FHWA LTPP Guidelines For Measuring Bridge Approach Transitions Using Inertial Profilers

Chapter 1. Profiling Bridge Sections

Background

The bump at the end of the bridge has long been studied for highways and railways, yet experts from across the transportation industry continue to identify it as one of the most prevalent substructure factors affecting bridge performance. Often, rideability is a subjective measurement used by State transportation departments to define the presence of a bump. User complaints typically drive maintenance schedules. However, the bump is not just an annoyance on the traveling public; the dynamic impact of vehicles resulting from the bump causes distress, fatigue, and long-term damage to the bridge deck. In addition, the bump also causes damage to vehicles and potentially creates an unsafe condition for drivers if this issue is not addressed in a timely manner. To ensure that the bump is within tolerable limits based on safety, rideability, and effects to long-term bridge performance for State transportation departments, tools are necessary to measure and assess the bridge approach transition. These products can ultimately be used to help State transportation departments manage and preserve their bridge inventory.

Profile Data Collection Considerations for Bridge Approach Transitions

The high-speed inertial profiler is an excellent tool to determine the smoothness or lack thereof as the result of differential differences between pavements, bridge approaches, and bridge structure. These devices can collect profile data without interruption to the travelling public, have sample rates of 1 inch or less, produce profiles that are consistent and repeatable, and provide datasets that are useable for producing numerous indices for riding comfort and pavement profile. The International Roughness Index (IRI), rolling straight edge (RSE) simulation, and diamond grinding simulation analysis have all been used to evaluate and provide solutions for issues with bridge approach transitions.⁽¹⁾ Profiling devices such as the Face Dipstick® or walking profiler (e.g., International Cybernetics’ SurPRO) could also be considered if only a few bridges were to be considered for testing. Testing with these devices requires a greater level of coordination because there is a need for lane closures and site preparation, which would most likely limit production to one site per test day. Of these two devices, the walking profiler would be the better alternative because a sample rate of less than 1 inch is possible. The elevations from these devices are generated without extensive filtering, which results in a product that is more representative of the true profile. A high-speed inertial profiler will detect a bump or dip at bridge approach transitions to a similar level of accuracy, but the longwave profile shape will not match that of the true profile. For profile data collection at bridge approach transitions, the high-speed inertial profiler can provide all of the information needed to detect bump or dip locations and produce statistics (e.g., IRI and RSE) to evaluate and compare bridge transition performance. Bump height can be obtained, but it will be a representative value. A comparison of testing devices is shown in table 1.

Table 1. Comparison of testing devices.

The bridge profile data collection procedure was developed using the Federal Highway Administration’s (FHWA) Long Term Pavement Performance (LTPP) high-speed inertial profiler. Survey units (see figure 1) contained three pavement profiling sensors located in the left and right wheelpaths and midlane, two 62.5-kHz texture-sensing lasers located in the left and right wheelpaths, ambient and surface temperature sensors, a Global Positioning System (GPS), and a forward direction photo logger.

Figure 1. Photo. Survey unit.

Although 25 ft (8 m) of pavement before and after the approach area is considered sufficient for bridge ride specifications, collecting data over a longer length provides a better understanding of bridge performance relative to localized pavement condition.⁽²⁾ These guidelines are based primarily on measurement procedures used within the LTPP Program for profiling weigh-in-motion (WIM) sites. The WIM data collection started 902 ft (275 m) prior to the WIM sensor and continued 97 ft (30 m) past it. This length was selected to measure how the bumps in the road affected truck movement and the resulting variability in dynamic loads that would affect the WIM sensing device. The primary concern for selection and calibration of WIM locations was to ensure that the location selected did not have a pavement profile that would induce unacceptable amounts of truck bounce as the truck approached the WIM scale. By collecting longer pavement lengths prior to the bridge interface, useful information could be used to model the dynamic effects of truck loads on the bridges. For many small, local bridges within the study group, which were primarily on rural roads, it would have been difficult to set up and collect profile data for lengths longer than 200 ft (61 m) prior to and past bridges. Based on this knowledge, the decision was made to use a distance from the bridge deck to the beginning of the test section of 200 ft (61 m) with the distance from the end of the bridge deck to the end of the test section also being 200 ft (61 m). This length should be sufficient enough to monitor the long-term changes that occur at the bridge location using inertial profilers.

For some of the bridge locations, it may be possible to collect profile data with a longer approach to the bridge location. This could be established as a pilot within the data collection if there is interest in modelling the dynamic effects of truck loads on bridges.

ProVAL is a profile viewing and analysis software used by many State transportation departments and consulting firms as a tool for calibrating inertial profilers and analyzing profile data.⁽³⁾ This software can use profile data from numerous profilers that are available within the North American market. ProVAL 3.5 has been recently updated to handle the .ARD file that is the current profile file generated from the LTPP profilers. Due to these added features within ProVAL, the use of this software for the processing of the bridge profile data is suggested. Some manipulation of the data may, however, be required for processing and presentation, but for the most part, ProVAL generates the reports for the bridge surveys.

Two 62.5-KHz texture-sensing lasers located in the left and right wheelpaths collected profile data at 0.012-inch (0.5-mm) intervals. The texture laser sensor output is a point-to-point measurement with a greater accuracy than that of the 16 kHz profile lasers. Unlike the profile laser sensors, which are integrated with the accelerometer output to generate a pavement profile, the output from these lasers is simply the step difference between samples. Some of the limitations of the texture laser are that there are spikes that are not part of the sampling within the profile data. Algorithms have been developed to remove these spikes, but at this time, there does not seem to be any consensus about the validity of the spike removal process. While options for spike removal exist, they are not currently used in the data collection and processing of LTPP profile data. The tight interval sampling provides an opportunity to identify discreet changes within the pavement surface with a greater level of accuracy than is available from the profile sensors because of the tighter sample rate and minimal filtering. Procedures are in place to use this information for determining pavement texture and are under consideration for determining faulting in portland cement concrete (PCC) pavements. It is possible to use this information to locate and measure the step at the transition between the pavement, approach structure, and bridge, but at this time, there is no widely available software that could be used for this purpose. As development in this area continues, the data collected could be useful in future analysis.

Bridge Sectioning and Site Information

A bridge profile section for purposes of this survey is defined as a section that includes a portion of pavement prior to the approach structure, bridge deck, and portion of pavement after the bridge. The length of the section will vary depending on the span length of the bridge deck. For some types of bridges, it may not be possible to visually determine the interface location of the approaches from the bridge. In general, the transition distance is within 25 ft (8 m) of the bridge structure. It is about 3 to 5 ft (0.9 to 1.5 m) from the face of the abutment wall to the end of the beams. The distance from the bridge deck to the beginning of the test section should be 200 ft (61 m), and the distance from the end of the bridge deck to the end of the test section should also be 200 ft (61 m). This distance may be modified at the discretion of the profile operator if any safety issues related to location conditions exist (e.g., stop signs within the profile area).

The bridge section should be marked as shown in figure 2 for bridge decks that are perpendicular to the paved roadway. Marking for bridge decks that are on a skew are shown in figure 3. Bridges that have both a perpendicular and skewed deck interface will require markings based on information provided from both figures. Monuments (in the form of nails or spikes) should be installed in the shoulders at the beginning and end of the test section as shown in the figures. These monuments will serve as section markers for future surveys. The monument at the beginning of the section should be located 200 ft (61 m) before the leading edge of the bridge deck. The monument at the end of the section should be located 200 ft (61 m) after the end of the deck, in the shoulder of the opposite lane. The distances measured should be accurate to within ±1 ft (0.305 m). Profile surveys should be done for all lanes in both directions.

1 ft = 0.305 m.

Figure 2. Drawing. Layout of bridge site with perpendicular joints.

1 ft = 0.305 m.

Figure 3. Drawing. Layout of bridge site with skewed joints.

A fourth-order Butterworth high-pass filter was used to determine profile elevations for these surveys with a cutoff wavelength of 300 ft (91 m). For the filter to have sufficient length to normalize the profile elevations, a minimum of 450 ft (137 m) of lead-in and lead-out were required (1.5 times the filter wavelength). For bridges on a curve, the minimum lead-in and lead-out required should be based on the lane with the shortest curve length. The survey unit had a default setting of 500 ft (152 m). To ensure the proper lead-in length and consistency in profile data collection, it is recommended that a cone or marker be placed as a reference for the profile operator to initiate the profile data collection. A permanent monument (in the form of nails or spikes) should be considered at these locations for ease of reference in future surveys. The speed of the profiler during the lead-in and lead-out should be consistent with that of the profile data collection. In other words, there should be no acceleration or deceleration in the lead-in or lead-out area.

To initiate the profile data collection and identify the start and end of the bridge deck, reflective stripes or cones should be placed at the start and end of the section and at the start and end of the bridge deck. The leave edge of the stripes should be next to the applicable monument. There are a number of options available to provide a reflective surface for the profiler photocell to initiate data collection. If no permanent or semipermanent reflective stripe is evident at the defined locations, the profiler crew can place cones with reflective tape near the edge of pavement and bridge deck. Alternatively, a reflective stripe (typically 2 inches (51 mm) in width) needs to be placed in the lane located perpendicular to the edge of pavement.

For State transportation departments willing to permanently mark the test location, preparations should be coordinated at the time of the first profile data collection to ensure that bridge section markings are consistent. This will allow for subsequent surveys to be repeated at the same locations. Alternatively, the profiler crew will be required to locate, measure, mark, and install the permanent monuments prior to the first survey.

Section location information can be painted near the outside shoulder for permanently marked sites at the discretion of the State transportation department. To locate a bridge, the profiler crew relies on the GPS longitude and latitude coordinates provided by the department or from a previous survey.

The profiler crew collects and records information regarding the roadway, bridge approach, and bridge structure on form 1A, as shown in figure 4. The following information would be useful in identifying test locations:

• The length and width of the bridge.
• The type and length of the approach.
• Pavement type (e.g., asphalt concrete or PCC).
• Pavement condition (e.g., cracking, deformations, and texture (with particular attention to the transition zones)).
• Landmarks or benchmarks (e.g., bridge posts).

Figure 4. Form. Perpendicular joint site layout (form 1A).

For bridges having a skewed interface with the roadway, the profile crew should use form 1B (see figure 5) and also provide the length of the skew and the distance between the start and end portion of the skew. Measurement procedures, measurement tools, and dimensional information (including all length and width information relevant to the survey) should be incorporated into the diagram as outlined in the corresponding form (1A or 1B).⁽⁴⁾ Comments specific to the site conditions and location should also be included.

Figure 5. Form. Skewed joint site layout (form 1B).

Pictures of the bridge, the bridge superstructure, pavement section, unique features (i.e., cracks in the pavement, vertical offsets, etc.), and location landmarks should be taken at the time of survey. Form 2A, as shown in figure 6, should be used to record the number and type of each picture. Form 2B (see figure 7) shows an overhead view of the photo locations. Pictures should not include any individual or vehicle license. In the event that these details are captured in any of the photos, either the photo should be deleted or the image blurred.

Figure 6. Log. Form 2A—photo log.

Figure 7. Log. Form 2B—photo diagram.