Firm Employs 3D Laser Scanning to Fix Bridge
The Philip J. Fahy Memorial Bridge, or the New Street Bridge, is a 1,500-linear-foot steel truss bridge that spans the Lehigh River in Bethlehem, Pa.
Currently, the bridge contains four vehicular traffic lanes, divided by a concrete median and a deteriorating pedestrian walkway, which was closed to the public in 2011. Going south to north, the bridge approach begins at the intersection of East Third Street and New Street, crossing over East Second Street, and then a Norfolk Southern railroad right of way before reaching the south side of the Lehigh River, accessed from East Second Street, and a secondary exit to East Second Street can be found on the southbound lanes. On the north side of the Lehigh River, Fahy Bridge spans over Sand Island, a public park, the Monocacy Creek, the D&L Canal, and an abandoned railroad right of way before reaching the New Street/Center intersection. Two stairwells provide access to the pedestrian walkway at East Second Street and at Sand Island.
After inspections in 2011 found deficiencies within the bridge deck and pedestrian walkway, the Pennsylvania Department of Transportation (PennDOT) proposed bridge rehabilitation improvements and traffic control, restoration of the pedestrian walkway and the creation of a dedicated bike lane in both directions. To aid in design efforts, HNTB Companies engaged Maser Consulting, P.A., a northeast and mid-Atlantic-based multi-disciplined civil engineering and land surveying firm with national satellite offices, in August 2012 to perform 3D high definition laser scans of Fahy Bridge’s super structure and surface elements along with roadway scans of New Street between East Third Street and Center Street, and ancillary entrance and exit ramps at East Second Street.
To create their engineering plans, HNTB’s objective was to utilize LiDAR data to develop a 3D model of the bridge for structural analysis and to develop 3D cross-sections of the bridge deck, sub-structure, bridge piers and abutments. Intersection improvements at East 3rd Street and Center Street were also proposed. In addition to utilizing QA/QC protocols and laser scanning best practices methodology, accurate and precise traverse control was essential to fulfill this objective.
The main intent of the scan was to capture the superstructure elements below the roadway surface. Any challenges that comes with scanning a bridge, especially a bridge of this length, must be identified. As with any terrestrial laser scanning project, there was a tremendous amount of pre-planning to identify issues such as site access, mobilization and survey control.
First, understand the subject of the scanning. Recognize the complexity of the elements that comprise the structure: the decking, bridge piers, abutments, plates, columns, stringers, girders, hinge points, etc., as well as their geometry. These components pose challenges in the scan data in the form of obstructions or shadows that affect the quality of the scanned object(s).
Second, know the surroundings. Examine the environment that surrounds the scan area, identify potential obstructions and pre-plan optimum scan positions and target locations. Pre-determining optimum scan locations is the key to effectively reduce the amount of scans necessary to capture critical bridge components. For example, the Lehigh River is a major obstacle. There was an initial concern that accurate scan data of mid-span components would not be captured. In addition, scan locations at the edge of the river were limited, as well as the distance the mid-span was from the edge of the river. Upon further investigation, low-tide elevations were in Maser Consulting’s favor, as well as a significant reduction of the river flow rate, enabling the company to set the scanner within the river bed and closer to mid-span.
Third, realize that the bridge is a moving object and is subject to vibration and displacement from live loads placed on the bridge by vehicular traffic.
“Scanning a large civil structure like the Fahy Bridge presents a unique set of challenges for surveying and laser scanning,” says Kevin Hannah, PLS, the laser scanning department manager for Maser Consulting. “Essentially, one is scanning a moving target, because of the bridge’s inherent dynamic movements. The first step toward minimizing movement is supposed to have the bridge closed 100 percent during scanning and surveying operations.”
To eliminate displacement in the scan data or “noise,” usually attributed to vibrations caused by vehicular traffic on the bridge, the bridge was closed to traffic and scans were performed on the bridge deck for two consecutive nights.
“When possible, traverse control was set on the bridge’s expansion joints, where movement is minimal,” Hannah says. “Below the bridge deck, scanning over a large expanse of water (the Lehigh River ), eliminated the possibility of additional scan target locations. Consequently, the bridges steel structure, which is geometrically rich, was utilized to assist registration.”
Last but not least, the most important aspect to consider with 3D laser scanning is a throwback to basic land surveying principles: having precise horizontal and vertical survey control, which has a direct effect on the quality of the registered scan data. Since a third-party firm was scheduled to provide horizontal and vertical coordinates for secondary traverse positions established by Maser Consulting, a kickoff meeting prior to commencement of scan activities was held at the site to coordinate scan locations. Pre-scan activities began by establishing horizontal and vertical control on both sides of the bridge, using multiple rapid static GPS observations and local geodetic monumentation. The third-party firm, performed its own traverse and located all of Maser Consulting’s primary control points. A control file and an adjustment report was provided to Maser, and thoroughly examined prior to final registration and unification of the cloud data.
Identifying these issues beforehand aided in determining optimum scan locations, thereby reducing the amount of time required to complete the bridge scans while increasing field efficiency.
The Leica Scanstation C10 time of flight laser scanner was implemented as the primary source of data capture for the project. A minimum of four targets were acquired at each scan setup. Each target scan was checked visually on the scanner interphase prior to storing. All target locations were positioned to maximize geometric strength at each scan setup. Successive scanner to scanner locations did not exceed a maximum of 200 feet. Target level bubbles were checked prior to being scanned during fieldwork. All project scan data was downloaded from the C10 internal drive and copied to the Maser Consulting secure file server at the Mount Arlington Offices. A total of 100 scans were completed for the project.
Scans within the Norfolk Southern Railroad right of way were dependent on flagman scheduling and at many times, obstructed by constant freight traffic.
Leica Cyclone software version 7.4.0 was utilized for all point-cloud processing and registrations. Point normals were calculated for each scan during import. Point-cloud registration constraints were computed primarily from target vertices (vertex to vertex method). To buttress the registration, secondary constraints were added by computing meshes for neighboring scans with sufficient overlap and geometry. All targets acquired were checked for accuracy prior to registration, including target labels, height and vertex centering. An independent registration was computed prior to computing a state plane or geo-referenced registration based on coordinates provided by others.
The independent registration yielded an average RMS (root mean square) error of the alignment of less than 0.015 feet for enabled constraints. The geo-referenced registration yielded an average RMS error of the alignment of for enabled constraints = 0.02 feet for enabled constraints. A rejection criterion of 0.04 feet was adopted for all registrations. Constraints with errors exceeding this value were excluded and disabled from the respective registrations. The final point cloud database (geo-referenced) was checked for accuracy by reviewing vertical cut planes from multiple angles through the bridge deck to check for “cloud shift” in registered point clouds. Noise resulting from sun glare and “thin” point cloud coverage at project limits were deleted prior to conversion to E57 file format.
As a secondary deliverable, Maser Consulting utilized the registered cloud data within Leica Cyclone to create a topographic survey of the bridge deck, adjacent intersecting roadways, and an active Norfolk Southern Railroad right-of-way, within PennDOT survey requirements in Microstation format.
“Creating surfaces utilizing scan data is much more precise,” says Bill Tsivikis, project surveyor. “We created a centerline alignment down the theoretical centerline of the road and generated point data in a cross sectional method. This allows us to create sections on an even interval and gives us the ability precisely provide data on grade breaks and other key points.
“The level of detail and accuracy you can achieve utilizing HDS (High Definition Scanning) is second to none,” Tsivikis says. “We were asked to provide clearances from the railroad tracks to bottom of bridge support I-beams. Having all that data on hand rather than sending someone back out to the field cuts down on field costs & is a huge time saver.”
All in all, approximately 4,000 linear feet of roadway surface area scans were in addition to the bridge underside, piers and abutments. Though scanning was postponed for several days due to inclement weather, actual field work was completed in 10 days, which included two night time scans of the bridge deck. Registration and processing of the cloud data, and creation of the topographic survey were submitted to HNTB two weeks thereafter. DGN and E57 files along with Leica TruView files were then provided to HNTB to begin the design process.
TruView is a browser plug-in from Leica Geosystems that provides users a 3D perspective of the point cloud data from the laser scanner’s point of view. Benefits include the ability to allow users to pan, zoom, rotate, obtain distance measurements and XYZ coordinates. A site map was created for HNTB showing scan world locations above and below the bridge, providing the ultimate user a unique vantage point of existing project conditions from each scan position.