Why would a bearing plate that further concentrates load onto small surface be used?

Here's a photo of a bearing plate where a bridge reinforced concrete beam meets the earthfill

The bridge span is about 20 meters long and consists of two reinforced concrete beams each resting on two bearing plates like the one shown - one plate for each end of each beam, total four plates. The bridge holds a railway track designed for 25 tons per axle cars. The bearing plate is made of cast iron (or maybe steel) and consists of two large parts joined through a hinge.

25 tons per axle cars means the bridge bears something like several hundred tons when a train is passing which we can assume causes at least one hundred tons per bearing plate shown. Yes, I just ignored the bridge weight.

Not only the plate upper and lower surfaces are rather small but the plate further concentrates the accepted load and transfers it onto the hinge through even smaller surface. Basically this rather small hinge alone accepts more than one hundred tons. And this is designed on purpose.

Why is the load deliberately concentrated instead of being distributed or at least transferred through some part with uniform section?

Because bridges and other structures are not static objects. They must be allowed to flex under varying loads and also accommodate changes in length from thermal expansion. The hinge pin allows changes in angle. and the sliding joint between the upper hinge plate and the flat plate on the bottom of the beam allows changes in length.

If the connections were rigid, these forces could destroy the structure over time.

• Is this for cases when the bridge bends downwards because of heavy load passing? – sharptooth Jun 7 '16 at 14:58
• @sharptooth Yes. And any other movement.Usually, at the other end of the bridge, there will be some sort of sliding plate arrangement. – Dave huh Jun 7 '16 at 20:34
• @sharptooth It is possible on some older bridges with a something called a rocker bearing, that over enough expansion and contraction or flexing cycles along with some freezing either due to temperature or corrosion, that the rockers will "walk" off their bearing pads, or will wind up listing/tilting so much that they become unstable. I inspected a bridge once that had its rocker pad ejected from the bearing and it was sitting on the ground 6' away. At first I did not know what it was when I picked it up, then I looked at the bearing and saw it had dropped and was not functioning as intended – Forward Ed Jun 24 '16 at 3:26

The reason is pretty simple. Steel is significantly stronger than concrete.

Nowadays we have high-performance concretes with $f_c > 100~\text{MPa}$ (and ultra-high-performance, which is substantially higher), but most ordinary structures don't use such high strength concrete. This bridge seems relatively weathered, so the concrete is probably at most $f_c > 40~\text{MPa}$ (probably even less).

Steel, on the other hand, nowadays has at least $f_y > 250~\text{MPa}$, often even more. I don't believe steel strength has evolved as quickly as concrete (correct me if I'm wrong), so the steel used on that bridge is probably at least equal to this.

The steel on that bridge is therefore some 7-8 times stronger than the concrete. So, whatever area the concrete requires to safely transfer the load to the steel (via the plates), the steel actually needs far less, so it can safely reduce its own dimensions. Buckling is controlled by the bracing all around the hinge.

As to why a hinge is used at all, that has to do with how the bridge was designed, as described in @DaveTweed's answer.

• The progress in concrete strength is amazingly nonmonotonic: research over the last 20 or 30 years has revealed that Roman concrete from BCE is incredibly strong, apparently due to mixing volcanic ash into the material. – Carl Witthoft Jun 8 '16 at 12:39
• @CarlWitthoft: Yes, but I'd argue (without any sources) that once concrete was rediscovered after the dark and middle ages, and especially after the scientific revolution, its strength has increased monotonically (but certainly not a constant rate). I doubt the bridge the OP is asking about is more than 50 years old. – Wasabi Jun 8 '16 at 12:48
• Steel has progressed as well. There are 2 GPA UTS heat-treated cast steels with 5-10% elongation, though they are a good deal more expensive than sub 1 GPa alternatives, due to required vacuum or argon shroud during melting and pouring. Also TRIP and TWIP steels with dramatically improved impact toughness and energy absorption (though admittedly that isn't much to do with strength). TWIP is up to 800 MPa, or up to 100% elongation. Yes, doubling in length before failure. Also quite expensive due to a complex combined heat treatment and forming process with tight chemistry. – wwarriner Jun 9 '16 at 2:26
• @CarlWitthoft : I would argue that the progress (or lack thereof) of concrete technology has more to do with market forces and distribution channels than technical innovation/discovery. The Norwegians have produced hollow stem drilling augers with concrete rather than steel, for example. But vested industry interests, ie local aggregate suppliers, Portland cement manufacturers, etc, and cost concerns are in fact some of the main obstacles to commercial adoption, in my view. – AsymLabs Jul 25 '16 at 14:57

If you go back to you basic engineering courses and look at you bending moment diagrams for beams, quite often they will be illustrated with pin roller supports. Pinned at 1 end only allowing rotation, and roller connection at the other allowing for rotation and horizontal translation. This makes the beam statically determinant.

When this bridge was originally built, elastomeric/rubber bearing pads and a few others did not exist as an option. This design emulates our formulas that we use for design, or rather the formulas work with this arrangement. So this type of configuration is good. It allows us to use our formulas as intended, keeps the design simple, and used the technology of the time. In addition, as mentioned in other posts, it allows for rotation at the support as a result of live load, variations in dead, or sag after removal of shoring (assumed shored construction. If the girder were lifted into place after casting, the large steel plate embedded into the concrete allows for some inaccuracies in measurement of the span and even placement. It also allows for support should the beam shift slightly due to vibration or earthquake.

Note, you will also see a similar configuration with steel girders on various types of bearings. I believe the term shoe or shoe plate will get used, though that may be more for buildings than bridges.

On an aside

When it comes to "RAIL" bridges, the vast majority in North America designed to AREMA will consist of simple spans whether it be single span or multi-span bridges. I found this statement funny while on my AREMA course as I just finished inspecting about a dozen rail bridges in my city with the vast majority not following this rule. In highway bridges you will tend to see continuous for live load and these bridges as a result these bridges are not statically determinant.