News:

Simutrans Wiki Manual
The official on-line manual for Simutrans. Read and contribute.

Inconsistent configuration for 375 electrostar

Started by Vladki, April 13, 2020, 11:16:15 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Vladki

BR class 375 consists of 4 vehicles: Front, Middle, Panto and Rear. Possible combinations are: FMR or FMPR. There are two versions: 375 (DC only) and 375/6 (AC/DC) and there is possible upgrade from 375 to 375/6. Now the inconsitency is in which vehicles are powered. F and R are powered in both versions, but M is powered only in 375 and P is powered in 375/6. That makes the upgrade of M to lose the engine, and P get the engine.
Moreover a 3 car formation of 375 has all units powered which is too much. Also the text description is funny:

375/6-Middle says "pantograph trailer" but the pantograph is painted on 375/6-Panto which is described as centre motor.

I think the proper order should be: FP(M)R, with FMR being powered, P unpowered with the image of pantograph.


Vladki

I have dived futher into the dats and wikipedia for reference, and prepared a bit larger patch. What it does?
- allowed configurations: FPR, FMPR for 375 and FPR, FMPR, FMPMR, FPMR for 375/6. The second center motor is coded as BR-377-7Middle2.
- all except Pantograph car are powered. Images fixed.
- costs adjusted so that powered units have same purchase and runningcost, unpowered are cheaper to buy, and free to run (per km)
- fixed_costs also adjusted to have the pantograph cheapest, but I'm not sure if they are correct. All cars in 375 have different fixed cost. in 375/6 it looks a bit more consistent (same price for front and rear, same price for middle motors; and much cheaper pantograph)

See commit: https://github.com/jamespetts/simutrans-pak128.britain/commit/c52350a0ccbcb115d5534bf5d4e118af5a063419
I tried to make a pull request but it includes my older suggestion for barge constraint fixes, which were not accepted.
So please be careful when applying.

TODO in translator - I cant modify english texts, so I put only suggestions
- translation swap for 375/6 pantograph and centre motor
- missing translation for 377/7 second motor

jamespetts

I have incorporated this fix - this is helpful.

I wonder whether the translator administrator might give Vladki administrator access Pak128.Britain-Ex on Simutranslator?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Vladki

#3
I have access to modify czech translations but not english :)

UPDATE: one more fix - middle power car was missing axles=4, and added gear to pantograph (even unpowered).
https://github.com/vladki77/simutrans-pak128.britain/commit/7b65ab202d2586ad31e75d177ce58073a5e95540

UPDATE2: I see similar problem with BR class 376. The constrains are not consistent, and probably were originally intended to allow for any combination.
Cars are Front, Middle (power), middle (Panto), Rear.
Realworld combo is: FMPMR
Simutrans allows: FPPPPPPPMMMMMMMMR  (with any amout of P's or M's in row, but not alternating them)

Suggestion for fixes:
- allow any combo: with shortest option: FR, and any combination of P and M in between.
- allow alternating FMPMPMPMR of any length, with shortest combo FPR
- allow alternating FMPMPMPMR of any length, with shortest combo FMPMR
- allow only real world combo FMPMR - would require distinguishing first and second M
I can prepare patch, but I'm do not want to decide which variant should go?

Sirius

#4
Imho it should either only allow the real-world combination or any number of M require exactly one P.
In the real-world (at least in Germany) most single electrification EMUs either have exactly one panto or one at the front and one at the back, so I don't think the alternating MP suggestion is a good one.

My suggestion for a fix would be to allow FMPM'R, with M being the same as M' apart from constraints. Either allow only that exact combination or any number of M and M'

I am not quite sure what would be the Pak128.Britainish way as there does not seem to be a consistent way for such cases in the pak.
Imho multiple units should allow for some not exactly realistic variations.

Both have their pros and cons.
Allowing a high level of customisation in train composition is much more difficult to balance can fight the uniquenes of a train and might be argued as unrealistic.
On the other hand, ingame circumstances are different from real-world circumstances. Allowing exact real-world combinations only will greatly restricts players ability to match their trains to the circumstances.
The realism aspect can be easily counter argued. Trains in the real-world are a product with many configurable variables. Just because it was ordered as a 5 car unit, does not mean any other composition would be unrealistic. For sure, it would be a different class in that case, so I would argue we should not offer train classes but instead offer "platforms" or "families", like "Siemens Velaro" instead of "Class 374"

Things go even further than just the number of cars but with the current features we cannot go much further.

E.g. in the real-world the transport companies of Cologne and Bonn said in the 1960s "we need train capable of operating under BOStrab (light rail operation law) and EBO (heavy rail operation law), a maximum speed of at least 100 km/h, equipped for road bound tracks and a few more things", so Duewag created the B100 series that matches exactly the required constraints.
In simutrans we canot say
"we need a vehicle capable of road bound and normal rail operation, at a speed of at least 100 km/h, AC and DC catenary electrification"
transport Company of Karlsruhe, ~1986

These are quire extreme examples but they clearly hit the point:
Such vehicles did not exist before. These were explicitly ordered and developed for that use case. We cannot do this ingame but it is imho quite important to allow as much flexibility as possible with the game engine to allow simucompanies to properly react to ingame requirements.

Apart from allowing flexible coupling of cars of a specific multiple unit class, I would go even further and offer train families instead of specific classes that can be coupled as desired.

Vladki

Quote from: jamespetts on April 15, 2020, 11:17:22 AM
I have incorporated this fix - this is helpful.

I wonder whether the translator administrator might give Vladki administrator access Pak128.Britain-Ex on Simutranslator?

I got translator rights for English, so the translation for 375/6 and 377/7 is fixed.

Phystam, I see there is japanese translation that needs to be fixed too - the translations for 375/6 Middle and Panto should be swapped. (Panto is really pantograph carriage now).

Vladki

#6
I'm looking further and found that all units of the electrostar family are in real world limited to these configurations: FPR, FMPR, FMPMR
F = front motor & cab
P = pantograph trailer (unpowered)
M = middle motor
R = rear motor & cab
Longer units, although probably possible are not in use in UK.
Some units were bought as 4-car unit, and 5th part was added later. The process took a few days and "new" units had to do a test ride. So with future recombination system, these should be recombined only in depot.

Current constraints in simutrans are a wild mess, allowing for unrealistic combinations, and also sometimes having powered pantograph car, etc. I'm about to prepare a big patch (hopefully not as big as the trams), to put this in order. The question about constraints is:
- should we allow all electrostars to have one of the 3 combinations, even if some classes really exist only in one or two of them?
- should we allow some further combinations like FMMMMPMMMMR ?


Oh no... This is gonna be as bad as trams....
runningcosts - mostly OK - proportional to power, with varying ratio. Unlike with trams, older elecrostars (357) are cheaper 0.10 c/km/kW, while modern are more expansive 0.15, and 0.20 (class 379 and 387).
purchase costs: most units have two prices for different parts, but quite unintuitively the front/rear cab part is often cheaper... I would fix this to be either: cab more expensive than middle part,  or powered more expensive than unpowered.
fixed_costs: That is complete mess... Will dig that later, but I would expect the front cab to be most expensive (drivers & conductors wages), then rear cab and powered parts, then pantograph (probably includes transformer), and unpowered DC only parts cheapest.

And the tractive effort - even a single source states 3 different values: https://districtdavesforum.co.uk/thread/22284/378-fifth-car (last post)
Quote
I suspect the extra coach will be an M with three motors as with all the other powered vehicles.

To my knowledge the motors used in the 376/378/379 are slightly different than those used in the 375/377 and are rated at 200kW rather than 250kW.

Some observations. A 378 accelerates at about the same rate as a 4 car 375/377 on DC. A 378s normal starting tractive effort is 150 kN whether on AC or DC. The difference is on DC this figure quickly falls off above 25 mph whereas on AC the full 150 kN doesn't drop off until 35 mph giving fantastic acceleration.

A 375/377 produces 132 kN starting tractive effort. 44 kN per powered vehicle or 22 kN per motor. The 378 based on 9 motors gives just over 17 kN per motor. When fully loaded a 378 can produce 180 kN starting tractive effort but even that is 20 kN per motor - less than a 375/377 (22 kN per motor) leading me to believe that the 376/8/9 are rated at 200 kW and the 375/377 are 250 kW.
150 kN divided by 9 motors is slightly below 17 kN (not over)... And what about the 180 kN when loaded?  And should this value be entered as is to be reduced by 80% gear, or modified to be 150 kN after applying the gear?

Sirius

#7
Quote from: Vladki on May 15, 2020, 06:19:49 PM- should we allow all electrostars to have one of the 3 combinations, even if some classes really exist only in one or two of them?
- should we allow some further combinations like FMMMMPMMMMR ?
Just as for trams, that's a general question, that should have a consistent answer for the whole pakset:
Do we prefer realism or gameplay?
And if we prefer realism, what do we consider realistic?
If we prefer gameplay, it's best to allow any combination, so players can react in ingame requirements. However, balancing this is more difficult.
If we prefer realism, there are two options:
1. Allow what was actually ordered.
2. Allow what was technically possible to order.

The latter is a little spongy, but imho that one should be preferred.

Anyways, that leaves room for discussion (in an other thread, at an other time) e.g. about Eurostar E300 or HST/IC125.

Vladki

I started with london overground class 378. As it has changed over time from FPR to FPMR to FMPMR, and we do not have an image with pantograph, I think that allowing any length with alternating MP is OK. But I have problem about capacity: https://www.whatdotheyknow.com/request/339955/response/848693/attach/3/Class%20378%20Data%20Sheet.pdf
Passenger accommodation: Please note that standing capacity figures exclude seating capacity
Seating capacity: (Number of full seats per train)186
Seating capacity: (Number of tip up and perch seats) 84
Wheelchair spaces 2 (but -6 tip up seats)
Maximum average load standing capacity (seated plus 25% crush)420
Maximum full load standing capacity (seated plus 70% crush)840
Theoretical crush standing capacity (seated not including perch seats) plus 7 standing customers per m2 in vestibule areas and 6 in all other: 906


This one was clearly intended for extreme crowds. Which value should we take for seats - normal only, or include the tip up and perch seats?
How much for standing? Trams had usually figures for 4 persons per m2... Lets say that the max figure is for 6.5 per/m2, then it would be cca 140 m2 -> 560 standing

jamespetts

Thank you for looking into this: this is most helpful.

As to realistic combinations, the general principle is to permit anything that would be permissible or realistically possible in reality, not to limit options to what people in fact chose in reality even if other options were available. Realism of this sort is not opposed to gameplay; rather, the gameplay consists precisely in building a transport network by overcoming realistic constraints; Simutrans-Extended is intended to be an immersive game.

Thus, in this case, if the manufacturers of these trains could have made different combinations of vehicles available but UK rail operators chose only a subset of what was available, then, ideally, we would want to offer the player everything that the manufacturer offered to the rail operators, but nothing that was not on offer.
As to seating capacity, I should generally ignore tip-up and perch seats and give the overcrowded capacity on the assumption that these seats are not being used.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Sirius

Thanks, I was hoping for that kind of statement.

Vladki

#11
OK thanks for claryfing. According to https://en.wikipedia.org/wiki/Bombardier_Electrostar

These were made almost only for UK, only one batch for south africa. Produced formations were: 3, 4, and 5 car long units, always with one unpowered (pantograph) car. Longer trains are possible by coupling two or more units together. So I will make the constraints this way: Front-(Middle)-Panto-(Middle2)-Rear. And adding Middle2 to all classes, especially those that really run in 5-car combination, and moving the 5th car introduction date to be the same as the rest of train (not waiting few years as was the case in reality).

Production of the trains ended in november 2017 when unit no. 387174 for Great Western Railway was completed at Derby[4] and was superseded by the Bombardier Aventra. So I will unite all retire dates of electrostars to this date. OK?

EDIT: digging into class 376: https://uktransport.fandom.com/wiki/British_Rail_Class_376 says that seating capacity is 344 (which is exactly what is specified in simutrans). But says that: In order to provide more standing room, the trains also have fewer seats, more handrails, and no on-board toilets (in comaprison to clas 375). But Class 375 has 296 seats (extrapolated to 5-car unit). Wikipedia says 376 has only 228 seats. (and a photo they are in 2+2 config). FInally found some videos and one car has 48 seats - so it seems that wikipedia is right this time (48*3+42*2=228)

EDIT2: first round of fixes pushed to new branch: https://github.com/vladki77/simutrans-pak128.britain/tree/electrostar-fixes

EDIT3: ohooo - found a funny constraint allowing classes 375 and 379 to be coupled together, but one is AC -only and the other DC-only :) So modifying to allow 375/6 which is dual voltage :)

Vladki

We have only 2 liveries for DC-only 375, but 4 for dual voltage 375/6. Should I add them to the DC only 375 too?

Rollmaterial

Yes, but only the Southern livery. First Capital Connect never operated DC-only units to my knowledge.


Sirius

I am unsure about this as our current feature set won't allow us to specify something like a maximum distance on battery and a reload time under electrification.

Same battery upgrade goes for West Midlands' CAF Urbos 3 btw.

Vladki

I was unable to find some authoritative source on what engines are on class 379 (power/tractive force/acceleration). Any help Freahk?

Sirius

I have no Idea, I'll have to search the archives for some technical data.
Might do so tomorrow afternoon. I have to sleep now, alarm clock rings in 6 hours to remind me of a little hike.

Vladki

I think I am finished with electrostars. Can anyone try https://github.com/vladki77/simutrans-pak128.britain/tree/electrostar-fixes ?
Allowed configurations are:
- Classes 375 (377) and 378 allow 3-, 4- and 5-car units, just as in real world.
- Class 376 in reality is only 5-car unit, but the game allows 3/4/5 cars.
- Classes 357, 387 and 379 in reality are only 4-car units, game allows 3/4 cars.
Should I add 5th car to these too?

Also classes 375 and 379 can be coupled together, although I'm not sure if that is used in real world. It does not make much sense in coupling different classes even if it would be possible, because they usually differ in electrification type, top speed, or presence of gangways...

Sirius

Well coupling these in the real-world makes more sense than in simutrans due to portion working.
Here, we have a Regional Express line that splits at a station where one train coninues to the Netherlands, where the other one continues to a destination in Germany. The first is equipped for operation in the Netherlands, for sure. The other one most often is not.

It's use ingame is limited to recycling old rolling stock, although even in that case it's arguable if it's not better to simply buy a new one that fits the requirements and scraping the multi electrification one.

jamespetts

We do want to be able to allow portion working in the future, so configurations that allow for this are to be encouraged. Do Vladki and Freahk consider this ready for integration yet?

Thank you all for your work on this.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Sirius

I think it's ready but this time I didn't have a detailled look at it, so I'd wait for Vladki.

Vladki

I think it is finished. I have already uploaded the translations

Vladki

Quote from: Vladki on May 21, 2020, 09:07:19 PMClasses 357, 387 and 379 in reality are only 4-car units, game allows 3/4 cars.
Should I add 5th car to these too?

James, Freahk, what do you think about allowing 5th car to these classes? It seems technically possible, only not ordered.

jamespetts

Thank you - now incorporated. As to a 5th car, if this really is not only theoretically possible but practically feasible, then this should be permitted.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

fam621

Quote from: Vladki on May 21, 2020, 09:07:19 PM
I think I am finished with electrostars. Can anyone try https://github.com/vladki77/simutrans-pak128.britain/tree/electrostar-fixes ?
Allowed configurations are:
- Classes 375 (377) and 378 allow 3-, 4- and 5-car units, just as in real world.
- Class 376 in reality is only 5-car unit, but the game allows 3/4/5 cars.
- Classes 357, 387 and 379 in reality are only 4-car units, game allows 3/4 cars.
Should I add 5th car to these too?

Also classes 375 and 379 can be coupled together, although I'm not sure if that is used in real world. It does not make much sense in coupling different classes even if it would be possible, because they usually differ in electrification type, top speed, or presence of gangways...

The electrostars with gangways can all be coupled together as irl it is possible for the trains software to be upgraded so that there is cross compatibility.

Vladki

#26
Quote from: thegamer7893 on May 24, 2020, 12:38:41 PMThe electrostars with gangways can all be coupled together as irl it is possible for the trains software to be upgraded so that there is cross compatibility.
In that case we are almost covered: gangways are on classes 375, 377, 379 and 387.   375 and 377 are so similar that for game purposes they are the same, 375 and 379 can be connected. Only 387 is not. It has higher top speed, but adding a constraint is not a problem. Other clases do not have gangways, yet they can connect with further units of the same class.

So I'll add 5th car to all classes and allow coupling of 375, 375/6, 379 and 387. EDIT - with the exception of coupling DC-only 375 to AC-only 379. ;-)

Sirius


Vladki

Added. Github pullrequest created, translations added too

jamespetts

Splendid, thank you: now incorporated.

It would be helpful if you could in future also add the translations to the Github repository since the files here will in any event need to be updated for all users who do not manually download the translations from Simutranslator.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Vladki

Quote from: jamespetts on May 25, 2020, 12:42:46 PMIt would be helpful if you could in future also add the translations to the Github repository since the files here will in any event need to be updated for all users who do not manually download the translations from Simutranslator.
How do I do that?

Vladki

#31
Testing the electrostars a bit and the seem quite sluggish. Of course... I did not divide the tractive effort by gear.
According to this site: https://districtdavesforum.co.uk/thread/22284/378-fifth-car
Classes 375/377 (and maybe 357 too) have two motors per car, rated at 250kW / 22 kN
Classes 376/378/379 have 3 motors per car rated at 200 kW, with tractive effort of 16.6 kN (378).
These values are put exactly as they are to dat files. But the tractive effort should be imho divided by gear...

Also no idea on tractive effort for 379 and 387. They probably have the same engines as 376/378, but 376/378 have top speed 120 km/h and 379 has 160 km/h and 387 even 177 km/h. The question is if they have different transmission ratio or not. It is IMHO probable they have different transmissions to give the suburban slow trains maximum accelereation. Something similar was done on czech locomotives in classes 371/372 and 363/362. They differ only in gear, and the traction effort is in roghly opposite ratio than top speed. (speed 120:160 = 3:4, tractive effort 275:205 = 4:3).   So that would give 12.5 kN for 379 and 11.4 kN for 387. But that would give them worse acceleration than class 375.

Oh, well more testing...


Wow this is a source: https://eversholtrail.co.uk/fleet/?class=electric-multiple-units  - acceleration for many many...

Anyway, they write that 357 has acceleration 0.62 * 172 tons = 106.6 kN / 3 power cars = 35.5 kN / gear 0.8 = 44.3 kN   ehhh what? One source says tractive effort 44 kN per car, and another exactly 0.8x less ?

And to make a complete mess in it https://en.wikipedia.org/wiki/Bombardier_Electrostar saus
Power output   
2x373 = 746 kW kW (377/3 only)
3x373 = 1,119 kW (others, third rail)
3x560 = 1,680 kW (AC mode)
And the first link says 3x500 or 3x600 depending on class

So what now?
Actually trying all electrostars in 4-car units, it seems not bad. But 3-car units are rather lazy...

jamespetts

Thank you for this research: this is very helpful.

It is very difficult trying to reconcile inconsistent sources. It is interesting that the tractive effort in one is exactly 0.8 of the other, since this 0.8 reduction factor is precisely that used by the "gear". I wonder whether something similar might be happening here? Perhaps you might use the higher of the two on this basis; but you might want to test the in-game acceleration to see which best matches the reported acceleration of each type of unit.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Vladki

#33
According to https://districtdavesforum.co.uk/thread/22284/378-fifth-car
Quote
A 375/377 produces 132 kN starting tractive effort.
A 378s normal starting tractive effort is 150 kN whether on AC or DC.
When fully loaded a 378 can produce 180 kN starting tractive effort
That's for the whole unit of 3-powered cars. Also once they say 150 kN whole train = 16.6 kN per motor, but then say that each motor is slightly above 17 kN... typo?

My thoughts:
Lets start with 375/377 (maybe also 357) ... 500 kW / 44 kN / 160 km/h  (1-powered car), Power:force 500:44 = 11.36;

376/378/379 - 600 kW / 50 kN / 120-160 km/h. I doubt that the 120 km/h have the same transmissions - tractive effort.
Pwer:force  600:50 = 12 . I'd use that for the 379, which has the same top speed. Same top speed, similar power/force ratio (12:1).   
For 376/378 I'd use the full load mentioned above - 60 kN, power/force ratio 600:60 (10:1). 
Speed ratio 379 vs 378 is 160:120 = 4:3 = 1.333; Force ratio 60:50 = 1.2, acceptable?

BUT on https://eversholtrail.co.uk/fleet/class-376/ they say max accel= 0.66 m/s/s * 172 t = 113.52 kN / 3 = 37.84 per car / 0.8 gear = 47.3  (not even the 50 kN I would expect...)
EDIT: 376 is 5-car formation so 0.66 m/s/s * 212 t / 4 powered cars / 0.8 gear = 43.7 kN .... What back again to 44 kN ???

And now the 387 which can go 177 km/h otherwise should be the same as 379. Speed ratio 177:160=1.1, so tractive force should be 50/1.1 = 45.5 kN (hmm back to original 44 kN ?)


I checked the acceleration graphs at https://eversholtrail.co.uk/fleet/. They are pretty useless - the scales are so sparse that you cannot find any precise values. And approximating the straight part of the graph gives me acceleration 0.83 for 376 and 0.7 for 375 :-)  So they are not consistent with themselves... Seems like not very trustworthy source...

Will try on game how they run...

Sirius

I made the experience that such collection pages are most often rather untrustworthy.
Data is often collected in a quick manner, not validating anything. The first match will simply be added to the list.

I'll do some research tomorrow, I planned to validate Eurostar data tomorrow anyway, as there seems to be some error either.