[Graham Plowman]

Kris,

Just a tip with this image:

http://imgur.com/ASoTshk

In the real world, that bridge would collapse. There is no chance it would support the signal box.

Always a good idea to find pictures of real locations to see how the bridge was constructed and model what we see, rather than what we think we see.

When engineering is ignored, it really spoils a model.

 

[Graham Plowman]

I would need to see the pictures you are quoting, but real bridges are not built like your model. The main girder beam always supports the whole structure in some way and the girder itself is always supported as both ends as a minimum. Real bridges are not built with the girder hanging as a cosmetic attachment!
I'd suggest that at the height of your bridge, there would probably be issues with the extra supports bending and collapsing without further re-enforcing - because the join halfway up between the two Hornby elevated track supports stacked on top of each other would be a weak/bend point.

I have written an article about one type of girder bridge construction here, but the principals are the same for all girder bridges: Modelling a Steel Bridge - Model Railways On-Line
You will note that I too have extra supports due to the length of the span.

This is a more recent picture of the bridge, a bit obscured in the distance. This pic appeared in Hornby magazine a few years ago:

Kind regards

 

[Graham Plowman]

Hmm the main girder would be the front one of a line of them behind and not simply supported by the single girder at the front, naturally I do not show these extra supports

Yes, that is true. In your case, there would be multiple girders parallel to the outer girders, possibly constructed as multiple spans next to each other. You may or may not see the girders between the parallel tracks, depending on how it was constructed. There would also be cross-member girders spanning between the parallel main girders.

The challenge with this of course, is that you have pointwork on your bridge which would have to support trains transitioning between parallel bridge spans.
In practice, the permanent way engineer would try to avoid placing turnouts on a bridge if possible, because such layouts would impact the construction (and cost) of the bridge. However, if there was a requirement that it had to be done, the engineers involved would need to engineer a viable solution at a reasonable cost.
This is the sort of thing that when it is 'done right' on a model layout, it 'looks right' and adds to the overall appearance of accuracy - makes it look like the modeller 'modelled what they see' rather than what they 'think they see'.