Mapping Voids, Debonding, Delaminations, Moisture, and Other Defects Behind or Within Tunnel Linings (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

387 a p p e N D I x x Data acquisition The basic principles of photogrammetry are illustrated in Fig- ure X.1: once a pair of photographs is acquired, the same point, P, is identified on each photograph, and the spatial coordinates of point P on the object (the ground in this case) are calculated by tracing two rays from the focal points Ol and Or of each pho- tograph, respectively, through the pixels that represent point P on each photograph. A patch of an object’s surface is therefore reconstructed from each pair of photographs: it is the patch portrayed by both photographs. The reconstructed surface can then be either scaled or scaled and georeferenced in a reference system of interest. Because each point on the reconstructed surface comes from a known pixel, each pixel can be exactly attributed to the rele- vant point of the reconstructed surface. Thus, the reconstructed surface is exactly textured with high-definition photos, allow- ing for a reliable and realistic virtualization of the object under consideration. This means that when the trace of a lining crack is digitized by following the trace pixels on the textured surface, the correct trace geometry on the underlying surface can be identified with certainty; this is not the case with laser scanner applications. Once a three-dimensional (3-D) model has been recon- structed, it can be scaled, or it can be scaled and georeferenced. A scaled model allows for crack, spalling, and visible moisture detection and measurement: crack length, aperture, location (relative to an arbitrary point); spalling area, depth, volume, and location; and moist area and location. Additionally, a scaled and georeferenced model allows for the following: • Change detection, that is, determining changes in crack lengths and aperture, spalling extent (area and depth), and moist area; • Determination of 44 Crack orientation, which is very useful for ensuring that the grouting holes actually intersect the crack; 44 Wall displacements such as convergence, tile delamina- tion, concrete delamination, and ceiling or floor sag- ging; and 44 Overall tunnel displacement, for example, in immersed tube tunnels, lifting caused by loss of ballast or sinking caused by debris discharge over the tunnel. The speed of photograph acquisition depends on the accu- racy required, the minimum distance between the cameras and the tunnel lining, and whether the tunnel is accessible to vehicles. As part of this research, special technology has been developed to achieve the performance detailed in Table X.1. accuracy of the 3-D Model and Information provided to Client An additional advantage of digital photogrammetry with respect to the laser scanner is that the bundle adjustment resid- ual is provided at each common point P (see Figure X.1), and then the root mean square (RMS) of the residuals is provided for each photograph, and for the entire model. These data provide detailed local and global information on the accu- racy of the model, which is not available in laser scanner applications. The pixel size on the lining is chosen before entering the tun- nel on the basis of the desired accuracy. For example, Table X.2 refers to the two-lane Liberty Tunnel in Pennsylvania, which is about 20 ft wide. In this tunnel, photographs were taken from the left lane, and the maximum distance from cameras to the opposite wall was about 12 ft. The a priori calculated accuracy of the photogrammetric model was 0.8 mm; the chosen pixel size on the lining (farthest distance to the camera) was about 1 mm by 1 mm. For each picture actually taken in the field, Table X.2 provides the RMS of the residuals, which is always less than 0.3 pixels, that is, 0.3 mm on the lining. The Active Points column refers to the number of common points between a given picture and all overlapping pictures. Digital Photogrammetry

388 Residuals are also calculated on: • Scale bars (scaled models): Table X.3 gives the residuals at the scale bars used at the Liberty Tunnel. The scale bar dimensions are provided with a National Institute of Stan- dards and Technology–traceable certificate of calibration. The residuals are better than 50 microns (0.002 in.). • Surveyed targets applied to the final lining (scaled and geo- referenced models): Table X.4 gives the residuals obtained on surveyed targets in the road enclosure of the Eisenhower Memorial Tunnel in Colorado. The residuals in each direc- tion are equal to about 0.5 mm (0.02 in.), and the overall, spa- tial residual is equal to 0.8 mm (0.03 in.). The overall accuracy (target survey and photogrammetric model) obtained in the divider wall at the Eisenhower Memorial Tunnel was equal to 1 mm. These results indicate that the following may be identified with confidence: • Bulging caused by incipient spalling and tile delaminations; • Subtle ceiling/floor movements that may indicate progres- sive failure of the support (e.g., roof collapse at the Central Artery/Tunnel in Boston); • Convergence of the tunnel walls; and • Overall tunnel displacement (e.g., in immersed tube tun- nels, lifting caused by loss of ballast or sinking caused by debris discharge over the tunnel). Figures X.2 through X.13 provide an example of a 3-D model of a tunnel lining—Liberty Tunnel in Pennsylvania—and its use in identifying lining defects. (Figures X.3 through X.8 show close-up views of the circled spalling event.) Figure X.14, Figure X.1. Photogrammetry principles. Table X.1. Speed of Photograph Acquisition Tunnel Type Taking Pictures from Taking Pictures of Lane Closure Required Accuracy Requireda Speed 2-lane road enclosure 1 lane Entire tunnel 1 lane from which pictures are taken 0.8 mm 1,200 ft/hr (360 m/hr) 2-lane road enclosure 1 lane Entire tunnel 1 lane from which pictures are taken 1 mm 2,400 ft/hr (720 m/hr) 2-lane road enclosure 1 lane Single wall (e.g., only tile panels) 1 lane from which pictures are taken 0.8 mm 2,000 ft/hr (600 m/hr) 2-lane road enclosure 1 lane Single wall (e.g., only tile panels) 1 lane from which pictures are taken 1 mm 3,300 ft/hr (1,000 m/hr) Air duct (20 ft [6 m] wide) Center Entire duct No 0.6 mm 660 ft/hr (200 m/hr) Air duct (20 ft [6 m] wide) Center Entire duct No 0.7 mm 1,000 ft/hr (300 m/hr) a Accuracy refers to photogrammetric model only.

389 Table X.2. Residuals and Active Points for Pictures Taken at Liberty Tunnel RMS Error (pixels) Active Points Name RMS Error (pixels) Active PointsName X Y Total X Y Total IMG_2272.JPG 0.116403 0.153484 0.192632 195 IMG_2287.JPG 0.164044 0.212679 0.268594 211 IMG_2273.JPG 0.117043 0.179321 0.214138 277 IMG_2288.JPG 0.117306 0.184640 0.218753 228 IMG_2274.JPG 0.131904 0.214972 0.252214 337 IMG_2289.JPG 0.097516 0.177364 0.202404 355 IMG_2275.JPG 0.145841 0.182533 0.233640 226 IMG_2290.JPG 0.097395 0.143055 0.173062 333 IMG_2276.JPG 0.119270 0.228402 0.257668 281 IMG_2291.JPG 0.114015 0.175516 0.209297 472 IMG_2277.JPG 0.082480 0.207591 0.223376 283 IMG_2292.JPG 0.113118 0.204120 0.233368 473 IMG_2278.JPG 0.098254 0.161682 0.189195 298 IMG_2293.JPG 0.133337 0.195348 0.236515 353 IMG_2279.JPG 0.114083 0.212339 0.241046 383 IMG_2294.JPG 0.110535 0.183812 0.214488 388 IMG_2280.JPG 0.135485 0.190771 0.233987 343 IMG_2295.JPG 0.112492 0.179191 0.211574 342 IMG_2281.JPG 0.123500 0.181342 0.219401 244 IMG_2296.JPG 0.109538 0.122035 0.163985 268 IMG_2282.JPG 0.132664 0.236927 0.271540 236 IMG_2297.JPG 0.111083 0.189163 0.219367 407 IMG_2283.JPG 0.107341 0.179687 0.209307 330 IMG_2298.JPG 0.122423 0.203170 0.237203 446 IMG_2284.JPG 0.109541 0.168065 0.200612 289 IMG_2299.JPG 0.114291 0.180656 0.213773 365 IMG_2285.JPG 0.153775 0.200796 0.252914 252 IMG_2300.JPG 0.117907 0.184573 0.219019 352 IMG_2286.JPG 0.176630 0.198991 0.266075 210 IMG_2301.JPG 0.119525 0.167282 0.205595 235 Table X.3. Residuals at Scale Bars Used at Liberty Tunnel Scale Bar Name First Point ID Second Point ID Distance (m) Accuracy (m) Residual (m) Scale bar 1 Point 1 Point 2 1.095956 0.000010 -0.000044 Scale bar 2 Point 3 Point 4 1.096029 0.000010 0.000029 Figure X.15, Table X.5, and Table X.6 exemplify the results pro- vided to the client at the end of the photogrammetric survey to document existing cracks and spalling events. Figures X.16 through X.21 illustrate the 3-D model of the clean air supply duct at the Eisenhower Memorial Tunnel. In this research, special lighting systems have been devised to evenly illuminate the lining even in dark situations, such as air ducts, and to ensure that the colors of the lining are reliably reproduced. The pipes attached to the divider wall have not been reproduced satisfactorily in three dimensions because pattern is needed in photogrammetry to find relative-only points, and steel pipes have very little, if any, pattern. Regard- less, one of the objectives of this application was to check the use of photogrammetry in surveying a divider wall with embedded steel hangers that cannot be examined by any non- destructive technique. Details of the divider walls are given in Figures X.17 and X.18, and a global accuracy of 1 mm was achieved, which ensures that progressive yielding of a hanger (or hangers) may be detected if surveys of this kind are car- ried out systematically. Figures X.19 and X.20 depict the model of the south wall as seen from the inside of the tunnel, where several cracks are evident. Some of the cracks have been digitized in Figure X.20: one is a typical construction (pour) joint, but the others are not, and their orientation allows one to infer the causes of distress in a specific area of the lining. Such inferences are difficult to make while inspecting the tunnel and mapping cracks by hand. Finally, Figure X.21 provides a detail of a cracked area of the lining. The provided model and the quantities obtained are completely objective and defensible, and may be used at any time during the operational phases of the underground infrastructures.

390 Table X.4. Residuals at Scale Bars Used at Eisenhower Memorial Tunnel (Road Enclosure) Control Point Names Number of Observations Image Point Residuals Control Point Residuals (m) X (pixels) Y (pixels) X Y Z 1 2 0.0067 0.0407 -0.0000 0.0004 -0.0006 2 2 0.0846 0.1426 0.0001 -0.0004 0.0006 3 2 0.1233 0.0489 0.0002 -0.0003 -0.0006 4 2 0.0458 0.1339 -0.0003 0.0003 0.0005 Control point RMS 0.0002 0.0004 0.0006 Total 0.0008 Figure X.2. Three-dimensional model of Liberty Tunnel inbound tube by the ventilation shaft. Figure X.3. Liberty Tunnel: Detail of spalling by ventilation shaft. Figure X.4. Foreshortened view of spalling by ventilation shaft. Figure X.5. Detail of exposed aggregate and rebar by ventilation shaft.

391 Figure X.6. Backside view of spalling by ventilation shaft to better appreciate depth and extent of spalling. Figure X.7. Closed polyline to determine area of spalling by ventilation shaft. Figure X.8. Information on spalling by ventilation shaft—coordinates of center point and area. Figure X.9. Overall view of a tunnel stretch by southern portal. Figure X.10. Overall view of tunnel stretch by southern portal with digitized cracks on left wall. Figure X.11. Detail of digitized shotcrete cracks on left wall by southern portal; also visible is a plane interpolated through crack trace.

392 Figure X.12. Detail of shotcrete cracks on left wall by southern portal with digitized cracks toggled off. Figure X.13. Close-up view of shotcrete crack on left wall by southern portal: detail of surface roughness. Figure X.14. Typical survey of existing cracks in a final lining. Figure X.15. Typical survey of spalling event in a final lining.

393 Table X.5. Example of Surveyed Crack Report for a Final Lining Center X (ft) Center Y (ft) Center Z (ft) Dip° Direction° Diameter (ft) Trace Length (ft) 7.218 34.186 0.098 90.0 0.0 11.188 10.925 9.383 13.944 -0.886 89.9 176.0 12.238 12.041 9.974 13.222 -3.839 88.0 84.0 1.444 1.411 Table X.6. Example of Surveyed Spalling Report for a Final Lining Center X (ft) Center Y (ft) Center Z (ft) Area Depth (in.) Volume (sq ft) Exposed Rebars? 23.845 151.513 3.515 3.51 2.4 0.702 Yes Figure X.16. Eisenhower Memorial Tunnel: Clean air supply duct, south-west ventilation building. Figure X.17. Eisenhower Memorial Tunnel: Clean air supply duct, south-west ventilation building, divider wall. Figure X.18. Clean air supply duct, south-west ventilation building, triangulated mesh of divider wall. Figure X.19. Clean air supply duct, south-west ventilation building, view of south lining wall from within duct. Figure X.20. Clean air supply duct, south-west ventilation building, view of south lining wall from within duct with digitized features. Figure X.21. Clean air supply duct, south-west ventilation building, detail of south lining wall from within duct.