Splicing 
If more than one web plate is used to build a longer bridge, splice plates are used to join the web plates. These are riveted flat panels. Splice plates are generally located about a third of the way in from the ends, but they aren’t placed in the center of the span.1
Webplates shall be spliced by one or more plates on each side; the splice plates shall have a section equal to at least threefourths of the web, and a pair of stiffeners shall be placed outside the plates. There shall not be less than two rows of rivets 31/2 inches apart on each side of the splice; rivets in splice plates shall not be spaced not over 31/2 inches apart for a distance of 14 inches at top and bottom and not over 41/2 inches between.2 .. it is found that the longest plate 66 in. x 1/2 in. that it is possible to get is 34 feet long; it is therefore necessary to splice the web. it will be spliced at the center, making each half 27 feet 11 inches long, nearly. 3 The first source says “Splice plates are generally located about a third of the way in from, the ends but they aren’t placed in the center of the span.” – while the last source says .. “… it will be spliced in the center ..“. The first source was published 1964 and the last one 1908. Perhaps they learned something in the intervening 54 years? Note: My bridge represents an On18 line that is using an older threefoot gauge line’s bridge. Since my layout is set c.1940’ish .. it would make sense that the bridge could date from 1908. That could mean that it happily was spliced at the center. In the 1908^{[2]} source they use a bridge about the length of mine but designed for standard gauge and they end up with a 1/2″ web. I figure that I can make a WAG and say ok .. for the same length bridge but narrow gauge then 7/16 inch web would work. “The splice shall have a section equal to at least threefourths of the web“. My bridge will have a web plate height of 60 in. FS (1.250 in. O Scale). The backs of the angles are spaced 1/4″ further apart then the height of the web to allow for irregularities in the edges of the web plates – so that would be ~1.258 in. over the backs of the angles. This is a ‘silly number’ really – as far as working in 1:48 but the point is that the main angles are spaced slightly taller then the web plate so there is no interference fro irregular plates with the flange plates. As far as modeling then .. fix the height of the main angles about .005″ higher then the web plate. I say ‘about’ as the critical figure is making the stiffeners and cross bracing tuck in under the main angels. (In other words – ensure that the cross frame height and stiffener height is the same and then ‘fit’ the main angles around those dimensions .. adjusting as I go.) So ..”just for fun”. A web plate 60 in. x 7/16 in. That size web has an area of 26.25 sq.in. Since each splice plate is required to have a sectional area equal to 75% of the web – then .75 X 26.25 =19.6875 sq.in. I will be using Evergreen .100 angle (4.8 in. FS). As the distance back to back of the flange angles is 1.258 in. (60.384 in. FS), the clear distance between the vertical legs is 1.258 – .100 .100 = 1.058 in. In “real life” they allowed 1/8 in. clearance of the top and bottom (good luck with that in 1:48!) leaves 1.058 – .003 = 1.055 in. (50.64 in.) for the height of the splice plate (calm down everyone – I KNOW I’m not going to get this kind of precision in 1:48 .. having fun playing with the pretty numbers) In the present case the area of the web is 26.25 in. then the area of each plate must be .75 x 26.25 = 19.6875 sq.in. As the plates are 1.058 in. deep (50.64 in. FS), the required thickness is 19.6875 / 50.64 = .389 inch. The nearest standard thickness (rounded up) would be 1/2 inch – although 3/8 plate would only be under about .020 in.. Remember – the actual web thickness of my bridge will be hidden – and I will be using .040 in. styrene for strength – but – I can “suppose” that it is in fact 7/16 in. plate .. and therefore I can use the 1/2″ thick splice plate since that WILL show .. or .. .010 in. styrene. (In which case amusing myself with thoughts of 3/8 plate is a little silly (like most of this math I suppose .. still .. having a great time playing with the numbers!) So. From all of the above what we is this:

Cooper’s Loading 
Just going to toss this in here for now before I forget.
Cooper’s E10 – Two Consolidation type steam locomotives with tenders weighing ??? tons followed by a series of cars (uniform load) 1,000 lbs. per. lineal foot. This was Cooper’s E10 loading because each driving axle weighed 10,000 lbs. Cooper’s E30 – Two Consolidation type steam locomotives with tenders weighing 106.5 tons followed by a series of cars weighing 3,000 lbs. per. lineal foot. This was Cooper’s E30 loading because each driving axle weighed 30,000 lbs. a Consolidation has a 280 wheel arrangement. Sooo .. two Consolidations would have 8 axles. At 30,000 per axle that’s a total of 240,000 lbs or 120 tons. Sooo .. the partial I have on the E10 means 80,000 lbs. or 40 tons. D&RGW 223 weighed 69,105 lbs with 59,330 on it’ drivers. That would work out to 14,833 lbs. per axle. That E30 says ‘weighing 106.5 tons (with tender)’. If we figure a tender at 75% of the weight of the engine then 106.5 X .75 = 60 tons. #223 put about 86% of the locomotive weight on the rails. Following that .. 86% of 60 tons is 44 tons. Divide that by 4 to get 11 tons/22,000 per axle. Well .. that kinda sorta gets us in the ballpark. In any case .. #223 would require a bridge with a E15 load rating .. if I have my numbers correct. Sooo .. working backwards. 30,000 lbs per axle is 15 tons. Multiply 15 x 4 and you get 60 tons to the rails. Sooo .. 60/.86= 69 tons. Further .. 10,000 lbs per axle is 5 tons. Multiply 5 x 4 and you get 20 tons to the rails .. and .. 20/.86= 23 tons of locomotive. Tenders weigh let’s say .. 75% of the loco .. so in this case .. 17.25 tons. Cooper’s E10 – Two Consolidation type steam locomotives with tenders weighing 40.25 tons followed by a series of cars weighing 1,000 lbs. per. lineal foot. This was Cooper’s E10 loading because each driving axle weighed 10,000 lbs. Locomotive weighs 23 tons and tender 17.25 tons. Hmm. 23 tons. That’s more like an early 440 American type in weight. I can use that then to apply the narrow gauge loading on the bridge. Now … if I can find where they punch in the numbers for the Cooper loading .. 
 The Model Railroader’s Guide to Bridges, Trestles & Tunnels, Jeff Wilson, 1964 [↩]
 Bridge specifications: Design of plate girders, International Correspondence Schools, 1908 – Sec.74Pg.15 [↩]
 Bridge specifications: Design of plate girders, International Correspondence Schools, 1908 – Sec.76Pg.39 [↩]