Carrying the Loads
The specification that was included in the Sydney Harbour Bridger contract provided to potential contractors included the weights that the Bridge was required to support. The four railway tracks, mentioned elsewhere, required for the proposed electric railway network were, of course, prominent together with the roadway between the pairs of tracks.
The list of different types of load followed customary engineering practice and the items are easily recognisable in modern structural design methods. The two main categories are “dead load” due to the weight of the supporting structure itself and that of the various types of traffic and other applied loads: “live load”. On a calm evening at average temperature, the Bridge is under only dead load at the time of the New Year’s Eve fireworks – barring the weight of the fireworks.
The specification required that the overall weight of the structure be allowed for, with a further two per cent for contingencies.
This introduces an important dilemma for the structural designer. How can the dead load be found for designing the structure when the strength calculations have yet to be made that include the live loads AND the dead load?
What this means is that, in many cases, a trial-and-error calculation system has to be adopted. A first estimate of the dead load - based on an initial sketch design - is used and the entire calculation is repeated a number of time, converging on the eventually correct answers.
It is quite clear that the experience of the designers plays an important part in this phase. Also, over many years approximate formulas have evolved for starting guidance in the case of simpler structures.
The experience factor becomes even more important in the cases of structures at or beyond the size limit of practice at the time.
Two loading conditions were stated for the Bridge: a local loading for cross-girders and stringers and a general requirement for the deck overall. In the first case a motor commercial vehicle was considered to be acting on an area of 330 square feet (30 sq. m). Outside such relatively concentrated loads, a general area load of “one hundred pounds per square foot” (4.8 kPa) was the requirement.
Although Bradfield, in his reports, had been clearly interested in multiple-unit electric railway coaches for the Sydney suburban system, the type of railway design load required by the specification consisted of two heavy electric locomotives followed by a long, uniform load of much lower intensity. This is recognisable as echoing the load system – “Cooper E”– used on North American railways then and now (http://www.tpub.com/content/armytransportation/TI-850-02/TI-850-020112.htm ), where two large steam locomotives haul a continuous load of much lower intensity per unit length. (Whilst steam locomotives now have much less presence, their occasional use must be allowed for.)
The railway loading was applied to all four tracks: two on each side of the roadway. The modern term for this level of loading is “heavy rail” as distinct from the smaller wheel loads such as found in the trams that took over the eastern pair of tracks, and now motor vehicles.
Deck System Impacts
The hard running surface of train wheels on steel rails results in a low shock-absorbing property – especially when compared with pneumatic tyres on made road surfaces. The impact is particularly noticeable at rail joints and other surface dislocations. Although extensive use of steam locomotives was not envisaged on the passenger lines, an impact factor was selected as 50% additional to static load on stringers, decreasing to 10% on the main bridge trusses. The corresponding values for the roadway were 25% and 10%. Please see Chapter 38.0 Railroad structures in Bureau of Structures website
The Tay Bridge collapse in Scotland in 1879 and the enquiry that followed caused engineers worldwide to look more closely at the effects of wind on structures. Much work was done in co-operation with meteorologists, resulting in national codes of practice.
- Tom Martin's Tay Bridge Disaster web pages
- The Tay Bridge Disaster by V.Ryan
- Forensic engineering: The Tay Bridge Disaster
The Australian code (AS 1170.2 – 2011) now contains 101 pages but things were somewhat simpler in the 1920s. It was required by the specification and contract that a “wind load normal to the bridge of 30 pounds per square foot (1.44 kPa) of exposed surface of the two trusses” be allowed for. This works out at slightly more than modern requirements. A similar type of allowance needed to be made for forces from wind blowing longitudinally to the Bridge.
During the enquiry into the Tay Bridge, it was suggested that the lateral force on a train crossing the bridge at the time of the collapse may have contributed to the failure. A reminder of this may be found in the requirement for the Sydney Harbour Bridge that “an exposed surface of a train” would be subject to a wind load of 300 pounds per lineal foot (446 kg per lineal m.).
Thermal expansion of the various components and the differences in expansion between different components and different materials were considered in the specification. A variation range of 67°C in the uniform temperature of the whole structure and of 28°C between the temperature of the steel and of the masonry was to be assumed.
Centrifugal and other forces
The northern approach to the Bridge has a horizontal curvature of 362 metres radius and the resulting lateral force exerted by two trains travelling at 80 km/h was to be taken into account.
A modern engineer would notice the omission from the specification of any consideration of seismic forces. The national Australian earthquake code did not appear until 1979.