2013 Session: 665

• Falling Weight Deflectometer Loading Pulse Duration and Its Effect on Predicted Pavement Responses
  Abstract: Falling weight deflectometer (FWD) testing is commonly used to characterize, through backcalculation, in situ moduli of flexible pavement layers. While an often-used tool, it is well known that the short duration impact load applied by the FWD is often much faster than load durations applied by traffic. This can lead to an overestimation of asphalt concrete (AC) layer moduli under traffic which can underestimate computed pavement responses for design. The objective of this study was to determine the ability of FWD testing to represent traffic loading and pavement responses at typical operating speeds. FWD tests were performed on nine sections containing instrumentation at the National Center for Asphalt Technology (NCAT) Test Track. The measured FWD pulse durations ranged from 0.030 to 0.050 seconds over a range of temperatures. Measured loading pulse durations from trucks moving at 45 mph were found to be two to three times greater than the measured FWD pulse durations at the same temperature. The mechanistic analysis software 3D-Move was used to model moving loads to produce a loading pulse equivalent to the loading pulse produced by the FWD. The equivalent speed was determined to be over 120 mph. Statistically significant differences between the simulated responses from the Test Track operating speed of 45 mph and at simulated responses from the equivalent FWD speeds were obtained.
  Authors: Leiva-Villacorta, Fabricio; Timm, David H.
  Authors: Leiva-Villacorta, Fabricio; Timm, David H.
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-2171
• Effect of Cross-Anisotropy of HMA Modulus on FWD Deflections and Embedded Sensor Stress-Strain
  Abstract: In this study, the effects of cross-anisotropy on asphalt pavement responses are examined. A dynamic Finite Element Model (FEM) is developed in ABAQUS to simulate pavement responses under Falling Weight Deflectometer (FWD) and truck loads on a pavement section at interstate I-40 (Mile Post 141) in New Mexico. This section has recently been instrumented with strain gauges, moisture probes, and pressure cells. Pavement response (i.e. stress, strain, deflection) from the instrumented section are compared to the FEM predicted values. Two combinations of cross-anisotropy are considered. First combination considers cross-anisotropy of modulus in every layers of the pavement, whereas the second combination considers it only in the Hot Mix Asphalt (HMA) layer. Time-deflection histories, stress, and strain are predicted from the FEM under FWD and truck loads. Results show that predicted deflections, stress, and strain are highly sensitive to cross-anisotropy. Predicted deflections, stress, and strain increase with a decrease in n-value, defined by the ratio of horizontal to vertical modulus of elasticity. Analyses are performed for two different shapes of loading area: semi-circle-rectangle and rectangle. It is shown that predicted stress and strain are larger for rectangular shape of loading area than for semi-circle-rectangle area. This study recommends including modulus anisotropy in FWD backcalculation and in pavement design.
  Authors: Ahmed, Mesbah Uddin; Tarefder, Rafiqul Alam; Islam, Md Rashadul
  Authors: Ahmed, Mesbah Uddin; Tarefder, Rafiqul Alam; Islam, Md Rashadul
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-3203
• Viscoelastic Genetic Algorithm for Inverse Analysis of Asphalt Layer Properties from Falling Weight Deflections
  Abstract: Falling Weight Deflectometer (FWD) is a nondestructive test typically used to backcalculate layer properties of in-service pavements. Most backcalculation methods assume the pavement to be layered elastic structure. However, asphalt pavements are multilayered viscoelastic systems and their response under small load level is primarily viscoelastic. Hence, although elastic analysis is computationally efficient, they don’t produce the fundamental material properties such as the complex modulus E* or creep compliance D(t). On the other hand, there are dynamic backcalculation methods that can consider the asphalt pavements as viscoelastic, however they are computationally inefficient and inaccurate if the FWD deflection time history is truncated. Furthermore, most existing methods either ignore the asphalt concrete (AC) temperature or assume the entire depth of the AC layer to be at a constant temperature. In this study, a new inverse analysis method is proposed to backcalculate linear viscoelastic pavement layer properties (i.e., E(t)) as well as the time-temperature superposition shift factor coefficients. In this method, FWD load-response history of a single FWD drop and variation in temperature along the depth of AC layer during the drop are used. Genetic algorithm based optimization scheme is offered to search for the pavement properties, which has the potential of converging to a global solution. As an example, two LTPP sections are selected for backcalculation. The results are positive, indicating that there is a possibility to backcalculate AC properties from a single FWD drop, provided the effect of dynamics is insignificant during the test.
  Authors: Varma, Sudhir; Kutay, Muhammed Emin; Levenberg, Eyal
  Authors: Varma, Sudhir; Kutay, Muhammed Emin; Levenberg, Eyal
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-4091
• Development of Time-Efficient Genetic Algorithm Backcalculation Method for Flexible Pavements
  Abstract: Non-destructive testing (NDT) methods, especially Falling Weight Deflection (FWD) test, are invaluable for assessing the structural load capacity of pavements. In order to determine the structural capacity of existing roads, engineers need accurate information on the thickness and elastic modulus of the various layers. Layer moduli and thicknesses are used to calculate load capacities to predict the performance of roads and select appropriate rehabilitation actions. Unfortunately, there is no analytical approach for back-calculation of FWD test output. Most of the existing back-calculation programs model deflection basin by minimizing a function which is related to difference of the calculated and measured deflections. Since the objective function is usually complex and has several minima, traditional calculus-based optimization techniques are often ineffective for obtaining the best fit. This paper, reviews different methods that can be used to back-calculate pavements layer moduli. Since genetic algorithms (GAs) have demonstrated significant potential for solving problems involving multi-modal search spaces, in this study GAs were employed to obtain the best deflection basin fit. Although GAs are strong search engines, they are not time efficient and require a long search time. The objective of this study is mainly to improve the time efficiency of GA-based search engine used for back-calculation. Using a novel method of coding and data structure, authors developed a time efficient back-calculation software, named MBGA: Moduli Back-calculation using Genetic Algorithm. Other issues such as the effect of pavement structure on the optimum value of GA parameters, and the effect of a rigid layer on MBGA convergence were investigated. MBGA performance was evaluated under various conditions for three and four-layer pavement systems. Results demonstrated that the new GA-based search engine not only performed well in finding the global optima, but also reduced the search time significantly. Since MBGA unlike the other GA-based programs requires reasonably short computational time, it can meet the engineering needs for practical use.
  Authors: Motamed, Arash; Tabatabaee, Nader; Motamed, Moggan
  Authors: Motamed, Arash; Tabatabaee, Nader; Motamed, Moggan
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-3637
• Effect of Cross-Anisotropy of HMA Modulus on FWD Deflections and Embedded Sensor Stress-Strain
  Authors: Ahmed, Mesbah
  Authors: Ahmed, Mesbah
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-3203
• Development of Time-Efficient Genetic Algorithm Backcalculation Method for Flexible Pavements
  Authors: Motamed, Arash
  Authors: Motamed, Arash
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-3637
• Falling Weight Deflectometer Loading Pulse Duration and Its Effect on Predicted Pavement Responses
  Authors: Leiva-Villacorta, Fabricio
  Authors: Leiva-Villacorta, Fabricio
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-2171
• Viscoelastic Genetic Algorithm for Inverse Analysis of Asphalt Layer Properties from Falling Weight Deflections
  Authors: Varma, Sudhir
  Authors: Varma, Sudhir
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-4091
• Viscoelastic Genetic Algorithm for Inverse Analysis of Asphalt Layer Properties from Falling Weight Deflections
  Authors: Kutay, Muhammed
  Authors: Kutay, Muhammed
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-4091
• Viscoelastic Genetic Algorithm for Inverse Analysis of Asphalt Layer Properties from Falling Weight Deflections
  Authors: Levenberg, Eyal
  Authors: Levenberg, Eyal
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: 13-4091
• Presiding Officer
  Authors: Carvalho, Regis
  Authors: Carvalho, Regis
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 665
  Paper Number: Z13-665