Why aren’t more DMUs built as DEMUs?

[GLC]

I was reading about the Voyager Project Thor (the failed plan to add a pantograph to Voyagers) on these forums, and it got me thinking about DEMUs in general. All the way from Sprinters to today, there seem to be way more mechanical DMUs over electric DMUs, but I was wondering why? Electric transmission is more efficient, so it would give either greater power or more efficiency over a mechanical transmission. I gather that an electric transmission system is typically more expensive, when compared to a mechanical one, but (as a complete outsider!), I would have thought a DEMU could have much more in common with a plain EMU (ignoring idiosyncrasies such as bus voltage for the controls etc), which would lead to cost savings.

For BR stuff, I can see the logic that since so many second gen DMUs had pieces cribbed from buses, it would be much easier/cheaper to take an engine/gearbox combo wholesale from Leyland, than try fit an engine into a PEP or a Mk 3 EMU or whatever. For newer stock, especially brand new CAF DMUs, not having an electric drivetrain surely limits future battery/OHLE conversion options, along with the loss of performance. Obviously hypothetical conversions are not paramount when ordering stock, but with a plan to decarbonise the railway in the next 15-20 years, this is well within the lifetime of the DMUs being delivered today.

Is it all just purely down to cost?

 

[hexagon789]

I was reading about the Voyager Project Thor (the failed plan to add a pantograph to Voyagers) on these forums, and it got me thinking about DEMUs in general. All the way from Sprinters to today, there seem to be way more mechanical DMUs over electric DMUs, but I was wondering why? Electric transmission is more efficient, so it would give either greater power or more efficiency over a mechanical transmission. I gather that an electric transmission system is typically more expensive, when compared to a mechanical one, but (as a complete outsider!), I would have thought a DEMU could have much more in common with a plain EMU (ignoring idiosyncrasies such as bus voltage for the controls etc), which would lead to cost savings.

For BR stuff, I can see the logic that since so many second gen DMUs had pieces cribbed from buses, it would be much easier/cheaper to take an engine/gearbox combo wholesale from Leyland, than try fit an engine into a PEP or a Mk 3 EMU or whatever. For newer stock, especially brand new CAF DMUs, not having an electric drivetrain surely limits future battery/OHLE conversion options, along with the loss of performance. Obviously hypothetical conversions are not paramount when ordering stock, but with a plan to decarbonise the railway in the next 15-20 years, this is well within the lifetime of the DMUs being delivered today.

Is it all just purely down to cost?

Electric transmission isn't always more efficient, the Danes refurbished their IC3 DMUs with new engines and a brand-new mechanical transmission sone years back which offers a 95% transmission efficiency (electric and hydraulic are typically ~85% on most modern installations).

 

[norbitonflyer]

A diesel electric power train takes up more space, which makes its use in a multiple unit tricky. It also seems that, at least historically, DEMU designs tended to go for one big engine and generator rather than lots of smaller ones, which meant it wouldn't fit under the frames but had to take up space at the expense of passenger accommodation. Look at the Southern DEMUs, and also the very limited choice of engines the designers of the Class 210 had because of the requirement for a through gangway. I think the Voyager/Meridian family (Class22x) were the first to fit a DE power train under the floor, and only Vivarail have tried it since.

 

[hexagon789]

I think the Voyager/Meridian family (Class22x) were the first to fit a DE power train under the floor

British Rail did so with a two-car trial railcar set in 1956. Two underfloor 450hp Paxman engines (one per car) and electric transmission, design speed was 90mph.

 

[edwin_m]

22x manage to fit an engine, alternator, traction electronics and body-mounted motors in the space between the bogies, admittedly on a 23m rather than 20m vehicle but one with a tilt profile that also reduces the available space underfloor. 80x do the same without the tilt profile but including whatever extra kit is needed to meet later emissions standards - although they had to raise the floor and I don't recall if the motors are body-mounted. So I don't think lack of space can be cited as a reason, particularly for a commuter-type unit that would need a smaller engine and where the motors could go on the bogies.

I suspect it's historically been mainly down to cost, but the answer would probably be different if anyone ordered any more DMUs today. With a strong likelihood of a hard deadline to take them out of service before their technical life expiry, the manufacturers and leasing companies would most likely go for a bi-mode, or at least a DMU with design provision to be converted later into a bi-mode or an EMU or even a hydrogen or battery hybrid.

 

[D365]

I think the Voyager/Meridian family (Class22x) were the first to fit a DE power train under the floor, and only Vivarail have tried it since.

Correct me if I'm wrong, but I think Hitachi have pulled it off pretty successfully

 

[JamesT]

I suspect it's historically been mainly down to cost, but the answer would probably be different if anyone ordered any more DMUs today. With a strong likelihood of a hard deadline to take them out of service before their technical life expiry, the manufacturers and leasing companies would most likely go for a bi-mode, or at least a DMU with design provision to be converted later into a bi-mode or an EMU or even a hydrogen or battery hybrid.

You'd think that, but the recent orders of the 195/196/197 classes suggest that deadline wasn't particular high up the specification list.

 

[Domh245]

but the answer would probably be different if anyone ordered any more DMUs today. With a strong likelihood of a hard deadline to take them out of service before their technical life expiry, the manufacturers and leasing companies would most likely go for a bi-mode, or at least a DMU with design provision to be converted later into a bi-mode or an EMU or even a hydrogen or battery hybrid.

You would have hoped so, but almost all the DMUs ordered recently (barring the 80x) have been the mechanical transmission CAFs, with rather limited scope for conversion!

 

[delt1c]

British Rail did so with a two-car trial railcar set in 1956. Two underfloor 450hp Paxman engines (one per car) and electric transmission, design speed was 90mph.

Would love to some pics and learn more

 

[Energy]

Correct me if I'm wrong, but I think Hitachi have pulled it off pretty successfully

Hitachi went for an engine in only some carriages although it is under floor.

 

[edwin_m]

You'd think that, but the recent orders of the 195/196/197 classes suggest that deadline wasn't particular high up the specification list.

You would have hoped so, but almost all the DMUs ordered recently (barring the 80x) have been the mechanical transmission CAFs, with rather limited scope for conversion!

The 195s and 196s were ordered before there was any commitment to abandon diesel, but do demonstrate that the economics still favoured mechanical transmissions fairly recently. Perhaps, like renewable energy, a government nudge was (and maybe still is) needed to point the industry towards developing a convertible design that is also price competitive.

I suspect Abellio saw the way the wind was blowing when they ordered the FLIRTs, which are the other convertible design.

Hitachi went for an engine in only some carriages although it is under floor.

True but not particularly relevant - like Bombardier they managed to accommodate a diesel-electric powertrain without putting any of it above floor level and with the forthcoming 810 they will have a version that matches the best available diesel performance and includes a bi-mode capability in a 24m bodyshell. If it's possible with a 125mph unit then it's also possible with a more run-of-the-mill unit that doesn't need as much power.

 

[Energy]

I suspect Abellio saw the way the wind was blowing when they ordered the FLIRTs, which are the other convertible design.

Although they ordered Flirts before they ordered the 196s for WMR, I imagine they got the 196s cheaply though.

 

[Bletchleyite]

The 195s and 196s were ordered before there was any commitment to abandon diesel, but do demonstrate that the economics still favoured mechanical transmissions fairly recently. Perhaps, like renewable energy, a government nudge was (and maybe still is) needed to point the industry towards developing a convertible design that is also price competitive.

To be fair, conversion of DMUs is possible, you just need something that produces rotation and a gearbox that turns it into the correct speed/torque for the transmission. Fitting a pantograph would be fairly hefty work due to the bodywork changes needed, but you certainly could fit some sort of battery or hydrogen arrangement, provided it went below the floor on the same type of raft.

 

[hwl]

Electric transmission isn't always more efficient, the Danes refurbished their IC3 DMUs with new engines and a brand-new mechanical transmission sone years back which offers a 95% transmission efficiency (electric and hydraulic are typically ~85% on most modern installations).

The transmission efficiency of DHMUs and DMMUs is very speed dependent unlike electric. While the Danes might have been able to get the transmission efficiency up to 95% at higher speeds they certainly won't be getting that at lower speeds where the torque converter is used (circa 17mph on UK mech transmission gearboxes for 100mph max speed (effectively the first gear) and typically up to 60-65% of max speed on Hydraulic transmissions). Hence Electric Transmission has a useful advantage for stopping services with comparatively more low speed running and acceleration and the potential for regenerative braking and battery storage /boost.

 

[hexagon789]

While the Danes might have been able to get the transmission efficiency up to 95% at higher speeds they certainly won't be getting that at lower speeds where the torque converter is used

I believe it is fully mechanical without a Torque Converter - it's described as a 12-speed mechanical gearbox, the original transmission was 6-speed iirc

Hence Electric Transmission has a useful advantage for stopping services with comparatively more low speed running and acceleration and the potential for regenerative braking and battery storage /boost.

I don't deny that, I was just trying to illustrate that mechanical transmission isn't woefully inefficient compared to electric etc

 

[Bletchleyite]

I believe it is fully mechanical without a Torque Converter - it's described as a 12-speed mechanical gearbox, the original transmission was 6-speed iirc

I'd be surprised if it didn't have a fluid coupling for starting out, though. (This is slightly different from a torque converter but follows similar principles).

 

[AM9]

I believe it is fully mechanical without a Torque Converter - it's described as a 12-speed mechanical gearbox, the original transmission was 6-speed iirc
I don't deny that, I was just trying to illustrate that mechanical transmission isn't woefully inefficient compared to electric etc

I think that HWL is saying that a mechanical transmission would likely be quite inefficient over the range of speeds/loads that most trains are expected to operate with in normal service.

 

[Bletchleyite]

I think that HWL is saying that a mechanical transmission would likely be quite inefficient over the range of speeds/loads that most trains are expected to operate with in normal service.

Why would that be the case? You just need enough ratios.

 

[AM9]

I'd be surprised if it didn't have a fluid coupling for starting out, though. (This is slightly different from a torque converter but follows similar principles).

12 gears would mean that the engine is working in a very limited power-band requiring frequent changes. If it didn't have a torgue converter, the frequency of those changes could reduce its reliability compared to a conventional 6-speed type.
That's the beauty of an electric transmission, i.e. full torque available a very low (zero) speeds and no mechanical parts other than a rotating three phase generator* and conventional EMU type bogie-mounted three phase induction/pm motors.
* even trains with fully mechanical or hydraulic transmission require rotary generators for control/emergency/hotel purposes.

 

[hwl]

I believe it is fully mechanical without a Torque Converter - it's described as a 12-speed mechanical gearbox, the original transmission was 6-speed iirc




I don't deny that, I was just trying to illustrate that mechanical transmission isn't woefully inefficient compared to electric etc.

Mech is certainly better than Hydraulic for most stopping and semi fast services IF you have the neutral option for coasting which not all of them do.

 

[AM9]

Why would that be the case? You just need enough ratios.

12 gears is approaching the flexibility of a CVT gearbox but at increased mechanical complexity and more frequent operation. It is still a far more imperfect transformer from a limited power band IC engine to zero to full speed axle rotation, and of course, it prevents any evolution from hydrocarbon energy to (probably) renewable electricity. Even before that, all deceleration has to be from friction, with no possibility of energy reclaimation - except to heat the train.

 

[hexagon789]

I'd be surprised if it didn't have a fluid coupling for starting out, though. (This is slightly different from a torque converter but follows similar principles).

I'll look up the source again, but it was in Danish and you always lose something using Google Translate!

I think that HWL is saying that a mechanical transmission would likely be quite inefficient over the range of speeds/loads that most trains are expected to operate with in normal service.

Wouldn't that depend on the set-up and the number of gears and their ratios? I'd have thought that a 12-speed would offer the ability to closely match the ideal torque curves of the engine?

Mech is certainly better than Hydraulic for most stopping and semi fast services IF you have the neutral option for coasting which not all of them do.

I'll try and find out if this does, but see my reply to Bletchleyite RE translation.

 

[hwl]

Why would that be the case? You just need enough ratios.

The big issues is efficiency @ low speeds e.g. <17mph with mech or < 40-55mph with most hydraulic, i.e. where the first gear is the torque converter. Above those speeds the efficiency is good.
Gear changes also dent efficiency, 6 is fine to cover higher speeds efficiently. You would need a lot of ratios to cover the low speed instead of a torque converter.

 

[Bletchleyite]

I'll look up the source again, but it was in Danish and you always lose something using Google Translate!

FWIW the difference between a fluid coupling and a torque converter is that the former just provides a level of slip so the engine doesn't stall on starting out (so a bit like a car's clutch but without the friction surfaces to wear out), whereas the latter actually magnifies torque. This difference is caused by the shape of the vanes on the rotating parts in the oil.

 

[hwl]

12 gears would mean that the engine is working in a very limited power-band requiring frequent changes. If it didn't have a torgue converter, the frequency of those changes could reduce its reliability compared to a conventional 6-speed type.
That's the beauty of an electric transmission, i.e. full torque available a very low (zero) speeds and no mechanical parts other than a rotating three phase generator* and conventional EMU type bogie-mounted three phase induction/pm motors.
* even trains with fully mechanical or hydraulic transmission require rotary generators for control/emergency/hotel purposes.

Looking at the efficiency of the transmission alone can be a bit of a red herring as it discounts auxiliary which have grown some what over time. As soon as you have modern electrical system (e.g. Stadler Flirts) and have a single alternator (compared to 3 at different alternators / voltages for older traction e.g. Voyager (Traction 3ph AC, 415AC "Hotel", 110DC for battery charging)) and electrically powered auxiliaries (cooling fans, pumps with computer control) and regenerative braking then the total system efficiency for electric transmission should be quite good.

 

[hexagon789]

FWIW the difference between a fluid coupling and a torque converter is that the former just provides a level of slip so the engine doesn't stall on starting out (so a bit like a car's clutch but without the friction surfaces to wear out), whereas the latter actually magnifies torque. This difference is caused by the shape of the vanes on the rotating parts in the oil.

Still can't find anything definitive but the following seems to be the case:

1. The original transmission was 5-speed (my mistake, I was slightly wrong)
2. The new transmission (fitted from 2006) is 12-speed
3. There is no retarder
4. The units have an EM track-brake
5. The transmission does not disengage when running with the traction power shut off
6. Efficiency was 85% with the original transmission
7. DSB wanted 95% efficiency with the new transmission
8. In practice it was found to be 94%
9. Losses for heating, lighting and air-con plus other auxiliaries are 2%

Nothing about a fluid coupling etc but there is a reference to a clutch(?).

 

[59CosG95]

I'd say another infrastructure-based reason is electro-magnetic interference (albeit not so much of a problem as it is for pure EMUs); the debacle with the 755s causing lighting on rural routes to go haywire stands to that. AC motors seem to be much more guilty of this than DC ones.
Case in point: 66s are DC-motored locos, while 92s are 3-phase AC motored locos with GTO inverters, which are infamously "messy" in terms of EMF. IGBTs are less "messy" but still not as clean, AIUI, as DC motors (but more efficient).

Hydraulic transmission units, meanwhile - no such problem.

 

[edwin_m]

Nothing about a fluid coupling etc but there is a reference to a clutch(?).

Maybe they've found a clutch that can work reliably at the ratings they need, and eliminating the fluid coupling gives the required efficiency.

 

[hexagon789]

Maybe they've found a clutch that can work reliably at the ratings they need, and eliminating the fluid coupling gives the required efficiency.

I've no idea, unless someone can find the specs on the transmission - "ZF AS 12-Speed (Rail)".

 

[edwin_m]

I'd say another infrastructure-based reason is electro-magnetic interference (albeit not so much of a problem as it is for pure EMUs); the debacle with the 755s causing lighting on rural routes to go haywire stands to that. AC motors seem to be much more guilty of this than DC ones.
Case in point: 66s are DC-motored locos, while 92s are 3-phase AC motored locos with GTO inverters, which are infamously "messy" in terms of EMF. IGBTs are less "messy" but still not as clean, AIUI, as DC motors (but more efficient).

Hydraulic transmission units, meanwhile - no sud ch problem.

This issue usually arises because electric trains draw current through the overhead line or third rail and in particular return it through the running rails. Thus there is a current of hundreds or thousands of amps going through the rails where the track circuits are trying to detect trains using a current of about one amp. As well as the DC or AC supply current this contains small components at the frequency of the motors themselves, which for AC motors varies with speed, and particularly with GTOs sits in the same frequency ranges used by track circuits. Thus there is the risk of the return current causing the track circuit to fail to detect a train (or more correctly to detect the absence of a train when one is present).

However this is normally only an issue for electric trains, because in a diesel, even with three phase drive, the current is in cables between the alternator, the traction electronics and the motor, so should never get into the track.

I'm not sure what the exact problem was with 755s - was it only when in electric mode, or possibly some sort of inductive coupling from cables within the train?