Topic: Modern rail vehicles for pak128.britain-ex

[jamespetts]

I'd like to increase the power of the Desiro UK units as I believe the figure used in the dat file is for power-at-rail and thus already accounts for losses that would otherwise be applied using 'gear'; but it's not high priority.

Interesting - may I ask what the source is for this?

[Vladki]

Vladki, what were your ideas on 800/801 constraints? If you like I could add them at the same time as the liveries.

Real formations according to wikipedia are:
5-car: DPTS-MS-MS-MC-DPTF
9-car: DPTS-MS-MS-TS-MS-TS-MC-MF-DPTF

But in simutrans we have:
name=BR-801Front  (DPTF)
name=BR-801MiddlePower  (MS)
name=BR-801MiddleTrailer  (TS)
name=BR-801MiddleComposite (non existent TC)
name=BR-801Rear (DPTS)

So we are missing MC (motor composite) and MF (motor first class), while TC (trailer composite) does not really exist (but probably could).
Also there is upgrade from TS to TC and back, which does not work properly, and should be removed.

I had two ideas about constrains. Flexible to allow any combination of motor and trailer cars, only forcing a motor car just next to front and rear cars, so that there is at least some power available...
More realistic: DPTF-(MF)-MC-MS/TS-MS-MS/TS-MS-DPTS
DPTF can be followed by MF or MC (or MS?)
MF must be followed by MC (or MS?)
MC can be followed by TS or MS
TS must be followed by MS
MS can be followed by TS, MS or DPTS

Here: http://www.hitachi.com/rev/pdf/2014/r2014_10_105.pdf is detailed seat plan of 5-car unit, so capacities can be checked and corrected if needed.

If you have sources for power-at-rail then you could set gear=100%, but add a comment with the source in the dat file. So that it is clear to anyone who looks at it later, that it is not a bug but a feature.

[Freahk]

If you have sources for power-at-rail then you could set gear=100%, but add a comment with the source in the dat file. So that it is clear to anyone who looks at it later, that it is not a bug but a feature.

I'd rather set the known engine power and adjust gear accordingly to result in the, in this case, also known power at rail. Feels more consistent to me, especially as the long-term objective should be to set the gear of all vehicles individually according to their actual loss factor.

[Vladki]

I'd rather set the known engine power and adjust gear accordingly to result in the, in this case, also known power at rail. Feels more consistent to me, especially as the long-term objective should be to set the gear of all vehicles individually according to their actual loss factor.

Of course, if we have both values that this is the ideal situation. I thought we have ONLY power-at-rail value.

[Freahk]

I'd use the quoted document with care.
I just had a more detailled look at it and it seems to mix up power at rail and installed traction power (engines maximum output)

As far as I know, the Velaro E has a total of 8800 kW engines installed, but the paper quotes that power as "Maximum power at wheel", so it is quite likely that further "powers at wheel" are actually the power of the installed engines.
In case of the Velaro E, the transmission losses are minimal due to the engine type, it's maybe 5% or even less, I once had a source about this and it was significiantly less than 20%, although I am not  sure about the exact number anymore.
Sorry for not putting the source here, I had posted those sources somewhere in the forums before, but I am too lazy to search for them now as it's not pretty relevant in this case anyway.

Just keep in mind, the quoted paper does not seem to be a pretty reliable source, as with any of such data collections, at least from my experience.
Those collections are nice to get started, but often unpreciese or even wrong in the details.

[jonbridg]

Ok, this gives me some peace of mind as max output power is what I've used on vehicles I needed new data for.

Contrary to what I said earlier classes 8xx will not be released first while I add the constraints, so as TPE Class 68, Mark 5a coaches and Class 397 are ready to go I've pushed them instead.

Interesting - may I ask what the source is for this?

The definition I found for power-at-rail is here:
https://enacademic.com/dic.nsf/enwiki/3181161

[jonbridg]

A summary of what I've been up to so far:

After much deliberation I have opted not to separate the 800 and 802 classes. My initial plan was to distinguish the 800 as cheaper-to-maintain, possibly with down-rated engines, and with restricted coupling constraints. This would be justified as the class was ordered to specifications laid down by the UK government.
802 on the other hand would be as Hitachi might present the product to a private customer, i.e. full-rated engines, and largely unrestricted constraints.
I tried this with separate vehicles, a number of common vehicles, and deleting liveries on vehicles that don't carry them in reality, but the effect in the depot window was confusing, particularly for new players. I also wanted to add new vehicles for variety and constraint purposes which resulted in a lot of duplication so, ultimately, I dropped the 800 side of the plan.
Proposed changes look like this:
Too speed: 200km/h (225 unrealistic in diesel mode)
Tractive effort: better than 65 (to reflect higher acceleration of 802, about 0.85m/s/s. Will be calculated using data published by Rail Performance Society and Eversholt Rail
New vehicles: front and rear motor standards  (constraint purposes), 18+58 composite (Hull spec) and Driving Trailer (compact; 10 extra seats, lesser catering. TPE/Hull spec)
All liveries will be supported.

I'll be ready to upload once the 68/397/Mk5a have been added to the pakset, to avoid a backlog for James Petts who is kindly merging the vehicles.
 

[jamespetts]

Excellent, thank you for this.

Can I check whether there is anything waiting to be merged at present of your work? It would be helpful if this could be indicated on this thread in future so that I can keep a track of what is awaiting merger.

Thank you again.

[Freahk]

and deleting liveries on vehicles that don't carry them in reality

I do not think this matches the spirit of the pakset, which is "design vehicles according to what would be possible in the real-world rather than what was actually ordered"
There's always much room for interpretation in this statement.

To my interpretation, adding more liveries to train that didn't carry these might not be a priority, but existing ones are totally fine and shouldn't be deleted.

[jonbridg]

To my interpretation, adding more liveries to train that didn't carry these might not be a priority, but existing ones are totally fine and shouldn't be deleted.

I agree. The deletions were temporarily made to a few vehicles, and then, only liveries I have created. These have now been reversed anyway. You won't find a Motor First vehicle on any real TransPennine class 802, but you will in Simutrans!

For the same reason I haven't changed kierongreen/TheHood's Virgin livery for Class 801, which is based on an early livery design that was not used in the real-world. A different Virgin livery was later applied to Class 800 and I have added this, but I don't know if it was ever used for Class 801.

Can I check whether there is anything waiting to be merged at present of your work? It would be helpful if this could be indicated on this thread in future so that I can keep a track of what is awaiting merger.

Of course, I've added a To Be Merged list in the original post. I haven't uploaded the .blends yet as I want to tidy them up before public release: the livery is packed within the .blends as a .png image, which is 'UV' mapped to the mesh using a material texture. My hope is that new liveries can be created using a 2D graphics program (which many more people seem to be familiar with) such as Photoshop, and Blender only used for final tweaking and rendering. I will write a more in-depth post explaining how it works soon.

[KneeOn]

These look fantastic and you've made serious progress in a very short space of time. That's commendable!

For what it's worth, I think keeping the various 80x as one unit until features directly relating to these come in to effect is the best way forward and agree with your decisions.

Keep the work up!

[jamespetts]

Thank you for this, and apologies for not having had a chance to look into this until now.

Unfortunately, there is a compile error:

Code: [Select]

ERROR: image_writer_t:  cannot open trains/./carriages/br-mk5a-fo-tpe_S.png

I suspect that you may have mistakenly left some of the image files out of the commit. I should be grateful if you could look into this so that I can test and integrate this.

Thank you very much for your work on this - this does look good.

[jonbridg]

Bother. I'm sorry, I copied the images to the 'railcar' folder instead of 'carriages'. I'll see if I can move them tomorrow.

[jamespetts]

Bother. I'm sorry, I copied the images to the 'railcar' folder instead of 'carriages'. I'll see if I can move them tomorrow.

Splendid, thank you.

[jonbridg]

Ok, I've moved the images, does that solve the problem?

[jamespetts]

Excellent, that does now work, thank you. Now incorporated.

[jonbridg]

Great, thanks very much.
Going away for a few days so there'll be no progress until I get back, and as the extra vehicles I've added to the 8xx classes allow me to add extra livery details (which I've yet to complete), the 8xx are still not complete to my satisfaction, I'm afraid.
However, I got bored whilst working on dat files and added a Cross Country livery to the Voyagers:



I've also re-rendered the Virgin variant with the new RGBA workflow. These are both ready to go.

Together with a dat file update to Class 222 (Not class 22 as it says in the commit, that's a misspelling) to make the constraints less restrictive, I've pushed these and updated the list in the original post.

There's no rush, of course, I'm happy for them to be added as and when time allows, this is my hobby after all

[jamespetts]

Excellent, thank you for that: now incorporated.

Bon voyage!

[ScotRail170434]

Nice work being done!

Is it possible top have a ScotRail update with the Class 320 added along with a Saltire variant of the 318 and a 156 update?

Would be nice to see on my WIP ScotRail fantasy network.

[jonbridg]

Hi Scotrail, I won't make any guarantees, but I'll keep those ideas in mind. I've a pretty long list of objects I'd like to do and it could be months before I get around to anything else.
That said, the shorter and cheaper Class 320 would make a useful addition alongside its faster cousins, and the great similarity of 320/321/322 would make it easier to create than some.

[ScotRail170434]

Hi Scotrail, I won't make any guarantees, but I'll keep those ideas in mind. I've a pretty long list of objects I'd like to do and it could be months before I get around to anything else.
That said, the shorter and cheaper Class 320 would make a useful addition alongside its faster cousins, and the great similarity of 320/321/322 would make it easier to create than some.

There’s already an SPT 318, so it’s just a relivery and the 321 is already in game, so the base units are there. I had a go at the 320 a while back, it worked functionally but looked odd graphically, I also done a 303, 101 and 156 which I might add to the repository if you wish.

[jonbridg]

Bon voyage!

Thanks, it was. Suitably refreshed from my break, I'm pleased to announce I've pushed the updated Class 801 featuring LNER livery, increased power and revised constraints. I've also RGBA-rendered the existing Virgin-proposed livery.
 

There’s already an SPT 318, so it’s just a relivery and the 321 is already in game, so the base units are there. I had a go at the 320 a while back, it worked functionally but looked odd graphically, I also done a 303, 101 and 156 which I might add to the repository if you wish.

Sounds good, are the 303/101/156 in SPT livery? What was odd graphically about the 320? 

[ScotRail170434]

Thanks, it was. Suitably refreshed from my break, I'm pleased to announce I've pushed the updated Class 801 featuring LNER livery, increased power and revised constraints. I've also RGBA-rendered the existing Virgin-proposed livery.
 
Sounds good, are the 303/101/156 in SPT livery? What was odd graphically about the 320?

The carriages had a gap between which I fixed but lost the blend and graphics etc due to switching to a new setup. Would be nice to see a better version. Aye SPT is correct.

[jamespetts]

Splendid, thank you for that: now incorporated.

I have amended some of the translation texts to make them more consistent with the general scheme of these texts especially as regards descriptions and capitalisation.

[Vladki]

Oh, I see the 801 has been modified. Thanks.

But I have a question.  Why there is a distinction between 801MiddlePower and 801FrontMS ?  Looking at the constraints I do not see any reason for FrontMS. Also we could do without RearMS, unless the difference in capacity 72 vs 88 is significant?

Then the First and Composite power cars should have the same running and maintenance costs as MiddlePower.
MiddleTrailer and  Rear should have the maintenance also like other Middle cars. Extra maintenenace makes sense only for the front car (driver, catering).

EDIT: Maybe the rearMS is the car with the auxiliary diesel engine? and thus it has reduced capacity and increased weight?

EDIT: and about 397 - the only car that has some per/km costs is front pantograph (unpowered). Usually it is the powered vehicles that have non-zero per/km costs, usually proportional to their power in kW. Now the 397 is extremely cheap compared to 801 or 395

Also - I cannot find any electric engine suitable for pulling the Mk5 coaches? (200 km/h)

[Freahk]

Also - I cannot find any electric engine suitable for pulling the Mk5 coaches? (200 km/h)

That's because the pakset is obviously very British.
The coaches were ordered to run with a class 68 locomotive, which is a 200 km/h fast diesel engine.

Afaik, there is no modern fast electric locomotive available in the UK, so it's not in the pak.

[jamespetts]

Freahk is correct - all the modern electric passenger trains in the UK are multiple units.

[jonbridg]

Hi Vladki, thanks for taking the time to look over these details.

But I have a question.  Why there is a distinction between 801MiddlePower and 801FrontMS ?  Looking at the constraints I do not see any reason for FrontMS.

The distinction between MiddlePower and FrontMS removes the need for MiddlePower-MiddlePower constraints, which could be misused. It prevents players forming long trains without trailer cars. which is fundamentally possible but unrealistic.
This arrangement also allows differences in livery along the train, as used by these Class 800 liveries:

Notice, on the blue unit, the three middle vehicles have different liveries. This may also be useful for the upcoming Avanti livery which looks complex.

Also we could do without RearMS, unless the difference in capacity 72 vs 88 is significant.
---
Maybe the rearMS is the car with the auxiliary diesel engine? and thus it has reduced capacity and increased weight?

The RearMS would indeed have the auxiliary engine. On Virgin/LNER 5-car units this vehicle also boasts a 'cafe bar' hence the reduced capacity (on real 9-car units the cafe vehicle is nearer the middle, but in Simutrans passengers can't tell the difference). Unless class-specific catering is implemented at some point, this cafe is actually redundant, so the reduced seating is purely for realism! If people would rather have more seating that's fine with me.

Then the First and Composite power cars should have the same running and maintenance costs as MiddlePower.
MiddleTrailer and  Rear should have the maintenance also like other Middle cars. Extra maintenenace makes sense only for the front car (driver, catering).
EDIT: and about 397 - the only car that has some per/km costs is front pantograph (unpowered). Usually it is the powered vehicles that have non-zero per/km costs, usually proportional to their power in kW. Now the 397 is extremely cheap compared to 801 or 395

I haven't paid much attention to the running costs, knowing that they would probably be changed as and when the pakset is properly balanced. Is there a standard formula for calculating running/maintenance costs? I believe I read somewhere that electricity costs 0.1c/kw in pak128.Britain?

Also - I cannot find any electric engine suitable for pulling the Mk5 coaches? (200 km/h)

This has been pretty well answered already. The only option is the 225km/h-capable Class 91 of 1989, which is outdated by the time Mk5 coaches appear. (Some day I hope to model Class 89 of 1986, a single prototype locomotive designed for 200km/h).  Currently the only single diesel locomotive capable of 200km/h in pak128.Britain is the Class 67; its successor, Class 68, is limited to 160km/h, though I suspect this is to reduce track wear. There were reports of a 177km/h bi-mode locomotive (Class 93) being introduced a year or two ago, but nothing has appeared as yet.

[Matthew]

I haven't paid much attention to the running costs, knowing that they would probably be changed as and when the pakset is properly balanced. Is there a standard formula for calculating running/maintenance costs? I believe I read somewhere that electricity costs 0.1c/kw in pak128.Britain?

The current pakset values are based on the formulae in this spreadsheet.

You not only can use this spreadsheet, you should, since it will ensure balance with the other vehicles in currently active games. It's unfortunate that none of us made you aware of it earlier, especially since you've mentioned that you dislike the .dat side of things.

This prompts two long-term questions.

Firstly, since this spreadsheet is in practice one of the most important standards of pak128.Britain-Ex, wouldn't it be better if it was added to the pakset's Github repository? The values for new vehicles can then be added to it as the vehicles are created. Having an up-to-date spreadsheet might also make it easier to carry out the eventual Great Economic Rebalancing.

Secondly, can you remember which thread you used to learn about pak128.Britain-Ex? If it's a stickied thread, perhaps the spreadsheet should be mentioned there.

[fam621]

Will there be a mini-buffet coach added as currently, there isn't one available with the trains pretty much running as at-seat catering only sets

[jonbridg]

Aha! Thanks Matthew, that's very useful, I'll refer to (and add to, if required) that in future. Which topic was that posted in? I hope it wasn't obvious, I'll be embarrassed if it is.

The 'sticky' topics I primarily used to create the vehicles were:

https://forum.simutrans.com/index.php/topic,15174.0.html Simutrans Extended Development >> Dat file reference for: Vehicles and Ways
The contents of this topic applies to all paksets, not just Pak128.Britain-Ex

https://forum.simutrans.com/index.php/topic,17510.0.html Pak128.Britain-Ex >> Step by step tutorial: making vehicles for the pakset.
This topic is primarily concerned with the graphics side of vehicle creation.

I learnt to calculate tractive effort and brake force from the recent topic on the recalibration of Electrostars; I think there was comment about the running costs of electric trains there.

Looking further the chief problem with the 397's dat is that I used the old 801 dat as a template. The result is that the end vehicles were altered from unpowered to powered without changing the costs. It should have cost the same as an 801, but that's irrelevant now that I have standard formulae to work from.

Will there be a mini-buffet coach added as currently, there isn't one available with the trains pretty much running as at-seat catering only sets

Probably not, in Simutrans the kitchen in the front vehicle is sufficient.

[DrSuperGood]

Apparently people need permission/licence to use my spreadsheet? Honestly do whatever you want with it.

It was generated from an old set of pakset data. As such some entries in it no longer match or are missing. Additionally changing anything inside it will not be used by the pakset automatically as I wrote a custom tool to merge the data into the pakset and then merged those changes into the git. The tool was not designed for an incomplete dataset so I am not sure what would happen if it was just run in its current state.

The balancing was very rough. People have pointed out the formula do not currently work within gameplay, especially for some vehicles with high powers but low power utilization as that was not factored in as doing so from a pure data view is not trivial. Additionally the formula should eventually be made obsolete once more features are added to extended such as wages and energy metering.

Personally for new additions I would recommend gut feeling and just throwing out some reasonably sensible numbers. As mentioned above everything will eventually be rebalanced once/if all the required features get added.

For reference, the spreadsheet was used to try and convert the pakset from speed bonus based balancing to something more physical. This is because fast trains were impossibly expensive to run at the time they were introduced as they were based around large speed bonuses to revenue which no longer existed. It also solved an issue where more powerful vehicles could be more efficient at hauling cargo at slower speeds that purpose built engines. That said this has its own problems with some vehicles being impossibly expensive to run due to having powerful engines but seldom using all the power. it also does not scale well with speed since faster vehicles have lower per km cost due to moving faster.

[Vladki]

The current pakset values are based on the formulae in this spreadsheet.

I can't find any formulae in that spreadsheet. However in my recent modifications of trams and EMU's I kept the ratio of runnigcost to power (without applying gear) to stay the same. E.g. for electrostars and azuma (801) it is 0.15 c/km per kW. Some other modern trains and trams have even 0.10 c/km per kW. Unpowered vehicles have 0 running costs.

For monthly maintenance I just used my gut feeling and kept the original values where they were sensible. Reasoning was: manned vehicles should be most expensive (driver, guard, catering). Powered vehicles also need more maintenance (engine checks and repairs) than unpowered. But I have no exact method, so I made fixes only when it was obvious copy/paste error or typo, etc...

[jamespetts]

I have now integrated Dr. Supergood's spreadsheet - thanks to Dr. Supergood and Matthew for that.

[Matthew]

Apparently people need permission/licence to use my spreadsheet? Honestly do whatever you want with it.

Thank you!

Personally for new additions I would recommend gut feeling and just throwing out some reasonably sensible numbers. As mentioned above everything will eventually be rebalanced once/if all the required features get added.

You modestly pointed out some weaknesses with the spreadsheet. And you're right that it's not necessary for some values: power and brake force, for example, are entered in real-life units and are often obtainable from online sources.

Where the spreadsheet shines, though, is the cost variables. Simucents do not bear any relation to real-life values,^§ but it's important for playability that paks have values that are broadly consistent with the rest of the pakset. For existing paks, DrSuperGood's monumental rebalancing has already achieved that. But your "gut feeling" about "reasonably sensible numbers" is the result of many (hundreds? thousands?!) of hours spent playing the pakset and playing about with the .dat files. There's nothing stopping people doing that, but the spreadsheet's cost formulae are a fast and convenient way to get cost values that are consistent with the interim economic balance. Exporting the values programmatically was very useful for the rebalancing but is less important when adding a handful of new vehicles.

I can't find any formulae in that spreadsheet.

Sorry, I linked to an older version. James has kindly integrated the latest version to the pakset here.

However in my recent modifications of trams and EMU's I kept the ratio of runnigcost to power (without applying gear) to stay the same. E.g. for electrostars and azuma (801) it is 0.15 c/km per kW. Some other modern trains and trams have even 0.10 c/km per kW. Unpowered vehicles have 0 running costs.

For monthly maintenance I just used my gut feeling and kept the original values where they were sensible. Reasoning was: manned vehicles should be most expensive (driver, guard, catering). Powered vehicles also need more maintenance (engine checks and repairs) than unpowered. But I have no exact method, so I made fixes only when it was obvious copy/paste error or typo, etc...

You know much more about both trams and Simutrans than I do, so you've surely chosen "reasonably sensible" values. But now that you have a chance to see the spreadsheet values for yourself, you can see that the formulae do try to use a "more exact method" using several factors. Using DrSuperGood's spreadsheet gives creators an easier and consistent option if they want it.

I have now integrated Dr. Supergood's spreadsheet - thanks to Dr. Supergood and Matthew for that.

Jonbridg's experience suggests that it would be helpful to edit the tutorial stickies to signpost it. If your modpowers allow you to do that, here is some draft text that you could use:

In the .dat reference, under "Vehicle economy" you could add:

Code: [Select]

(If you are contributing vehicles to pak128.Britain-Ex, you can find formulae for calculating runningcost and fixed_maintenance data in [url=https://github.com/jamespetts/simutrans-pak128.britain/blob/master/Pak128%20Britain.xlsx]a spreadsheet in the Git repository[/url]. For other paksets, consult existing .dat files.).


In the graphical tutorial, above "Stage 5 - Compiling the vehicles" you could add:

Code: [Select]

If you add new vehicles, then it is helpful to use use data values that are consistent with the rest of the pakset. You can find useful formulae for calculating runningcost and fixed_maintenance data, and many examples of other values, in your Git repository's [url=https://github.com/jamespetts/simutrans-pak128.britain/blob/master/Pak128%20Britain.xlsx]Pak128 Britain.xlsx spreadsheet.[/url]

§ That may have to be fixed in the Great Economic Rebalancing, but that's off-topic.