Application of a hybrid neural network structure for FWD backcalculation based on LTPP database

Pavement structure bearing capacity is an important evaluation parameter in pavement design, construction, maintenance management, and reconstruction, and is generally expressed by the pavement deflection value. Some of the current road bearing capacity detection equipment have high detection accuracy, but the detection speed is slow, they cannot achieve real-time continuous detection; and some detection speeds are fast, but the measurement accuracy is easily affected by the pavement roughness and vehicle vibration. Moreover, the detection result is the pavement displacement, which cannot directly reflect the comprehensive modulus of the pavement structure. In this paper, firstly, a two-stage jump mechanical model of “machine-pavement” system is established in order to develop a device that can simulate the real driving load and continuously test the bearing capacity of pavement structure, and the main factors affecting the acceleration response of vibrating drums were determined through analysis. Then, a finite element model of the “machine-pavement” system was established to overcome the difficulty in obtaining the parameters such as vibrating weight, equivalent stiffness, and equivalent damping of pavement structure in numerical solution of dynamic model. Next, the mean value A of the maximum acceleration signal of the vibrating drum and the coefficient of variation acv of the maximum acceleration signal were selected as evaluation indicators to analyze the change trend of the maximum acceleration of the vibrating drum with the excitation frequency and excitation force under different composite modulus of pavement structures. Finally, the relationship between the composite modulus E of the pavement structure and the maximum acceleration A of the vibrating drum was obtained by simulating the pavement structure with the composite modulus ranging from 100 MPa to 2900 MPa, and the accuracy of this relationship was verified by field tests. The research showed that the acceleration signal of the vibrating drum had a good fitting relationship with the bearing capacity of the pavement structure when the testing device with the vibrating drum mass of 100 kg, the exciting frequency of 60 Hz, and the exciting force of 650 N jumped on the pavement structure, and the error was about 20% after comparing with the results of Benkelman beam testing, which basically met the engineering requirements. Therefore, the device can be used to continuously detect the bearing capacity of pavement structures.

show abstract

“…is needed, and these details are hard to know. Moreover, the solution results are not unique in the back-calculation process [4,5].…”

Section: Discussionmentioning

confidence: 99%

“…At present, a large number of scholars have studied the pavement structure layer modulus back-calculation based on FWD deflection basin information and temperature correction [3][4][5]. Many scholars have also studied the traffic speed deflectometer and compared its detection results with those of FWD [6][7][8].…”

Section: Traffic Speed Deflectometer 2000smentioning

confidence: 99%

Development of a Continuous Testing Device for Pavement Structure Bearing Capacity

et al. 2022

View full text Add to dashboard Cite

show abstract

“…With the popularization of computer‐aided engineering, the artificial neural network (ANN; Sharma & Das, 2008) method and genetic algorithm (GA; Fwa et al., 1997) method were proposed for modulus back‐calculation. The ANN method currently has the fastest computing speed, but training an ANN takes a long time, and ANN performance depends on how successfully it has been trained (Han et al., 2022). The GA approach has high computational accuracy and global convergence.…”

Section: Introductionmentioning

confidence: 99%

Modeling and validation of impact forces for back‐calculation of pavement surface moduli

Zhao

Liu

Wang

et al. 2023

Computer aided Civil Eng

View full text Add to dashboard Cite

When an impact load is sufficiently small, its influence on the pavement structure is mainly from the surface layer material. To explore the influence depth of an impact load and back‐calculation of the pavement surface modulus, both numerical calculation and experimental testing were conducted, and the results are presented in this paper. The numerical calculation was performed through a DEM‐FDM‐coupled model. After a new modulus back‐calculation algorithm is formed by analyzing the numerical modeling results, experimental tests were also conducted for verification, and the results were analyzed. The max value of the falling weight impact force was closely related to the elastic modulus of the material, and the influence depth was controllable. Finally, it is proved that the method can be used to calculate the surface layer modulus and estimate the surface layer thickness.

show abstract

“…The most effective method of nondestructive testing, used to determine the mechanical characteristics of structural layers in road pavement, is the method of determining the modulus of elasticity of the layers, based on solving the inverse problem of restoring the required parameters by maximum vertical displacements (bowl of deflections) recorded experimentally under impact loading. Recent years have witnessed fundamentally new approaches to solving this class of problems: with the use of artificial neural networks (Han et al, 2021;Saltan et al, 2013;Vyas et al, 2021;Wang et al, 2021), genetic algorithms to adjust the theoretical and experimental fields of vertical displacements (Fwa et al, 1997;Le and Phan, 2021;Park et al, 2010;Tsai et al, 2004;Varma et al, 2013;Wang et al, 2019;Zhang et al, 2021), new approaches to dynamic deformation analysis (Bazi and Assi, 2022;Booshehrian and Khazanovich, 2018;Cao et al, 2020;Lee et al, 2018;Zhang et al, 2019;Zhao et al, 2015), consideration for the wave nature of deformation (Al-Adhami and Gucunski, 2021;Chatti et al, 2017;Marchant and Papagiannakis, 2010;Quan et al, 2022;Zaabar et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Determining Viscoelastic Characteristics of the Elements of Multi-Layer Structures Based on Energy Dissipation Analysis

Tiraturyan

Uglova²,

Akulov³

2022

A&J

View full text Add to dashboard Cite

The deteriorating operating conditions of roads is one of the most important problems facing specialists in the road industry. It is mainly associated with a reduction in the rigidity of road pavements made as extended multi-layer structures. To identify the causes of rigidity reduction, non-destructive testing is used, which is based on solving the inverse coefficient problem of restoring the elastic constants based on the response on the surface. Purpose of the study: We aimed to provide a rationale for a new pavement condition indicator that would take into account the history of deformation and loading from a source on the surface, based on which it would be possible to solve the inverse problem of restoring the elastic and viscous characteristics of pavement layers. Methods: To do that, we performed mathematical modeling of the stress-strain state of a multi-layer medium based on the solution of a system of dynamic Lame equations. Viscosity is taken into account by introducing tangents of the angles of wave energy losses in the materials of layers. Results: The results obtained in modeling for the first time made it possible to establish the relationship between changes in the modulus of elasticity as well as tangents of the angles of energy losses in pavement layers and the amount of energy dissipation in the structure. Discussion: It should be noted that it is possible to switch from the bowl of maximum dynamic deflections as the main pavement condition indicator to the analysis of hysteresis loops on the road structure surface, recorded at different distances from the point of load application and being an analogue of the full bowl of dynamic deflections showing the history of the test object loading.

show abstract

Application of a hybrid neural network structure for FWD backcalculation based on LTPP database

Cited by 35 publications

References 26 publications

Development of a Continuous Testing Device for Pavement Structure Bearing Capacity

Development of a Continuous Testing Device for Pavement Structure Bearing Capacity

Modeling and validation of impact forces for back‐calculation of pavement surface moduli

Determining Viscoelastic Characteristics of the Elements of Multi-Layer Structures Based on Energy Dissipation Analysis

Contact Info

Product

Resources

About