Matthew Yarnold

Temperature-Based Structural Identification of Long-Span Bridges

AbstractTemperature-based structural identification (TBSI) is a quantitative structural evaluation approach that relies on responses resulting from temperature fluctuations. Through this approach, the transfer function that defines how thermal induced strains give rise to global displacements and restrained member forces can be captured. This input-output relationship is highly sensitive to mechanisms that pose modeling challenges, such as boundary and continuity conditions, and thus is quite valuable within the model updating process. The method follows the traditional structural identification (St-Id) framework with a priori modeling, experimentation, and model calibration steps appropriately modified to allow for the measurement and simulation of temperature-induced responses. TBSI was evaluated through the use of simulations and laboratory experiments and then implemented to identify an arch bridge. In addition, a comparative study was performed with an independent evaluation of the same bridge using ...

Advanced Visualization and Accessibility to Heterogeneous Monitoring Data

The proper management and visualization of data is crucial for the success of structural health monitoring (SHM). The article discusses how SHM is a promising tool for the better management of infrastructure. General principles for SHM data management are researched, established, and proposed by the authors, and an original solution for data management based on these principles is presented. The article discusses how data management is especially challenging when heterogeneous data are involved and combined with camera images. Various sensors based on different technologies can measure many parameters such as strain, tilt, weather, etc., whereas live cameras can visualize traffic response and all of the data streams can be registered both statically and dynamically. Data management is even more complex if multiple users access the data and have diverse backgrounds and interests (e.g., the owner/manager of the structure, operator, responsible engineer, and academic). The proposed principles were implemented in novel data management software and applied to a signature bridge for validation purposes. The article discusses how feedback from interested groups including managers, operators, engineers of record, and academics are used for validation.

Forensic Investigation of the Route 61 Bridge

Sensing technology has the ability to provide information that can answer difficult questions with regard to constructed bridge behavior, improving the way complex structures are maintained. A detailed evaluation of the State Route 61 Bridge in Clinton, TN is prime example. The Route 61 Bridge is a 3-span continuous steel girder bridge that includes several complex features; the structure has a kinked-girder alignment, steel integral pier caps, skewed substructures, and one integral abutment. It was built in 1985, rehabilitated significantly in 1996, and is again scheduled for a rehabilitation in the upcoming year. The current structure has a failed expansion joint, severely tilted expansion bearings, and includes substantial concrete degradation at the ends of the structure. As a result, a detailed assessment was performed that included the application of sensing technology for field measurements along with numerical finite element modeling. Vibrating wire technology was utilized for measurement of relative bearing movement as well as girder strains over a period of three months. In addition, global movement was measured at select times with survey equipment. After collection and review of the data, it was determined that unexpected thermal movement was the primary cause of the increased bridge deterioration. Recommendations were made for the upcoming rehabilitation to increase the life of the structure.

Evaluating the coefficient of thermal expansion using time periods of minimal thermal gradient for a temperature driven structural health monitoring

Structural Health Monitoring aims to characterize the performance of a structure from a combination of recorded sensor data and analytic techniques. Many methods are concerned with quantifying the elastic response of the structure, treating temperature changes as noise in the analysis. While these elastic profiles do demonstrate a portion of structural behavior, thermal loads on a structure can induce comparable strains to elastic loads. Understanding this relationship between the temperature of the structure and the resultant strain and displacement can provide in depth knowledge of the structural condition. A necessary parameter for this form of analysis is the Coefficient of Thermal Expansion (CTE). The CTE of a material relates the amount of expansion or contraction a material undergoes per degree change in temperature, and can be determined from temperature-strain relationship given that the thermal strain can be isolated. Many times with concrete, the actual amount of expansion with temperature in situ varies from the given values for the CTE due to thermally generated elastic strain, which complicates evaluation of the CTE. To accurately characterize the relationship between temperature and strain on a structure, the actual thermal behavior of the structure needs to be analyzed. This rate can vary for different parts of a structure, depending on boundary conditions. In a case of unrestrained structures, the strain in the structure should be linearly related to the temperature change. Thermal gradients in a structure can affect this relationship, as they induce curvature and deplanations in the cross section. This paper proposes a method that addresses these challenges in evaluating the CTE.

Identification of steady-state uniform temperature distributions to facilitate a temperature driven method of Structural Health Monitoring

Structural Health Monitoring seeks to characterize the performance of a structure from combinations of recorded sensor data and analytic techniques. Temperature is normally considered noise in this analysis, obstructing the goal measuring the elastic response of the structure. While these elastic loads do help characterize a portion of structural behavior, the thermal loads on a structure can induce comparable strains to these elastic loads. Characterizing a relationship between the temperature of the structure and the resultant strain and displacement can provide a deep understanding of the structural condition. In order to begin characterizing this 3-dimensional relationship, time periods with relatively steadystate, uniform temperature distributions need to be identified from the measured data. These periods of uniform temperature distribution in the structure show a thermal response as free as possible from thermal gradients across the structure. These steady-state periods help create a signature of the structure when analyzed with the relevant strain and displacement measurements of the structure. An algorithm for finding these uniform distributions was created to identify these desirable time periods with data of interest. Finding time periods with a completely uniform temperature distribution can be unreasonable, so a suitable temperature interval was chosen to produce a set of data with a reasonable approximation to a uniform distribution, while still providing a large enough set of data to produce meaningful results. These time intervals provide the necessary temperature, strain, and displacement measurements to characterize a signature for the structure, providing a more in-depth analysis in SHM.

Temperature-Driven Assessment of a Cantilever Truss Bridge

Temperature-driven assessment has the potential to advance our understanding of long-span bridge behavior. The novel approach researched as part of this study is the identification and monitoring of a long-span cantilever truss bridge using the inputoutput temperature relationship. The goal is to use this relationship to identify and monitor unknown quantifiable information with regard to an existing structure using the structural identification process (e.g. boundary conditions, continuity conditions, force distribution, etc.). Since many structural parameters on long-span bridges are highly sensitive to temperature loads, a structure such as the Hurricane Bridge is a prime candidate for this type of monitoring. The Hurricane Bridge is a four-span, cantilever truss bridge over the Caney Fork River in DeKalb County, Tennessee, with a total length of approximately 1787 feet. It was built in 1949 and rehabilitated in 2011. The rehabilitation included widening the deck, strengthening various truss members, and installing a “catch system” consisting of four stainless steel rods around each vertical at the cantilever locations. This critical structure has a great deal of uncertainty related to performance and remaining service life. Therefore, a temperature-driven monitoring system has been designed and implemented to reduce the uncertainty associated with the: “catch system”, pin and hanger effects at cantilever locations, and bearing mechanisms. The sensing technology of this system is comprised of fifty-six vibrating wire strain gages, eight vibrating wire displacement gages, and sixty-four thermistors. Long-term data collection is on-going; however, preliminary results are presented and tasks for future research are explored. INTRODUCTION & OBJECTIVES In recent years, engineering practices have transformed from a mindset of replacement to rehabilitation with regard to many structurally deficient bridges. Much of this motivation stems from funding and the amount of bridges in need of repair at this time. According to the American Road and Transportation Builders Association and the 2015 National Bridge Inventory released by the Federal Highway Administration, “there are nearly 204 million daily crossings on 58,495 U.S. structurally deficient bridges in need of repair” (ARTBA 2016). Due to the increasing number of deficient bridges, monitoring techniques are being utilized more often in order to prioritize the bridges based on their performance and need for intervention. Currently, the most prevailing technique for monitoring long-span bridges is ambient vibration monitoring. Using this method, modal parameters such as natural frequencies, mode shapes, and damping can be determined and tracked for a structure. Although this method has been utilized, ambient vibration monitoring also has challenges associated with it (Catbas 2007). Ambient vibration monitoring has difficulty dealing with environmental effects such as seasonal temperature change since they can mask damage (Peeters and De Roeck 2001). Therefore, a significant challenge for this type of approach is removing the temperature effects. The prevailing reason for the limited success of ambient vibration monitoring of long-span bridges is the limited sensitivity to structural damage (Brownjohn et al. 2011). Alternatively, a temperature-driven concept, where thermal “loads” are treated as the excitation and the corresponding static responses are correlated, shows promise to mitigate many of the shortcomings of ambient vibration monitoring (Yarnold and Moon 2015; Kromanis 2016). Logistically, a temperature-driven approach can be performed continuously over a period of time with minimal data storage and time synchronization requirements. In addition, the equipment is relatively inexpensive and generally self-sustaining with little need for man-power resources once the system is installed and operational. The results can be recorded throughout the structure’s changing environments and can potentially identify structural changes that occur as a result of seismic, wind, ice, impact, or similar nature. This is primarily due to the fact that a temperature-driven baseline is highly sensitive to many changes of structural systems (Yarnold and Moon 2015; Laory et al. 2013). Temperature-driven monitoring is particularly useful for large structures. Long-span bridges, for example, are more responsive to thermal loads than live loads, making the results easier to identify. Figure 1: Structural Identification (St-Id) Process The novel approach researched as part of this study is the identification and monitoring of a long-span, cantilever truss bridge using the input-output temperature relationship. The goal is to use this relationship to identify and monitor unknown quantifiable information with regard to an existing structure (e.g. boundary conditions, continuity conditions, force distribution, etc.) using the structural identification (St-Id) process shown in Figure 1 (Yarnold et al. 2015). “St-Id is the process of creating and updating a model of a structure based on its measured static and/or dynamic measured response which will be used for assessment of the structure’s performance for informed decision making” (Catbas et al. 2013). As shown in the figure, the process can be expanded upon to incorporate a temperature-driven approach. The temperature-driven concept is further explained below followed by illustration of the comprehensive design and implementation for the cantilever truss bridge study. TEMPERATURE-DRIVEN CONCEPT Since long-span bridges have a high sensitivity to thermal effects, everyday temperature exposure can excite a response from the structure. The temperature-driven concept utilizes this cause-and-effect relationship to develop a behavioral signature for the bridge. This process is detailed in Figure 2 below. The temperature variations (input) are quantifiable and can be measured simultaneously with the member strains, displacements, and/or rotations (output) that the bridge experiences in response to the thermal load. Once the behavioral signature has been determined, it can be used to update a model to represent the current condition of the structure. This process can be used for both St-Id as mentioned previously and structural health monitoring (SHM) for long-term performance tracking. Figure 2: Temperature-Driven Concept ASSESSMENT OF THE HURRICANE BRIDGE Bridge collapses are not prevalent in today’s age; however, they can happen. One such occurrence was the I-35W bridge collapse in Minnesota in 2007. This structure was a long-span, steel truss bridge that experienced a catastrophic failure due to a poor design and lack of redundancy (National Transportation Safety Board 2008). Motivated by this disaster, Tennessee Department of Transportation initiated a review of similar bridges in Tennessee, one of which being the Hurricane Bridge shown in Figure 3. Located in DeKalb County, Tennessee, the Hurricane Bridge is a four-span, Warren deck truss bridge that is approximately 1787 feet in total length. Two suspended sections comprise the middle of the bridge as shown in Figure 4. One end of each section rests atop the middle pier while the other end is connected to a cantilever and the rest of the bridge via a pin and hanger detail. This bridge was built in 1949 by the U.S. Army Corps of Engineers and was rehabilitated in 1977 and 2011. The primary goals of the 2011 rehabilitation were to widen the deck, strengthen several structural members, and install a “catch system” to increase redundancy at the cantilever locations. The “catch system” consists of four, 3-inch diameter stainless steel rods installed around each of the vertical hanger members at the cantilever locations to essentially “catch” the suspended section in the event of a failure. The “catch system” is not a commonly used rehabilitation method; therefore, a large degree of uncertainty exists regarding the behavior. Recall, the intent of this study was to use a temperature-driven monitoring approach to minimize the uncertainty of the bridge with regard to the behavior of the pin and hanger, the “catch system”, and the bearings. Figure 4: Hurricane Bridge Overview Figure 3: Hurricane Bridge Following the St-Id process shown previously, an element-level 3D finite element model of the Hurricane Bridge was created. The model includes all primary superstructure and substructure components as shown in Figure 5. A thermal load was applied to the entire structure, and the bearing mechanisms were characterized by connection elements with variable translational stiffness. The variable stiffness elements allowed for simulation of varying stiffnesses of continuity conditions such as bearings and the pin and hanger connections. Figure 5: Finite Element Model of Hurricane Bridge After the finite element model was complete and checked, design of the temperature-driven experiment was performed. The sensing equipment used for this project required the ability to capture the results for any scenario and the ruggedness to withstand prolonged weather exposure. Due to increased demand for monitoring assessments, sensing equipment specifically designed for this purpose was readily available. Vibrating wire strain and displacement gages were decided upon and used to identify the behavior of the bridge. Vibrating wire gages measure frequency from the excitation of a small wire within the gage. The frequency can then be directly correlated to the strain or displacement being measured. Sensitivity studies were performed for various scenarios to determine the optimum location for sensing equipment. For example, Figure 6 shows two scenarios compared to the “As Drawn Conditions” specified in the original and rehabilitation plans. The “As Drawn Conditions” have free movement at the pin and hanger and the expansion bearings of Pier 5 and Pier 7. The scenario “Pier 7 Bearings Seized” has free movement at the Pier 5

Local Buckling Analysis of Trapezoidal Rib Orthotropic Bridge Deck Systems

This paper compares the behavior of local buckling in trapezoidal rib orthotropic bridge deck systems. The primary objective of this paper is to compare the condition of a uniform stress pattern (column) and that of a non-uniform stress pattern (beam-column). The former is current practice while the latter is recommended based on the current study. The presence of thin steel plate members within the deck system causes local buckling to be a valid concern. Parametric analyses were performed using the finite element method to compare local buckling of the rib walls (webs) and deck plate by varying the corresponding width-to-thickness ratios. Since the rib walls have the highest width-to-thickness ratio, they were the primary focus of this research. Generally, this type of deck system is analyzed under axial compression due to global forces. However, bending moments from the local loading of the deck weight and vehicles are typically present in the system. Therefore, the bridge deck system was analyzed under...

Temperature-Based Model Updating of Bridge Structures

Temperature-based (TB) model updating has been researched as a novel approach to inform numerical models of constructed systems. The overall TB concept utilizes temperature variations as a measureable forcing function for bridges and thus can be used to obtain a complete input-output relationship (or transfer function). This is achieved through instrumentation of both the critical members and movement mechanisms. Studies were conducted on existing bridges where the temperature (input) was measured at select locations along with the corresponding strains, displacements, and rotations (output). Then TB model updating was performed to refine a numerical model according to the measured input-output relationship so it more closely represented the actual behavior of the structural system. This paper discusses the approach for TB model updating along with the primary benefits and drawback of the methodology.

Bearing Assessment with Periodic Temperature-Based Measurements

Bearing performance is closely tied to overall structural performance and serviceability issues and, as a result, demands considerable attention to condition assessment. That the life cycle of bearing systems is far less than that of bridges has been well documented. Eventually, the bearings break down, with resultant undesirable forces and deformations. Currently, visual inspection is the most common method used to assess bearings and other movement systems. The correlation is limited, however, between visual appearance and the functionality of movement systems. Thus visual inspection falls significantly short in such assessments; this situation necessitates the application of sensing technology for in-depth evaluations. This paper discusses a method, termed periodic temperature-based assessment (PTBA), which aimed to supplement visual inspection through measured input–output temperature responses without the use of a finite element model. A framework was developed for a general approach to PTBA. Strategies also were researched for the measurement of the input–output temperature relationship for bridge structures related to movement systems. In addition, the PTBA method was demonstrated on a long-span, steel, tied-arch bridge to evaluate its expansion bearings.

Preparing Engineers for Evaluation of Constructed Systems

Current engineering education places a significant focus on design of new, idealized structures. However, many graduating engineers are entering the workforce and are faced with evaluation of existing complex systems. To address this learning gap, a new graduate course was recently added at Tennessee Technological University titled, “Structural Identification of Constructed Systems”. Structural identification is the process of creating, then updating a physics-based structural model based on its measured static and/or dynamic measured response which is then used for evaluation of the structure’s performance as well as making critical decisions. The new course taught students the fundamentals of Structural Identification through hands-on experience, which included several laboratory experiments and projects. This paper presents the primary characteristics of the course along with sample student projects.