Guidance needed on plate girder bridge for Eastwood Town

[gordon s]

I would be very grateful for some advice on a feature I plan to incorporate on my layout, Eastwood Town which is currently under construction (again!).  The two outer tracks are raised above the four inners by 70mm to allow a line to cross over the main running lines to a goods storage area.

 

As you would expect there was no problem in creating this in Templot, but now having looked at it, I will need some assistance.  The whole line is on a curve and from memory plate steel bridges on a curve were probably built in straight sections and then laid to form a series of straight trackbed sections with the line remaining curved. First thoughts with my limited knowledge would suggest this will need some realignment of the two pairs of low level lines to possible accommodate some more supports as I fear the unsupported spans will be just too wide.

 

Were bridges ever built using curved sections?

 

Does anyone have some ideas how this could be accomplished or will some major rework be necessary?

 

Of course if all else fails, I could just ignore the fundamentals of mechanical engineering and fabricate something that will do the job, but totally freelance....

 

I don't really want to go down that route unless it's impossible to resolve, so would be grateful if anyone could assist and come up with a workable and feasible design.

 

[allan downes]

Hi Gordon.

 

I would doubt very much that they ever built plate girder bridges on a curve - they would rather have detoured the approach roads rather than attempt that and spanned the railway if possible at the narrowest point. Certainly many bridges skewed across the tracks at an alarming angle but after having seen your awesome layout and what you are capable of, I'm sure you would be forgiven for using a little modellers licence and do what YOU think is best and I, for one, would certainly not be complaining !

 

Cheers.

Allan

 

PS :any chance of a few pictures Gordon ?

[APOLLO]

Knew I'd seen a curved plate girder bridge. Barnstaple Jcn, Ilfracombe line.

 

http://farm9.staticflickr.com/8529/8557626112_3748f49329_z.jpg

 

 

 

Brit15

[2ManySpams]

Gordon,

 

The closest tightly curved gider bridge I can think of was the LSWR one across the River Taw in Barnstaple. It started off running broadly perpendicular on one bank but by the time it reached the other it had turned through 90deg.

 

It comprised of a series of shortish girders. http://www.robertdarlaston.co.uk/Railways2_files/image088.jpg

 

The piers remained parallel to the bank,   

Knew I'd seen a curved plate girder bridge. Barnstaple Jcn, Ilfracombe line.

 

http://farm9.staticflickr.com/8529/8557626112_3748f49329_z.jpg

 

 

 

Brit15

Great minds think alike!

[2ManySpams]

Note however that curved girders are nowhere near as strong as a straight girder as so need lots of bracing between pairs of beams. They therefore tend to be used in short spans or require deeper / wider beam sections compared to a straight beam. 

 

Here's the science bit: 

This is because the top section of a (simply supported) beam under load is in compression whilst the bottom section is in tension. If you take a plastic ruler and curve it and then push along its length (mimicking the compression section) the curve will increase. If you take a plastic ruler and curve it and then pull along its length (mimicking the tension section) the curve will decrease. The curved beam therefore has very complex forces and turning moments working on it and effectively it wants to rotate out of vertical. Not great for supporting trains!

[allan downes]

Downes wrong again !!!

[unravelled]

I think that a pair of bridges would work. They would need to be almost double track width, to cope with the curve. The left hand (as plan) girders would need to be longer than the right, probably with abutments/pillar faces tangential to the adjacent track they are crossing. Looking on the web, I did find a 3 span bridge crossing 6 tracks, the ECS flyover at Old Oak Common. Although this is a straight bridge, it has to be wide enough to accommodate a reverse curve in the track. Slightly North West of OOC, there are a couple of curved single track bridges taking the Overground over the main line at Willesden Junction, though these are practically at right angles to the tracks below.

 

Hth

 

Dave

[jukebox]

Gordon;

 

I would have thought Holloway Bank has a very similar alignment and configuration to what you need:

 

http://www.signalbox.org/forum_files/kxdistants%5BPA%5D.jpg

 

A pair of girders could span the uppermost three tracks, whilst a flyover type arrangement could get you over the lower pair at a shallow angle...

 

 

Cheers

 

Scott

 

edited to add dodgy photoshop effort

[92220]

Hi Gordon,

I'll bring round a copy of "Bridges for Modellers", which has endless useful information about different types of bridge, prototype practice and variations from that, and also how to build them in model form.

Best wishes,

Iain

 

Typo edited...

[The Stationmaster]

Downes wrong again !!!

I wonder?  As far as I can recall every 'curved' steel girder bridge I've come across has consisted of a series of straight sections but the shorter those sections are the less obvious it is that they are straight.  It would be interesting to see an engineering drawing of the Barnstaple bridge to see how 'curved' the sections are.  Walnut tree Viaduct in South Wales had a very distinct curve at the south end (pics below) and the track also curved slightly at the north end but was built entirely of straight girder sections.

 

http://kriscarter.files.wordpress.com/2011/03/wt05_1958_midlandexplorerboy.jpg

http://archive.rhondda-cynon-taf.gov.uk/treorchy/image.php?i=40174&r=2&t=4&x=1

http://www.ebay.co.uk/itm/WALNUT-TREE-VIADUCT-WALES-2-2-8-1959-/200937737039

http://treharrisdistrict.tech-hosts.co.uk/railwaysphotos.php

[2ManySpams]

I wonder?  As far as I can recall every 'curved' steel girder bridge I've come across has consisted of a series of straight sections but the shorter those sections are the less obvious it is that they are straight.  It would be interesting to see an engineering drawing of the Barnstaple bridge to see how 'curved' the sections are.  Walnut tree Viaduct in South Wales had a very distinct curve at the south end (pics below) and the track also curved slightly at the north end but was built entirely of straight girder sections.

 

http://kriscarter.files.wordpress.com/2011/03/wt05_1958_midlandexplorerboy.jpg

http://archive.rhondda-cynon-taf.gov.uk/treorchy/image.php?i=40174&r=2&t=4&x=1

http://www.ebay.co.uk/itm/WALNUT-TREE-VIADUCT-WALES-2-2-8-1959-/200937737039

http://treharrisdistrict.tech-hosts.co.uk/railwaysphotos.php

 

Hi Mike, I'll have a look in my LSWR books to see if I can find any plans or descriptions. I know what you mean about most curved bridges consisting of straights (for the sound engineering reasons explained in an earlier post), this is the approach I used on the metal viaduct for Black Country Blues. Given what is 'normal' and 'everyday' I'd be inclined to go for a series of short straights forming a curve in this case. 

[The Stationmaster]

Further delving has rediscovered the film that was shot from a train crossing the bridge in early years and it does show that some of the girders are definitely curved - but they are very short -

 

[2ManySpams]

I'm looking for a photo in one of my books, colour I'm fairly sure, showing a bloke sitting on one of the beams with a cutting torch. I recall that it may show the track bed removed but can I find it? 

 

Anyway lawns need cutting before the rain and thunderstorms arrive....

[2ManySpams]

Further delving has rediscovered the film that was shot from a train crossing the bridge in early years and it does show that some of the girders are definitely curved - but they are very short -

 

 

Cracking video Mike - the signalman in the box at Barnstaple Town is very enthusiastic with his flag!!

[Arthur]

I wonder?  As far as I can recall every 'curved' steel girder bridge I've come across has consisted of a series of straight sections but the shorter those sections are the less obvious it is that they are straight.

]

I'd agree Mike, I'll come back to the Barnstaple bridge. It is possible to roll curved sections but it is more costly, similarly, you can fabricate curved section by riveting or welding the flanges to a curved plate, the web, in which case the flanges need to be rolled in two planes to curve them. Again costly, and other than the aesthetics of a nice curve, I can see no reason to bother. Curved sections would be shorter than straight ones, increasing the number of support piers needed under each section joint, cost again, plus a cluttered underside, okay spanning an estuary, not so good spanning other tracks. Also, whereas with straight sections, a pier at each end would be sufficient, with long curved sections would you also need intermediate piers?

 

I suspect that the Barnstaple bridge is wrought iron, not steel. Wrought iron, being essentially hand made, was produced and rolled in relatively small amounts, so you don't have the ability to produce very long, straight, sections anyway. Much wrought iron section was fabricated too, look at the number of rivets, so the increased cost of producing a curved bridge vs straight, in wrought iron, would be much less so than in steel. In steel however, long, straight and deep sections were relatively cheap, the cost of curving them would seem an unnecessary luxury.

 

So, with the caveat of 'never say never!!, I would say that a large bridge with curved sections would be rare.

 

I don't know when the last wrought iron bridge was built in this country, before 1900 I would suspect.

 

So for Gordon's bridge, I'd say it would be steel, with three or so long straight sections spanning the running lines with minimal piers, installed where they could be fitted.

[2ManySpams]

Agree completely Arthur.

 

Despite the fact there seems to be a curved(ish) viaduct at Barnstaple, I would also go with a series of offset curves as you suggest.

 

Modern steel fabrication plants with CAD/CAM can produce just about any shape that architects can come up with now....for aesthetic reasons, they are however less structurally efficient and more costly! 

[ozzyo]

Crossing.jpg

 

Hello Gordon S,

 

could you move the point back a bit and get a straight section over the first two roads then a curved section from what would be the pier between the two sets of running lines. This should give a shorter section of curved line on the bridge. The down side would be that the curve into the goods yard would be sharper.

 

I've had a bit of a play at drawing two bridges on your track plan. The first one has a very wide start to the first bridge (I'm not that keen on that), the second bridge is not that bad in being just about being parallel between the two sides.

 

If you could move the track centres on the first pair of bottom tracks to get a large pier between them. You could then get the first and second bridges more parallel (the green lines) between the two sides.

 

If you could I'd try and get the bridge to go across the four bottom lines to have the inner abutment between the two points on the inner curve. If you did that you could have a curved squew masonry bridge.

 

Most of the time the real railway would only use curved bridges if there was no other option. Forces on the outer rail etc. As you can see on the footage of the Barnstaple bridge the rails would be laid on baulk limbers. On a bridge like this going over four tracks the running line may have a "check" rail (this is not the correct name, I can't remember it at the moment) on both the inner and outer rails of the curves. At the same time it could also have a "check" rail (see above) on the out side of the running rail.

 

These tended to have a larger run in than you would see on a point, the middle ones sometimes going to a vee between the running rails. The purpose of these rails was 1st. to bring any stock that was off the road back into something like gauge, and 2nd. it also help to keep the track more ridged sideways. These rails could also be seen on straight bridges some times with only the inner rails and some times with the outer rails as well. I don't recall any with just the outer rails, that's not to say that it didn't happen.

 

OzzyO. 

[gordon s]

Thanks for all your input guys.  I'm going to have a look at the plan again this afternoon.  I can possibly straighten things out a bit although I suspect this will mean going below my 36" radius rule...

 

As it's a freight only route it may not be a problem although 9F's will need every inch possible.  Running a curved trackbed is a doddle.  It's just the length of the spans and size of piers that will break all the laws of mechanics and if possible I want to avoid that.

 

Let me fire up Templot and see what comes out....

[gordon s]

No time like the present, so a bit of tweaking and acceptance of a 31" radius B8 turnout from the outer loops has realigned the curve to make it a much sharper angle across the lower level main lines, which will allow much better positioning of the support pillars and much shorter spans.

 

Thanks again.  I knew it wasn't right and had to be done, but thankfully it has turned out fine.  I'll do something similar with the second set of bridges that will allow access from the opposite direction. 

 

 

 

[unravelled]

This plate girger bridge spans the canal on the approach to St Pancras. Measuring, using google maps, gives a girder length of about 25m, and a span of about 20m. Brick counting suggests a depth of about 2.5m. Double track.

 

 

 

Dave

[ozzyo]

No time like the present, so a bit of tweaking and acceptance of a 31" radius B8 turnout from the outer loops has realigned the curve to make it a much sharper angle across the lower level main lines, which will allow much better positioning of the support pillars and much shorter spans.

 

Thanks again.  I knew it wasn't right and had to be done, but thankfully it has turned out fine.  I'll do something similar with the second set of bridges that will allow access from the opposite direction. 

 

Crossing 2.jpg

 

Sorry to say but the upper junction on the L/H side with a point on the bridge does not work for me, if its in a scenic section try to bring the point to section 9/3 or 7/2 in 7/2 it would be over solid ground not on a bridge.  Then try bringing the point that's at  12/1 to 14/1. You may have to move the king point to section 5/9 to get this.

 

OzzyO.

[jamie92208]

On Green Ayre we've got Greyhound Bridge which was on a similar curve to Barnstaple with a 10 mph restriction and check rails to cross the River Lune.  It consisted of 8 spans each of about 60'.   I've got the original plans and the girders were 6' deep with the track height supported about 2' up off the bottom.  The bottom flange was 18" and the top flange 2'3" to create a walkway.   Built in 1911 it's made up of several small plates usually 9' by 6' with reinforcing T girders every 3'.  These are made larger at the joints between the plates.   As at Barnstaple, and on your layout the piers were made to always be in line with the river (or railway traffic) flow.  This means that each girder was in fact of different length and no section had both sides parallel.   On the model I have had to increase the width a bit to allow for overhang of vehicles at the ends and in the centre.   If you go through my layout thread you should see some pictures on the way.   I would go for a three span bridge.

 

What is interesting is that the upper and lower flanges increase in depth at the centre of ech span from 1 ply (of half in plate) to up to 7 ply.   This is visible on the lower plate but on the upper one they put L girders along the edges and then filled it with ashphalt to create the walkway so it looks to be of constant depth.

 

Jamie

[Siberian Snooper]

Hi Gordon

 

The approach spans to Mr Brunnel's masterpiece the Royal Albert Bridge are short sections of girder bridge, I cant remember the radii of the approaches, but a severe restriction applies. There are also lots of short span girders on the approaches  to London Bridge Station from the Thames direction.

 

Cheers SS

[gordon s]

Thanks Oz for your comments.  I've probably misled you a little when talking about the second access and the section of the plan I showed probably added to that confusion.  

 

All the storage roads (upper - goods and lower - passenger stock) are effectively off scene and whilst I had originally thought in terms of a second bridge, I now realise a short tunnel would be a far better way of dealing with that section.  The pointwork over the lower tracks was dictated by the curves and my own minimum radius choice.  As it is the turnout will have to have the Tortoise motor remote mounted and controlled via a tube and wire connection, as there is no room beneath that turnout to mount a motor without clashing with the lower tracks.

 

I guess we all pick up on design cues from previous layouts and this has always stuck in my mind.  It all stems back from childhood days sitting in a Quad Art on it's way to KX at Holloway bank.

 

So here's a pick from an earlier version of ET showing the twin tunnels and the bridge crossing over.  This one was 140mm from ground level.  The current design will be just 70mm in height, but follows a similar appearance.

 

I will be printing off this section of the revised layout this morning and see if the realignment solves all the issues of span and supports.  Fingers crossed.....

 

[trisonic]

I enjoyed this thread and actually learned some stuff!

 

Best, Pete.