v 
jJ 
solution, In a tedious process many details have to be checked, evaluated and tried out before a 
method with a high probability of success may be formulated, Selecting a method appropriate to 
the specific problem at hand - more often than not by a process of elimination - is the first and 
most decisive step ; the resulting procedure will usually be specialized due to the basic conditions 
and it will be unique, i.e. it cannot be transferred in full to any other special photogrammetric 
problem. 
The photogrammetric evaluation of the models of the prestressed cable net roofs of the 
sports facilities for the Olympiad 1972 at Munich shall illustrate some practical aspects of the 
method-finding process, Reference is made to Table I, which depicts the main considerations in- 
volved. Although the objects and the photography have been described by Bóttinger [1] brief mention 
is made here for completeness' sake, 
Basic Conditions 
The environment was the Institute for Light Weight Structures (IL) of Prof, Frei Otto. 
The micro-climate of the room changed with the outside atmosphere. The working space was very 
limited and the floor could not be kept free of vibrations. 
The models had a very complex structure being built of steel wire and steel tubes and 
steel rods, The connections of wires and those of wires and other steel parts were either lashed 
or soldered, The dimensions of the models, built at a scale of 1:125, were 1,5 x 2.0 m? for the 
athletic arena, the swimming hall and the olympic center and 1," x 3.3 m? for the stadium. The 
vertical extension of all models was about 0.7 m. 
The object data requested were the space coordinates of all wire intersection points and 
system points (see[1] ). 
The photogrammetrically determined coordinates of the nets were to be the input data for : 
calculation and mapping of the cutting pattern of each roof, The largest model, the stadium, had o 
about 3. 800 points to be measured. The following considerations lead to an understanding of the 
accuracy requirements. The prestress of the net cables had a tolerance of 15 %, which is equiva- 
lent to a deviation of 0, 6 tons from the initial stress necessary to prevent unloading of cables under 
heavy loading e.g. by snow. With the known elasticity and the area of cross section of the cables, 
this tolerance may be converted to a permissible change in length of 10 mm for a cable of 50 m 
totallength using Hooke's law, Thus, theoretically, individual lengths had to be determined to 
t 0,08 mm at the scale of 1:125, Since, however, the models could not be manufactured with un- 
limited precision, the photogrammetric point determination had to stay within the model tolerances 
of about + 0,3 to * 0, 5 mm, Initially it was intended to transfer the models, exactly as they were 
built, including possible deviations from theoretically stipulated relations between geometry and 
forces, into reality. Therefore a homogeneous set of coordinates, i.e. a digital model of the phy- 
sical model, had to be determined by the photogrammetric process. The internal accuracy of this 
digital model had to comply with the accuracies derived from the prestress tolerances. As the 
works progressed this basic concept was abandoned giving rise to grave consequences, Since this 
had no bearing on the photogrammetric method employed it shall suffice to mention that after the 
change of concept the photogrammetrically determined digital models served as input data for the 
computation of figures of equilibrium, whose cutting pattern was calculated and mapped from the 
equilibrium figure's coordinates. 
Initial Decisions 
Evaluation of the basic conditions led to the formulation of an analytical method of evalua- 
tion of the models, i.e. to close-range stereo-photography and measurement in a stereo-comparator. 
The necessary tools were available and they consisted of : 
- two modified SINAR NORMA 13/18 plate cameras with objectives WILD 60071, 11/115 mm, 
- a GALILEO VEROPLAST Bi-Camera with AEROGON 6. 3/150, 
- a ZEISS stereocomparator PSK fitted with interchangeable punch-tape and punch card output. o 
- and a computer program for the orientation of independent pairs of photographs (the computer 
program will be treated in more detail later on). 
BO ~~ 
E 3 
ZÜl-— ! —— 
  
|