[Csantucci]

I just downloaded release 1485, saw the superelevation box and checked it and ...wow! Trains tilt in curves. In fact, due to the fact that tracks are laid horizontally up to now, it's more a tilting than a superelevation box. But it's perfect for our Pendolino HS tilting train!
Two points:
1) in the log file a huge amount of log lines simply showing the names of the track and road shape files appear; it would be nice if they could disappear (even if I know that I can disable the log file
2) I did not notice the tilting when in cab (at least I believe). Everything remained horizontal and vertical as usual.

I also noticed that the oscillations now are more "harmonic" and with lower frequency.

[James Ross]

This feature should have been labelled as HIGHLY experimental (and ideally not included just yet). This is not a feature suitable for release right now.

[JTang]

These features are turned-off by default. A new tab of "Experiments" was added to the Option choice box, where you can set the values for these features. As James said, they are for experiments and discussions only.

[_o_OOOO_oo-Kanawha]

Nice feature in development, but fraud with possible problems.
As in Railworks and the prototype, superelevation needs transitions to and from straight track into curves and back.
In Railworks one has to define the correct easement sections while laying track, as it cannot be added to an existing track without extensive cutting and welding.
Superelevation can be added to existing curves but trains negotiating those curves jolt and sway unrealistically.

I believe in OR, the superelevation in currently in the rolling stock, i.e. it is more of an active tilt. When viewed from the cab, it probably makes for a nice visual effect, as does a gentle cabsway.

When viewed from the outside by discriminating eyes, the lack of "proper" superelevation in the track is noticeable, even if it is only like 4-6 degrees at best.
If OR's devs want to implement superelevation by a global track parameter, the lack of easements in and out of the curves becomes apparent.
In theory, one could compute the amount of superelevation from the radius of the curved track automatically, allowing for some natural easement if there are gently sweeping curves and transitions. But since most track is laid down Marklin/Lionel style, there is a sudden transition from straight into curved track and back, which is visually distracting at speed and becomes more pronounced when superelevation or true cabsway is added.

So the prospect of more realistic train dynamics is promising, but please do not make the same mistakes as RSC. Our OpenRails dev's know better, listen better and in the end implement better.

[Csantucci]

Hi James,
we understand that you all are in an experimental phase. I think it's wise that your experiments, including JiJun's ones are shared with us. I'm sure he will get useful feedback.

_o_OOOOetc: yes, what's implemented now is only a beginning, but it's a good one. And a big point is that what has been done is 100% compatible with existing routes and rolling stock.

JiJun: maybe you already know it, but I would like to remind here that the maximum tilt of a Pendolino is 8 degrees (plus track superelevation). It would be nice if this were be taken into account when defining the various levels of tilting.

[JTang]

If you follow the train and get closer to the front, you will notice the track is indeed tilted. The code generates those tracks dynamically, thus better looking tracks is possible by providing a track profile file. The amount of elevation can be computed, I just need to learn it.

The biggest problem is indeed how to determine a curve consists of multiple curved sections. The current implementation treats every section independently, and will raise and lower the track within one session. It needs a pre-traversal of tracks at loading time to identify long curves.

[Csantucci]

Thanks JiJun, than it is more sophisticated than I noticed.
By the way the strings with the track shape file names appear in the logfile even if logging is disabled.
Here a screenshot of the superelevation:
http://www.interazioni-educative.it/Varie/Sopraelevazione.jpg
I find it GREAT!
I only miss the tilted world view when in cabview.

[JTang]

The newer version has those printing removed. The tilted cabview is more complex than I thought thus I do not plan to tackle this at the moment.

[Csantucci]

OK, thank you JiJun.

[conductorchris]

This is pretty darn cool and I hope it works out. Thanks for working on it!
Christopher

[steamer_ctn]

Hi,

Whilst I haven't seen this feature in action yet, it sounds like a good addition. However, in reality, I would expect that the amount of superelevation will vary from route to route depending on route speed limits, railway design practices, etc.

Therefore is it possible to set superelevation values on a route by route basis?

One possible way to do this might be to add a new parameter(s) to the trk file (outside the last bracket) which will be ignored by MSTS, but read by OR.

The two options are:

i) Parameter which sets the maximum value to be used in the route, eg - SuperMax (3.0)

or

ii) A series of statements that define the superelevation to be used for principle curves in the route, eg Super ( 300, 3.0 ) - where the 1st figure is the curve radius in metres, and the 2nd figure is superelevation in degrees. Curves outside the values defined would use the nearest value defined.

Thus some level of route individualization might be possible without upsetting core MSTS operation.

Is this a possibility?

[gpz]

I think the problem is that the superelevation varies curve-by-curve, not route-by-route. There might be different curves in a single route with and without any superelevation as well.

[captain_bazza]

This is fascinating work.

Cheers Bazza

[Csantucci]

I made some advertising about the superelevation feature in the italian Trainsimhobby forum, and responses are very positive. Some further picture was posted there. But an attentive member noted what JiJun has stated from the beginning, that is that up to now the superelevation is computed track element by track element, so that sometimes superelevation in a long curve does not have a standard behaviour.

[dennisat]

gpz, on 13 March 2013 - 01:27 AM, said:

I think the problem is that the superelevation varies curve-by-curve, not route-by-route. There might be different curves in a single route with and without any superelevation as well.

Not only does it vary curve by curve but also by speed limit. If your route is limited to 50mph there's no point having superelevation for 100mph - you'll get a lot of inside rail wear and flange wear. There's also the problem of transition - you can't switch to superelevation abruptly, your passengers will end up at one side of the train or you may even derail it. This would be particularly acute on reverse curvature. There's obviously a maximum - a train must not fall off a curve when it's stationary! The minimum is of course no superelevation; curves in freight yards and complex station layouts would have no superelevation because of the difficulty of arranging different rail levels between points and crossings.

To be authentic, and rail enthusiasts are nothing if not nitpickers, the route author needs to specify the superelevation when the route is built.

Dennis