due to nucleate condensation at short (effective) cutoff.

It does occur to me that we could build a comparatively cheap inertial 'steam dryer' with an electric motor spinning vanes in the dome instead of using purely passive means like the Elesco steam dryer approach. That might address some of the potential carryover issues, and perhaps lube or lock problems due to water, at high steam demand rates.

Please tell us more about the cushioning pistons - i'm not familiar with that design.

Dave

I have a picture of the general arrangement in the steam_tech Photos somewhere. There is Meiningen's 'flavor' and a slightly abridged version using Wagner-style valving.

Basically, the valve head is counterbored, so it incorporates a 'cylinder'. A small piston head with appropriately-sized 'dashpot' vents is placed on the valve rod at each end, so that it is engaged but not fully bottomed when the valve heads hit their stops during steam readmission. The combination of ring friction and steam displacement 'cushions' the shock of the heads hitting the stops (thereby, in theory, eliminating both the 'clink' and the need to manipulate throttle and cutoff delicately to avoid accelerating the valve head halfway down the valve cylinder to slam destructively into the hard stop).