Zulfiqar, Cabieses, Khan, Mikhail
Life­Cycle­Cost Model for the Design of a Bridge Vibration Monitoring System (LCC­BVMS) Ahsan Zulfiqar Miryam Cabieses Andrew Mikhail Namra Khan Department of Systems Engineering and Operations Research George Mason University Fairfax, VA 22030­4444
1
Zulfiqar, Cabieses, Khan, Mikhail
Table of Contents 1.0 Context
1.1 Current Bridge Inspection System
1.2 Bridge Inspection Process
1.3 Structurally Deficient Bridges
1.4 Periodic Manual Inspection Historical Data
1.5 Causes of Bridge Component Failure
1.6 Bridges Types
1.7 Bridge Components
1.8 Structural Vibration
1.9 Main Components and Failure Types
2.0 Stakeholder Analysis
2.1 Primary Stakeholders
2.1.1 Federal Highway Administration
2.1.2 Department of Transportation
2.1.3 Design Engineers
2.1.4 Construction Team
2.1.5 Inspection Team
2.2 Secondary Stakeholders
2.2.1 Bridge Users
2.3 Stakeholder Interactions
2.4 Stakeholder Tensions
3.0 Problem Statement
4.0 Need Statement
5.0 Requirements
5.1 System Requirements
5.2 Design Requirements
5.3 Component Diagram
6.0 Proposed Solutions
6.1 Bridge Vibration Monitoring System: Concept of Operations
6.2 Accelerometers
6.3 Proposed Alternatives
6.3.1 Periodic Manual Inspection
6.3.1.1 Manual Bridge Inspection
6.3.2 Event-Based BVMS
6.3.2.1 Hybrid Wired Accelerometer with Manual Inspections
7.0 Method of Analysis
8.0 Simulation
8.1 Simulation Requirements
8.3 Simulation Results
2
Zulfiqar, Cabieses, Khan, Mikhail
8.4 Design of Experiment
9.0 Life-Cycle cost
9.1 Current Periodic Manual Inspection
9.2 Event-Based BVMS Cost Calculations
9.2.1 Hybrid Wired Accelerometer with Manual Inspections
9.2.2 Hybrid Wireless Accelerometer with Manual Inspections
9.2.3 BVMS Equipment Cost
9.2.5 BVMS Total Cost
10.0 Business Case
10.1 Total BVMS cost vs Periodic cost
10.2 Inspection Savings
10.3 Value Hierarchy
10.4 Utility Vs. Cost
11.0 Conclusions and Recommendations
12.0 Project Management
12.1 Budget Analysis
12.2 Work Breakdown Structure
12.3 Project Plan
13.0 References
3
Zulfiqar, Cabieses, Khan, Mikhail
1.0 Context 1.1 Current Bridge Inspection System The Federal Highway Administration (FHWA) is an agency within the U.S Department of Transportation whose mission is to provide stewardship over the design, construction, maintenance and preservation of the Nation’s highways, bridges and tunnels. This department is under the administration of 607,380 bridges, which include the regions of D.C and Puerto Rico as states. The average age of these bridges is 42 years, however, most exceed the lifetime they were built to have. Hundreds of thousands of civilians use bridges as a means for transportation. In order to guarantee civilians with safety, FHWA conducts bridge inspections every two years. The current bridge inspection process is manual, involving visual and aural inspections that are both time consuming and costly [15]. The FHWA requires evaluation of all bridges however; it is costly having a bi­annual inspection cost of $2.7 billion for the U.S. where the average inspection cost per bridge ranges from $4,500­$30,000. This requires closing lanes for the span of the inspection, which can take 1 to 3 days causing traffic congestions. With the rapid growth of highway transportation and the dispersed age distributions of bridges, fatigue damage is increasing faster than the growth in inspection and repair (How, C. 2014). The bridges’ structural deficiencies demonstrate an exponential growth resulting due to detection of the deficiencies after the periodic inspections. With the bridges’ infrastructure deteriorating, there is an increase in maintenance cost as well as ​
an increase in the frequency of the inspection cycle from 2 years to 6 months. 4
Zulfiqar, Cabieses, Khan, Mikhail
1.2 Bridge Inspection Process Figure 1: General Bridge Inspection Procedure The bridge inspection process starts when an inspector is informed about a bridge that needs to be inspected. The inspector gathers any historical data about the bridge that needs inspection. The inspector visits the site to determine the equipment needed for inspection and places a reservation for the equipment. Afterwards, the inspector coordinates with the traffic controller for the bridge inspection on a certain date for safety precautions [1]. The FHWA administers the process and regulates it through a manual bridge inspection. On average, 2 to 3 inspectors do inspections in a day; however, it varies depending on the complexity of the bridge and its length [9]. Each part of the bridge is inspected and rated by an inspector including the four main components; deck of the bridge, truss, superstructure, and substructure. These parts are rated subjectively according to the Bridge Inspector’s Reference Manual. Depending on the rating of the 5
Zulfiqar, Cabieses, Khan, Mikhail
condition, an action is taken. The rating varies from good condition, replacement or repair order or ultimately closing the bridge [1]. The inspectors record the data obtained during the inspection and store it once they return to the base. After the inspection process has finished and the data has been stored, the maintenance personnel have to follow the records collected to repair or replace the parts of the bridge, if necessary. The bridge inspection is an ongoing process therefore, despite the fact that they are inspected every two years, every year they inspect half the bridges thus, the system never becomes idle [1]. A sample inspection sequence for a bridge inspection is presented in ​
Figure 2​
. It is imperative to develop a sequence for the inspection since it ensures safety and thorough inspection of the bridge. Figure 2: Sample Sequence for Bridge Inspection [1] 6
Zulfiqar, Cabieses, Khan, Mikhail
1.3 Structurally Deficient Bridges Over time, the effects of nature and man cause bridges infrastructure to deteriorate more and more, if unaddressed, the smallest of cracks can result in devastating collapses. A major concern are structurally deficient bridges which are bridges rated in poor condition classified as such due to the fact that its load carrying capacity is significantly below current design standards. According to the American Society of Civil Engineers as of 2013 over 12% of the bridges in twenty­one states in the United States have structurally deficient bridges shown on Figure 3 [6]. These bridges are a prominent cause for concern due to the lack of safety they provide to the public. Inspecting these structurally deficient bridges prior than the 2­year period is essential in order to detect, and prevent bridge failures increasing the maintenance cost associated with the inspection process [9]. Figure 3: Deficient Bridges in the U.S. 7
Zulfiqar, Cabieses, Khan, Mikhail
1.4 Periodic Manual Inspection Historical Data The Virginia Department of Transportation (VDOT) offered historical data of 17 bridges consisting of 21 inspections since 1973. This data provides us with the number of defects found per inspection per bridge, the total number of repairs done per inspection per bridge, and the frequency of detection of defects (How, C. 2014). The defects identified indicate an exponential increase over the years with a slightly downward curve shown due to the bridges being repaired. The data shows that it can take up to 8 years for a repair to be made from the time the defect was identified. This all depends on the bridge engineers’ prediction of how long the bridge can remain as is without being repaired. This delay, in response, has an impact on the bridge’s behavior. When a defect is identified in a specific component of a bridge, it infers that it is not handling the stress that it is designed to handle causing another component to handle that stress. The other component is now handling more stress than it is designed so it fails, which causes a chain reaction in increasing defects found in bridges. Figure 4 shows the number defects found for each of the 17 bridges from 1973 to 2014. Figure 4: Periodic Repairs Events for 17 different Bridges 8
Zulfiqar, Cabieses, Khan, Mikhail
1.5 Causes of Bridge Component Failure There are many factors that cause bridge failures such as high winds, poor weather conditions, maximum loading, vibration amplification, applied stress and general wear and tear. However, the main cause of failure is due to the delay in inspections. The delay in inspection causes a delay in the detection of defects which in effect causes a delay in the time it takes to repair the defect. As shown in the figure below, an identified defect takes on average about 8 years to get repaired. The effectiveness of the external factors affecting the bridge can be minimized by identifying defects earlier than the periodic inspection process. Figure 5: Years to repair identified defect 9
Zulfiqar, Cabieses, Khan, Mikhail
1.6 Bridges Types The FHWA administers 607,380 bridges that vary in the type of material it is consisted of, the usage of the bridges as well as its structure. Under the classification of structure there are approximately more than 16 types defining the parts that need to be inspected, the time required, and the necessary equipment for the inspection. Bridges classified by material define the age and how it needs to be replaced or repaired. The final classification of bridges is defined by use, which demonstrate the frequency that the capacity of the bridge is utilized. The four main structure of bridges are the Arch, Beam, Suspension, and Cantilever. One of the oldest types of bridge in existence is the Arch bridge as seen in figure 7. They date back to the pre­Roman era and have been extensively in use because of their natural strength. An Arch bridge can be visualized as two parts of a whole, the curved bottom across the landscape and a flat span. The weight on an arch bridge is transferred outward along the curve of the arch to the supports at each end. As a consequence, the supports, called abutments keep the ends of the bridge from spreading out. Figure 7: Arch Bridge As depicted in figure 8, a beam bridge is the most basic type of bridge that can be constructed. It consists of a horizontal beam that is supported on the ends. Any weight that is on the bridge is transferred from the horizontal beam to the vertically supported 10
Zulfiqar, Cabieses, Khan, Mikhail
ends. A beam bridge is also called a girder bridge, and the supported ends are often known as piers. The beam itself is built from materials that are strong, yet flexible, so that it is able to support its own weight as well as the weight on the bridge. When a weight pushes down on the horizontal beam, the beam stretches at the bottom and compresses on the top. This is commonly called tension and compression. Figure 8: Beam Bridge Suspension bridges are majestic due to the span distances they cover. They usually appear light and strong, and can run longer than any other kind of bridge. As the name suggests, a suspension bridge suspends the causeway from huge main cables, which extend from one end of the bridge to the other as seen in figure 9. These cables rest on high towers and are secured at each end by anchorages. The towers help the main cables to be extended over long distances, and hence derives the majestic look. In the case of these bridges, the cables to the anchorages carry the weight of the bridge. The anchorages in turn are embedded in either solid rock or heavy concrete blocks. Inside these anchorages, cables evenly distribute the load and prevent them from breaking free. 11
Zulfiqar, Cabieses, Khan, Mikhail
Figure 9: Suspension Bridge A cantilever bridge is the last major type of bridge where two horizontal structures support a third horizontal structure on which a load is carried. The two horizontal structures must be anchored, and this must be done well. Cantilever bridges are a modified form of beam bridge, where the support instead of being placed at the end, is placed somewhere in the middle of the span as shown in figure 10. A cantilever is a structure that is unsupported at one end but supported at the other, very similar to diving boards in a swimming pool. This configuration made longer spans possible and wider clearance beneath. Figure 10: Cantilever Bridge 12
Zulfiqar, Cabieses, Khan, Mikhail
1.7 Bridge Components There are four main bridge components which vary depending on the kind of bridge structure. These components are the deck, superstructure, substructure and truss. The deck ​
is the roadway portion or surface of a bridge as shown in figure 11. It is where the most external load is applied due to the traffic on the bridge consisting of expansion joints, sidewalks and railings, and shoulders. This bridge component is composed of concrete which is critical to know when performing the inspection. In concrete, inspectors look for things such as cracks and its cause which involve structure, flexure, shear, crack size, nonstructural, and crack orientation. Figure 11: Components of a Bridge The superstructure ​
are the components that take the load from the deck exerted by external forces and transfer them to the substructure. It consists of the floor beams, girders, bearing and stiffeners as shown in figure 11. The superstructure is generally composed of metal and it is where the most common failures are found. A failure in one of the superstructure may result in the collapsing of the bridge, therefore inspectors tend to pay close attention to these components. 13
Zulfiqar, Cabieses, Khan, Mikhail
The substructure, as seen in figure 11, are the components that receive the loads from the superstructure and transfers them to the ground. The substructure is generally made of concrete, consisting of abutments or end­bents, columns, wall piers or interior bents, footings and piling. Trusses are organized as straight elements that are connected at the ends by hinges to develop a secure arrangement. The truss consists of a grouping of triangles that are manufactured from straight and steel bars On application of loads on the truss joints, forces are communicated to the truss elements. The steel truss bridge members are in compression or tension making this component possess a high ratio of strength to weight. 14
Zulfiqar, Cabieses, Khan, Mikhail
1.8 Structural Vibration All bridges undergo some form of dynamic loading, which cause them to
vibrate. Bridges have always been subjected to dynamic influences due to vehicles
driving over them. These dynamic influences can lead to deterioration of the bridge
caused by the rapid growth of highway transportation. The dynamic loading that all
bridges undergo vary from the forces of the bridge structure, to the car engine of the
vehicle on the deck of the bridge, to the external forces/vibrations caused by the
weather such as the wind.
Structural vibration is a repetitive motion that can be measured and
observed in a structure, which can help characterize the behavior and performance
of the structure as it responds to vibrations or dynamic loading. Unwanted vibration
can cause fatigue or degrade the performance of the structure. Therefore it is
desirable to eliminate or reduce the effects of vibration. However, in other cases,
vibrations are unavoidable and sometimes even desired helping decipher what the
changes in vibration mean. Understanding the effect on the structure is essential in
order to control and modify the vibration or if necessary to minimize the structural
response.
There are four methods of analyzing vibrations, which include free, forced,
sinusoidal, and random vibration. The free vibration is the frequency at which the
structure wants to naturally vibrate at. It is the natural response of a structure to
some impact or displacement. The response is completely determined by the
properties of the structure, and its vibration can be understood by examining the
structure’s mechanical properties. The forced vibration is the response of a
structure to repetitive forcing function that causes the structure to vibrate at the
frequency of the excitation. Sinusoidal vibration causes the structure that is excited
by the forcing function to vibrate at a pure tone with a single frequency such as a
15
Zulfiqar, Cabieses, Khan, Mikhail
waveform. Finally, random vibration is composed of a complex combination of
external forces applied to the structure.
The vibration that bridges exhibit through different external forces are used
to predict the condition of the bridge. Each bridge is unique in its vibration. This
uniqueness comes from the mass of the bridge, the stiffness of its components, the
material that the bridge is built from and other parameters that are specific to each
bridge such as its shape, length and number of lanes. Other dynamic influences
occur due to the traffic that goes on the bridge, having an effect on the vibration of
the bridge.
Any change in the structure of the bridge will have an impact on that
vibration of the bridge; the cause of this change could be due to a deficiency.
Knowing that the deficiency will cause a change in vibration allows the bridge to be
inspected with this method as it detects the condition of the bridge from the change
in vibration. However, the changes in frequency vary allowing the detection of what
parameter is actually being affected to be known. For instance, Figure 12 shows the
red line to be the natural frequency obtained from the free vibration analysis
method, whereas the blue line demonstrates a phase shift to the left stating that the
stiffness of the structure has decreased. However, if the phase shift had been to the
right it will show that the structure has become stiffer. Thoroughly analyzing these
diagrams is a prominent concern in order to identify the condition of the bridge
structure.
16
Zulfiqar, Cabieses, Khan, Mikhail
Figure 12: Natural Frequency (red) vs. Phase Shift (blue)
1.9 Main Components and Failure Types As a bridge gets older, its behavior changes due to the change in its actual components. A change in the material of the component can affect the behavior of the bridge. These changes are caused by external factors that lead to failures. The main failure types are found on the deck, substructure, and superstructure. Associated with these components is the kind of material the component is made of as well as the most common failure types that correspond to these components. These main failure types are Collision, Section Loss, High Stress, and Fatigue with percentages of 13.05%, 3.26%, 20.65%, and 16.3% respectively [7]. Although the current inspection method is manual, these can all be inspected and detected using the Bridge Vibration Monitoring System (BVMS). 17
Zulfiqar, Cabieses, Khan, Mikhail
Figure 13: Main Components & Failure Types 2.0 Stakeholder Analysis A stakeholder is a group, organization, member or system who can be affected by a system. The primary stakeholders of the bridge inspection systems are the Federal Highway Administration (FHWA), District Department of Transportation (DDOT), Bridge Design Engineers, Bridge Construction Team, and the Inspection Team which are directly involved with the construction and inspection of the bridge. The secondary stakeholders are groups of people that are not directly involved with the construction and inspection but are impacted indirectly such as the traveling public or the Bridge Users. 18
Zulfiqar, Cabieses, Khan, Mikhail
2.1 Primary Stakeholders 2.1.1 Federal Highway Administration The Federal Highway Administration (FHWA) is an agency within the U.S Department of Transportation established in 1966 ​
that supports State and local governments in the design, construction, and maintenance of the Nation’s highway system and various federally and tribal owned lands. Through financial and technical assistance to State and local governments, the Federal Highway Administration is responsible for ensuring that America’s roads and highways continue to be among the safest and most technologically sound [11]. Therefore, FHWA’s main aim is to improve safety, mobility, and livability, and to encourage innovation. 2.1.2 Department of Transportation Department of Transportation (DOT) ​
serves the U.S. by ensuring a fast, safe, efficient, accessible and convenient transportation system that enhances the quality of life for residents and visitors by ensuring that people, goods, and information move efficiently and safely [12]. ​
Specific DOTs are involved in the planning, designing, construction, and maintenance for the states streets, alleys, sidewalks, bridges, traffic signals, and street lights. Also managing and making improvements to the street system to facilitate traffic flow through the state. 2.1.3 Design Engineers Designers and engineers include the groups that design the entire system providing a cost effective solution that minimizes time. The designers are responsible for the design of the bridge that needs to be constructed. Their main objective is to design a strong and light bridge that provides safety and longevity. 19
Zulfiqar, Cabieses, Khan, Mikhail
2.1.4 Construction Team The construction team includes the contractors who are responsible for building and constructing bridges that are safe for the travelers. 2.1.5 Inspection Team The inspection team includes the Inspecting Engineers, the managers and the contractors who detect the defected parts and areas of the bridge. The inspection team monitors the bridge at intervals and detects any defective parts for safety. There are two different types of inspectors. The team leaders who are experienced specifically for these inspections and the general inspectors that are trained due to insufficient amount of inspectors. Team leaders receive a minimal salary wage whereas the general inspectors may get paid in an hourly rate. Their main objective is to provide safety through inspections, minimize inspection time and cost and maximize their pay. 2.2 Secondary Stakeholders 2.2.1 Bridge Users The bridge users include the passengers, commuters and any transporting goods. They are not directly involved in the inspection or construction of these bridges, nonetheless, the bridge partial shutdown for these inspections or maintenance extends their travel time. 2.3 Stakeholder Interactions The engineering design of the system, as illustrated in Figure 14, starts with the owner or DOT hiring a consulting engineering company to design the desired bridge. The 20
Zulfiqar, Cabieses, Khan, Mikhail
design engineers then design the bridge and sends it back to the owner. The owners of the bridge then bid the bridge project to contractors, tell them the plan and sends them the bridge design to have the bridge built. The contractors bid competitively, and the winning contractor constructs the bridge. After the construction of the bridge, the owners hire inspection engineers and contractors to inspect the bridge [13]. Figure 14: Stakeholder Interactions (grey) and Stakeholder tensions (colored) 2.4 Stakeholder Tensions There are several tensions that arise amongst the stakeholders involved in the bridge inspection system due to their differing objectives. As depicted in Figure 14, based on the objectives we can see that both FHWA and DOT would support an alternate bridge inspection system which would reduce the costs and assure public safety. This 21
Zulfiqar, Cabieses, Khan, Mikhail
alternate system would ​
eliminate traffic congestion caused by the closure of lanes which ultimately dissatisfied bridge users. ​
The bridge users support the inspection, restoration, and protection of the bridges; however, as previously stated they disagree with the current process of closing lanes. An alternate inspection system that does not involve manual inspection and contribute to traffic congestion would be highly favored by bridge users. Although the bridge designers and the Construction team are not directly affected there are some tensions due to the fact that the inspection team can hold the designers liable if the bridge were to not behave the way it was designed to behave or if the bridge was not constructed properly [13]. ​
On the other hand, the inspection team would oppose an alternative bridge inspection system since it would eliminate manual labor and thus the need for these inspectors. This would either result in job losses or major financial concerns for the inspectors. Furthermore, the inspection team could also be held liable by the department of transportation is something on the bridge was not inspected correctly. These tensions between the stakeholders define our problem and present a need for an alternative system. 22
Zulfiqar, Cabieses, Khan, Mikhail
3.0 Problem Statement Bridges play a key role in the national transportation system, and the ability to assess their condition is vital for safe transportation operations. ​
One of the problems, as discussed in the context, is the very high biannual inspection cost. Costing the U.S., $ 2.7 billion for one inspection cycle. Another major problem is that the periodic biannual inspections cause a delay in detection of deficiencies since a possible defect isn’t detected until the next inspection cycle. This causes the repairs to be performed late on the bridge which leads into the next major problem. The delay in repairs causes a lag time in the repair times that puts stress on other components of the bridge. When a component on the bridge is not repaired early enough, it leads to other bridge components’ failures. Therefore, overtime there is an exponential growth in the defects found on a bridge as shown earlier. 23
Zulfiqar, Cabieses, Khan, Mikhail
4.0 Need Statement There is a need for the FHWA to improve the efficiency of the periodic bridge inspection process. The cost of the periodic bridge inspection process is very high. Therefore, the FHWA needs to reduce the total inspection cost which includes traffic control cost, labor cost, and equipment cost by decreasing the rate of inspection. Having less frequent inspections will reduce the use of inspecting equipment, hiring labor and traffic congestions reducing the total cost of inspections per year. Additionally, there is a need for a system that can detect deficiencies right when they occur rather than waiting for the next inspection to happen. Therefore, there is a need for a bridge monitoring system that will monitor the bridge on a continuous basis so all deficiencies are detected right as they occur. This will allow repairs to be identified earlier and therefore decrease bridge failures. 5.0 Requirements 5.1 System Requirements The requirements of this system were derived from the statement of need in
order to eliminate the stakeholder tensions. The prominent objective is to reduce
the cost and time of inspection by reducing the frequency of inspections. The
Mission of the Bridge Vibration Monitoring System (BVMS) is to monitor all bridge
components and detect when deficiencies occur on the bridge. The BVMS shall
reduce the number of inspection performed on a bridge and increase the rate of
detection of deficiencies. The BVMS shall also be available at all times, improve
24
Zulfiqar, Cabieses, Khan, Mikhail
safety for all inspectors and ride users, and be able to communicate all the data
collected to the base.
5.2 Design Requirements The new system shall consist of the following functional components: Acceleration detection sensors, Digital Acquisition Unit (DAU), communication between sensors and DAU, Base monitoring unit, and communication between DAU and base monitoring unit. The system shall alert the DAU with frequencies of each accelerometer each day and send natural frequencies captured by the sensors to the DAU via a communication network system. The system shall also alert the base if the frequency of the accelerometer captures a deficiency for 7 consecutive days. Additionally, the system shall use fourier analysis to convert the vibration data from the time domain to the frequency domain. 5.3 Component Diagram The major components of the Bridge Vibration Monitoring System are shown in figure 15. The BVMS components start with accelerometers that experience vibrations that output electrical signals. This output is sent to the DAU where it is converted from a time domain to a frequency domain using fourier transform analysis. Then the frequency is analyzed and compared to the reference vibrations fingerprint frequency. If there is a change in the natural frequency from the reference vibrations for 7 consecutive days, the DAU alarms the base monitoring unit. However, if there is no change detected from the natural frequency compared to the reference vibration it loops back and and monitors again. 25
Zulfiqar, Cabieses, Khan, Mikhail
Figure 15: Component Diagram for Bridge Vibration Monitoring System 26
Zulfiqar, Cabieses, Khan, Mikhail
6.0 Proposed Solutions 6.1 Bridge Vibration Monitoring System: Concept of Operations The proposed alternative is an Event­Based Bridge Vibration Monitoring System (BVMS) substituting the Periodic Inspection Process. This transitions the process from inspecting bridges biannually to when a deficiency is detected. The Event­Based Inspection is a hybrid process that uses accelerometers to identify if a deficiency exists and manual inspection to inspect the entire bridge. Allocating inspections to bridges that require inspections, rather than periodically inspecting bridges that do not require it. Therefore, this transition does not project more cost due to a tradeoff of inspecting bridges in sound condition bi­annually and structurally deficient bridges more frequently to inspecting bridges where a defect is known to exist. The BVMS allows the deficiency on the bridge to be detected in the instance it was created. The figure below shows the event based system. The placed accelerometers captures the components unique vibration frequencies. These vibration frequencies are compared to the reference fingerprint frequencies acquired by Dr. Lattanzi. This comparison identifies the condition of the bridge component and sends that information to the data acquisition unit (DAU). If the condition remains in a fair or poor condition for 7 consecutive days it sends the data from the DAU to the base. Figure 16: Bridge Vibration Monitoring System 27
Zulfiqar, Cabieses, Khan, Mikhail
6.2 Accelerometers Electronic sensors that convert vibration motion into electrical signals can measure structural vibrations. Electronic motion sensors are devices that detect when the original condition has changed. The sensor has a normal state and reports when the standard status is disturbed. A considered alternative is accelerometers, which is an electromechanical device that measures vibration or acceleration forces. Accelerometers designed to measure vibration are based on the piezoelectric effect. Piezoelectric accelerometers are used when a mass applies force (pressure) to a nonconducting material creating a high­impedance charge resulting in a voltage across the material (Fig. 17). Essentially, this effect generates an electrical output that is proportional to the applied acceleration generating an opposed buildup of charged particles on the piezoelectric material. This charge is proportional to the applied force. This applied force alters the alignment of positive and negative ions, which results in an accumulation of these charged ions on opposed sides. Accelerometers obey Newton's law of motion, F=ma. Therefore, the total amount of accumulated charge is proportional to the applied force, and the applied force is proportional to acceleration. Figure 17: General Accelerometer Dynamics 28
Zulfiqar, Cabieses, Khan, Mikhail
Some requirements for accelerometers are current excitation, AC Coupling, Grounding, Filtering, and Dynamic Range. The current excitation is the external current needed to power the amplifier. This is necessary when choosing a data acquisition system since one needs to know the excitation voltage the accelerometer requires or the excitation may not be sufficient for the accelerometer to work. The accelerometer requires AC Coupling since it acquires a signal from both the direct current (DC) and alternating current (AC) where the DC portion offsets the AC portion from zero. The AC Coupling removes the DC offset by having a capacitor in series with the signal. This eliminates the long­term DC drift that sensors have due to age and temperature effect, increasing the resolution and the usable dynamic range of the system. Another requirement is properly grounding the system due to the fact that vibration measurements are highly susceptible to noise especially in an electrically conductive surfaces allowing noise to enter the ground path of the measurement signal through the base of the accelerometer. One can avoid improper grounding by grounding either the signal conditioning input or the sensor but not both. If the sensor is grounded, it must connect differentially. If the sensor is floating, the signal conditioning systems should be connected inverting input to ground. Adding a low­pass filter is also required since it reduces signals with frequencies higher than the cutoff frequency and passes signals with a frequency lower than the certain cutoff frequency. This identifies the correct range of frequencies required not allowing samples that distort the measurement. Compliant materials, such as a rubber interface pad, can create a mechanical filtering effect by isolating and damping high­frequency transmissibility. Here is where the frequency range is established. The last requirement is a dynamic range, which is a measure of how small you can measure a signal relative to the maximum input signal the device can measure, expressed in decibels. Thus, the input range and the specified dynamic range are important for determining the needs of the system. 29
Zulfiqar, Cabieses, Khan, Mikhail
Choosing an accelerometer that complies with the above requirements will allow accurate data to be collected when incorporating them for the bridge inspection process. BDI accelerometers are specifically designed for dynamic structural monitoring and measure frequencies in 3­dimensions. Obtaining the structures’ vibration in time domain, however, using fourier transform to convert it into frequency domain (Lattanzi, 2014). In the frequency domain, the natural frequency of the component is determined per day for each placed accelerometer on the bridge. However, we must first identify the optimal location to place these accelerometers. The most suitable location will vary per bridge due to its complexity. Nonetheless, incorporating these electronic sensors in these certain parts of the bridge will help analyze if any part inspected has changed from its original state and if so, by how much, stating the level and type of the vibration. This will allow us to know if these parts of the bridge are deficient since the accelerometer will be modified to a particular specified range allowing us to see any variation in its condition. The level of variation according to its recorded previous condition would allow us to quantify the change and ultimately evaluate if the inspected part needs to be repair, replaced, restored or ultimately left unchanged. 30
Zulfiqar, Cabieses, Khan, Mikhail
6.3 Proposed Alternatives 6.3.1 Periodic Manual Inspection 6.3.1.1 Manual Bridge Inspection As previously stated, the current process is a periodic manual inspection process as shown in the Figure 18. The actual on­site inspection takes about 1 to 3 days depending on the length and complexity of the bridge. Additionally the cost of inspection for one bridge ranges on average from $4,500 to $30,000 per inspection. Some advantage associated with this process are the fact that inspectors are comfortable with this process and the quality of inspection is said to be 95% reliable [9]. However, this method involves manual labor incurring costs, it is time consuming, and hazardous for the personnel. Figure 18: Periodic Manual Inspection 6.3.2 Event­Based BVMS The Event­Based alternative is to place the accelerometers at the fracture critical locations of the bridge with a certain distance that will cover and measure frequencies of 31
Zulfiqar, Cabieses, Khan, Mikhail
the entire bridge. If a deficiency occurs around or in the area that the accelerometer is placed, it is more susceptible to be detected. As mentioned previously, the accelerometers can be placed on the bridge ensuring coverage for all areas and components of the bridge. The distance at which these accelerometer will be placed can be determined by the range of data transmission for a specific accelerometer. The fracture critical locations with the highest shear and moment establish where the placement of the accelerometer should be which is where a deficiency is more likely to happen. By using the Theodore Roosevelt Bridge as a sample bridge, 31 fracture critical locations were identified which are the piers for high shear and the middle of the span for high moment. Therefore, the Event­Based BVMS has two proposed alternatives 1) Hybrid Wired Mounted Accelerometer with Manual Inspections and 2) Hybrid Wireless Mounted Accelerometer with Manual Inspections. 6.3.2.1 Hybrid Wired Accelerometer with Manual Inspections The first proposed Event­Based alternative is a Hybrid Wired Mounted Accelerometer with Manual Inspections. This alternative involves installing these accelerometers mounted onto the bridge with cables that will serve as both a power source and a communication system as seen in figure 18. One cable is connected to a power source through all the accelerometers to supply power. Additionally, a method for data collection is needed. This communication of data is accomplished through an additional cable. A second cable is ran through the accelerometers connected at the end to Data Acquisition Unit (DAU) where all the data will be stored. Analyzing the data gives the condition of the bridge components that were monitored. If the condition obtained shows any deficiency for 7 consecutive days, the base monitoring unit will be alarmed. The 7 day threshold eliminates possible external forces that can affect the frequency reading of the accelerometers. Replicating the days of the week, accounting for different traffic patterns throughout a week. After an alarm, the base monitoring unit will then send inspectors to manually inspect the whole bridge. 32
Zulfiqar, Cabieses, Khan, Mikhail
The time the actual onsite inspection will take is the same as the current periodic manual on site inspection: 1 to 3 days. Moreover, the cost of the actual onsite inspection is $4,500 per inspection and the acquisition cost for the wired sensors is $77000. An advantage of this design is that it is easy to install and allows for fast power supply and data transmission. However, installing cables requires many accessories such as clamps, tapes, and adhesives for secure connection which increases the cost [4]. Furthermore, another shortcoming of this method is that using accelerometers will require the inspectors to acquire new technological skills to be able to interpret the data gathered. Figure 18: Hybrid Wired Mounted Accelerometer with Manual Inspections 6.3.2.1 Hybrid Wireless Accelerometer with Manual Inspections Consequently, the second proposed Event­Based alternative is a Hybrid Wireless Mounted Accelerometer with Manual Inspections. The concept for this design is the same as the concept of the first design alternative however, in this case the cables for power and communication will be replaced by battery power and the communication between the devices will be wireless as seen in figure 19. This alternative consists of mounting these sensors connected to a low power communication system that will transport the acquired data to a transportable device capable of taking the collected data from the bridge site to the ground base station. 33
Zulfiqar, Cabieses, Khan, Mikhail
Figure 19: Hybrid Wireless Mounted Accelerometer with Manual Inspections This alternative includes a communication system that transmits the information to the data acquisition unit. For this Event­Based alternative, the accelerometers are mounted onto individual microcontrollers that provide power and data transmission. The microcontrollers have batteries for power supply, antennas for communication, and a memory storage to store the frequency data from the accelerometers. These sensors form a wireless sensor network which is a network of spatially connected sensors that monitor physical or environmental conditions [14]. The accelerometers on the bridge communicate with each other through the antennas and send their data down to the data acquisition unit. Like the wired sensor alternative, the time for the actual onsite inspection is 1 to 3 days and the cost of the onsite inspection is $4,500 per inspection. Additionally, there is an acquisition cost of $75,000 and a concurrent cost of $1000 per year. This method provides quick real time data through the antennas as it is all a part of the network. However, some of the major shortcomings of this alternative is that it also, like the first Event­Based alternative, requires the inspectors to acquire new skills. Another concern, is that the integrity and security of the data is questioned since it is all part of a network. 34
Zulfiqar, Cabieses, Khan, Mikhail
7.0 Method of Analysis The method of analysis is a queuing system used to calculate the total number of defects found in year i. Using the periodic manual inspection historical data, the probability of identifying a defect per bridge in year i and the probability of the time it takes to repair the identified defect per bridge in year i were calculated. Providing a table of years to fix a defect. If the defect is fixed in year i, it is taken out of the queuing system, however, at the same time there is an arrival of a newly generated defect in year i. Providing the total cumulative defect found per bridge in year i. Since this process is done bi­annually, there is a feedback loop to repeat this process every two years. Figure 20: Method of Analysis 35
Zulfiqar, Cabieses, Khan, Mikhail
8.0 Simulation 8.1 Simulation Requirements The simulation requirements are as follows: 1. The simulation shall use the periodic historical data for deficiencies found on a bridge to generate the number of defects. 2. The simulation shall compare the probability of finding the number of defects on a bridge to the randomly generated probability which uses a uniform distribution to identify the number of defects found bi­annually. 3. The simulation shall use the probability of the time to repair a defect found to assign the number of years it will take to repair an identified defect. 8.2 Simulation Model Figure 21: Probability of identifying a defect per year per bridge 36
Zulfiqar, Cabieses, Khan, Mikhail
The probability of identifying a defect per bridge in year i demonstrates a multinomial distribution with ​
26 inspections for 51 years for 17 bridges displayed in the horizontal axis of the chart. Having a fixed success probability of identifying a defect for that given year displayed in the vertical axis of the chart. This probability uses the analog of Bernoulli trials with its corresponding probability for finding the defect or not, leading to the total number of defects observed in the year of the inspection. These probabilities are used in order to perform a monte carlo simulation in which a uniform random number generator generates the probability of finding the number of defects for that given year for a particular bridge for 51 years repeated for 100 bridges. Figure 22: Probability of repairing a defect per year per bridge The probability of the time it takes to repair a defect was incorporated in order to identify the total number of defects found after several years after the periodic repairs. The histogram identifies that the average time it takes to repair a defect is 8 years. This is also used in the monte carlo simulation in which a uniform random number generator generates the probability of repairing the already identified defects for the 100 bridges. Giving an output of the lag time between the identified repair to the actual repair. After running the model for 100 iterations the average total number of defects per year and average time to repair a defect per year for the 100 bridges was calculated. 37
Zulfiqar, Cabieses, Khan, Mikhail
Outputting the Bridge Number, the year the defect was found per bridge, the total number of defects found per year per bridge and the mean time to repair the defect found per year per bridge. The simulated historical data was submitted to a sensitivity analysis. 8.3 Simulation Results Figure 23 demonstrates a sample output from the simulation. Displaying the defects found per year and the years to repair the defects found. The number of defects found per year were found through averaging out the number of defects found for 100 iterations for each year shown. The same iterations were done for the repairs. The “Historical Data” output shows the number of defects found per year and the repair time for those defects using historical data. The “Repairs happening 20% earlier” output shows the number of defects found per year and the repair time when the repairs are happening 20% earlier on the bridge. There are 7 defects lefts on the bridge at year 51 when repairs are done 20% of the time earlier compared to 12 defects left using historical data. This is due to the fact that the bridge components with the deficiencies are not able to handle the stress until they are repairs; hence, other components on the bridge end up handling that stress that they are not designed for, leading to more deficiencies. This domino­effect leads to an unsafe bridge with an exponential increase in deficiencies. 38
Zulfiqar, Cabieses, Khan, Mikhail
Figure 23 : Sample Simulation Results 8.4 Design of Experiment In order to see whether a decrease in the time to repair a bridge has an impact on the life­span of the bridge, the total defects remaining on a bridge after 51 years were analysed using sensitivity analysis. The average number of defects remaining after 51 years for 100 different bridges were calculated using the equation shown in Figure 24. Figure 24: Equation used to calculated total defects remaining after 51 years The design of experiment, shown in Figure 25, was done through performing sensitivity analysis on the results obtained from the simulation. The average number of 39
Zulfiqar, Cabieses, Khan, Mikhail
defects remaining after 50 years for 100 bridges are 10.13 when the “Mean time to repair a defect found” is decreased by 20%, when compared to the defects remaining when the “Mean time to repair a defect found” is calculated using historical data. Figure 25: Design of Experiment 40
Zulfiqar, Cabieses, Khan, Mikhail
9.0 Life­Cycle cost The Life Cycle Cost (LCC) that is considered in this study is based on a comparison of two alternatives: the current periodic manual inspection and the proposed Event­Based BVMS. The BVMS alternative has two options: Wired Accelerometers and Wireless Accelerometers. The only difference between these two options in the BVMS alternative is the way that these accelerometers are connected and the way data is transferred between the accelerometers and the Data Acquisition Unit (DAU). The Wireless Accelerometers option allows the data to transfer between accelerometers and the DAU wirelessly via the micro­controllers. The other option in the BVMS alternative is the Wired Accelerometers; in this option the data will be transferred between accelerometers via wires that are connecting the accelerometers and the DAU. A 50­year operation time for each alternative is considered which is based on the life expectancy of a bridge. In order to make a comparison between costs of different alternatives a common bridge has to be chosen to eliminate variables that are not counted for in the model. In this case, the Theodore Roosevelt Bridge was chosen as a common domain for all alternatives to calculate the 50­year life cost. 41
Zulfiqar, Cabieses, Khan, Mikhail
9.1 Current Periodic Manual Inspection The first alternative is the current periodic manual inspection. The cost for this alternative is calculated using the following equation: The cost for each inspection for Theodore Roosevelt Bridge is estimated to be $30,000 based on expert’s estimation [1]. Numbers of inspection per bridge is estimated to be 20 inspections in the first 40 years of the bridge’s lifetime based on the biannual inspection regulation and an annual inspection for the last 10 years of the bridge’s lifetime due to the deterioration in the bridge’s condition. This yields a total of 30 inspections that are needed for a bridge during the 50­year lifetime. Using the above equation, cost for manual inspection will result in a total of $900,000 for 50­year inspections. 42
Zulfiqar, Cabieses, Khan, Mikhail
9.2 Event­Based BVMS Cost Calculations As mentioned earlier the BVMS alternative is split into two options based on the way the accelerometers are connected: Wired Accelerometers and Wireless Accelerometers. The next section will compare the cost of the two options. 9.2.1 Hybrid Wired Accelerometer with Manual Inspections The cost of this option is calculated based on two factors: the equipment cost and the inspections needed. An important factor to take in consideration is the equipment replacement, since these equipment are subject to failure due to their life span. The other factor that contributes to the cost of this alternative is the inspection cost. The details for the equipment cost and the inspection cost are specified in the BVMS Equipment Cost section & BVMS Inspection Cost section. Furthermore, the total cost for this alternative was calculated for the 50­year life span in order to be comparable with the 50­year cost for the current manual inspection. 9.2.2 Hybrid Wireless Accelerometer with Manual Inspections The cost for this option was also based on two factors: the equipment cost and the inspections needed. A replacement cost for the equipment was also considered and a total of 50 years operation cost was calculated in order to compare the three alternatives. The total cost is calculated for the cost of equipment for the 50­year life span and adding to that the 50 years total expected inspections needed. The total cost is discussed in details in the BVMS total cost section. 9.2.3 BVMS Equipment Cost There are two equipment costs: The equipment cost for wired accelerometers alternative and the equipment cost for the wireless accelerometers alternative. The BVMS 43
Zulfiqar, Cabieses, Khan, Mikhail
equipment cost for the wired alternative is calculated based on the following equation: The total cost is associated with “set” terminology because this set will need replacement based on its life span expectancy. Best, expected and worst costs were calculated for the equipment cost based on the life expectancy of the equipment set where the worst lasts 5 years, the best lasts 10 years and the expected lasts 8 years. The following graph shows the equipment cost for the wired alternative through a 50­year lifespan. Figure 26: Total equipment cost for the Wired Accelerometers alternative within 50 years 44
Zulfiqar, Cabieses, Khan, Mikhail
Additionally, the cost for the wireless accelerometers alternative equipment cost throughout the 50­year lifespan is calculated based on the following equation: The “set” is again needed in this case due the fact that equipment will need to be replaced based on the life expectancy of the set. The 50­year estimated cost was calculated based on the life expectancy for the equipment. The following graph shows the best, expected and worst costs for the BVMS wireless option. Figure 27: Total equipment cost for the Wireless Accelerometers alternative within 50 years 45
Zulfiqar, Cabieses, Khan, Mikhail
9.2.4 BVMS Inspection Cost The number of inspections needed was calculated based on historical data that was shown in previous sections. A total cost for inspections was calculated for the 50 years life span based on the following equation: Based on the simulated data for 100 bridges the number of inspections needed per bridge for its lifetime was detected. Then, the total cost for inspections needed for that bridge was calculated and analyzed. The analysis illustrated that the costs for these bridges were distributed normally with a mean of $267,600 and a standard deviation of $48,500. The number of unnecessary inspection cannot be detected accurately which is due to the way the historical data was collected. VDOT started archiving the historical data in year 1973. However, most of these bridges were built in the 1950s and 1960s, which is a gap of 4 to 20+ years with no information about the conditions of these bridges. An observation leads to a conclusion that the recorded historical data shows the absence of defect in the first 4 years from the recorded historical data. Furthermore, 4 bridges that were built in 1961 have recorded defects in the first year from the recorded historical data in 1973. That observation leads to a conclusion that the unrecorded historical data is zero defects present from the year they were built to the year VDOT have started archiving the inspection data. Using these historical data analysis and the simulated data, the costs for inspections needed per bridge were calculated and categorized into best, expected and worst case scenarios from the cost perspective. The best and the worst cases are the two ends of the normal distribution with 95% confidence. The range for the best and the worst costs are $170.5k and $364.7k respectively and the expected cost falls within the range 46
Zulfiqar, Cabieses, Khan, Mikhail
$267,600 and $270,000. These two values were detected based on the mean for the normal distribution that is associated with the total inspection cost for the bridges and the mode for the total lifetime costs for the bridges. 9.2.5 BVMS Total Cost The total cost that is included in this study is focused on the inspection cost and the inspection processes and regulation improvement to eliminate the unnecessary inspections in a bridge’s life time. The maintenance cost, which is the maintenance of the bridge monitoring system, is included for the proposed system in the life expectancy of the equipment that will be placed on the bridges to monitor the vibrations. The total cost of the two proposed options were calculated based on the following common equation: The 50 years inspection cost is the number of inspections that take place in a bridge’s lifetime multiplied by the price of each inspection based on the Simulated data. The 50­year Equipment cost was calculated using the following equation: The following graph explains the prices for BVMS Wired accelerometers option with best, expected and worst costs compared to the cost for the current manual process. 47
Zulfiqar, Cabieses, Khan, Mikhail
Figure 28: Total cost for the Wired Accelerometers alternative in 50 years operation time The following graph explains the prices for BVMS Wireless accelerometers option with best, expected and worst costs compared to the cost for the current manual process. Figure 29: Total cost for the Wireless Accelerometers alternative in 50 years operation time 48
Zulfiqar, Cabieses, Khan, Mikhail
10.0 Business Case 10.1 Total BVMS cost vs Periodic cost Based on the data obtained from calculated life cycle costs, it is evident that the BVMS has reduced the cost from the current inspections. The BVMS can cause savings up to 40.2% using the wireless alternative in the best case scenario and 20.2% in the expected case scenario. Additionally, as seen in the table below, in the worst case scenario where the wireless set has to be replaced every 5 years, the proposed system is costing 20% more than the current periodic inspection. Similarly, with the wired alternative the BVMS has reduced the cost from the current inspections cost with 37.7% cost savings in the best case scenario and 17% cost savings expected scenario. Additionally, in the worst case scenario, the proposed system will cost 25% extra than the current periodic inspection method. The cost analysis between the proposed BVMS system and the Current periodic inspection is summarized in the following table. 49
Zulfiqar, Cabieses, Khan, Mikhail
Table 1: Total Savings 10.2 Inspection Savings Analyzing just the manual inspection cost for the proposed BVMS system, it can be seen that it reduces cost by 60% from the current periodic inspection in the worst case of the bridge’s behaviour. In the expected number of inspections needed on a bridge throughout its lifetime, the cost savings will be 70% from the current periodic inspection. Additionally, in the best case scenario the saving is calculated to be 81% from the current periodic inspection. The cost savings shown occur due the fact that that inspections happen less often. The inspection process itself doesn’t change but the frequency of the inspections is the factor that decreases the cost from the current periodic inspection to the proposed event based inspection system. The change in the frequency of inspection is the main cause for the cost savings that occur with the proposed system at the expected and the best case scenarios. The following table summarizes the cost difference in inspection between the periodic inspection system and the event based system. Inspection Cost Event­Based vs. Periodic Worst Expected Best Event­Based 0.365M 0.268M 0.17M Periodic 0.9M 0.9M 0.9M Table 2: Inspection cost 50
Zulfiqar, Cabieses, Khan, Mikhail
10.3 Value Hierarchy Figure 32 demonstrates the most attributes that are used to measure the utility of the current periodic inspection system compared with the proposed Event­Based inspection system. The three most important attributes derived from the systems requirements are: Availability, Safety, and Communicability. Availability compares how available the alternative is. Safety compares how safe the alternative is for the Bridge Users/Inspectors. Communicability is how easily the on­site inspection data communicated with the bridge engineers. Figure 32: Value Hierarchy The weights for the three attributes shown in Figure 33 were obtained from a survey distributed to our sponsor, Dr. Lattanzi, an experienced Bridge Inspector. These attributes were then used to calculate the utility for the two alternatives. This resulted in 51
Zulfiqar, Cabieses, Khan, Mikhail
the Event­Based Alternative having a higher score of 8.9 compared to the current periodic bridge inspection system. Figure 33: Utility Score Calculations 10.4 Utility Vs. Cost The utility score was compared to the cost of the two different systems to compare and contrast which system will be a better recommendation. The utility vs. cost for the two alternatives are shown in Figure 34. For the BVMS, the best, expected, and worst case scenarios are plotted. Figure 34: Utility Vs. Cost 52
Zulfiqar, Cabieses, Khan, Mikhail
The Event­Based BVMS provides a higher utility and lower cost for the best and expected case scenarios; however, the cost is higher in the worst­case. 53
Zulfiqar, Cabieses, Khan, Mikhail
11.0 Conclusions and Recommendations The periodic bi­annual bridge inspections required under the current bridge inspection system are not reliable and are high in cost. This problem is solved with the proposed Event­Based bridge inspection system. The Event­ Based system provides a higher utility of 8.9 compared to the utility of 5.0 for the current periodic system, and a cost savings of up to 40.2%. Therefore, it is recommended that the proposed Event­Based system to be implemented. Out of the two Event­Based systems, the option with the wireless sensors with low­power communication systems is recommended due to the fact that it will not require the installation of wires for power source and communication. 54
Zulfiqar, Cabieses, Khan, Mikhail
12.0 Project Management 12.1 Budget Analysis Our individual hourly rate is $40.00 per hour, however, with the GMU overhead cost, which is a factor of 2, our total rate per person is $80.00 per hour. We arbitrarily estimated the time to be spent on the project per week, for 38 weeks using as reference the time spent per week for the first 3 weeks into the project. The estimated time to be spent on the project is 2104.05 hours with a total budgeted cost of $168,324.00. We estimated a worst and best case scenario by choosing the probability to be 15 percent, giving us a 85 percent probability of reaching our estimated goal. As the project evolved however, we’ve put more hours than was planned. The chart below shows the actual cost slightly on the same margin as our worst case scenario. Figure 35: Budget Analysis 55
Zulfiqar, Cabieses, Khan, Mikhail
The SPI and CPI ratio shows that our schedule performance index is below 1, indicating that we had been under schedule throughout the project due to the fact that the scope of the project was changed several times. The cost performance index demonstrates that we are above the projected cost. This ultimately shows that despite our time being spent on the project no process has actually been made. We need to be more productive leading to our set goal. Figure 36: Cost Performance Index (CPI) vs. Schedule Performance Index (SPI) 56
Zulfiqar, Cabieses, Khan, Mikhail
12.2 Work Breakdown Structure Figure 37: Work Breakdown Structure 57
Zulfiqar, Cabieses, Khan, Mikhail
12.3 Project Plan Figure 38: Project Plan 58
Zulfiqar, Cabieses, Khan, Mikhail
13.0 References [1] T. J. Ryan, J. E. Mann, Z. Chill, and B. Ott, “Bridge Inspector’s Reference Manual.” Federal Highway Administration, Dec­2012. [2] Mark Serridge and Torben R. Licht, “Piezoelectric Accelerometer and Vibration Preamplifier Handbook” Theory and Application Handbook, Nov­1987. [3] Guy Kulwanoski and Jeff Schnellinger, “The Principles of piezoelectric Accelerometers.” Sensors Online, Feb­2004. [4] “Accelerometer Mounting and Installation Techniques”. Wilcoxon Research, 2008. [5] MetroMess, “Accelerometer Mounting”, Manfred Weber. [6] “2013 Report Card for American Infrastructure,” 2013. [7] Cook, Wesley, "Bridge Failure Rates, Consequences, and Predictive Trends" (2014). All Graduate Theses and Dissertations. Paper 2163. http://digitalcommons.usu.edu/etd/2163 [8] FHWA Archives data for “Deficient Bridges by State and Highway System” http://www.fhwa.dot.gov/bridge/deficient.cfm [9] C. How, “Virginia Department of Transportation (VDOT),” 20­Oct­2014. 59
Zulfiqar, Cabieses, Khan, Mikhail
[10] “Bridge Construction” Department of Civil Engineering, University of Delaware http://www.ce.udel.edu/courses/CIEG%20486/Bridge.pdf [11] “What we Do”, (Federal Highway Administration) http://www.fhwa.dot.gov/ [12] “About DOT”,( Department of Transportation) http://www.dot.gov/
[13] Lattanzi, David.“George Mason University,” 26­Oct­2014. [14] Margaret, Rouse.“Wireless Sensor Network (WSN)”, (Search Data Center), searchdatacenter.techtarget.com
[15] D. Lattanzi, “Design Of An Autonomous Bridge Inspection System.” . 60