The Camelback type of locomotive could cause serious injury/death if a main rod were to break and come swinging through the bottom of the cab.

Matt- my post was answering the question about what happened with the Mallard during its record setting run in 1938.

I agree with your assessment of the report of the damage to Tornado.

Yeah, I heard about that. Poor Tornado.

The Winter 2008 PRRT&HS Keystone shows in December 1944, PRR had tested N&W J 610 on the Fort Wayne Line between Crestline and Chicago. 610 was the last of the 6 wartime J's (605-610) that had been built without streamlining or roller bearing rods as Class J-1. 610 was upgraded to Class J in November 1944 and delivered to PRR on December 4, 1944. The Ft. Wayne was PRR's racetrack and where a J with 70 inch drivers was least suited, but it was also the only place the big engines would meet the clearances.

The 610 "handled such trains as 57, 77, 56, 78 and 28 with ease, hauling 15 and 16 cars." On December 10, 1944, 610 suffered a left valve gear failure with a bent connecting rod, bent eccentric rod, loosened eccentric crank and damaged reverse gear quadrant. Cause was lack of lubrication in the left valve chamber. Repairs took 8 days and the testing continued until January 3, 1945.

Later in 1945, PRR ordered 50 production model T1 4-4-4-4's. EMD E7A's 5900-5901 arrived September 20, 1945.

Phil Mulligan

I suspect Matt and EJ are correct; the 'parallel' with valve overheating is likely similar to that experienced with the N&W J in high-speed testing on PRR. Does not take much excessive friction overheat to destroy the 'rest' of the available tribology and then start heating the valve spool into more and more runaway friction heating, faster than the contact with the liner can carry it away (or expand the ID of the liner). As I recall, the cage of the affected valve was heat-blued, some of which might have occurred 'after the failure' as there shouldn't have been oxygen in the steam either at intake or exhaust in sufficient quantity to form the oxide film.

Mallard's big end had identifiable 'design weakness' but I cannot remember exactly what it was; it was documented in at least one British reference and I suspect Hugh Odom can find and describe it. That reference indicated that the detail design of the big end was revised afterward and it was indicated that there would have been no 'repeat' of the failure had subsequent runs to 125mph been made.

On a new build, such as the Tornado or the T-1, what would the incremental effort be, in terms of possible locomotive design modifications and cost, to include real-time sensors at key components? Data could be fed to the cab or to the tool car for a crew member to monitor.

70000 posted an email purportedly from the tour operator which stated, in part, "The locomotive has run many miles since winter maintenance with no obvious issues in this area. Whilst we have been asked if the higher speed was a factor in the failure, at present we believe it not to be. It probably occured slightly earlier in the journey than if we had been running at lower speed, but it would most likely still have happened.”

Depending on how the lubricant is fed to the component, and on the characteristics of the lubricant, is it possible that conditions exists such that, above a certain speed, as-built the lubrication requirements exceed the system's ability to deliver adequate lubrication?

On the Tornado, is the middle cylinder between the frames a full pressure cylinder or is this a larger diameter, lower pressure compound cylinder?

Real time sensing of lubrication has already been done to some extent on several existing locomotives.

For instance, the N&W 611 and I'd guess all the other class J locomotives were equipped with a cab mounted gauge panel which showed the flow from individual lubricator feeds, although it was apparently not restored when the engine was restored and I believe it was later removed. Something like this would show if there was a problem with lubrication flow to the valves.

Grand Canyon 4960 had (still has?) digital temperature readouts from the driving axle bearings, and I believe there was at least one other engine with a similar setup.

My guess is that speed would rarely, if ever, be the cause of a lubrication breakdown. Mechanical lubricators for steam locomotive valves and cylinders are generally driven off the valve gear such that more oil is delivered at long cutoffs, which would be at lower speeds. As the cutoff is reduced (engine speeds up), less oil is delivered, which seems to show that less lubrication is required at high speeds than at low speeds. Lubrication can break down due to excessive superheat temperatures, but my guess it was simply a reduction or interruption in oil flow caused by a clogged feed or something similar.

Tornado uses simple steam expansion; all 3 cylinders are the same size.

Doesn't seem likely, as the engine was tested at speeds up to 100MPH just last year:
http://www.rypn.org/forums/viewtopic.php?f=1&t=40595

Mr. Storzek - I saw the high speed test videos. Very impressive.

Whodom - Thank you for the information. I know the mechanical feeders are triggered by the valve gear, and that the travel is tied to the travel of the valve gear, so, as you stated, as speed is increased and the cutoff is reduced, a smaller amount of oil will be introduced for each quantity of steam per cycle but at a greater rate of cycles per time period. For ease of the model, I am assuming that the shorter travel of the oil valve somewhat corresponds to the reduced amount of steam per cycle as the cutoff is reduced so that the vapor pressure of the oil in the unexpanded steam is somewhat constant. At the most reduced cutoff, the steam is expanding the most and so the vapor pressure of the oil when the cylinder is fully expanded (at full travel) is at its lowest, and so its mean free path to the surfaces that require lubrication is the greatest. As speed is further increased, the oil has less time to get to where it is needed. So I thought perhaps there could have been a speed where, at fully reduced cutoff and thus minimum oil vapor pressure in the expanded steam, the required amount of oil could not get to where it is needed in the short cycle time available. It will be interesting to see what the skilled steam craftsman in the Tornado organization find when they go over the locomotive.

As far as sensors go, I was thinking more along the lines of temp and pressure sensors and perhaps some strain gauges. I did not know if a basic oil flow monitor would show a slight decrease in flow capable of possibly causing issues at such high speeds/cycling rates.

Update from the A1 Trust:

This drawing of the inner cylinder's motion was also posted:



Link to full update: https://www.a1steam.com/category/news/

There was a further update on what they have found so far, together with likely work required, which was uploaded to their news page on Friday 20th for those interested..........
https://www.a1steam.com/category/news/

Good Morning,

Hugh, thank you for posting the internal layout drawing of the center cylinder driveline.

Considering the center main rod, two very large counterweights valve eccentric, all confined between two frame members with limited space, makes the thought of all this whirling at very high speed a bit hair-raising!

This has been a very interesting discussion on lubrication. I would be interested in investigating the cab gauge panel to monitor lubrication rates which was installed in the 611. Assuming this was all mechanical, it would be interesting to understand how this system worked.

For our reciprocating gas compressors we have DNFTs (Digital No Flow Timers) to shut down the compressor if it operates too long without a lube pulse. This is connected to the first spool shuttle valve on the distribution block. The distribution block can be a temperamental if you get air trapped in the oil lines. The DNFT could generate false alarms but at least it would alarm to you of a lube flow issue. These are electronic timers which we have the luxury of tying to the unit PLCs. How the scheme would survive in a steam locomotive is hard to tell. Our two and four throw compressors are pretty violent in regards to vibration. Perhaps a steam locomotive could be a worse application due to weather conditions, road and engine vibration.

This is unfortunate and appears to be a lube point failure, however, If it is a design issue, it will be interesting to see how the A1 trust addresses it.

A question I have for the forum......is less lubrication at shorter cut-off desired? Would it make sense to tie the lubricator crank arm to the crosshead so you would get more lubrication flow at high-speed? Or would additional lubricant cause a problem?

It will be interesting to see what conclusions are made out of this investigation.

Regards,

Randy

Update from the A1 Trust- they’re already manufacturing replacement parts for the damaged valve gear. They’re continuing to investigate the complete cause of the failure.

https://www.a1steam.com/2018/04/25/moti ... ir-update/

I suppose one of the advantages of a new build is quicker turn around for replacement parts, with the patterns for the castings and the drawings all readily available. Are the people who are fabricating the combination lever and union link, Stephenson Engineering at Atherton, the same entity that fabricated the originals? This underscores the importance of a project like the T-1 in the US in preserving manufacturing capability and knowledge.

Latest update from the A1 trust on the repairs to Tornado...
https://www.a1steam.com/2018/07/16/torn ... 16th-july/

Note the reference to the current hot and very dry weather we are currently having in the UK which has restricted/eliminated steam operations on several Preserved lines, and meant diesel assistance is also required on any mainline tours. No sign of any appreciable rainfall on its way for most of the Country in the next few weeks, so worth bearing in mind in case you happen to be visiting the UK in the near future!