Resilient moduli of demolition wastes in geothermal pavements: Experimental testing and ANFIS modelling

https://doi.org/10.1016/j.trgeo.2021.100592Get rights and content

Highlights

  • Thermal conductivity characterization for predominant types of recycled demolition wastes.

  • Deformation and thermal characterization of recycled demolition wastes for geothermal pavements.

  • Intelligent ANFIS-based model for predicting the resilient modulus of recycled demolition wastes.

Abstract

Construction and demolition (C&D) waste materials have been used in a wide range of civil engineering applications, particularly as unbound pavement materials. A comprehensive understanding of the deformation and thermal properties of C&D materials is necessary for their usage in novel applications related to heat transfer in pavement unbound layers, such as geothermal pavements. This research study focused on developing a correlation between the resilient modulus (MR) and thermal conductivity of C&D materials for geothermal pavement applications. The thermal conductivity of C&D materials, namely recycled concrete aggregate (RCA), crushed brick (CB), reclaimed asphalt pavement (RAP), and waste rock (WR), was evaluated at different moisture contents and dry densities. The MR and permanent deformation responses of C&D materials were characterized at the optimum moisture content (OMC), 85%OMC, and 70%OMC, using the repeated load triaxial (RLT) test. An intelligent model was developed for predicting the MR of C&D materials incorporating thermal conductivity, physical properties, confining stress, and deviator stress as input parameters using adaptive neuro-fuzzy inference system (ANFIS) approach. The developed ANFIS model had excellent performance in predicting the MR of C&D materials, with R2 = 0.99 for both training and testing datasets. The ANFIS model was converted into a mathematical relationship, which can be used by researchers and practitioners for estimating the MR of C&D materials.

Introduction

Recycled waste materials are being widely used as sustainable materials in civil engineering applications such as unbound pavement layers [36], [37], [46], [84], pervious concrete [16], [55], [56], [72], [81], geothermal pavement [7], [33], railway capping and ballast [43], [61], [67], [68], and seismic-isolation foundation system [25], [26], [42]. Recycling and reusing waste materials in construction activities reduces the carbon footprint of projects and has multiple environmental and economic benefits [34], [79], [80], [83].

Construction and demolition (C&D) waste materials are typically utilized in pavement base and subbase to provide structural support and uniformly distribute the applied traffic loads to the subgrade [34], [40]. The sustainable use of C&D materials such as recycled concrete aggregate (RCA), crushed brick (CB), reclaimed asphalt pavement (RAP), waste rock (WR), and their combinations with supplementary materials such as glass and plastic for pavement base/subbase layers has been investigated in several studies [22], [32], [34], [36], [37], [46], [83], [87].

Energy geo-structures have received considerable attention in recent decades as sustainable solutions for providing new sources of energy [51], [59]. Geothermal pavement is an innovative and low-cost pavement system, formed by inserting heat exchangers in the pavement unbound layers [7], [33]. The usage of C&D materials for construction of geothermal pavements further enhances the environmental benefits of the system, saves energy, and diverts a substantial amount of wastes from landfills [7]. The circulation of the fluid in pipes installed in the C&D unbound layers extracts the accumulated energy and reduces the pavement surface temperature [7]. Thermal and deformation responses of materials used for construction of geothermal pavements have a significant impact on the long-term performance and efficiency of such systems.

The thermal conductivity of soils, rocks, and pavement materials has been recently investigated in several studies [3], [12], [23], [27], [48], [60]. The divided bar method, introduced by Benfield [15] and Bullard [19], is one of the most reliable and accurate approaches for determining the thermal conductivity of granular materials [6], [14], [17], [27], [33], [48], [70]. Based on this method, thermal conductivity is measured when the sample reaches steady-state thermal equilibrium [14], [33], [47]. However, despite the wide application of C&D materials in transportation infrastructure projects, few studies have investigated the thermal conductivity of these materials, which is useful in thermal and thermo-mechanical analysis of unbound pavement base/subbase layers. Ghorbani et al. [33] have evaluated the thermal conductivity of C&D materials using the divided bar method. Moisture changes in unbound pavement layers due to rainfall, freeze–thaw action, and variations in groundwater level can significantly affect the thermal conductivity of materials and thus the performance of the geothermal pavement systems. Therefore, it is of importance to study the thermal conductivity of C&D materials in different moisture and density levels, which will aid researchers in the thermo-mechanical analysis of unbound pavement base/subbase layers and will enable these materials to be readily used in geothermal pavements.

The unbound pavement materials undergo resilient and permanent deformations due to the applied traffic loading. The deformation performance of unbound pavement materials needs to be characterized in terms of both resilient and permanent deformation responses [53], [78]. The resilient response of pavement materials is typically examined by performing the repeated load triaxial (RLT) test in different stress combinations, while a large number of cycles is applied to the sample for permanent deformation characterization. The gradual accumulation of permanent deformation is an important concern in unbound pavement layers which can result in excessive rutting and structural failure of the pavement system [22], [37]. A comprehensive comparison and understanding of the deformation responses of predominant types of C&D materials are required for their sustainable and wide use as unbound granular materials in transportation infrastructure projects.

The resilient modulus (MR) is currently used by several standards and specifications as a key input parameter for determining the thickness of pavement layers [1]. The MR of unbound granular materials is influenced by several factors such as aggregate type and shape, moisture content, and material gradation [36], [37], [52], [71], [85]. To date, several prediction models have been developed for predicting the MR of pavement materials through regression analysis, to avoid the complicated and expensive procedure for conducting RLT tests. These models relate the MR of unbound granular materials to stress state parameters such as confining and deviator stresses, or a combination of stress state parameters and physical properties [49], [54], [66], [82].

One of the major drawbacks of traditional regression-based models is that they are typically unable to consider the effect of several influential parameters in a single model. In addition, a complicated and time-consuming regression analysis procedure is required to incorporate more input variables in such models. Artificial intelligence methods have been utilized as efficient tools for tackling complicated problems in various fields of geotechnical engineering [28], [36], [37], [62], [63], [65], [69], [75]. However, the application of artificial intelligence methods for predicting the MR of C&D waste materials as unbound pavement materials is limited to date.

In this research, thermal conductivity of C&D materials was evaluated at different moisture contents and dry densities. The permanent deformation behavior and resilient modulus of C&D materials were evaluated at OMC, 85%OMC, 70%OMC, and the results were discussed. An ANFIS model was proposed relating the MR of C&D materials to thermal conductivity, physical properties, confining stress, and deviator stress. The developed ANFIS model was converted to a practical formula for calculating the MR of C&D materials.

Section snippets

Experimental characterization

The predominant types of C&D materials including recycled concrete aggregate (RCA), crushed brick (CB), reclaimed asphalt pavement (RAP), and waste rock (WR) were used in the laboratory characterization stage. C&D materials were collected from recycling facilities across the state of Victoria, Australia. A series of scanning electron microscopy (SEM) images were taken to study the microstructure and surface topography of C&D materials. Fig. 1 presents the physical appearance of utilized

Thermal conductivity

Fig. 4 (a-d) presents the results of thermal conductivity against dry density for C&D materials at various moisture contents. An increase was observed in the thermal conductivity of C&D materials as the dry density increased. An evident increase was also noted in the thermal conductivity with increasing the moisture content. As the moisture increased in the samples, the voids were filled with water and thermal conductivity increased due to the higher conductivity of water compared to air. It

Conclusions

This research study focused on the permanent deformation and MR responses of C&D materials for geothermal pavement applications. The thermal conductivity of C&D materials including recycled concrete aggregate, crushed brick, reclaimed asphalt pavement, and waste rock were measured at different ranges of moisture content and dry density using the divided bar method. The thermal conductivity of C&D materials was found to increase, as the moisture content and the compaction energy increased. The

CRediT authorship contribution statement

Behnam Ghorbani: Conceptualization, Project administration, Methodology, Formal analysis, Software, Data curation, Investigation, Writing - original draft. Arul Arulrajah: Conceptualization, Supervision, Funding acquisition, Writing - review & editing. Guillermo Narsilio: Conceptualization, Funding acquisition, Writing - review & editing. Suksun Horpibulsuk: Conceptualization, Funding acquisition, Writing - review & editing. Melvyn Leong: Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was supported under Australian Research Council’s Linkage Projects funding scheme (project number LP170100072). The second and fourth authors would also like to acknowledge the support from National Science and Technology Development Agency (NSTDA), Thailand under Chair Professor program (P-19-52303).

References (87)

  • B. Ghorbani et al.

    Experimental investigation and modelling the deformation properties of demolition wastes subjected to freeze–thaw cycles using ANN and SVR

    Constr Build Mater

    (2020)
  • B. Ghorbani et al.

    Dynamic characterization of recycled glass-recycled concrete blends using experimental analysis and artificial neural network modeling

    Soil Dyn Earthquake Eng

    (2021)
  • E.M. Golafshani et al.

    Predicting the compressive strength of normal and high-performance concretes using ANN and ANFIS hybridized with Grey Wolf Optimizer

    Constr Build Mater

    (2020)
  • F. Gu et al.

    Characterization and prediction of permanent deformation properties of unbound granular materials for pavement ME design

    Constr Build Mater

    (2017)
  • A. Jessop

    The effect of environment on divided bar measurements

    Tectonophysics

    (1970)
  • D. Kusumawardani et al.

    The influence of aggregate shape properties on aggregate packing in porous asphalt mixture (PAM)

    Constr Build Mater

    (2020)
  • F. Lekarp et al.

    Modelling permanent deformation behaviour of unbound granular materials

    Constr Build Mater

    (1998)
  • J.-X. Lu et al.

    Sustainable design of pervious concrete using waste glass and recycled concrete aggregate

    J Cleaner Prod

    (2019)
  • J.S. McCartney et al.

    Energy geotechnics: Advances in subsurface energy recovery, storage, exchange, and waste management

    Comput Geotech

    (2016)
  • M. Naeini et al.

    Cyclic behavior of semi-rigid recovered plastic blends in railway track substructure

    Transp Geotech

    (2021)
  • B.T. Pham et al.

    A novel approach for classification of soils based on laboratory tests using Adaboost, Tree and ANN modeling

    Transp Geotech

    (2021)
  • S. Rajaei et al.

    Rubberized alkali-activated slag mortar reinforced with polypropylene fibres for application in lightweight thermal insulating materials

    Constr Build Mater

    (2021)
  • A.A. Shahmansouri et al.

    Artificial neural network model to predict the compressive strength of eco-friendly geopolymer concrete incorporating silica fume and natural zeolite

    J Cleaner Prod

    (2021)
  • H. Soliman et al.

    Permanent deformation behavior of unbound granular base materials with varying moisture and fines content

    Transp Geotech

    (2015)
  • A. Tasalloti et al.

    Experimental investigation of the mechanical behaviour of gravel-granulated tyre rubber mixtures

    Constr Build Mater

    (2021)
  • C.S. Vieira et al.

    Use of recycled construction and demolition materials in geotechnical applications: A review

    Resour Conserv Recycl

    (2015)
  • D. Wang et al.

    A mixture of coal wash and fly ash as a pavement substructure material

    Transp Geotech

    (2019)
  • AASHTO. AASHTO guide for design of pavement structures;...
  • AASHTO T307. Determining the resilient modulus of soils and aggregate materials, standard specifications for...
  • A. Antriasian et al.

    Longitudinal heat flow calorimetry: a method for measuring the heat capacity of rock specimens using a divided bar

    Geotech Test J

    (2014)
  • A. Arulrajah et al.

    Thermal performance of geothermal pavements constructed with demolition wastes

    Geomech Energy Environ

    (2021)
  • ASTM D698. Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort (12 400...
  • ASTM D1557. Standard test methods for laboratory compaction characteristics of soil using modified effort. West...
  • ASTM D6913/D6913M. Standard test methods for particle-size distribution (gradation) of soils using sieve analysis;...
  • AustRoads. Determination of permanent deformation and resilient modulus characteristics of unbound granular materials...
  • A. Beck

    A steady state method for the rapid measurement of the thermal conductivity of rocks

    J Sci Instrum

    (1957)
  • Benfield A. Terrestrial heat flow in Great Britain. Proc Royal Soc London Series A Math Phys Sci...
  • Bhutta MAR, Hasanah N, Farhayu N, Hussin MW, bin Md Tahir M, Mirza J. Properties of porous concrete from waste crushed...
  • F. Birch

    Flow of heat in the Front Range, Colorado

    Geol Soc Am Bull

    (1950)
  • BS 812-105.1. Testing aggregates. Methods for determination of particle shape. Flakiness index;...
  • Bullard EC. Heat flow in South Africa. Proc Royal Soc London Series A Math Phys Sci...
  • J. Carlson et al.

    Determining thermal conductivity of paving materials using cylindrical sample geometry

    J Mater Civ Eng

    (2010)
  • CEN EN 13286-7. Unbound and hydraulically bound mixtures-part 7: cyclic load triaxial test for unbound mixtures....
  • Cited by (14)

    • Prediction of UCS and CBR behavior of fiber-reinforced municipal solid waste incinerator bottom ash composites using experimental and machine learning methods

      2023, Construction and Building Materials
      Citation Excerpt :

      Assessing and describing an appropriate link between the selected algorithm and the dataset is crucial. These models have effectively been used to expect the behavior of UCS and CBR values in concrete and soil stabilization-related studies [34–37]. Considering the implications and robustness of the presented models, it can be concluded that the newly created AI-based models, predominantly ANN and ANFIS, can forecast the UCS and CBR values of soils and can be used for other real-time geotechnical engineering problems.

    • Strength and permanent deformation properties of demolition wastes, glass, and plastics stabilized with foamed bitumen for pavement bases

      2022, Construction and Building Materials
      Citation Excerpt :

      RCA has been recognized as an alternative pavement material with superior properties than those of other C&D wastes and natural aggregates. Physical, strength, and stiffness properties of RCA in unbound pavement layers have been investigated in several studies [3,6,11,12]. The presence of adhered cement mortar to the aggregates is the primary contributing factor to enhanced stiffness and strength properties of RCA [8].

    • The prediction analysis of compressive strength and electrical resistivity of environmentally friendly concrete incorporating natural zeolite using artificial neural network

      2022, Construction and Building Materials
      Citation Excerpt :

      Computational intelligence is no longer confined to particular computational areas. In recent years, different CI methods like genetic programming (GP) (e.g., [64–66], evolutionary approaches (e.g., [67–69], neural network (NN) (e.g., [70–72], and fuzzy logic (FL) (e.g., [73–75] have shown promising performance to handle complex concrete strength prediction and modelling. For a comprehensive review about using CI techniques to predict the concrete material characterisation, interested readers refer to Rafiei et al. [76].

    • Geotechnical assessment and large scale direct shear testing on recycled brick aggregates

      2022, Materials Today: Proceedings
      Citation Excerpt :

      It was suggested that the soil-CDW mixture is a reliable material for pavement work. Ghorbani, B. et al. (2021) [18], investigated the correlation between thermal conductivity and resilient modulus of four type of C& D waste materials namely WR, CB, RCA, and RAP. The repeated load triaxial test results were used to develop an intelligent model based on adaptive neuro-fuzzy inference system (ANFIS) approach and a mathematical relationship was proposed to estimate the resilient modulus of C&D waste.

    View all citing articles on Scopus
    View full text