alstomVery interesting read.CSX has been storing their AC6000CW fleet lately, more than likely due to the economy, and I was wondering why put those units in storage, and leave units such as C40-8 narrow cabs, and C44-9W's out. I know the AC60's are fuel efficient, quite reliable and are some of the most prized units on CSX's fleet. But what has been said makes sense. If you look at it logically, you could preserve fuel by using the AC60's, but it is a poppycock issue to bring up. Two AC6000CW's could produce more horsepower/tractive effort than three GEVO's, while consuming a smaller amount of fuel. In the long run, this would ultimately save on money, and cut back on the emissions that everyone is so worried about.
Very interesting read.
CSX has been storing their AC6000CW fleet lately, more than likely due to the economy, and I was wondering why put those units in storage, and leave units such as C40-8 narrow cabs, and C44-9W's out. I know the AC60's are fuel efficient, quite reliable and are some of the most prized units on CSX's fleet. But what has been said makes sense.
If you look at it logically, you could preserve fuel by using the AC60's, but it is a poppycock issue to bring up. Two AC6000CW's could produce more horsepower/tractive effort than three GEVO's, while consuming a smaller amount of fuel. In the long run, this would ultimately save on money, and cut back on the emissions that everyone is so worried about.
If the GEVOs are DC then it is possible that the AC6000CWs could produce more TE, but they would have 1200 less hp.
The AC6000CW fleet on Union Pacific has not been "scrapped", but has been rebuilt, or downgraded to 4,400 hp. UP didn't feel the need to have 6,000 hp units. There are quite a few standard AC60's on the UP roster, if I remember correctly, kept primarily in DPU and coal service, and not paired up. The downgraded AC60's are known as AC60/44CW's.
GE will no longer guarantee the availability of replacement parts for the HDL diesel after 12/31/09, after that date GE will sell you parts that they have, but when they are gone your done. This is why CSX and BHP arer having their AC6000CWs re-engined with the 16-cyl. GEVO diesel. The Union Pacific chose not to do this, instead they are replacing the 16-cyl. HDL angine with the lower hp. 16-cyl FDL engine. This in effect makes them the same as the un-converted "Convertible" AC6044CWs. Any UP AC6000CW numbered in the 75xx series still has the 6000 hp. HDL diesel, any AC6000CW numbered in the 69xx series now has the 4390 hp. 16-cyl. FDL diesel. Any AC6000CW look-alike numbered in the 70xx series or 71xx series,was built as a AC6044CW and will stay as built.
The good news, looking at this from a railfan's point of view, is that the CSX AC6000CW's seem to be slowly returning from storage. CSX units 621 and 627 were spotted pulling a mixed freight today. I know that the 621 was in storage for a little while, and I believe 627 was stored for some time in Rice Yard at Waycross, GA. Hopefully more will come out.
carnej1 For what it's worth EMD is also coproducing a SD90MAC-H equivalent in and for China and claims to improved the 265 H but time will tell (and see above about North American customers)..
For what it's worth EMD is also coproducing a SD90MAC-H equivalent in and for China and claims to improved the 265 H but time will tell (and see above about North American customers)..
The EMD 16V265H engine is currently the dominant engine in the Marine market for its power range right now. So basic problems must have been solved.
With regards to Paul's earlier comments about settling for lesser availability, take the case of CP's last batch of ES44ACs which did not favorably impress CP. Three brand-new ES44ACs were assigned to a transcontinental Intermodal out of Montreal, before the train reached the first crew change point the train stalled because one of the locomotives died, not normally a problem as two ES44AC could easily move the train, except that the air compressors on the other two locomotives had seized up and they couldn't pump the air up to release the brakes. A Dogcatch crew had to be sent out of Sudbury with a set of replacement power. It was about this time that CP discovered that GE had substituted Gardner-Denver air compressors for CP's favored Wabco type. Needless to say GE had to replace the compressors in all 40 locomotives. Many failed before GE got them all changed. Gardner-Denver has manufactured large numbers of compressors for locomotives, all Soo Line power is equipped with them, I don't know whether this was a bad batch, or if this particular model has compatiblity problems with the GEs, but for GE to change the model specified by CP was not a good decision.
SactoGuy188 The problems with the SD90MAC-H and AC6000CW was that both EMD and GE were literally pushing the limits of prime mover technology, and as such both these models suffered a lot of serious reliability problems. (Mind you, I think GE has a better chance to develop an ES60AC model since the Deutz HDL prime mover suffered fewer problems than the 265H prime mover EMD developed.) I would not be surprised that GE does look again at building an ES60AC model, especially with tightening EPA emission rules, which may require trains with smaller locomotive consists.
The problems with the SD90MAC-H and AC6000CW was that both EMD and GE were literally pushing the limits of prime mover technology, and as such both these models suffered a lot of serious reliability problems. (Mind you, I think GE has a better chance to develop an ES60AC model since the Deutz HDL prime mover suffered fewer problems than the 265H prime mover EMD developed.) I would not be surprised that GE does look again at building an ES60AC model, especially with tightening EPA emission rules, which may require trains with smaller locomotive consists.
Being done. As mentioned earlier in the thread the CSX and BHP AC60 fleets have been rebuilt with 16 cyl. GEVO engines and GE is building (actually coproducing with a Chinese manufacturer) a "Sino-fied" 6,000 HP Evolution series locomotive for China Railways...
As for a domestic new- build ES60AC it does not seem that there is customer interest presently (even CSX).
"I Often Dream of Trains"-From the Album of the Same Name by Robyn Hitchcock
If you're talking today's ES44AC/ES44DC from GE and SD70M-2/SD70ACe from EMD, they're generally pretty reliable units for one reason: these new locomotives don't push the reliability limits of both the prime mover and traction motor. Remember, both the current GE Evos and the current EMD SD70 models are based on technology that have been around since the early 1990's with the SD70M/SD70MAC and C44-9W/AC4400CW models, so the newest locomotives should have a proven reliability record.
tleary01 I do not know the specific adhesion-management software used on either type of locomotive, though both EMD and GE specification tables in their locomotive service manuals stated that the maximum (theoretical) tractive effort for each locomotive was 200 Kilo-Pounds. In reality I have never observed more than 183 Kilo-Pounds indicated while pulling a train. In the locomotive service manuals EMD and GE state the maximum tractive effort a 200 K# but I have never seen an AC locomotive get there, instead the tractive effort is limited to the maximum attainable with the existing adhesion when at full throttle and speed below 12 mph. The nominal axle load weight for EMD and GE AC drive locomotives is stated at 70,000 pounds per axle or a total nominal weight of 420,000 pounds with approximately 1/2 fuel and sand supplies.
I do not know the specific adhesion-management software used on either type of locomotive, though both EMD and GE specification tables in their locomotive service manuals stated that the maximum (theoretical) tractive effort for each locomotive was 200 Kilo-Pounds.
In reality I have never observed more than 183 Kilo-Pounds indicated while pulling a train. In the locomotive service manuals EMD and GE state the maximum tractive effort a 200 K# but I have never seen an AC locomotive get there, instead the tractive effort is limited to the maximum attainable with the existing adhesion when at full throttle and speed below 12 mph.
The nominal axle load weight for EMD and GE AC drive locomotives is stated at 70,000 pounds per axle or a total nominal weight of 420,000 pounds with approximately 1/2 fuel and sand supplies.
My frame of reference is CSXT; and the reason for my question was that CSXT never tested an SD90MAC. So I know nothing about their actual performance.
CSXT operates AC4400CWs and ES44ACs with nominal weights of 432,000 pounds and software that allows the units to reach a maximum TE of 200,000 pounds. Each traction motor can reach 36,000 pounds TE; but each unit has an overriding limit of 200,000 pounds to ensure that two-unit consists do not produce more than 400,000 pounds. The units do not produce that level of TE routinely (which is presumably why GE was willing to increase the per-motor TE cap from 30,000 pounds to 36,000 pounds); however CSXT relies of them to reach the 200,000-pound level (especially when they are in helper service and the rails have been conditioned by the passage of the train) when necessary to prevent stalls.
CSXT does not operate EMDs capable of producing 200,000 pounds TE; however in testing sessions it has achieved that level of TE from SD70ACes which had nominal weights of 432,000 pounds and were configured with software that had been developed for the SD90MAC.
The boneyards are filled with failed businesses that did just what Paul is suggesting. Let the quality slip, save some quick bucks, then die. Let's see, GM comes to mind. Once you get a reputation for poor quality it's hard to change it. People remembered the Vega for a long long time. Trucking companies cursed GM's Detroit Diesel engines for poor quality and GM no longer makes heavy duty truck engines.
One very important thing I learned in B school was that for every customer who complained there were many other disatisfied customers who were silently looking for somewhere else to take there business. Do not go that route.
Because most disatisfied customers don't complain and just go away silently, they're hard to quantify. This often allows the dang fools who want to intentionally degrade quality to prevail. They've got numbers on the money "saved". Their opponents often can only say "We're pissing off the customers." The numbers often win. And they keep "winning" right up until the doors close.
Coffee is a great example. There are two basic types of coffee beans. One is more expensive and tastes better. The other is cheaper but produces a bitter taste. It was ever so tempting to let quality slide and use more and more of the cheaper beans in the mix. Eventually, the coffee companies were in trouble as people drank less and less coffee. But, they had saved dollars in the mean time.
It took Starbucks to reintroduce quality coffee to America and revive coffee sales. Others followed Starbuck's lead. I know Starbucks now has its own problems. Having a quality product doesn't guarantee success. But having a poor quality product does guarantee eventual failure. Intentionally degrading the quality of your product or service is corporate suicide.
I do not know the specific adhesion-management software used on either type of locomotive, though both EMD and GE specification tables in their locomotive service manuals stated that the maximum (theoretical) tractive effort for each locomotive was 200 Kilo-Pounds. In reality I have never observed more than 183 Kilo-Pounds indicated while pulling a train. When the locomotive is actually moving the train at throttle 8 the tractive effort will increase from the stall maximum because a moving wheels can be driven into "creeping adhesion" where the wheel is driven slightly faster than the linear rail speed, but not actually slipping. During the creep condition the wheels "sing" with a varying pitch as the traction control system hunts for the maximum adhesion without causing slip, the traction control does the same at stall but wheel creep improves adhesion. In the locomotive service manuals EMD and GE state the maximum tractive effort a 200 K# but I have never seen an AC locomotive get there, instead the tractive effort is limited to the maximum attainable with the existing adhesion when at full throttle and speed below 12 mph.
The nominal axle load weight for EMD and GE AC drive locomotives is stated at 70,000 pounds per axle or a total nominal weight of 420,000 pounds with approximately 1/2 fuel and sand supplies. (a standard SD40-2 has a nominal weight of 396,000 pounds/66,000 pounds axle weight). What is not usually considered is how much fuel weight affects tractive effort which was another "deal breaker" for the 6,000 HP locomotive. A 6000 HP locomotive will consume 50% more fuel than a 4000 HP locomotive under the same operating conditions so that the fuel load of each locomotive has a 50% greater impact on maximum tractive effort because an AC locomotive's maximum tractive effort is really limited by available adhesion and is relative to wheel/axle load weight and rail conditions.
As a note to my previous message, when a stall test or moving at full throttle below 10 MPH the GE AC drive locomotives originally were much more agressive than the SD90 in attempting to drive the traction motors and instead of sitting or moving smoothly with strong tractive effort and only an occasional short over-shoot, the GE AC locomotives would literally "jump" producing a strong shock effect to the carbody which caused some secondary failures to components within the carbody. GE has since tuned the drive control software to minimize this effect, but that is another story.
oltmanndOK. Thanks. Any interchangeable parts? Crank? Pistons? Con rods? Piston carrier? Fuel injection equipment?
OK. Thanks. Any interchangeable parts? Crank? Pistons? Con rods? Piston carrier? Fuel injection equipment?
There can't be very many interchangeable parts. GE has decided to end support for the HDL, and all three owners of locomotives equipped with the engine are making or have made moves to eliminate locomotives powered by that diesel motor. CSX is having their AC6000CWs re-engined with the 16-cyl GEVO diesel, as did BHP in Australia. The Union Pacific is going the other way and is well along in re-engining their AC6000CWs with a 16-cyl 7FDL GE diesel rated at 4390 hp. BHP is done with their 8 conversions, I think both CSX and UP will be done by the end of the year.
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
oltmannd carnej1Regarding CSX's AC60 fleet I would bet they would alll be retired if GE had not repowered them (on their own dime for the most part) with 16 cylinder GEVO prime movers. Isn't the GEVO basically the marketing name for the current version of the HDL?
carnej1Regarding CSX's AC60 fleet I would bet they would alll be retired if GE had not repowered them (on their own dime for the most part) with 16 cylinder GEVO prime movers.
Isn't the GEVO basically the marketing name for the current version of the HDL?
No. Although the GEVO did evolve from the HDL there are some major differences/improvements. One major difference is that the GEVO uses a single turbocharger versus the HDlL's two...
Other than the occasional nice spreads that Trains does on new locomotives (great graphics) discussions of the details of diesel locomotives, how they look and how they work, tends to make me sleepy. Never-the-less, I still have some views, as useless as they may be.
On the subject of 4,000 and 6,000 HP units, I've had the vague impression that the fact that a fleet of 4000hp units provide a somewhat better chance of matching HP to the requirements of a given train has caused some disfavor of the idea of a 6000HP fleet. If you need 7000HP, and only have 6's, you get 12,000 on the train, rather than 8000HP with two 4000 HP units. Admitedly, I have no notion of the difference betwee the cost to own and operate the two sizes, so maybe that is not a big issue. Not that the crew would care.
Given the proliferation of types and sizes, there must be days at the power desk and terminal engine facility when someone dreams of having nothing but a string of 1800HP units, fully equiped with all the modern technology, just waiting in line to fill the power requirements of the next train. Will that be 2 or 4, sir? Guess that is why they make the big bucks.
As for road failures, a post above asked if the marginal cost of maintaining high reliability is justified by the avoidance of the cost of road failures. I can't answer with any knowledge, but I recall a forum post way back that authoritively gave some very high costs for dog catch crews. If you figure that a road failure on a line runnig near capacity causes delays all over the place, with more crews going on hours, and the messing around to get the dead train powered up, maybe the higher cost of reliability is justified. Can't be sure, but I am inclined to think that locomotive builders and railroad planning people have probably pushed the numbers around a bit.
"We have met the enemy and he is us." Pogo Possum "We have met the anemone... and he is Russ." Bucky Katt "Prediction is very difficult, especially if it's about the future." Niels Bohr, Nobel laureate in physics
tleary01 By now a many "facts" have been presented concerning this subject. I want to provide some real numbers derived from actual train tests and operation. . an AC4400 and AC6000, and by the way an SD90 (with the same axle load weight) will produce the same stall/low speed tractive effort under the same adhesion conditions.
By now a many "facts" have been presented concerning this subject. I want to provide some real numbers derived from actual train tests and operation. .
an AC4400 and AC6000, and by the way an SD90 (with the same axle load weight) will produce the same stall/low speed tractive effort under the same adhesion conditions.
In the comparison between the AC4400CW and the SD90MAC, what adhesion-management software did the AC4400CW have; and what was its nominal weight?
Don't confuse reliability vs availability.
During the late Dash 8 era, GE had some stats showing their locomotive "mission reliability" was 3 road failures per loco per year.
Availability, which would include the scheduled quarterly and annual inspections, trips to the wheel truing machine and any failures, would be good at 96%. If you throw in trips to the backshop, wreck repair, etc., a good overall fleet availability number would be 92%-94%.
Paul the answer to your question will depend upon other factors, such as; is the lower availability due to more frequent breakdown, or waiting for parts,, or waiting for shop time, or time needed to reach shop. Offsetting points to consider, will the extra inventory of parts, shop costs, and extra locomotives needed, be greater than the savings in purchase price, will we need a larger mechanical force to repair the more trouble prone locomotives, etc.
By now a many "facts" have been presented concerning this subject. I want to provide some real numbers derived from actual train tests and operation.
The tonnage a locomotive can pull is a result of the tractive effort that it can produce. As track speed increases more horsepower is needed to produce tractive effort. Once the locomotive speed increases to the point of maximum tractive effort at full available horsepower a further increase in speed will be accompanied by a reduction of tractive effort until the reduced tractive effort is just enough to keep the train moving at that balance speed. The advantage of high horsepower locomotive is to move that balance speed to a higher value. This is a simplified statement that does not consider wheel to rail adhesion that is not always ideal, then tractive effort will have to be reduced to keep the wheels from slipping.
Considering that an AC drive locomotive can produce higher tractive effort than a DC drive locomotive with the same horsepower and adhesion conditions and that an AC drive cannot force the wheels into a slipping condition, an AC drive locomotive will pull more tonnage and acheive a higher balance speed than a similar powered DC drive locomotive. Increased horsepower, 6000 vs 4400 drives that balance speed higher because it keeps tractive effort higher as speed is increased. In theory this is just what a railroad would want for high speed intermodal trains. The theory is good, the engines and drive systems just didn't hold up to the high power demanded of them.
Conversely, at starting speeds and slow drag speeds when tractive effort is maximum and not limited by horsepower, then it is a matter of the traction motor's capability to produce tractive force which is limited on AC drive locomotives by wheel to rail adhesion. The maximum tractive effort produced by a six axle AC locomotive is not determined by engine horsepower, an AC4400 and AC6000, and by the way an SD90 (with the same axle load weight) will produce the same stall/low speed tractive effort under the same adhesion conditions.
I have performed the following test on all 3 types of AC locomotives: Apply full service brake to the train so that the locomotive cannot move the train, place the throttle directly into run 8, and observe the indication of the tractive effort meter on the operator's display. All 3 types will smoothly and rapidly build tractive effort to 175 - 180 Kilo-Pounds and stretch the couplers then stop moving. Every few seconds one of the wheels will make a "ping" as it rotates about 1/4 inch and stops. The tractive force is so strong that the top of the wheel bearing adapter is almost touching the truck and there is no gap between the coils of the journal springs. An AC drive locomotive can sit in that condition until the locomotive runs out of fuel and will not harm its traction motors.
When accelerating a train or when a grade slows the AC drive locomotive, if adhesion remains good, that high tractive effort will be sustained at speeds below approximately 12 mph. Simple math will tell you that 3 AC drive locomotives can exceed 500,000 pounds of tractive effort under ideal adhesion conditions, more than a grade E coupler knuckle is rated for. To employ full tractive effort from 3 AC drive locomotives, distributed power is necessary. However, even when a single AC drive locomotive is working as a remote on the rear of the train its high tractive effort at low speeds is capable of lifting lightly loaded cars near the end of the train while in a curve. In this case distributed power can be set for the CTE mode (Controlled Tractive Effort) so that the remote locomotive tractive effort is limited to 120,000 pounds maximum.
BN went 5:3 with the SD70MACs vs SD40-2s. You can figure about 35% adhesion for the SD70MACS and 18-21% for the SD40-2s (Conrail only figured 18% for them, but part of that was Flexicoils vs HTC)_
The light weight of the AEM7s uis why the DC version is only good for 8 or 9 cars and the AC version is good for 12 or 13.
BTW, the speed TE curve for an AEM7 is flat at the adhesion limit up to nearly 100 mph.
Railway Man oltmannd Um...not quite. It didn't start out that way, anyway. Way back just at the dawn of AC power, the AAR formed an ad-hoc committee to write specs for a test fleet of AC units. The plan was for all the roads to pitch in enough to get a reasonalbly sized order placed - say 20 units. There was concensus on the committee that a 6000 HP AC unit was the goal. They'd have the same ratio of HP to TE as an SD40-2/C30-7 and an SD60/C40. - you could do 2:1 with the former and 3:2 with the latter and get exactly the same train performance, so they'd be just as flexible in application as the existing fleet. That whole plan was short-circuited by the BN's purchase of a zillion SD70MACs for drag service. [snip; emphasis added - PDN.] Interesting to hear this from the other side of the big river, and from the mechanical department perspective. Approaching it from my side (network planning), we couldn't make the building blocks work. From our perspective the comparison was never with SD40-2s/C30-7s or SD60s/C-40s, but with C44-9Ws and SD70M-2s. (Paul - the example I threw out was just to show the ratios. It's not an actual train plan.) RWM
oltmannd Um...not quite. It didn't start out that way, anyway. Way back just at the dawn of AC power, the AAR formed an ad-hoc committee to write specs for a test fleet of AC units. The plan was for all the roads to pitch in enough to get a reasonalbly sized order placed - say 20 units. There was concensus on the committee that a 6000 HP AC unit was the goal. They'd have the same ratio of HP to TE as an SD40-2/C30-7 and an SD60/C40. - you could do 2:1 with the former and 3:2 with the latter and get exactly the same train performance, so they'd be just as flexible in application as the existing fleet. That whole plan was short-circuited by the BN's purchase of a zillion SD70MACs for drag service. [snip; emphasis added - PDN.]
Um...not quite. It didn't start out that way, anyway.
Way back just at the dawn of AC power, the AAR formed an ad-hoc committee to write specs for a test fleet of AC units. The plan was for all the roads to pitch in enough to get a reasonalbly sized order placed - say 20 units. There was concensus on the committee that a 6000 HP AC unit was the goal. They'd have the same ratio of HP to TE as an SD40-2/C30-7 and an SD60/C40. - you could do 2:1 with the former and 3:2 with the latter and get exactly the same train performance, so they'd be just as flexible in application as the existing fleet.
That whole plan was short-circuited by the BN's purchase of a zillion SD70MACs for drag service.
[snip; emphasis added - PDN.]
Interesting to hear this from the other side of the big river, and from the mechanical department perspective. Approaching it from my side (network planning), we couldn't make the building blocks work. From our perspective the comparison was never with SD40-2s/C30-7s or SD60s/C-40s, but with C44-9Ws and SD70M-2s.
(Paul - the example I threw out was just to show the ratios. It's not an actual train plan.)
RWM
That's OK; the challenge of matching the number and capabilities of locomotives to the actual needs of the service remains, regardless of the specific numbers for any particular loco or service.
In fact, that problem gets exacerbated as the individual units have bigger numbers - tractive effort and/or horsepower - the "building blocks" are so big and sizable that it's got to be hard to find a good fit or balance between an integer number of units ( 2-2/3 units, anyone ?), and varying train sizes that just aren't that neat when trying to meet a specific service schedule. Balancing all of those factors - HP/ ton ratio and hence total HP, TE req'd, costs, etc. has to be hard when the "increments" or quantum "steps" are so large and inflexible.
Interestingly, it's those same SD70MACs that oltmannd mentions that Al Krug writes about in his "Drawbar Limits / Coupler Forces" essay at: http://www.alkrug.vcn.com/rrfacts/drawbar.htm He notes that they regularly produce 120,000+ lbs. TE, whereas the older SD40-2's were good for only 90,000lbs. each. His explanantion of "Tractive Effort vs. Horsepower" is also pretty good exploration of the subject at:
http://www.alkrug.vcn.com/rrfacts/hp_te.htm
Finally, for another extreme to meet a different "service plan", consider Amtrak's ASEA-EMD AEM-7's and MARC, NJ Transit, & SEPTA's ALP-44's - 7,000 HP, with not much weight - only 202,000 lbs. = 101 tons ! - on 4 axles in B-B trucks. Clearly a locomotive for short, light trains that will be going fast - up to 125 MPH !
- Paul North.
Great counter-point ! Here's my response, which may illustrate it better:
But I already get that 99.5+% reliabiilty from my "commuter car" - a '95 Olds sedan - with only routine scheduled maintenance and annual PA state mechanical inspection, and fixing what turns up then. Only 2 road failures in 11 years - both were for 2- 3 days at most (arguably 1 was operator-caused, at that).
So why should I spend $20,000 or so as a one-time lump-sum - or $250 to $300 per month if I finance it - for a new car, to get just marginally better reliability ? I could instead bank the savings, and call a taxi on the rare occasion when the chariot breaks down. A few minutes later than normal into work once in a long while is no big deal where I work now - no more so than a slightly late train is on the railroad. And / or rent a replacement car when it's in the shop for said inspection and maintenance. Withal, I'd still be - and am - thousands of dollars ahead each year.
So my question remains: So wouldn't or couldn't the railroad consider doing basically the same thing, writ larger ?
Paul_D_North_JrAs a matter of "least total cost economics", has the reliability goal - as exemplified by the above data - been pushed too far ? Have we gone beyond the point of diminishing returns with that ?
Lets assume 96% reliability.
Would you by a new car if I assured you that it would only spend 2 weeks of every year you own it broke down in a shop? (excluding normal servicing and inspection time) That's 96% reliability.
Would you by a new car if I assured you that it would only spend 1 day a month, every month of every year you own it broke down in a shop? (excluding normal servicing and inspection time) That's over 96% reliability.
I know my wife gets upset when if one of our vehicles is down more than a couple days a year (99.5% reliability).
So, no I don't think we are at the point of diminishing returns yet.
Dave H. Painted side goes up. My website : wnbranch.com
oltmannd Drawbar strength is a bit of a grey area. It's definitely a fatigue life proposition, but the max load is only felt by the first car, so there's this whole probability function that has to be part of the calculation. I've never seen anybody try to keep fatigue life statistics for drawgear. Most of the work I've seen was done using computer simulations of particular trains and then some back-of-the-envelope engineering judgment. Rule of thumb at Conrail was you could use 2 ACs on anything and 3 on unit trains with grade E. I got into a big arguement with an EMD regional sales manager over whether we could get a knuckle on a fairly small train with 3 AC units. I said "yes". He said "no". He didn't understand F=ma.
Drawbar strength is a bit of a grey area. It's definitely a fatigue life proposition, but the max load is only felt by the first car, so there's this whole probability function that has to be part of the calculation.
I've never seen anybody try to keep fatigue life statistics for drawgear. Most of the work I've seen was done using computer simulations of particular trains and then some back-of-the-envelope engineering judgment. Rule of thumb at Conrail was you could use 2 ACs on anything and 3 on unit trains with grade E.
I got into a big arguement with an EMD regional sales manager over whether we could get a knuckle on a fairly small train with 3 AC units. I said "yes". He said "no". He didn't understand F=ma.
Don (and others) -
Have you read Al Krug's "seat-of-the-pants" experience-based discussion of this, at the URL below ? [though it seems to be off-line / "Access denied !" at the moment]
"Railroad Facts and Figures" - "How Much Force Can a Coupler Withstand?" at: http://www.alkrug.vcn.com/rrfacts/drawbar.htm
He would understand, I think. Also says the train usually breaks about 10 cars back ["notwithstanding"], despite the theoretical maximum pull at the 1st car. Interesting stuff.
Railway Man Current models are delivering in the high 90s availability rates. That measures the amount of time the locomotive is available for service vs. the amount of time it is not. Time not available includes refueling, FRA inspections, program maintenance, and unplanned maintenance. Thirty years ago we were seeing availability rates in the high 80s. About 8 years ago, I kept track of how many locomotive failures I had over a six-month period during my shift on my territory that caused train delays. Average trains per shift per day was 35. Average days per week 5.5. Total trains measured, slightly more than 5000. After eliminating collisions with vehicles, rocks, trees, large animals, and washouts, locomotives running out of fuel, and crew mistakes such as forgetting to release handbrakes on a DPU, the answer is ... zero. However, we would never let a train out of a terminal with known poor performers, either. The average train delay for a locomotive failure for any cause that I was then experiencing was in excess of 8 hours, it almost invariably required a recrew, and almost invariably delayed other trains up to 1-2 hours. Sometimes it affected only 1 or 2 other trains, sometimes it affected 20 or more. Train delays are extraordinarly costly because the event quickly ripples into the entire subdivision, the end terminals, and upsets crewcalling plans, yard blocking plans, service plans, car trip plans, and locomotive utilization plans, and usually affects train performance incentives and other service guarantees. It's difficult to measure the total cost because the effects linger for days and are subtle. But it is high. [3 paras. snipped; emphasis added - PDN.]] RWM
About 8 years ago, I kept track of how many locomotive failures I had over a six-month period during my shift on my territory that caused train delays. Average trains per shift per day was 35. Average days per week 5.5. Total trains measured, slightly more than 5000. After eliminating collisions with vehicles, rocks, trees, large animals, and washouts, locomotives running out of fuel, and crew mistakes such as forgetting to release handbrakes on a DPU, the answer is ... zero. However, we would never let a train out of a terminal with known poor performers, either.
The average train delay for a locomotive failure for any cause that I was then experiencing was in excess of 8 hours, it almost invariably required a recrew, and almost invariably delayed other trains up to 1-2 hours. Sometimes it affected only 1 or 2 other trains, sometimes it affected 20 or more. Train delays are extraordinarly costly because the event quickly ripples into the entire subdivision, the end terminals, and upsets crewcalling plans, yard blocking plans, service plans, car trip plans, and locomotive utilization plans, and usually affects train performance incentives and other service guarantees. It's difficult to measure the total cost because the effects linger for days and are subtle. But it is high.
[3 paras. snipped; emphasis added - PDN.]]
OK, so now we've got some real-life experience-based statistical data to work with. What you've done is to pull out the non-mechanical failures, so that what's left is purely "road breakdowns" = zero ! Kudos to those Mechanical Dept. and Power Desk folks ! Now, from that I'll take "devil's advocate" role here and pose the challenge: As a matter of "least total cost economics", has the reliability goal - as exemplified by the above data - been pushed too far ? Have we gone beyond the point of diminishing returns with that ? Allow me to expand:
To help understand my point, this analysis is like the joke about the engineer (my kind, not the locomotive kind): When he was asked about the proverbial glass with water to a level that is halfway up it - "Is it half-full, or half-empty ?", replied with "Neither - but it's either twice as large as it needs to be, or it's being used only at half-capacity !" So it is here.
This is not airplanes, where the tolerance for in-flight failures has to be as near to zero as possible, at the considerable alternative expense of a lot of hanger time for inspections and preventative maintenance, etc. In contrast, the railroad can tolerate a few road failures from time to time, with effects that are costly, but not disastrous. In fact - as the narrative above makes clear - the railroad already has to do that anyway !
Even with a fleet of entirely fault-free locomotives, there are enough delays resulting from a myriad of other random causes - weather and water, "outside" factors such as idiots at grade crossings and trespassers, rocks, trees, animals, derailments caused by equipment, track, or operator failure, etc., etc. - that a "perfect" smooth and fluid, scheduled operation just isn't going to happen consistently, much as that might be desirable. There's always going to be some chance and occurrences of disruption that's going to have to be mitigated and "worked-around" by adjusting and modifying the operations accordingly.
In view of that, what's the point - what is to be gained ? - by trying to attain a theoretical standard of "no locomotive road failures" ? It requires spending a lot of money in design, construction, testing, trouble-shooting, and maintenance to achieve that, maybe also making the locos heavier and more complicated than they otherwise need to be. But I'll bet that reliability could be dialed back a little bit and some savings realized therefrom, without drastically increasing the costs of the road failures.
Yes, road failures will increase somewhat, if that is done. But we've seen that the railroad already has to cope with that kind of thing on a regular basis, so it's not like an occasional locomotive failure is going to introduce disharmony into an otherwise clock-like operation. And using the example provided by RWM above, what if only - say, 1 road failure a month - had resulted from slightly less reliable locomotives over those 6 months ? Meanwhile the savings on maintenance of the multiple locomotives on those 5,000 trains - at say, a mere $10 per train - would be $50,000 over the same time period ? That means that each of the half-dozen directly delayed trains would have $8,300+ of savings to use to offset all of the direct and indirect costs of that one road failure per month. Would that savings be enough to cover those costs ? I don't know the precise answer, and I realize this is a "slippery slope" - I'm not advocating a "no (or low) maintenance" policy and a Penn-Central-type of band-aid operation.
Perhaps someone more knowledgeable than me has looked at the matter, and made an informed and intelligent decision that a "zero tolerance" for failure is the best way to go - certainly it has a simple, basic appeal to it. But I doubt if that's the ideal answer - rarely does a standard of "perfection" result as the optimum solution for the trade-off between minimizing maintenance costs and failure costs, at the point where the sum of those 2 independent costs is at it lowest.
Any other throughts or responses to this ?
oltmanndDrawbar strength is a bit of a grey area. It's definitely a fatigue life proposition, but the max load is only felt by the first car, so there's this whole probability function that has to be part of the calculation.
CSXT's standard, for locomotive application purposes, is 400,000 pounds. Its high-TE AC4400CWs and ES44ACs are limited to 200,000 pounds of TE per unit to reduce the risk of two-unit consists breaking drawbars.
oltmanndUm...not quite. It didn't start out that way, anyway. Way back just at the dawn of AC power, the AAR formed an ad-hoc committee to write specs for a test fleet of AC units. The plan was for all the roads to pitch in enough to get a reasonalbly sized order placed - say 20 units. There was concensus on the committee that a 6000 HP AC unit was the goal. They'd have the same ratio of HP to TE as an SD40-2/C30-7 and an SD60/C40. - you could do 2:1 with the former and 3:2 with the latter and get exactly the same train performance, so they'd be just as flexible in application as the existing fleet. That whole plan was short-circuited by the BN's purchase of a zillion SD70MACs for drag service. UP, CP, CSX and Conrail tried the higher horsepower, general service variety. UP even opted for that "convertible" set-up that would allow the 6000HP engine to plop down in place of the 16-710. UP and CP had a horrible time with the 6000 HP EMD engine. CSX had trouble with their GE's but stuck with it and pairs of AC60s seem to be rather happily applied to intermodal trains around their network although CSX seems rather content to purchase DC machines (ala BNSF) for their higher HP/ton service of late. Conrail's plans were cut off at the knees by the merger. And thus endith the saga of the 6000 HP AC locomotive (for now, anyway)
UP, CP, CSX and Conrail tried the higher horsepower, general service variety. UP even opted for that "convertible" set-up that would allow the 6000HP engine to plop down in place of the 16-710.
UP and CP had a horrible time with the 6000 HP EMD engine. CSX had trouble with their GE's but stuck with it and pairs of AC60s seem to be rather happily applied to intermodal trains around their network although CSX seems rather content to purchase DC machines (ala BNSF) for their higher HP/ton service of late.
Conrail's plans were cut off at the knees by the merger.
And thus endith the saga of the 6000 HP AC locomotive (for now, anyway)
What I find odd consider the history of the SD90MAC/AC6000CW is that the Chinese National; Railway is planning on operating their new 6,000 HP EMD and GE units in 3 unit 18,000 HP lashups for heavy haul service.
I know that at least one of the Australian iron ore railroads bought AC6000CW's for low speed ore train service but is that a waste of horsepower (i.e I've always read that higher horsepower per axle is most helpful in high speed freight service and that a 6,000 HP locomotive is not efficient at low speeds)? The same operator has since bought 4,4400 HP GEVOS.
Regarding CSX's AC60 fleet I would bet they would alll be retired if GE had not repowered them (on their own dime for the most part) with 16 cylinder GEVO prime movers. In fact CSX had earlier looked at rebuilding them into AC44000CW equivalents with FDL 16 engines (one engine was so converted, IIRC).
Some thoughts that occurred to me while reading through this thread:
1.) Note that the "service design" ==> operating doctrine described by RWM above seems to be driven by essentially utilizing all of the available tractive effort capacity of the coupler-drawbar system (see excerpts from Al Krug below), and then adding horsepower as needed for the speeds desired or demanded for the particular service.
2.) Also, it seems to me that the 450,000 lb. tractive effort standard pretty much leads to having to use a DPU configuration to keep the drawbar stress within allowable limts.
Staying with RWM's figures (for consistency): A single 420,000 lb. unit - 6 axles at 70,000 lbs. each, pretty much a practical maximum - can only exert 150,000 lbs.of tractive effort if it's adhesion factor is 36 % or better. So a single unit of 6,000 HP (or even a gazillion HP) on only 6 fully-loaded axles is only going to use about 42 % (150,000 / 360,000) of the pulling capabilty of the coupler-drawbar system. Unless it's a really light train that just has to be sent out now, that's a waste of potential train that could have gone along - with the same crew - just by adding another locomotive or two and enough cars to use up the rest of the coupler-drawbar capacity.
If another (2nd) locomotive is added - then 300,000 of that 360,000 DB capacity is being used (83 %), but there's still some 17 % going to waste. Add a 3rd locomotive to use up the rest - but then the engine consist might be over the allowable load on the couplers and drawbars - that's why most railroads have limits on the number of powered axles on the head end. So either go to a DPU arrangement to spread out the TE, or plan on using a little less than all of the available TE in the worst case, going up the ruling grade. As long as the locos are going faster than the speed of their max TE rating, by their constant HP nature that lower TE will happen automatically anyway: speed x TE = a constant, and decreasing the speed increases the potential TE (but only within the adhesion limits), and vice-versa.
From Al Krug's "Railroad Facts and Figures" - "How Much Force Can a Coupler Withstand?" at: http://www.alkrug.vcn.com/rrfacts/drawbar.htm
"For the later Grade E couplers I have seen various strengths listed in publications from 330,000 lbs to 500,000 lbs. The current BNSF Air Brake book lists them at 390,000 lbs. An AAR book I have lists them at 360,000 lbs "working strength" and 500,000 lbs "shock load". . . .
Standard Grade E drawbars and knuckles are NOT good for 450,000 lbs!"
He goes on to say that the BNSF's SD70MACs have Tractive Effort meters instead of amp-meters, and that he regularly sees them in the 120,000+ lbs. range per unit, so 3 units are safe for the 360,000 lb. coupler limit. But while the SD40-2s are only good for 90,000 lbs. each, 5 of them are 450,000 lbs. - hence the second quote above.
3. With regard to single-unit steam locomotive operation, and the risk of breakdown with them stranding a train and tying up the line: Recall that many locomotive districts were only 100 or so miles long, and then they were changed out for another freshly fueled and inspected loco. That's a lot less time and territory for things to go wrong before a shop will look at and fix anything needed. Also, a lot less distance for a relief loco and crew to have to go to retrieve a cripple and pull it in, so that would happen sooner and hence minimize the delay time and ripple effects on other trains and the system. While there were extended-distance runs (mainly passenger) on roads such as NYC, ATSF, SP, UP (and others), those were the exception, and then usually only for modern power with a good service record.
Railway Man The problem is this: there are very service designs in the U.S. that economically favor a locomotive consist with 12,000 hp but only 300,000 lbs of continuous tractive effort, as opposed to a locomotive consist with 12,000 hp and 450,000 lbs. continuous tractive effort, or multiples thereof. The 6,000 hp locomotive has been a locomotive in search of purpose and need. If it had one, there would be a lot more incentive to solve their technical problems. RWM
The problem is this: there are very service designs in the U.S. that economically favor a locomotive consist with 12,000 hp but only 300,000 lbs of continuous tractive effort, as opposed to a locomotive consist with 12,000 hp and 450,000 lbs. continuous tractive effort, or multiples thereof. The 6,000 hp locomotive has been a locomotive in search of purpose and need. If it had one, there would be a lot more incentive to solve their technical problems.
Our community is FREE to join. To participate you must either login or register for an account.