STRUCTURAL DEFORMATION MEASUREMENT OF A MODEL 
BOX GIRDER BRIDGE 
Peter j Scott 
University College London 
Introduction 
In early 1977, a 1:12 scale model of a multiple 
box girder bridge was tested through its entire 
working load range to unserviceability levels and 
final collapse. The test was one of a series 
instituted in the United Kingdom by the Department 
of the Environment in a programme of research into 
elevated urban motorway design and the main pur- 
pose was to study a bridge with a bifurcation. 
The bridge was built and tested in the Department 
of Civil Engineering at Imperial College, London. 
Deformations were recorded photogrammetrically 
throughout the test and the measurements were 
used to locate the load stage at which buckling 
had started in local areas. Development of the 
buckling throughout the test provided the rare 
opportunity to study the ultimate load behaviour 
of an entire structure. 
The Design of the Bridge 
  
Fig. 1 is a general view of the completed 
structure. The model simulated a typical mul- 
tiple box girder bridge of medium span. The 
bottom and side faces of each box were made of 
steel plate and the reinforced concrete deck acted 
compositely with the steel by forming the top of 
the boxes. This can be seen in Fig.2. The chosen 
scale of 1:12 was smaller than anything realisti- 
cally attempted before. The problem that it posed 
for photogrammetry was that since 1 mm was signi- 
ficant in the deformation at full scale, the 
model would have to be measured to a tolerance of 
better than 0.1 mm. In the event, a compromise 
of +0-2 mm in the quoted deformation was stated as 
the standard which photogrammetry should hope to 
achieve. 
Under increasing vertical load, the bottom 
face of the box would show some local deformation 
over the internal stiffeners (Fig.3). The pur- 
pose of photogrammetry was to monitor the bottom 
compressive face in order to detect the load point 
at which these buckling deformations appeared, 
and to measure their subsequent rate of develop- 
ment. The density of photogrammetric measurement 
is timited only by the size of premarked target re- 
quired, and the targets are very cheap and easy to 
install. Perhaps the greatest advantage of photo- 
grammetry, however, lies in the fact that a large 
number of targets can be recorded on photographs. 
After final failure has occurred the local areas of 
interest are known and only the targets in these 
areas need be measured. 
  
45 
À