Page 3 of 10

Re: New Rail Types

Posted: 04 Feb 2010 20:26
by Zuu
Zephyris wrote:Oooh, current a railtype callback. Now that's a good idea :D 100% flexibility for grf authors!
100% flexibility for the grf authors, but what information will the users get? And remember that the AI can't read the manual on a website of a set. An AI author can't possible hardcode support for all possible grf logic since both AIs and GRFs are created over time and will remain for long time.

It will be a lot of work to get a spec that is as flexible as a railtype callback, but hopefully the most important and most used aspects can get into the spec as in properties that can be read by OpenTTD and exposed to the users (players and AIs). Otherwise the GRF authors will indeed have 100% flexibility but the users can't predict how the engines will work before having bought them and done extensive experiments with them. Something that the player can do once or twice and then remember when they start a new game but the AI will have to do for every new game.

Now I think this will be my last post about this or I will sound as all I do is complaining. I do like the new tracks and possible future new road types. Hopefully also there will be information for the players and AIs to figure out how the things work before buying/building them in the most important cases.

Re: New Rail Types

Posted: 04 Feb 2010 20:28
by michael blunck
snail wrote: [...] The scheme I thought about was similar to the one Michael proposed. I think it could work for continental European sets, for reasons I will point out below:
Well, we won´t have to work out a detailed scheme beforehand. This can be postponed until new sets are using new rail types, either as stand-alone track sets or integrated into a rail vehicles set. The only thing needed would be a unique label in case compatibility should be achieved.
snail wrote: What do you guys think?
Well, yes. Only some comments ATM. Bear in mind that each historic/realistic/whateveryouwanttocallit set may have different needs. Insofar, numbers would differ between countries in any way.

IMO, your NG and "early" SG axle load values are too high, at least for middle European countries (don´t know about french track), max axle loads were significantly lower. I.e. most Bavarian steamers around 1860 had max axle loads of 12t, 14t around 1880 and only since 1900 steamers with axle loads of 16t were built. Likewise even in the 1940s, axle loads of 20t were quite uncommon with the DRG, nearly all track only allowed 18t (at best). Lastly, even today, the RIC "standard" for newly built track in Europe is 22.5t ("D4"). To my knowledge, only Sweden is using 25t track ("E").

I´d like to repeat myself, there´s no need to come to a conclusion about characteristics of certain track types, we´d only have to define some labels.

Well, continuing:
snail wrote: * Early tracks would have a relatively high speed (120 km/h), but that's because the Crampton engines of 1845 were already capable of running that fast.
Question is if they really did. :cool: Bavaria also had early Cramptons but neither am I able to find such high max speeds (instead 70/80/90 km/h) nor would it be a sound assumption that passengers trains would indeed having travelling by that speed.
snail wrote: * All track types can have catenary; the latter is actually pre-built for high-speed tracks (in order to save one slot). Each rail type will have its own catenary graphics;
Well, maybe not all, but this is a really good idea, having a number of different catenary graphics.
snail wrote: * A combination of third rail and catenary is possible on wooden-sleepers tracks (this was the case in reality), allowing 3rd-rail and catenary-powered trains to share the same tracks;
Well, quite uncommon, n´est-ce pas?
snail wrote: * Rack rail could be interesting too, but how to implement it in TTD? In reality, the rack system allows steeper grades, but in the TTD world, all uphill tracks have the same steep grade...
That´s a problem which would indeed need some thought. OTOH, a swiss railway set would unquestionably need rack vehicles. :cool:
snail wrote: Another rail type we didn't list is threephase catenary. The Italian set I'm thinking of will definitely need it.
Indeed there are many different (and incompatible) electrification systems, best thing would be to handle them only graphics-wise. See e.g. my old post for an overview of Swiss electrification systems.

And yours! Déjà-vu somebody? :cool:

regards
Michael

Re: New Rail Types

Posted: 04 Feb 2010 20:49
by Badger
michael blunck wrote:
snail wrote: * A combination of third rail and catenary is possible on wooden-sleepers tracks (this was the case in reality), allowing 3rd-rail and catenary-powered trains to share the same tracks;
Well, quite uncommon, n´est-ce pas?
There are a couple of places in the UK where passenger trains swap from catenary to 3rd rail and vice versa, and a few stretches of line where both modes are present.

Re: New Rail Types

Posted: 04 Feb 2010 21:15
by michael blunck
Badger wrote:
michael blunck wrote:
snail wrote: * A combination of third rail and catenary is possible on wooden-sleepers tracks (this was the case in reality), allowing 3rd-rail and catenary-powered trains to share the same tracks;
Well, quite uncommon, n´est-ce pas?
There are a couple of places in the UK where passenger trains swap from catenary to 3rd rail and vice versa, and a few stretches of line where both modes are present.
snail on Aug 21, 2006 wrote: The problem is, that in France there were many engines that could run both on 3rd rail tracks and under catenary (had both pantos and 3rd rail shoes). So what I was thinking is, to do two sets of sprites for those engines, one with pantos up (to display when the engine is under catenary) and one with panto down (when it drives on a 3rd rail track).
regards
Michael

Re: New Rail Types

Posted: 04 Feb 2010 21:26
by Badger
Ah, that would explain the french. Ta!

Re: New Rail Types

Posted: 04 Feb 2010 21:39
by michael blunck
Badger wrote: Ah, that would explain the french. Ta!
Hehe. :cool:

In fact, having both 3rd rail and catenary (with highly differing voltages!) with the same track is uncommon because of severe technical drawbacks, which I´ve explained somewhere on this or the German forums some years ago, IIRC.

The other way round (snails post from 2006) makes more sense to me, though.

regards
Michael

Re: New Rail Types

Posted: 04 Feb 2010 21:48
by Zephyris
I predict some set vs set fighting!

Re: New Rail Types

Posted: 04 Feb 2010 22:17
by Snail
Fighting? I'd rather call it a constructive discussion (albeit quite animated :D )

Memories, memories... I eventually ended up drawing the engines I was talking about, both with pantos up and pantos down :p

Perhaps we could avoid having both 3rd rail and catenary on the same square: we could have engines, however, that should be able to keep running from a catenary-powered square to a 3rd-rail-powered one, although at a *very* limited speed (like 5 km/h). Let's think of this: if a line of track is electrified with catenary up to some point, and with 3rd rail from there on (this could happen at stations, for instance), an "amphibious" engine should be able to reach the last catenary-powered square, slow down to a really low speed, transition to the first 3rd-rail-powered square, change the sprite (visually, the panto goes down), and then speed up again. Of course, other electric, non-amphibious engines could not proceed further. Diesels and steamers would instead continue as if nothing happened.
This actually happened in the Maurienne mountain line in southeast France.
Would this be achievable?
Michael Blunck wrote:Well, we won´t have to work out a detailed scheme beforehand. This can be postponed until new sets are using new rail types, either as stand-alone track sets or integrated into a rail vehicles set. The only thing needed would be a unique label in case compatibility should be achieved
Well, maybe it's a bit too early to establish a new rail types scheme with all the details. What I meant was that, if we had at least some rough guidelines, we could prevent weird effects such as having narrow gauge trains running on monorail tracks, as Foobar was mentioning.
Every set would have its own peculiarities (rack rail in the Swiss set -I definitely agree with Michael, this is a must- , threephase catenary in the Italian set, and so on), but there definitely are some widespread track types, such as third rail, high speed etc. that could be coded the same way across the sets. This would ensure some degree of compatibility.
I agree that using the same labels might be a good way to solve this.

Concerning the Crampton engines, well, it's true we don't know whether the max speed of 120 was ever reached during normal service. Maybe we could code such engines with a max speed of 100 km/h or even 90. However, their low power would cripple them so much that such a high speed could be reached only when pulling very few cars on a straight, flat track. Plus, they would be really expensive to buy.

I agree that a max axle weight of 16 tons for 1830 would be a little bit too much. However, we are "limited" to 16 track types including monorail, maglev and the various electrification systems: this doesn't open up much space to define new track types. We have to keep in mind that, for instance, PLM's class 121 engines of the 1870's had a max axle load of 14.8 tons, so they will need a track they can run on. And the Atlantics of 1900's max axle weight was about 18 tons so, again, they will need a suitable track.
To be more accurate, we should start with a max axle weight of about 12 tons in 1830, then define another track with a max of 15 tons in about 1870, and then one with a max of about 18 tons around 1900. Then we would need a max of 20~21 tons around 1930, when the big steamers start getting introduced (the French 241's max axle weight was a whopping 20 tons). This would cause a wide proliferation of track types; that's why I proposed a simpler scheme.

Of course, each set could define its own tracks with their specific limits, but it would be nice if there were some degree of uniformity at least across railroad-wise similar countries (it would improve compatiblity).

Re: New Rail Types

Posted: 04 Feb 2010 22:23
by Zephyris
Fighting? I'd rather call it a constructive discussion (albeit quite animated :D )
Exactly what I mean really really; excited and enthusiastic fighting :)

Re: New Rail Types

Posted: 05 Feb 2010 02:43
by Eddi
Snail wrote:Perhaps we could avoid having both 3rd rail and catenary on the same square: we could have engines, however, that should be able to keep running from a catenary-powered square to a 3rd-rail-powered one, although at a *very* limited speed (like 5 km/h).
i can't see this being either realistic or useful. I haven't seen trains switch from catenary to 3rd rail, but between phase-separated catenary systems, trains travel at high speed through an isolated section. getting to a standstill at that point would be fatal.

Re: New Rail Types

Posted: 05 Feb 2010 02:48
by NekoMaster
HOw do I get this to work? I added metro tracks and narrow guage rails but they only replace existing rail types, im using OpenTTD r19009M (A patched build I made)

Re: New Rail Types

Posted: 05 Feb 2010 02:53
by Eddi
you don't. the feature is not complete yet, and the grfs are not updated to work with this.

Re: New Rail Types

Posted: 05 Feb 2010 02:56
by NekoMaster
Eddi wrote:you don't. the feature is not complete yet, and the grfs are not updated to work with this.
Then how are people in previous posts getting it?

Re: New Rail Types

Posted: 05 Feb 2010 04:39
by ostlandr
The nightlies now have the ability to vary the default steepness. in 1% increments. :twisted:

Actually, the cool thing about rack locomotives is that their TE isn't limited by their weight. Although there are rack diesels also, this thread may help explain it: http://www.tt-forums.net/viewtopic.php?f=26&t=31980

FYI I have seen pictures of rack locomotives where the circle of the crank pin was actually larger in diameter than the diameter of the drive wheel tread! The drive rod ends actually dip below the surface of the rail. That effectively "gears down" the locomotive without adding complex gearing (the rack pinion(s) would be on the axles between the bearings) in order to sacrifice speed for TE.

American style geared locomotives (the Shay, Heisler and Climax) use reduction gearing and powered trucks to both sacrifice speed for TE and make 100% of the locomotive's weight available for adhesion, but they are still limited by weight on drivers, which is limited by the axle load the track will bear. This is usually very low for primitive logging railways, although some large Shays such as Cass Scenic's "Big Six" were used on steep, winding coal mine branches- this one worked a NINE PERCENT grade by adhesion! http://www.cassrailroad.com/shay6.html
michael blunck wrote:[rack railways]
maquinista wrote: Some trains can increase their tractive effort if there is a rack. This could be done with a new callback.
Well, point is that vehicles for rack rail don´t have higher TE than non-rack vehicles. It´s simply the friction coefficient (rack/pinion) which is higher than that for wheel/rail of a non-rack vehicle.

I wouldn´t like to see a rack locomotive in game with an absurd high TE value. What´s needed here is simply a different kinetic model with µ~1 instead of 0.3 or so. Only for mixed rack/normal rail one could think about the need of a new callback, but o/c not for changing TE.

OTOH., it´s questionable if there´s any need for rack rail in TTD with its equal slope gradient. Would it be possible to have rails on steep slopes?

regards
Michael

Re: New Rail Types

Posted: 05 Feb 2010 04:49
by NekoMaster
If we did include a rack rail, it would have to be so that when you have a rail go up a slop with out "stair stepping" it, it would act as a rack rail

stair steping

Code: Select all

""\
    ""\
        ""\
            ""\
                ""\
Strait slope (steep)

Code: Select all

\
  \
    \
      \

Re: New Rail Types

Posted: 05 Feb 2010 05:15
by ostlandr
[speed limits]

Here's an article from 1898 giving the tech specs and details of US vs UK high-speed rail. The New York Central's famous "999", which set the world speed record in 1891 with an unofficial top speed of 112 mph/188 kph, is pitted against the GWR's "Worcester."
Top speeds of 93 mph/150 kph are entirely reasonable for this era- if the track, the grades and the curves allow.

http://www.catskillarchive.com/rrextra/stamen.Html

[rail type cost vs speed]

I understand that "vanilla" OTTD doesn't really care about track speed limits, but for users of UKRS/NARS/NARS2 and similar sets, being able to lay cheap, low-speed track on branch lines as opposed to expensive high-speed track on main lines (see article linked above) would be fun and useful.

[rail (and bridge?) type axle weight limits]

Interesting idea. This would link in with the above. You lay a branch line "on the cheap" in the early days of steam, then later on you want to buy heavier locomotives. Suprise! They aren't available in the depot for this rail type. You'd have to spend some capital to upgrade to the next heavier rail type, using that handy-dandy button already on the toolbar.

And possibly when the new locomotive gets to that old wooden trestle, it turns around and you get the message "train X is lost." The bridges will also only support certain weights, as opposed to just having speed limits. (Is this even possible in the code?) Maybe just have certain bridge types not be upgradeable to certain rail types? Or get really evil and have the bridge collapse. :twisted:

Re: New Rail Types

Posted: 05 Feb 2010 06:27
by NekoMaster
Irl, theres still A LOT of wooden bridges, which yes they have a low speed limit, but they hold even heavy stuff like todays modern diesels and electrics

Re: New Rail Types

Posted: 05 Feb 2010 06:30
by Snail
Eddi wrote:
Snail wrote:Perhaps we could avoid having both 3rd rail and catenary on the same square: we could have engines, however, that should be able to keep running from a catenary-powered square to a 3rd-rail-powered one, although at a *very* limited speed (like 5 km/h).
i can't see this being either realistic or useful. I haven't seen trains switch from catenary to 3rd rail, but between phase-separated catenary systems, trains travel at high speed through an isolated section. getting to a standstill at that point would be fatal.
Well, I haven't seen it either, but I'd imagine it would be done at low speeds. When switching between catenary and 3rd rail, the pantograph needs to be raised (or lowered) and the 3rd rail shoes lose (or gain) contact with the 3rd rail. If this is done at high speed, I imagine it could result in damage to the structures or the locomotive parts...

Then again, I'm no engineer so I could be wrong. My point is, some sets would definitely need amphibious locomotives which could operate on both 3rd rail and catenary-powered tracks.

Re: New Rail Types

Posted: 05 Feb 2010 08:44
by michael blunck
Snail wrote: My point is, some sets would definitely need amphibious locomotives which could operate on both 3rd rail and catenary-powered tracks.
Which should be quite easy by properly setting that "compatible rail type list".

OTOH, taking a look at Peters specs, you´d have to specify the loco for a single "track type" ("electrified rail" or "3rd rail"), but only with a (not yet existing) callback you could make the loco react upon the actual track type, lowering pantos or whatnot.

regards
Michael

Re: New Rail Types

Posted: 05 Feb 2010 09:35
by peter1138
How to make new rail types available: clearly the old "there must be a vehicle of the new type" method is lacking some what. I'm thinking of a monthly availability callback, called for each rail type, which would allow a lot of freedom on availability:

* the current date
* what other rail types are available
* what other rail types have vehicles are available
* etc

The callback would return a yes/no answer, or if it fails would use the old "is a vehicle available" method. Alternatively it could be a property, just another list of rail type labels, where "if this label is available, make this rail type available too". But that's less flexible.