Train Gearing and Maximum Speeds

[mr_jrt]

Context:
I was having a ponder about a reply to a thread on the MML recently, and it got me thinking about the situation with Thameslink where it runs on 125mph fast lines, but the 700s have a max speed of 100mph, costing capacity. Short of some pretty impressive infrastructure interventions, the only way to avoid impacting the capacity on the fast lines is to stay off them entirely, or downrate them to the fastest common speed. In Thameslink's case, I was also wondering about the Peterborough services, which as a trade-off run with 110mph stock into the surface platforms at Kings Cross, and their equivalent on the MML, the EMT Connect services which also run with 110mph stock into the surface platforms at St. Pancras. Thameslink's Bedford services run on the fast lines and as far as I know, consume additional paths (so are presumably flighted to optimise this), but is this less of an issue as so few miles of the MML fast lines south of Bedford are currently rated for 125mph?

As far as I'm aware, the two gearing optimisation variables are basically acceleration and top speed, with the two extremes being metro where the train is optimised for rapid acceleration, but the lowest top speed (60mph for the underground, for example), or intercity, where they have the highest top speed (>125mph), but the worst acceleration. Given the need for identical performance profiles to make the most of capacity in the Thameslink core, I can see why the 110mph services cannot continue through the core, or have to consume extra paths on the fast lines if they use 700s.

Finally, the point:
Is there a reason why trains cannot have gearboxes with multiple sets of gearing? i.e. Metro-style gearing to get up to 60mph rapidly, then something like a clutch to disengage the gearing and coast whilst switching to the higher speed gearing? I presume there has to be a reason otherwise modern EMUs wouldn't have the set maximum speeds that they do (75/90/110/125) without needing to be re-geared to go faster. Is it just cost reduction, or is it actually impossible? I can't imagine why the 700s wouldn't have this if it were possible, given how new they still are and the constraints on the lines they run on.

 

[norbitonflyer]

A clutch and gearbox would add considerable weight and complexity to the system - there is a reason why diesel locomotives don't use them.

On the older Underground trains there was a "weak field" control which reduced the strength of the windings in the motors, reducing the back EMF and thus improving the speed, but at the expense of acceleration. Used on the outer reaches of the subsurface lines. Whether such a thing is technically possible with AC traction motors I wouldn't know.

 

[Nym]

Weak fielding on LU stock was actually an additional 2 or 3 notches at the top of the sequence.
BR locomotives such as the 47s also had this feature.

Also, some trains do have gearboxes, noteable amongst these being the Class 172s

 

[contrex]

I do know that some French locos had two available ratios for the the traction motor gearing - freight and passenger. Switchover could only be done at a standstill, by moving a lever on each bogie.

 

[norbitonflyer]

Also, some trains do have gearboxes, noteable amongst these being the Class 172s

Yes indeed, many shunters (01, 03-06), and most 1st generation DMUs. But none of those were diesel-electric.

 

[hexagon789]

I do know that some French locos had two available ratios for the the traction motor gearing - freight and passenger. Switchover could only be done at a standstill, by moving a lever on each bogie.

All monomotor SNCF classes as far as I know, both diesel-eletrics and electrics.

Weak fielding on LU stock was actually an additional 2 or 3 notches at the top of the sequence.

Camshaft BR EMUs are the same (and the 'Thumper' DMUs for that matter), there is more than one weak field stage, but the power handle has only one weak field option - if selected it'll simply progress through each stage in turn unless the power handle is moved to a lower setting before the next stage comes in, then the unit will remain in the current weak field stage.

 

[Snow1964]

Diesel-electric really needs to be split into dc electric motors and ac motors as they do not behave same.

With dc motors, the gearing is more critical, because there tends to be a maximum speed, or more precisely top end acceleration falls off as the motor approaches top speed as the magnetic fields do not switch quickly so the rotor gets ahead of ideal, and the next field coil puts up some resistance to spinning, so has what could be described as natural balancing speed as it's max

With ac motors, which are lighter, the acceleration rate is more constant, the speed of the motor is more to do with how the electronics switch the current, than being based on rotational speed of the commutator on dc motors. Being lighter it is easier to make them spin faster, (or if same size as a dc motor, would be lot more powerful).

With ac motors can easily design it to have top speed say 10-20% higher than service speed, and still have decent acceleration, without the old trade off a dc motor.

With dc motors it often took an eternity to accelerate to the fastest 5-10mph as rate of acceleration falls off so sharply and there was no way of getting it faster, and cutting out half the fields (weak field) only tended to add a few percent to speed.

 

[hwl]

A clutch and gearbox would add considerable weight and complexity to the system - there is a reason why diesel locomotives don't use them.

On the older Underground trains there was a "weak field" control which reduced the strength of the windings in the motors, reducing the back EMF and thus improving the speed, but at the expense of acceleration. Used on the outer reaches of the subsurface lines. Whether such a thing is technically possible with AC traction motors I wouldn't know.

Weak field is unnecessary on modern 3 phase AC systems with asynchronous motors.

Context:
I was having a ponder about a reply to a thread on the MML recently, and it got me thinking about the situation with Thameslink where it runs on 125mph fast lines, but the 700s have a max speed of 100mph, costing capacity.

Plenty of sub 100mph, 100/105/110/115mph south of Bedford.

Short of some pretty impressive infrastructure interventions, the only way to avoid impacting the capacity on the fast lines is to stay off them entirely, or downrate them to the fastest common speed. In Thameslink's case, I was also wondering about the Peterborough services, which as a trade-off run with 110mph stock into the surface platforms at Kings Cross, and their equivalent on the MML, the EMT Connect services which also run with 110mph stock into the surface platforms at St. Pancras. Thameslink's Bedford services run on the fast lines and as far as I know, consume additional paths (so are presumably flighted to optimise this), but is this less of an issue as so few miles of the MML fast lines south of Bedford are currently rated for 125mph?

As far as I'm aware, the two gearing optimisation variables are basically acceleration and top speed, with the two extremes being metro where the train is optimised for rapid acceleration, but the lowest top speed (60mph for the underground, for example), or intercity, where they have the highest top speed (>125mph), but the worst acceleration. Given the need for identical performance profiles to make the most of capacity in the Thameslink core, I can see why the 110mph services cannot continue through the core, or have to consume extra paths on the fast lines if they use 700s.

The third option on modern 3 phase AC systems (which have less gearing issues than older DC or single phase AC) is more power (to increase torque) which is effectively a non starter on 3rd rail.

Finally, the point:
Is there a reason why trains cannot have gearboxes with multiple sets of gearing?

the torque, rpm and the resulting gear geometry effectively rules this out (How many electric cars or HGVs have multiple gear options)

i.e. Metro-style gearing to get up to 60mph rapidly, then something like a clutch to disengage the gearing and coast whilst switching to the higher speed gearing? I presume there has to be a reason otherwise modern EMUs wouldn't have the set maximum speeds that they do (75/90/110/125) without needing to be re-geared to go faster.

the range of max speeds is about optimising energy efficiency too...

Is it just cost reduction, or is it actually impossible?

Impossible

I can't imagine why the 700s wouldn't have this if it were possible, given how new they still are and the constraints on the lines they run on.

 

[boiledbeans2]

Weak field is unnecessary on modern 3 phase AC systems with asynchronous motors.

Technically, it already exists on all inverter drive AC motors.

Field (or flux) weakening is used above the base speed of the motor. This reduces the max. torque but increases top speed.

The microprocessor turns field-weakening on/off automatically depending on the motor speed, and therefore cannot be manually selected by the driver.

So the operation would be different to DC motors, but it still exists in AC motors

 

[mr_jrt]

I don't know what proportion of axles are driven on a 700, but I'm reasonably confident it's not all of them. Rather than having the complexity of full gearboxes, could you instead convert the required number of non-driven axles to driven axles, but geared for a different profile, with the ability to switch between them, with the other set just freewheeling? I (obviously) don't know if freewheeling is something the existing axles can do, and I can't imagine it being that useful in current operation, but you never know unless you ask. It obviously doubles the amount of traction equipment in the bogeys, but if it means you can have both high acceleration and pelt down the fast lines at full whack there has to be a value on that.

...or thinking about it, would simply powering twice the axles abet all geared for a high speed profile improve the acceleration so much that it all becomes a moot point?

 

[dastocks]

the torque, rpm and the resulting gear geometry effectively rules this out (How many electric cars or HGVs have multiple gear options)

There are Lexus/Toyota hybrids that do this. I own a Lexus GS450h that switches between a low and high range automatically. It's an extension of the Prius hybrid system using a small 2 speed auto-box on the output allowing a higher top speed.

 

[boiledbeans2]

...or thinking about it, would simply powering twice the axles abet all geared for a high speed profile improve the acceleration so much that it all becomes a moot point?

That's about it. More motors = more torque = more acceleration, even if they were all the same gearing.

The railways used to be the leaders in traction technology. Today, it's the EV industry, e.g. Tesla with 'Ludicrous Mode'.

 

[AM9]

I don't know what proportion of axles are driven on a 700, but I'm reasonably confident it's not all of them. Rather than having the complexity of full gearboxes, could you instead convert the required number of non-driven axles to driven axles, but geared for a different profile, with the ability to switch between them, with the other set just freewheeling? I (obviously) don't know if freewheeling is something the existing axles can do, and I can't imagine it being that useful in current operation, but you never know unless you ask. It obviously doubles the amount of traction equipment in the bogeys, but if it means you can have both high acceleration and pelt down the fast lines at full whack there has to be a value on that.

...or thinking about it, would simply powering twice the axles abet all geared for a high speed profile improve the acceleration so much that it all becomes a moot point?

The 700s have 50% of the wheels driven, partly because that is more than enough to provide the required performance. The trains are also pretty light which also contributes to their acceleraion, e.g. a 12-car 700/1 weighs just 410 tonnes. Doubling the number of motored bogies would add at least 12 tonnes, impairing that acceleration performance.

 

[PyrahnaRanger]

How many electric cars or HGVs have multiple gear options

My BMW hybrid has 8 gears - the motor works through the auto box alongside the engine; if you really floor it, both are fed in at the same time. If this happens when you aren’t expecting it, a change of trousers may well be needed.

BMW explain it here:
BMW Types of Hybrid Explained

As far as HGVs go, I’ve only ever driven an old (1987) Seddon Atkinson. It had effectively a 16 speed box, made up of a lo-range 4 speed with a splitter, and then a hi-range 4 speed with a splitter. It was very impressive to drive, and on one occasion we pulled a 12T coach out of Dover quicker than the coach could normally do on it’s own…

 

[D365]

Can we take a step back and determine whether there actually is a problem as the OP suggests..?

 

[gingertom]

I don't know what proportion of axles are driven on a 700, but I'm reasonably confident it's not all of them. Rather than having the complexity of full gearboxes, could you instead convert the required number of non-driven axles to driven axles, but geared for a different profile, with the ability to switch between them, with the other set just freewheeling? I (obviously) don't know if freewheeling is something the existing axles can do, and I can't imagine it being that useful in current operation, but you never know unless you ask. It obviously doubles the amount of traction equipment in the bogeys, but if it means you can have both high acceleration and pelt down the fast lines at full whack there has to be a value on that.

...or thinking about it, would simply powering twice the axles abet all geared for a high speed profile improve the acceleration so much that it all becomes a moot point?

problem in 3rd rail land is if you double the number of motors you double the current take. The direct consequence of which is the need to upgrade power feeder stations. Also can the pick up shoes handle the (vastly) increased current?

 

[hwl]

Technically, it already exists on all inverter drive AC motors.

Field (or flux) weakening is used above the base speed of the motor. This reduces the max. torque but increases top speed.

The microprocessor turns field-weakening on/off automatically depending on the motor speed, and therefore cannot be manually selected by the driver.

So the operation would be different to DC motors, but it still exists in AC motors

I was trying to keep things simple as it isn't used in most case (only very high speed range) but it isn't used in most cases so I didn't mention it.

I don't know what proportion of axles are driven on a 700, but I'm reasonably confident it's not all of them.

50%

Rather than having the complexity of full gearboxes, could you instead convert the required number of non-driven axles to driven axles, but geared for a different profile, with the ability to switch between them, with the other set just freewheeling?

You would have more at a same but different gearing, you wouldn't switch
Freewheeling is called coasting and virtually all rolling stock can do it apart from some diesel mechanical MUs

I (obviously) don't know if freewheeling is something the existing axles can do, and I can't imagine it being that useful in current operation, but you never know unless you ask.

It is incredibly useful

It obviously doubles the amount of traction equipment in the bogeys,

No it is not doubled because a single gearing ratio would be used with just few more motors rather than double.

but if it means you can have both high acceleration and pelt down the fast lines at full whack there has to be a value on that.

...or thinking about it, would simply powering twice the axles abet all geared for a high speed profile improve the acceleration so much that it all becomes a moot point?

Can we take a step back and determine whether there actually is a problem as the OP suggests..?

There isn't!, but it appears he won't look at the sectional appendix on his own to reach that conclusion because he has already come to a different one...

problem in 3rd rail land is if you double the number of motors you double the current take. The direct consequence of which is the need to upgrade power feeder stations. Also can the pick up shoes handle the (vastly) increased current?

Precisely it is non starter for 3rd rail land as I pointed out in post 8...

 

[AM9]

Precisely it is non starter for 3rd rail land as I pointed out in post 8...

Moreover, south of City Thameslink, all running is limited to 100mph maximum owing to the unsuitability of 3rd rail for much over metro speeds. Even north of London, most of the route to Luton, where the tracks are shared with EMR ssrvices, is limited to 100mph anyway, and the fastest TL trains need the acceleration that they have been supplied with.
For the OP: the class 700 design is very closely aligned to the routes and services that they run, giving a 100mph train that can successfully integrate with MML IC traffic yet accommodate peak passenger loads with a tare weight of about 220kg per passenger. In some situations, the slower acceleration of the class 220s (and poosibly the 810s) can delay the TL services.

 

[mr_jrt]

Can we take a step back and determine whether there actually is a problem as the OP suggests..?

My specific line of thinking was how do you enable regional services to use the fast lines in the most effective way, with a view to these then freeing up high level platform capacity by being sent through the TL core. I wanted to learn more about the technical aspects which you all have done wonderfully, and I thank you for that. I limited the scope to just how and why things already are to justify posting in the main section as an educational aid.

Now we've been shunted into speculative, things can be a bit freer - the services I had in mind were:
A) The ECML outers to Peterborough and things like the Cambridge Flyers, basically anything using the ECML fast lines.
B) The Corby services, and possibly, the Leicesters as well.
..and C) The Bedford services that already make use of the MML fast lines, which although there isn't a whole lot of 125 south of Bedford, I recall reading the OHLE was being upgraded to enable more of it, so it could become more of a consideration.

Basically, shove as much as you can into the core to reduce the need for platform capacity on the surface, but the need to use the fast lines yet also have an identical (and optimised) acceleration profile in the core precludes this.

 

[hwl]

My BMW hybrid has 8 gears - the motor works through the auto box alongside the engine; if you really floor it, both are fed in at the same time. If this happens when you aren’t expecting it, a change of trousers may well be needed.

BMW explain it here:
BMW Types of Hybrid Explained

As far as HGVs go, I’ve only ever driven an old (1987) Seddon Atkinson. It had effectively a 16 speed box, made up of a lo-range 4 speed with a splitter, and then a hi-range 4 speed with a splitter. It was very impressive to drive, and on one occasion we pulled a 12T coach out of Dover quicker than the coach could normally do on it’s own…

How many non hybrid electric vehicles is the relevant comparator as we are talking about electric trains.
Cars:
Tesla - No
BYD (apart from the driving school car as this has to have manual transmission with fake stall for driving tests!) - No
Lucid - No

HGV:
Tesla - No
MAN - No
Volvo/Renault - 2 speed (urban/non-urban mode) their current motor designs and electronics are bit underwhelming so the gear box may change in the future, they are also probably trying to keep the cost down at the moment.
DAF/PACCAR - 2 speed (urban/non-urban mode) their current motor designs and electronics are bit underwhelming so the gear box may change in the future, they are also probably trying to keep the cost down at the moment.
Mercedes - 2 speed (urban/non-urban mode) their current motor designs and electronics are bit underwhelming so the gear box may change in the future, they are also probably trying to keep the cost down at the moment.

Moreover, south of City Thameslink, all running is limited to 100mph maximum owing to the unsuitability of 3rd rail for much over metro speeds. Even north of London, most of the route to Luton, where the tracks are shared with EMR ssrvices, is limited to 100mph anyway, and the fastest TL trains need the acceleration that they have been supplied with.
For the OP: the class 700 design is very closely aligned to the routes and services that they run, giving a 100mph train that can successfully integrate with MML IC traffic yet accommodate peak passenger loads with a tare weight of about 220kg per passenger. In some situations, the slower acceleration of the class 220s (and possibly the 810s) can delay the TL services.

On your last point - The 810s and regenerative braking on electric will improve the fast service performance compared to the 222s. The OP has been looking at the acceleration of the wrong rolling stock...

The fast service time penalty for stop due to poor acceleration is a big issue.

 

[PyrahnaRanger]

How many non hybrid electric vehicles is the relevant comparator as we are talking about electric trains.

Sorry, but you didn’t make this clear from your post - for most purposes, my car is classed as an EV, for example, there are no tax or ULEZ payments levied against it, you can park in EV only bays, you can use charging points in public spaces…

 

[dastocks]

How many non hybrid electric vehicles is the relevant comparator as we are talking about electric trains.
Cars:
Tesla - No
BYD (apart from the driving school car as this has to have manual transmission with fake stall for driving tests!) - No
Lucid - No

Actually Tesla tried a 2-speed transmission on early versions of the Roadster but had problems with it.

The Porsche Taycan (an all-electric EV) has a 2 speed transmission for the same reason as the higher-end Lexus hybrids: you get the low speed performance without spinning the main traction motor too fast at higher road speeds.

This would be the reason to do something similar with a train.

 

[hwl]

Actually Tesla tried a 2-speed transmission on early versions of the Roadster but had problems with it.

They learnt... (do the Plaid models have gearing?)

The Porsche Taycan (an all-electric EV) has a 2 speed transmission for the same reason as the higher-end Lexus hybrids: you get the low speed performance without spinning the main traction motor too fast at higher road speeds.

Road vehicles have higher rolling resistances than rail which has pushed most road vehicle manufacturers towards PM motors as the primary motors which run into the same issues as traditional DC motors or older single phase synchronous AC motors on rail requiring field weakening or multiple gears to get good torque across the full vehicle speed range. They are effectively a very large brushless cordless power tool. The down side of PM motors is that you have to always apply the field in exactly the right way and put some power in or out continuously while doing this which is ok if you have higher rolling resistances (when not braking), this also leads towards single pedal control systems which are very easy to implement.
Many higher performance electric cars (and trucks e.g Tesla and MAN with single speed truck gearing but more motors than the others) use multi motor solutions with different motor design for the second of two or third of three motors, for example Tesla (and others) us three phase AC asynchronous motors on the front wheels of the all wheel drive models with the front motors only occasional used for acceleration and braking but coast most of the time with no power applied or taken out resulting in a lighter than rail type duty cycle for these motors (the three phase asynchronous motors are happier switched off than any other design - which is important for rail).
Three phase AC motors are a (vector) control algorithm nightmare which is one reason many in the road industry might have avoided them for the primary motors (the German automotive manufacturers are more keen to make this this than anyone else).

This would be the reason to do something similar with a train.

Three phase asynchronous motors have fewer issues with larger speed ranges than PM motors so 125mph in rail is usually fine above this may need compromises. This could easily be achieved with a mix of 4-pole and 2-pole three phase asynchronous traction motors without needing gearing changes or fixing the other perineal issue of squirrel cages not being strong enough to increase the max rpm (i.e. the max motor rpm limit is mechanical not electromagnetic).

 

[Railperf]

Context:
I was having a ponder about a reply to a thread on the MML recently, and it got me thinking about the situation with Thameslink where it runs on 125mph fast lines, but the 700s have a max speed of 100mph, costing capacity. Short of some pretty impressive infrastructure interventions, the only way to avoid impacting the capacity on the fast lines is to stay off them entirely, or downrate them to the fastest common speed. In Thameslink's case, I was also wondering about the Peterborough services, which as a trade-off run with 110mph stock into the surface platforms at Kings Cross, and their equivalent on the MML, the EMT Connect services which also run with 110mph stock into the surface platforms at St. Pancras. Thameslink's Bedford services run on the fast lines and as far as I know, consume additional paths (so are presumably flighted to optimise this), but is this less of an issue as so few miles of the MML fast lines south of Bedford are currently rated for 125mph?

As far as I'm aware, the two gearing optimisation variables are basically acceleration and top speed, with the two extremes being metro where the train is optimised for rapid acceleration, but the lowest top speed (60mph for the underground, for example), or intercity, where they have the highest top speed (>125mph), but the worst acceleration. Given the need for identical performance profiles to make the most of capacity in the Thameslink core, I can see why the 110mph services cannot continue through the core, or have to consume extra paths on the fast lines if they use 700s.

Finally, the point:
Is there a reason why trains cannot have gearboxes with multiple sets of gearing? i.e. Metro-style gearing to get up to 60mph rapidly, then something like a clutch to disengage the gearing and coast whilst switching to the higher speed gearing? I presume there has to be a reason otherwise modern EMUs wouldn't have the set maximum speeds that they do (75/90/110/125) without needing to be re-geared to go faster. Is it just cost reduction, or is it actually impossible? I can't imagine why the 700s wouldn't have this if it were possible, given how new they still are and the constraints on the lines they run on.

You have to remember that LNER Class 801 can accelerate faster to 100mph than a GN Class 387. And maybe only fractionally slower than a Thameslink Class 700. The Class 810 Aurora units are predicted to be similar performance maybe slightly slower than a Class 801. So right now, the 140mph capable Intercity units offer as good acceleration as a 100mph suburban unit.
So the real issue now is the paths lost by trains switching between slow and fast lines at Leagrave and Harpenden, and delays caused by late running trains at those locations. Don't forget The switching from Fast to Slow at Kentish Town either. Now added to that is the complication of EMR Connect switching between fast and slow at Wellingborough or Kettering South. I'm aware of northbound EMR Intercity services catching up the Connect services for sure - which could be avoided if Connect services were switched over closer to Bedford. Another issue is the very slow approach southbound of all Bedford stopping trains due to track and platform layout. Bedford could really do with a platform on the Up fast lines - that would help journey times. By the time pathing and performance allowances are added in to account for this, the Intercity services are effectively averaging speeds similar to the slower services.