National Academies Press: OpenBook

Guidelines for Dowel Alignment in Concrete Pavements (2009)

Chapter: Chapter 3 - Findings and Applications

« Previous: Chapter 2 - Research Methodology
Page 15
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 15
Page 16
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 16
Page 17
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 17
Page 18
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 18
Page 19
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 19
Page 20
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 20
Page 21
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 21
Page 22
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 22
Page 23
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 23
Page 24
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 24
Page 25
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 25
Page 26
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 26
Page 27
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 27
Page 28
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 28
Page 29
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 29
Page 30
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 30
Page 31
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 31
Page 32
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 32
Page 33
Suggested Citation:"Chapter 3 - Findings and Applications." National Academies of Sciences, Engineering, and Medicine. 2009. Guidelines for Dowel Alignment in Concrete Pavements. Washington, DC: The National Academies Press. doi: 10.17226/14249.
×
Page 33

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

15 This chapter describes the results of field studies, laboratory testing, and analytical modeling conducted in this project. These results were used to develop the design and construction guidelines provided as Attachment A to this report. 3.1 Field Testing MIT Scan-2 data from 60 projects located in 17 states were used to evaluate typical values of dowel alignment and position. These data included measurements of over 2,300 one- or two-lane joints and more than 28,000 dowel bars. Data that appeared to have been affected by nearby metallic objects, such as tie bars or traffic loops, were not included in the analyses. The relationship between dowel misalignment and pave- ment performance also was evaluated. Dowel misalignment was characterized by its type (e.g., embedment length, vertical translation or concrete cover, and rotational tilt) and magni- tude. Pavement performance was characterized by observed distresses such as transverse joint faulting and transverse crack- ing. The performance of some joints also was evaluated using FWD testing. The results of these analyses are presented below. 3.1.1 Typical Misalignment The misalignment levels measured in this study give insight into the level of alignment that can be achieved using current construction practices. The achievable level for which there is no observable effect on performance is used in the development of the guidelines. All rotational misalignments (i.e., horizon- tal skew and vertical tilt) are expressed as a deviation from alignment over 18 in. [457 mm], which is the length of a typical dowel. 3.1.1.1 Vertical Translation Dowel bars are assumed to be embedded at the mid-depth of a slab. A vertical deviation from this position is consid- ered vertical translation. Negative vertical translation indi- cates that the dowel bar is closer to the surface and positive vertical translation indicates that the dowel bar is closer to the base. Average measured vertical translation for individual projects ranged from −1.1 in. to +0.9 in. [−28 mm to 23 mm]; the dis- tribution of these averages is presented in Figure 3.1, which shows that, although 63% of the projects are within the typical DOT-specified vertical translation limits of ± 0.5 in. [±13 mm], it is possible for entire projects to have an average vertical translation level of more than 0.5 in [±13 mm]. Over 95% of the projects have an average vertical translation level within ± 1.0 in. [± 25 mm] of the slab mid-depth. This can be a result of using dowel baskets of incorrect height, the use of an improperly set DBI and/or concrete mix-related issues, or the placement of pavement that is thicker (or thinner) than spec- ified. The vertical translation distribution of the dowel bars within each nominal thickness is shown in Table 3.1. When absolute values are considered, the average vertical depth deviation for all projects is 0.46 in. [12 mm] and the standard deviation is 0.6 in. [15 mm]. As mentioned previously, some of the variability in vertical translation or concrete cover is due to differences between designed and as-built slab thicknesses, and the computation is based on the nominal design thickness. For example, if the design thickness is 10 in. [254 mm], then the dowel bar is expected to be located at a depth of 5 in. [127 mm]. However, if the as-built thickness is 10.24 in. [260 mm] and the dowel bar is located at mid-depth (5.12 in. [130 mm] from the pave- ment surface), then a deviation of +0.12 in. [3 mm] would be assumed. The vertical translation manifests itself most notably in reduction of the concrete cover either from the top or the bottom surface. Table 3.2 shows the average concrete cover, dowel depths, and corresponding standard deviations for various projects with nominal concrete thickness ranging from 8 to 12 in. [203 to 305 mm]. C H A P T E R 3 Findings and Applications

3.1.1.2 Longitudinal Translation The longitudinal center of a dowel bar is designed to be at the location of the transverse joint saw cut. Therefore, any deviation in longitudinal dowel placement with respect to the joint axis is considered longitudinal translation (i.e., reduced dowel embedment on one side of the joint). Key factors influ- encing longitudinal translation are joint marking and saw cut operations. The average longitudinal translation for all projects was 0.86 in. [22 mm], and the standard deviations within individ- ual projects ranged from 0.4 in. to 1.9 in. [9 mm to 49 mm]. This suggests that the dowel bars were placed in their accu- rate longitudinal positions in some projects and in varying longitudinal positions in other projects. The average standard deviation for all projects was 0.9 in. [23 mm]. The maximum average longitudinal translation was 1.9 in. [49 mm], resulting in a lowest average embedment length of 7.1 in. [180 mm] for the 18-in. [457-mm] long dowels. The standard deviation of longitudinal translation for all of the individual dowels was 1.2 in. [30 mm]; Figure 3.2 shows the longitudinal translation distribution. Over 91 and 98% of all bars are within ±2 in. [50 mm] and ±3 in. [75 mm] from the transverse joint, respectively. 3.1.1.3 Horizontal Skew The MIT Scan-2 unit determines the positions of the two ends of a dowel bar and computes the horizontal skew as a deviation from the longitudinal axis over the length of the dowel. The horizontal skew can be positive or negative depend- ing on the horizontal angle of the dowel bar relative to the longitudinal joint. The absolute values were used for mean analysis and the actual values were used for standard deviation analysis. Average absolute horizontal skew for individual projects ranged from 0.13 in. to 0.41 in. [3.3 mm to 10.4 mm], and the average absolute horizontal skew for all projects was 0.23 in. [5.9 mm]. The horizontal skew standard deviations for indi- vidual projects ranged from 0.1 in. to 0.34 in. [2.6 mm to 8.7 mm], and the average standard deviation was 0.19 in. [4.7 mm]. The average horizontal skew for all bars from all projects was 0.24 in. [6.1 mm] with a standard deviation of 0.21 in. [5.3 mm]. Less than 80% of all bars were within 3⁄8 in. [9 mm]. Figure 3.3 shows the horizontal skew distribution for all bars from all projects. Almost 90, 98, and 99.5% of the dowel bars have horizontal skew less than 0.50 in. [13 mm], 0.75 in. [19 mm], and 1.0 in. [25 mm], respectively. 3.1.1.4 Vertical Tilt The MIT Scan-2 unit pinpoints the vertical positions of the two ends of a dowel bar and determines the vertical tilt as the vertical deviation from the longitudinal axis with respect to 16 0% 5% 10% 15% 20% 25% 30% 35% < -1in. -1 to -0.5 in. -0.5 to 0.0 in. 0.0 to +0.5 in. +0.5 to 1.0 in. > 1.0 in. Pe rc en t o f S ec tio n s Vertical Depth Deviation, in. Figure 3.1. Measured vertical translation distribution. 8 9 10 11 12 # of Projects 3 7 11 26 5 # of Dowel Bars 1036 4847 7321 9529 2388 Dowel Diameter, in. 3 X 1.25 4 X 1.25; 3 X 1.5 3 X 1.25; 8 X 1.5 2 X 1.25; 24 X 1.5 5 X 1.5 Construction 3 Basket 2 Basket; 5 DBI 1 Basket; 10 DBI 20 Basket; 6 DBI 3 Basket; 2 DBI Average Depth, in. 3.76 4.56 5.13 5.47 5.94 Standard Deviation 0.38 0.57 0.49 0.69 0.46 < -1 in. 0.0% 0.0% 9.1% 3.8% 0.0% -1.0 to -0.5 in. 25.0% 10.0% 0.0% 23.1% 0.0% -0.5 to 0 in. 50.0% 40.0% 18.2% 23.1% 40.0% 0 to 0.5 in. 25.0% 30.0% 54.5% 30.8% 40.0% 0.5 to 1.0 in. 0.0% 20.0% 18.2% 19.2% 20.0% >1 in. 0.0% 0.0% 0.0% 0.0% 0.0% Concrete Thickness (in.) Dowel Bar Statistics Dowel Bar Deviation Distribution Table 3.1. Vertical dowel translation for sections with different thicknesses.

the length of the dowel. Although the vertical tilt can be pos- itive or negative depending on the vertical angle of the dowel bar relative to the surface of the slab, the absolute values were used for mean analysis, and the actual values were used for standard deviation analysis. Average absolute vertical tilt for individual projects ranged from 0.11 in. to 0.51 in. [2.9 mm to 13.1 mm], and the average vertical tilt for all projects was 0.24 in. [6.1 mm]. The stan- dard deviation for individual projects ranged from 0.1 in. to 0.53 in. [2.7 mm to 13.5 mm], and the average standard devi- ation for all projects was 0.19 in. [4.9 mm]. The average vertical tilt for all bars from all projects is 0.23 in. [6 mm] with a standard deviation of 0.21 in. [5.4 mm]. These values are nearly identical to the horizontal skew values. Approximately 80% of all bars were within 3⁄8 in. [9 mm]. Figure 3.4 shows the vertical skew distribution for all bars from all projects. About 91, 98, and 99% of dowel bars had vertical tilt less than 0.50 in. [13 mm], 0.75 in. [19 mm], 1.0 in. [25 mm], respectively. 3.1.2 Effect on Pavement Performance Distress data were collected for 37 pavement sections, many of which had almost no distresses. Some projects exhibited minor shallow surface spalling (less than 0.5 in. [13 mm] deep) that apparently did not result from dowel misalignment but 17 8 9 10 11 12 Projects 3 7 11 26 5 Dowel Bars 1036 4847 7321 9529 2388 Dowel Diameter, in. 3 X 1.25 4 X 1.25; 3 X 1.5 3 X 1.25; 8 X 1.5 2 X 1.25; 24 X 1.5 5 X 1.5 Construction 3 Basket 2 Basket; 5 DBI 1 Basket; 10 DBI 20 Basket; 6 DBI 3 Basket; 2 DBI Average Depth, in. 3.76 4.56 5.13 5.47 5.94 Standard Deviation 0.38 0.57 0.49 0.69 0.46 < 2.0 in. 0.0% 0.0% 0.0% 0.0% 0.0% 2.0 to 2.5 in. 0.1% 0.1% 0.0% 0.0% 0.0% 2.5 to 3.0 in. 1.0% 2.8% 0.0% 0.0% 0.0% 3.0 to 3.5 in. 26.4% 3.3% 0.8% 0.1% 0.0% 3.5 to 4.0 in. 42.7% 5.4% 2.9% 1.4% 0.0% 4.0 to 4.5 in. 27.2% 29.1% 3.9% 8.1% 0.4% 4.5 to 5.0 in. 2.7% 38.3% 27.7% 16.3% 1.2% 5.0 to 5.5 in. 0.0% 19.0% 46.1% 23.8% 13.1% 5.5 to 6.0 in. 0.0% 1.6% 16.1% 28.5% 44.7% 6.0 to 6.5 in. 0.0% 0.2% 2.2% 15.2% 27.9% 6.5 to 7.0 in. 0.0% 0.1% 0.2% 5.8% 11.4% 7.0 to 7.5 in. 0.0% 0.0% 0.1% 0.8% 1.3% 7.5 to 8.0 in. 0.0% 0.0% 0.0% 0.0% 0.1% > 8.0 in. 0.0% 0.0% 0.0% 0.0% 0.0% Concrete Thickness (in.) Dowel Bar Depth Distribution Dowel Bar Statistics Table 3.2. Measured depth of dowel bars from slab surface for various concrete thicknesses. 0% 5% 10% 15% 20% 25% 0.0 to 0.25 in. 0.25 to 0.5 in. 0.5 to 0.75 in. 0.75 to 1.0 in. 1.0 to 1.25 in. 1.25 to 1.5 in. 1.5 to 1.75 in. 1.75 to 2.0 in. 2.0 to 2.25 in. 2.25 to 2.5 in. 2.5 to 2.75 in. 2.75 to 3.0 in. 3.0 to 6.5 in. Pe rc en t o f B ar s Longitudinal Translation, in. Figure 3.2. Distribution of longitudinal translation. 0% 10% 20% 30% 40% 50% 60% 70% < 0.25 in. 0.25 to 0.50 in. 0.50 to 0.75 in. 0.75 to 1.00 in. 1.00 to 1.25 in. 1.25 to 1.50 in. 1.50 to 3.50 in. Pe rc en t o f B ar s Horizontal Skew, in. Figure 3.3. Distribution of horizontal skew.

was more likely due to saw cut timing. To evaluate the effect of dowel misalignment, the performance of sections with high and low levels of misalignments were compared. 3.1.2.1 Comparison of Section Performance Dowel alignment of the 37 projects was classified in Groups A, B, and C with regards to the four misalignment categories: (1) vertical depth deviation, (2) longitudinal trans- lation, (3) horizontal skew, and (4) vertical tilt. For analysis purposes of each misalignment category, projects with place- ment accuracy in the bottom third were placed in group C, projects with placement accuracy in the top third were placed in Group A, and the rest were placed in Group B. Only two sections were included in Group C in all four categories, eight sections were included in Group C in three of the four categories, only one section was included in Group A in all four categories, and seven sections were included in Group A under three of the four categories. Tables 3.3 and 3.4 give the percentage of slabs with differ- ent distresses for the different projects of Groups A and C, respectively. No clear trend was observed with respect to any of the distresses between the Group A and Group C sections. In fact, the only project that was included in Group A under all four categories (1-OH3) had the highest amount of trans- verse cracking (48% of the slabs). Because of the differences between the projects in factors such as design, traffic, age, climate, and materials, project-level analyses were conducted to examine the development of dis- tresses in the different slabs of a specific project containing joints with varying degrees of dowel misalignment. 3.1.2.2 Project-Level Analysis Dowel alignment levels are not uniform within each project, so the effects of dowel misalignment on the distribution of distresses within the sections were analyzed. Two types of project-level analyses were conducted. In one analysis, joints or slabs with high levels of distresses and joints or slabs with no significant distresses were identified, and the dowel mis- alignments for the two groups of joints were then compared. The second analysis involved sorting the joints with respect to misalignment level from highest accuracy to lowest accuracy and comparing the distresses of those joints or adjacent slabs. Examples of these analyses for transverse cracking and joint faulting are presented here. Appendix B provides more detail on the analysis for these distresses and joint opening. Transverse Cracking and Joint Spalling. Of the 37 sec- tions surveyed, 26 projects did not exhibit any transverse cracking, and 33 projects did not exhibit any high-severity spalling at the joints. Only six projects (1-AZ3, 1-AZ9, 1-CA3, 1-IL2, 1-OH3, and 1-WI2) had a considerable amount of trans- verse cracking or high-severity joint spalling (greater than 10% of slabs or joints); they could be considered for project-level analysis. All other projects could not be used for project-level analysis because they did not have significant amounts of such distresses. 1-OH3 was not used for project-level analysis because it had an unusually high percentage of slabs with transverse cracking (48%), such that no difference in joint and slab performance could be attributed to the varying levels of mis- alignment (nearly all joints had a cracked adjacent slab). 1-CA3 was excluded from the analysis because it included dowel retrofitted joints, and most of the cracking probably occurred 18 0% 10% 20% 30% 40% 50% 60% 70% < 0.25 in. 0.25 to 0.50 in. 0.50 to 0.75 in. 0.75 to 1.00 in. 1.00 to 1.25 in. 1.25 to 1.50 in. 1.50 to 4.00 in. Pe rc en t o f B ar s Vertical Tilt, in. Figure 3.4. Distribution of vertical tilt. Project 1 - AZ 8 1 - AZ 6 1 - WI 1 1 - MN 2 1 - MN 1A 1 - OH 1 1 - AZ 4 1 - WI 3 1 - IL 2 1 - MN 1B Slabs Cracked (Transverse) 0 0 6 0 0 0 3 7 14 0 Spalling (Major) 0 0 0 0 0 0 0 15 3 0 Corner Breaks 0 0 0 0 0 0 0 0 0 0 Slabs Cracked (Longitudinal) 0 3 0 6 0 0 3 0 3 0 Joint FDR 0 0 12 0 0 0 0 0 0 0 Midpanel FDR 0 0 0 0 0 0 0 0 0 0 Slabs with distress, percent Table 3.3. Slabs with distresses in Group A sections.

before the joints were retrofitted. Therefore, the analysis was performed on the other four projects (1-AZ3, 1-AZ9, 1-IL2, and 1-WI2). Thirty percent of the slabs in project 1-AZ3 exhibited trans- verse cracking, and none of the joints had any major spalling. A statistical analysis was conducted to compare the dowel alignment of joints adjacent to slabs that exhibited transverse cracks with that of joints adjacent to slabs that did not exhibit any transverse cracking. The test section had 33 joints, 16 of which were adjacent to slabs with transverse cracking (Group A) and 17 of which were adjacent to slabs without any transverse cracking (Group B). Student’s t-tests were conducted to deter- mine whether there were any statistically significant differences between the two sets of joints with regard to the average absolute values of vertical and longitudinal translation, vertical skew, and horizontal tilt at the individual joints. There was no statistical difference in average vertical trans- lation, average longitudinal translation, or average vertical tilt between joints that are adjacent to slabs exhibiting transverse cracking and joints adjacent to intact slabs. However, there was a statistically significant difference between the two groups with respect to horizontal skew. Contrary to expectations, the joints adjacent to the intact slabs had higher levels of average horizontal skew than the joints adjacent to cracked slabs. This suggests that factors other than misalignment contributed to cracking. Moreover, the actual levels of misalignment of both groups are less than 0.32 in. [8 mm], which is well within typical specification tolerances, and should not cause joint lockup. The analyses for test sections 1-AZ9, 1-IL2, and 1-WI2, presented in Appendix B, also showed that there is no statisti- cally significant difference between the alignments of dowels in joints adjacent to cracked slabs and alignments of dowels adjacent to uncracked slabs for these sections. Therefore, the results of the project-level analyses suggest that, within the nonextreme limits of dowel translations (vertical and horizontal) and rotations (vertical tilt and hor- izontal skew) measured in this study, there appear to be no differences in the amounts of transverse cracking and joint spalling as a result of dowel misalignment. Faulting and LTE Analyses. Many of the evaluated proj- ects had only small levels of faulting ranging from 0 to 0.1 in. [0 to 3 mm] at most of the joints. The low faulting could be attributed to the use of relatively large dowel bars (1.25- and 1.5-in. [35- and 38-mm] diameter) and the young age of the pavement sections. For the faulting analyses that follow, only older pavements (> 10 years) that exhibited some significant amount of faulting (mean faulting > 1 mm) were considered. Faulting measurements were taken in the wheel path and at the slab edge, and the maximum of the two values was used for analysis. Vertical Translation. Analysis was conducted to compare faulting and LTE at joints with dowels that were centered within ±0.25 in. [±6 mm] (on average) of mid-depth with those that had dowels centered more than 1.0 in. [25 mm] closer (on average) to the pavement surface. The average vertical transla- tion at each joint in each project was computed with respect to the mid-depth of the pavement. For a given project, the average faulting of all joints with the smaller level of vertical translation (Group 1) was paired with the average faulting of all joints with the higher level of vertical translation (Group 2). Ten projects were considered in this analysis. The same analysis procedure was used for joint LTE; five projects were considered. The relatively high P-values suggest that there are no statistically significant differences in faulting or LTE between the two groups (i.e., joints with average vertical translation < ±0.25 in. [±6 mm] and joints with average vertical trans- lation > 1.0 in. [25 mm] closer to the slab surface). Note that the faulting levels considered in this study were extremely low, and the vertical translations greater than 1.0 in. [25 mm] closer to the surface were observed on thick slabs with sufficient cover (i.e., the dowels were still 4 to 5 in. [102 to 127 mm] from the surface of the slab). Longitudinal Translation. Analysis was conducted to compare faulting and LTE at joints with dowels that were centered within ±0.5 in. [13 mm] (on average) of the transverse joints with those that had dowels that were centered greater than 2.0 in. [51 mm] (on average) from the transverse joints. Although longitudinal translations over 3 in. [76 mm] were 19 Project 1 - NC 1 1 - NC 4 1 - NC 3 1 - CA 31 - IN 2 1 - OH 31 - OH 4 1 - WI 2 0 0 0 0 24 48 Spalling(Major) 0 3 0 40 Corner Breaks 0 0 0 6 0 Slabs Cracked (Longitudinal) 0 0 0 0 0 Joint FDR 0 0 0 0 0 Midpanel FDR 0 0 0 0 6 Slabs with distress, percent Slabs Cracked (Transverse) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Table 3.4. Slabs with distresses in Group C sections.

observed on some newly constructed pavements, these proj- ects were not included in the performance analysis because of their age. For a given project, the average faulting of all joints in Group 1 (joints with dowels that were centered within ±0.5 in. [±13 mm], on average, of the transverse joints) was paired with the average faulting of all joints in Group 2 (joints that had dowels that were centered greater than 2.0 in. [51 mm], on average, from the transverse joints). The same analysis was conducted for joint LTE. Fourteen projects were considered in this analysis, but only four projects provided sufficient data points for this paired t-test. The high P-values suggest that there is no statistically sig- nificant difference in faulting or LTE between the two groups (i.e., joints with average longitudinal translation < ±0.5 in. [±13 mm] of the transverse joint and average longitudinal translation > ±2.0 in. [±51 mm] of the transverse joint). Note that the faulting levels measured in this study were extremely low, and none of the joints considered in this study had sig- nificant levels of average longitudinal translation (> 3 in. [76 mm]); the average minimum embedment length was 7 in. [178 mm] or more. Therefore, the effects of higher longitudinal translation on faulting and LTE cannot be determined on the basis of this data set. However, other studies (Burnham, 1999) showed that embedment lengths of 2.5 in. [64 mm] or less resulted in higher levels of faulting at these joints. Vertical Tilt. Analysis was conducted to compare faulting and LTE at joints with dowels that had vertical tilts less than ±0.25 in. [±6 mm] (on average) with those that had dowels with vertical tilts greater than ±0.75 in. [±19 mm] (on average). The average vertical tilt at each joint at each project was computed. For a given project, the average faulting of all joints in Group 1 (joints with dowels that had vertical tilts less than ±0.25 in. [±6 mm]) was paired with the average faulting of all joints in Group 2 (joints that had dowels with vertical tilts greater than ±0.75 in. [±19 mm]). The same analysis was conducted for joint LTE. Fourteen projects were considered in the analysis, but only four projects provided sufficient data points for this paired t-test. The P-value of 0.024 calculated for faulting suggests that there is a statistically significant difference in faulting between the two groups (joints with average vertical tilt < ±0.25 in. [±6 mm] and joints with average vertical tilt > ±0.75 in. [±19 mm]). The joints with higher average vertical tilts had higher levels of average faulting. Note that the faulting levels are extremely low, and only a small number of joints at each section had average tilt > ±0.75 in. [±19 mm]. However, there was no statistically significant difference in LTE between the two groups as indicated by the relatively high P-value of 0.474. Horizontal Skew. Analysis was conducted to compare faulting and LTE at joints with dowels that had horizontal skews of less than ±0.25 in. per 18 in. [± 6 mm per 457 mm] (on average) with those that had dowels with horizontal skews greater than ±0.75 in. per 18 in. [±19 mm per 457 mm] (on average). The average horizontal skew at each joint at each project was computed. For a given project, the average faulting of all joints in Group 1 (joints with dowels that had horizontal skews of less than ±0.25 in. per 18 in. [± 6 mm per 457 mm]) was paired with the average faulting of all joints in Group 2 (those that had dowels with horizontal skews greater than ±0.75 in. per 18 in. [± 19 mm per 457 mm]). The same analysis was con- ducted for joint LTE. The same procedure was followed for joint LTE, where the average LTE of all joints in Group 1 was paired with the average LTE of all joints in Group 2 for each project. Fourteen projects were considered in this analysis, but only four projects provided sufficient data points for this paired t-test. The P-value of 0.45 calculated for faulting suggests that there is no statistically significant difference in faulting between the two groups (joints with average horizontal skew < ±0.25 in. [± 6 mm] and joints with average horizontal skew > ±0.75 in. [± 19 mm]). The P-value of 0.11 calculated for LTE, however, suggests that there is moderate statistical significance in the differences in LTE between groups of joints with these differ- ent levels of horizontal skew. The joints with higher average horizontal skews had slightly lower joint LTE. It should be noted that (1) the faulting levels are extremely low, (2) only a small number of joints at each section had average skew > ±0.75 in. [±19 mm], and (3) a small number of sections provided data for LTE comparisons. 3.1.3 Summary of Field Study Analyses Review of the field data from 60 projects indicated the following ranges for dowel misalignments in the majority of joints: • Vertical translation: ± 0.5 in. [± 13 mm] for pavement that is 12-in. [305-mm] thick or less; • Horizontal skew: ± 0.5 in. per 18 in. [± 13 mm per 457 mm]; • Vertical tilt: ± 0.5 in. per 18 in. [± 13 mm per 457 mm]; and • Longitudinal translation: ± 2 in. for 18-inch dowels [± 51 mm per 457 mm]. These ranges of misalignment represent tolerances that are easily achieved in the field. Furthermore, dowel misalignment within these ranges on slightly higher levels does not appear to affect pavement performance significantly. 3.2 Laboratory Testing This section summarizes the results of dowel pullout and shear tests conducted to evaluate the effects of dowel mis- alignment on joint lockup and dowel efficiency. 20

3.2.1 Modified Pullout Testing 3.2.1.1 Results Overview It was observed that greasing or not greasing the dowels greatly influences pullout force as shown in Figure 3.5 for dowels embedded 6 in. in the same beam. Ungreased dowel requires a significantly higher force to cause pullout failure. Embedment length also had a significant effect on pullout force. Figure 3.6 shows the pullout force versus relative dowel and displacement for two aligned dowels (to illustrate the variability in pullout force) and for a dowel with 3 in. [76 mm] of embedment. The figure shows that the dowel with lower embedment length required a lower pullout force than either of the aligned dowels and illustrates the large variability in pullout force. Inspection of the interface between the dowel and concrete surface after each pullout test indicated slight surface paste chipping for dowels embedded with 1 in. [25 mm] of tilt and spalling damage for dowels embedded with 2 in. [51 mm] of rotation. Figure 3.7 presents the distribution of the maximum forces required to pull out dowels embedded with different types and levels of misalignment in 4 groups. One group includes prop- erly aligned, ungreased dowels with 6 in. [152 mm] of embed- ment. Another group includes properly aligned dowels, dowels with 2 in. [51 mm] of rotation, and dowels with 4 in. [102 mm] of rotation, all greased with 9 in. [229 mm] embedment. The third group is similar to the second group, except the dowels had 6 in. [152 mm] of embedment. The fourth group includes unrotated greased dowels with 2 and 4 in. [51 and 102 mm] embedment length and dowels with 3 in. [76 mm] of embed- ment and 2 in. [51 mm] of rotation. 3.2.1.2 Trends To test for statistically significant differences in pullout forces between the various groups of dowels, Student t-tests were conducted (details of all these tests are provided in Appendix C). The analysis confirmed that greased dowels with 6 in. [152 mm] of embedment require significantly lower pullout forces than similarly embedded ungreased dowels, and even greased dowels with 9 in. [229 mm] of embedment require a lower mean pullout force than that of ungreased dowels with 6 in. [152 mm] of embedment. It can be observed in Figure 3.7 that rotational misalignment up to 2 in. per 18 in. [51 mm per 457 mm] dowel length did not have a significant effect on pullout force, but rotational 21 0 2,000 4,000 6,000 8,000 10,000 12,000 0 0.05 0.1 0.15 0.2 0.25 0.3 Pu llo u t F o rc e, lb s. Relative Dowel Displacement, in. 6 in. embedment with no grease 6 in. embedment with grease Figure 3.5. The effect of greasing dowels on pullout force versus displacement. 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 0.05 0.1 0.15 0.2 0.25 Pu llo ut F o rc e, in . Relative Dowel Displacement, in. Aligned Dowel with 3 in. embedment Figure 3.6. Effect of embedment length on pullout force versus displacement. Figure 3.7. Distribution of maximum pullout forces for greased and ungreased dowels with varying degrees of misalignment.

misalignments of 4 in. per 18 in. [102 mm per 457 mm] dowel length had a significant effect. The analysis further illustrated that there is no statistically significant difference between the means of pullout forces for the aligned and 2 in. [51 mm] rotated dowels, while the 4 in. [102 mm] rotated dowels required significantly larger pullout forces. This suggests that dowels that are not properly greased or dowels that experi- ence extreme rotation would increase longitudinal restraint at the joints. Also, because of the reduced dowel-concrete contact area, a lower pullout force is required for a reduced embedment length. For example, the 9-in. [229-mm] embedded dowels require a significantly larger pullout force than the 3-in. [76-mm] embedded dowels. It has been reported that some distresses may develop pre- maturely due to joint lockup caused by dowel misalignment (Tayabji, 1986). However, analysis of the pullout data obtained in this study has shown that moderate misalignment of indi- vidual dowels did not have a significant effect on the maximum required pullout force, and greasing the dowels prior to embed- ment reduced the required pullout force significantly. 3.2.2 Shear-Pull Testing Although dowels transfer load through shear and moment mechanisms, numerous studies have shown that the shear mechanism dominates and the moment transfer mechanism can be neglected (Guo et al., 1996). The MEPDG structural analysis model assumes that dowels transfer the load in shear only. The shear-pull test was used to evaluate the ability of dowels with various misalignments to resist a shear load after being subjected to the pullout test, and it simulates the ability of a dowel to transfer a wheel load (in shear) after the joint has been opened due to slab contractions caused by temperature change or shrinkage. Shear performance measures (such as shear stiffness and shear capacity) were used to evaluate the effectiveness of each dowel-concrete system in resisting applied shear loads. Shear capacity is defined as the load at which the concrete around the dowel experiences shear failure. Shear stiffness is defined as the relationship between changes in shear force in relation to changes in relative dowel displacement. An example of shear force versus relative dowel displacement for dowels with 2, 3, 4, and 9 in. [51, 76, 102, and 229 mm] of embedment is shown in Figure 3.8, which illustrates how the ultimate shear can be affected by dowel misalignment. The figure shows that a 9-in. [229-mm] embedded dowel has a higher ultimate shear force than any of the dowels with a lower embedment length. Also, there is no loss of shear stiff- ness in the dowel with 4 in. [102 mm] of embedment until the system fails at a load of about 7 kips [31 kN]. For the 2 and 3 in. [51 and 76 mm] embedment cases, the dowel not only has a lower ultimate shear capacity (about 5 kips [22 kN]) but there is also a loss in shear stiffness from almost the beginning of load application. 3.2.2.1 Trends Shear stiffness and ultimate shear capacity can be used to compare the effects of different types and levels of misalign- ments. For example, Figure 3.9 shows that vertical tilt of up to 2 in. [51 mm] did not have a significant effect on shear stiffness or ultimate shear capacity while 4 in. [102 mm] of vertical tilt greatly reduced the shear stiffness and ultimate shear capacity. This suggests that the shear capacity decreases as vertical tilt increases above 2 in. [51 mm]. Because the shear test was performed after the pullout test, the extreme loss in stiffness experienced by the 4-in. [102-mm] vertically-tilted dowel was probably caused in part by the 22 0 2000 4000 6000 8000 10000 12000 0 0.02 0.04 0.06 0.08 0.1 0.12 Sh ea r Fo rc e, lb s. Relative Displacement, in. 9 in. embedment 4 in. embedment 3 in. embedment 2 in. embedment Figure 3.8. Illustration of reduced ultimate shear capacity and loss of stiffness with increased dowel misalignment. 0 2000 4000 6000 8000 10000 12000 0 0.02 0.04 0.06 0.08 Sh ea r Fo rc e, lb s. Relative Dowel Displacement, in. Aligned 1 in. vertical tilt 2 in. vertical tilt 4 in. vertical tilt Figure 3.9. Vertical tilt shear pull shear capacity.

damage at the dowel-concrete interface that occurred during the pullout test. Figure 3.10 shows the relative dowel displacement versus shear force for dowels with various embedment lengths. This figure shows that reducing the embedment length from 9 in. [229 mm] to 6 in. [152 mm] had little effect on dowel shear behavior, reducing the embedment length further to 4 in. [102 mm] had little effect on shear stiffness but resulted in a large reduction (26%) in shear capacity. Further reducing the embedment to 3 in. [76 mm] produced an additional 12% reduction in shear capacity (38% total reduction with respect to the 9-in. [229 mm] embedment condition) and a 63% reduction in shear stiffness. The reduction in embedment to 2 in. [51 mm] resulted in shear capacity and shear stiffness of 56% and 30%, respectively, of the values for 9 in. [229 mm] embedment. The average shear capacity values for aligned dowels and dowels with concrete cover reduced from 3.25 to 1.25 in. [83 to 32 mm] due to vertical translation of 2 in. [51 mm] were 9.3 and 4.3 kips (42.2 and 19.1 kN, respectively). Thus, the ultimate shear capacity was reduced by more than 50% due to this reduction in concrete cover. When comparing the shear force versus relative dowel displacements for an aligned dowel to that for a dowel with reduced embedment length, a dowel with reduced concrete cover, and a dowel with both reduced embedment length and reduced concrete cover, a compounding effect of misalign- ments is observed. The decrease in concrete cover results in a large decrease in ultimate shear capacity with little loss of shear stiffness while the reduction in embedment results in modest losses of both shear capacity and stiffness. However, the combination of both misalignments results in large reductions in both ultimate shear capacity and shear stiffness. The repeated shear load testing also revealed that the shear capacity of a dowel subjected to repeated loading was sig- nificantly lower than that for a dowel subjected to a single displacement-controlled loading. A comparison of shear force versus displacement for a 2 in. [51 mm] vertically translated dowel (i.e., with concrete cover of 1.25 in. [32 mm]) to that of an aligned dowel (i.e., with concrete cover of 3.25 in. [83 mm]) shows that (1) the shear stiffness of both dowels was reduced after 14,000 load cycles and (2) the stiffness of the 2 in. [51 mm] vertically translated dowel exhibited a secondary decrease as the load approached 3 kips [13 kN] (as shown by the rapid decrease in slope of shear load versus displacement curve above 2.5 kips [11 kN]). This observation suggests that failure of the dowel with reduced concrete cover and subjected to repeated loading started at loads approaching 3 kips [13 kN], which is significantly lower than the failure due to single load applica- tion of 4.7 kips [21 kN]. Single load tests showed that reduced dowel diameter causes lower shear stiffness. Similar shear performance trends were noted for aligned dowels and dowels with reduced concrete covers. As expected, lower shear capacity and shear stiffness were measured for the 1.25-in. [32-mm] diameter dowels than those for 1.5-in. [38-mm] diameter dowels with the same alignment. The effects of dowel misalignment on performance observed in the laboratory study can be summarized as follows: • Presence of greasing significantly affects pullout force. • Dowel rotation as extreme as 2 in. per 18 in. [51 mm per 457 mm] does not affect dowel shear capacity. • Reduction of dowel embedment length to 3 in. [76 mm] or less significantly affects shear capacity. • Reduction in concrete cover from 3.25 to 1.25 in. [83 to 32 mm] causes large reduction in ultimate shear force. • The combined effect of low concrete cover and low embed- ment length is greater than the effect of either one of the two misalignments. 3.3 Analytical Modeling The ABAQUS beam-dowel model presented in Chapter 2 was used to perform computer simulations to augment the results of the laboratory study and to further investigate the effects of dowel misalignment on joint behavior. 3.3.1 Finite Element Beam Model A validated beam model will allow consideration of dowel misalignment cases other than those tested in the laboratory. The shear force required to cause 0.05 in. [1.3 mm] of relative displacement was defined as the dowel shear capacity because the shear force required to cause this displacement according to the analytical model was similar to the shear capacity level 23 0 2000 4000 6000 8000 10000 12000 0 0.02 0.04 0.06 0.08 0.1 0.12 Sh ea r Fo rc e, lb s. Relative Displacement, in. 9 in. embedment 6 in. embedment 4 in. embedment 3 in. embedment 2 in. embedment Figure 3.10. Effect of embedment length on shear force versus displacement.

causing failure in the laboratory (see Figure 2.13). Therefore, the ultimate shear forces for dowel misalignments that could not be investigated in the laboratory can be estimated analyt- ically as the shear force corresponding to 0.05 in. [1.3 mm]. Table 3.5 gives the ultimate shear capacities for different levels of longitudinal translation, concrete covers, vertical tilt, and dowel diameter. The table shows that the shear capacity of a dowel is reduced by increasing levels of longitudinal transla- tion (reduced embedment length), increased by increasing concrete cover, and not affected by the magnitude of vertical tilt. These results are similar to those observed from the laboratory tests except for the case of 4 per 18 in. [102 per 457 mm] tilt. It also should be noted that the beam test does not account for the effect of interaction with multiple dowels which could be important for vertical tilt. The data also show that reductions in dowel diameter reduce the shear capacity and shear stiffness. 3.3.2 Finite Element Slab Model The laboratory tests indicated that dowel rotations up to 2 in. [51 mm] or less per 18 in. [457 mm] of dowel length did not result in significantly different dowel-concrete system responses under loading. A similar trend also was observed from the analytical simulations. However, previous studies (Khazanovich et al., 2001) have shown that rotations of adja- cent dowels or multiple dowels in a single joint can influence the behavior of the joint. To investigate this effect, a slab model with multiple embedded dowels was considered. The material parameters obtained from the beam model were used to simulate dowel performance using the slab model for the following four cases of rotational combinations: • Case 1: All dowels rotated by the same amount, but adjacent dowels are rotated in opposite directions. • Case 2: Each dowel tilted with the same magnitude and in the same direction. • Case 3: The dowel in the wheel path aligned properly, and each other dowel rotated with the same magnitude and direction. • Case 4: The dowel at the wheel path rotated, and each other dowel aligned properly. For each case, joint LTE was calculated as the ratio of the corner deflection of the unloaded slab to the corner deflection of the loaded slab. Table 3.6 shows LTE for these four cases with different levels of dowel rotation. LTE is lower for the oppositely misaligned dowels especially as the dowel tilt exceeds 1 in. per 18 in. [25 mm per 457 mm]. Thus, although the mean misalignment level of Case 2 was higher than Case 1 (the mean misalignment is zero in Case 1), Case 1 results in a lower joint LTE. Therefore, the mean misalignment level is required for characterizing the translational misalignments, but the standard deviation of the dowels is required to describe the rotational joint misalignment. Table 3.6 also shows that, for the same magnitude of dowel tilt, the LTE is lower for Case 4 than for Case 3, especially at the higher misalignment levels. Thus, the alignment of the dowel in the wheel path (critical dowel) has a more significant effect on the LTE than the alignment of the other dowels in the joint. 3.4 Pavement Performance Modeling 3.4.1 Faulting The equivalent dowel diameter concept requires the consid- eration of adjustment factors for each type of misalignment. The development of such factors is presented in this section. 3.4.1.1 Embedment Length Adjustment Factor The finite element model was used to obtain the embedment length adjustment factor. A series of finite element runs were 24 Embedment, in. Shear Capacity, lbs. Concrete Cover, in. Shear Capacity, lbs. Rotation, in./18 in. Shear Capacity, lbs. Diameter, in. Shear Capacity, lbs. 2 4900 3.25 10400 0 10400 1.0 5600 3 6600 4.25 12000 0.5 10300 1.125 6800 4 8000 5.25 13400 1.0 10200 1.25 8000 5 9100 7.25 14600 1.5 10300 1.375 9200 6 9900 2.0 9700 1.5 10400 9 10400 Longitudinal Translation Vertical Translation Vertical Tilt Dowel Diameter Table 3.5. Shear capacities for different levels of misalignment. Dowel Tilt, in./18 in. Case 1 Case 2 Case 3 Case 4 0.5 84.5 86.0 85.7 84.9 0.75 84.1 85.3 85.3 84.4 1 82.0 83.5 85.1 83.2 1.25 81.0 82.9 85.1 1.5 79.5 82.1 85.2 82.3 Average LTE 82.2 84.0 85.3 83.7 LTE, percent - Table 3.6. LTE predictions for various levels of dowel rotation.

made. The first series was performed for a 1.5-in. [38-mm] diameter dowel with embedment lengths varying from 2 to 9 in. [51 to 229 mm] and for dowel diameters ranging from 1.0 to 1.5 in. [25 to 38 mm] with a 9 in. [229 mm] embedment. Figures 3.11 and 3.12 present the relationships between embed- ment length and shear capacity and between dowel diameter and shear capacity, respectively. By equating the shear capacity of a misaligned dowel with the shear capacity of an aligned dowel of reduced diameter, an equivalent reduced dowel diameter could be determined. For example, a 1.5-in. [38-mm] dowel with embedment of 5 in. [127 mm] has a shear capacity of 9000 lb [40 kN] (Figure 3.14), which is equivalent to that of a 1.4-in. [36-mm] diameter dowel with embedment of 9 in. [229 mm]. The adjustment factor then is calculated by dividing the corresponding dowel diameter by the nominal dowel diameter. Therefore, the adjustment factor for an embedment length of 5 in. [127 mm] is 1.4/1.5 = 0.933. Figure 3.13 presents the computed adjustment factor remb for a range of the embedment lengths Lemb. This relationship can be presented by the following equation: where Lemb is the embedment length in inches. Equation 6 is applicable for embedment lengths between 2 and 6.9 in. [51 and 175 mm]. An adjustment factor of 1 should be assumed for embedment lengths greater than 6.9 in. [175 mm] and 0 for embedment lengths less than 2 in. [51 mm]. In cases where the embedment length of the dowels varies along the joint, this procedure will result in a different equiv- alent dowel diameter for each dowel in the joint. However, because the MEPDG faulting model assumes the same diam- eter for all dowels, a single equivalent dowel diameter that accounts for all dowels needs to be estimated. Finite element modeling shows that the LTE of the joint is affected by misalignment of the dowel in the wheel path approximately as much as the combined effect of the same level of misalignments for all of the other dowels in the joint (see Table 3.6). Therefore, the following procedure should be used for a joint with variable dowel embedment lengths: 1. Compute an adjustment factor for each dowel in the joint. 2. Determine the mean adjustment factor for all of the dowels in the joint. 3. Determine the mean adjustment factor for the three dowels in the critical wheel path (for example, the right wheel path in the truck lane). 4. Use the average of the two values obtained in Steps 2 and 3 as the adjustment factor for the joint. r L Lemb emb emb= − + +0 010 0 167 0 3242. . . (6) 25 0 2000 4000 6000 8000 10000 12000 0 2 4 6 8 10 U lti m at e S he ar Fo rc e, lb s. Embedment Length, in. Figure 3.11. Dowel shear capacity versus embedment length for 1.5-in. [38 mm] diameter dowels. 0 2,000 4,000 6,000 8,000 10,000 12,000 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 U lti m at e S he ar Fo rc e, lb s. Dowel Diameter, in. Figure 3.12. Dowel shear capacity for various dowel diameters (embedment length equal to 9 in. [229 mm]). 0.0 0.4 0.8 1.2 0 1 2 3 4 5 6 7 8 9 Co rre ct io n Fa ct or (r em b) Embedment Length, in. Figure 3.13. Adjustment factor (remb) versus embedment length.

3.4.1.2 Vertical Translation (Low Concrete Cover) Adjustment Factor Laboratory beam testing was conducted only for nonverti- cally translated dowels and dowels with a vertical translation of 2 in. [51 mm] (representing concrete covers ranging from 1.25 in. to 3.375 in. [32 mm to 86 mm]). The finite element beam model was used to extend the results of the laboratory tests to concrete covers up to 7.25 in. [184 mm]. Dowels 1.25 in. and 1.5 in. [32 mm and 38 mm] in diameter were used in the analysis for concrete covers of less than 5.25 in. [133 mm], and only 1.5-in. [38-mm] diameter dowels were used for con- crete covers of 5.25 in. [133 mm] and greater (because 1.5-in. [38-mm] diameter dowels are commonly used in thick pave- ments that result in these large concrete covers). This analysis produced the following relationship between dowel shear capacity and concrete cover: where DSC = the dowel shear capacity in lbs; d0 = the dowel diameter in inches; and CC = the concrete cover in inches. Figure 3.14 presents the relationship between concrete cover and shear capacity for two dowel diameters obtained from model simulations and the laboratory test data. Figure 3.14 illustrates the reduction in shear capacity due to a reduction in concrete cover. The field measurements conducted in this study showed that vertical translation of up to 0.5 in. [13 mm] should be expected for concrete thick- nesses of 12 in. [305 mm] or less, and a study conducted by MTO (MTO, 2007) concluded that translation of up to 1 in. DSC CC CC d= − +( )153 3 25032 0. (7) [25 mm] should be expected for concrete thicknesses above 12 in. [305 mm]. Therefore, reduction in shear capacity should be considered only if the reduction in concrete cover exceeds the normal variability level, as represented by the following equation: where ΔDSC = the change in dowel shear capacity in pounds due to reduction in concrete cover beyond normal variability; CC = the concrete cover in inches; and CCref = the reference level concrete cover in inches. CC and CCref can be computed as follows: where HPCC = the designed PCC thickness in inches; d0 = the designed dowel diameter in inches; and Ddepth = the depth of the dowel in inches. However, if the computed CCref > 3.5 d0, then 3.5 d0 should be used as the CCref. This maximum value was selected based on the results showing no increase of dowel shear capacity for concrete cover exceeding 3.5 times the dowel diameter. If CC is equal to or greater than the reference concrete cover (CCref), no reduction in effective dowel diameter should be considered (i.e., rcc = 1.0). If CC is less than 2 in. [51 mm], the adjustment factor should be considered to equal 0 (i.e., rcc = 0) because of high spalling potential. Effective dowel diameters should be considered for intermediate values. Figure 3.15 presents the relationship between dowel diam- eter and dowel shear capacity obtained from finite element analysis for HPCC equal to 8 in. [203 mm]. The relationship between the reduction in normalized shear capacity and the reduction in dowel diameter from d0 to d is presented as follows: Thus, the adjustment factor for concrete cover rcc can be pre- sented as follows: r d d DSC d cc = = − 0 0 1 1 9628 Δ (11) ΔDSC d d= −( )9628 0 (10) CC H d H CC H ref PCC PCC ref P = − − ≤ = 2 2 0 5 120 . , .for in CC PCC PCC d H CC H d 2 2 1 0 12 2 2 − − > = − 0 0 for in (9). , . − −H DPCC depth2 ΔDSC CC CC CC CC dref ref= − −( )+ −( )[ ]153 3 25032 2 0. (8) 26 0 4000 8000 12000 16000 0 1 2 3 4 5 6 7 8 Sh ea r Ca pa ci ty , lb s. Concrete Cover, in. 1.5 in. Laboratory 1.5 in. ABAQUS 1.25 in. Laboratory 1.25 in. ABAQUS 1.5 in. Equation 7 1.25 in. Equation 7 Figure 3.14. Shear capacity versus concrete cover.

Substituting Equation 8 into equation 11 results in the following adjustment factor for each individual dowel: However, the adjustment factor should be assumed to be zero for concrete covers less than 2 in. [51 mm] because low concrete cover can cause spalling around the dowel. Even if spalling is not visible, the ability of the dowel to transfer the load will be diminished. Figure 3.16 presents the cal- culated dowel diameter adjustment factors versus vertical translation for combinations of PCC thickness and dowel r CC CC CCcc = − − ∗( ) + ∗( )+ (1 153 3 2503 153 32. .ref ref )⎡⎣ − ∗( ) ⎤⎦ 2 2503 9628CC (12) diameter. The figure indicates that the adjustment factor decreases as the vertical translation increases; the decrease is less drastic for thicker concrete slabs because of the larger concrete cover. Finite element analysis showed that the LTE near the slab corner is affected by the dowel in the wheel path as much as all of the other dowels in the joint combined. Therefore, the following procedure should be used for a joint with variable concrete cover: 1. Compute an adjustment factor for each dowel in the joint. 2. Determine the mean adjustment factor for all of the dow- els in the entire joint. 3. Determine the mean adjustment factor for the three dow- els in the critical wheel path (for example, the right wheel path in the truck lane). 4. Use the average of the two values obtained in Steps 2 and 3 as the adjustment factor for the joint. 3.4.1.3 Rotation (Horizontal or Vertical Tilt) Adjustment Factor Dowel rotation in the form of vertical tilt and horizontal skew can have adverse effects on the performance of concrete pavement joints. Increased restraint to joint opening and closing due to dowel rotation may cause micro-damage and minor spalling around dowels (as observed in the laboratory tests) that reduce joint LTE. Laboratory tests and analyses have shown similar effects for vertical tilt and horizontal skew. Therefore, the equivalent dowel diameter concept can be applied in a similar manner to both types of misalignment. The equivalent dowel diameter concept used to account for the effects of translational misalignments also is used to account for the effects of rotational misalignments. As noted earlier, relative rotations of dowels (e.g., opposite misalignment) have a greater effect on the joint performance than rotational mag- nitude. The effect of rotational misalignments on joint LTE was determined by analyzing slabs with multiple dowels. In this analysis, the corner deflections of the loaded and unloaded slabs were computed for various combinations of dowel misalignment. The joint LTE was calculated by dividing the unloaded slab corner deflection by the loaded slab corner deflections, and the nondimensional joint stiffness, JStiff,was determined using the following equation (Khazanovich and Gotlif, 2002): in which LTE is expressed as a percentage. JStiff LTE= − ⎛ ⎝⎜⎜ ⎞ ⎠⎟⎟ − 1 0 01 0 012 1 17786 . . . (13) 27 0 4000 8000 12000 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 U lti m at e S he ar Fo rc e, lb s. Dowel Diameter, in. Change in Shear Capacity Change in Dowel Diameter Figure 3.15. Shear capacity versus dowel diameter. 0.5 0.6 0.7 0.8 0.9 1 1.1 -6 -4 -2 0 2 4 6 Vertical Translation, in. 13 in. thick - 1.5 in. dd 10 in. thick - 1.5 in. dd 13 in. thick - 1.25 in. dd 10 in. thick - 1.25 in. dd 8 in. thick - 1.25 in. dd Co rre ct io n Fa ct or Figure 3.16. Concrete cover adjustment factors (rcc) versus vertical translation for combinations of dowel diameter and PCC thickness.

Finite element analysis was performed for each of the four rotational misalignment cases described in Section 3.3.2, and the joint stiffness was determined. Linear regression analysis was used to develop the following relationship: where MeanTilt = the average tilt of the dowels in the joint in inches per 18 in.; StDTilt = the standard deviation of the tilt of the dowels in the joint in inches per 18 in.; and WPTilt = the maximum dowel tilt in the critical wheel path in inches per 18 in. JStiff0 is the computed stiffness of the joint with aligned dowels; this value is presented in Table 3.7 for each dowel diameter. These joint stiffness values account for the contri- butions of the dowels to the stiffness of the joint, but not those of aggregate interlock, foundation support, or other factors. The joint LTE can be predicted using the following equation (Crovetti, 1994): Figure 3.17 presents the load transfer efficiencies using Equations 14 and 15 versus the load transfer efficiencies computed from the finite element analysis. These values correlate well. Figure 3.18 shows the sensitivity of predicted LTE to the level of tilt for uniformly and oppositely tilted 1.25-in. [32-mm] diameter dowels in a joint. The LTE decreases with increases in tilt level. The LTE values are lower for oppositely tilted dowels than for uniformly tilted dowels. Similar observations were noted in earlier studies (Khazanovich et al., 2001). The field study showed that construction practices should permit the installation of dowels with tilt no greater than 0.5 in. per 18 in. [13 mm per 457 mm] of dowel length. Such a level of misalignment did not affect pavement performance. LTE JStiff = + ∗( )− 100 1 1 2 0 849 % . . (15) JStiff JStiff MeanTilt StD= − × − ×0 0 20623 0 61796. . Tilt WPTilt− ×0 86862. (14) Based on these observations, joints with oppositely misaligned dowels with rotation of 0.5 in. per 18 in. [13 mm per 457 mm] dowel length were used to represent the nominal condition. Thus, any combination of dowel misalignment that results in a LTE equal to or greater than the nominal LTE will not affect the pavement performance; rotations that result in less than the nominal LTE will have adverse effects on the joint performance. Table 3.8 gives the nominal load transfer efficiencies obtained from the finite element slab model for various dowel diameters; lower LTE is obtained for smaller dowel diameters. The pre- dicted LTE can be expressed in terms of the diameter of a properly aligned dowel, using the following relationship: d e LTE= 0 0103 0 0582. . (16) 28 Dowel Diameter (in.) JStiff0 1 6.537 1.125 7.447 1.25 8.461 1.375 9.601 1.5 10.894 Table 3.7. Computed stiffness for various dowel diameters. y = 1.0001x R² = 0.9376 79 80 81 82 83 84 85 86 87 79 80 81 82 83 84 85 86 87 Pr ed ic te d L TE (% ) Finite Element Model Computed LTE (%) Figure 3.17. Computed versus predicted LTEs for various rotational misalignments. 74 76 78 80 82 84 86 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 Pr ed ic te d LT E (% ) Tilt, in./18 in. Case 2: Uniformly Tilted Case 1: Oppositely Tilted Figure 3.18. Predicted LTEs for oppositely and uniformly tilted dowels.

By substituting Equation 15 into Equation 16, the adjustment factor for rotational misalignment (vertical tilt or horizontal skew), rrot, is obtained as follows: If the adjustment factor rrot is greater than 1, a value of 1 should be assumed. 3.4.1.4 Faulting Prediction After computing the adjustment factors for all misalignment types, the equivalent diameter for the doweled joint should be computed using Equation 4. The effect of dowel misalignment on the reduction of pavement life with respect to faulting and/or the reduction in predicted faulting reliability then can be determined using the MEPDG faulting model. 3.4.2 JPCP Cracking Joint lockup can cause transverse cracking. As discussed in Section 2.3.3, if joint lockup causes significant longitudinal stresses, then it can be accounted for by the MEPDG crack- ing model. Earlier studies identified the possibility of high longitudinal stresses as a result of rotationally misaligned dowels. One of these studies (Khazanovich et al., 2001) did not consider concrete damage in the immediate area of the dowel and over- estimated the joint capacity to resist opening. Another study (Prabhu et al., 2006) indicated the possibility of large longi- tudinal stresses for unrealistically large joint openings. To consider the effect of possible joint lockup on longitu- dinal stresses, the mid-slab stresses for aligned and misaligned dowels were considered. For the misaligned case, each dowel was tilted by 2 in. [51 mm] over its 18 in. [457 mm] length, and the dowels were configured so that adjacent dowels were rotated in opposite directions to provide the greatest potential for joint lockup. Figure 3.19 presents the computed longitudinal stresses for various joint openings that represent different degrees of thermal contractions. r d d d LTErot = = ( ) 0 0 0 0103 0 0582 . exp . (17) For small joint openings, the slab with misaligned dowels experienced slightly higher stresses than the slab with aligned dowels at the joint. However, above a certain amount of joint opening, the slab with aligned dowels experiences consistently higher stresses than the slab with misaligned dowels. This phenomenon is caused by damage in the concrete surrounding the misaligned dowels, which reduces the ability of the joint to resist opening. Based on these observations, it can be concluded that dowel misalignment itself is not a sufficient cause for joint lockup and it does not cause significant additional stresses in the longitudinal direction. Also, the MEPDG cracking model can be used without modification to account for dowel misalignment. Several conditions were evaluated using the finite element slab model to investigate the effect of friction at the dowel- PCC interface. These conditions included a joint with aligned dowels and a joint with two levels of misalignment (0.25 in. [6 mm] over 18 in. [457 mm] and 2 in. [51 mm] over 18 in. [457 mm]) for very high friction between the dowel and PCC. Figure 3.22 shows the computed longitudinal stresses for these cases and the low friction cases for both an aligned and a misaligned dowel. Figure 3.20 indicates that aligned dowels with high friction result in significantly higher longitudinal stresses and higher resistance to joint opening than misaligned dowels with low friction. Also notice that both the low and high friction cases for misaligned dowels did not converge for the entire range of joint openings (due to failure of concrete around the dowels). The analysis showed very high stress concentrations at the surface in the vicinity of the dowels, which may cause micro- cracking or micro-spalling in the concrete around the dowels and lead to joint lockup. 29 Dowel Diameter, in. LTE 0.75 0.73 1 1.01 1.125 1.14 1.25 1.26 1.375 1.36 1.5 1.46 Table 3.8. LTE versus dowel diameter for slab model simulations. 0 5 10 15 20 25 30 35 40 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 St re ss , p si. Deflection, in. Aligned Misaligned Figure 3.19. Longitudinal stresses for slabs between joints containing aligned and rotationally misaligned dowels.

While misalignment can result in higher longitudinal stresses, joint lockup is more greatly affected by friction between the dowel and PCC. The laboratory tests identified the lack of dowel grease as a cause for higher dowel-concrete interface friction and significant increase of the maximum pullout force. 3.4.3 Joint Spalling In addition to joint performance considerations, the location of dowels with respect to the top PCC surface may present constructability and safety concerns. In some cases, joints are not saw cut to the prescribed depth to avoid cutting through the dowels located too close to the surface. In this case, joints will not be formed and random cracking occurs. Also, a com- bination of vertical translation and tilt may result in low con- crete cover at the end of the dowel. This condition may cause a high stress concentration in the concrete at the dowel end, possibly leading to spalling and exposure of the dowel to the pavement surface. To address safety concerns related to the low concrete cover, it is recommended to saw cut through the dowels at the joint if any dowel end is located within 2 in. [51 mm] of the top pavement surface. This minimum concrete cover requirement is supported by the findings of a full scale laboratory study in which a slab with 2 in. [51 mm] of concrete cover above the dowel performed as well as a slab with 3 in. [76 mm] of cover for up to 10 million load cycles (Odden et al., 2003). Also, the laboratory testing conducted in this project showed that a dowel with 1.25 in. [32 mm] of concrete cover sustained a shear force of 4.5 kip [20 kN], which is greater than the force required to be transferred by the critical dowel in a pavement joint loaded by a 60-kip [267 kN] tandem axle load (based on ISLAB2000 analysis). The minimum recommended concrete cover should allow for adequate consolidation of concrete around dowels and may be increased for aggregate with maximum size greater than 1 in. [25 mm] based on local experience. Also, the min- imum concrete cover may be increased if high possibility for dowel corrosion exists. If the concrete cover is less than the minimum required concrete cover, the equivalent dowel diameter should be set equal to zero (i.e., undowelled joint is assumed.) 3.4.4 Pavement Performance Prediction The procedure for predicting pavement performance based on the measured dowel misalignment level requires that an equivalent dowel diameter be determined for each joint in a pavement using the MEPDG procedures. The following obser- vations were considered in formulating the performance pre- diction procedure: • The MEPDG prediction models were calibrated using the data contained in the LTPP database for 500-ft [152-m] long pavement sections. • Joints with severe dowel misalignment that are concentrated in a certain portion of the pavement section will have a greater effect on performance than similar joints distributed randomly along the pavement project. • Nondestructive dowel location devices could be used to investigate dowel misalignment in pavement projects. • The equivalent dowel diameter of each joint can be deter- mined using the equations provided in Chapter 3. • The mean value of a pavement design parameter (pavement thickness, subgrade resilient modulus, etc.) along the section is used. The following procedure is suggested for analyzing the effects of dowel misalignment on the performance of a uniform pavement project: 1. Use dowel alignment measurements to calculate the equiv- alent dowel diameter in each joint of the pavement project using the procedure described in Section 3.4.1. 2. Establish “uniform sections” approximately 500 ft. [152 m] long for the purposes of analysis and evaluation such that all joints in the section have similar equivalent dowel diameters. A series of several joints in any section with relatively uniform equivalent dowel diameters that are substantially different from the others in uniform sections may be evaluated as a separate uniform section. 3. Compute the mean equivalent dowel diameter for each section. Two or more adjacent sections with no significant difference in equivalent dowel diameters can be combined into a single section. 30 0 50 100 150 200 250 0 0.01 0.02 0.03 0.04 0.05 St re ss , p si. Deflection, in. High Friction; 1 in./18 in. Rotation High Friction; 0.125 in./18 in. Rotation High Friction; Aligned Low Friction; Aligned Low Friction; 2 in./18 in. Rotation Figure 3.20. Effect of dowel-PCC friction on longitudinal concrete stresses.

4. Perform MEPDG computations for each uniform section using the calculated mean equivalent dowel diameter for the section. If certain adjustment measures such as dowel retrofitting are performed, the effective dowel diameters of the retrofitted joints should be recalculated and the pavement performance predictions computed. 3.5 Examples of Application of The Equivalency Concept 3.5.1 Example 1. Assessment of a Single Joint The following example illustrates the calculation of the effect of dowel misalignment on joint performance for a joint in an 11-in. thick pavement. The joint is assumed to contain 12 dowels with 18 in. [457 mm] length and 1.5 in. [38 mm] diameter with the following features: 1. The saw cut is 4 in. [102 mm] away from the designed loca- tion, resulting in 4 in. [102 mm] of longitudinal translation and 5 in. [127 mm] of embedment length for all dowels. 2. The dowel basket used for the placement was 0.75 in. [19 mm] taller than was required for the mid-depth dowel placement, resulting in 0.75 in. [19 mm] vertical trans- lational displacement towards the pavement surface and reduced concrete cover from 4.75 in. to 4 in. [121 mm to 102 mm] for all dowels. 3. The dowels were placed with the rotational misalignment (vertical tilt and horizontal skew) given in Table 3.9. The procedure for determining the equivalent dowel diam- eter for this joint involves the calculation of four adjustment factors corresponding to the types of dowel misalignment assumed in this example. 3.5.1.1 Embedment Length Adjustment Factor Since the embedment length is greater than 2 in. [51 mm] and less than 6.9 in. [175 mm], the adjustment factor for the longitudinal translation and related reduction in embedment length is calculated using Equation 6 as follows: 3.5.1.2 Vertical Translation (Low Concrete Cover) Adjustment Factor The reference concrete cover and the actual concrete cover (CC) are calculated using Equation 9, as follows: CCref is also limited to a maximum of three times the dowel diameter, or 4.5 in. [114 mm] in this example. Thus, the calculated value for CCref of 4.25 in. [108 mm] is used. The adjustment factor for the loss in concrete cover is calculated using Equation 12 as follows: 3.5.1.3 Vertical Tilt Adjustment Factor For the vertical tilt values provided in Table 3.10, mean vertical tilt of 0.2 in. [5 mm], standard deviation of the verti- cal tilt of 0.633 in. [16 mm], and wheel path dowel vertical tilt of 0.5 in. [13 mm] were calculated. The joint stiffness can be calculated using Equation 14 as follows: The LTE of the joint can be calculated using Equation 15 as follows: LTE = + ( ) =− 100 1 1 2 10 03 85 51 0 849 . . . % . JStiff = − ×( )− ×(10 8942 0 20623 0 2 0 61796 0 633. . . . . ) − ×( ) =0 86862 0 5 10 03. . . rcc = − − ∗( ) + ∗( ) + ∗( )⎡1 153 3 4 25 2503 4 25 153 3 4 2. . . .⎣ − ∗( ) ⎤⎦ =2503 4 9628 0 968. CC CC ref = − − = = − − − 11 2 1 5 2 0 5 4 25 11 2 1 5 2 11 2 . . . . in. 4 75 4. = in. remb = − ( ) + ( )+ =0 010 5 0 167 5 0 324 0 9092. . . . 31 Dowel Bar Number Vertical tilt, in./18 in. Horiz. Skew, in./18 in. 1 -0.44 -0.26 2 0.50 -0.32 3 -0.34 -0.32 4 -0.80 -0.38 5 -0.54 -0.48 6 1.46 -0.27 7 -0.54 -0.39 8 0.46 -0.33 9 -0.54 -0.47 10 -0.54 -0.43 11 -0.54 -0.44 12 -0.54 -0.42 Table 3.9. Assumed dowel misalignments.

The adjustment factor for vertical tilt can be obtained from Equation 17 as follows: 3.5.1.4 Horizontal Skew Adjustment Factor For horizontal skew values provided in Table 3.10, mean horizontal skew of 0.38 in. [10 mm], standard deviation of the horizontal skew of 0.073 in. [1.9 mm], and maximum wheel path dowel horizontal skew of 0.32 in. [8 mm] were calcu- lated. The joint stiffness can be calculated using Equation 14 as follows: The LTE of the joint can be calculated using Equation 15 as follows: The adjustment factor for horizontal skew can be obtained from Equation 17 as follows: Because the adjustment factor should not exceed 1, an adjustment factor of 1.0 should be assumed. 3.5.1.5 Assembly of Calculated Adjustment Factors The equivalent dowel diameter (deq) for the joint is obtained by multiplying the original dowel diameter (d0) by the adjust- ment factors for concrete cover, embedment length, vertical tilt, and horizontal skew as follows: Since the concrete cover for each dowel was greater than the minimum required concrete cover, no further reduction of the equivalent dowel diameter is needed. Therefore, to account for the effects of the misalignment in this joint, the pavement d r r r r deq emb cc vt hs= × × × × = × × ×0 0 909 0 968 0 996 1. . . × = 1 5 1 31 . . in. rhs = ×( ) =0 0103 1 5 0 0582 86 03 1 02 . . exp . . . LTE = + ( ) =− 100 1 1 2 10 49 85 98 0 849 . . . . JStiff = − ×( )− ×10 8942 0 20623 0 38 0 61796 0 073. . . . .( ) − ×( ) =0 86862 0 32 10 49. . . rvt = ×( ) =0 0103 1 5 0 0582 85 51 0 995 . . exp . . . should be treated as if it had dowels with a diameter of 1.31 in. [33 mm] (and not the actual 1.5-in. [38-mm] diameter). 3.5.2 Example 2. Assessment of a Pavement Section The following example illustrates the calculation of the effect of dowel misalignment on the performance of a 540-ft. [165-m] pavement section with an 11 in. [279 mm] thickness. The pavement section has 30 joints, each of which contains 12 dowels with 18 in. [457 mm] length and 1.5 in. [38 mm] diameter. The pavement was designed for the following per- formance criteria (after 20 years at 90 percent reliability): • Transverse cracking not to exceed 12% of cracked slabs. • Mean joint faulting not to exceed 0.12 in. [3 mm] • IRI not to exceed 190 in./mile [3.0 m/km]. The equivalent dowel diameters were calculated for each joint (results are shown in Table 3.10). Because the pavement section is less than 1000 ft [305 m], the mean equivalent dowel diameter was computed for the entire pavement section result- ing in 1.41 in. [36 mm]. This equivalent dowel diameter was then used in an MEPDG simulation to predict faulting and IRI for the project. Figures 3.21. and 3.22 present the predicted faulting and IRI, respectively, for the as-designed pavement (dowel diameter of 1.50 in. [38 mm]) and for a similar pave- ment with 1.41 in. [36 mm] dowels. These results indicate that the predicted faulting and IRI of the project are within the specified acceptance thresholds. However, analysis of the MEPDG run output files (not pre- sented here) showed that because of dowel misalignment the reliability of faulting and IRI not exceeding the performance threshold decreased from 96.7 to 91.9% and from 92.5 to 91.0%, respectively. 32 Joint # Equivalent Dowel Diameter (in.) Joint # Equivalent Dowel Diameter (in.) Joint # Equivalent Dowel Diameter (in.) 1 1.31 11 1.5 21 1.21 2 1.5 12 1.22 22 1.5 3 1.41 13 1.5 23 1.5 4 1.14 14 1.5 24 1.27 5 1.5 15 1.49 25 1.5 6 1.1 16 1.5 26 1.5 7 1.5 17 1.5 27 1.5 8 1.5 18 1.23 28 1.5 9 1.5 19 1.05 29 1.37 10 1.5 20 1.5 30 1.5 Table 3.10. Equivalent dowel diameter for each joint in the pavement section.

33 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0 2 4 6 8 10 12 14 16 18 20 22 Fa u lti ng , in . Pavement age, years Performance Threshold 1.5 in. 1.41 in. Figure 3.21. Predicted faulting for the as-designed pavement project. 80 120 160 200 240 0 2 4 6 8 10 12 14 16 18 20 22 IR I, in . / m ile Pavement age, years Performance Threshold 1.41 in. 1.5 in. Figure 3.22. Predicted IRI for the as-designed pavement project.

Next: Chapter 4 - Conclusions and Suggested Research »
Guidelines for Dowel Alignment in Concrete Pavements Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Report 637: Guidelines for Dowel Alignment in Concrete Pavements examines the effects of dowel misalignment on concrete pavement performance, and highlights measures for reducing misalignment and its adverse effect.

Appendixes A through D to NCHRP Report 637 are available online and provide detailed information on the literature review, laboratory and field test results, and finite element analysis.

Appendix A: Review of Literature and Other Relevant Information

Appendix B: Field Testing Results

Appendix C: Laboratory Testing Results

Appendix D: Finite Element Analysis

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!