| Author: | Sui, Xin |
| Title: | Quality assurance of flexible and semi-flexible pavement using ground-penetrating radar |
| Advisors: | Leng, Zhen (CEE) |
| Degree: | Ph.D. |
| Year: | 2026 |
| Department: | Department of Civil and Environmental Engineering |
| Pages: | xix, 140 pages : color illustrations |
| Language: | English |
| Abstract: | Quality assurance (QA) is a critical component of pavement construction, ensuring the volumetric properties and surface conditions of paving materials, such as in-situ density, layer thickness, modulus, and smoothness. Traditional QA procedures rely on destructive core sampling and subsequent laboratory testing. This method is detrimental to pavement structures and only provides limited pavement information at discrete locations. Recently, advancements in multidisciplinary testing equipment and methods have led to the increasing adoption of non-destructive testing and evaluation (NDT&E) techniques, such as falling weight deflectometers (FWD), infrared thermography (IR), intelligent compaction (IC), nuclear/non-nuclear gauges, acoustic emissions (AE), lasers, and ground-penetrating radar (GPR). Among these, GPR has emerged as an attractive tool for measuring volumetric properties such as density, thickness, and moisture content (MC) due to its reasonable accuracy, rapid survey speed and extensive coverage. GPR operates by transmitting and receiving electromagnetic (EM) waves to detect variations in the dielectric properties of pavement materials or defects. These dielectric properties, particularly the dielectric constant, can be estimated from GPR signals (A-scans) and correlated with volumetric properties. Despite its potential, GPR has yet to gain full acceptance as a reliable QA tool due to the concern on its reliability and noise interference. The prediction errors of GPR can be caused by either suboptimal prediction models between dielectric properties and volumetric properties or from noise during GPR surveys. For example, material heterogeneity is frequently overlooked when estimating dielectric constants, resulting in inaccurate in-situ density predictions. Additionally, vibration effects caused by irregular antenna movements can distort the true reflection amplitude of GPR signals, making multi-layer thickness profiling challenging. Furthermore, the correlation between GPR attributes and the moisture content (MC) of flexible pavements remains inadequately established, limiting the full potential of GPR for MC monitoring. On the other hand, while QA procedures using GPR for flexible pavements have been well-documented, the novel semi-flexible pavement (SFP), comprising porous asphalt (PA) skeletons filled with cement grouts, still lacks standardized QA protocols despite its growing adoption for its superior rutting resistance. This study aims to fill these gaps by enhancing GPR-based QA accuracy for flexible pavements and proposing a novel QA framework for SFP construction. To improve GPR-based in-situ density predictions of flexible pavements, this study developed an enhanced model by introducing a random pore distribution coefficient into the Al-Qadi-Lahouar-Leng (ALL) density prediction model. This modification mitigates errors caused by the assumption of material homogeneity in the surface reflection (SR) method for dielectric constant estimation. Finite-difference time-domain (FDTD) simulations incorporating mesoscale pore characteristics were used to optimize the coefficient for open-graded PA, gap-graded stone mastic asphalt (SMA), and dense-graded asphalt concrete (AC) mixtures via non-linear curve fitting. Laboratory GPR tests on Superpave gyratory compaction (SGC) specimens demonstrated that the revised ALL model outperformed the conventional version, improving prediction accuracy by 6.7% for PA, 2.7% for SMA, and 1.9% for AC mixtures. To address interference of noise from antenna vibrations during multi-layer thickness profiling, an antenna height shifting correction method was developed. Simulations revealed a third-order polynomial relationship between antenna height and GPR signal reflection amplitude. Based on this, a height-corrected surface reflection (HC-SR) method was proposed and validated through field GPR tests and laboratory measurements of field cores. The HC-SR method improved thickness prediction accuracy by 13.3%, 19.9%, and 17.4% for the first, second, and total asphalt layers, respectively, compared to the traditional SR method. An automatic MC prediction model, the Leng-Sui (LS) model, was developed for wet asphalt mixtures. The model's coefficients were optimized using macro-to-meso-scale EM modelling and non-linear curve fitting. Laboratory tests on SGC specimens verified the LS model's superior performance, achieving an average absolute MC prediction error of 0.2%, outperforming the conventional models, such as the MISO model (1.6%) and the Al-Qadi-Cao-Abufares (ACA) model (0.8%). This approach provides a practical, high-accuracy solution for continuous MC monitoring of asphalt pavements. For SFP construction, this study introduced an automatic method to evaluate the grouting rate, quantified as the volumetric ratio of grout to air voids within the PA matrix. A novel dielectric constant prediction model, the Leng-Wang-Sui (LWS) model, was developed based on EM mixing theory for grouting rate estimation. Sensitivity analyses, numerical simulations of meso-scale GPR-SFP models, and laboratory tests validated the model, achieving an average relative estimation error below 7.0%. A practical QA workflow integrating GPR with the LWS model was proposed for on-site SFP grouting rate assessments. Overall, this study aims to enhance GPR's reliability during the QA procedure of flexible and semi-flexible pavement as a full replacement of traditional coring method. To achieve this, revisions were made upon existing models (e.g., ALL model and SR method) and novel models were established (e.g., LWS model and LS model). Deliverables of this study contribute to the field, and potential forward-looking implications on pavement QA events. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14491