Steam & Excursion > 3-cylinder steam sound?

The first recordings I ever heard of a 3-cylinder steam locomotive are the ones at the following link, UP 4-12-2 #9009 recorded by Howard Fogg in 1954:

http://www.utahrails.net/up/up-4-12-2.mp3

The first sequence ("easing around a freight derailment on temporary track") features exhausts that aren't at all evenly spaced. The second sequence ("we heard 9009 leave town later") has exhausts that are more or less evenly spaced, but the sixth exhaust is notably louder than the other five.

I've recently come across LP's with recordings of other UP 4-12-2's and SP 4-10-2 #5021. In those recording, the time signature is still definitely six beats to the measure instead of four, but the beats are fairly evenly spaced and of roughly equal volume, unlike what we hear 9009 doing.

Can anyone explain what's going on in those 9009 sequences that makes the exhausts uneven in the first and pulsating in the second?

Evan_Werkema Wrote:
-------------------------------------------------------
> Can anyone explain what's going on in those 9009
> sequences that makes the exhausts uneven in the
> first and pulsating in the second?

It could be as many as three things.

1. On a two-cylinder locomotive, all four exhaust passages from the pistons to the blastpipe are usually reasonably close to the same length. Since the exhaust steam travels the same distance, the exhaust beats match up well with valve events. On most 3-cylinder saddles, the middle cylinder exhaust passages will be of a different length than the outside cylinders - they may both be shorter, or one may be shorter and one longer than the outside cylinders (depending on the angle of the bore, etc.). Since the exhaust steam travels different lengths, the exhaust beats won't be regular.

2. Not all 3-cylinder engines had their cylinders set at 120-degree intervals. I can't recall which engine, but I remember reading about one that had the middle cylinder crankpin set at 117 degrees relative to one cylinder (and therefore 123 degrees to the other).

3. If the valves aren't square, that will add to the 'offbeat' sound.

The difference is more noticeable at slow speeds than at high speed, for two reasons: the exhaust beats are farther apart at slow speed, and at high speed the exhaust from one cylinder will actually 'pull' the exhaust of the next cylinder along, speeding it up and shortening the time required for it to get to the blastpipe.

JAC



Edited 1 time(s). Last edit at 02/03/11 06:01 by johnacraft.

Evan_Werkema Wrote:
-------------------------------------------------------
> I've recently come across LP's with recordings of
> other UP 4-12-2's and SP 4-10-2 #5021. In those
> recording, the time signature is still definitely
> six beats to the measure instead of four, but the
> beats are fairly evenly spaced and of roughly
> equal volume, unlike what we hear 9009 doing.

UP 9009 sounds just like a worn out 3 cylinder engine should, while your other recordings are probably engines who's valve gears are in good shape. One thing about Gresley valve gear is that any lost motion in the two outside Walschaerts valve gears is magnified by the levers for the 3rd cylinder's valve.

Notice the recording was made in 1954 and I would imagine maintainence on them was down to a minimum.

I purchased that record when it was new and haven't played it in years, no turn table. Sure brings back memories. Thanks for posting.
Blainem

> 2. Not all 3-cylinder engines had
> their cylinders set at 120-degree
> intervals.

Simple 3-cyl engines are always? supposed
to have "cylinders" set at 120 degrees--
that is, they reach front dead center at
equal 120-degree angles of driver rotation.
Since the 4-12-2's center cylinder is
inclined 9.5 degrees, the center crank
has to be 9.5 degrees away from the
120-deg position.

Evan,

These guys are all right to one extent or another.

9009 sounds dreadful, especially at low speed. She was very clearly completely out of time. I am fortunate to have a number of recordings of 4-12-2's and not a single one sounds anywhere near as badly as poor 9009. She was one of the "bald-face" engines in which the Gresely gear operating the valve for the inside cylinder was replaced with a Walschaerts gear, arranged via addition of a new eccentric crank on the right side which was constructed such that it had pins arranged to drive two eccentric rods, which of course operated two links. As usual, these links drove radius rods in the normal manner, but the gear for the inside cylinder utilized a swinging link attached to the front of the radius rod to transfer motion across the engine, behind the cylinder saddle, to operate the valve for the inside cylinder. Union Pacific referred to these eight 4-12-2's as "3rd link" engines. While the gear driving the valve was changed, the valve dimensions did not change, nor did the crank pin on the main wheel driven by the inside cylinder (2nd main axle).

John makes a good point, too, about the length and size of steam/exhaust passages. They definitely differed on the 9000's. Thus in any six-exhaust revolution two exhausts are always louder than the other four. Even when in good dimension, they will sound "ONE-two-three-FOUR-five-six." I have always believed this difference contributes to guys referring to their sounds as "off-beat", even if they're very nearly perfectly spaced.

John E Bush
Omaha

To add to both the John Craft and John Bush statements above, I can tell you that all those 3 cylinder German 2-10-0 and 4-6-2 steam locomotives have that same "3 cylinder beat" thru out their speed range. Starting a heavy train is obviously the most noticeable.

Lima made quite a few three-cylinder engines for logging service.

Stan :)

Does the 9000 in Pomona still have the Gresley gear, or does it have the triple Walschaerts?

9000 was never converted. In fact, the entire program ceased when the remaining engines in classes UP-1 & 2 (9000-14) received new Gresely levers fitted with the SKF roller bearing assemblies which all the other 9000's received new when constructed later, 1928-30. Those bearings made a material difference in the lost motion arena.

John Bush
Omaha

ALCo had the rights to the so-called Gresley derived valve gear (actually invented by an assistant, Harold Holcroft, but refined by Gresley) in the U.S. Gresley's love of the derived valve gear for the inside cylinder was a major Achilles heel of his designs. Any wear in the joints caused unequal steam distribution in the third cylinder, and therefore increased (or decreased) piston thrust on the inside main bearing. They were high-maintenance locomotives everywhere the derived valve gear was used.

The gear was applied to several large locomotive designs during the three-cylinder fad in America during the mid-1920s. Most famous and longest-lived of these were the UP 4-12-2s. 88 were built, 1926-30. They successfully increased the speed of freight trains in the flatlands but were a maintenance nightmare. Eight of the first 15 were rebuilt with a double-Walschaert valve gear. Those built from 1928 onward had fewer problems because they used roller bearings in the joints, resulting in far less wear.

With regard to sound, Gresley three-cylinder locomotives usually had a distinctive three-part exhaust beat, even if the valve gear was in tip-top condition. There are hundreds of sound recordings of ex-LNER locomotives that will illustrate this.

Vinton Wight's recordings, now available on CD from Smithsonian-Folkways, contain sounds of UP 9007 (Sounds of Steam Locomotives, vol. 1) and 9052 (Sounds of Steam Locomotives, vol. 5). The latter sequence is of most interest. 9052 sets out some cars at Lincoln NE, then leaves town -- more than four minutes of sound. 9052's exhaust is heavily syncopated, suggesting that not all the wear problems were solved by using roller bearings on the Gresley valve gear joints.



Edited 1 time(s). Last edit at 02/19/18 22:09 by jpp452.

Now you've got me wondering what all those old Santa Fe (and other) four cylinder compounds must have sounded like. It would be rare if a sound recording of those existed today.

There are recordings out there of Mallets, such as N&W 2-8-8-2's, in the 1950's. The high pressure cylinders exhaust into the low pressure cylinders so make no sound. The low pressure cylinders exhaust is a four square exhaust pattern, but with a very low exhaust pressure sound - less chuff, more whoosh.

Other four cylinder compounds (Vauclain, Balanced, Tandem, etc.) would sound similar. A two cylinder cross compound would also be a low pressure exhaust, but only two exhausts per driver revolution.

-LD