The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
559 
could not be placed at the very outer side of the cantilever slabs. 
But it is difficult to compare the mean values of the target 
sphere residuals to the mean positive and mean negative 
residuals, which resulted from the deformation analysis by the 
software Geomagic Qualify, due to an unknown weighting of 
mean positive and mean negative residuals. However, it must be 
considered that the deformations detected by TLS only describe 
the deviation of the cantilever slab. Hence, the implementation 
of absolute vertical displacements is indispensable for the 
detection of settlement and tilting of the bridge girder. 
4. COMPARISON OF TLS AND PRECISE LEVELLING 
The load test on the Felsenau viaduct enabled the comparison of 
different geodetic measurement methods. Below, the results of 
the measurements by the precise levelling are summarised, and 
the results of TLS are compared with the results of the precise 
levelling. 
4.1 Deformation measured by precise levelling 
The reference height point close to the test field was determined 
for each loading situation by a precise levelling with 13 setups 
from a height transfer reference outside of the Felsenau viaduct. 
The accuracy for the reference height point was calculated to 
0.36 mm (la). Furthermore, the height of the bolts in the test 
field was measured by a single observation due to efficiency 
reasons. The accuracy of the relative height determination for 
the bolts was calculated to 0.15 mm (la). The resulting 
accuracy of a single measurement of a bolt in the test field was 
computed to 0.50 mm (la). 
equidistance: 0.5mm 
Figure 8. Deformations between initial situation and loading 
situation P2 of bridge girder detected by precise 
levelling. 
Figure 9. Vertical displacements of sphere centres between 
different loadings and initial situation measured by 
precise levelling. 
Figure 8 shows the area-wide vertical displacement between the 
initial situation and the loading situation P2. Settlements and 
tilting of the bridge girder can be clearly detected. Furthermore, 
a deflection of the cantilever slabs is visible for the outer side of 
the slabs close to the loads. 
For the precise levelling, the vertical displacements between the 
initial situation and the situations with different load conditions 
are listed in Figure 9. Hence, bolts 104 and 204 performed the 
largest deviations with 10.8 mm and 10.9 mm. 
4.2 TLS versus precise levelling 
The TLS data was recorded in a local system without any 
connection to the outside of the Felsenau viaduct. Hence, local 
deformations as the deflection of the cantilever slab were 
detected. In contrast to TLS, the precise levelling was 
connected to a transfer point outside of the Felsenau viaduct. 
Absolute deformations of the bridge girder could be detected. 
For the comparison of TLS data with data of the precise 
levelling, a transformation of the TLS data into the precise 
levelling height system was required. At least, the settlement 
and tilting of the bridge girder had to be added to the TLS 
measurements. The additional vertical displacements of the 
reference targets of the terrestrial laser scanner were calculated 
by interpolating the vertical displacements of bolts for precise 
levelling which were installed close by. By analysing the 
vertical displacements of the TLS reference points, the 
settlement and tilting of the bridge girder could be determined 
under the assumption that the TLS reference targets remained 
stable to each other. Table 2 lists the calculated settlements and 
tilting of the bridge girder for the different loading situations. 
Settlement 
Tilting 
[mm] 
[°] 
PI 
-0.85 
0.0051 
P2 
-3.35 
0.0111 
P3 
1.65 
0.0040 
Table 2. Settlement and tilting of the bridge girder for the 
corresponding load situations derived from the 
vertical displacements measured by precise levelling. 
The vertical displacements between the different loading 
situations and the initial situation were transformed with the 
corresponding transformation parameters (Table 2). Figure 10 
presents the transformed vertical displacements of the targets 
for TLS. 
Figure 11 shows the differences between the transformed 
vertical displacements measured by the terrestrial laser scanner 
and measurements by the precise levelling for the targets in the 
test field. As a result, there are differences up to 3.5 mm. The 
displacements are normally distributed and no systematic 
deviation is detectable. 
The mean residuals, respectively the mean vertical 
displacements, which were measured by precise levelling as 
well as TLS, are presented in Table 3. For TLS, the mean value 
refers to the transformed vertical displacements. The ranges of 
the mean values are similar for precise levelling and for TLS.