The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
2.3 Precise levelling 
The precise levelling was performed with the Trimble DiNi 
digital level (http://www.trimble.com) and an invar precision 
bar code levelling-staff. The a priori-accuracy (la) for the 
height measurement is set to 0.3 mm per 1 km of double 
levelling. Due to the night-time measurements, the invar 
precision bar code levelling-staff had to be lighted by a 
floodlight to enable the measurements by the precise level 
(Figure 4). The precise levelling was established to measure 
absolute vertical displacements of the bridge girder. 
Figure 4. Precise levelling of reference points on the viaduct. 
2.4 Tacheometry 
The measurements by the tacheometer were performed from the 
valley floor at a distance of about 150 m from the object. Prior 
to the load test, bolts for tacheometer prisms were installed 
underneath the cantilever slabs and the lower slab of the box 
girder. This enabled the measurement of profiles of the bridge 
girder. The main purpose of the tacheometer measurements was 
to provide an additional measurement method to the precise 
levelling and the detection of special deformation behaviours of 
the bridge superstructure. The measurements by tacheometer 
are not further analysed or discussed below. 
2.5 Measurement setup 
Due to the large height of the bridge girder above ground 
(approximately 60 m), the measurements by the terrestrial laser 
scanner as well as the precise levelling were performed on the 
Felsenau viaduct. Prior to the load tests, measuring bolts were 
embedded into the deck slab. In advance, short rods could be 
screwed on the bolts during the load tests. The bolts were 
connected with the concrete slab through the asphalt. 
=^====5==30===5== 
202 
: - ' © 
= y=ss= 
203. 205 105 
■©^© © 
103 102 
© © 
106 
■~© ■ ♦ 
• 207 © 201© 
’"‘*'”•©204 104© 
^©101 
©107 
A scanner 
G reference target 
O target for laser scanning and bolt for precise levelling 
* bolt for precise levelling 
Figure 5. Measurement setup for terrestrial laser scanner and 
targets for TLS and precise levelling. 
For the precise levelling, more than 40 bolts with the respective 
short rods were set up (Figure 5). The bolts were measured by 
precise levelling for each period of the load tests as well as for 
the initial situation. Furthermore, an additional precise levelling 
was measured each time between the test area and a height 
transfer reference outside of the Felsenau viaduct. This was 
essential for the detection of bridge girder settlements. 
The terrestrial laser scanner Imager 5006 was placed on a heavy 
tripod in the middle of the bridge girder (Figure 3). The heavy 
tripod was used to reduce possible movements and torsions of 
the tripod. The height of the tripod was set to approximately 
2 m to minimize small angles of incident of the laser beam on 
the surface. Nevertheless, small angles of incidence could not 
be avoided due to the large extensions of the test area. In 
addition to the Imager 5006, five reference targets were setup 
for registration purposes of the 3D-point clouds. Hence, four 
reference targets were setup as well in the middle of the bridge 
girder, and the fifth reference target was established on the 
opposite side of the test field on the cantilever slabs (Figure 5). 
The maximum range from the scanner to the reference targets 
was about 17 m. White spheres made of wood with a diameter 
of 15 cm were used as reference targets. 
In addition to the reference targets, further targets were set on 
several bolts or short rods, which were used for the precise 
levelling. The additional targets are labelled in Figure 5. The 
targets were coated spheres made of Styrofoam with a diameter 
of 12 cm. These additional targets were mainly established for 
the analyses of the relation between the measurement of precise 
levelling and TLS. 
Generally, the scans were performed with the scanning 
resolution “high” (0.036° for horizontal and vertical angular 
resolution) and the targets were additionally scanned with the 
resolution “superhigh” (0.018° for horizontal and vertical 
angular resolution). Due to the relative deformation monitoring 
of the Felsenau viaduct by TLS, only cantilever slider 
deflections were expected to be detected. Tilting and 
settlements of the bridge girder could not be monitored by the 
presented measurement configuration by the terrestrial laser 
scanner. An absolute height reference was needed for the 
detection of absolute bridge girder deformations. 
3. PROCESSING TLS DATA 
Processing the TLS data included the registration and filtering 
of the 3D-point clouds as well as the determination of 
deformations by comparing the 3D-point clouds with load on 
the Felsenau viaduct to the initial situation. For the analyses, an 
area-wide deformation analysis and a discrete analysis with 
respect to targets on the object were carried out. The 
carriageway, i.e. the cantilever slabs were scanned up to a range 
of 20 m from the station of the terrestrial laser scanner. The 
maximum distance for the measurements on the carriageway 
was limited by the angle of incident of the laser beam on the 
surface and the black colour of the asphalt, which influenced 
the backscatter of the laser light. 
3.1 3D-point cloud registration 
Five spheres with a diameter of 15 cm were used as reference 
targets for registration purposes of the 3D-point clouds. Before 
registering the 3D-point clouds, the spheres had to be modelled 
by fitting a sphere with known diameter into the 3D-point cloud 
according to the least-square method. Hence, the mean absolute 
error was 0.7 mm and the standard deviation 1.0 mm. These