Evaluation of the Foundation and Wingwalls of Skewed Semi-Integral Bridges with Wall. Abutments. A thesis presented to.

Transcription

1 Evaluation of the Foundation and Wingwalls of Skewed Semi-Integral Bridges with Wall Abutments A thesis presented to the faculty of the Russ College of Engineering and Technology of Ohio University In partial fulfillment of the requirements for the degree Master of Science Jibril Shehu June Jibril Shehu. All Rights Reserved.

2 2 This thesis titled Evaluation of the Wingwalls and Foundation of Skewed Semi-Integral Bridges with Wall Abutments by JIBRIL SHEHU has been approved for the Department of Civil Engineering and the Russ College of Engineering and Technology by Eric P. Steinberg Associate Professor of Civil Engineering Dennis Irwin Dean, Russ College of Engineering and Technology

3 3 ABSTRACT SHEHU, JIBRIL, M.S., June 2009, Civil Engineering Evaluation of the Wingwalls and Foundation of Skewed Semi-Integral Bridges with Wall Abutments (109 pp.) Director of Thesis: Eric P. Steinberg The use of jointed type bridges has increased the maintenance cost of bridges. With the introduction of the jointless bridges such as semi-integral bridges, the high maintenance cost is eliminated. Prior research has shown that skewed semi-integral bridges tend to rotate towards the acute corners as the atmospheric temperature increases through the season. As a result of this rotation, forces are generated and transferred to the wingwalls of the bridge. In the state of Ohio, wingwalls are currently only designed to act as retaining walls for the adjacent embankment soil, and not to resist the forces exerted on them by the superstructure due to thermal expansion. In addition, wingwalls are aligned more closely parallel to the longitudinal axis of the bridge than having straight standard wingwalls. Thus, the wingwalls become more subjected to bending forces than axial compression forces. The wingwalls of two semi-integral bridges were instrumented in order retrieve data relating the thermal expansion of the superstructure to the bridge movement as well as wingwall and wall abutment tilt. Two finite element models were created for the analytical aspect of this research. The first model was created in order to gain more insight pertaining to the effect that different skew angles and span lengths have on the magnitude of the forces generated and transferred to the wingwalls due to thermal

4 4 expansion. The second model was created to study the distribution of stresses in the wingwalls and wall abutments. The field data as well as the results the analytical evaluation is presented herein. Approved: Eric P. Steinberg Associate Professor of Civil Engineering

5 5 ACKNOWLEDGMENTS I would like to dedicate this thesis to my family: Shehu A. Mafara, Talatu Shehu, Kamaruddeen Shehu, Sanusi Shehu, Hadiza Shehu, Nurudden Shehu, Fatima Shehu and Mohammed Mainasara Sa idu who s moral and financial support made my college career possible. I will never forget the strength, motivation and constant encouragement you all provided me with during the extent of my study. I would also like to thank my grandparents, Malam Haruna Abdulkareem and Baluba Haruna Abdulkareem whose constant prayers have put me through the obstacles I faced during my study. This thesis would not have been possible without the support, encouragement and patience of my graduate research and faculty advisor, Dr. Eric P. Steinberg, under whose supervision I chose this topic and began the thesis project. Thank you for giving me this wonderful opportunity to study and conduct my research under your supervision. I would like to thank my graduate chair Dr. Shad M. Sargand for giving me the opportunity to do my graduate study at Ohio University as well as provided me with technical assistance throughout my thesis project. In addition, I would like thank Dr. Teruhisa Masada and Dr. Xiaoping Shen for being members of my comity and for providing me with the technical assistance to complete the thesis. I would also like to thank Issam Khoury whose relentless technical support facilitated my project. In recognition of all your help and support, I would like to thank my friends Bella Anne Ndibuisi, DeVon A. Ortiz, Musonda Kapatamoyo, Oludamilola A. Daramola and Suleman Wadur. The motivation and kindness that you all provided played a major role to the successful completion of my college career. I cannot end without thanking my

6 6 colleagues and friends in the Civil Engineering graduate office. Your assistance throughout the extent of my project is highly appreciated.

7 7 TABLE OF CONTENTS Page ABSTRACT... 3 ACKNOWLEDGMENTS... 5 LIST OF TABLES... 9 LIST OF FIGURES CHAPTER 1: INTRODUCTION Scope Integral and Semi-Integral Bridges Objectives Outline CHAPTER 2: BACKGROUND Literature Review Project Location and Bridge Background CHAPTER 3: FIELD EVALUATION Instrumentation Selection and Operation Measurement of Expansion and Contraction of Bridge Wingwall and Wall Abutment Tilt Angle Measurements Temperature Measurements CHAPTER 4: ANALYTICAL EVALUATION System Analysis Analysis of Wingwall and Wall Abutment... 58

8 8 CHAPTER 5: RESULTS Field Results Bridge Movement Abutment and Wingwall Tilt Analysis Results Forces and Stress in Wingwalls at Wingwall/Diaphragm Joint Stress Distribution on Wall Abutment and Wingwall CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS Instrumentation Scheme and Data Acquisition The Project Recommendations BIBLIOGRAPHY

9 9 LIST OF TABLES Page Table 1: Soil Stiffness Conversion to Spring Stiffness Table 2: PEJF Test Result Table 3: Measured East Bound Internal Temperatures and Joint Gap Table 4: Change in East Bound Internal Temperatures and Gap Displacements Table 5: Measured West Bound Internal Temperatures and Joint Gap Table 6: Change in West Bound Internal Temperatures and Gap Displacements Table 7: Measured East Abutment and Wingwall Tilt Table 8: Measured West Abutment and Wingwall Tilt Table 9: Change in East Abutment and Wingwall Tilt Table 10: Change in West Abutment and Wingwall Tilt Table 11: Forces in East Wingwall Table 12: Stresses in East Wingwall Table 13: Forces in West Wingwall Table 14: Stresses in West Wingwall Table 15: Analysis East Wall Joint Movement (45 o skew, and k soil = lb/in 3 ) Table 16: Analysis West Wall Joint Movement (45 o skew, and k soil = lb/in 3 )... 90

10 10 LIST OF FIGURES Page Figure 1: Semi-Integral Abutment and Diaphragm (Elevation View) Figure 2: Semi-Integral Abutment and Diaphragm (Side View) Figure 3: Forces on Bridge causing Horizontal Plane Rotation Figure 4: Turned-Back wingwall (Parallel Alignment to Bridge Longitudinal Axis) Figure 5: Standard Straight Wingwall position Figure 6: Semi-integral/Integral Abutment type Skew Vs Bridge length (ft) Limitations (Bridge Design Manual, 2007) Figure 7: Project Location Map Figure 8: Superstructure Framing Layout (MUS ) Figure 9: Top Bolted Field Splice Figure 10: Bottom Bolted Field Splice Figure 11: Abutment and Footing Section Figure 12: Observed Distress of West (Wall West Bound Bridge) Figure 13: Observed Distress of East Wall (East Bound Bridge) Figure 14: Propagated Cracks of West Wall (West Bound Bridge) Figure 15: Instrumentation Plan (East Bound Bridge) Figure 16: Instrumentation Plan (West bound Bridge) Figure 17: Top view of ID-C Indicator Figure 18: Side view of ID-C Indicator and Metal Targets Figure 19: Digi-tilt sensor and Readout device (Masada, 2007)... 41

11 11 Figure 20: Digital Strain Meter Figure 21: Target Location (East Bound Bridge) Figure 22: Target Location (West Bound Bridge) Figure 23: Tilt Reference point locations (East Bound Bridge) Figure 24: Tilt Ref. points Location (West Bound Bridge) Figure 25: Field Set-Up for Data Acquisition (Masada, 2007) Figure 26: Reference plate hanging on the wall Abutment (Masada, 2007) Figure 27: Deck Divided into Rhombus Shell Elements Figure 28: Girder Sections (Black) divided same size as shell Elements Figure 29: Backfill Modeled as Linear Springs Figure 30: Wingwall and Wall Abutment Model Figure 31: Lateral Earth Pressure (ksi) Distribution Figure 32: Pressure (psi) from Diaphragm due to Thermal Expansion Figure 33: Pressure (psi) from the Self-Weight of Bridge Deck Figure 34: Schematic Diagram of Wall and Backfill Figure 35: Change in East Joint Movement 1 versus Abutment Internal Temperature Figure 36: Change in East Joint Movement 2 versus Abutment Internal Temperature Figure 37: Change in East Joint Movement 1 versus Wingwall Internal Temperature Figure 38: Change in East Joint Movement 2 versus Wingwall Internal Temperature Figure 39: Change in East Joint Movement 1 versus Average Internal Temperature Figure 40: Change in East Joint Movement 2 versus Average Internal Temperature Figure 41: Measured East Joint Movement 1 versus Abutment Internal Temperature... 74

12 12 Figure 42: Measured East Joint Movement 2 versus Abutment Internal Temperature Figure 43: Measured East Joint Movement 1 versus Wingwall Internal Temperature Figure 44: Measured East Joint Movement 2 versus Wingwall Internal Temperature Figure 45: Measured East Joint Movement 1 versus Average Internal Temperature Figure 46: Measured East Joint Movement 2 versus Average Internal Temperature Figure 47: Effect of Span Length on Forces for Different skews (k soil = 35.7lb/in 3 ) Figure 48: Effect of Span Length on Forces for Different skews (k soil = lb/in 3 ) Figure 49: Effect of Span Length on Forces for Different skews (k soil = lb/in 3 ) Figure 50: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 15 o ) Figure 51: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 30 o ) Figure 52: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 45 o ) Figure 53: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 15 o ) Figure 54: Display of Results for Superstructure Analysis Figure 55: Close -up Display of x and y Forces Figure 56: Deflected Shape of the Wall Abutment and Wingwall Figure 57: Stress (ksi) Distribution on Front Side Figure 58: Stress (ksi) Distribution on Backside... 99

13 13 CHAPTER 1: INTRODUCTION 1.1- Scope The use of jointed type bridges has increased the overall maintenance cost of bridges due to expansion joint leakage. One of the major causes of this joint deterioration was observed when water carrying de-icing salts leaks through the expansion joints. With the introduction of the jointless bridges such as semiintegral and integral bridges however, the high joint maintenance cost is eliminated. In addition, the added simplicity in the construction of integral and semi-integral bridges has led them to become more popular in recent years (Bettinger, 2001). Oesterle and Lotfi (2005) also confirmed that in addition to reduced maintenance cost, jointless bridges improve riding quality, promote lower impact loads, reduce snowplow damage to decks and approach slabs, as well as improve the seismic resistance of the bridge Integral and Semi-Integral Bridges Integral bridges are those designed such that the superstructure (deck, girder and diaphragm) are rigidly connected to the substructure through bonded construction joints between the diaphragm and abutment, and at the abutment/foundation interface (Steinberg, Sargand, and Bettinger, 2004). In semi-integral bridges, the deck, girders, approach slab, and diaphragm act together as a single unit. However, flexible bearing surfaces such as elastomeric

14 14 pads are used in place of the bonded construction joints used for integral abutment bridges. A semi-integral bridge is illustrated in an elevation view (Figure 1) and side view (Figure 2). These flexible bearing surfaces (see Figure 1 and Figure 2) provide more flexibility at diaphragm/abutment interface. Thus the magnitude of the forces transferred to the foundation is theoretically decreased (Steinberg, Sargand, and Bettinger, 2004). The two bridges analyzed in this project were skewed semi-integral bridges and were located in the state of Ohio. In Ohio, the concrete encasement (diaphragm) in a semi-integral bridge is supported on elastomeric pads which in turn rest on the abutments (Steinberg, Sargand, and Bettinger, 2004). This is illustrated in Figure 1 and Figure 2. In order to prevent unwanted materials such as rocks, soil and trash from restraining the movement of the diaphragm/abutment joint and diaphragm/wingwall joint, polystyrene is used to fill the joint s gap. The structural behavior of these types of bridges is affected by both the temperature changes and loading conditions imposed on the bridges.

15 15 Bridge Deck Approach Slab Girder Backfill Diphragm Polystyrene Elastomeric Pad Fill Abutment Pile Cap Figure 1: Semi-Integral Abutment and Diaphragm (Elevation View)

16 16 Diaphragm Wingwall Abutment Polystyrene Elastomeric Pad Pile Cap Figure 2: Semi-Integral Abutment and Diaphragm (Side View) It is known that a bridge expands as it undergoes an increase in temperature. This increase in temperature causes the bridge diaphragms to be pushed outwards and into the soil, which causes the soil pressure to be increased. The impact of the combined effect of the expansion of bridge coupled with the backfill soil pressures is still uncertain (Metzger, 1995). Prior research has also shown that skewed semi-integral bridges tend to rotate as the ambient air temperature increases through the season. One analysis showed that the superstructure of a skewed semi-integral bridge will tend to rotate in the horizontal plane (Figure 3) unless otherwise retrained by guide bearings (Burke, 1994a). The magnitude of this rotation is greater for bridges with higher skews and occurs sooner for bridges with longer span length than those with shorter span

17 17 length (Burke, 1994a). As a result of the bridge trying to rotate, forces are generated and transferred to the wingwalls of the bridge. 1.2 Objectives The main focus of this research project was to utilize the results of the field analysis, as well as the computer analysis, in order to achieve the following: - Evaluate conditions of wingwalls and wall abutments - Asses the cause of the observed distress in the walls - Monitor the movement of the bridges due to temperature changes - Begin to develop guidelines to assist in achieving improved design of skewed semi-integral bridges - Improve the understanding of the soil-structure interaction between superstructure, substructure, and embankment soil due to changes in temperature 1.3 Outline Chapter 2 of this report gives background information about the conducted research project. It also provides a summary of the prior research done on similar topics.

18 18 Chapter 3 of this thesis discusses the field evaluation aspect of the research. It explains the instrumentation scheme of the project as well as the installation process employed to instrument the two bridges. This chapter also describes the instruments used to measure all the parameters required for the project including how the instruments were used for data acquisition and the reasons for choosing such instruments. The chapter discusses in detail the procedure and assumptions made in order to analyze the field data acquired. Chapter 4 of this report describes the analytical evaluation aspect of the research project. It describes the computer analyses carried out in the project as well as the computer models used in the analyses. This chapter discusses in detail the procedure and assumptions made in order to perform the computer analysis. Chapter 5 presents all of the results associated with the analyses carried out in chapter 3 and 4. Plots and tables were used in this chapter to draw meaningful conclusions and provide helpful recommendations in the proceeding chapter. Chapter 6, with the use of the analysis results in chapter 5, draws conclusions with regards to the aforementioned objectives of the project and also provides recommendations that could be used in the future by design engineers.

19 19 CHAPTER 2: BACKGROUND 2.1 Literature Review The amount of research conducted on the design and behavior of skewed semi-integral bridges due to thermal expansion is very limited. As such, information pertaining to the effect it has on wingwalls and foundations is difficult to find. Substantial amount of research work however, has been done on the pressures and forces acting on the abutment of skewed integral bridges due to thermal expansion (Bettinger, 2001). Steinberg, Sargand and Bettinger (2001) conducted research to determine the forces exerted in the wingwalls of skewed semi-integral bridges in Athens County and Tuscarawas County, Ohio. The research was conducted with the aim of gaining a better understanding of the effect that changing ambient temperature has to the distribution of forces on the wingwalls of skewed semi-integral bridges. The wingwall/abutment joints of two bridges in Ohio were instrumented to monitor the movement of the bridges as well as the forces generated in the wingwalls due to thermal expansion. The Tuscarawas County Bridge was a single span bridge with a span length 87ft and a roadway width of 32ft and a skew angle of 65 o. It was made up of composite reinforced concrete deck supported by steel girders. The backfill material for this bridge was sandy soil. The Athens County Bridge however, was a four span continuous steel girder semi-integral bridge, also with a composite reinforced concrete deck. The outer and inner spans were 70ft

20 20 and 87ft long respectively, and the roadway width was 40ft wide and the skew angle was 25 o. In addition to the instrumentation of the bridges, computer analysis was carried out in order to validate the data collected in the field. SAP 2000 was used to model the Athens County Bridge to determine the forces generated and exerted on the wingwalls. In addition, bridges were analyzed with multiple skew angles and span lengths in order to determine the impact a greater skew or longer span length has on the forces generated against the wingwalls. The research discovered that significant wingwall force up to a maximum of 35.7kip in magnitude was experienced by the Tuscarawas County Bridge during the course of the study. The maximum longitudinal movement recorded for the Athens County Bridge was in and a maximum movement into the wingwall of in was also measured. Based on the data recorded there was no direct correlation between the ambient temperature and the generated forces for the Tuscarawas County Bridge. Analytical results revealed that as the bridge skew angle increases, so does the reaction at the wingwall. Also for a larger skew angle, the increase in the wingwall force was higher as the backfill stiffness increased (Steinberg, Sargand, and Bettinger, 2001). In 1994, Martin P. Burke (b) documented his work on research he conducted to investigate the longitudinal, lateral, and rotational movement of semi-integral bridges. An insight on the behavior of skewed semi-integral bridges due to thermal expansion can be gained by making reference to Figure 3. In response to rising temperatures of the superstructures in semi-integral bridges, an

21 21 elongation (ΔL) is experienced by the bridge. Thus, a force P E is developed in the bridge as a result of this expansion. The backfill in turn reacts to produce resistive compressive passive soil pressures on both diaphragms of the bridge. The resultant of these compressive passive soil pressures is denoted by P P. The generated force P E of a skewed semi-integral bridge with skew angle θ is resisted by the longitudinal component (P P secθ) of the resisting compressive passive force (P P ) of the soil being compressed behind the diaphragms. The lateral component (P P tanθ) of the resisting passive force helps to resist the frictional backfill force and the wingwall reaction (P P tanδ). Figure 3: Forces on Bridge causing Horizontal Plane Rotation In skewed semi-integral bridges, the resultant forces of the passive soil pressures developed at both ends of the bridge due to longitudinal expansion of the bridge are not concurrent (Burke, 1994b). Burke (1994b) also noted that a moment couple that could potentially lead to the rotation of the bridge towards its acute

22 22 corners in the manner shown in Figure 3 is therefore produced due to the two forces. In order for the superstructure of a skewed semi-integral bridge to be stable, the force couple system that is causing rotation as described above must be resisted by an equivalent force couple system. This, as presented by Steinberg, Sargand and Bettinger in 2004 can be stated in equation 1 as: PpLsinθ = PptanδLCosθ + (additional force couple from wingwalls) (1) Where: PpLsinθ = Force couple developed from passive pressure. PptanδLCosθ = Frictional resistive force couple In the state of Ohio, wingwalls are currently designed to act only as retaining walls for the adjacent embankment soil (Bridge Design Manual 2000), and not to resist the forces exerted on them by the superstructure due to thermal expansion. In addition, wingwalls are recently constructed more in the turned back (aligned more closely parallel to the longitudinal axis of the bridge) position cantilevered from the superstructure (Figure 4) as opposed to constructing the wingwalls in the standard straight (Figure 5) position. Though this parallel alignment (turned-back position) of the wingwall provides an additional longitudinal restraint by way of backfill-wingwall friction or shearing resistance of backfill for wingwalls with rough surfaces due to the increase in confining pressure imposed by the turned back wingwalls (Burke, 1994b), it is important to note that the wingwall now becomes more subjected to bending stresses (Figure 4) than axial compression (Figure 5) because of the manner in which the forces are applied on the wingwalls.

23 23 Figure 4: Turned-Back wingwall (Parallel Alignment to Bridge Longitudinal Axis) Figure 5: Standard Straight Wingwall position In this document, the state of Ohio s concept for semi-integral bridges was also explained by utilizing a semi-integral abutment bridge that was developed and adopted. Burke (1994b) also outlined and discussed the distinctive characteristics of semi-integral bridges that should be noted and provided for by design engineers. With the principal goal of eliminating the bridge deck joints, designers in the state of Ohio adopted the design of continuous integral concrete bridges and

24 24 continuous integral steel bridges which began over 6 decades and 3 decades ago, respectively (Burke, 1994b). There were however, few exceptions in the sense that joints were provided at the ends and center of bridges with span length longer than 600ft. In addition, bridges with a skew angle greater than 30 o, those longer than 300ft, curved bridges, and those with wall or stub abutments were also provided with joints. Burke (1994b) also explained that due to the inadequate functional quality and durability of the provided deck joint sealing systems and the constant maintenance associated with it, the designers in Ohio were forced to innovate ways of combining the attributes of integral construction to those of bridges with movable joints. As a consequence, the semi-integral bridge design concept was adopted. As can be seen from Figure 6, prior to the development and adoption of semi-integral bridges, the application range of deck-jointed bridges was limited to 400ft span length with no skew or 200ft span length with a maximum skew angle of 30 o. With the introduction of semi-integral bridges however, the limits were expanded to the extents shown in Figure 6 (Burke and Gloyd, 1994). Several characteristics outlined by Burke (1994b) need to be recognized and provided for by design engineers. However, for the purpose of this thesis, only longitudinal, lateral and rotational restraint were discussed.

25 25 Figure 6: Semi-integral/ /Integral Abutment type Skew Vs Bridge length (ft) Limitations (Bridge Design Manual, 2007) As Burke (1994b) described, the bulk part of the longitudinall restraint for bridges, most especially those without fixed piers comes from the approach slab-subbase friction, shearing resistance of elastomeric bearings and the compressive resistance of the superstructure backfill. As aforementioned, Burke (1994b) explained that turned-back wingwalls also provide additional longitudinal restraint through backfill-wingwall rough surfaces. This additional resistance is a consequence of the friction or shearing resistance of backfill for wingwalls with increase in confining pressure imposed on the backfill by the turned back wingwalls. Thus, the higher the confining pressure on the backfill, the higher the backfill-wingwall friction and shearing resistancee of backfill. This in-turnn

26 26 provides the additional longitudinal restraint. It is important to recognize that only half of the frictional resistance of the bearings and the frictional resistance of the approach slabs are available to restrain any longitudinal forces during the cold weather. This is because the superstructure contracts and pulls away from both the abutments and backfill, thus the compressive resistance of the superstructure backfill does not come into play. For this reason, granular backfill at the abutments is more desirable. The lateral restraint of the bridge superstructure is provided by the superstructure-backfill and approach slab-backfill interaction. Also the compressive resistance of the filler in the lateral joints (diaphragm/wingwall) and the shearing resistance of the elastomeric bearings in the bridge seat provide additional lateral restraint to the superstructure (Burke, 1994b). Burke (1994b) added that for situations where substantial lateral resistance is required, such as skewed structures, structures exposed to stream flow pressure or earthquake forces, guide bearings are necessary. For superelevated bridges (bridge seat joint and elastomeric bearings are sloped), lateral guide bearings are also necessary (Burke, 1994b). Based on the analysis for the rotational restraint of the superstructure of skewed semi-integral bridges discussed earlier in this research, it can be surmised that the superstructures of skewed semi-integral bridges have a tendency to rotate in the horizontal plane unless otherwise restrained by guide bearings (Burke, 1994b). Burke (1994b) also pointed out that if the shear resistance of an idealized granular backfill and backfill-structure frictional resistance are used, it can be proven that, unless both

27 27 abutments of the bridge are otherwise provided with guide bearings, the superstructure of semi-integral bridges with a skew angle greater than 15 o will be unstable. Oesterle and Lotfi (2005) discussed and presented the experimental and analytical results of a research conducted to investigate the transverse movement in skewed integral abutment bridges. The experimental aspect of the study involved testing of bridge components and field monitoring bridge structures and models. In addition, a field survey of 15 jointless bridges was also conducted. The analytical aspect on the other hand investigated the effects of skew angle on the response of joint-less integral abutment bridges to longitudinal expansion. During the analysis phase, a bridge with a skew angle with respect to the bridge deck was considered. When a skewed integral bridge experiences an increase in length due to thermal expansion, the passive soil pressure that develops behind the abutments has a transverse component (Oesterle and Lotfi, 2005). Oesterle and Lotfi (2005) also added that the soil friction on the abutment is expected to resist the transverse component of the passive soil pressure within certain limits of the skew angle. Although the range of the maximum skew angle used by states is from 0 o to no limit, a commonly used maximum skew angle is 30 o (Oesterle and Lotfi, 2005). In situations where the soil friction is insufficient however, depending on the transverse stiffness of the abutment, either significant transverse forces or movements could be generated. The analyses showed that by doing simple statics, equation 2 has to be satisfied in order to attain rotational equilibrium.

28 28 F af (L cos θ) = P p (L sin θ) (2) F af = P p tan δ (3) Where: F af = Soil/abutment interface friction P p = Passive soil pressure θ = Skew angle δ = Angle of interface friction By substituting (3) into (2) it was proven that δ = θ in order to achieve rotational equilibrium. Thus, this means that as long as the skew angle of an integral bridge does not exceed the angle of interface friction, the superstructure of the bridge can be held in rotational equilibrium (Oesterle and Lotfi, 2005). Generally, the backfill of integral abutment bridges are granular soils. Based on the granular friction angle range of 22 o to 26 o provided by NCHRP Report 343, a reasonable conservative skew angle limit below which transverse forces or movement are not expected is 20 o (Oesterle and Lotfi, 2005). When this value is exceeded however, provisions must be made to accommodate the generated transverse force or movement. By incorporating the lateral resistance of the abutment, equation 2 is found to be;

29 29 (F a + F af) ) (L cos θ) = P p (L sin θ) (4) Thus by substituting (3) into (4), the force required to resist transverse movement is found to be; F a = P p (tan θ tan δ) (5) Note that F a is the sum of abutment lateral resistance and passive pressure on the substructure surface perpendicular to the abutment. The experimental results of this research were used to estimate the expected movement of the bridge in situations where provisions are not made to resist the movement. A ft three span steel-plate girder bridge located in Knox county Tennessee with a skew of o was instrumented. The span lengths were ft, 208ft and 68.08ft. There were a total of 12 girders spaced at 135in supporting a 9.25in deck. Eighteen HP 10 x 42 piles were used to support the 13ft tall east abutment wall. The wingwalls which were aligned parallel to the longitudinal axis of the bridge were supported on three piles. A measured end movement of 0.781in was utilized to calculate various important parameters. Based on this end movement, a bridge expansion of 2.26in was calculated. In addition, the center of rotation from the east end was determined to be 143.8ft, and the angle of rotation was calculated to be The analysis also determined that the transverse movement at the acute and obtuse corners ware 0.688in and 0.092in respectively. Sargand, Masada and Engle (1999) in their research titled Spread Footing Foundation for Highway Bridge Applications presented the results of their investigation on 30 bridge spread footings at 5 highway bridge construction sites

30 30 in the state of Ohio. The research was carried out in an attempt to validate the premise that shallow spread footings can be used in supporting highway bridge structures on both cohesive and cohesionless soils. Sargand, Masada and Engle (1999) pointed out that in addition to a lack of movement guidelines for spread footings, the belief in the notion that spread footings settle more and require higher maintenance, an uncertainty in the selection of performance prediction method together with the uncertainty in the subsoil properties used in settlement prediction caused design engineers to be reluctant to use spread footings. Thus, their research was also geared towards verifying the performance prediction methods of these shallow spread footing foundations. In order to accomplish one of the objectives of this research pertaining to evaluation of field performance of the spread footings, the bridge foundations were instrumented to measure certain key parameters. These parameters were; overall settlement, differential settlement, tilting of the abutment wall/pier column (which is of particular interest to this thesis), and contact pressure at the base of the footing. The settlements were monitored using the level survey method, the contact pressures at the base of the footings were measured using earth pressure cells, and the tilt angle measurements were measured using an accelerometer. The result of the study indeed validated the premise that spread footings could be used to successfully support highway bridge structures on both cohesive and cohesionless soils as long as favorable subsoil conditions exist (Sargand, Masada, and Engle, 1999). The range of total settlement in this study was 5 to 36mm (0.19 to 1.43in), the

31 31 differential settlement was in most cases less than 30% of the total settlement and the maximum recorded contact pressure at the base of the footings was usually less than 250kpa (36psi) unless there is a presence of bedrock or weathered rock. Finally, tilting of the abutment wall/pier column was within ±0.3 o. 2.2 Project Location and Bridge Background This research project involved the instrumentation of two bridge structures east and west bounds of State Route (S.R) 16, located in Muskingum County near Frazeysburg Ohio. The instrumented bridges were part of a project which involved the upgrading of a 2.36 mile stretch of S.R 16 (see Figure 7) from a two lane to four lane highway. The bridges pass over Raiders Road and have the bridge number MUS The project was completed and opened to traffic in Project Location Figure 7: Project Location Map

32 32 The east and west bound single span bridges had a total span length of 140ft and 139ft from center to center of bearing, respectively. The bridges were also 42ft wide from toe to toe of barriers. They were both semi-integral bridges with a skew angle of 45 o. Figure 8 below shows the superstructure framing layout of the west bound bridge. The girder sections of the bridges were grade 50 ASTM A572 welded steel plate girders each divided into three segments, i.e one interior and two exterior segments. The segments were rigidly connected with bolted field splices. A diagram of the top splice is shown in Figure 9, and the bottom splice is shown in Figure 10. Each segment had a total number of five girder sections spaced 10ft apart. The girder sections of the interior segments were 69.5ft in length and had a top flange plate size of 20 X1-3/8, a web plate size of 54 X5/8, and a bottom flange plate size of 22 X2-1/2. The exterior segments were 34.75ft in length and had plate sizes of 20 X1-3/8, 54 X5/8, and 22 X1-7/8 for the top flange, web and the bottom flange, respectively. Figure 8: Superstructure Framing Layout (MUS )

33 33 Figure 9: Top Bolted Field Splice Figure 10: Bottom Bolted Field Splice The deck was 9 in depth and was made of 4.5ksi reinforce concrete. The deck spans along the full length of the bridge before running into 25ft long approach slabs on both ends of the bridge. The deck was supported by the girders, which also rest on a wall abutment at each end of the bridge. As can be seen from

34 34 Figure 11, the 19.1ft highh wall abutments were supported on a 4ksi, 3ft deep and 14.5ft wide strip foundation footing. The wall abutments, which spanned the full width of the bridge, were tapered from bottom to top with the far face of the wall vertically straight and the near face slanted from the base with a width of 4ft to the top with a width of 3ft. Figure 11: Abutment and Footing Section

35 35 After the bridge was open to traffic, signs of distress were observed as cracks at the abutment/wingwall interface of the west wall of the west bound bridge as well as the east wall of the east bound bridge. Pictures of the observed distresses are shown for the west wall (Figure 12) and the east wall (Figure 13), respectively. Initially, the cracks were patched but only proved to resurface again. Over time, the cracks were observed to have propagated as shown in Figure 14. Figure 12: Observed Distress of West (Wall West Bound Bridge)

36 Figure 13: Observed Distress of East Wall (East Bound Bridge) 36

37 Figure 14: Propagated Cracks of West Wall (West Bound Bridge) 37

38 38 CHAPTER 3: FIELD EVALUATION In order to investigate the effect that the changes in ambient air temperature has on the forces generated in the wingwalls of the bridges, the wingwalls and abutments of two semi-integral bridges were instrumented. This was done by developing an instrumentation plan to retrieve data such as temperatures and bridge expansion that was used to estimate the forces exerted on the wingwalls by the diaphragm. In addition, the wall abutments of the bridges were instrumented to monitor the rotation of the abutment walls in order to evaluate the performance and condition of the strip foundation system used to support the bridge. A finite element model was also created to gain more insight on the behavior of the semi-integral bridges under thermal expansion as well as to compare the finite element results with the recorded field data. This thesis presents field results from the instrumentation of the two bridges to monitor the expansion of the bridge as well as the analytical results of the computer analysis. 3.1 Instrumentation Selection and Operation The wingwalls of both bridges were instrumented to determine the effect of thermal expansion/contraction of the bridges on the movement at the wingwalls. The wall abutments were also instrumented to monitor their tilt which was suspected to be due to either differential settlement and/or thermal effects. A plan view of the instrumentation can be seen in Figure 15 and Figure 16.

39 39 Figure 15: Instrumentation Plan (East Bound Bridge) Figure 16: Instrumentation Plan (West bound Bridge) In order to carry out the objectives of the research project, the following parameters and instruments were used to measure and monitor the behavior of the bridges. The instruments used in this project were based on the monitoring demands of the project as well as prior experience with similar projects.

40 40 Measured Parameters: (i) Thermal induced expansion and contraction of the expansion joint at wingwall/diaphragm interface (ii) (iii) Tilt angle measurements of abutments and wingwalls Internal temperature of instrumented wingwalls and abutments The instruments used to measure the aforementioned parameters were a Mitutoyo ID-C indicator (see Figure 17 and Figure 18) for thermal induced expansion and contraction of joints, a digi-tilt tilt meter (see Figure 19) for tilt angle measurements of abutment and wingwalls, and a TC-21k digital strain meter (see Figure 20) to read thermocouples for internal temperature measurements of the wingwalls and abutments. The above mentioned instruments have been used by Ohio University in previous projects and were found to be reliable and also produce useful results. Figure 17: Top view of ID-C Indicator

41 41 Figure 18: Side view of ID-C Indicator and Metal Targets Figure 19: Digi-tilt sensor and Readout device (Masada, 2007)

42 42 Figure 20: Digital Strain Meter 3.2 Measurement of Expansion and Contraction of Bridge The expansion and contraction measurements of the expansion joint at the wingwall/wall abutment interface were taken for both bridges. The instrumentation layouts of the targets for the expansion and contraction measurements of both the west and east bound bridges are illustrated in Figure 21 and Figure 22 below.

43 Figure 21: Target Location (East Bound Bridge) 43

44 44 Figure 22: Target Location (West Bound Bridge) Digimatic indicator metal targets (see Figure 18) were installed on the walls of the bridges on each side of the expansion joint. To install the metal targets, the desired distance between the two targets was measured using a tape measure, and holes were drilled at those locations. The concrete particles from the drilling were cleaned out to ensure a strong bond between the metal targets and the concrete. The targets were then embedded into the walls and held in place with the use of epoxy. For the east bound bridge, two sets of targets were installed 8in and 10in apart, while on the west bound bridge only one set of targets installed 8in apart, was used. The installation of the west bound bridge metal targets was done several months after the initial installment of instrumentation due to difficulties in gaining access to the height

45 45 of the desired location. In addition to that, the topography of the ground surface adjacent to the location at which the installation was to take place was very steep and considered dangerous for the installation crew. As mentioned earlier, the expansion and contraction measurements were taken with the ID-C indicator (see Figure 17 and Figure 18). In order to take a measurement, calibration of the micrometer on the ID-C indicator to the appropriate target spacing was required, i.e 8in or 10in. A decrease in the pre-set gauge length indicated contraction of the bridge while an increase in the gauge length indicated an expansion of the bridge. The resolution and accuracy of the instrument are in and in, respectively (Bettinger, 2001). 3.3 Wingwall and Wall Abutment Tilt Angle Measurements The measuring of the angle of tilt for the wingwalls and abutments was attained by establishing measurement stations on the bridges. A total of six stations were set up with three stations on each bridge. Figure 23 and Figure 24 show the locations at which the stations were established on the walls of each bridge. The reference points circled in black with the arrow pointing to the left were located further to the left in the picture. For each bridge, one station was established on the wingwall and the other two were established on the wall abutments.

46 46 Figure 23: Tilt Reference point locations (East Bound Bridge) Figure 24: Tilt Ref. points Location (West Bound Bridge)

47 47 In order to establish a reference station, a rebar locator was used to determine the location of the rebars in the walls. This was done in order to assure availability of depth into the walls up to 2in, as well as to prevent drilling into the rebars in order to avoid decreasing the structural integrity of the wingwalls or wall abutments. A vertical distance of 2.5ft was measured between the two points at which holes were drilled. The concrete particles were cleared after drilling to establish proper bond between the concrete and the stainless steel reference points. Using epoxy, two stainless steel reference points were embedded 2in into the wingwall and wall abutments. As described by Masada (2007), in order to take a tilt reading at a given measuring station, a stainless steel ball on a threaded shaft was screwed into each of the established reference points. A reference plate was then positioned and held against the steel joints, and with the use of an accelerometer, readings were taken. The field set up for the data acquisition is illustrated in Figure 25. In addition, Figure 19 shows the components of the data acquisition equipment (readout device and accelerometer) while Figure 26 shows the reference plate hanging on the wall.

48 Figure 25: Field Set-Up for Data Acquisition (Masada, 2007) 48

49 49 Figure 26: Reference plate hanging on the wall Abutment (Masada, 2007) The Digi-Tilt tiltmeter (see Figure 19) manufactured by Slope Indicator (Seattle, Washington) was used in the tilt measurements. The system comprises of a sensor (accelerometer) and a digital readout device. According to Masada (2007), the tilt-meter sensor has a range from -30 o to +30 o and a sensitivity of o. One single measurement using the device consisted of two different readings, i.e a positive and negative reading. In order to calculate the angle of tilt (θ) of the walls from the true vertical, the positive and negative readings obtained in the field were applied to the following equation as presented by Masada (2007).

50 50 ( + Reading) ( Re ) 1 ading θ ( rad. ) = sin 4 (6) A positive θ value indicated that the wall tilted away from the backfill behind it, whereas a negative θ value indicated that the wall tilted into the backfill. It should be noted that the movement is based on an initial reading taken at the completion of the instrumentation since the bridge was constructed long before the reference points were installed. 3.4 Temperature Measurements In order to take the internal temperatures of the wingwalls and wall abutments, Omega type T (Copper Copper-Nickel) thermocouples were embedded into the walls. These thermocouples have a maximum temperature range of -270 to 400 o C (-454 to 752 o F) and a tolerance of 1.0 o C (Omega Engineering, 2006). A total of two thermocouples were installed each on the west and east bound bridges. One thermocouple was embedded on each wingwall and wall abutment and were both held in place using epoxy. A Digital Strain meter (TC-21k model 232) was used to record the temperature readings. A picture of this device can be seen in Figure 20. Different locations were used to determine if thermal variations existed between the wall abutments and wingwall.

51 51 CHAPTER 4: ANALYTICAL EVALUATION In performing the analytical aspect of this research, a computer software program was used to carry out a finite element (FE) analysis. SAP 2000, a structural Engineering software that is capable of performing both FE analysis and design was utilized to analyze the response of the bridge deck and the wingwall. Due to the fact the bridges have already shown signs of distress and have also experienced movement before instrumentation, the initial conditions were unknown and thus could not be fully accounted for in the analysis. For this reason, it is very difficult to make a direct comparison between the computer analysis and the field analysis. Since several assumptions were made to account for the initial conditions of the bridges, the computer analysis is thus considered parametric and was performed for the purpose of gaining more insight on the magnitude of the forces that could potentially be generated and transferred to the wingwalls, and to better understand the distribution of the stresses on the wall panels. 4.1 System Analysis Finite element analyses were performed on bridge systems using multiple bridge spans with different skew angles and varying backfill soil stiffness values. Span lengths of 139ft, 200ft, 400ft and 600ft each with skew angles 15 o, 30 o, 45 o and 60 o were all modeled in the computer software with the different backfill soil stiffness values. Though according to the 2007 Ohio Bridge Design Manual, the maximum span length allowable

52 52 for a semi-integral bridge with a combination of skew angle from 0 o -50 o is 400ft (see Figure 6), a 600ft span was analyzed in order to investigate this specified limitation. Likewise, the maximumm allowable skew angle as presented in the 2007 Ohio Bridge Design Manual is 50 o (see Figure 6). The justification for this limitation needed to be investigated as well; thus the 60 o skew angle was analyzed. The 139ft long bridge deck modeled was that of the bridge monitored in the field. It should be noted that the length of 139ft is the span length from center to center of the bearing support, and this is true of all the lengths of the models analyzed. Several assumptions were made in order to model the deck of the bridge into the FE software. The results of this analysis did not take into account the shearing resistance of the bearing pads as well as the frictional resistance of the backfill on the superstructure. The entiree length of the bridge was modeledd for analysis by dividing the deck into small rhombus shaped thin-shell elements due to the skew (seee Figure 27). The deck shells were created using a 4.5ksi normal modulus of elasticity (E) for the concrete was weight (150pcf) concrete. The value of 3.82 X 10 3 ksi as determine using the compressive strength of the concrete and equation 7 below. E = f ' c (7) Figure 27: Deck Dividedd into Rhombus Shell Elements

53 53 As explained by MacGregor and Wight (2005), the range of values for poison s ratio (ν) of concrete is usually from 0.11 to 0.21 and the coefficient of thermal expansion (α) of normal weight concrete is from 5 to 7 X 10-6 strain/ o F. MacGregor and Wight (2005) also added that an all around value of 5.5 X 10-6 strain/ o F may be used. Thus, default values of 0.2 and 5.5 X 10-6 strain/ 0 F were used in the analyses for poison s ratio and coefficient of thermal expansion, respectively. The girder sections were grade 50 ASTM A572 welded steel plate girders, which were created in the program based on the dimensions of the plates given in the drawings. This was done because the sizes of the girders used in the project were not available in the software library. According to Metal Suppliers Online, a mean value of thermal coefficient is 6.7 X 10-6 strain/ o F and an elastic modulus of 30 X 10 6 psi were used for the material. In addition, a yield strength value of 50 ksi was used since the material is grade 50. All girder sections were divided into small elements of the same size as the rhombus thin-shell elements (See Figure 28). Figure 28: Girder Sections (Black) divided same size as shell Elements

54 54 The backfill soil was modeledd as linear springs (see Figure 29) to resist motion in the X and Y directions. This was done to mimic the restraining effect the backfill soil had on the abutment walls. Though the stiffness value for the granular backfill material was not presented in the literature, stiffness values for sands were assumed (Bowles, 2005). Stiffness values of 35.37lb/in 3, lb/in 3 and lb/in 3 were assumed based on similar research conducted by Bettinger (2001).These soil stiffness were then converted into spring stiffness/support (kip/in) in the X and Y directions as can be seen in Table 1. It should be noted that the behavior of soil is much more complex than the assumptions made in order to create the models for the analysis. However, for the purposes of this research, the assumptions made in creating the models were adequate to assist in comparing the analyses results with the field data. Figure 29: Backfill Modeled as Linear Springs

55 55 The calculations carried out in converting these stiffness values are as follows: Contact Area (A C ) between the wall abutment and the embankment WA H A 42 ft 6.33 ft AC = = (8) cos( θ ) cos( θ ) Where: θ = skew angle (degrees) W A = Width of Abutment H A = Height of Abutment Total Spring Stiffness (TSS) SS A TSS = C (9) cos(θ ) Where: TSS = Total Spring stiffness SS = Soil Stiffness A total of nine springs were evenly spaced along the width of abutments for all the models with the exception of 139ft and 200ft with a skew angle of 60 o. These two models had a total of 17 springs evenly spaced along the width of the abutment. Thus, the spring stiffness per support (SpS/sup) = TSS/9 or TSS/17 where appropriate. These values were then broken down into their respective X and Y components using: X comp = ( SpS / sup) cos( θ ) (10) Y comp = ( SpS / sup) sin( θ ) (11)

56 56 Table 1: Soil Stiffness Conversion to Spring Stiffness The 2in thick preformed expansion joint filler (PEJF) used between the wingwall/abutment interface was also modeled as a linear spring at the acute corners of the bridge model. To determine the stiffness of that spring, a compression test was performed by Ohio University. The results of this test are presented in Table 2. The test required that the PEJF material be able to support 250 psi of compressive stress at 50% compression of the material. The results of Table 2 were determined using abutment/wingwall interface dimensions of the bridge. The required wingwall area for the east and west bound bridges were in 2 and in 2 respectively. The PEJF test sample used was 4in X 4in in dimension and thus had an area of 16in 2. As can be seen from Table 2, the magnitude of force required to compress the PEJF by 1in respectively for the east and west bound bridges were average values of

57 57 709kips and 851kips for the two separate tests performed. The stress associated with these forces as determined in the Lab was 259psi. Since this stress exceed the maximum required stress of 250psi in the PEJF, the forces at each wingwall required to compress the PEJF by 1in were back calculated using a compressive stress of 250psi at the abutment/wingwall interface. Spring stiffness values at the acute corners of the east and west bound bridges were thus determined to be 685 kip/in and 820 kip/in respectively. These values of 685 kip/in and 820 kip/in were used instead of 709kip/in and 851kip/in because they represent the maximum stress value of 250psi in the PEJF. In addition 685kip/in and 820kip/in are conservative values since based on the test, the material can withstand 259psi of stress, but the value was reduced to 250psi. Table 2: PEJF Test Result

58 58 The supports at the approach ends of the decks were modeled as rollers to allow for thermal expansion and contraction. According section of the Ohio Bridge Design Manual, a base construction temperature of 60 o F shall be assumed for design purposes, thus a 60 o F temperature change was imparted on the deck and girders of the bridge models. 4.2 Analysis of Wingwall and Wall Abutment An FE model of the wingwall and wall abutment (See Figure 30) was made by utilizing SAP. The analysis was carried out to investigate the magnitude and stress distribution pattern at the abutment/wingwall interface and on the surface of the wingwall and wall abutment. The Wingwall and wall abutment model was broken into 1,728 solid elements and the base support of the walls was fixed. The solid elements were created using a 4.5ksi normal weight (150pcf) concrete as given in the project drawings. As aforementioned, the value of modulus of elasticity (E) used was 3.82 X 10 3 ksi.

59 59 Wingwall Wall Abutment Figure 30: Wingwall and Wall Abutment Model In this model, only 9ft of both the wingwall and wall abutment was used. This was done to reduce the model size and the associated computing time of the analysis. A poison s ratio (ν) and coefficient of thermal expansion (α) values of 0.2 and 5.5 X 10-6 strain/ o F, respectively (MacGregor and Wight, 2005) were used. The loads (See Figure 31 - Figure 33 ) considered during this analyses include; the self weight of both the wingwall and wall abutment, the weight of the deck and girders (dead load) supported by each abutment, the lateral earth pressure from the

60 60 backfill acting on both the wingwall and wall abutment which were calculated using the equations below and the 250psi pressure load from the diaphragm acting on the wingwall at the abutment/wingwall interface. Figure 31: Lateral Earth Pressure (ksi) Distribution

61 Figure 32: Pressure (psi) from Diaphragm due to Thermal Expansion 61

62 62 Figure 33: Pressure (psi) from the Self-Weight of Bridge Deck The following equations (Budhu, 2008), were used to calculate the lateral earth pressure of the backfill against the wingwall and wall abutment. The Rankine s theory of active earth pressure (soil pushing against wall) was used. This is because the passive case (wall pushing against soil) is an extreme situation and unlikely to be the mode in this case. Passive case is possible when the deck expands at higher temperatures. Figure 34 shows a schematic diagram of the wall and embankment soil.

63 First, equation 12 was used 63 to determine the active earth pressure coefficient. Using a value of β = 0, η= 2.9 o and φ = 30 o a value assumed within the 28 o -33 o typical range of φ for granular soils (Budhu, 2008), K ar was determined to be This value was found to be close to the typical K ar value of (Bowles, 1996) determined for the same parameters. Figure 34: Schematic Diagram of Wall and Backfill K ar = cos( β η) cos 2 η( cos β 2 1+ sin φ' 2sinφ' cosϕ a sin φ' sin β ) (12) Where: K ar = Rankine s Active Earth Pressure Coefficient = Internal Friction Angle

64 P soil = K ar γ h 64 sin β ϕ sin 1 a = β + 2η (13) sinφ' The pressure was then determined using equation 14 (14 and a γ value of 120 pcf. This γ value was an assumed value within the 106pcf 125pcf typical range of γ for sands (Bowles, 2005) P soil = K γ h (14) ar Where: P soil = Soil pressure at a given height γ = Unit weight of soil = 106 to 125pcf (average 120 pcf) according to Bowles (2005) h = height The soil pressure acting on the face of each solid element was determined by first calculating the pressure at the top and bottom of each solid element, and then taking the average between the two pressures. The difference between the pressure at the top and bottom of each solid element was small and decreased as the depth of the wall increased from the top to bottom. Thus the elements were considered small enough to assume that the pressure distribution within a solid element was constant throughout the element. The analysis of the walls for stress distribution proceeded the completion of the application of the loads.

65 65 CHAPTER 5: RESULTS 5.1 Field Results The measured parameters, which include - (i) rotation of wall abutments; (ii) wingwall/diaphragm joint gap; and (iii) internal temperatures of the wingwall and wall abutments, were tabulated. Plots of the joint gaps versus the internal temperatures of the wingwall, wall abutments as well as the average between the wingwall and wall abutment temperatures were also created in order to determine if a correlation that existed between the aforementioned parameters. The movement of the east and west bound bridges was monitored by measuring the gap of the expansion joint in conjunction with measuring the internal temperatures of the wingwalls and wall abutments. The measured bridge movement data for the east bound bridge is presented in Table 3 and Table 4, and that of the west bound bridge is presented in Table 5 and Table 6. This data includes the internal temperatures of the south (east bound) and north (west bound) sides of the wall abutment, the wingwall, and an average of the two internal temperatures. In addition, the lengths of the target gap 1(targets spaced approximately 10in apart) and 2 (targets spaced approximately 8in apart) are presented in the tables as well. Table 4 and Table 6 present the difference in the subsequent readings taken relative to the first reading taken after the completion of the installation of the targets. The difference in reading was calculated as the subsequent readings

66 66 minus the initial reading. Thus negative values of T represent decrease in temperatures relative to the first reading, and negative values of the gap movements depict a closing of the joint gaps and an expansion of the bridge superstructure. In addition, positive values of T represent an increase in temperatures relative to the first reading, and positive values of the gap movements depict an opening of the joint gaps and a contraction of the bridge. The plots of the aforementioned data are shown in Figure 35 - Figure 46. These plots show the relationship between the internal temperatures and bridge movement. The tilting of the east and west bound wall abutments as well as wingwalls were also measured. The data recorded pertaining to this response is presented in Table 7 -Table 10. It is important to note the temperature taken on 7/11/2008 from NOAA was assumed to be the temperature of the south end of the east bound wall abutment. The remainder of the temperature was determined by adding or subtracting the difference of the initially recorded temperatures accordingly. Table 3: Measured East Bound Internal Temperatures and Joint Gap

67 67 Table 4: Change in East Bound Internal Temperatures and Gap Displacements Table 5: Measured West Bound Internal Temperatures and Joint Gap Table 6: Change in West Bound Internal Temperatures and Gap Displacements * Air temperatures used from NOAA instead of incorrect internal temperatures

68 68 Figure 35: Change in East Joint Movement 1 versus Abutment Internal Temperaturee

69 69 Figure 36: Change in East Joint Movement 2 versus Abutment Internal Temperaturee

70 70 Figure 37: Change in East Joint Movement 1 versus Wingwall Internal Temperaturee

71 71 Figure 38: Change in East Joint Movement 2 versus Wingwall Internal Temperaturee

72 72 Figure 39: Change in East Joint Movement 1 versus Average Internal Temperature

73 73 Figure 40: Change in East Joint Movement 2 versus Average Internal Temperature

74 74 Figure 41: Measured East Joint Movement 1 versus Abutment Internal Temperaturee

75 75 Figure 42: Measured East Joint Movement 2 versus Abutment Internal Temperaturee

76 76 Figure 43: Measured East Joint Movement 1 versus Wingwall Internal Temperaturee

77 77 Figure 44: Measured East Joint Movement 2 versus Wingwall Internal Temperaturee

78 78 Figure 45: Measured East Joint Movement 1 versus Average Internal Temperature

79 79 Figure 46: Measured East Joint Movement 2 versus Average Internal Temperature

80 80 Table 7: Measured East Abutment and Wingwall Tilt Table 8: Measured West Abutment and Wingwall Tilt Table 9: Change in East Abutment and Wingwall Tilt

81 81 Table 10: Change in West Abutment and Wingwall Tilt 5.2 Bridge Movement The distance measured between the targets installed on the east and west bound bridges together with the associated temperatures are given in Table 3 - Table 6. The opening and closing of the wingwall/diaphragm joint in response to the aforementioned internal temperatures can be better understood by making reference to Table 4 and Table 6. As mentioned earlier, the negative values indicate opening of the joint gap relative to the initial reading. In addition, a positive value indicated closing of the joint relative the first readings shown in Table 3 and Table 5. By making reference to Table 3, it can be seen that the largest east bound joint gap closing measured during the course of the research occurred at joint gap 2 and was in. This corresponds to the largest temperature increase recorded, and the bridge was also expected to expand the most during that period. The largest east bound joint gap opening measured occurred at joint gap 1 and was in. This value corresponds to the largest temperature decrease recorded during the period of data acquisition. During this

82 82 period, the bridge was expected to contract the most. Though the installation of targets on the west bound bridge came later during the research, some data was collected with regards to opening and closing of the joint gap. As can be seen in Table 6, the west bound joint gap opening measured was in. By comparing Table 3 with Table 5, and then Table 4 with Table 6, it is apparent that the wingwall of the west bound bridge experienced more changes in internal temperatures than that of the east bound bridge. In addition, the north side of the wall abutment for the west bound bridge also experienced more changes in the internal temperatures than the south side of the wall abutment for the east bound bridge. This is partly because the bridge superstructure provided more shade to the reference points of the east bound bridge.thus it would be expected that the joint gap opening and closing of the west bound bridge would be greater than that of the east bound bridge. This is evident by making reference to the last recorded values in Table 4 and Table 6. This could be as a result of the position of the sun at the time of data acquisition since the bridges were instrumented on opposite sides. The plots of the data in Table 3 - Table 6 are presented in Figure 35 - Figure 46. Figure 35 - Figure 40 represent the plots of the recorded joint gap displacements versus the recorded internal temperatures, and Figure 41 - Figure 46 represent the plots of change in the recorded joint gap displacements versus the change in the internal temperatures relative to the first reading after completion of instrumentation installation. It is evident from the plots of Figure 35 - Figure 40

83 83 that as the temperature increased, the joint gap decreases. This is the expected response of the joint to temperature increase. The plots also show a linear regression curve fit to each of the data points. The R 2 values and the linear regression equation are displayed on each plot. The R 2 value ranged from to for Figure 35 - Figure 40 and to for Figure 41 - Figure 46. This proves that there is a good and consistent data fit by the linear regression lines. Also by making reference to Figure 41 - Figure 46, it can be surmised that as the change in temperature increased in the positive X- axis (warmer temperatures), the change in gap displacements increased in the negative direction with the negative sign signifying expansion of the bridge and thus closing of the gap. Also as the change in temperature increased in the negative X-axis (cooler temperatures), the change in gap displacements increased in the positive direction with the positive sign signifying contraction of the bridge and thus opening of the gap. 5.3 Abutment and Wingwall Tilt The initial tilt-meter measurements were taken on October 18 th This was the date the tilting reference stainless steel studs were installed. Several other measurements were subsequently taken and the data collected is summarized in Table 7 - Table 10 for both east and west bound bridges. As stated earlier, the tilt angle measured using the tiltmeter is the angle with respected to the vertical.

84 84 In Table 7 - Table 10 the positive measurements indicate that the wall is moving away from the backfill, while negative measurements indicate that the wall is moving towards the backfill. According to the data presented in these tables, the tilt angles measured ranged from o to o (Abutment) and o to o (Wingwall) for the east bound bridge. In addition, the tilt range for the west bound bridge are o to o (Abutment) and o to o (Wingwall). From a structural standpoint, it was expected that when the bridge undergoes thermal expansion, the abutments would move towards the backfill while the wingwall moves away from the backfill. This behavior was observed for the west bound bridge as shown by the recorded field data in Table 10. However, by making reference to Table 9, it can be seen that though the abutment moved into the backfill, it did not follow that the wingwall move away from the backfill. In fact, the data showed that the wingwall moved in the same direction (towards/away) from the backfill as the abutment in all cases. Thus it can be surmised that the movement of the wingwalls and abutments towards and away from the backfill is more a very complex one, and possibly has to do with the complex soil behavior as well. Further research is required in order to fully understand the reasons why the wingwalls and abutments moved in the direction they did. 5.4 Analysis Results Two separate finite element analyses were performed in SAP The first analysis involved a model of the bridge superstructure in order to investigate the magnitude of the wingwall forces and stresses generated at the wingwall/diaphragm joint due to thermal expansion. The second analysis was a

85 85 model of a section of the wingwall and wall abutment to investigate the distribution of the stresses on the walls. The results of the first set of analyses carried out on the superstructure are present in Table 11 -Table 16. In addition, plots were generated as shown in Figure 47 - Figure 49 to investigate the effect of varying the length of superstructure length on the wingwall forces for different skews. Also, Figure 50 - Figure 53 were used to investigate the effect of varying the length of the superstructure length on the wingwall for different soil stiffness. It should be noted that only the forces for the west wingwall were plotted because they were larger in magnitude due to a bigger contact area between the diaphragm and wingwall. In addition, the plots would be the same with the exception of the magnitude of the forces.

86 Table 11: Forces in East Wingwall 86

87 Table 12: Stresses in East Wingwall 87

88 Table 13: Forces in West Wingwall 88

89 89 Table 14: Stresses in West Wingwall Table 15: Analysis East Wall Joint Movement (45 o skew, and k soil = lb/in 3 )

90 90 Table 16: Analysis West Wall Joint Movement (45 o skew, and k soil = lb/in 3 ) Plot of West Wingwall Force Vs Span Length (k soil = 35.37lb/in 3 ) Wingwall Force (kip) deg 30 deg 45 deg 60 deg Span Length (ft) Figure 47: Effect of Span Length on Forces for Different skews (k soil = 35.7lb/in 3 3 )

91 91 Plot of West Wingwall Force Vs Span Length (k soil = lb/in 3 ) Wingwall Force (kip) Span Length (ft) 15 deg 30 deg 45 deg 60 deg Figure 48: Effect of Span Length on Forces for Different skews (k soil = lb/in 3 ) 1000 Plot of West Wingwall Force Vs Span Length (k soil = lb/in 3 ) Wingwall Force (kip) Span Length (ft) 15 deg 30 deg 45 deg 60 deg Figure 49: Effect of Span Length on Forces for Different skews (k soil = lb/in 3 )

92 92 Wingwall Force (kip) Plot of West Wingwall Force Vs Span Length (Skew = 15 o ) Span Length (ft) k = (lb/in 3 ) Figure 50: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 15 o ) Wingwall Force (kip) Plot of West Wingwall Force Vs Span Length (Skew = 30 o ) Span Length (ft) k = (lb/in 3 ) Figure 51: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 30 o )

93 93 Plot of West Wingwall Force Vs Span Length (Skew = 45 o ) Wingwall Force (kip) Span Length (ft) k = (lb/in 3 ) Figure 52: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 45 o ) Wingwall Force (kip) Plot of West Wingwall Force Vs Span Length (Skew = 60 o ) Span Length (ft) k = (lb/in 3 ) Figure 53: Effect of Span Length on Forces for Different Soil Stiffness (Skew = 15 o )

94 Forces and Stress in Wingwalls at Wingwall/Diaphragm Joint The finite element analyses of the bridge superstructure were performed after the completion of the model. The magnitudes of the generated forces were displayed in the form of x and y spring reactions as shown in Figure 54. Figure 55 shows a close-up display of x and y wingwall forces. Figure 54: Display of Results for Superstructure Analysis Figure 55: Close -up Display of x and y Forces To determine the magnitude of the force and stresses generated and transferred to the wingwall due to thermal expansion, the resultant of each set of x

95 95 and y spring forces and the associated stresses were calculated and summarized in Table 11 Table 14. It is important to remember that as mentioned earlier, these wingwall forces were generated without taking into account the shearing resistance of bearing pads as well as the frictional resistance of the backfill on the superstructure. This could be reason for the high magnitude of wingwall forces generated from the analyses. According to the results of these analyses, and by making reference to Table 11 and Table 13as well as Figure 47 - Figure 53, it can be seen that at a lower soil stiffness (35.37 lb/in 3 ), the forces in the west wingwall increased as the skew angle increased. At higher soil stiffness ( lb/in 3 and lb/in 3 ), the west wingwall forces initially increased; however, the rate of increase, decreased as the skew angle increased. This is evident in the plots by the decrease in slope noticed as the skew increased. Figure 47- Figure 49 reveal that for a given soil stiffness value, the magnitude of the wingwall force increased as the span lengths increased. The plots also revealed that for a given soil stiffness, higher skew angles yielded larger wingwall forces at longer span length. This implies that in order to minimize the magnitude of the generated wingwall forces for a given soil stiffness, lower skew angles are more desirable over higher skew angles, especially for longer span length. By making reference to Figure 50 - Figure 53, it can be observed that the magnitude of the wingwall force increased as the span length increased. In addition, the plots also showed that for a given skew angle, higher soil stiffness

96 96 values yielded larger wingwall forces. This implies that in order to minimize the magnitude of the generated wingwall forces for a given skew angle, soil with a lower stiffness value is more desirable at the backfill. 5.6 Stress Distribution on Wall Abutment and Wingwall A finite element analysis was also performed toward understanding the distribution of the stresses on the wall panels. Once the necessary assumptions and restraints were imposed, and the application of the aforementioned loads was completed, the modeled structure was then analyzed. Figure 56 shows the deflected shape of the modeled structure at the completion of the analysis.

97 97 Original Position Wingwall Wall Abutment Figure 56: Deflected Shape of the Wall Abutment and Wingwall By making reference to the deflected of the structure, it can be seen that the front side of the wingwall is in compression while the back side is in tension. This is due to the direction of the applied lateral earth pressure acting on the back side as well as the 250psi lateral load from the diaphragm. The 250psi pressure load is the maximum transferable pressure through the PEJF. In addition, it can be seen that the top of the wall abutment went down a little due to the applied deck self-weight and also, just like the wingwall, the front side is in compression and

98 98 backside is in tension due to lateral earth pressure of the backfill acting on the backside. Figure 57 and Figure 58 show how the stresses are distributed on both the front side and the backside of the wall panels. Stress Concentration Figure 57: Stress (ksi) Distribution on Front Side

99 99 Figure 58: Stress (ksi) Distribution on Backside By making reference to Figure 57 and Figure 58, the portions of walls experiencing the highest stress concentration are revealed. Of particular interest to this research was the stress concentration in Figure 57. The portion that experienced the highest stress concentration in Figure 57 is the location of the observed distresses (see Figure 12 -Figure 14). This high stress distribution could be a result of the multiple directions in which the structure deflected (see Figure