Current and Proposed New-Builds

It's David Wardale.

In Jos Koopmans book "The fire burn's more brightly..", he says that there was an error in the design of the Giesl; the basic proportions were incorrect. With this knowledge it might be possible to construct an improved Giesl.

 

That may be true although I would question what would be the point given that a Lempor is a relatively easy to manufacture device and a Giesl isn't!

For those that have not got close up and personnel with a Giesl - apart from the obvious elongated chimney the nozzle itself is a quite a complex casting, true it's not impossible to make its just more work for a limited benefit.


Sent from my iPad using Tapatalk

 

If you want a cheap solution why don't you look at the result of the GWR 6023 "King Edward II". No new chimney, just a new blastcap,
it only needs understanding how fluid dynamics works in chimneys!

 

I agree, and also I prefer the external appearance of the Lempor.

 

Jos, I agree that the 4 nozzle blastpipe on 6023 is a good low cost solution, and I hope it will be successful when fully tested.

Am I correct in thinking that a Lempor has a more efficient diffuser, and therefore overall will produce a lower back pressure?

 

My understanding of that point is as follows.

When the steam enters a superheater element, it starts at the saturation temperature of water at whatever the boiler pressure is. It flows through the superheater elements at a particualar mass-flow rate and emerges at some increased temperature.

Around the outside of the superheater element, within the flue, furnace gases enter the flue at whatever the firebox gas temperature is and flow at a particular mass-flow rate, emerging into the smokebox at reduced temperature.

There is therefore an energy transfer from the flue gases to the steam. That is a dynamic process, not an equilibrium process. The rate of energy transfer, and therefore the temperature at which the steam attains, is dependent on:

- The mass-flow rate of steam through the element
- The mass-flow rate of furnace gases through the flue
- The temperature differential between the two gases.

A well-designed superheater will have its proportions chosen to achieve the optimum rise in temperature when the two mass flows are at the desired level, and crucially when the furnace gases are at their normal working level. Therefore, any deviation from those (especially cool furnace gases) will dramatically reduce the rate of energy transfer to the steam.

The crucial point when starting is that the firebed is at perhaps 800 degrees C.The gases therefore emerge above the fire at that temperature, cool down somewhat on hitting the brick arch, and are relatively cold when entering the flue. At the same time, the mass flow rate of steam through the elements is possibly higher than normal working (you use more steam starting than running). So you get a double whammy for energy transfer: you are trying to heat more steam than the superheater is designed for, and are doing it with a gas flow of furnace gases that is hundreds of degrees colder than it is designed for. Under those conditions there is reduced energy transfer to the steam.

After a period of working, the firebed will heat to about 1400 - 1500 degrees C. At that point, heat transfer is optimised within the superheater. But it takes time. The flue gases can't approach that temperature until you have heated up somewhere between several hundredweight and a ton or so of rock (aka coal) from 800 to 1400 degrees; have heated up several hundredweight of concrete in the brick arch up to its peak temperature; heated up a considerable mass of copper in the firebox; and heated up several hundredweight of steel superheater elements from the steam saturation temperature to the superheated steam temperature. Collectively (and especially the firebed and brick arch) that is a huge thermal mass: it doesn't change temperature instantaneously. It requires the furnace to be working under a high draft for a considerable period of time, and with a few exceptions, most heritage line running don't see that level of sustained draft to raise the firebed temperature. On the mainline, people often seem to assume that a "cold" fire will take twenty or thirty minutes to really heat up, and not many heritage lines have 20 - 30 minutes of sustained effort.

(Obviously repeated shorter efforts will eventually raise the temperature of the firebed and brick arch, and they don't cool down instantaneously either. My feeling on the Bluebell, where an "up" journey requires about 25 minutes of hard running over a 40 minute period is that it takes until near the tunnel on the first trip of the day for the superheater to really start providing a significant benefit. On the second and third trips, the residual elevated temperature of the brick arch (which cools down slowly) means that you achieve that position quicker, and those trips are notable in being easier to fire).

Tom

 

Why "fully tested"? The characteristics of a front-end are fairly predictable, it is almost a linear function. So if it is working properly at a certain regulator/cut-off setting you can be sure that it is working in another setting. As far as I was concerned, after the tests at Didcot I was absolutely sure it would do fine at the SVR and will do so during a 75 mph main line loaded trip.

As for a Lempor, for the last 35 years there has not been published proper test results, I sympathize with Prof. Dr A. Giesl-Gieslingen who noted that already in his book of 1986. The Lempor is longer and could have a wider diffuser, both lower the exit velocity. Given the constant row about the nozzle inclination I have the feeling that the present application is flawed. Since my recent recalculation of the Rugby results inclusive of the 9F Giesl test I am also very sure that the Porta theory is flawed also. It is time to have some facts about the Lempor not buckets full of opinions.
Kind regards
Jos

 

I'm just thinking that the banking engines at, for example, Tebay would have been subject to the same "duty cycle".

 

Seconded!

 

The SNCF 2-12-0 experimental compound was tried with normal superheat plus reheat between HP and LP and gave coal consumption that was very low.
World class actually.
It was then tried without high temperature superheat and used 1% coal more.Then no superheat at all and it used 6% more coal.
Chapelon claimed that a boiler without superheater things can be made 6% more efficient.
The testing from around 1910-20 stating that superheat saves 20% was probably wrong.
The real reason for steam/fuel improvement was that superheating needed much better valving (Big piston valves) to survive hot steam.
The 2-12-0 had steam jacketed HP cylinders as it has been customary at sea for ages.
The final steam locomotive improvement could have been to keep all cylinder surfaces at boiler-temperature all time.
Making a cirkulating pump is no big deal today but would have been a constant source of bother when steam was locomotion.
May I suggest an two (-inside-)- cylinder compound 2-8-2 tank with Boiler water heated cylinders all the time in steam?And based on a spare USATC S160?

 

Trying not to take sides because any improved draughting system is only as good as its installation, lots of empirical work has been done on Lempor to the extent that workable dimensions can be theorized even if the true nature of the function cannot be formulized.
But it seems to me that the golden rule is multi nozzle is better than single and that is the prime mover; Nozzle geometry, kylchap splitters, mixing chambers etc... add cost and even at the point where the optimum geometry for atypical state of working is happened upon, I think Jos Koopmans could argue that as pripherals they are the subject to a diminishing return beyond a simple multi nozzle arrangement.
Any re examining of the draughting arrangement is likely to result in a change of appearance if the draughting was wrong in the first place but Geisl apart that's not likely to be noticeable. For the actual cost of fitting a multinozzle blastcap and a slight redesign of your chimney shape to me its a no brainer to have this done if you are a new build or even just renewing components. I find the lack of faith... disturbing

 

Amongst those of us who are advocates of modern steam power it seems to be an article of faith that Porta was right about everything, and his theories beyond question. Theories are supposed to be tested and proven before being accepted as irrefutable fact, but due to applications of Porta's innovations postdating the steam age there is a pretty small body of evidence. These results mostly on locomotives under the care of very highly skilled and technically minded engineers-"pet" projects, as it were.
It is not the same as the evidence of a class of broadly identical locomotives in daily service on general traffic based at a dozen different running sheds for several years, with diverse duties, crews, variable coal, water, maintenance etc.

 

Probably because the Rio Turbio Railway's the most dramatic example of Porta's stuff working, and continuing to work pretty well for some 30 years after his departure and only really going wrong when the mining company was privatised and asset stripped in the 90s. Horrible coal and water in Patagonia and he managed to add another 260hp to the tiny 2'6" locomotives, letting them haul trains of 1700 tonnes on the regular.

 

Have any of Porta's developments been used with any long-term success in the UK, or elsewhere outside Argentina?

I was just reading over on the Lynton & Barnstaple thread about the trouble they are allegedly having with a "Porta wheel profile" - is that the same Porta? I recall the Ffestiniog's experimentation with a gas producer firebox on one of the Penrhyn Hunslets did not last very long.

 

An earlier posted mentioned the modified Hunslet austerities, these had KylPor exhaust and GPCS mods with underfeed stokers. How long these lasted I'm not sure. Think the underfeed stoker was supposed to allow single manning - didn't work terribly well though and so the system would not have been considered a success no matter how well they pulled/steamed.
Its been commented that Hunslet didn't fully implement the necessaries for the GPCS; - stoker method and firegrate inappropriate and insufficient secondary air resulting in an odd shaped firebed which would become airborne when working very hard.

 

The key word here is "theories". Until backed up by solid evidence out on the track, that's all any novel suggestion remains. In the case of Porta's work, much has been proven in service already. Though comparative maintenance costs aren't an area with which I'm at all well up on, I note that that the Alfred County Railway (which was then attempting fully commercial operations) were certainly sufficiently encouraged to fork out for a second NGG16A.

It seems a good place to mention that although heritage lines aren't about making massive profits, the bottom line is still a consideration. Although coal mightn't be the top item on the expenditure list, it sure as hell ain't a negligible cost either.

In the case of superheating, whilst the system clearly takes time to reach it's optimal operating temperature as @Jamessquared mentions. Once it's done so, it does unquestionably represents a performance improvement. As the Bluebell's saturated stud tend to be older, smaller locos, dare I suggest any savings to be had by not having a superheater to drain the workshop's coffers would be dwarfed by the neccessity of double heading quite a few service trains?

I'd be genuinely interested to see a direct comparison of the pros and cons of superheat v saturated - including maintenance burden - under (max speed 25mph) heritage railway conditions, as was done for mainline, on the LBSCR under Marsh, with the I3 class, though how (and just 'cos there are a few about), say, a Black 5's owners feeling about stripping the superheater out might be another matter entirely .... even it were an easy task, which you can be sure it wouldn't be.

If superheater maintenance overheads were sufficiently excessive to dictate a return to saturated steam, I'd have thought the pretty savvy management on the Ffesterbahn would've ripped superheaters out some time ago and AFAIK, new Fairlie "James Spooner" will be including the feature. The warming up issue is a bit of a red herring anyway .... unless a working timetable calls for inordinately long layovers, regaining temperature will be considerably faster than for the first run of the day. Either it's no more a problem than the entire boiler coming to working temperature, or I'm missing something.

When it comes to ejectors, the only observation I feel qualified to make is that the Giesl design, incorporating intricate moving components, is by definition going to be a more complex setup than Lempor pattern. There's simply more to go wrong .... and naturally enough, Sod's Law applies in full measure.

"Lyn's" 'Porta' wheel profile is an interesting one (I've just gone back through the 'Lynformation' bulletins, but can't find a reference to driving wheel profile). AIUI, the intention was to gain extra traction. Giving the idea a shot on a line with so much 1:50 - especially for a four-coupled loco - sounds reasonable enough to me. Is there some fuller description of what exactly did or didn't happen to convince the L&B to fork out for reprofiling ahead of the loco's entry into full service? I ask, as it seems that either (a) Porta's theory was wide of the mark, or (b) the profile as applied to 'Lyn' somehow didn't actually comply with the theory. In either event, some clarification would be useful.

 

The intricate moving components on a Giesl are no such thing, essentially the only moving parts are two plates which sit either side of the nozzles that allow the area of the main blast nozzles to be reduced in area. Once adjusted (each piece has two slots and are located on studs with nuts) to meet with the blast required they can be left alone - well I say left alone they soon become immovable in the harsh conditions that exist in a smoke box. The complexity of such a ejector is the shape of the thing with the differing angles of the main nozzles coupled with the steam pipes to the blower - the blower has two small nozzles between each of the main blast nozzles.


Sent from my iPad using Tapatalk

 

I remember seeing some of Porta's suggestions for Tornado, and they were just ideas. They would have needed a great deal of development to get them working, if they worked at all. Similarly, I've seen a drawing of Porta's proposed "Lemprex" exhaust. It was very complex, including a reservoir to smooth out the steam flow. I can understand what he was thinking, but (in my opinion) it would have required a great deal of work to get it working, let alone provide any useful advantage over existing types.

On the other hand, and as described in Wardale's book "Red Devil", the Lempor had "de Laval" convergent-devergent nozzles. These were intended to operate in supersonic mode during the initial exhaust release, and subsonic (low back pressure) mode in between "the chuffs".
I'd never paid much attention to these nozzles before this, but when I looked into it I thought it was quite a clever idea.

I'm not surprised that a mathematical analysis has never been published, it would be a nightmare to try to analyse such a non-linear system mathematically. Don't let me discourage anyone from trying though.

I understand the 4-8-4 Niagaras on the NYC also had de Laval nozzles, and there may have been others.

 

The "Red Devil" was designed with a very high blast pressure, above the sound barrier. Using DeLaval orifices was a must since standard orifices would have
released the steam at sound velocity and not beyond. However such a high blast pressure is avoided like the plague in more standard locomotives.
Kind regards
Jos

 

They're tolerance critical and they're adjustable. In my books, that makes 'em more vulnerable than certain other flavours which are merely tolerance critical, but fixed. As you say .... they do their job in extremely harsh conditions, so I stand by my comment.