National Academies Press: OpenBook

Aesthetic Concrete Barrier Design (2006)

Chapter: Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment

« Previous: Chapter 4 - Aesthetic Concrete Barrier Design Guideline Development
Page 32
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 32
Page 33
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 33
Page 34
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 34
Page 35
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 35
Page 36
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 36
Page 37
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 37
Page 38
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 38
Page 39
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 39
Page 40
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 40
Page 41
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 41
Page 42
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 42
Page 43
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 43
Page 44
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 44
Page 45
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 45
Page 46
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 46
Page 47
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 47
Page 48
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 48
Page 49
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 49
Page 50
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 50
Page 51
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 51
Page 52
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 52
Page 53
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 53
Page 54
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 54
Page 55
Suggested Citation:"Chapter 5 - Simulation and Preliminary Aesthetic Design GuidelineDevelopment." National Academies of Sciences, Engineering, and Medicine. 2006. Aesthetic Concrete Barrier Design. Washington, DC: The National Academies Press. doi: 10.17226/13888.
×
Page 55

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.

32 CHAPTER 5 SIMULATION AND PRELIMINARY AESTHETIC DESIGN GUIDELINE DEVELOPMENT INTRODUCTION Opening geometry in see-through rails, if improperly designed, can have a devastating effect on the rail’s crash performance when struck by a vehicle. Likewise, in mono- lithic concrete barrier surfaces that are not see-through, sur- face discontinuities, protrusions, or depressions in the face of the barrier can introduce vehicle instability and/or snagging. Surface discontinuities, protrusions, or depressions in the face of the rail or at rail openings may be acceptable, provided their depth and/or geometry do not produce excessive vehic- ular snagging and excessive decelerations. The effect of architectural surface treatments is little understood and could have significant safety-related effects. Native area stones can be applied as a veneer to enhance the appearance of concrete barriers. To date, the FHWA’s guide- lines for vertical-faced, crash-tested stone masonry guardwall state that maximum projections should not extend beyond 38 mm of the neat line, deep raked joints should be 50 mm thick, and mortar beds should be 50–75 mm thick. Stone that creates protrusions greater than described is not consid- ered crashworthy. Based on aesthetics and stone availabil- ity, a smoother stone face may be used, such as Class A or B masonry. In addition to native stone, alternative methods of form- ing concrete walls and barriers provide designers with a wide range of possible architectural treatments in the form of patterns and textures. Caltrans tested several architec- tural surface treatments applied to the Type 60 single-slope (9.1-degree) concrete barrier and identified several textures and patterns that could be applied to the single-slope con- crete barrier. Crash testing of single-slope median barrier with aesthetic surface treatments by Caltrans resulted in the first set of guidelines for the aesthetic surface treatment of concrete bar- riers. As a result of the Caltrans study, recommendations for allowable surface asperity geometry on the face of single- slope and vertical-face barriers were developed. The guide- lines, which were approved by the FHWA in acceptance let- ter B-110, permit the following types of surface treatments: • Sandblasted textures with a maximum relief of 9.5 mm. • Images or geometric patterns cut into the face of the bar- rier 25 mm or less and having 45-degree or flatter cham- fered or beveled edges to minimize vehicular sheet metal or wheel snagging. • Textures or patterns of any shape and length inset into the face of the barrier up to 13 mm deep and 25 mm wide. Geometric insets with an upstream edge with an angle of up to 90 degrees should be less than 13 mm deep. • Any pattern or texture with gradual undulations that have a maximum relief of 20 mm over a distance of 300 mm. • Gaps, slots, grooves, or joints of any depth with a max- imum width of 20 mm and a maximum surface differ- ential across these features of 5 mm. • No patterns with a repeating upward sloping edge or ridge. • Any pattern or texture with a maximum relief of 64 mm, if such pattern begins 610 mm or higher above the base of the barrier and all leading edges are rounded or sloped to minimize any vehicle snagging potential. No part of this pattern or texture should protrude below the plane of the lower, untextured portion of the barrier. Prior to the Caltrans study, there was a lack of any guidance regarding acceptable surface treatment of concrete barriers at the national level, and little or no uniformity existed in aes- thetic barrier design among the states. While the Caltrans study addressed single-slope and vertical-face concrete barriers, there was no design guidance for widely used safety shape concrete barriers. The primary objective of this research was to develop guidelines for the aesthetic surface treatment of New Jersey and F-shaped concrete barriers (herein generally referred to generically as safety shape barriers) based on bar- rier impact performance. The guidelines are intended to aid engineers and designers in choosing aesthetic surface treat- ments for concrete safety shape median and roadside barriers that will not adversely affect crashworthiness. When considering the geometry of surface asperities, variables include the depth, width, and shape of the relief or recess. Due to the number and range of these variables, it was economically impractical to conduct a parametric investigation based solely on crash testing. However, the researchers believed that a parametric investigation could be performed using finite element computer simulations that can provide a detailed assessment of the three-dimensional impact response associated with the introduction of specific aesthetic treatments.

33 A pilot study was conducted to demonstrate the feasibility of using finite element simulation for this research. Simulation was used as a tool to develop a set of preliminary guidelines that defined relationships between different design parameters for aesthetic surface treatments on safety shape barriers. Once these preliminary guidelines were established, a full-scale crash-testing effort was conducted. The initial crash-tested configurations were picked based on simulation results. The results from these crash tests were analyzed in conjunction with the preliminary guidelines to determine the asperity geometries to be evaluated in subsequent crash tests. This pro- cedure maximized the information available for adjusting the preliminary guidelines to yield the final design guidelines. Simulations were performed using LS-DYNA. LS-DYNA is a general-purpose, explicit-implicit, nonlinear finite element program capable of simulating complex nonlinear dynamic impact problems. LS-DYNA has been used extensively in simulations involving vehicular impacts with roadside safety appurtenances, including safety shape barriers. The decision to choose this explicit finite element code for this research was based on several reasons, including: • The availability of vehicle models that correspond to NCHRP Report 350 design test vehicles (mainly the 820C and 2000P vehicles). These vehicle models have been used for roadside safety applications for the last 6 or more years, and their fidelity and limitations are reasonably understood. • The ability to model the geometry of the safety shape barriers and the details of the aesthetic surface treatment (which affects the mechanics of the vehicle-barrier inter- action) with a high degree of fidelity. • The availability of a large contact algorithm library. These contact algorithms provide means to model vehicular col- lisions with roadside objects. PILOT STUDY AND FINITE ELEMENT MODEL VALIDATION In all types of modeling, approximations must be made when trying to represent reality. If finite element analysis is to be used to assess the effects of surface treatments on con- crete median barriers, the vehicle and barrier models must be capable of capturing the dynamic response associated with a concrete barrier impact. This capability was investigated by performing a pilot validation study. The pilot study had the following three objectives: • Perform finite element simulation of previously available crash tests so as to identify potential modeling problems. • Make necessary changes to improve the performance of the vehicle models and validate them for use in evaluat- ing the performance of barriers with surface asperities. • Identify surrogate measures for assessing OCD. Accurately representing the geometry of a concrete barrier and added surface asperities is fairly straightforward. How- ever, one of the limitations associated with using current finite element analysis codes to model concrete barriers relates to the material behavior. At the time of undertaking this research, there were no available robust concrete material models that could accurately and efficiently capture the failure/fracture of concrete. Even though the FHWA was at that time sponsoring the development of such a material model, it was not available in time for use in this project. However, since the concrete bar- rier profiles of interest in this project had been successfully crash tested and their structural adequacy was not at issue, this was not considered as a significant limitation. For this reason, modeling the concrete barriers as a rigid material without fail- ure was considered a reasonable and practical assumption. Most of the effort devoted to the validation effort therefore focused on the vehicle models. The validity of the improved 820-kg passenger car and 2,000-kg pickup truck vehicle models was established by comparing the results of simulations with the results of full- scale crash tests. It should be noted that an accurate compari- son of a simulation with a successful test does not necessarily constitute validation. It is important that some of the tests selected for use in the validation study include relevant failure modes. The two most critical failure modes associated with the performance evaluation of longitudinal barriers are vehic- ular instability (i.e., rollover) and OCD. While the validation study focused on these two evaluation criteria, other vehicular acceleration-based criteria were also analyzed and compared. 820C Vehicle Model The researchers identified historical crash tests that could be used to assist with vehicle model validation in the pilot study. The number of crash tests useful for this purpose was very limited. One of the first simulations that the researchers performed was that of Caltrans Test No. 582.(19) In this test, a 1990 Geo Metro impacted a single-slope concrete barrier with an inclined fluted surface at a speed of 100 km/h at an angle of 20 degrees. These testing conditions conform to the impact conditions for Test 3-10 in NCHRP Report 350 (see Figure 33). The slope of the single-slope barrier was 9.1 degrees from vertical, and the overall height of the barrier Figure 33. Caltrans single-slope barrier with fluted surface texture.

34 was 1.42 m. The surface of the barrier was modified to incor- porate inclined flutes or ribs. The flutes were oriented at a 45- degree angle from the ground, rising in the direction of vehi- cle travel. The cross section of each flute was 19 mm high and 19 mm wide. The flutes were spaced 50.8 mm on the cen- ter along the length of the barrier. The vehicle was redirected but rolled over as it exited the barrier. The vehicle model used in the initial simulation of this impact was the reduced Geo Metro model that was developed by the National Crash Analysis Center (NCAC) under FHWA sponsorship. This model contains approximately 16,100 ele- ments. Initial simulation results did not show a good correla- tion with the test results. Several changes were made to the original model to improve its performance in interacting with the surface asperities. Changes focused primarily on the vehi- cle’s front suspension and the tires. The suspension was mod- ified to include deformable control arms and some of the other linkages for the suspension mechanism (see Figure 34). Sim- ulation results with the modified 820C vehicle model showed better correlation with the Caltrans fluted-surface, single- slope concrete barrier test results. In addition to the Caltrans fluted barrier, a smooth single- slope barrier was used as a baseline system to validate the modified Geo Metro model. A comparison of vehicle dynam- ics between crash tests and finite element simulations of the Caltrans fluted single-slope barrier and those of standard single-slope barrier follows. Caltrans Single-Slope Barrier with Angled Flutes Figure 35 shows sequential images comparing the actual crash test of the fluted single-slope barrier with the finite ele- ment impact simulation of the barrier using the modified Geo Metro model. A comparison of roll, pitch, and yaw angles versus time is shown in Figures 36 through 38, respectively. A significant improvement in correlation of the roll angle was achieved, but with minor divergence in the pitch angle correlation. Single-Slope Barrier After the modified Geo Metro model demonstrated the ability to capture interaction with the inclined asperities on the fluted single-slope barrier, a baseline simulation using a smooth-faced single-slope barrier was performed. The pur- pose was to verify that the changes made to the Geo Metro model did not adversely affect other areas associated with concrete barrier impacts. Figures 39 through 41 compare the angular displacements of the vehicle obtained from the crash test(20) and simulation of the single-slope barrier. Improved correlation was observed with the modified model for both the roll and pitch behavior. Both models showed good correlation with the test data for the yaw angle. Summary of the 820C Vehicle Model Validation At the start of the simulation study, a significant effort was put into the improvement and validation of the 820-kg, small- car model for impacts into single-slope barriers with and with- out surface asperities. As described above, results from full- scale crash tests performed by Caltrans were used to help assess validity of the model for this purpose. The suspension on the original reduced Geo Metro model was extensively (a) (b) Figure 34. 820C front suspension: (a) modified model; (b) actual vehicle.

Figure 35. Sequential comparison of test and simulation of angle fluted barrier. Figure 36. Comparison of roll angles from crash data with vehicle simulations for the angle fluted barrier. 35

36 Figure 37. Comparison of pitch angles from crash data with vehicle simulations for the angle fluted barrier. Figure 38. Comparison of yaw angles from crash data with vehicle simulations for the angle fluted barrier.

Figure 40. Comparison of pitch angles from crash data with vehicle simulations for the single-slope barrier. Figure 39. Comparison of roll angles from crash data with vehicle simulations for the single-slope barrier. 37

38 modified, and this modified version of the Geo Metro was con- sidered to be adequately validated against single-slope barrier tests with and without surface asperities (i.e., angled flutes) to proceed with its use in this project. At the interim panel meeting, the focus of the research changed from single-slope barrier to New Jersey safety shape barriers. Consequently, the validation of the small-car finite element model had to be revisited. The number of crash tests into rigid concrete safety shape barriers with 820-kg cars was found to be very limited. Some New Jersey safety shape bar- rier tests that were identified were conducted on a modified- barrier profile in the early 1980s under NCHRP Report 230. The barrier modification consisted of a 75-mm pavement overlay in front of the barrier that covered the 75-mm reveal/lip at the bottom edge of the barrier. Further, the tests were conducted with a different vehicle (i.e., Honda Civic) at a 15-degree angle rather than the 20-degree angle currently specified in NCHRP Report 350. These factors limited the usefulness of these tests for validating the Geo Metro model for NCHRP Report 350 impacts into a New Jersey safety shape barrier. A reference to a 1981 test of an unmodified New Jersey safety shape barrier at a 20-degree impact angle was identi- fied. The test was conducted by Dynamic Science, Inc., under FHWA contract DOT-FH-11-9115. Despite considerable efforts by the research team, including consultation with the TTI librarian and the FHWA, the report and film for this test could not be located. TTI researchers ultimately simulated impacts of the Geo Metro into an F-shape barrier and the New Jersey safety shape barrier with the 75-mm reveal/lip covered by a pavement overlay. It was discovered that for these safety shape barriers, neither the original NCAC model nor the TTI-modified model exhibited adequate correlation. Although correlation was achieved with the single-slope barrier, the safety shape barri- ers interact differently with the vehicle’s tires, wheels, sus- pension, and so forth, and a model validated for one barrier shape will not necessarily work for another barrier shape. Further, the tests that were simulated were conducted with a Honda Civic rather than a Geo Metro. Therefore, it is not known how much of the observed differences between the tests and simulations were attributed to differences in vehicle- barrier interaction versus differences in vehicle type. While the validation effort for the 820C vehicle was going on, TTI researchers were working in parallel on the 2000P vehicle model validation, details of which follow in subse- quent sections in this chapter. As mentioned previously, the 2000P pickup truck design vehicle is believed to be more crit- ical than the 820C in regard to evaluation of OCD and stabil- ity in impacts with concrete barriers. As an example, con- sider the Texas T411 aesthetic bridge rail shown in Figure 42. An impact into this barrier with an 820-kg passenger car at Figure 41. Comparison of yaw angles from crash data with vehicle simulations for the single-slope barrier.

97 km/h and 21.2 degrees was successful, while an impact into this barrier with a 2000-kg pickup truck at 101 km/h and 24.9 degrees failed due to excessive OCD (see Figure 43). Since the small car was not considered to be the critical design vehicle from the standpoint of evaluating OCD or sta- bility, the researchers planned to use the pickup truck as the primary vehicle for developing the preliminary guidelines. The role of the small-car model was to be limited to check- ing the preliminary guidelines established by the pickup truck. Therefore, rather than undertaking another extensive effort to improve the validity of the Geo Metro model for impacts into the New Jersey safety shape barrier while retain- ing sufficient fidelity to detect surface asperities, the research team shifted its focus to the development of preliminary guide- lines based on parametric simulations with the pickup truck model. Once the preliminary guidelines were established based on the parametric simulations conducted with the pickup truck, the research team sought input from the project panel regard- ing the panel’s desire to revisit the validation of the small-car model for impacts with safety shape barriers. The research team believed that with additional time and resources, improved correlation of the Geo Metro for impacts into safety shape barriers could be achieved through further modification to the model. However, the benefits derived from such an effort needed to be weighed against the cost of the effort and the delay it would have imposed on the full-scale crash-testing program. Use of the small-car model was limited to providing a check of the preliminary guidelines established by the pickup truck. This objective could also be accomplished with a full- scale crash test. This approach was approved by the project panel. Conse- quently, further validation of the small car for the New Jersey shape barrier was discontinued, and a crash test was performed to check the validity of the guidelines for the small car. Details of the crash-testing phase are presented in Chapter 6. 2000P Vehicle Model Initial validation efforts for the 2000P vehicle were carried out with the reduced element pickup truck model that was developed by the NCAC. Simulations with the vehicle impact- ing a smooth-surface, single-slope barrier and a New Jersey shape barrier were performed and compared with available crash test data.(21,22) The correlation between test and simula- tion was not considered acceptable. Certain vehicle suspension parts in the reduced vehicle model (e.g., control arms) are modeled as rigid materials. The lack of deformability in the front suspension was believed to be the primary cause of the observed discrepancies between test and simulation. TTI researchers then simulated these crash tests using the NCAC detailed pickup truck model. This model, which contains approximately 54,800 elements, incorporates a deformable front suspension. A comparison of the vehicle dynamics resulting from the simulation and crash test showed reasonable correlation for both barriers. Figures 44 through 46 compare the roll, pitch, and yaw displacements, respectively, for the single-slope barrier. Figures 47 through 49 provide a similar comparison of angular displacements for the New Jersey safety shape barrier. While reasonable correlation was obtained for both barriers, it can be seen from these figures that the single-slope barrier showed better correlation. This is likely due to the more prominent role of the vehicle suspension in impacts with the New Jersey shape barrier and limitations in the suspension model of the finite element pickup truck. Having demonstrated reasonable correlation, the detailed NCAC pickup truck model was selected for use in the devel- opment of the preliminary design guidelines. Surrogate Measure of OCD As mentioned previously, a common cause of barrier failure in a crash test is excessive OCD. As an example, OCD failure Figure 42. Texas T411 aesthetic bridge rail. Figure 43. Pickup truck after impact with Texas T411 bridge rail. 39

40 Figure 44. Comparison of roll angles of crash data with detailed pickup truck vehicle simulation on the single-slope barrier. Figure 45. Comparison of pitch angles of crash data with detailed pickup truck vehicle simulation on the single-slope barrier.

Figure 46. Comparison of yaw angles of crash data with detailed pickup truck vehicle simulation on the single-slope barrier. Figure 47. Comparison of roll angles of crash data with detailed pickup truck vehicle simulation on the New Jersey safety shape barrier. 41

42 Figure 48. Comparison of yaw angles of crash data with detailed pickup truck vehicle simulation on the New Jersey safety shape barrier. Figure 49. Comparison of pitch angles of crash data with detailed pickup truck vehicle simulation on the New Jersey safety shape barrier.

was the most predominant type of failure in the Caltrans study, “Crash Testing of Various Textured Barriers.”(19) There has been little research performed to assess or improve the ability of vehicle models to accurately capture and predict OCD resulting from a barrier impact. Part of the pilot study con- ducted under this project was devoted to assessing the ability of existing vehicle models to predict OCD, either through direct measurement of the maximum deformation inside the passenger compartment (similar to the procedure used in a crash test), or by means of a surrogate measure correlated against the OCD measurements obtained in full-scale crash tests. Several crash tests of concrete barriers with the 820C pas- senger car and the 2000P pickup truck were identified. How- ever, the number of useful crash tests was limited, especially for the small car. This was because OCD was not measured and reported prior to the publication and adoption of NCHRP Report 350 and many of the small-car compliance tests with standard concrete median barrier shapes were conducted before NCHRP Report 350. All of the identified concrete barrier crash tests with measured OCD were modeled and simulated. Each simulation was set up to collect several potential measures of OCD. The objective was to determine a measure that would demonstrate the best correlation with the maximum OCD reported in the crash tests. Simulated Barrier Designs Oregon Bridge Railing. The Oregon bridge rail is a con- crete beam and post bridge rail similar to the Texas T411. When the impact performance of this barrier was evaluated with a pickup truck, the OCD was 475 mm, which signifi- cantly exceeded the 150-mm limit imposed by the FHWA.(23) Therefore, this test served as one of the failure points in the OCD pilot study. Figure 50 shows an image of the rail con- structed for the crash test and the associated LS-DYNA model used in the simulation of the system. Deep Cobblestone Barrier. The deep cobblestone barrier (shown in Figure 51) is a single-slope barrier with a random cobblestone surface treatment. This barrier was tested by Caltrans as part of its project to develop guidelines for aes- thetic surface treatments for single-slope barriers. The bar- rier failed the test due to excessive OCD of the pickup truck caused by the interaction of the wheel with the cobblestones. The maximum amount of relief on the cobblestone surface was 64 mm. For the simulation, the cobblestone surface was modeled using hemispherical and ellipsoidal shapes with the same depth as the actual surface treatment. Because this was one of the few pickup truck tests with a solid (i.e., without win- dows) concrete barrier that failed due to excessive OCD, it provided a useful data point for correlation of the surrogate OCD measures. Shallow Cobblestone Barrier. After the failure of the deep cobblestone barrier, the depth of the cobblestone sur- face treatment was reduced to 19 mm and retested by Cal- trans. Typical relief of the shallow cobblestone surface is shown in Figure 52. In the pickup truck crash test of this bar- rier, the drive shaft became dislodged from the transmission. Although the vehicle remained upright during the test, this type of damage was considered by Caltrans to represent a potential rollover risk. As a result, Caltrans decided that the barrier did not meet NCHRP Report 350 evaluation criteria. However, since the shallow cobblestone reduced the maxi- mum OCD of the vehicle to within acceptable limits, this test illustrated the effect of surface asperity depth on vehicle (a) (b) Figure 50. Oregon bridge railing: (a) actual; (b) simulation model. 43

44 response and represented another useful data point for devel- oping a surrogate measure for OCD. Cobblestone Reveal Barrier. An alternative treatment developed by Caltrans to address the OCD problems asso- ciated with the deep cobblestone barrier was to provide a smooth reveal at the bottom of the barrier. The 610-mm-tall reveal, which had a smooth, sandblasted finish (see Figure 53), was intended to reduce the snagging contact between the barrier and wheel assembly and, thereby, reduce the result- ing OCD. This test successfully passed NCHRP Report 350 criteria and provided another point for use in establishing the thresholds for a surrogate OCD measure. This barrier also possessed some similarity to the safety shape barriers that were to be addressed in this study, since the surface as- perities were to be applied to the upper-wall portion of the safety shape barrier, while the toe of the barrier was to be left smooth. Single-Slope Barrier, New Jersey Safety Shape Barrier, and Modified Texas T203 Bridge Rail The standard single-slope barrier, New Jersey safety shape, and Texas T203 bridge rail were also modeled and evaluated. Each of these tests had acceptable OCD (i.e., < 150 mm) and met NCHRP Report 350 guidelines. These “passing” crash tests provide confidence in establishing a “passing” threshold for the selected surrogate OCD criterion. Results The first and most obvious measure of OCD was to take a direct measurement of the maximum deformation to the floorboard and toe pan of the vehicle in a manner similar to that used in crash test evaluation. An example of the OCD generated in the pickup truck model is shown in Figure 54. The deformation shown in Figure 54(b) is caused by induced buckling resulting from compression of the floorboard. (a) (b) Figure 51. Cobblestone barrier: (a) actual; (b) simulation model. (a) (b) Figure 52. Shallow cobblestone barrier: (a) actual; (b) simulation model.

Direct measurements of OCD were obtained from the simu- lations and compared with measured full-scale crash test OCD values. As shown in Table 1, there is some correlation observed between the simulation and the test data. However, the relia- bility of predicting the outcome of a test, based on a single num- ber from simulation, was not considered very high. In an actual crash test, the wheel, wheel well, fender, and other parts may contact the floorboard and cause additional OCD. The accurate representation of this mode of deformation requires failure in one or more components of the suspension that are not repre- sented in current vehicle models. For this reason, direct mea- surement was not used as a measure for predicting OCD. As mentioned above, much of the OCD in a barrier crash test results from the wheel and suspension assembly being deformed and shoved back into the toe pan area. An option was set into the model to collect the direct impact forces between the wheel and barrier. These forces were evaluated using several criteria. The XY and XYZ resultants of the peak force, peak 10-ms moving average force, impulse over the time of initial impact, and total impulse were all computed and analyzed to investigate their correlation to OCD measure- ments. These measures of contact force between the wheel and the barrier have been tabulated in Table 2. The correlation between these measures and actual OCD measurements was found to be poor. The amount of variation that exists in these data between acceptable and failed crash tests was not ade- quate to confidently use these forces as a surrogate measure for OCD. This is possibly due to the unreliable values of force between parts undergoing such severe deformation. Finally, the internal energies of the vehicle parts in the crushed region of the vehicle were obtained and checked for correlation to OCD measurement. The internal energy in a part is related to the overall deformation experienced by the part. Internal energies obtained from the floorboard and wheel well showed the best correlation to the actual crash test re- sults among the measures evaluated (see Table 3). Between these, the truck floorboard was selected as the surrogate measure of OCD because it had slightly better correlation, (a) (b) Figure 53. Cobblestone reveal barrier: (a) actual; (b) simulation model. (a) (b) Figure 54. Buckling floorboard: (a) undeformed; (b) deformed. 45

46 TABLE 1 Direct measurements for truck OCD study Pass/FailName Oregon Fail 475 170 Cobblestone Fail 160 225 Cobblestone with Reveal Pass 98 50 Single Slope Pass 140 50 Modified T203 Pass 130 80 Shallow Cobblestone Pass 133 105 New Jersey Pass Not Reported 80 Crash Test OCD [mm] Direct Measurement [mm] TABLE 2 Wheel to barrier contact forces and impulses for truck OCD study XYZ Resultant XY Resultant Crash Test OCD [mm] Total Impulse [N-s] Max 10 ms Moving Avg. [N] Max Force [N] Impulse [N-s]Name Oregon Fail Fail Fail Fail Pass Pass Cobblestone Single Slope Cobblestone with Reveal New Jersey Modified T203 Shallow Cobblestone Oregon Cobblestone Single Slope Cobblestone with Reveal New Jersey Modified T203 Shallow Cobblestone Pass/Fail Pass Pass Pass Pass Pass Pass Pass Pass 475 160 98 140 130 133 475 160 140 130 133 98 Not Reported Not Reported 1,290,000 1,340,000 1,290,000 1,176,000 276,000 498,000 228,000 263,000 901,000 278,000 229,000 290,000 910,000 510,000 459,000 450,000 195,000 164,000 197,000 231,000 459,000 455,000 438,000 195,000 196,000 164,000 123,000 449,000 28,800 29,900 34,500 10,700 15,200 14,500 14,800 12,100 12,800 12,800 21,100 18,700 18,700 27,700 28,300 31,900 35,900 14,300 14,600 10,600 15,100 12,000 12,700 11,900 19,800 18,000 18,000 40,100 TABLE 3 Internal energies for truck OCD study Crash Test OCD [mm] Floorboard Part [N-mm] Wheel Well Part [N-mm] Name Pass/Fail Oregon Fail Fail Pass Pass Pass Pass Pass Cobblestone Cobblestone with Reveal Single Slope New Jersey Modified T203 Shallow Cobblestone 475 160 98 140 Not Reported 130 133 9,826,000 10,783,000 782,400 721,300 1,130,000 1,172,000 2,150,000 14,140,000 11,040,000 3,980,000 2,469,000 2,870,000 3,300,000 7,540,000 because its deformation is less influenced by contact with other parts of the vehicle, and because use of floorboard deformation is more intuitively appealing given the nature of OCD that occurs in a crash test. Conclusions for the 2000P Study As mentioned above, the internal energy of the floorboard of the pickup truck was selected as the most appropriate surrogate measure for evaluating OCD. Using the internal energy from the simulations and the reported OCD values from the crash tests, thresholds for the surrogate measure were established. As shown in Figure 55, the passing limit was selected as 2,200 N-m and the failure limit was tenta- tively set at 10,700 N-m of internal energy in the floorboard of the pickup truck. The failure point that occurs at an internal energy of 7,100 N-m is associated with the Oregon bridge rail. It is

noted that the interaction of the vehicle with this rail is con- sidered to be substantially different than what typically occurs in an impact with a “solid” barrier. In the test of the Oregon barrier, the frame rail of the pickup truck protruded inside one of the “windows” and snagged severely on the inside of one of the concrete balusters. As a result, instead of the load going to the floorboard as it does in most OCD failures, the load was directed to the frame. Therefore, the Oregon bridge rail crash test was not taken into considera- tion when selecting the failure limit. The outcome of impacts with solid barriers in which the internal energy of the floorboard is between 2,200 N-m and 10,700 N-m is largely unknown due to lack of crash test data with a sufficient range of OCD values. As is described in Chapters 6 and 7, the full-scale crash tests were designed to adjust these energy thresholds and reduce the size of the region with unknown performance. GENERALIZED SURFACE ASPERITY DEFINITION In order to model asperities on the barrier surface for evaluation in the parametric simulation effort, it was nec- essary to adequately define their geometry. Almost all sur- face asperities can be placed within one of three categories: perpendicular, rounded, or angled surface interruptions. These generalized types of surface asperities are shown in Figure 56. The angled or inclined asperity can be defined in terms of a depth d and angle θ, either of which can be varied to achieve a different profile. The perpendicular asperity is a subset of the angled asperity with θ = 90 degrees. The rounded asperity can be approximated as an angled surface asperity by selecting an effective angle θ. The illustration shown in Figure 56 uses a tangent to the rounded surface Figure 56. Generalized types of surface asperities. Figure 55. Passing and failing crash tests OCD versus internal energies of floorboard. 47

48 at half the depth d to define an effective angle θ. Because the angled asperity is the most general, it is the type of sur- face asperity used in the parametric study to develop guide- lines for the aesthetic treatment of concrete safety shape barriers. For the purpose of this research, asperities were defined as the portion of the barrier that was recessed into the barrier sur- face. In other words, an asperity is a depression in the surface of the barrier. Thus, another critical dimension to be included in the parametric study was the width of the asperity, W, which was defined as the distance between the outer edges of the asperity spacing, as shown in Figure 57. The distance between two adjacent asperities was defined as the asperity spacing (Ws). For the parametric studies presented in this chapter, an asperity spacing of 25 mm was used. The asperities were created by depressing the surface of the barrier profile at the desired asperity locations. The orig- inal barrier profile was, therefore, unchanged in the regions between asperities, defined in Figure 57 as the asperity spac- ing (Ws). The asperities began at the top of the “toe” of the safety shape barrier and continued vertically to the top of the barrier. The toe of the barrier remained smooth, as shown in Figure 58. PRELIMINARY AESTHETIC DESIGN GUIDELINES All simulations in the parametric study to establish the pre- liminary aesthetic design guidelines were performed with the 2000P pickup truck impacting a rigid New Jersey safety shape barrier following Test 3-11 of NCHRP Report 350. The impact conditions for Test 3-11 involve the 2000-kg pickup truck impacting the barrier at a speed and angle of 100 km/h and 25 degrees, respectively. The parametric study was performed using 45-degree, 90-degree, and 30-degree asperity angles (θ). The asperity width (W) and depth (d) were systematically varied for each of these angles. The impact performance associated with each simulation run was assessed based on the established surrogate OCD thresholds. As previously mentioned, the passing and failing internal energy limits were selected as 2,200 J and 10,700 J, respectively. The results were used to establish pre- liminary relationships that identified asperity configurations having impact performance considered to be “acceptable,” “marginal/unknown,” and “unacceptable.” If for a simulated asperity configuration, the truck floorboard internal energy was more than 10,700 J, the configuration was marked as “unacceptable.” Floorboard internal energy value between Figure 57. Surface asperity geometry variables. Figure 58. Truck with fluted New Jersey shape barrier.

2,200 J and 10,700 J implied that the asperity configuration was marked as “marginal/unknown.” If the floorboard internal energy was less than 2,200 J, the configuration was marked as “pass” or “acceptable.” Other than the baseline simulation with the New Jersey bar- rier without asperities, none of the simulated configurations resulted in truck floorboard internal energy of less than 2,200 J. Consequently, “unacceptable” and “marginal/ unknown” were the only two regions identified in the prelim- inary guidelines. The “pass” region was later identified with the use of full-scale crash testing, details of which follow in the next chapter. It is worth mentioning that asperity configura- tions (with very small depths and large widths) can be selected such that they would result in floorboard internal energies of less than 2,200 J. Such configurations can be used to establish a “pass” region in the preliminary guidelines. However, the asperity depth and width values for such configurations would have no practical significance for the aesthetic surface treat- ment of concrete barriers and hence were not simulated. Simulation results for the truck floorboard internal energy for the 45-degree, 90-degree, and 30-degree asperity angles are presented in Tables 4 through 6, respectively. Simulations with no depth refer to a smooth-faced New Jersey safety shape barrier. The tables present floorboard internal energy for dif- ferent asperity configurations. Using the results for the simu- lated values of W and d, a curve denoting the failure threshold was plotted for each asperity angle. The corresponding rela- tionships for the 45-degree, 90-degree, and 30-degree asperity angles are shown in Figures 59 through 61, respectively. For convenience of use and comparison, the relationships for the 45-degree, 90-degree, and 30-degree asperity angles have been combined in Figure 62. The curves shown in Figures 59 through 62 provide only a failure line, above which the asperity geometries were pre- dicted to fail to meet impact performance criteria and below which the impact performance was unknown. As discussed previously, the existence of a region of unknown performance is due to a lack of crash tests of rigid barriers with measured OCD values corresponding to the regions below the failure line. This region of unknown performance was reduced through a judiciously selected full-scale crash-testing phase, the details of which are presented in the next chapter. Examining the shape of the guideline curves, the effects of asperity width (W) and depth (d) on barrier performance appear logical. When the asperity width (W) is small, the vehi- cle engages more asperities during its contact with the barrier. This in turn presents more resistance to vehicle sliding on the barrier and causes more snagging and damage to the vehicle. Consequently, we see a reduction in the allowable asperity depth (d ) for these smaller widths. As the width (W) of an asperity increases, the allowable depth (d) also increases up to a limiting or controlling value. It can also be seen that as the angle becomes shallower, the failure line moves upward to higher asperity depths. Thus, the 30-degree asperity angle results in a larger “marginal/ acceptable” region than the 45-degree asperity angle. Simi- larly, the “marginal/unknown” region significantly reduces with the increase in asperity angle from 45 degrees to 90 degrees. For the 90-degree asperity angle, it can be seen that for asperity widths less than 500 mm, the internal energy TABLE 4 Parametric study results for a 45-degree angle of asperity Asperity Width (W) [mm] Truck Floorboard Internal Energy [J] Asperity Depth (d) [mm] Passing Limit = 2,200 J Failure Limit = 10,700 J Run Vehicle Pass/Fail 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck 555 555 555 555 555 555 555 280 280 280 280 180 180 180 80 80 80 30 30 5 0 15 6.5 0 0 0 15 15 65 15 0 1,108 4,341 6,986 8,397 12,835 15,939 18,318 1,108 8,965 1,108 9,158 1,108 1,108 4,149 8,900 14,680 15,507 11,844 17,182 90 40 6.5 52.5 52.5 27.5 27.5 Pass Pass Pass Pass Pass Fail Fail Fail Fail Fail Fail Fail Marginal Marginal Marginal Marginal Marginal Marginal Marginal 49

50 of the truck floorboard exceeds the failure limit for almost all practical asperity depths. In the simulations, the 90-degree asperity angle induces more severe snagging and resistance to sliding, which causes more damage to the vehicle. An analogy can be drawn between the 90-degree asperities and splices in tubular steel rail members that have demonstrated the potential for severe snagging and increased OCD in full-scale crash tests. One must bear in mind that all simulated asperities (45-degree, 90-degree, and 30-degree asperities) were modeled as rigid. In an actual impact, spalling or fracture of the concrete asperities may occur. This spalling can serve to reduce the snagging forces below levels that would be induced by a rigid asperity. The degree of snagging reduction was difficult to quantify with- out additional test data. For the 90-degree asperity angle, as the asperity width (W) increases beyond 500 mm, the allowable depth (d) increases up to a limiting or controlling value. Another interesting point to note is that as the width (W) decreases to a value of 30 mm or less, a significant increase in the asperity depth (d) can be achieved. This is because even though d is large, the small width of the asperity reduces the potential for vehicle parts to intrude into the asperity. Thus, the opportunity for vehicle TABLE 6 Parametric study results for a 30-degree angle of asperity Asperity Width (W) [mm] Truck Floorboard Internal Energy [J] Asperity Depth (d) [mm]Run Vehicle Pass/Fail 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck 100 100 200 200 200 400 400 400 400 600 600 600 600 600 50 50 100 75 0 25 0 1,108 4,476 1,108 6,238 7,652 1,108 4,465 7,538 11,341 4,701 5,985 1,108 3,952 6,724 25 25 75 25 50 0 0 Pass Marginal Marginal Marginal Marginal Pass Marginal Pass Marginal Marginal Pass Marginal Fail Marginal Passing Limit = 2,200 J Failure Limit = 10,700 J TABLE 5 Parametric study results for a 90-degree angle of asperity Asperity Width (W) [mm] Truck Floorboard Internal Energy [J] Asperity Depth (d) [mm]Run Vehicle Pass/Fail 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck Truck 5 30 30 30 30 55 55 55 125 280 280 280 400 400 500 500 500 580 580 15 0 12.5 6.5 0 15 40 6.5 6.5 0 0 40 2,157 1,108 3,257 6,077 6,049 1,108 17,497 25,000+ 18,250 30,000+ 1,108 1,108 1,108 8,909 8,653 1,108 24,240 12,000+ 21,000+ 2.5 15 0 6.5 40 0 2.5 Pass Fail Fail Pass Pass Fail Fail Fail Pass Fail Pass Marginal Pass Marginal Marginal Fail Pass Marginal Marginal Passing Limit = 2,200 J Failure Limit = 10,700 J 19 Truck 580 27.5 14,000+ Pass

Figure 59. Depth versus width guideline for a 45-degree asperity angle. Figure 60. Depth versus width guideline for a 90-degree asperity angle. 51

Figure 61. Depth versus width guideline for a 30-degree asperity angle. Figure 62. Overlaid depth versus width guideline curves. 52

53 snagging and cumulative vehicle damage is reduced. Even if some intrusion into the asperity occurs, the depth of the intru- sion is again limited by the small width of the asperity. This effect is illustrated by the vertical failure line in Figure 60. In summary, simulation results for the 90-degree asperity angle indicate that only a very limited set of asperity configu- rations with this angle can be used for aesthetic barrier design. Further investigation was performed with a full-scale crash test, details of which are presented in the next chapter. It can be seen that for shallower asperity angles, much greater asperity depths can be achieved. Note that some of the simulated asperity depths for the 30-degree asperity angle may not be practical from a design standpoint. How- ever, because not all desired aesthetic surface treatments can be anticipated, a wide range of asperity depths has been included in the analysis to more completely define the rela- tionship between asperity depth and width for shallow-angle asperities. It is noted that a standard concrete barrier design will have a functional limit on the maximum asperity depth that can be accommodated without exposing the reinforcing steel or leav- ing insufficient concrete cover. In order to maintain the desired clear cover (typically 37.5 mm to 50 mm) for reinforcement steel when significant asperity depths are incorporated into the design, the barrier may need to be widened beyond the width required to satisfy strength requirements. The exact nature of the curves beyond an asperity width of 600 mm (denoted with a dashed line) is not completely defined. However, the curves should reach a limiting asper- ity depth as the asperity width increases. As the asperity width continues to increase, a point will be reached where the vehicle is only engaging or interacting with a single asperity during redirection. At this point, there will be a crit- ical asperity depth that will no longer be influenced by asperity width. It is worth noting that the failure lines shown in Figures 59 through 62 are placed at the midpoint between simulated sur- face asperity depths at a given asperity width that resulted in marginal and unacceptable OCD (as defined by the internal floorboard energy). However, the increase in internal floor- board energy is not linearly related to asperity depth. This can be observed by comparing the internal energy values for dif- ferent asperity widths and depths. The truck floorboard internal energy data for the 45- degree, 90-degree, and 30-degree asperity angles is pre- sented in three-dimensional graphs in Figures 63 through 65, respectively. For a given asperity width, the increase in internal energy with increase in the asperity depth can be Figure 63. Truck floorboard internal energy values at different configurations for a 45-degree asperity angle.

Figure 64. Truck floorboard internal energy values at different configurations for a 90-degree asperity angle. Figure 65. Truck floorboard internal energy values at different configurations for a 30-degree asperity angle. 54

55 15 mm, the internal energy increases from 4.3 kJ to 17.2 kJ as the asperity width decreases from 555 mm to 80 mm. Since the internal energy of the truck floorboard is believed to be related directly to the OCD, these graphs were helpful in understanding the relationships and trends associated with the asperity parameters. Such information was considered when developing the full-scale crash test plan. Having gained a reasonable insight into the effect of differ- ent asperity parameters on OCD and having established the preliminary guidelines presented above, the researchers initi- ated the crash-testing phase of the project. Details of this phase of the project are presented in the next chapter. visualized. It can be seen that as asperity width decreases, the relative difference in internal energy between two simi- lar asperity depths increases. For example, with reference to Figure 63, it can be observed that for an asperity width of 555 mm, the internal energy increases 63% (from 4.3 kJ to 7.0 kJ) as the asperity depth increases from 15 mm to 27.5 mm, respectively. For an asperity width of 80 mm, the internal energy increases by 93% (from 8.9 kJ to 17.2 kJ) when the asperity depth increases from 6.5 mm to 15 mm. Similarly, for a specific asperity depth, the floorboard inter- nal energy increases as the asperity width decreases. For example, with reference to Figure 63, for an asperity depth of

Next: Chapter 6 - Crash Testing and Further Evaluation of Preliminary AestheticDesign Guidelines »
Aesthetic Concrete Barrier Design 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 554: Aesthetic Concrete Barrier Design provides guidance for the aesthetic treatment of concrete safety shape barriers.

  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!