1102 Box Canyon Road, Fallbrook, CA 92028
This yarn is inspired because of a happenstance at two consecutive Old Time Thresheree shows I happened to attend in the not-too-distant past. Everything was going along hunky dory with all the exhibitions in full attractive demeanor, right up to the parade sequence, when along labored a fine old steamer with the most terrible case of asthma that I have ever heard in my life! She was pounding on only one end and, had I been blindfolded, I would have taken her for one of those old 1910 Titan gas tractors that could really shake the earth as they rolled by. As this venerable specimen passed our reviewing stand, it became at once evident wherein all the difficulty arose. This fine animal was equipped with a Woolf valve gear and also a rocker shaft to increase the valve throw beyond block drive. As she banged her way along the route leftward in front of us, it became apparent that this rocker was being exercised by an extreme over-length in the adjustment between the rocker and the block drive. This caused the upper section of the rocker to oscillate too far rearward, with the result that only the head end of the cylinder was receiving and exhausting steam; the rear port hardly being uncovered at all which most detrimentally transformed her into a single-action job not at all intended by the manufacturer. A monkey wrench and ten minutes should have brought her back into the 'ball park.' Actually, a good steam man can set a valve very closely by simply adjusting the valve travel in one direction until it is clearly evident that she is coming short on one end; then adjusting travel in the opposite direction until the same unbalance becomes evident. By counting the turnbuckle turns between extremes, and splitting the difference, it will be found that operation is way above reproach. This, of course, is a quick and dirty method to be resorted to in the field in order to keep the crews busy until after closedown when the engineer's duties seem to never end.
But now, to digress a bit in general on valve gears, which, after all, are about the most complicated element of the old steamer other than the injector (which, since it appears to have worn out and will not respond to cleaning any more, must be replaced) and the force-feed lubricator (which at most may require cleaning to remove a bit of debris from under the discharge valve, as evidenced by the fact that it has been pumping oil all day over the sight feed but the reservoir is still full!).
Ever since Watt discovered that a cam or eccentric could replace the labors of an attendant who was originally required to operate the valves on the New comen pumping engines, and engines became double acting, the subject of proper valve setting also became a subject to be mastered by the operating engineer. Now, today, we may take so much for granted. If it be a simple mill engine, all that is required is to equalize the valve travel over opposite ports; then adjust the eccentric for the desired lead. If a tram and marks have not already been provided, it requires but removal of the valve compartment cover and a few course measurements. Rocking valves are a bit more tedious, but true to form while going to Stevenson link motion is a repetition of the first case above requiring only that each end of the link be matched in accordance with direction of motion. With 'open' method of link drive as was common in locomotive practice, it is interesting to note that these old locos would run at speed in either direction with the 'Johnson Bar' hooked up to true center (the ports became slightly uncovered to inlet steam at dead centers).
But, eventually, and originally in marine practice, along came the development of the 'radial' type of valve gear. There are quite a raft of patents applicable to this Rube Goldberg contraption; but they all accomplish the same duty with the same effectiveness. These gears soon became applicable to railway practice because of their accessibility, and were competitive with the Walschaerts type of outside valve gears (which later were often mis classed as of radial type). The Baker and Southern were the only two radial types of locomotive gear in extensive practice. As Stevenson dropped out mostly because of its inaccessibility, Baker and Walschaerts absorbed most of the field. However, Young (who also designed a rotary locomotive valve) did find a small installation for a valve gear which resembled Walschaerts, except that, instead of utilizing a crank drive for the link off the main driver, it utilized the cutoff drive from the crosshead on one side to not only furnish that function but also, through a short connecting link, drive a rocker shaft that extended across the locomotive frame and supplied drive for the valve on the opposite side of the locomotive. All valve drive functions were thus derived from only crosshead source.
It is well to point out here that, if a single fixed eccentric is to be utilized on a reversible engine, it must necessarily be set at 90 degrees from crank position. But an eccentric (alone) set at the position does not allow setting at proper lap of the valve. Hence, since proper eccentric lead is in the neighborhood of 120-135 degrees, an auxiliary motion is required. This is ordinarily supplied by the auxiliary drive off the crosshead. If you recall your plain geometry, the resultant vector derived from two other vectors 90 degrees apart, may result in a final vector anywhere between zero and 90 degrees (depending upon the respective length of the combined vectors). By utilization of a rocker linkage, the resultant vector may become one of 90 to 135 (or more) degrees. Thus, with Walschaerts, we combine 90 degrees off the main crank throw with that of zero (or 180) degrees off the crosshead and wind up with our desired 120-130 degrees.
A unique scheme of utilizing a single eccentric for both forward and reverse action is found in the Russell, which shifts the eccentric from +120 degrees (or whatever may be necessary) to -120 degrees. This gear may also be 'hooked up' as with any other motion EXCEPT that of Marsh-bless his heart. With the latter, one cannot work his steam expansively except through governor action, which is generally regarded as uneconomical. For this results in 'wiredrawing' of the steam; in which case working pressure is reduced through an orifice without accomplishing useful work. For, when any gas is expanded through throttling action it becomes cooled according to fitting 'laws' of thermodynamics. And while on the subject of Russell, with its double-ported valve, it must be realized that even though twice as much (inlet) valve is uncovered per lineal length of motion, there is yet wiredrawing near opening and closing of this port. Double-porting is irrelevant to the exhaust port; - that end has to take care of itself. And, while we are on the subject of Russell, it may also be noted that this firm's method of tandem-compounding was regarded as superior to Mr. Woolf's in that each cylinder had its respective valve and steam chest; which not only avoided long wasteful steam portages from high to low pressure (there can be no power developed through expanding steam in long passages between cylinders) but also allowed the low-pressure valve chest to act as a receiver as utilized in cross-compound cylinders. I have diverted a bit off our prime subject here because I have been rather unable to submit written matter more often; and have thus been putting my thoughts into words. Please excuse the folly. Back to gears!
I was going to ask 'Where does the nomenclature RADIAL come from with respect to steam engine valve gears?' Well, if we refer to some of the old cyclopedias and patent records regarding valve gears, we will find that at some early point in time marine engineers must have deplored double eccentrics and linkages for every cylinder on their old (already cumbersome) multiple expansion engines. So some ingenious chap who was really quite smart at the time devised a (movable pivoted) swinging linkage which, by means of changing the pivotal point of the (radius) link could cause the valve motion to change 180 degrees. These old marine engines were of vertical type, of course, to fit the ship's enclosure. But the gear found widespread adaptation in both railroad and traction usage, both of which utilized horizontal engines. There were many variations found in this gear; some of which appeared to do hardly more than change a point of support or suspension. But, like many ideas, some of which were mechanically impossible (the Patent Office could not have required working models in every instance for the inventor) the variations found themselves protected for at least seventeen years. And now, since this manuscript is framed for the dear old steam traction engines, let me direct your attention to Figure One, which I accomplished with an old set of drafting instruments furnished with an American Correspondence School course in electrical engineering way back in 1919 and which proved so much over my pumpkin head, without a close instructor, that I was forced to give it up rather pronto after having been coerced into paying some $400 for it! I have used the set ever since; but it is doubtful that I have derived $400 worth of usage out of it in 72 years. Anyhow, back to our first sketch.
This is a line drawing design layout for a genuine radial type gear and is identical to that as employed on both my Monarch double-cylinder engines; one of which is installed on Luellabelle (see the cover of IMA, Nov-Dec. 1981) and the other, one of the finest engines you ever could see, is off their traction, and is now on the exhibit steam line over at our fine museum in nearby Vista. The Monarch line was manufactured in Groton, N.Y. and Luellabelle's original (roller) came off line in 1925. But now back to our Fig. 1 which is not drawn to design scale but simply intended for illustration.
'A' is the crankshaft; 'B' is the crankpin located at OUTER DEAD CENTER; and 'C is the crosshead wrist pin. Now, the central position of travel of the source end of the RADIAL arm 'D' and the load end of eccentric strap 'E' is shown at 'F'. As will be pointed out later, 'F' position is also the center of travel for the sliding block in the later Woolf version of this original design. 'G' is the movable anchorage point (by the reverse lever) for the Radial link. 'G' may thus be moved (in three or four increments) to G-l, one direction of rotation for the engine, or to G-2 for reverse operation. The displacement of 'G' is usually accomplished by a large bell crank which is anchored at the same position as 'F' and therefore is not shown on sketch for simplicity. This arm must be the same length, center-to-center, as 'D'; otherwise (at center throw of the eccentric) the valve would be disturbed from its present location. If it is detected in practice that such slight movement of the valve is thus caused, it is termed 'slip' and must be corrected as described later.
Now, one remaining remark at this point. The anchorage for the bell crank arm (which is hidden behind, and coincides with 'D', and is also the center of sliding block travel in Woolf design) is usually mounted vertically above the crankshaft (it is mounted below in Reeves design). However, it may also be mounted ahead of the crankshaft line as shown, by the angle 'X'. In any case (Case?) the angle 'Y' must be preserved as 90 degrees; for, remember, we are keeping our eccentric throw at 90 degrees from the crank throw. The right-angle takeoff accomplished this, while the direct angle between crankpin and eccentric is 'in phase' at zero degrees (except our case of displacement through angle 'X'). It is also called to attention that, if a rocker arm is employed in the valve stem drive, the eccentric might be mounted at 180 degrees from the crankpin (or the position of the reverse lever made opposite).
Now, only one item remains for consideration; the takeoff for the valve stem drive through the linkage arm 'H'. This fixed appendage T to the eccentric strap must be mounted to same, in a dimension not only offset from the strap, but also at an appropriate proportional intermediate distance between the ends of same in order to provide the desired 'lead' for steam admittance to the cylinder. It is seen that the resultant motion at this point is a combination of both eccentric and the swinging link 'D' as it rocks up and down through the end points' J' and 'K' and through the arcs 'L-l' or 'M, M-l' according to direction of rotation of the engine. 'N, N-l' of course is the line of travel when the reverse lever is placed on center if the crankshaft be rotated.
The working layout for extension T, the wrist connection to 'H' of which describes the oblique ellipse like path 'O' was originally determined by pattern layout, and really did not present a serious task. Today, if such need were felt, its layout, together with all other elements of such gear, would become the product of a bit of computer manipulation through usage of graphics and aid design software programs. And while I have been struggling over ten years with these fantastic electronic contraptions, I am not prepared as yet to tackle this sort of problem.
While 'O' is rather an odd figure, it suggests that steam distribution might not be the same at both ends of the cylinder; and may not even yield the same available power from the cylinder in forward motion compared to reverse. This is true; but when we recall that distribution is measured in proportional valve opening for corresponding points of piston travel, and the piston travels further on the head end than the crank end for quarters of shaft revolution, we can readily appreciate the value of an indicator diagram when making a setup to realize the same amount of cylinder power developed between both ends. At the Case factory (and I suppose at others as well in the Good Old Days) all steam engines were required to measure up to an excellent indicator diagram on final test. Yes, the angularity of the connecting rod and the ips-switching of valve gears play great antics. An article on the first-mentioned topic was written by this author for IMA several years ago; while a splendid article on forward and reverse power was written in our same publication more recently. Please don't make me dig those editions out of my files right now.
We may now readily see how Mr. Woolf decided to modify things a bit, by simply replacing 'D' and its supporting structure with a grooved crosshead guide, so to speak; and placing a sliding shoe within the slot way so as to move up and down in a fashion resembling the arcs 'L' and 'M'. To retain similarity, both Case and Avery called attention to their 'curved block' in their advertising. However, it is my belief that Aultman-Taylor and some other manufacturers utilized straight blocks. If the RADIAL arm 'D' had been of infinite length, then, according to geometrical precepts, the resultant arc 'N, N-1' for instance, would become a straight line. But the utilization of a curved block affords more acceleration in combination with the eccentric sweep, such that the two ends of the drive, working in parallel, thus cause quicker valve opening and closure with less wiredrawing of steam for which good feature radial type gears are noted. However, 'hooking up' a radial gear does not allow for advancing the valve action at a speed similar to that rendered by the Stevenson link motion.
At this point, we might examine some of the quirks which should be checked when examining valve setting to determine whether the eccentric may have slipped or excessive wear occurred. This is a one-two-three operation in that order. Referring to Figure 1, which depicts a layout for the reverse head bracket, the block guide shaft is in upper bore 'P' while the crankshaft runs in 'A'. When the crank is placed on center as in Figure 1, and the eccentric is at half travel (the angularity of the eccentric may be accounted for similar to that of the main connecting rod) and the reverse lever is rocked through neutral position, the valve should exhibit no movement during this lever movement. If there is 'slip' in valve movement, it means that point 'P' in the upper bearing of Figure 2 has likely fallen due to wear or mis adjustment. The remedy for this error is to adjust the shims provided at 'Q' to raise or lower 'P' until minimal slip is observed. (In Figure 1 this must allow the distance F to J to equal that of F to K.) To digress a bit further at this point, it must not be construed that, while the eccentric is at full travel in Fig. 1, and the strap 'E' is at half travel (point 'F') this should lead us to believe the valve is also at half-travel. For the latter should be at 'lead' setting for the crank end of cylinder as shown. The lead is thus caused by the offset in the strap extension T, which substitutes for the lap-and-lead lever from the crosshead in the Walschaerts type of gear. In fact, the two gears bear very close resemblance in action, for, when the eccentric crank of the Walschaerts is at mid travel, the piston is at full travel, and the lead lever just uncovers the port at that end of piston travel.
Secondly, let us tackle Figure 3 which illustrates the linkage between the reverse lever and the block guide. A turnbuckle or other means should be provided at this juncture to allow for adjustment such that, with the reverse lever on center, when turning the engine over, the valve should exhibit the same closure over either cylinder port.
Thirdly, referring to Figure 4, is shown a linkage adjustment, in this case through a reverse rocker, between the eccentric strap takeoff and the drive to the valve stem. It is simply an equalizing turnbuckle (or other length-adjusting arrangement) which allows for setting the full opening of one cylinder port to the same amount for the opposite port. From indicator diagrams, it may be found that a bit of 'fudging' is required for port opening and/or lead to obtain a well balanced power distribution for both ends.
The above procedures were also prescribed by that fine New England trained mechanic who came to dwell in Oregon a long time ago: Charles 'Pop' Arnold. And I am reminded to say that, in those older days, we always looked to Rhode Island, Connecticut, New York and Pennsylvania or New Jersey to find the best. We had a splendid journeyman railway machinist when I was an ambitious kid boiler maker and machinist helper on the C&NW in the Dakotas whom I shall always remember: Ray 'Yankee' Cooke; like the Rhode Island locomotive builder. Our chief boilermaker, Peter Paluzak, also hailed from around the Polish center in New Jersey. But after WWI, the cultural center of our country began moving westward and it has never ceased to this day. Had the recession of 1920 never occurred, likely I may have remained in the railway service through my active life. But, like the old saw goes, 'Nothing changes like change', and I have been spinning ever since.
Hah, I just glanced up at one of the railway locomotive pictures on the wall: a Southern lokey with their Joy radial valve gear. Its unusual arrangement reveals the suspension linkage (as shown at 'G' in Figure 1) mounted horizontally above the cross head wherein 'G' is slid back and forth by the reverse lever in what resembles a curved second crosshead guide. There are so many variations in this type of gear; but let no one confuse you that Walschaerts is of this class. That old boy simply combined two crank motions which were out of phase by 90 degrees to derive the differential angle of drive as explained previously (it is worth repeating). In the electrical-electronic field such combination of vectors is utilized most extensively in demodulating color television signals, changing digital to analog information, etc.
Had our dear old steam locomotives not been superseded by Diesel power, it is likely that the poppet valve designs which were cam operated as driven by gearing (at that time being developed by Caprotti, Italy, and Franklin, USA) would have resulted in making automobile mechanics out of every chap who aspired to becoming a steam engineer; for these locomotive designs would have been carried over into our traction engine fields. They actually were incorporated in the later stationary steam engine designs as exemplified by Ames in their Uniflow engines-one of the closest approaches to steam turbine efficiency.
Well, this brings me to the point where I am reminded that another old Shanty Irishman once stated to me during an explanatory discussion, 'Now I have told you all I can about something I know nothing about.' At any rate, please do not call my attention to some old steamer which is parading down the show path with one governor ball missing, lest I may feel constrained to knit another yarn. Of course, I am always appreciative of criticism which begets more learning; so let any hailstones fly which may. It's all fun, you know.
So good going in there, you fine fellows; and keep those wonderful relics huffing and puffing. See you there!