All I can say is “God bless the Woodward Governor Company!”
One last point the quoted horsepower figure for a locomotive is a conversion of the electric power output from the main generator (alternator), the Brake Horsepower rating of the diesel engine itself is higher. In otherwords an SD60 is producing a little over 2.8MW at full throttle.
Yes, it makes a lot of sense.
Just want to add for anybody pondering these questions: Drawbar horsepower = tractive force x train speed. This must be equal to driver (driving wheels) horsepower = driver torque x driver rpm. This in its turn (forgetting some losses) is translated to the prime mover (pm) horsepower = pm torque x pm rpm. It is the locomotive control system’s job to set the prime mover horsepower at any moment to what is desired at the drawbar (increase/decrease tractive force to change train speed or to maintain constant speed regardless grade). Both torque and rpm are adjusted in low throttle settings. In higher up the pm rpm is kept constant and only the torque adjusted metering fuel and air.
The above ignores the traction motors for simplicity as mere intermediates. Their torque and rpm follow, for any drawbar condition, the corresponding driver values adjusted by the gear ratio: product of torque and rpm remains unchanged. (I admit that this is a simplification, gear losses should be taken into account.)
Conversions from the drawbar values to rotating axle values become simple, if you use watt, newton meter and revolutions per second instead of horsepower, lb ft and rpm (newton meter = watt second).
This analysis of drawbar horsepower (tractive effort times train speed equals power of the “prime mover”) works for a steam locomotive, but not for a diesel electric because at starting, just before the train moves, thediesel-electric’s prime mover may be generating significant horsepower but the drawbar horsepower would be zero. It is only once the diesel electric has accelerated the train to the speed where the traction motors can convert all the horsepower of the prime mover into useful work that the drawbar horsepower equals (neglecting conversion losses) the prime mover horsepower. At very low speed much of the prime mover’s horsepower could be wasted as heat generated in the traction motors and the overall efficiency of the locomotive could be quite low.
Normally, an engineer would limit this inefficiency by not notching up the speed of the diesel prime mover until the train had begun to accelerate and the current to the traction motors had begun to drop. Newer locomotives do all this automatically, adjusting the engine speed to obtain maximum fuel efficiency. The early diesel electric locomotives such as the EMD FT were all manual machines. The engineer had a load meter which was essentially an ammeter marked with variously colored zones showing how long the traction motors could be operated at any current rating. It was up to the engineer to make sure nothing got fried.
I once had an operating manual for the EMD F3, which had the same kind of controls as the FT. There were very explicit instructions for the engineer to excercise caution at low speeds and prolonged operation at high tractive effort (high current) levels so as not to damage traction motors. This condition would most likely occur when starting a heavy train on an adverse grade. The locomotive might be able to start the train by working the traction motors in the time-limited zone but if the train could not be accelerated so as to drop the current within safe contin
I don’t understand (1) how AC-traction “locks all the axles together at the same speed” and (2) how this would maximize tractive effort under difficult rail conditions unless each axle were subject to the exact same rail conditions.
How AC induction motors “lock all the axles together” is a little difficult to understand at first, but not after a bit of explanation.
We’re all familiar with the characteristics of a permanent magnet DC motor, which behaves a lot like an ordinary direct current traction motor. With no load the motor turns very fast with the no-load top speed limited to the point where the back emf (the voltage the motor is generating whenever it is turning, a voltage that opposes the voltage that is making the motor turn) nearly balances the forward emf applied by the power source. The only difference is the relatively small amount of current, and therefore power, required to overcome bearing and windage losses. If we load the motor, it will slow down quite a bit in order to “pull the load”, which it does because the reduction in speed means there is less back emf available to counteract the applied voltage and the current draw goes up. It is this increased current multiplied by the applied voltage that is the power to pull the increased load. The more load is applied, the slower the motor turns and the more current it draws until the motor finally stalls. That represents the maximum torque the motor can generate. But it won’t generate this high torque for too long or it will burn up due to too much current through the windings. Remember, at stall the current is not limited by the back emf because there is none. It is limited only by the ohmic resistance in the windings.
This means that the load characteristic of a DC motor is not very “stiff.” It takes a relatively large change in speed to make a relatively small change in current draw and thus power and torque. On a diesel-electric with DC transmission, each axle has its own motor. If one axle begins to slip due to uneven weight distribution or a slippery spot on the rail, it will speed up quickly with the small reduction in load due to the slipping. Once it really starts to slip its abili
So, basically, all of the axles turn at the same speed until it’s necessary for the speed of a slipping axle to change in order to correct the slip by reducing the torque. I understand that concept.
And I understand the wheel-diameter issue in relation to EMD units. However I don’t understand why wheel diameters would have to be matched exactly on GE units since they have a separate inverter for each axle.
Jay, you have a good point. GE’s latest AC transmissions utilize individual inverters for each axle for exactly the reason you state. Each axle’s inverter operates at the proper frequency to match that axle’s drive wheel diameter. This eliminates drive wheel diameter variation as a problem. Even though each inverter may operate at a slightly different frequency, the effect is the same as I described formerly.
GE’s latest control scheme for the inverters incorporates smart wheel slip control with each inverter changing its frequency on a real time basis under microprocessor control to minimize wheel slip. This gives the best possible adhesion factor under all track conditions and is reflected in the very high tractive effort these locomotives are capable of delivering.
EMD’s AC drive technology uses a single inverter per truck, so each inverter powers two or three axles, but both inverters operate at the same frequency. This system reflects the simpler design I described earlier, the inverter doesn’t have to change frequency when an axle suffers wheel slip. The frequency stays the same, the slipping axle merely speeds up a bit which lessens the slip between the rotating magnetic field and the motor rotor which automatically reduces the torque and limits the wheel slip. It doesn’t eliminate the slip, merely limits it to a very small value where tractive effort doesn’t suffer much. The missing load due to the slipping axle is shifted to the other axles, with their stator/rotor slip increasing just enough to increase their torque to take up the load shed by the slipping axle. When the slippery axle regains its footing the whole system equalizes out again automatically with no need to change the frequency.
In GE’s advanced scheme, the control system can detect if an axle begins to slip. The ultimate in control allows a slipping axle to reduce power while operating at the same frequency (this is done by modifying the voltage sen
You are right that in starting from a standstill the prime mover power is at first not shown as traction motor power output but consumed as heat loss. I felt that for the sake of clarity I can lump that together in what I expressed “(forgetting some losses)”. My argument is that the moments when the prime mover power goes mostly to heating traction motors (+ some other losses) represent a small fraction of a locomotive’s working hours. Once the locomotive starts gaining speed and higher voltage has to be applied to counter the increasing back emf of the traction motors, more power is drawn from the prime mover. This increase, however, goes to useful work and the losses’ share of the total diminishes. DC series motor represents a very low ohmic resistance at a standstill. Therefore only a low voltage is needed to drive through it enough current to produce the torque and thus the tractive force allowed by the adhesion. So only a fraction of the prime mover’s full horsepower can be applied at first (and consumed as heat). How significant the loss power is (absolute and relative to the useful power) during the cycl
[quote user=“Alan Robinson”]
How AC induction motors “lock all the axles together” is a little difficult to understand at first, but not after a bit of explanation.
We’re all familiar with the characteristics of a permanent magnet DC motor, which behaves a lot like an ordinary direct current traction motor. With no load the motor turns very fast with the no-load top speed limited to the point where the back emf (the voltage the motor is generating whenever it is turning, a voltage that opposes the voltage that is making the motor turn) nearly balances the forward emf applied by the power source. The only difference is the relatively small amount of current, and therefore power, required to overcome bearing and windage losses. If we load the motor, it will slow down quite a bit in order to “pull the load”, which it does because the reduction in speed means there is less back emf available to counteract the applied voltage and the current draw goes up. It is this increased current multiplied by the applied voltage that is the power to pull the increased load. The more load is applied, the slower the motor turns and the more current it draws until the motor finally stalls. That represents the maximum torque the motor can generate. But it won’t generate this high torque for too long or it will burn up due to too much current through the windings. Remember, at stall the current is not limited by the back emf because there is none. It is limited only by the ohmic resistance in the windings.
This means that the load characteristic of a DC motor is not very “stiff.” It takes a relatively large change in speed to make a relatively small change in current draw and thus power and torque. On a diesel-electric with DC transmission, each axle has its own motor. If one axle begins to slip due to uneven weight distribution or a slippery spot on the rail, it will speed up quickly with the small reduction in load due to the slipping. Once it re
As a rough order of magnitude estimate, the losses in the electrical transmission of a diesel-electric are on the order of 10% or so. The best units have a bit lower losses and older units would have losses quite a bit higher. The magnitude of the loss also depends on the exact operating point at which it is measured. For example, before the train starts moving the losses are essentially 100%. The number of around 10% would represent the loss when the locomotive was operating at it’s best efficiency point. These losses are primarily I squared R losses due to the resistance in the windings of the traction motors and the alternator/generator and iron losses in the magnetic circuits of the traction motors and the alternator/generator.
Newer units have lower losses partially because they have more sophisticated computerized control schemes. Older units would have higher losses and those losses could be accentuated by sloppy operation. This could be caused by an engineer not paying attention to the points at which he notched up or down the throttle or at what points he made transitions from one type of traction motor connection to another. This would be the equivalent of making inappropriate gear selections in a manual transmission automobile which we know can have a substantial effect on fuel efficiency. Since the days of the F7/GP7 the transistions were automated, but the early automation systems were fairly clumsy. The newer microprocessor based systems are considerably more efficient not to mention more reliable.
The newest units have such low losses that it is difficult to imagine where else there is room for efficiency improvements in the electric drive system. Some minor efficiency improvements may be made to traction motors and alternators/generators. Further major efficiency improvements will almost certainly come from improvements in the thermal efficiency of the prime mover (not an easy thing to do) or in recapturing braking energy otherwise wasted via means of hybrid drive systems so tha
One area for improvement in transmission efficiency is replacing the induction motors with synchronous motors, and this would be most noticeable at low speeds where the relative difference between rotational and synchronous speeds are greatest in the induction motor.
There’s some room for improvement in the inverters, a major source of loss in inverters is reverse recovery charge in the free-wheeling diodes. This can be almost eliminated by replacing those diodes with SiC Schottkey diodes.
I also agree with the comment earlier in this thread that AC traction motors and inverters will, within a few years, be cheaper than DC traction motors. The trade-in market may keep DC motors around for a while.
You are right about synchronous AC motors offering some advantages for rain traction applications, and they are being used in some subway and commuter equipment in Japan, at least. But the squirrel cage induction motor is pretty hard to beat for plain old ruggedness and high power in a small package. One advantage of the synchronous design is that it isn’t so important for motor efficiency for very small gaps between rotor and stator to be used. This means that synchronous motors are somewhat easier to build and keep well maintained in terms of mechanical tolerances. I wonder how long we will wait to see a synchronous drive on a diesel electric in this country?
Inverters will show small incremental improvements in efficiency, but again, when efficiencies already run better than 95%, it becomes ever more difficlult to squeeze out any part of those remaining few percent. I think where we will primarily see advances is in power capability, robustness under shock and vibration loads, surge capability, reliability and ease of servicing. This is also true of the inverter applications in other industrial fields.
I hope I didnt inadverdently start a flame war. I did, however, learn some good info on how wheel slippage is prevented and quite a bit other stuff.
[censored]
Ok guys, I’m a real novice here and hope someone can answer a question for me? Is tractive effort a calculated value or is it determined by some kind of testing?
I’m learning a lot here and appreciated the posts,
Wayne
Depends what you mean by “tractive effort”. You might see “tractive effort” listed for each model in some roster book-- it might give the weight for each unit and then divide that by 4 and call the result the unit’s “tractive effort”. Obviously the unit’s actual capability won’t be that easy to predict.
Ok, thanks for the info. Gonna have to go back and re-read this very interesting post. And I thought quantum mechanics was hard…this is way more complicated than ever imagined.
Wayne
I suppose the answer is ‘both’ - at low speeds maximum tractive effort is basically limited by the mechanical design of the trucks and the wheelslip control system, so really this has to be measured e.g. by putting a test car between the locomotive and the train to measure the coupler/drawbar force (=tractive effort). At higher speeds it’s power limited, so if the transmission efficiency etc is known it can be calculated fairly accurately.
Modern AC drive locos know how much tractive effort is being generated - they can generate such high tractive effort that it has to be limited in some situations e.g. DPU operation so that the coupler/drawgear limits are not exceeded. Jay Potter’s article in Trains recently about the ‘heavy’ AC locos on CSX was quite interesting…
Tony
Tony,
Thanks for the explanation, this makes sense to me. I never thought about the engineering complexity of either steam or diesel locomotives and this posting has been very interesting to me. I’ll look for Jay Potter’s article in Trains, it sounds interesting.
Wayne