C&O 2-6+6-6 reached 46 mph with 14075 tons?

[timz]

Dreyfusshudson

The resistance equation I use for the loaded coal train is R=1.7+0.03V+0.000187V2 lbs/long ton

So you figure with no engine pulling it the train will reach 65.7 mph on 0.2% down, just by gravity-- and 102.2 mph on 0.3% down.

Dreyfusshudson

about the drawbar pull being different going up and downhill

You figure the engine produces 7074 IHP and 6371 DBHP on the level at 40.42 mph. If that's true, then when it's producing 7074 IHP at 40.42 mph on a 2% upgrade, its DBHP will be around 4200; with the same IHP at the same speed on a 2% downgrade its DBHP will be around 8500.

BigJim

the "booster" feature (optional) [on the N&W 2-8+8-2] was being utilized and its purpose of which was to add heat from the live steam to the HP's exhaust, not to raise receiver pressure

Far as we can tell from Jeffries' book, receiver pressure in booster mode was the same as receiver pressure in simple.

[BigJim]

Dreyfusshudson

What is clear is that the high TEs they could deliver were achieved when admitting live steam to the LP cylinders, so they were operating in ‘semi’ simple mode, so in that sense they were not really true Compounds.

This is only applies when the "booster" feature (optional) was being utilized and its purpose of which was to add heat from the live steam to the HP's exhaust, not to raise receiver pressure per-se. I would suggest that you find a copy of "N&W Giant of Steam" by Col. Lewis Ingles Jeffries.

[Dreyfusshudson]

To Big Jim: The only source I have on the N&W Ys is Warden’s book, which is short on data, so there’s not a lot I can analyse. The Ys were clearly very trusted machines, and did a super job hauling huge amounts of coal right to the end. The intriguing thing about the Ys is their adhesion ratio; If developing 152600lbs TE in simple mode, it comes out at 3.6; at 170000lbs, it is 3.2, but extra ballasting was added to the front raising this to 3.4. The A has 3.4, the J has 3.6. Everyone else thought you needed 4+. So what’s going on? Did the N&W know that 4+ wasn’t a good rule? Or did they have some knowhow that allowed them to bend the rule? This question was raised here in 2008, but not answered. Now 170000lbs far exceeds the Big Boy’s 133500lbs, so yes, they would lug more coal than anything else in the country, so no wonder they were revered.

What is clear is that the high TEs they could deliver were achieved when admitting live steam to the LP cylinders, so they were operating in ‘semi’ simple mode, so in that sense they were not really true Compounds. The question to my mind is what would have happened if you had used a simple articulated with 170000lbs TE. The more modern eight axle simples had TEs which maxed out at 140000lbs, presumably because of adhesion ratio considerations. It would not have been difficult to create one with 170000lbs TE, but would it have slipped? Or does the Compounding give some other kind of operational advantage?

If you have any information on what the cut off ratios were between HP and LP cylinders at various speeds, I could have a go at modelling the engine behaviour, to see what kind of efficiency benefits were possible. Quite apart from the problem with losses between the cylinders and wheelrims with French Compounds, there was a secondary problem with condensation on the LP cylinder walls, which was only eliminated at much higher inlet steam temperatures than needed for simples to avoid this problem, at least 700 deg F. At less than this there will (according to modern theory coupled to Altoona data) be very significant condensation in the LP cylinders, which given the Ys massive size will be large enough to balance the benefit from Compounding. So, in the absence of any data, I would wager that the Ys were no more efficient than any similar simple when operating in Compound mode.

Not that I think this matters that much; operational success does not depend much on the finer points of efficiency, as far as I can tell.

[Dreyfusshudson]

 

The resistance equation I use for the loaded coal train is R=1.7+0.03V+0.000187V² lbs/long ton; where V is the average speed over the previous 0.05 mile segment (not shown), not the speed at that point. I’m not saying this is correct, only that it fits the data to hand reasonably well, and it can’t be far wrong.

 

I don’t think I understand your point about the drawbar pull being different going up and downhill. It’s different because the Cylinder power (IHP) is different, because the engine is working faster going downhill, and at constant cut off this means cylinder power will be greater. Am I missing what you are saying?

 

[Dreyfusshudson]

 

Further to the earlier query, I did some simulations of a C&O 614 test run from Handley to Hinton, the steeper stretch on the trip from Huntington. As Wardale writes, even with 5000 tons of coal 614 was not especially taxed. I ‘drove’ it following two rules: evaporation rate no more than 60000lbs/hr and speed not to exceed 40mph.  Once 40 mph is reached on the level grades out of Handley, 614 bounds along quite easily at this speed in ca 35% cut off, ca 3000IHP. The steepest grades are about 0.4%, mostly between mp 417 and 399, and speed falls to 16mph in 55% cut off on this stretch. If the engine was not in the best of condition, considerably higher values will have been needed to sustain this level of power. I am not allowing for use of the trailing truck booster in this, and the fact that 614 was down to 2mph at one point after this broke underlines that it certainly wasn’t in the best of health- it really should have had no problem.  I covered the 73 miles to Huntington in just under three hours, average ca 25mph; Wardale says the whole trip average was about 30 mph, but maintain 40 mph on the easier stretch from Huntington to Handley will not have been a problem. I have evaporation as 6.4 lb water/ lb coal, 20% too high, which is some kind of measure of the condition of the loco I think. In good nick, I estimate it could have hauled 50% more, minimum speed about 11mph.

 

On the George Washington, working to a speed limit of 60 mph, (probably quite a bit higher than the rule book!), a J3a would be barely out of 1^st gear with 900 tons on this stretch, about 58mph start to stop, 32000lbs/hr evaporation with feedwater heater, no problem with exhaust pressure. Even climbing the 1 in 88 westbound to Alleghany, it would have cleared the summit in the low 30s without working at a steam rate which would cause excessive backpressure. On the Cardinal runs, with 1500 tons of load I suspect the time loss will have been due to the inability to accelerate from rest at the same rate as a pair of diesels, much lower TE.  

 

So, this was not an unfair or overly onerous set of tests, but the fact that the locomotive was not in the best condition made it a lot harder work than it might have been. The restricted exhaust opening will not really have been an issue on the C&O.

 

[BigJim]

Dreyfusshudson

As for Compounding, I have spent the last 5 years researching French Compounds,

You should have been studying the N&W Y6 class instead. But, then, you don't like compounding, so what's the use?

[Dreyfusshudson]

Thanks, Paul for raising these points.

I haven’t really studied the test results on C&O 614 in Wardale’s book, because the data is a bit sketchy, and the locomotive was, as you observe, in rundown condition.  However, looking at it again, it raises some interesting points, most particularly his strictures about how restricted the exhaust of 614 was.  Elsewhere he states Porta’s view that US exhausts ‘could not have been worse’; I researched this point fully, and concluded that whilst under extreme tests conditions, such as the 6600IHP tests of the Niagara and T1, back pressures were truly abysmal, if you looked at the steam rates used in practice they were acceptable, this a view gleaned from analyses of NYC, MILW, UP and ATSF running. I believe it to be generally true. It may have been possible to gain a fractional increase in efficiency on some designs from even lower back pressure, but nothing of a game changing nature. However, 614 is a possible exception, and this is maybe what coloured Porta’s view. Wardale quotes back pressures in the range 14 to 28 psi, and with a free nozzle area of the size and discharge coefficient he quotes, this would correspond to blastpipe flows of 52000 lbs/hr and 74000lbs/hr at 70 mph, approximately 3400 and 4100IHP respectively, not a lot from so large a grate. As tested, 614 was fitted with live steam injectors, so this will have been the cylinder rate. Since the loco was in run down condition, the evaporation needed may have been significantly higher than this. He makes a stab at the maximum IHP as being 3600, but this will have been at lower speed, and at 30 mph I make the power at 28 psi back pressure about 3800, in line with this. Thinking about it, I suspect that even with 11 heavyweight coaches on the George Washington climbing eastwards towards Hinton, a J3 would not have needed much more than 3000ihp given the limited speed possible, and with 10% recycle thanks to a feedwater heater, this would have meant that under its normal hard working conditions, the back pressure was ok, in other words, there was nothing wrong with the design for its intended purpose. It may well be however that what was good enough in terms of steam rate for the George Washington was not good enough for 5000 tons of coal, so the back pressure weakness was exposed.

I will need to ‘build’ a J3a and the line from Huntington to Hinton to test these ideas out fully. I will report back in due time on what emerges.

On the boiler efficiency point, my model gives 6.4lbs water/lb coal at 52000lbs/hr, 5.4 lbs water lbs coal at 74000lbs/hr, so the quoted 5.4 lbs water/ lb coal is in the right ball park, bearing in mind as noted above that the actual evaporation required to achieve these cylinder rates may have been higher. Wardale says the coal was well-nigh perfect, so you can’t blame this, though he does say a different mechanical stoker to normal was used, and if you look at his pictures of the exhaust, you can only conclude that combustion conditions were far from the text book standard my boiler model assumes. As a generality, in UK road tests were able to reproduce test plant results when the crews’ driving and firing were very carefully supervised, but tests where the crews were doing what comes naturally generally came out worse. So, my conclusion so far is that 5.4 lbs water/ lb coal was not a bad return, given that, according to Wardale, Ross Rowland was not driving with economy in mind!

On the Big Boy point, I think the SSC will indeed have been around 15 at 6800 IHP when working at 50 mph in the 40-50 mph range. You have spotted correctly that I divided lbs/hr by 1000 to get tons not 2000! Apologies. Hopefully the Table below comparing performance at 50mph clarifies- these numbers are all ‘of the order of’.

	Cut-off	Steam	IHP	SSC	Lb water	Coal				lb water/	Drawbar Eff
	%	lb/hr		lbs/ihp-hr	sqft/ hr	lb/hr	t/hr	lb/edhp-hr	lb /ihp-hr	lb coal
Big Boy	35	101860	6922	14.72	611	24580	12.3	4.06	3.55	4.14	5.38%	Utah coal
Big Boy	35	101860	6922	14.72	611	13627	6.8	2.25	1.97	7.47	8.30%	App coal
H8	43	102216	6902	14.81	681	16105	8.1	2.67	2.33	6.35	6.96%	App coal

 

I don’t think smaller driving wheels would have helped, in that they would have increased TE, and therefore to maintain your adhesion factor you would have gone to smaller cylinders, so, since limiting tonnage up the ruling grade is all about TE, no difference really, unless you want to argue that larger drivers are inherently slippier.

As for Compounding, I have spent the last 5 years researching French Compounds, having discovered that in their normal working regime they were less efficient than equivalent simples; this is very clearly what all their data says. Only at impracticably high rates achieved on tests do benefits come through. Chapelon never mentions this. The reason is that, according to the test plant at Vitry, the losses between the cylinders and wheel rims were far higher than on a simple, as measured at Rugby. The Rugby data conform to accepted models, the French data don’t. So, despite better cylinder efficiency, less power turns up at the drawbar. I have batted this around with experts over here, including those who knew Chapelon and are sympathetic to his views, looking for an explanation, without ever coming to ground, so it’s all rather unsatisfactory. Was it a specific problem with the French designs? Or is it inherent in Compounds? And why was compounding a success in marine applications? Whatever the answer, I’m not a fan of the complexity of Compounding.

[timz]

The assumed train-resistance-vs-speed in dreyfusshudson's table isn't a smooth curve-- it increases 46 lb from 33.11 mph to 33.55 mph and then 271 lb from 33.55 to 33.69 mph, and it's higher at 40.42 mph than at 40.79 mph. If we draw a parabola thru his 33.11-mph, 40.42-mph and 47.10-mph resistances then A turns out to be 24312 lb, B is 235.65386 and C is 4.013680. Based on that resistance, the train's balance speed with no engine pulling it is 64.6 mph on 0.2% downgrade and 96.5 mph on 0.3%.

The 2-6+6-6's drawbar pull on 0.333% down is maybe 9000 lb more than it is on 0.2% up-- does your table know that?

[Paul Milenkovic]

Dreyfusshudson

Why then did a Big Boy not out pull an H8 at speed? The reason is clear in the UP data- abominable boiler efficiency! It was raising only 4-4.5 lbs steam/ lb coal fired even at very modest water rates around 80000lbs/hr, this from a 150 sqft grate, 10% larger than the H8. The calorific value of the coal is not stated, so the true thermal efficiency of the boiler is not calculable, but it must have been very poor. Now in my view this has absolutely nothing to do with the design, everything to do with the quality of coal used. Boiler efficiency is determined by two things, the heat transfer efficiency (% heat produced finishing up raising steam) and the % of coal fired which is lost unburned. Heat transfer efficiency in boilers is always good, and with its exceptionally long tubes, Big Boy heat transfer will have been excellent. The problem lies entirely in the unburned coal losses. Now Appalachian and UK hard coals of ca 13500 Bthu /lb show very similar unburned losses (lbs/sqft grate/hr), somewhat more for mechanical stoking than hand firing. That is, the burning characteristics of these coals under high draught are similar, likewise French coal. I have attempted to back calculate what unburned coal losses were on the Big Boy, assuming the coal used was 11800Bthu/lb. At 600lbs/sqft/hr the answer is about 75lbs/sqft/hr, or nearly 6 US tons/hr. With Appalachian coal you would lose about 25 lbs/sqft/hr at this rate. If the UP coal had even lower Calorific value than assumed, the computed unburned losses would not be quite so high, but still way above those achieved with good quality coal (e.g. 65lbs/sqft/hr at 10500Bthu/lb). So whatever was fed to their firebox was barely worthy of the name ‘coal’.

The 6000DHP at 45 mph quoted by Kratville suggests maximum Big Boy cylinder power was about 6800 IHP. I do not have an estimate of their superheat, but I suspect it was not particularly good, on account of the relatively low combustion rates/sqft achieved, and the over long flues. If anyone has any data, I would be pleased to receive it. With a superheat of 680 deg F I estimate the Big Boy would consume about 104000lbs/hr steam in 35% cut off at 45 mph to deliver 6800IHP, say 94000lbs/hr water or 665lbs/sqft/hr. Even at this relatively modest rate, coal consumption would be about 23 short tons an hour, compared to a tender capacity of 28 tons. In other words, there was a very practical limit to boiler output, hence maximum IHP and DHP. An H8 with Appalachian coal would burn about 14 short tons an hour at the same rate.

Oh yeah, what about running a passenger 4-8-4 locomotive (we are not allowed to call it a "Northern" on the C&O) on that route dragging coal?

Think that is crazy?  Ross Rowland's ACE project did just that as a publicity stunt to call attention to their proposed ACE 3000, and this effort backfired as the fuel consumption was horrendous, at least according to Wardale's "Red Devil" book.

Rowland's locomotive the 614 in those "tests" was averaging 5.4 lbs evaporation per pound on high-btu Appalachian coal, giving 45 percent boiler efficiency, very comparable results to the 4.5 lbs/lb on the proportionately lower BTU coal you mention. 

Here is my question.  If the Big Boy was producing 6800 indicated ("cylinder") HP on 104,000 lbs/hr evaporation, that works out to about 15 lbs/hp-hr, a respectable figure for Super Power generation steam.  And if they achieved 4.5 lbs evaporation per pound of coal, that is 3.3 lbs of coal/hp-hr?  Which would be equivalent to just under 3 lbs of coal/hp-hr with a proper higher-BTU coal?  Again, a respectable figure of Super Power steam?

My question then is 3.3 lbs/hp-hr at 6800 (indicated) HP should come out to 11.2 short tons/hour, not the 23 tons you quote?

Now there are a lot of person-year to dog-year conversions on Wardale's data because he insists on using metric units (you English invented those units, why can't you use them?).  But it seems that the 614 on those trials was consuming 6-10 lbs of (high BTU Appalachian) coal per HP hour.

Besides leaking water into the fire space and consuming heat at 1000 btu/lb of water and having a high back pressure exhaust system that saps cylinder efficieny, the coal consumption of the 614 suggests it was worked quite "hard", which is very heroic when people talk about things but made the trials a public relations disaster when all of that condensed vapor, smoke, and cinders came flying out of the stack to remind people why Diesels replaced steam in the first place.

It has already been said that maybe the H-8 was misapplied to coal drags (let alone the 614!).  Do you suppose that something with smaller drivers and perhaps compound expansion could have produced the required tractive effort with much less evaporated steam and in turn less combustion loading on the grate, giving coal consumption closer to 2 lbs/indicated hp-hr?

[Dreyfusshudson]

I was wondering if the following table might help explain what is going on. It shows an analysis of the motion of the 14000 ton train over an undulating line, hauled by an H8 working in constant cut off (ca. 50%), over a series of 0.05 mile stretches. The analysis takes no account of the fact that when a gradient transient occurs, the gradient resistance will change only slowly, because some of the train is still on the previous gradient; it is a gross simplification, but not an unhelpful one. It likewise does not take account of curvature.

Route	gradient	Speed	IHP	DHP	Vehicle Res	Gravity Res	acceleration
Miles		Mph			HP	HP	HP
69.5	500	33.69	6288	5711	3309	5067	-2665
69.55	500	33.57	6273	5698	3292	5049	-2643
69.6	500	33.46	6257	5685	3275	5032	-2622
69.65	500	33.34	6242	5672	3258	5014	-2600
69.7	500	33.23	6227	5659	3241	4997	-2579
69.75	500	33.11	6212	5646	3224	4980	-2558
69.8	100000000	33.22	6223	5657	3224	0	2434
69.85	100000000	33.33	6238	5671	3240	0	2431
69.9	100000000	33.44	6252	5683	3255	0	2427
69.95	100000000	33.55	6266	5695	3271	0	2424
75	100000000	40.42	7074	6371	4354	0	2017
75.05	-300	40.79	7116	6410	4390	-10160	12179
75.1	-300	41.15	7156	6442	4452	-10251	12241
75.15	-300	41.51	7193	6471	4514	-10342	12298
75.2	-300	41.86	7228	6500	4576	-10431	12355
75.25	-300	42.21	7263	6527	4637	-10519	12409
76	-300	47.01	7694	6855	5523	-11727	13059
76.05	100000000	47.04	7686	6843	5555	0	1288
76.1	100000000	47.07	7687	6843	5560	0	1283
76.15	100000000	47.10	7689	6845	5566	0	1279

 

At Distance 69.5 (Column 1) the train is on a 1 in 500 upgrade (Column 2). Speed is 33.69 mph (Column 3), and the H8 is developing 6288 HP in the cylinders. (Column 4). The train is decelerating, (Column 3).  Cylinder power in a given cut off decreases as speed decreases and so as the train decelerates, cylinder power falls, and with it the DHP (Column 5). As the train decelerates, the power needed to roll the train along (Vehicle Res) decreases (Column 6). A massive amount of power is being used to overcome gravity on the train (Column 7), the net being that the power available for acceleration is negative, (Column 8) hence the deceleration. If the train continued to decelerate, the acceleration power would eventually fall to zero, and a balancing speed on that gradient would be achieved. In some circumstances the end result might be a stall.

When level track is reached at 69.8, the balance changes dramatically. No power is needed against gravity, so there is now power available for acceleration. As the train accelerates, the cylinder power rises, and with it DHP. The change in gradient causes no dislocation in DHP. (Though the dynamometer car might show odd behaviour at this point for reasons given earlier)

The power available for acceleration is low relative to the weight of the train, and after 5 miles of near level track at distance 75, the train has struggled up to but 40.4 mph; the cylinder power and hence DHP have risen accordingly. At this point, suppose the track begins to fall at 1 in 300. The power needed against gravity becomes hugely negative, and so the power available for acceleration is huge, and a mile further on at distance 76, the speed is already 47 mph. Back to level track at this point, the train reverts to a very slow acceleration.

Now obviously, with a train over a mile long, the transition between gradient power and acceleration power would be much smoother than shown here, and the motion of the train different to that shown. The train would not accelerate as quickly at the top of the grade, but it would also not accelerate as quickly when a downgrade is reached. The reason I am relatively relaxed about this is that I hold a view that, for all sorts of reasons, it is only meaningful to talk about power developed over segments of several miles, and over several miles, things tend to balance themselves out. As discussed earlier, if you are looking for a spike on a dynamometer trace to claim a power record, all you need to do is to make a significant advancement of cut off; whether the power developed can be sustained is then a matter for the boiler to decide, but the record book is written.

[Dreyfusshudson]

 

Thanks tmz.

 

On the coal dock, I've now been sent the track plans, and firstly it is clear that MP 0 is where the tracks diverge, not at the switches, so my MP distances are ca 0.5 miles out, and secondly there was indeed a big coal dock at Mp 34.6! This is about 1.3 miles i.e. ca. 1 train length north of the road bridge, and looking at Google Earth carefully, you can see there is just about room for two tracks to the west of the now single track main.

 

As to the coal dock at Robbins, the line was built from Limeville to Waverly in 1917 to connect with the N&W there, at ca. mp 29. The C&O then built its own line north to Columbus from that point 10 years later. Since there will have been a need for a coal dock in 1917, maybe Robbins was the original one? But why would the C&O build a new one with the end of steam near? This link above gives some detail on Waverly, mostly N&W, and shows the same photo I posted, siting the dock at Gravel Washer Road near Waverly, but this is several miles from the tracks. http://www.waverlyinfo.net/n---w---c---o-in-pike-co.-ohio.html Someone out there must know the detail, but it doesn't add too much to the overall discussion, I think.

 

On rolling resistance, gravity is pretty potent stuff. Over here, there is about 10 miles of 1.25% going north from London to Glasgow in the Scottish lowlands, pretty straight. In 1961, the late running Midday train to London cleared the southbound summit at 60 mph, and the driver shut off steam and evidently ‘forgot’ to apply the brakes, with the result that the bottom of the bank 10 miles further on was passed at 105mph. (Well documented; The speed limit was 75mph at the time). 10 more miles of the same would have taken them well past the world speed record for steam; the easy way to beat this is find a long stretch of steep downhill and let gravity do the work. My equations suggest that the 14000 ton train would take 30 miles to accelerate from 40 to 54 mph without steam on 0.2%. My guess is the balancing speed would be in the high 50s, not sure about your 65mph. With 30 miles of 0.3% the equations say you would be up to 75 mph, of 0.4% to 93mph. Money to be made selling brake blocks to the C&O for trains heading east from Alleghany!

 

As I wrote earlier, the resistance equation I use gives a rolling resistance requiring 3500DHP to maintain 35 mph on the level; it requires 5200DHP at 45 mph. I haven’t checked Davis’s estimate, but if as you say he requires double this, then that says that you would need 7000DHP at 35 mph, and 10400DHP at 45 mph, the latter way beyond H8 capability, the former requiring 7500 IHP which is right at the limit of what an H8 could achieve, nothing left for any acceleration. This is just not credible, given what the H8s achieved. I would bet, though haven’t checked that you simply could not lift the loads up to Alleghany eastbound at this resistance. I feel we are missing each other on this point.

 

My estimate of 33 mph is based on the (incorrect) assumption that the whole train was on a 0.2% gradient for the whole stretch; as I wrote my purpose was not to calculate the actual dhp at any point (which I believe is impossible), but to establish the level of cylinder power needed that had a hope of producing a peak value of 7498DHP.   Are you attempting a more sophisticated calculation which is taking into account the actual gradient pull the train experiences over a specific stretch of track, being on several different grades at once, and includes the actual curvature? This would be far beyond what I am attempting, and could explain the differences. I am happy to forward you my spreadsheets which would allow you to ‘drive’ the train and see the basis of my calculations.

 

[timz]

I'm using the Dave Cramer copies of the Nov 1960 C&O charts-- more info from them below.

Yeah, sure sounds like there was a coal dock  near Coal Dock Rd (39.03435N 82.91636W). But the 1960 chart says "coaling station" at MP 34.6, and if you go to historicaerials.com and search for 39.142N 82.879W and choose the 1960 aerial you'll see what I assume was this coal dock. Far as I can tell from the chart the bridge at 39.1473N 82.86545W was about MP 35.35, and this structure is about 0.75 mile south of that bridge. I'm not enough of a C&O fan to know when or if this coal dock replaced the one near Robbins.

So we'll forget about the 2-6+6-6 stopping for coal at MP 34.6.

So you think it can pass the top of the grade MP 54.9 at 33 mph. (Chart shows the top of the grade at the midpoint of the curve, maybe 400 ft north of the bridge over the N&W main at 39.37116N 82.97439W.) Which means you think the train's resistance is less than half of Davis-formula resistance. If resistance were that low, you could place the train on a long 0.2% grade and it would eventually accelerate to 65 mph-- with no engine pulling it. Just gravity. Think that's true?

Or are you assuming the 2-6+6-6 is more powerful than the drawbar pulls I chose above? What is your assumed drawbar-pull-vs-speed curve, and your assumed train resistance?

The chart shows Milepost 0 at about 38.7300N 82.8829W, between the two curves. The bridge at 39.02672N 82.91614W south of Coal Dock Rd scales at MP 22.30. The bottom of the long climb, MP 42.2, is about 39.2288N 82.8529W, midway between the two slight curves to the right. Elevation there is 603.1 ft; elevation at the summit at MP 54.9 is 691.9. Total curvature in the 12.7 miles is about 322 degrees, so the average compensated grade is 0.15%.

Then at MP 58.71 elevation is 688.4; the curvature in that 3.8 miles is enough to make the compensated grade zero. The bridge at 39.44077N 82.9750W is MP 59.7; the overpass at 39.45915N 82.97124W is mp 61.0. The track then curves left to the undergrade bridge at 39.4649N 82.9709W, MP 61.43, elevation 678.5.

[Dreyfusshudson]

With respect to timz’s post, I think a couple of points need clarifying here. I inferred from the earlier correspondence that MP 0 on this line is at or near the junction of the C&O lines to Cincinnati and Columbus at Limeville, where the new line, originally to Waverly started. In the present day, the tracks to Columbus separate from those to Cincinnati about 0.5 mile south of where they actually peel apart. All that follows is based on the assumption that MP 0 is at this separation. My arbitrary point in Russell for the start of the northbound test runs is at the west end of a large building in the yards there, 15.75 miles south of Limeville Junction. It is said that on some runs, the test was terminated at Robbins, 39 miles to the north. At 39.1 miles north of my arbitrary start point, the line crosses ‘Coal Dock Road’, ca MP 23.4, so even though Google has not heard of Robbins, I’m pretty certain that’s where it was. This picture shows the coal dock there, looks like it straddled the main. But was it the north or southbound one? (Google Earth shows there are still a couple of loaded hoppers in the sidings there, maybe for the off chance that an H8 turns up?)

http://cohs.org/repository/archives/thumb/cohs/cohs-5956.jpg

The test trains running all the way to Columbus consumed 22-25 tons of coal; the tender capacity was 25 tons, so a top up point would be needed for all but the most daring of crews. So, in answer to your question ‘Did the train stop at the Coal dock?’ I would say ‘I don’t know, but it seems very likely’. The question then is, was there another coal dock?

I’m not clear about the location of the coal dock at MP 34.6 you refer to. On my assumed MP scale, this is about 0.8 miles north of the bridge over Route 355, near Waverly. It is in open countryside, right next to the river, with no hint of a coaling facility remaining.

Now if the northbound coaling dock was in Robbins, the train would progress much as I wrote on the relevant stretch, because the sharp dip down to the river four miles further on would quickly get the train into the mid 40s. If the northbound coal dock was ca 12 miles further on as you suggest, then that’s a very unfortunate location, because the line rises 80’ in the next twenty miles, which is a big ask if you are hauling 14000 tons from a standing start! The train could not get up to the mid 40s on the level stretch on to MP 42.5, and realistically, the speed could not reach this value until the dip near MP 59, and the ‘cruising at 6600-6900 DHP in the mid forties’ scenario could not then commence until after the rise near mp 61. If you can clarify this point, it would be much appreciated.

On your point about a 2-8-2 being able to haul 14000tons at 35 mph, well, on the level the rolling resistance of the train at the resistance I deduced requires about 3500 DHP. Altoona tests show an ancient L1s 2-8-2s could not deliver this from its 70 sqft grate, but the slightly more modern I1s 2-10-0 is a bit more efficient and might get close at its boiler limit of 58000lbs water/hr. However, the big point is that this leaves you absolutely nothing to deal with any gradients- and with loads as large as this old man gravity extracts fearsome, unavoidable taxes. On the simulation I did where I quote a minimum speed of 33 mph I have the gradient as a very gentle 0.2% at this point; but the H8 is not capable of sustaining 33 mph on even this grade with a delivered DHP of 5800; the train is decelerating still. To sustain 33 mph there would require ca 8500 DHP, and even in full gear the H8 cannot develop this at this speed.

[Dreyfusshudson]

Thanks for to all for the supportive comments.

Mention of the Big Boys raises the question as to why these machines, which had a larger grate and higher tractive effort were shown by Kratville to deliver a mere (!) 6000DHP at 45-50 mph. (Not sure if this has been dissected previously, if so I apologise). Using a similar process to that for the H8, I have ‘driven’ 4016 from Ogden up to Wahsatch to simulate its test run. With 3883 tons, 4016 was working hard, but not quite flat out on the steeper stretches, and indeed the load limit for the Big Boys was later raised to 4200 tons. Maximum DHP was just over 6000 on the 4016 run, at 41 mph. If you want to use dhp as your comparative measure of power, then at a given boiler rate it reaches a maximum between 40 and 50 mph, so that’s the speed range to look at (for all locomotives I know of, it is maximum boiler output rather than tractive effort which limits power developed at speed).

If we are comparing haulage capacity at low speed, where the boiler limit is not in play, all that needs to be considered is tractive effort (TE) which tells you, pretty accurately, what the maximum possible load on the ruling grade is, adhesion permitting. At a steady 14 mph on the 1 in 88, 4016 was delivering 4300 equivalent drawbar horsepower, about 3800 actual dhp, requiring a tractive force of about 120000lbs in its cylinders, about 90% of its TE of 135000lbs, and more than an H8 with 110000lbs TE could deliver. To see what this might mean, I ‘drove’ a Big Boy from Hinton to Alleghany on the C&O, and found it could lift about 7000 tons, magnificent, but not enough to eliminate a helper district, so no more economically attractive than an H8.

Why then did a Big Boy not out pull an H8 at speed? The reason is clear in the UP data- abominable boiler efficiency! It was raising only 4-4.5 lbs steam/ lb coal fired even at very modest water rates around 80000lbs/hr, this from a 150 sqft grate, 10% larger than the H8. The calorific value of the coal is not stated, so the true thermal efficiency of the boiler is not calculable, but it must have been very poor. Now in my view this has absolutely nothing to do with the design, everything to do with the quality of coal used. Boiler efficiency is determined by two things, the heat transfer efficiency (% heat produced finishing up raising steam) and the % of coal fired which is lost unburned. Heat transfer efficiency in boilers is always good, and with its exceptionally long tubes, Big Boy heat transfer will have been excellent. The problem lies entirely in the unburned coal losses. Now Appalachian and UK hard coals of ca 13500 Bthu /lb show very similar unburned losses (lbs/sqft grate/hr), somewhat more for mechanical stoking than hand firing. That is, the burning characteristics of these coals under high draught are similar, likewise French coal. I have attempted to back calculate what unburned coal losses were on the Big Boy, assuming the coal used was 11800Bthu/lb. At 600lbs/sqft/hr the answer is about 75lbs/sqft/hr, or nearly 6 US tons/hr. With Appalachian coal you would lose about 25 lbs/sqft/hr at this rate. If the UP coal had even lower Calorific value than assumed, the computed unburned losses would not be quite so high, but still way above those achieved with good quality coal (e.g. 65lbs/sqft/hr at 10500Bthu/lb). So whatever was fed to their firebox was barely worthy of the name ‘coal’.

The 6000DHP at 45 mph quoted by Kratville suggests maximum Big Boy cylinder power was about 6800 IHP. I do not have an estimate of their superheat, but I suspect it was not particularly good, on account of the relatively low combustion rates/sqft achieved, and the over long flues. If anyone has any data, I would be pleased to receive it. With a superheat of 680 deg F I estimate the Big Boy would consume about 104000lbs/hr steam in 35% cut off at 45 mph to deliver 6800IHP, say 94000lbs/hr water or 665lbs/sqft/hr. Even at this relatively modest rate, coal consumption would be about 23 short tons an hour, compared to a tender capacity of 28 tons. In other words, there was a very practical limit to boiler output, hence maximum IHP and DHP. An H8 with Appalachian coal would burn about 14 short tons an hour at the same rate.

What then if we take the Big Boy back to Russell, and head north with 14000 tons, and fire it with Appalachian coal? Well, it could in round terms produce 10% more steam than an H8, and hence roughly 10% more power. So on a short term burst at 900lbs/sqft/hr, as 1608, it would develop approaching 9000IHP in 50% cut off at 45 mph, and over 8000DHP, thereby accelerating to over 50mph. 

One has to ask at this stage however, ‘What use would this have been, given the extra coal burned?’ Not a lot, I suspect. The underlying point is that US freight designs were generally designed to haul maximum possible, or maximum desired tonnage up the ruling grade, with speeds of 10-15 mph being acceptable in this context. This requires highest possible TE. The boiler pressure + cylinder capacity required to deliver this means that at speed, the cylinders could produce phenomenal amounts of power, as much as the boiler will permit, but with maximum operating speeds in the 40-50 mph range the tonnage/ gradient combinations that required this were, I suspect, relatively rare. Massive coal trains on flat divisions is one opportunity. With merchandise freights of, say 5000 tons, 7500 IHP could only be developed where there were long stretches of 0.3% gradient or greater. That is, I suspect, the overriding priority for freight designs was low speed haulage capacity, and there was for the most part more than enough power available from the boiler to haul normal tonnages at acceptable speeds elsewhere.

If 4014 returns to the mainline, it will be freed from the restraint of its customary coal, for it will burn oil. I am not sure this change will allow it break any records however. I know very little about oil firing, but I do know that the ATSF took the boiler limit of its oil fired Baldwin passenger locomotives as no more than 600lbs/sqft/hr- (similar to the practicable operating limit of coal fired types), which may indicate that it is not possible to fire oil at higher rates.

More generally the Big Boy vs H8 debate that surfaces from time to time is actually sterile. Both were excellent machines that did the jobs they were designed to do in their operating environment well. The Big Boy has greater low speed haulage capacity than the H8, but this would not have brought benefit in the H8’s operating territory. If fired with decent coal, it would also have been able to develop more power than an H8 at speed, but even the 6700DHP of the H8 could only be applied in limited circumstances.

At least, that’s the way it looks from over here.

[timz]

Dreyfusshudson

with 14075 tons, it bounces along quite happily at 43-46 mph developing, as reported 6600- 6900DHP, until a relatively steep uphill stretch brings it down to about 33mph

That's if you assume train resistance less than half of Davis-- e.g. 33000 lb drawbar pull is needed to roll the 14075-ton train at 35 mph on the level. A hefty 2-8-2 could manage that, if train resistance was really that low.

Another question: did the test train stop at the coal dock at MP 34.6? If so, they couldn't reach top speed by MP 42.2, the foot of the climb.

[CSSHEGEWISCH]

Also remember that the reputed 7498 drawbar HP produced by the H-8 is routinely duplicated by pairs of GEVO's or SD70ACe's on any number of trains every day in North America.

[Dr D]

14,000 tons o' stuff = C&O 1600

Lets see that's 222 capt'n Ahab chasin Sperm Whales each 67 feet long and weighing in at 63 tons - load for a "whale" of a steam engine!

Or 1868 each of - tyrannosaurus rex - of which giant meat eat'n dinosaurs was a mere 7.5 short tons - and that's a train load of nasty passengers for the "tyrant king" of mighty railroad steam locomotion.

Or - let's see - 222 mighty Abrams M1 battle tanks superior in armament protection and electronics - at 68 short tons - thats enough battle equipment to fill a Iraqi Walmart parking lot with a whole lot of hurt from the "hurt locker" - and more than enough for the C&O 1600 weapon of freight railroad warfare!

Or - 5600 Cadillac SRX automobiles - each at 2.5 short tons - thats one big load of big buck limos for the - mile long steam locomotive the true "cadillac" of steam freight power!

Or - let's see - 3 C&O 1600's in one mighty lashup of freight tripple headin' power could haul one RMS Titanic steamship fully loaded from coal bunkers to smokestack weighing in at 46,328 tons - long, short or metric?  - I can just see three mighty Allegheny 1600 draggin the RMS mail ship down the main line - truely a "titanic" feat for a brace of the true "titanic" locomotive of railroad steam power.

Or - At 57,900 short tons it would take a brace of 4 of the afore mentioned C&O 1600 freight railroad main line "battleships" of steam locomotion to haul the 16" gunned USS MISSOURI World War II battleship down the equivalent prescribed main line - Oh yah!

Lets see - C&O had 60 of these monsters - That's enough to haul the whole US fleet of 15 US Navy battlewagons - slidin down the main line - a veritable fleet at sail of mighty steam power

Or - A veritable flotilla of locomotives - to do the work of C&O railroadin across the tide bound Piedmont, Blue Ridge and Appelachia of Kentucky, Michigan, Ohio, Virginia and West Virginia of America.

14,000 tons - haul a whole lot o' stuff!

Doc

[Firelock76]

C&O 1604 is on display indoors at the B&O Museum now, or at least it was six or seven years ago when I saw it.  That is, I TRIED to see it but there were kids swarming all over and in the cab like bees!  Oh well, we've got to manufacture the next generation of railfans somehow!

I'll tell you, see one of those things "up close and personal" and like a Big Boy the sheer size of it will take your breath away.

I won't comment on the 46mph speculation but it's no surprise to me it could pull 14,000 tons o' stuff.

[Dr D]

Of all the mighty 1600's built - 60 for the Chesapeake & Ohio - and 8 for the Virginian Railroad - only two remain - one "hero" C&O 1601 - the second produced - and kept heated indoors in absolutely mint condition in the Henry Ford Museum in Dearborn Michigan.  And one "beater" C&O 1604 which spent years rusting outdoors on the scrap line in Russel, KY. and was later given to the Virginia Museum of Transportation where it was in effect drowned in a flood and afterwards thankfully ended up still stored out of doors in the B&O Transportation Museum.  Thankfully, these two remain of the first 4 built and unlike all the Pennsylvaina Railroad T1 engines and New York Central Railroad Niagara and Hudsons.

Of the - biggest - most mighty - strongest - gigantic - ultimate power - steam locomotives ever built are the 20 Union Pacific "Big Boy" of which 8 survive and these two mighty Cheasapeake & Ohio "Allegheny" of the 68 built.  All but three "Allegheny" lived normal lives and were scrapped - except the two survivors that remain - and one drama queen C&O 1642 went up in a boiler explosion in June 1953 due to suspected low water while working at grade under power killing the crew.  Apparently the boiler ended up quite a ways from the site of the explosion.  One "Big Boy" drama queen UP 4005 similarly was wrecked and went down on its side in a wreck in April 1953 at 50 mph also killing the crew.

Every time I look at that firebox supported by a 6 wheel truck and contemplate the power in that boiler - I reel to that much potential bottled on wheels hurtling at 45 mph with that much loaded coal behind.  It was sheer poetry in 20th century engineering and as massive a steel construction - a glory to behold.

And to think UP 4014 is likely to steam again in the next few years!  And to think C&O 1601 remains existant today - in such a perfect condition - as to also so run again!

Doc

[kgbw49]

Mr. 

Thank you for that excellent analysis! The reputation of the H-8 and the Lords of Steam at Lima will live on!

[erikem]

Dreyfuss,

Looks like a reasonable analysis to me, there was certainly enough hot water in the boiler of an H-8 to support an extra ~600-900DBHP over max continuous DBHP for a couple of minutes.

 - Erik

[Dreyfusshudson]

timz

 

Overmod

calculate an average grade for each section of the test train -- each car, if you feel so inclined.

 

 

No one has ever felt so inclined. To calculate the train's total grade resistance you say: this many cars, total this many tons, are on 0.2% up; that many cars, that many tons are on level; the other many cars, the other many tons are on 0.4% down. Total will be this many times four pounds minus the other many times eight pounds.

 

 

Overmod

you might calculate the likely slack action

 

 

Guess you must mean, cars 130 to 160 are on 0.4% down at momentarily-constant 40 mph and so will be pushing on car number 129 with X pounds of force at that moment, assuming Davis resistance. How will that estimate help you evaluate anything?

 

 

 

 

Not sure if this thread has been forgotten, but I’ll try to add to the debate. My take on the Q&A so far is that the dynamometer car behind 1608 may have measured a transitory 7488DHP, but this may not accurately reflect what the locomotive was actually doing- it could be an artefact caused the uncertain mechanics of a long train on undulating track. I agree with this- the dhp analysis is horribly complex and beyond present capabilities, I think. So, I suggest, we need to think about a different question, namely ‘Could 1608 have delivered 7488 DHP with this load on this track?’  

 

 

 

I have done a number of analyses of the test runs of 1608 from Clinton Forge to Alleghany, Hinton to Alleghany, and northwards from Russell. I use computer models of engine power and efficiency, Boiler outputs and efficiency, and locomotive and train resistance to ‘drive’ the test trains over the stretches concerned. The models have been validated extensively against masses of UK test data collected in the 1950s. They predict results from Altoona collected 40 years prior to that very well. I am beginning to suspect that the laws of physics and chemistry are the same on both sides of the Atlantic. I don’t have detailed C&O gradient profiles, but for reasons you have discussed these are pretty useless when very long trains are involved. I estimate heights along the way from Google Earth, which seems better than it used to be- within a few feet of true at each point. Over a reasonable distance the estimated gradient is therefore pretty much spot on. For example, I get the westbound climb to Alleghany at exactly 1/88 over 12 miles.

 

 

 

My objective is to see if I can reproduce the reported coal and water consumptions. I treat wagon resistance as an unknown- as Sgriggs observes, Johnson cannot be relied on; I adjust the wagon resistance to deliver the reported mean dhp. These exercises have all sorts of uncertainties in them, and if you get within 5% of the reported coal and water values you are doing very well.  The simulations I have done climbing to Alleghany westward with empties and eastward with loads are sufficiently close to what is reported for me to be confident that the models are basically sound in this environment too. 1608 was flat out, if not quite on its knees in near full gear climbing the 1 in 88 at about 13mph with the empties; it really wouldn’t have been practicable to take even 200 tons more. It could have taken perhaps 500 tons more loaded coal eastbound.

 

 

 

The test report makes clear that 6600-6900DHP was regularly sustained on some H8 test runs. At 45 mph, my model of locomotive resistance gives about 700HP. (It would be good to have independent confirmation of this, for they may be special factors in the US the model does not take into account; but there is no reliable data I am aware of). This means the cylinder power required would be around 7500. North of Robbins Ohio, the line rise very gently up the Scioto valley. The line is about 650’ above sea level there. It falls quite sharply to 600’ 5-6 miles further on, down to the riverside. After about 15 miles at near dead level, it then climbs about 100’ over the next 27 miles to Circleville. Driving 1608 in 50% cut off on this stretch with 14075 tons, it bounces along quite happily at 43-46 mph developing, as reported 6600- 6900DHP, until a relatively steep uphill stretch brings it down to about 33mph, rising back to about 46mph at the foot of the downgrade mentioned in the earlier correspondence. If left in this cut off speed would fall by a few mph on the upgrade near mp61, and with it IHP and DHP, back to the bottom end of the range around 6600. But what if cut off were advanced to 55% for 90 seconds or so climbing the grade? Well cylinder power would rise to over 8000, and dhp, albeit for a few seconds only, would hit 7500.

 

 

 

Now I want to stress that all these DHP estimates neglect all the factors you have discussed, and my analysis is really only valid for train of less than a couple of hundred yards length. My only real conclusion is that there seems to be good evidence that an H8 can sustain 7500hp in the cylinders at this speed, and a piece of short term showmanship could easily take this above 8000, in which case 7500DHP is certainly possible.

 

 

 

The engine model say that 7500IHP at these speeds needs about 120000lbs/hr steam to the cylinders, say 110000lbs water evaporated/hr, or about 800lbs/sqft grate/hr, 20000lb coal/hr. This is not unreasonable with this quality of coal- US draughting was able to deliver 1000lbs water/sqft grate/hr as on e.g. the Niagara and T1 tests.  Boiler efficiency is falling rapidly at 800lbs/sqft/hr, so this is likely the sensible practical limit of the boiler- 700lbs/sqft/hr or just under 7000IHP would perhaps be a more realistic estimate of maximum daily capability (The feats of the T1 and Niagara at higher rates are magnificent, but no guide to their daily capabilities- the NYC thought a Niagara was worth no more than 4500IHP in service even though it achieved 6600IHP on test!) Could you up the H8 steam rate short term from 120 to 130000lbs/hr for a minute or so to deliver 8000+IHP?- almost certainly. And if you really wanted to, could you sustain this for longer, to prove a point?- well, quite possibly in my view, but thus would not be a serious measure of true daily capability, just one for the record books.

 

Overall, I’m inclined to let 1608 rest in peace, with its laurels intact.

 

 

[timz]

Overmod

calculate an average grade for each section of the test train -- each car, if you feel so inclined.

No one has ever felt so inclined. To calculate the train's total grade resistance you say: this many cars, total this many tons, are on 0.2% up; that many cars, that many tons are on level; the other many cars, the other many tons are on 0.4% down. Total will be this many times four pounds minus the other many times eight pounds.

Overmod

you might calculate the likely slack action

Guess you must mean, cars 130 to 160 are on 0.4% down at momentarily-constant 40 mph and so will be pushing on car number 129 with X pounds of force at that moment, assuming Davis resistance. How will that estimate help you evaluate anything?

[Dr D]

Regarding much of the performance of the C&O H-8. 

It might be helpful to review the 1940's era dynameter car "weighing head" for measuring drawbar pull.  Quoting Ralph Johnson from Baldwin Locomotive,

"...the modern railway dynamometer consists of a 'weighing head', which connects the drawbar through a properly proportioned lever to pistons operating on the hydraulic principle.  These then transfer through a special diaphrams the variations due to compressive action set up by the movement of the drawbar to a movable pen on the chronograph table...The recording paper is arranged to move at a speed proportionate to the speed of the car...

...as the distance traveled is in a measurable period of time, the foot pounds of work divided by this period will indicate the horsepower developed at the drawbar...

Oil can be used in the hydraulic dynamometer system but as oil is subject to variation in viscosity due to temperature changes, it is better to use a mixture of glycerine and alcohol.  Both rubber and leather are used for diaphrams.  The bearings in the lever are made frictionless by the use of steel pins, carefully hardened, surounded by roller bearings.  The pistions in the weighing head are supported by ball bearings with very close clearance which eliminate any loss due to frictional contact between the pistons and the cylinder bore.  The piston movement on maximum drawbar pulls is only six-thousandths of an inch.  The larger cylinder towards the front of the car takes the buffing shock and can register to 1,250,000 pounds with no loss of accuracy.  The rear cylinder can measure any drawbar pull up to 500,000 pounds. If other limits are desired than the standard one mentioned the cylinder diameters can be changed.

...a telephone circuit connects the dynanmometer car to the engine cab so that coal, throttle, and reverse lever positions, etc. can be signaled back and recorded on the paper by magnet operated pens, and communications maintained with the cab.

Visible pressure gauges are usually placed at the front of the table so that instantaneous readings can be taked of drawbar pull, speed, train line pressures, and locomotive steam pressure.

Dynamometer cars have also been built with the chronograph table compartment raised into a cupola, thus allowing the weighing head and transmission unit to be placed underneath on the main deck of the car, out of the way...in many ways this is a very desirable plan.  It allows the table operator to see much better and gives more room around the table."

--------------------

Regarding the use of the dyanamometer car, Johnson writes,

"A complete code for testing locomotives both in the laboratory and on the road has been issued by the American Society of Mechanical Engineers and adopted as a standard practice of that Society in 1926.  This code with its supplementary sections on 'Definitions and Values', 'General Instructions' and 'Instruments and Apparatus' covers the subject thoroughly."

"Locomotve road tests are inherently less accurate than laboratory tests.  Under the usual conditions of road service thre are bound to be wide fluctuations in speed, drawbar pull, and rate of firing, all fundamental factors in performance; and even under the most rigid control, much of this variation will inevitably remain and exercise an important influence on the results.  In a locomotive, cut-off and speed, for example, vitally affect the steam consumption, and boiler performance likewise varies greatly with the rate at which the boiler is driven.  Road tests should be thoroughly prepard for, conducted with care, and the results carefully analyzed, or they are likely to be misleading.  The use of a dynamometer car is of great help in synchronizing the various records and makes the results much more reliable."  

-------------------------

Again Johnson writes concerning records,

"The number of observers required for a test depends upon the nature of the data to be obtained.  When making an efficiency test at least six observers should be located on the locomotive, two for taking indicator diagrams and any other data that can be taken from the front end, two for cab data and two for coal and water records.

A suitable signal arrangement should be installed so that observations may be properly timed, and where desirable, taken simultaneously.

In the dynamometer car at least two observers are required, one to record the time of each start and stop, location points such as mile-posts, stations, bridges etc., and some to record all information on the dynamometer record and keep track of indicator cards.  When testing articulated locomotves, the engine force is increased to take indicator cards for all four cylinders."

"At the end of each test a summary report should be drawn up roughly to determine if the run was of value, in order that a sufficient number of successful trips can be made to insure bringing out the desired information.  Each observer can average or total his readings, and modern dynamometer cars are equipped with registering counters that enable the operator to make a calculation of the evaproration, drawbar horsepower, and coal per drawbsar horsepower within a few minutes after the run.

If possible, extra men should compute the results and make up a final report on each run, keeping as close behind the road work as possible.  It is only by knowing the results of the previous runs that the man in charge of the test can intelligently plan the subsequent work...

The final report of the test will be drawn from these individual run reports."

-------------------

"Train and Weather Data - all details pertaining to the train such as length and weight, number and kind of cars, distribution of loaded or empty cars and pertinent or unusual conditions regarding lubrication, braking equipment, etc., should be recorded.

The velocity and direction of the wind may be ascertained with an anemometer.  Details of the weather such as rain, snow, temperature, etc., should be noted.  If an anemometer is not available the velocity of the wind may be estimated..."

-------------------

"Drawbar Horsepower - This can be measured and integrated accurately by means of the dynamometer car and is the best measure of actual locomotive performance.  However, as a basis for estimating locomotive performance or design it is valueless because it excludes an unknown amount of work done by the locomotive in moving its own weight, which varies widely according to the load behind the tender, the grade and acceleration."

--------------------------

"RESISTANCE - When hauling a train, a locomotive develops a certain amount of horizontal pull, known as tractive force, which is equal to the resistance of the train to motion.  The total tractive force which is developed in the cylinders of the locomotive, is equal to the total resistance of the locomotive, tender and train.  The available tractive force, or that measured a the tender drawbar is equal to the total tractive force less that required to move the locomotive and tender and is balanced by the resistance of the train.

Resistance at any specified speed is ordinarily measured in pounds per ton, and is equal to the amount of horizontal force required to keep one ton (2000) pounds of train moving at that speed.  If at any speed the locomotive is of sufficient capacity to develop a greater tractive force than required, it is capable of accelerating the train to a higher speed.  A balance is reached when the maximum tractive force that the locomotive can exert at any given speed, is equal to the total resistance of the locomotive and train at that speed."

---------------------

Hope this helps give a context for much of the above discussion.

Dr. D

[Overmod]

timz

You presume the C&O "worked over the grade profile"-- how? And calculated "the estimated change in each car's resistance"-- how?

It's tedious, but:

Look at the grade profile, and calculate an average grade for each section of the test train -- each car, if you feel so inclined.  Use this as a factor in calculating car resistance (on level or upgrade) or gravitational momentum increase (after subtracting a factor for rolling and air resistance, etc. -- use relevant parts of the Ðavis formula) on sufficient downgrade.  This corresponds to a particular moment, or short period, of observed test time at the dynamometer car.  Lather, rinse, repeat for subsequent test-time points.

With a little additional work, you might calculate the likely slack action in the consist for a couple of seconds after the datum point for a given resistance calculation, with the assumption that grade (and resulting grade-determined factors) will not change much in that short a time, but consist dynamics certainly might.

Doing this with a computer, of course, makes the work much less tedious... once you have the model put together.  I'd be surprised if some of the more-determined aficionados of train-simulation programs that have 'physics engines' in them have not already done some of this.

[timz]

Overmod

Can you work over the grade profile, calculate the estimated change in each car's resistance, and sum to get a resistance figure? Yes, and I'd presume the C&O test data includes this kind of analysis.

You presume the C&O "worked over the grade profile"-- how?  And calculated "the estimated change in each car's resistance"-- how?

I'm guessing they attempted no such analysis. What would they do with the result?

Overmod

other factors that could produce -- and I would argue in some very significant cases, did produce -- fluctuations in actual load that throw off the accuracy of the dbhp calculations.

Think the "other factors" were at work between MP 42 and MP 61? The "actual load" at the tender drawbar fluctuated?

[Overmod]

AnthonyV

f I am interpreting their comments correctly, Paul and Tim are stating that the instantaneous net tractive effort and therefore traction power can be determined by applying Newton’s second law to the locomotive (and tender, if applicable) if the instantaneous locomotive acceleration and velocity, grade, and drawbar force are known.

Two questions: Can this approach be used to determine the locomotive DBHP vs. velocity curve?

Sorta.  Remember that in the '40s there weren't sensitive accelerometers, so you had to deduce it from speed data.  Yes, you could insert a factor for acceleration into the calculations for horsepower.

On the other hand, people since Lomonosov have recognized the importance of steady load and steady speed in determining actual dbhp.  The Russians had long continuous grades in the steppes which could be used to get horsepower at various speeds with a minimum of tinkering; the Germans and others used brake locomotives to provide reasonable train resistance without long-train effects; the British (LMS, I think) used the equivalent of dynamic braking to allow 'fine-tuning' of applied resistance.  In my opinion you should NEVER measure horsepower where the locomotive, and perhaps any part of the train, is on a downgrade.

 

How would the behavior of the rest of the train come into play using this approach?

You'd look at the various accelerations (and changes in momenta) for different parts of the train, over the time range of interest, calculate the effect on the resistance, and incorporate into your calculations.  The ringer here is that you don't know the accelerations back in the train; you only recognize the resultant when a 'node' or run-in reaches the rear coupler of the dynamometer car.  The instrumentation isn't measuring that (at least, on most dynamometer cars).

Can you work over the grade profile, calculate the estimated change in each car's resistance, and sum to get a resistance figure?  Yes, and I'd presume the C&O test data includes this kind of analysis.  Problem is that it's theoretical and not "actual" resistance; there are other factors that could produce -- and I would argue in some very significant cases, did produce -- fluctuations in actual load that throw off the accuracy of the dbhp calculations.

[AnthonyV]

Overmod

 

 

Paul Milenkovic

"Behind the dynamometer car-- what do we care what's going on back there? If the engine's drawbar pull is 90000 lb, then it's 90000 lb, . . ."

You, sir, get an A in Engineering Mechanics along with a call to teach the class next semester.

 

 

If that gets an A, then can I teach the phlogiston theory in your school's thermodynamics course for one?

Of course it matters what the train is doing back there.  The dynamometer isn't measuring the locomotive's drawbar pull, it's measuring the strain between the tender drawbar and the dynamometer car, probably either electrically or with hydraulics.  Now if you get run-in from the train, the effective "drawbar pull" the device is measuring has little to do with what the locomotive is delivering at that moment; likewise, if slack runs out you're going to see a very high reading on the dial and graphs, again nonrepresentative of what the engine is producing.

Meanwhile, when the load on the locomotive is reduced, its power is being used to accelerate only its own mass, so it can reach a higher speed than it would if actually pulling the train.  Train resistance will pull that speed down again... but the force measured as the train pulls on the drawbar will create a false impression if you ASSume it's all being generated by the steam rather than yanking the locomotive to a lower speed.

Meanwhile, if my drawbar TE curve has nonlinearities in it (due to shock, slack action, etc.) and I then use a planimeter to integrate under it for relatively short time periods, might I expect to see some resulting figures that... might not exactly reflect the reality of the dbhp actually developed by the locomotive?

 

Overmod

 

I agree with you that the drawbar force cannot be used in a mindless fashion to determine the drawbar horsepower.

 

If I am interpreting their comments correctly, Paul and Tim are stating that the instantaneous net tractive effort and therefore traction power  can be determined by applying Newton’s second law to the locomotive (and tender, if applicable) if the instantaneous locomotive acceleration and velocity, grade, and drawbar force are known.

 

Two questions:

 

Can this approach be used to determine the locomotive DBHP vs. velocity curve?

 

How would the behavior of the rest of the train come into play using this approach?

 

Thanks

Anthony V.

[AnthonyV]

timz

Good point-- they couldn't measure grade since they can't separate it from acceleration, but maybe they could get the combined grade-acceleration correction for the drawbar pull?

Your guess as good as mine, and my guess is no such device. They measured speed and calculated accel from that, and got grade from the track charts, which couldn't be perfectly accurate.

 

Good point Tim.  I wasn't thinking about acceleration when I asked the question.

Anthony V.

[timz]

Overmod

if you get run-in from the train...

Yes, if the pull at the dynamometer car's rear coupler dropped to zero, then the pull at its front coupler would plunge too. Think that will happen when the engine's drawbar pull is 50000 lb and the train is accelerating?

Overmod

if slack runs out...

After it runs in-- think it will run in, all the way forward to the dynamometer car, while the train is accelerating?

Overmod

when the load on the locomotive is reduced [due to slack run-in], its power is being used to accelerate only its own mass, so it can reach a higher speed than it would if actually pulling the train.

It will reach a higher speed than the rest of the train? Because it's no longer pulling the train? How much higher speed?