Stiffness characterisation of full-scale airfield test pavements using computational intelligence techniques

The falling weight deflectometer (FWD) is a non-destructive test equipment used to assess the structural condition of highway and airfield pavement systems and to determine the moduli of pavement layers. The backcalculated moduli are not only good pavement layer condition indicators but are also necessary inputs for conducting mechanistic based pavement structural analysis. In this study, artificial neural networks (ANNs)-based backcalculation models were employed to rapidly and accurately predict flexible airport pavement layer moduli from realistic FWD deflection basins acquired at the U.S. Federal Aviation Administration's National Airport Pavement Test Facility (NAPTF). The uniformity characteristics of NAPTF flexible pavements were successfully mapped using the ANN predictions.     

Airfield pavement deterioration assessment using stress-dependent neural network models

In this study, an artificial neural network (ANN)-based approach was employed to backcalculate the asphalt concrete and non-linear stress-dependent subgrade moduli from non-destructive test (NDT) data acquired at the Federal Aviation Administration's National Airport Pavement Test Facility (NAPTF) during full-scale traffic testing. The ANN models were trained with results from an axisymmetric finite element pavement structural model. Using the ANN-predicted moduli based on the NDT test results, the relative severity effects of simulated Boeing 777 (B777) and Boeing 747 (B747) aircraft gear trafficking on the structural deterioration of NAPTF flexible pavement test sections were characterized. The results indicate the potential of using lower force amplitude NDT test data for routine airport pavement structural evaluation, as long as they generate sufficient deflections for reliable data acquisition. Therefore, NDT tests that employ force amplitudes at prototypical aircraft loading may not be necessary to evaluate airport pavements.     

Co-variance matrix adaptation evolution strategy for pavement backcalculation

The falling weight deflectometer (FWD) is the foremost and widely accepted tool for characterizing the deflection basins of pavements in a non-destructive manner. The FWD pavement deflection data are used to determine the in situ mechanical properties (elastic moduli) of the pavement layers through inverse analysis, a process commonly referred to as backcalculation (B/C). Several B/C methodologies have been proposed over the years, each with individual strengths and weaknesses. Hybrid methods (combining two methods or more) are recently proposed for overcoming problems posed by stand-alone methods, while extracting and compounding the benefits that are individually offered. This paper proposes a novel hybrid strategy that integrates co-variance matrix Adaptation (CMA) evolution strategy, Finite element (FE) modeling with neural networks (NN) non-linear mapping for backcalculation of non-linear, stress dependent pavement layer moduli. The resulting strategy, referred as CMANIA (CMA with neural networks for inverse analysis) is applied for asphalt pavement moduli backcalculation and is compared with a conventional B/C approach. Results demonstrate the superiority of this method in terms of higher accuracy, achieving nearer to global solutions, better computational speed, and robustness in predicting the pavement layer moduli over the conventional methods.     

Simulating pavement structural condition using artificial neural networks

This article aims to investigate the feasibility of incorporating of an artificial neural network (ANN) as an innovative technique for modelling the pavement structural condition, into pavement management systems. For the development of the ANN, strain assessment criteria are set in order to characterise the structural condition of flexible asphalt pavements with regards to fatigue failure. This initial task is directly followed with the development of an ANN model for the prediction of strains primarily based on in situ field gathered data and not through the usage of synthetic databases. For this purpose, falling weight deflectometer (FWD) measurements were systematically conducted on a highway network, with ground-penetrating radar providing the required pavement thickness data. The FWD data (i.e. deflections) were back-analysed in order to assess strains that would be utilised as output data in the process of developing the ANN model. A paper exercise demonstrates how the developed ANN model combined with the suggested conceptual approach for characterising pavement structural condition with regard to strain assessment could make provisions for pavement management activities, categorising network pavement sections according to the need for maintenance or rehabilitation. Preliminary results indicate that the ANN technique could help assist policy decision makers in deriving optimum strategies for the planning of pavement infrastructure maintenance.     

Long-term deflection of cracked composite beams with nonlinear partial shear interaction — A study using neural networks

A new study of the short- and long-term deflections of simply-supported composite beams using finite element analysis and artificial neural networks (ANNs) is presented. In this study, two ANN models are developed and trained using the results of a finite element model developed by the authors in a companion paper. The finite element model accounted for the nonlinear load–slip relationship of shear connectors as well as the creep, shrinkage, and cracking of concrete slabs. The effects of creep and shrinkage of the concrete slab are considered only for non-cracked concrete. A large database representing a wide range of different design parameters was constructed for the purpose of training and verifying the two ANN models. It was found that the two ANN models were capable of predicting deflections of composite beams not used as part of the training process. The ANN models were then used to evaluate the effects of non-geometric design variables on the short- and long-term deflections of simply-supported composite beams. Finally, the short- and long-term deflections computed based on the approaches given in the AISC specification and Eurocode 4 were assessed using the results of the finite element model. It was found that the AISC approach underestimates short-term deflections and overestimate long-term deflections when compared with the results of the finite element method.     

Fuzzy logic modeling of deflection behavior against dynamic loading in flexible pavements

Flexible pavements are especially affected by moving vehicles. As a result of the moving vehicles, the pavement starts to deteriorate. For the determination of the structural capacity of the pavement, non-destructive testing equipments are used. These are mainly Benkelman beam, Dynaflect and falling weight deflectometer (FWD). In such a process, the most important thing is to analyze the collected data. In general, linear elastic theory and finite element method are used for this purpose. Since linear elastic theory and finite element method are time consuming, a fuzzy logic approach is used for the elimination of this drawback during the course of this study. Results indicate that the fuzzy logic approach can be used for the modeling of the deflection behavior against dynamic vehicle loading for flexible pavements. The fuzzy model is able to predict the deflection behavior against dynamical loading. The new approach can capture the non-linearity of surface deflection behavior.     

非线性局部剪力交互作用下的开裂组合梁长期变形神经网络研究

结合有限元分析和人工神经网络,提出一种新的思路,研究简支组合梁的短期和长期变形。本文建立两个神经网络模型,采用相关论文中有限元模型的结果进行样本训练。有限元模型考虑了抗剪连接件的非线性荷载-滑移关系,以及蠕变、收缩和混凝土板的裂缝。而对没开裂的混凝土只考虑了蠕变、收缩的影响。为训练及验证两个神经网络模型,建立了一个包括不同设计参数的大数据库。研究发现,两个神经网络模型均能预测组合梁的变形。因此,神经网络模型可用以评估非几何设计参数对简支组合梁的短、长期变形影响。最后,根据AISC规范和欧洲规范4方法计算简支组合梁的短、长期变形,并与有限元模型结果进行比较。结果表明,与有限元方法相比,AISC方法低估了短期变形而高估了长期变形。     

Laboratory evaluation of WMA mixture for use in airport pavement rehabilitation

The concept of warm mix asphalt (WMA) gives a promise for rehabilitating airport pavement to realize quick turnover to traffic after construction, however, laboratory and field data in terms of the performance-related properties are significantly lacking for using WMA in airfield in Japan. To fill this gap, three WMA mixtures (different gradations) were systematically investigated compared with the conventional airfield used hot mix asphalt (HMA) through a series of laboratory tests in terms of wheel tracking test, submerged wheel tracking test, raveling test, static bending and fatigue bending test. These WMA mixtures were made at two production temperatures (30 and 50 °С lower than the normal, respectively) by incorporating a commercially sold additive. Results showed that overall, the WMA mixture with a coarse gradation produced at the temperature 30 °С lower than the normal exhibited a comparable performance compared with the control HMA mixture, and it was further recommended for use in airport pavement rehabilitation.     

浦东国际机场道面混凝土的质量控制

浦东机场的第二跑道是国内在软土地基上建造的第一条4F级跑道,对跑道的质量要求也是国内一流的。本文结合工程实际,分析了混凝土抗折强度的影响因素,总结出混凝土内在和外在质量控制要点,针对各个施工环节进行全过程质量控制,确保道面混凝土工程质量达到设计要求。     

Assessing artificial neural network performance for predicting interlayer conditions and layer modulus of multi-layered flexible pavement

The objective of this study is to evaluate the performance of the artificial neural network (ANN) approach for predicting interlayer conditions and layer modulus of a multi-layered flexible pavement structure. To achieve this goal, two ANN based back-calculation models were proposed to predict the interlayer conditions and layer modulus of the pavement structure. The corresponding database built with ANSYS based finite element method computations for four types of a structure subjected to falling weight deflectometer load. In addition, two proposed ANN models were verified by comparing the results of ANN models with the results of PADAL and double multiple regression models. The measured pavement deflection basin data was used for the verifications. The comparing results concluded that there are no significant differences between the results estimated by ANN and double multiple regression models. PADAL modeling results were not accurate due to the inability to reflect the real pavement structure because pavement structure was not completely continuous. The prediction and verification results concluded that the proposed back-calculation model developed with ANN could be used to accurately predict layer modulus and interlayer conditions. In addition, the back-calculation model avoided the back-calculation errors by considering the interlayer condition, which was barely considered by former models reported in the published studies.     

Estimation of flexible pavement structural capacity using machine learning techniques

The most common index for representing structural condition of the pavement is the structural number. The current procedure for determining structural numbers involves utilizing falling weight deflectometer and ground-penetrating radar tests, recording pavement surface deflections, and analyzing recorded deflections by back-calculation manners. This procedure has two drawbacks: falling weight deflectometer and ground-penetrating radar are expensive tests; back-calculation ways has some inherent shortcomings compared to exact methods as they adopt a trial and error approach. In this study, three machine learning methods entitled Gaussian process regression, M5P model tree, and random forest used for the prediction of structural numbers in flexible pavements. Dataset of this paper is related to 759 flexible pavement sections at Semnan and Khuzestan provinces in Iran and includes “structural number” as output and “surface deflections and surface temperature” as inputs. The accuracy of results was examined based on three criteria of R, MAE, and RMSE. Among the methods employed in this paper, random forest is the most accurate as it yields the best values for above criteria (R=0.841, MAE=0.592, and RMSE=0.760). The proposed method does not require to use ground penetrating radar test, which in turn reduce costs and work difficulty. Using machine learning methods instead of back-calculation improves the calculation process quality and accuracy.     

Fuzzy Representation of Pavement Condition for Efficient Pavement Management

Abstract: Many government agencies and private consulting companies manage large pavement networks in terms of infrastructure condition assessment and maintenance planning. Efficient pavement management is supported by pavement management systems (PMSs), which includes models for pavement condition assessments considered “valuable” by agency's engineers. The objective of this article is to define a pavement condition model able to overcome surveyors’ subjectivity in rating distresses and thus provide meaningful pavement conditions for the agencies to employ in project planning. The article proposes a fuzzy inference model for calculating pavement condition ratio (PCR) specifically tailored on the Alabama Department of Transportation Pavement (ALDOT) guidelines and policies. Applied to several surveyors’ ratings, the proposed model has the ability to smooth distress extent differences among surveyors producing PCR values within acceptable range of variability. The proposed approach has the intention of not only enhancing pavement condition characterization but also to exploit the opportunity made available by automation in the collection and interpretation of pavement data which are anyway characterized by an inherent subjectivity.     

级配碎石层在2种典型路面结构中的计算分析

以长沙市武广片区道路试验段为依托工程,通过理论分析和3D有限元数值分析、工程应用与后期跟踪检测相结合等技术手段,对级配碎石层应用于2种典型道路路面结构(柔性结构和倒装结构)的性能差异进行了研究。利用大型有限元分析软件ABAQUS建立路面结构的3D模型,充分考虑了级配碎石层的弹塑性,对柔性结构、倒装结构在单次平面均布矩形荷载作用下的力学响应进行了分析;通过变化级配碎石层的回弹模量和厚度,对2种路面结构的力学指标峰值进行了敏感性分析。试验段检测结果表明,采用3D有限元模型分析路面结构能较真实地反映路面结构的实际力学响应。     

Pillar design by combining finite element methods, neural networks and reliability: a case study of the Feng Huangshan copper mine, China

This paper presents a mine pillar design approach by combining finite element methods (FEMs), neural networks (NN) and reliability analysis. This practical approach is presented by examining an actual cylindrical mine pillar in a copper mine and taking into account uncertainties in ore pillar material parameters including modulus, Poisson's ratio, density and uniaxial compressive strength. The ore pillar had to be able to safely and effectively support a drilling room that occupied an open space of 3.8 m high and 55 m long and 20 m wide and at a depth of 360 m below ground surface. Three-dimensional FEM was used to simulate the mining operations and to estimate average pillar compressive stress at each operation step. A pillar performance function was established in implicit form taking into account pillar strength and pillar dimension. NN was incorporated in the FEM to substantially reduce the number of finite element calculations in establishment of the relationship between pillar compressive stress and basic random variables. Trained NN was then used to generate a database for the implicit performance function. The database was used to determine the reliability index and failure probability for each trial pillar diameter. Relationship between pillar reliability index and each of the coefficients of variation of the basic random variables was used for optimal design of pillar diameter. The optimal pillar design was used in the mining construction and functioned well.     

Neural Network for Rapid Depth Evaluation of Shallow Cracks in Asphalt Pavements

Abstract:   Rapid and nondestructive evaluation of pavement crack depths is a major challenge in pavement maintenance and rehabilitation. This article presents a computer-based methodology with which one can estimate the actual depths of shallow, surface-initiated fatigue cracks in asphalt pavements based on rapid measurement of their surface characteristics. It is shown that the complex overall relationship among crack depths, surface geometrical properties of cracks, pavement properties, and traffic characteristics can be learnt effectively by a neural network (NN). The learning task is facilitated by a database that includes relevant traffic and pavement characteristics of Florida's state highway network. In addition, the specific data used for the NN model development also contained laser-scanned microscopic surface geometrical properties of cracks in 95 pavement sections and pavement core samples scattered within five counties of Florida. Relatively advanced training algorithms were investigated in addition to the Standard Backpropagation algorithm to determine the optimal NN architecture. In terms of optimizing the NN training process, the "early stopping method" was found to be effective. The crack depth evaluation model was validated based on an unused portion of the database and fresh core samples. The results indicate the promise of NN usage in nondestructive estimation of shallow crack depths based on crack-surface geometry and pavement and traffic characteristics .     

Integrity analysis of single and multi-layer thin cement mortar slab structures using the spectral analysis of surface wave NDT method

The non-destructive testing method using spectral analysis of surface waves (SASW) has mainly been developed and used for many years in the fields of geotechnical engineering and highway engineering, such as examining the material properties of pavement systems, soil media, etc., under an infinite half-space condition. In the recent decades, extensive research in this area has been focused on understanding the applicability and limitations of the SASW method. The method consists of the generation, measurement and processing of dispersive surface waves. In an SASW test, the surface of the media under consideration is subject to an impact, using, for example, a 12-mm steel ball to generate surface wave energy at various frequencies. Two vertical accelerometer receivers detect the energy transmitted through the testing media. By recording signals in digitized form using a data acquisition system and processing them, surface wave velocity can be obtained using a dispersion curve. Based on the dispersion curve, the shear wave velocity can also be plotted using forward modeling, which can be related to various material properties. This paper presents a modified experimental technique for the non-destructive evaluation of compressive strength of in-place single-layer concrete slabs through a correlation with the surface wave velocity. The paper also presents the relationship between the theoretical and experimental compact dispersion curves when the SASW test is applied to multi-layer thin cement mortar slab systems with a finite thickness. The test results show that the surface wave velocity profile obtained from the theoretical dispersion curve has lower values than the profile obtained from the experimental compact dispersion curve under the condition of a finite thickness due to different boundary conditions and reflections from the boundaries. Based on the measured response, an experimental study was conducted to examine if the dispersive characteristics of Rayleigh waves exist in the multi-layer cement mortar slab systems. This study can be utilized in examining structural elements of general concrete structures, and can also be applied to the integrity analysis of concrete structures with a finite thickness.     

路面结构对路基不协调变形的力学响应研究

采用有限元分析方法,考虑旧路路面结构利用与不利用两种典型工况,对拓宽公路因新旧路基不协调变形引起的脱空问题及其对路面结构的影响进行了数值分析,结果表明旧路面不利用时路面结构的受力状况均好于旧路面利用时路面结构的受力状况。     

Evaluating subcontractor performance using evolutionary fuzzy hybrid neural network

This paper developed an evolutionary fuzzy hybrid neural network (EFHNN) to enhance the effectiveness of assessing subcontractor performance in the construction industry. The developed EFHNN combines neural networks (NN) and high order neural networks (HONN) into a hybrid neural network (HNN), which acts as the major inference engine and operates with alternating linear and non-linear NN layer connections. Fuzzy logic is employed to sandwich the HNN between a fuzzification and defuzzification layer. The authors developed and applied the EFHNN to assess subcontractors performance by fusing HNN, FL and GA. Enhancing subcontractor performance assessments are crucial in terms of providing to general contractors information on historical contractor performance essential to guiding a selection of appropriate subcontractors for a specific current or future subcontracting need. Results show that the proposed EFHNN may be deployed effectively to achieve optimal mapping of input factors and subcontractor performance output. Moreover, the performance of linear and non-linear (high order) neuron layer connectors in the EFHNN was significantly better than performances achieved by previous models that used singular linear NN.     

Application of semi-analytical finite element method to analyze the bearing capacity of asphalt pavements under moving loads

To facilitate long term infrastructure asset management systems, it is necessary to determine the bearing capacity of pavements. Currently it is common to conduct such measurements in a stationary manner, however the evaluation with stationary loading does not correspond to reality a tendency towards continuous and high speed measurements in recent years can be observed. The computational program SAFEM was developed with the objective of evaluating the dynamic response of asphalt under moving loads and is based on a semi-analytic element method. In this research project SAFEM is compared to commercial finite element software ABAQUS and field measurements to verify the computational accuracy. The computational accuracy of SAFEM was found to be high enough to be viable whilst boasting a computational time far shorter than ABAQUS. Thus, SAFEM appears to be a feasible approach to determine the dynamic response of pavements under dynamic loads and is a useful tool for infrastructure administrations to analyze the pavement bearing capacity.     

Multivariate adaptive regression (MARS) and hinged hyperplanes (HHP) for doweled pavement performance modeling

Standard neural networks in infrastructure performance modeling cannot handle discontinuities in the input training data set, and the performance can in some cases be an issue in the presence of higher frequency and higher order non linearity in pavement condition, traffic and other environmental data. This makes the traditional neural network more of a “black box” with limited physical explanation of the results. This paper is a comparative analysis between multivariate adaptive regression and hinged hyperplanes for doweled pavement performance modeling.