Hydraulic Transmission

[computerSaysNo]

I’ve been reading up and trying to understand the principles and advantages/disadvantages of hydraulic transmissions such as the Voith T211, as used in Sprinter and Turbostar classes, amongst others.
The first stage of the transmission is the torque convertor. When there is a big difference between the engine output speed and the final drive speed (i.e. wheel speed), it converts engine power into torque allowing for better acceleration. From reading this article, it seems that torque convertors can only multiple the original engine torque by two to three times:

In addition to the very important job of allowing your car come to a complete stop without stalling the engine, the torque converter actually gives your car more torque when you accelerate out of a stop. Modern torque converters can multiply the torque of the engine by two to three times. This effect only happens when the engine is turning much faster than the transmission.

My first question is; do manual-style gearboxes for road vehicles multiply the original engine torque by more than 2-3 times? If so, why was a fluid coupling chosen for rail vehicles instead of a road vehicle-style box?
My second question relates to trains travelling at higher speeds. The article says:

At higher speeds, the transmission catches up to the engine, eventually moving at almost the same speed. Ideally, though, the transmission would move at exactly the same speed as the engine, because this difference in speed wastes power. This is part of the reason why cars with automatic transmissions get worse gas mileage than cars with manual transmissions.

To counter this effect, some cars have a torque converter with a lockup clutch. When the two halves of the torque converter get up to speed, this clutch locks them together, eliminating the slippage and improving efficiency.

My question is, why was a second transmission stage (fluid coupling) chosen for rail vehicles, instead of a simple lockup clutch?


The second part of my questions relates to the fluid coupling; this takes over from the torque convertor at higher speeds. However, this article says that, in a fluid coupling, there will always be “slip” (i.e. a difference in input and output speeds). How does this manifest in rail vehicles? As I’ve read that in this stage of the transmission the engine speed is directly proportional to the speed of the train regardless of the load on the engine, whereas I would have thought that at higher loads the slippage would be greater and so the engine RPM would be increased?

Thanks in advance for any replies (there probably will be follow-up questions!).

 

[Taunton]

If so, why was a fluid coupling chosen for rail vehicles instead of a road vehicle-style box?

The 1960s Modernisation Plan dmu, plus all the smaller shunters like Class 03, had just this, it was the gearbox out of trucks and buses (especially the latter) of the era, built by a company called Self Change Gears in Coventry, sometimes called the "Wilson" gearbox after its inventor. There was no separate clutch, but you did need to close the throttle, change gear manually, and open the throttle again.

In later times I was struck by how reliable such an arrangement was over the multiple-unit control wires, which I never experienced any troubles with. The device for changing direction, which of course road vehicles do not need, was not as robust.

 

[matchmaker]

The 1960s Modernisation Plan dmu, plus all the smaller shunters like Class 03, had just this, it was the gearbox out of trucks and buses (especially the latter) of the era, built by a company called Self Change Gears in Coventry, sometimes called the "Wilson" gearbox after its inventor. There was no separate clutch, but you did need to close the throttle, change gear manually, and open the throttle again.

In later times I was struck by how reliable such an arrangement was over the multiple-unit control wires, which I never experienced any troubles with. The device for changing direction, which of course road vehicles do not need, was not as robust.

A small correction to the above - the "Wilson" gearbox was a preselector type with a gear change pedal. The Self Changing Gears as fitted to locos and DMUs was a semi-automatic type, similar to the Pneumocyclic fitted to Leyland buses. Wilson was one of the founders of SCG, although it was taken over by Leyland in the 1950s.

 

[cadder toad]

This is only a partial answer as I can't remember(if I ever knew!) and I'm very much open to correction. Hopefully it will encourage a better answer from someone who does know. I seem to remember the Ladybird Book of Trains having some good pictures to explain some of this.

1. Car gearboxes multiply the torque by a factor of four I would guess, reducing the speed similarly. (Engine power = torque x rotational speed) The question of using a torque convertor rather than a gearbox is one of cost, weight and reliability. The answer will vary depending on application and the power to be transmitted. Sprinters are around 2-3 times more powerful than a car. Torque convertors are lighter than a gearbox of the same power. Production volumes make gearboxes cheaper and that advantage is lost at higher power ratings, as fewer are made.

2. A torque convertor is a fluid coupling with a variable speed, by changing the oil in circulation. Do sprinters have both a torque convertor and fluid coupling? Its possibly to allow a smooth take up of power rather than starting with a severe jolt, and consequent load on engine bearings and so on.

 

[hexagon789]

torque convertor is a fluid coupling with a variable speed, by changing the oil in circulation. Do sprinters have both a torque convertor and fluid coupling?

Yes, they do. The Sprinter transmission is explained in detail in the Driving Instructions a copy of which is viewable on the locodocs website here:

150/1 150/2 DMMU DMU General Information and Driving Instructions 33056/114 2 1987

Page 17 has a diagram, while pages 18 and 19 provide a detailed description of how it works.

 

[E27007]

There is hyfrostatic drive, found in mainline stock suxh as Tampers, with hydrostatic drive a diesel engine powers a hydraulic pump to pressurise hydraulic oil, the pressurised oit is directed into hydraulic motors to turn the wheels, a feature of hydrostatic is the diesel engine runs at 90 to 100% of engine rpm when the ubit is above walking pace , oterwise a risk of danage to the main pump.

 

[Irascible]

My first question is; do manual-style gearboxes for road vehicles multiply the original engine torque by more than 2-3 times? If so, why was a fluid coupling chosen for rail vehicles instead of a road vehicle-style box?
My second question relates to trains travelling at higher speeds. The article says:

My question is, why was a second transmission stage (fluid coupling) chosen for rail vehicles, instead of a simple lockup clutch?


The second part of my questions relates to the fluid coupling; this takes over from the torque convertor at higher speeds. However, this article says that, in a fluid coupling, there will always be “slip” (i.e. a difference in input and output speeds). How does this manifest in rail vehicles? As I’ve read that in this stage of the transmission the engine speed is directly proportional to the speed of the train regardless of the load on the engine, whereas I would have thought that at higher loads the slippage would be greater and so the engine RPM would be increased?

Thanks in advance for any replies (there probably will be follow-up questions!).

Torque converters ( fluid couplings too ) are extremely simple devices - other than the bearings on the shafts, there isn't really much else, and they also provide very smooth transitions. You can hear the characteristic dip in revs when a unit moves away from standstill, that's the oil filling the torque converter - there's no wear from slipping the transmission as you would get with a friction clutch, and no jerking like you'd get from a dog gear. That's pretty much the answer to "why a fluid coupling" also, plus you just drain oil from one side & stick it in the other to "change gear". Excess energy put in in all of them that isn't put into the wheels is going to just heat the oil up. At higher loads the relative speed of the engine vs wheels will be higher, it doesn't mean the engine revs higher.

 

[2192]

Does a unit fitted with a gear box do more miles per gallon than a similar unit fitted with hydraulic transmission?

 

[hexagon789]

Does a unit fitted with a gear box do more miles per gallon than a similar unit fitted with hydraulic transmission?

On stop-start stuff mechanically transmission is generally more efficient, on longer distance runs at speed with long distances between stops hydraulic can be more efficient because of the ability to freewheel for considerable periods. (I believe that was the conclusion drawn between a conventional hydraulic 159 and the unit experimentally fitted with a ZF gearbox.)

 

[gabrielhj07]

A further question:

What is the reason for different units taking different approaches to fluid transmissions? I believe the class 180, for example, has a three-stage transmission, whereas the class 158 has one stage of fluid transmission before the direct drive kicks in (I think a lockup clutch of sorts).

 

[hexagon789]

A further question:

What is the reason for different units taking different approaches to fluid transmissions? I believe the class 180, for example, has a three-stage transmission, whereas the class 158 has one stage of fluid transmission before the direct drive kicks in (I think a lockup clutch of sorts).

I know 185s don't lock-up (very similar transmission to 180s but different speed ratios), each stage is a torque converter unlike 158s which have a torque converter and a fluid coupling (direct drive) stage.

 

[Tynwald]

180 and 185 have Voith T312. They have a torque converter and 2 fluid couplings. They have hydrodynamic brake as well. 175 have a T211 transmission.

 

[supervc-10]

On stop-start stuff mechanically transmission is generally more efficient, on longer distance runs at speed with long distances between stops hydraulic can be more efficient because of the ability to freewheel for considerable periods. (I believe that was the conclusion drawn between a conventional hydraulic 159 and the unit experimentally fitted with a ZF gearbox.)

I wonder if the ZF transmissions could be made to freewheel? I know a lot of DSG/automated manual type transmissions in cars will freewheel- had a Skoda Octavia with a DSG transmission that would freewheel when in 'eco' mode. They just disengage the clutch when the computer decides that freewheeling is better than engine braking.

 

[gabrielhj07]

I wonder if the ZF transmissions could be made to freewheel? I know a lot of DSG/automated manual type transmissions in cars will freewheel- had a Skoda Octavia with a DSG transmission that would freewheel when in 'eco' mode. They just disengage the clutch when the computer decides that freewheeling is better than engine braking.

That sounds rather unpredictable.

 

[Bletchleyite]

That sounds rather unpredictable.

Trains don't need engine braking (and it isn't that effective with diesels anyway), so a simple freewheel would work fine. These are fitted to all first generation DMUs so it's clearly possible on an epicyclic gearbox.

 

[gabrielhj07]

Trains don't need engine braking (and it isn't that effective with diesels anyway), so a simple freewheel would work fine. These are fitted to all first generation DMUs so it's clearly possible on an epicyclic gearbox.

On a car I meant. On a train I'm sure it would be a great idea.

 

[Bletchleyite]

On a car I meant. On a train I'm sure it would be a great idea.

You get hardly any engine braking effect on a small-engined diesel car anyway, so it's probably not that pronounced. Petrols generally have more engine braking because not only do you have the compression but also the air intake is "throttled" (hence the term "throttle" for the accelerator), so there is a braking effect on two out of four strokes (intake and compression), not only the compression stroke. On a diesel the accelerator just controls the amount of fuel going in.

 

[Irascible]

Trains spend a lot of time coasting, so it's of more benefit being able to freewheel than use the engine for braking. It'd be beneficial if the engines could stop-start more, but then you need to provide extra services for lubrication.

 

[coppercapped]

There are several factors to be taken into account is deciding what type of transmission should be used for a particular application: initial cost; reliability; size; fuel efficiency and so on.

Taking the last point first an engine's best specific fuel consumption (in units such as kg/bhp-hr) is to be found when the engine is running at a speed at or near its maximum torque output. So any transmission which can best match maximum torque engine speed to road speed will give the lowest fuel consumption for the journey. Modern electric transmissions can do this very well and mechanical transmissions with many gear ratios also come close. Hydrodynamic transmissions such as the Voith are efficient as soon as the transmission has shifted to the fluid coupling - when the 'slip' between driving and driven members is minimal - but below that speed the efficiency of the torque converter drops even though the engine torque is still being magnified.

The trial with the ZF gearbox mentioned above took place some ten years ago and demonstrated fuel savings of some 15% on the 'start-stop' section of the Waterloo-Exeter route west of Salisbury and somewhat less on the stretch from Salisbury to Waterloo with only two stops. The reason reported at the time as to why no freewheel was fitted was that it was a time limited trial and because the ZF gearbox was dimensionally different to the Voith box considerable reengineering would have been needed to get it to fit. Modern ZF boxes include an internal freewheel, see the ZF website at https://www.zf.com/products/en/rail/trends/repowering_rail/repowering_rail.html

The reason that railway applications avoid dry-plate clutches is that they wear and need to be replaced from time to time, using hydrodynamic drives with no clutches avoids such down time and maintenance costs even at the cost of marginally lower fuel efficiency.

At the time when the design decisions were being made the hydrodynamic drive ticked all the boxes for railcar use up to about 1000bhp. They were compact, reliable (basically 'fit and forget') and being available in a range of sizes depending on the input torque perfectly adequate for the job.

You get hardly any engine braking effect on a small-engined diesel car anyway, so it's probably not that pronounced. Petrols generally have more engine braking because not only do you have the compression but also the air intake is "throttled" (hence the term "throttle" for the accelerator), so there is a braking effect on two out of four strokes (intake and compression), not only the compression stroke. On a diesel the accelerator just controls the amount of fuel going in.

My 2 litre Golf diesel with a fully automatic mechanical transmission (a DSG) has quite noticeable engine braking. My late father's torque converter automatic had essentially none. As there is no restriction to the airflow into the engine, being permanently connected to the road wheels on overrun the engine sucks in the full quantity of air which is then compressed so absorbing some of the kinetic energy of the vehicle. A petrol engine does not do this, the throttle valve is shut so the air intake tracts are under a partial vacuum and the compressive effects are minimal.

 

[hexagon789]

You get hardly any engine braking effect on a small-engined diesel car anyway, so it's probably not that pronounced. Petrols generally have more engine braking because not only do you have the compression but also the air intake is "throttled" (hence the term "throttle" for the accelerator), so there is a braking effect on two out of four strokes (intake and compression), not only the compression stroke. On a diesel the accelerator just controls the amount of fuel going in.

Engine braking is very pronounced on 172s, so much so I was told they have to be kept under power even going downhill sometimes.

I wonder if the ZF transmissions could be made to freewheel? I know a lot of DSG/automated manual type transmissions in cars will freewheel- had a Skoda Octavia with a DSG transmission that would freewheel when in 'eco' mode. They just disengage the clutch when the computer decides that freewheeling is better than engine braking.

AFAIAA they can, and ZF do offer transmissions with integral "freewheel"

 

[notadriver]

I’ve noticed with electric or the ZF 6 speed transmission, full power can be applied from a stand still or near stand still. Torque converter transmissions require a lower power input often restricted by electronics until a higher speed is reached - between 10 and 20 mph. This becomes noticeable when starting on rising gradients and even more when an engine is out.

 

[hexagon789]

I’ve noticed with electric or the ZF 6 speed transmission, full power can be applied from a stand still or near stand still. Torque converter transmissions require a lower power input often restricted by electronics until a higher speed is reached - between 10 and 20 mph. This becomes noticeable when starting on rising gradients and even more when an engine is out.

The original instructions for Sprinters was start in Notch 1 or 2 to fill the torque converter, then go straight to 7 and release the brake.

Obviously no longer accepted technique, but can definitely take full power from a stand if adhesion allows it.

I believe the 159s however had a modification to prevent full power being given, even if selected, below either 10 or 12mph.

170s have a similar "half-power" restricter below about 10mph.

185s are limited to Notch 4 (out of 5) output to 20mph, due to the large amount of torque produced by the engine.

 

[notadriver]

The original instructions for Sprinters was start in Notch 1 or 2 to fill the torque converter, then go straight to 7 and release the brake.

Obviously no longer accepted technique, but can definitely take full power from a stand if adhesion allows it.

I believe the 159s however had a modification to prevent full power being given, even if selected, below either 10 or 12mph.

170s have a similar "half-power" restricter below about 10mph.

185s are limited to Notch 4 (out of 5) output to 20mph, due to the large amount of torque produced by the engine.

How do you know what 185s are limited to ?

 

[coppercapped]

On starting a diesel powered train (except very low power machines such as the Class 03 shunter) the limiting factor is the adhesion limit and not the power output from the prime mover. So unless there are special conditions, such a starting up an incline, power is always wasted below the point where the adhesion limit crosses the constant power curve. Modern electric transmission control circuits will limit the power supplied from the diesel engine regardless of what the driver demands until the constant power part of the tractive effort-speed curve is reached so saving fuel.

In the absence of engine power control, such as is the case with older designs of hydrodynamic transmissions, applying full power from the start will simply heat up the oil - which is not very economical!

I don't have any figures to hand for DMUs, but for some modern locomotives the adhesion limit crosses the constant power curve at about 5mph for the Class 66, 10mph for the Class 68, 35mph for the Class 93 on 25kV (10mph in hybrid mode) and 10mph for the Class 91. Incidentally above about 32-33mph the Class 91 puts down more tractive effort than the Class 66, at 75mph it's about double. (These figures are from a ROG's presentation to an IMechE meeting last year).

So the conclusion is that it is wasteful to apply full power at speeds below the point where the adhesion limit crosses the constant power curve.

 

[supervc-10]

That sounds rather unpredictable.

Off topic (and I'm to blame- sorry mods) but it's quite cleverly done in the VW Group cars. If you take your foot completely off the accelerator it engine brakes, but a tiny throttle application will coast.

The trial with the ZF gearbox mentioned above took place some ten years ago and demonstrated fuel savings of some 15% on the 'start-stop' section of the Waterloo-Exeter route west of Salisbury and somewhat less on the stretch from Salisbury to Waterloo with only two stops. The reason reported at the time as to why no freewheel was fitted was that it was a time limited trial and because the ZF gearbox was dimensionally different to the Voith box considerable reengineering would have been needed to get it to fit. Modern ZF boxes include an internal freewheel, see the ZF website at https://www.zf.com/products/en/rail/trends/repowering_rail/repowering_rail.html

Interesting to hear! Thank you. I wonder if a more long-term plan for 158/159s would incorporate a freewheel mode. 15% is nothing to be sniffed at, and I wonder if the 158 operators like GWR operating some more stop/start type routes would be interested too.

 

[Tynwald]

so they have a torque convertor. take the oil out of it and you disengage the transmission.

 

[notadriver]

The ZF transmission definitely provides greater acceleration and economy over a hydraulic one but it does seem to hunt between the gears at lower speeds in a way a Bus or coach with that same transmission doesn’t. Can anything be done to lessen this behaviour ?

 

[hexagon789]

How do you know what 185s are limited to ?

It's in the traction manual.

The ZF transmission definitely provides greater acceleration and economy over a hydraulic one but it does seem to hunt between the gears at lower speeds in a way a Bus or coach with that same transmission doesn’t. Can anything be done to lessen this behaviour ?

Extend the changeover speed ranges further?

 

[coppercapped]

Off topic (and I'm to blame- sorry mods) but it's quite cleverly done in the VW Group cars. If you take your foot completely off the accelerator it engine brakes, but a tiny throttle application will coast.



Interesting to hear! Thank you. I wonder if a more long-term plan for 158/159s would incorporate a freewheel mode. 15% is nothing to be sniffed at, and I wonder if the 158 operators like GWR operating some more stop/start type routes would be interested too.

These days any conversion is more likely to be to a diesel/battery hybrid. In any event it is unlikely as the remaining life of the 158s is unlikely to be long enough to make financial sense as the fuel cost savings have to be offset against the costs of the re-engineering design work and the actual conversion. The trains are already 30 years old.