Bridge Lengths

[AP]

Just a query - I've discovered that under pakBritain v102, bridge lengths are quite limited - brick to 9 tiles, timber to just 5 tiles. Was wondering why the change was made from the earlier versions? Seems a little limiting, when you compare to what we could do before....

[The Hood]

I agree - actually on my setup, and the sources released on sourceforge there is a bug in the dat file which means that the limit isn't enforced - if it is in the release version I assume James Petts fixed it.  I will increase this limit for the next release, but as it isn't critical that won't be for a while.  Any ideas what is a more sensible limit (eventually there will be different bridge types which have different functions)

[AP]

I don't think their should be a limit, not on length anyways. Much of the suburban rail network in London was built entirely on arched brick viaducts - the limit is the funds of the company rather than the technology.

Presumably the land value meant it was not possible or desirable to buy land for embankments (in urban areas) - using bridges enables industry to locate underneath, or in simutrans, it allows the avoiding of mass demolition (expensive and reduces catchment for stations).

It is arguable that the pricing should vary dependent on the height of the viaduct - a low arch being cheaper than a high one but I don't know if that is possible or indeed if it is already done.

Similarly, I see no tecnnological reason for a length limit on a timber viaduct.

I would however, imagine a height limit on a trestle-type timber viaduct (which is what we have) should exist. Ideal would be if the design mutated to a more expensive brunel-type timber fan bridge design (as used extensively in Cornwall), for higher spans, but perhaps that is me dreaming.

I don't know that a brick viaduct needs that much of a height limit, either; playing simutrans with "alpine" type terrain needs bridges of a decent height (as my first image shows).

Thoughts?

(EDIT- wouldn't it be nice if we could build viaducts on the diagonal and round corners...  )

[The Hood]

Thanks - don't think simutrans has a max. height option.  As for diagonal viaducts/corners - there is the option of building elevated ways, with a maximum height of 1 above ground - they are on the to-do list for pak128.Britain but I think there are some pak128 ones lying around somewhere...

[VS]

Re pak128 - not that I know about them. Maybe Timothy.

[The Hood]

I think they were an add-on (probably Timothy), but it was a long time since I played...

[jamespetts]

There is a hidden complexity with bridge length limits: whilst a brick arch viaduct can indeed be of any sort of limit across land, the same is not true for deep water, which generally requires a spanning bridge rather than a viaduct type. This cannot be simulated in Simutrans, of course, so perhaps retaining the limit, but making it fairly long, would be helpful? The same or a similar graphic could be used for elevated railway track (and road, and tramway...) to use as urban viaducts as AP suggested.

There is further hidden complexity with height differentials in cost: if the feature were to be implemented in Simutrans, it would have to differentiate between high spanning bridges that have no supports other than on the banks (where the cost would not be materially greater, and where there would be no height limit) and high viaducts (where the cost of building them higher would be proportionately greater, and there would be a limit of height, namely the weight that the lower parts of the support columns could withstand).

[VS]

What's so hard - spanning bridges would simply have no height limit Or am I missing something?

With possible restrictions to height and/or length, bridges could be roughly differentiated - spanning with length limit, and supported with height limit.

Re costs, why not keep it simple - pay for every tile of body + every tile of pillar. I must confess that I never count how much a bridge would cost, I simply compare the prices per tile and speed, and simply use the better suited. If I had to consider even the last penny, surely there is more to think than building a bridge

[AP]

whilst a brick arch viaduct can indeed be of any sort of limit across land, the same is not true for deep water, which generally requires a spanning bridge

Although over shallow water, there is no limit, of course (see Florida Overseas Railroad with the famous Seven Mile Bridge).

If I had to consider even the last penny, surely there is more to think than building a bridge

I find very hilly maps quite fun; balancing the desire for a smooth and level railway (=fast, profit) against the cost of raising terrain to form embankments (expensive but finite), against the maintenance of bridges and tunnels (cheaper to build but ongoing cost).

[jamespetts]

Very interesting points. Certainly, an absence of any height limit were taken to be an indication that a bridge was a spanning type, then the cost could well be de-linked from the height whereas otherwise it would be linked. That still leaves the issue of supported bridges and water, however, which I suspect is trickier to solve.

Balancing a smooth and level railway with the cost of construction is indeed a very interesting balance to strike. One way of looking at it is considering whether the line should be designated as a primary route, a secondary route or a branchline, and considering expenditure accordingly. In Simutrans-Experimental at least there is a further factor: if one is building in days of steam, then one should avoid hills at all costs, because steam locomotives in particular have trouble climbing hills. In later times, hills are less problematic, but corners more so (through speed restrictions), in modern times, one has to consider whether to rebuild one's curvy but flat network into one with fewer corners but more hills.

[VS]

James, I think the ground shape continues even under water, as players can see from different water tones. Thus it should be possible to simply calculate absolute height from solid ground anywhere.

Perhaps paying for pillars is too complicated to carry out, if they do not extend below water level. In that case, maybe it would be easier to pay for every tile for height everywhere on the bridge, regardless of pillars' presence.

Anyway, these are just brainstorming ideas.

[jamespetts]

VS,

good points both. Paying for height (unless that option is disabled by setting something like "max_height=0" in the .dat file) would be an adequate approximation and would be easier to implement and understand than pillars. The point about the sea height is also sensible, with the one caveat that it is much harder to build in water than on land, so a high viaduct on land might well be impossible to build in a depth of water equivalent to the valley. Perhaps there is room for a setting prohibiting certain kinds of bridges from crossing areas below sea level?

[VS]

Oh, that's what you meant - not calculation, but reflecting true costs.