During the building of the bridge many different forms of testing were carried out to ensure the quality of materials being used and to ensure that structural members were capable of withstanding the applied loads. When the bridge was complete it was tested to determine that no one member, or group of members, within the structure were likely to be over-stressed during service.
a. Tensile testing of steel
Shown below is a stress-elongation curve from a tensile test to failure for a test specimen of the silicon steel used in the Sydney Harbour Bridge. This shows the superior performance of the silicon steel (UTS >577 MPa for this sample) compared to the 420 MPa UTS for 1020 mild steel.
This test graph clearly shows the characteristic 'necking' behaviour of steel at the yield point Y. When the cylindrical test specimen yields, it suddenly stretches and narrows (necks) over a short length without any extra load being applied. The curve from Y to U then shows that as load continues to be applied to it, the test specimen then elongates (stretches) by a large amount until it fails (breaks) at point U, its ultimate tensile strength. When the steel finally fails it releases a great deal of stored energy, mainly in the form of sound - in other words, it sounds like a gun going off in the testing lab!
b. Compression testing of steel structural components
The twin arches of the bridge are made up of chord beams loaded in compression. To ensure that these beams were capable of absorbing the applied loads, scale-model beams were manufactured using the same methods as the full size beams. These beams were then tested to destruction in a specially built Avery 1250 ton (1270 tonne) compression testing machine. See image below.
The applied loads and deformation of the beams were monitored during testing and the resultant stress and strain values scaled up to determine the expected full-size load carrying capacity. Long slender beams such as these fail by buckling, not by pure compression. The images show typical buckling failure after testing.
c. Load testing
When the Bridge was completed, Dr Bradfield arranged for it to be loaded in various configurations using many steam locomotives on the railway lines (western side) and tramlines (east). This is much greater than the expected loading in normal day-to-day use of the Bridge.
While the Bridge was loaded, surveyors measured the deflection in the Bridge from reference points on land. Strain measurements were also taken, using strain gauges (instrumentation for measuring strain directly rather than deflection) within individual beams and chords.
The deflections obtained from these measurements were then plotted to an exaggerated scale on an elevation drawing of the bridge.
As mentioned in the section on stress and strain - within the elastic limit stress is directly proportional to strain. Therefore the degree of deflection of any one joint gives a measure of the stress developed at that joint. This could then be compared with the calculated stress to determine any likely areas of excess stress.
The figures obtained were compared with the theoretical deflections calculated by engineers. The results were found to be within tolerance.