Bridge question for the engineers

[The Fatadder]

I have a Rail over river bridge to build on my layout and have a question.

 

The crossing is wider on one side of the bridge than the other, I am building a guarded bridge, should the two beams be the same length or the minimum required to span each side?

 

Hopefully the photos make it a bit more clear

[eastglosmog]

In my experience, what would normally happen in your situation is that a new, straight, channel would be cut for the river, there would be parallel abutments for the bridge and girders of the same length.  However, exceptions are to be found (as always),

[The Fatadder]

Something like this?

Shaded in blue

 

Or in a less likely situation would you get something like this, with two equal beam lengths but it’s the trusses angles between them, I assume not given presumably it would be a lot weaker

Edited April 28, 2018 by The Fatadder

[giz]

Two unequal spans would be unlikely because if they were both the same section size they would deflect by different amounts resulting in a slope across the width under load.

 

In theory one could be made stiffer than the other to compensate but I've never seen this done as it would be difficult to achieve under all loading conditions.

 

Your first solution would be cheaper in real life but only if the river flow wasn't great enough to cause scour problems. For a faster flowing river the skew option would probably be prefered. The longer span wouldn't be weaker as deeper/heavier beams would be used to compensate.

Edited April 28, 2018 by giz

[mikeg]

From a design perspective the abutments would be built away from the river parallel to the line of the river, thus giving some protection from any scouring during any flooding that might occur.

 

From observation of local bridges in the Lea Valley the distance from the river edge to the abutments is in the region of 6-8ft.

 

Hope this is helpful

 

Mikeg

[eastglosmog]

Something like this?

Shaded in blue8004DBE6-AC6A-4904-BB1A-3079A176543A.jpeg

 

Or in a less likely situation would you get something like this, with two equal beam lengths but it’s the trusses angles between them, I assume not given presumably it would be a lot weaker

Second option is more likely - girder sections would be made large enough to cope with the stresses.  Nowadays, the opening would be designed to cope with peak once in a 100 year or worse food flows, but back in Victorian times they would have made a best guess how big it should be based on what they could see locally.  Quite possible that the new channel would have been lined with stone and so have less resistance to flow than the old natural earth and weed lined channel - this would increase conveyance and reduce scour.

[The Fatadder]

Thanks all

 

On one side the channel has been cut out with a brick retaining wall, which would agree with the faster flowing river. Will go with option 2

[Il Grifone]

The Victorian method generally was to calculate/guess the loading and then add 50%.

 

This method failed dramatically with the Tay Bridge which was designed for 60 m.p.h. wind pressure (after seeking advice on this factor). 95 m.p.h. wind pressure brought it down. (Not being built to specification and with unsuitable materials liable to fatigue failure didn't help. In fairness, this was not fully understood at the time - hindsight is 20/20.)

 

We know a lot more now and the margin has been reduced to 10%.

 

I doubt that a new channel would be cut at a right angle as shown, as this would produce severe scouring.

 

The studies of André Chapelon into fluid dynamics were revolutionary in this respect.

[sharris]

If you do a google picture search for "Skewed bridge trusses" you will find plenty of examples that might be of help. 

Edited April 30, 2018 by sharris

[The Johnster]

Remember that, if you are altering the flow of your river, then the bend next to that building will need a retaining wall to stop the river eroding the bank and undermining it.  The main channel of a river always runs to the outside of a bend and undercuts it while matter is deposited on the inside of the bend; remember ox-bow lakes in school?  Bet you never thought that bit of your education was ever gonna be of any use, didja?

Edited May 1, 2018 by The Johnster

[The Fatadder]

All very interesting thanks

 

I have a retaining wall ready to install down the whole length of the river, though the idea was originally more focused on protecting the river from clay contamination. Now I know it surveys a dual purpose

[The Johnster]

Put the retaining walls of the erosion side; they are not needed on the deposition side.  Check out the Severn Valley Railway's ordeal at Stern's Crossing to illustrate what can happen!

[eastglosmog]

If you are modelling an industrial area, it would be quite in order to put the walls on both sides of the river - see the River Taff through Merthyr Tydfil for example.  A river in full flood is capable of eroding both sides of its channel and the railway company would be likely to want to protect the bottom of the embankment leading to the bridge, even if they did not need to build a retaining wall to steepen the overall slope (the space you have between river and railway is limited).

[Shadow]

Some diagrams here, rail over road though.

 

http://www.lusas.com/case/bridge/currie_road_bridge.html

[KalKat]

f I recall correctly, the River Brent where it passes under the Warncliffe Viaduct (Hanwell Viaduct) is carried in a brick lined course of a U shape cross section.

 

 

Emma

[John_Miles]

If you do a google picture search for "Skewed bridge trusses" you will find plenty of examples that might be of help. 

 

Skewed doesn't necessarily imply girders of different lengths. Post #4 gets it right, girders of different lengths lead to large secondary moments (i.e. bending across the span rather than along it) which most bridge designers would prefer to avoid.

[sharris]

Skewed doesn't necessarily imply girders of different lengths. Post #4 gets it right, girders of different lengths lead to large secondary moments (i.e. bending across the span rather than along it) which most bridge designers would prefer to avoid.

I was actually thinking of post 3 picture 2 when I wrote that - equal lengths but offset.

[eastglosmog]

Whilst equal length girders was by far the most common situation, it was not universal.  Attached photo shows where the Up line of the Merthyr Tredegar and Abergavenny crossed the entrance to Llanelly Quarry in the Clydach gorge - markedly differing lengths were required for the inner and outer girders.

Actually, even on skew bridges with same length girders, deflection would differ on either side, because the point of application of the load on each girder would be at differing distances from the bridge abutment, giving differing deflections as the loco passed over. 

PS I have a suspicion that equal length girders may also have been used as calculations would only be needed for one girder not multiple girders - an onerous task in pre-computer days.

Edited September 3, 2022 by eastglosmog
Restore photo

[giz]

 

Actually, even on skew bridges with same length girders, deflection would differ on either side, because the point of application of the load on each girder would be at differing distances from the bridge abutment, giving differing deflections as the loco passed over. 

The FE analysis linked to in post 14 shows this quite well, see the displacment contour plot. However, since delection is related to the cube of the span, unequal spans will have a much greater effect so I suspect would be avoided if at all possible.

 

I just had a look through Bridges for Modellers by L.V. Wood and whilst there are plenty of examples of skew bridges there are none in the book with unequal length girders.

[The Johnster]

If you are modelling an industrial area, it would be quite in order to put the walls on both sides of the river - see the River Taff through Merthyr Tydfil for example.  A river in full flood is capable of eroding both sides of its channel and the railway company would be likely to want to protect the bottom of the embankment leading to the bridge, even if they did not need to build a retaining wall to steepen the overall slope (the space you have between river and railway is limited).

 

Many rivers are 'canalised' in urban areas in this way, but it didn't stop me spotting a kingfisher on the stretch opposite Merthyr Bus station by Rhydycar Centre some years ago.  Another feature of this stretch, again common in urban and especially industrial settings where water is taken off for industrial use, is a weir; any such stretch where water is held back, effectively dammed, from it's normal flow, has a very different effect on the erosion of it's banks, which is more or less curtailed altogether.  Often both banks on such stretches are built up and the river's course artificially straightened in order to free up land on each side for building factories or whatever.