See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/361844331

Applications of Ground Penetrating Radar for the Assessment of Newly Paved

Roads_ Master Thesis & Appendix 2022

Thesis · July 2022

DOI: 10.13140/RG.2.2.17571.53287

CITATIONS READS

0 1,983

1 author:

Ahmed Mahmoud Mohamed Ali

National Research Institute of Astronomy and Geophysics | NRIAG |
3 PUBLICATIONS 1 CITATION

SEE PROFILE

All content following this page was uploaded by Ahmed Mahmoud Mohamed Ali on 08 July 2022.

The user has requested enhancement of the downloaded file.

UNIVERSITY OF SZEGED

MSC THESIS

Ahmed Mahmoud Mohamed Ali

Szeged
2022
 UNIVERSITY OF SZEGED
FACULTY OF SCIENCE AND INFORMATICS
Department of Geoinformatics, Physical and Environmental Geography

MSC THESIS

Applications of Ground Penetrating Radar for the Assessment of

Newly Paved Roads

Submitted by:
Ahmed Mahmoud Mohamed Ali

Supervisor: Dr. György Sipos

Department of Geoinformatics, Physical and Environmental Geography,

Szeged
2022
 Abstract
Pavement condition assessment is essential as quality control to ensure that the pavement was
built correctly, which is critical for transportation infrastructure management. Maintaining the
quality of the transportation infrastructure at a level that provides comfortable, safe, and cost-
effective driving for road users requires frequent diagnostics. Ground-penetrating radar (GPR)
has broadened the scope and increased the certainty of the information obtained from pavement
examinations. GPR technology can continuously and rapidly examine pavement performance
and structure with 100% coverage using different techniques. GPR is a non-destructive
geophysical technology based on electromagnetics that is commonly used in pavement studies
as quality control, relying on the propagation of radar pulses. The study aimed to focus on layer
differentiation, layer thickness mapping, dielectric constant within pavement layers derived
from the amplitudes and time of flight of the subsurface layers, and pavement condition
evaluation derived from GPR data analysis. To employ these objectives, a GPR survey has
been carried out on two recently paved roads in Szeged, Hungary, using an air-coupled horn
antenna of 1 GHz central frequency with deeper resolution and 2 GHz central frequency with
a shallower resolution with their suitable parameters. In addition, the 1 GHz ground-coupled
central frequency antenna is used in a way comparable to the 1 GHz horn antenna. PaveScan
provided essential applications for the paving process, such as poor uniformity and substantial
dielectric changes. Coring was used to calibrate and constrain GPR performance, and its bulk
density data were measured to compare with dielectric permittivity. During pavement
compaction, density data were collected using a nuclear density gauge to monitor changes in
the amplitude of the reflected waves. Several data processing procedures that employ the GPR
reflected signal to estimate the dielectric characteristics of structures studied, hence calculating
their thicknesses, have been developed to increase GPR prediction capabilities. Data
interpretation comprised large data sets provided as longitudinal profiles from a survey of each
road lane. The survey findings highlight that the GPR as a non-destructive technique (NDT)
delivers accurate and reliable findings when evaluating the condition of newly paved roads.

Keywords: Ground Penetrating Radar (GPR); Air-coupled antenna; Ground-coupled

antenna; Amplitude; Thickness; Dielectric permittivity; Core samples; density; Compaction
Evaluation.
 List of Abbreviations

GPR Ground penetrating radar

NDT Non-destructive testing
GSSI Geophysical Survey Systems Inc.
EM Electro-magnetic
TWT Two-way travel time
TOF Time-of-flight
SR Surface reflection
ns Nanoseconds (10-9 seconds)
GHz Gigahertz (109 Hz {1 billion Hz})
RDM Rolling density meter
HMA Hot mix asphalt
εr Dielectric constant (or relative permittivity)
 Contents

Chapter 1 Introduction ............................................................................................................... 1

1.1 Overview and problem statement ................................................................................ 1

1.2 The aims of the study .................................................................................................. 3

1.3 The structure of the thesis ........................................................................................... 3

Chapter 2 Literature review ....................................................................................................... 4

2.1 Pavement Overview ......................................................................................................... 4

2.2 GPR principle: component and physical background ...................................................... 5

2.2.1 Propagation of GPR pulses ........................................................................................ 7

2.3.3 Types of antennas ...................................................................................................... 8

2.3 Application of GPR on pavement assessment.................................................................. 9

2.3.1 Calculations of dielectric permittivity using various methods ................................ 10

2.3.1.1 Calculation of dielectric according to interval time and known thickness ....... 11

2.3.1.2 Calculation of the dielectric from the surface reflection amplitude ................. 11

2.3.2 Estimation of the layer thickness using various methods ........................................ 13

2.3.2.1 Calculation of layer thickness ........................................................................... 13

2.3.3 PaveScan RDM........................................................................................................ 14

2.4 Nuclear density gauge: Density property ....................................................................... 16

Chapter 3 Material and methods .............................................................................................. 17

3.1 Study area ....................................................................................................................... 17

3.2 Methods and data collection and processing .................................................................. 19

3.2.1 Acquisition of data from GPR air-launched antennas ............................................. 20

3.2.1.1 Horn antenna distance mode ............................................................................. 21

3.2.1.2 Horn antenna time mode ................................................................................... 23

 3.2.2 GPR ground coupled antennas data acquisition ...................................................... 25

3.2.3 GPR data processing ................................................................................................ 27

3.2.4 Acquisition of PaveScan RDM 2.0 data .................................................................. 30

3.2.5 Acquisition of nuclear gauge data ........................................................................... 33

3.2.5 Coring extraction and data collection ...................................................................... 34

Chapter 4 Results and discussion ............................................................................................. 38

4.1 Thickness analysis .......................................................................................................... 38

4.1.1 Thickness analysis using air-launched horn of the Vásár road ............................... 38

4.1.2 Thickness analysis using air-launched horn antenna of the Cserje sor road ........... 43

4.1.3 Comparison of the thicknesses of various antennas of the Vásár road ................... 44

4.1.4 Comparison of the thicknesses of various antennas of the Cserje sor road ............. 48

4.2 Dielectric analysis .......................................................................................................... 49

4.2.1 Heat map of the dielectric constant ......................................................................... 49

4.2.2 Dielectric constant comparison of various methods on Vásár road ........................ 53

4.2.3 Dielectric constant comparison of various methods Cserje sor road....................... 55

4.3 Evaluation of the compactness of asphalt layers using various antennas ...................... 56

4.4 Coring data analysis ....................................................................................................... 58

Chapter 5 Conclusion ............................................................................................................... 62

References ................................................................................................................................ 64
Acknowledgement ................................................................................................................... 71
Appendixes .............................................................................................................................. 73

Appendix A .......................................................................................................................... 73

Appendix B .......................................................................................................................... 81

Appendix C .......................................................................................................................... 83

Statement.................................................................................................................................. 84
 List of figures

Fig. 1: Aims of the study............................................................................................................ 3

Fig. 2: Flexible Pavement Layers in Road Construction ........................................................... 5
Fig. 3: Photograph of a GPR (SIR4000) equipped with 400 MHz antenna .............................. 6
Fig. 4: GPR reflections from a basic pavement structure and profile generation ...................... 8
Fig. 5: Reflections from layer interfaces on pavements ........................................................... 13
Fig. 6: The GPR reflections from pavement layers.................................................................. 14
Fig. 7: GSSI Pavescan RDM Components on Cserje Sor Road .............................................. 15
Fig. 8: Nuclear density gauge components and two operation modes ..................................... 16
Fig. 9: Study area location map of Vásár and Cserja sor ......................................................... 17
Fig. 10: Positioning aluminium foils during asphalt construction ........................................... 18
Fig. 11: The research workflow of the connection between objectives and methods ............. 19
Fig. 12: A standard GPR component (SIR 30 unit), GPS and antennas arrangement ............. 20
Fig. 13: Distance between a GPS antenna and each of the horn antennas............................... 21
Fig. 14: Summary of procedures of using GPR with horn antenna and Sir 30 on site. ........... 22
Fig. 15: Field survey of the air-coupled horn antenna measurement, illustrating photos........ 23
Fig. 16: The location of GPR data that was collected during compaction ............................. 24
Fig. 17: Distributions of the measured profile by the 1GHz ground-coupled antenna for Vásár
and Cserje sor roads ................................................................................................................. 26
Fig. 18: GPR system component and data collection of the 1GHz antenna ............................ 27
Fig. 19: GPR data processing steps for horn and ground antennas ………………………….29
Fig. 20: PaveScan RDM Asphalt Density Assessment Tool from GSSI ................................ 31
Fig. 21: Pavescan calibration ................................................................................................... 32
Fig. 22: Pavescan data collection ............................................................................................. 33
Fig. 23: The CPN MC-3 PORTAPROBE® data collection .................................................... 34
Fig. 24: Core extraction in the field site ................................................................................. 35
Fig 25: Measurements on the core sample in the dry condition in the lab ............................. 36
Fig. 26: Measurements of the core sample in the wet condition in the lab.............................. 37
Fig. 27: The thickness of every picked layer of Vásár road ................................................... 39
Fig. 28: Comparison and correlation of the depth of the picked layer of air coupled horn
antenna measurements of Vásár road (2GHz) ........................................................................ 43
Fig. 29: The thickness of piking of every layer of the air-coupled horn antenna and ground
coupled antenna measurements (1 GHz) of Vásár road........................................................... 45
Fig. 30: Comparison and correlation of depth of the air-coupled horn antenna and ground
coupled antenna measurements (1 GHz) and percentage difference of Vásár road ............... 46
Fig. 31: Transverse measurement along Vásár road using SIR 3000 with 1 GHz antenna in
certain parts of Vásár road ....................................................................................................... 47
Fig. 32: The arrangement of channels of (1 and 2 GHz) antennas on the measured road's right
and left lanes ........................................................................................................................... 49
Fig. 33: Horizontal heat maps of dielectric permittivity along Vásár road ............................. 51
Fig. 34: Horizontal heat maps of dielectric permittivity along Cserje sor road ...................... 52
Fig. 35: Dielectric comparison of air-coupled horn antenna antenna (2 GHz) and PaveScan in
the Vásár road, right lane. ....................................................................................................... 54
Fig. 36: Dielectric comparison of air-coupled horn antenna (2 GHz) and Pavescan in the
Cserje sor road, left lane ......................................................................................................... 56
Fig. 37: The change in amplitude, dielectric and density change during the compaction
process in each layer ................................................................................................................ 58
Fig. 38: Comparison of the dielectric and density of multiple methods in dry condition ....... 60
Fig. 39: The measurement of the core samples in dry and wet conditions .............................. 61
 List of tables

Table 1 Range of dielectric constant value of various materials ............................................. 11

Table 2 Parameters of data collection used for 1GHz and 2GHz air-coupled horn antennas . 22
Table 3 Parameters of data collection used for 1GHz ground coupled antenna ..................... 25
Table 4 Data collection parameters used for 1GHz ground coupled antenna on core samples
.................................................................................................................................................. 37
Table 5 The mean thickness measured by air-coupled horn antenna of each profile
corresponding to each layer ..................................................................................................... 40
Table 6 Location of the core samples on the measured GPR profiles ..................................... 40
Table 7 Mean thicknesses of the GPR data measured by measured by air-coupled horn
antenna of the second and first binder courses correspond to core samples ............................ 41
Table 8 Comparison of the percentage difference in dielectric permittivity between the Vásár
and Cserje sor roads ................................................................................................................. 50
Table 9 Percentage difference between air-coupled horn antenna and PaveScan for dielectric
determination of Vásár road..................................................................................................... 54
Table 10 Percentage difference between air-coupled horn antenna and PaveScan for dielectric
determination of Cserje sor road .............................................................................................. 55
 Chapter 1 Introduction

1.1 Overview and problem statement

Indeed, appropriate road pavement evaluation becomes critical to enable safe traffic
movements of citizens and commodities. Furthermore, a reliable transportation network is
critical to economic progress (Rasol et al., 2022). Paved highways are key transportation
network features that must be well designed and preserved regularly to maintain performance
(Khamzin et al., 2017). Compaction is a critical component in the construction of long-lasting
asphalt pavements. Even if all other mixed design requirements are satisfied, insufficient
compaction might result in early irreversible deformation, excessive ageing, and moisture
damage (Wilson & Sebesta, 2018). Reduced in-place density has been linked to a reduction in
structural performance and, conversely, the textural features of the asphalt surface layer (Bell
et al., 1984; Linden et al., 1989; Crispino et al., 2007;). Due to the crucial nature of preserving
these facilities for mobility, economic, and social progress, transportation infrastructure testing
and diagnostics are critical. For this purpose, the accurate and timely assessment of structural
condition problems becomes critical to ensure transportation system safety, time-saving
practices, continuously full coverage, and prevent excessive repair and maintenance expenses.
Optimising the procedures employed for such assessments will result in more accurate
information on the pavement's state (Evans et al., 2006).
Therefore, the application of GPR in the assessment of roadways from the early stage
during paving is essential to enhance customary testing strategies, for example, "visual
examination, coring and inspecting, deflection and dynamic effect checking frameworks"
(Solla et al., 2021). GPR is the only technique capable of providing relevant subsurface
information at speeds comparable to highway speeds (Hall et al., 2001). GPR technology will
enable accurate and stringent quality control to ensure that all road constructions are built to
the required design thicknesses, compaction, and material quality (Pitoňák & Filipovsky,
2016).
This research enlightens the investigating an existing but developing technology (GPR)
while evaluating methods for enhancing the data obtained from it and optimising its usage in
an established engineering field. Unless the techniques are carefully and strategically adapted
to the specifics of each project job, they will not be applied optimally. For the past 30 years,

1
ground-penetrating radar (GPR) technology has been used to examine pavement performance
and structure. Ground penetrating radar (GPR) is a diagnostic instrument used to identify
pavement structure layers and pavement structure flaws.
GPR and other techniques offer a wide range of applications and crucial parameters in
modern pavement engineering as a complete inspection to maintain the pavement's
performance and prolong the pavement's life. Investigating various pavement properties using
both nondestructive (e.g., non-destructive testing, or NDT) and destructive methods can assist
in evaluating pavement conditions. Intrusive procedures can involve the acquisition of core
samples to validate layer thicknesses and the retrieval of material for additional laboratory
investigation. Although intrusive techniques can provide incredibly helpful data on pavement
qualities and condition, they have the disadvantage of being time-consuming to extract material
from the pavement and incurring additional time and money to restore the pavement after
analysis. By using non-intrusive methods whenever possible, pavement properties can be
surveyed without causing damage to the structure and can often be evaluated in a swifter, less
time-consuming manner than intrusive testing, for instance, GPR, nuclear density gauge, and
PaveScan RDM (Evans, 2010). Finally, the information collected during the study is compiled
and used to analyse the state of the pavement to extend its life (Evans et al., 2007).
Since then, obtaining comprehensive information and parameters will require a mix of
methods relevant to pavement investigations. These parameters include the assessment of
pavement layer thickness, dielectric, and the compaction density, voids, and moisture
penetration in pavement systems (Al-Qadi & Lahouar, 2005; Benedetto et al., 2015; Diamanti
& Redman, 2012; Loizos & Plati, 2007; Tosti et al., 2018). Transportation infrastructure can
be affected by defects in either of these elements, reducing its lifespan and affecting the comfort
of automobiles within it. Hence, an early pavement inspection enhances maintenance and repair
procedures, lowering maintenance costs and extending the lifetime of road pavements (Rasol
et al., 2022).

2
1.2 The aims of the study

The thesis's broad objective was to enhance pavement investigation capabilities by

improving the approaches and processes used to gather information from GPR and other
techniques.
The study's specific aims were (Fig. 1):

Fig.1: Aims of the study

1.3 The structure of the thesis

Chapter 1 introduces the context and the problem statement of the study. The research
objectives have been identified; Chapter 2 provides a theoretical framework for the ideas and
technology of GPR. An explanation of the physics of GPR is presented along with a discussion
of the relevance of material dielectric characteristics. Furthermore, the previous literature
corresponding to the objectives of the present study was reviewed; Chapter 3 describes the
research methodology used to achieve the study objectives and includes the research activities.
In addition, by explaining the specifics of the work done to develop techniques for optimising
the GPR data collection; Chapter 4 depicts the result of the study that achieved the desired
goals. Besides, the discussion of these results is presented in different approaches. Chapter 5
demonstrates the conclusion and recommendations regarding this research.

3
 Chapter 2 Literature review

2.1 Pavement Overview

Certainly, paved roads are vital transportation network elements designed to maintain
a sufficient performance level under the circumstances of timely maintenance and preservation.
Characterization of pavement quality is critical to the management of transportation
infrastructure. The success of a road transportation system is closely tied to the quality of the
pavements, which must be stiff and functional throughout their life, with no distress that
impedes traffic flow and security (Pais et al., 2019). Therefore, systematic condition
monitoring is critical for cost-effective pavement system maintenance, management of
maintenance costs, extending the service life of pavements, and lowering driving (Benedetto
& Pensa, 2007; Li et al., 2015).
Paving asphalt, also known as bitumen or tar, is a petroleum substance used as a binder
known an asphalt binder to generate mixes that are subsequently utilised to build driving
surfaces such as highways, runways, and parking lots. The idea is to create a smooth surface
that can resist significant loads without deforming or breaking. In addition, these combinations
also contain aggregate (usually crushed rock, clean gravel, or sometimes crushed glass the
important feature of the aggregate is that it is impermeable and has long-term chemical
stability). One of the primary goals of a good mix and structural design is to ensure that the
pavement can bear traffic loads without deforming or degrading to the point that it is no longer
useable within the design time (Mallick & El-Korchi, 2011). Poor asphalt pavement conditions
and increased traffic frequently result in poor riding quality, tire degradation, and even traffic
accidents. Deficiencies in the pavement surface or structural faults contribute to catastrophic
road accidents, unmanageable maintenance costs, and a significantly reduced service life (Li
et al., 2015).
Generally, the pavement system is made up of multiple layers (Fig. 2), including the
hot mix asphalt layers (HMA), the base course (aggregated dolomite layer) and the sub-base
course (an unbound sand layer). The natural subsoil supports this structure (the subgrade layer)
(Solla et al., 2021). Notably, the “sub-base” is above the subgrade. It is a layer of unbound
compacted aggregate that protects the subgrade and gives the upper layers of the pavement a

4
base on which to build. Certainly, the foundation of the pavement structure is made up of the
sub-base and the sub-grade. The primary structural layer of the pavement, called the “base” (or
“road base”), is built above the sub-base. This base is made of ‘dolomite’, a crushed rock
substance. The base layer is meant to withstand and disperse the loads imposed on the pavement
by cars so that the foundation materials are not destroyed. The pavement surface layer sits on
top of the foundation layer and is commonly used to produce a comfortable and safe driving
surface for automobiles (Evans, 2010).

Fig. 2: Flexible Pavement Layers in Road Construction (Bhushan Mahajan, 2020)

Regarding heated mix asphalt (HMA), it is a very flexible hot substance that may be
poured as a viscous fluid during construction. It is made by combining asphalt binder and
aggregate in varying amounts. This heated mixture is then poured into a paving machine, which
extrudes the flexible material into a uniform layer on the road or area to be covered. Following
a consistent and flat HMA coating, a roller or similar compacting equipment compresses the
flexible, pliable material. As the substance cools, a solid, impermeable surface—the asphalt
pavement—forms (Mallick & El-Korchi, 2011).

2.2 GPR principle: component and physical background

The GPR comprises a control unit, a transmitting antenna, a receiving antenna, and a
sampler/recorder, which samples the reflected signals and converts them into a computer-

5
storage format. In addition, a distance measurement device, usually an odometer wheel, is also
included in many systems (Fig. 3). However, some systems, instead of a wheel, employ GPS
or a total station, while others can use both (Utsi, 2017).

Fig. 3: Photograph of a GPR (SIR4000) equipped with a 400 MHz antenna

The control unit produces short electromagnetic pulses transmitted through a

transmitter into a medium by pulse radar systems. When the pulse hits an electric interface in
the medium, that is, with contrasting properties or significant dielectric contrast, some energy
is reflected to the receiver. In contrast, the remaining is transmitted forward until it approaches
the time window defined in the settings, proving that GPR is an electromagnetic survey
technology often used in the reflection mode. The reflected energy is recorded and shown as a
waveform with amplitudes and the time elapsed between transmission and reflection. The time
it takes for the wave to return to the GPR equipment and the number of reflected signals induced
by differences in material qualities is (Persico, 2014; Cao & Al-Qadi, 2021; Solla et al., 2021a).
Consequently, as the antenna moves over the ground, a two-dimensional picture (known as a
radargram or B-scan) of the observed reflections is created in the form of an XZ graphic
representation. The x-axis shows horizontal distance (antenna movement) along the survey

6
line. The z-axis represents the two-way travel time of the emitted and received pulses or
represents the depth of the reflectors. The depth of the reflector can be computed if the time
required to propagate to a reflector and back is measured, and the velocity of signal propagation
in the medium is known (Solla et al., 2021). The capacity of the GPR to record and display
details on targets in a distinct manner on radar pictures is referred to as resolution (Rasol, 2021).
The reflection coefficient principles determine the amount of reflected radar energy, which is
determined by the contrast in dielectric characteristics of the materials. The reflection
coefficient, 𝛾𝑖 , describes the strength of the reflected field, can be calculated by Eq. (1). When
εr2 > εr1, as in the case of an air-filled vacuum in a dielectric substance, the reflection
coefficient has a positive value (Evans et al., 2006).
√𝜀𝑟,𝑖 − √𝜀𝑟,𝑖+1
𝛾𝑖 = (1)
√𝜀𝑟,2 + √𝜀𝑟1,𝑖+1
εr,i , εr,i+1 - Relative permittivity of the pavement layers i and i+1 respectively,
starting with the top layer.

2.2.1 Propagation of GPR pulses

Figure 4 depicts the transmission of a GPR pulse through the subsurface and its
reflection back to the surface through the receiver antenna. In addition, the figure shows the
reflection at each interface that has a difference in the dielectric constant ε of the pavement
layers and the amplitude strength at these interfaces.
The resultant profile is shown on a greyscale or colour scale, with different shades or
colours corresponding to different magnitudes of reflected amplitudes (Pitoňák & Filipovsky,
2016). A variety of factors can influence the transmission and propagation of GPR signals, but
the electromagnetic characteristics of a material, namely its dielectric permittivity ε, magnetic
permeability μ, and electrical conductivity σ, are particularly essential. Electrical conductivity
describes the movement of free charge (producing electric current) in the presence of an electric
field. Energy is dissipated because of resistance to charge flow. Dielectric permittivity
describes the movement of charge confined within a material structure in the presence of an
electric field. The storage of energy in the substance arises from charge displacement; it is also
the main influence on signal attenuation and penetration depth. Magnetic permeability
describes the response of inherent atomic and molecular magnetic moments to a magnetic field.
It is negligible because it is equal to 1 in non-metallic materials such as the asphalt in this study

7
( Li et al., 2013; Olhoeft, 2003). For GPR examinations on conventional pavement materials,
the dielectric characteristics are the most relevant (asphalt, concrete, granular materials, etc.)
since they influence the wave velocity and reflection coefficient and the resolution of the
acquired GPR data. The GPR signal is therefore weakened as it travels across the asphaltic
pavement. Attenuation rates in dry, intact asphalt are rather modest compared to attenuation
rates in wet, porous asphalt, which are much higher (Barnes & Trottier, 2004; Laurens et al.,
2005). As a result, inter-layer and base layer reflection amplitudes are usually smaller in lesser
quality pavements and larger in better-quality pavements.

Fig. 4: GPR reflections from a basic pavement structure and profile generation. Adapted
from Li et al. (2013)

2.3.3 Types of antennas

Electromagnetic antennas enable GPR to send and receive EM waves by turning EM

waves into current and vice versa. Thus, the voltage corresponding to the current may be used
to measure a GPR signal's amplitude (Evans, 2010). Because the depth of interest underneath
pavements might vary, it is advisable to employ two antenna systems, one with a high-
frequency antenna and the other with a low-frequency antenna (Malvar & Cline, 2002). When
considering the same GPR system, the antenna frequency significantly impacts the signal-
penetrating depth and data resolution: the higher the antenna frequency, the lower the
penetrating depth and the higher the resolution (Maser et al., 2006; Lai et al., 2014). GPR
antennas operate in either a ground-coupled or air-coupled mode to couple electromagnetic

8
waves into the pavement structure under test. The latter is the most used configuration for
pavement surveying.
Air-coupled antennas are mostly horn antennas with a central frequency of 1 or 2 GHz.
According to GSSI (Geophysical Survey Systems Inc.), the depth range of 1GHz and 2GHz
horn antennas is 0.9 and 0.75 m, respectively, and both of them are used with the SIR 30 system
to gather data at highway speeds (GSSI, 2018a). The most significant benefit of air-coupled
systems is their repeatability since antenna coupling is unaffected by changes in asphalt
qualities (Saarenketo, 1997). An additional benefit is that, since they are located above the
pavement, they can gather data at full speed (up to 100 km/h) without disrupting traffic
(Saarenketo, 2006). Hence, this will reduce the time, cost, and risk. Thus, around 1990, the
trend was toward using a horn antenna, which allows for repeated measurements and faster
measurement rates (Saarenketo, 1997).
Ground-coupled antennas cover a wide range of core frequencies, from 80 MHz to 1500
MHz, with the signal transmission of up to 20–30 meters in transportation infrastructure
assessments. These antennas are in touch with the pavement. In fact, ground-coupled systems
have a distinct advantage in signal propagation (Solla et al., 2021), and ground-coupled
antennas are smaller than air-coupled systems (Evans, 2010). Another benefit is the improved
vertical resolution. Ground-coupled systems typically collect data at speeds of 5–30 km/h
(Saarenketo, 2006).

2.3 Application of GPR on pavement assessment

The use of ground-penetrating radar (GPR) in transportation infrastructure studies dates

back to the early and mid-1970s. GPR has been employed in a wide range of civil engineering
applications, including road pavement inspection and assessment for repair and maintenance
(Rasol et al., 2022). Moreover, GPR has the most applicability among geophysical engineering
methods that offer information on the physical characteristics of a place, which may then be
applied to roadway concerns (Anderson & Ismail, 2006). According to (Evans, 2015),
assessing the characteristics and features of newly paved roads is a typical need in many
sections of the built environment. GPR measurements may be used to gather data on the asphalt
structure, such as the material layer (Al-Qadi & Lahouar, 2005; Maser et al., 2006) and
subsurface deficiencies, based on received signals and the two-way travel times (Chen &
9
Scullion, 2008; Saarenketo & Scullion, 1994). GPR's main advantages are gathering data at
high speeds, constantly, using mobile acquisition devices, and in a non-contact mode.
Furthermore, it allows the prospected road infrastructure region to remain operational during
the GPR evaluation, lowering costs and user annoyance (Solla et al., 2021). GPR may be used
to estimate asphalt pavement thicknesses using accessible ground truth data (i.e., core control)
to restrict and validate the interpretation of GPR data (Willett et al., 2006).

2.3.1 Calculations of dielectric permittivity using various methods

In fact, the dielectric constant is the most widely utilized GPR parameter to describe
asphalt mixtures. It can be related to the material composition aggregate, binder, and air-void
content (Fernandes et al., 2017). The dielectric permittivity (or dielectric constant), ε, is a
critical pavement material attribute that describes a medium's ability to allow electromagnetic
energy to pass through it (Maser; Scullion, 1991). This reflection occurred because of the
change of contrast of layers properties that interpret into dielectric constant ε. The larger the
dielectric difference between the two pavement layers, the greater the quantity of reflected
energy (Li et al., 2013). The magnitude of the dielectric constant is proportional to the strength
(amplitude) of the reflected fields (Solla et al., 2021).
Porubiaková & Komačka (2015) mentioned that the dielectric constant is not
constant for road building materials and due to probable heterogeneity of material composition,
can be found in a range (Table 1). The range of dielectric constants for asphalts is vast, and it
is vital to select the proper value when determining the thickness of asphalt layers. As a result,
cores are visibly extracted from the pavement to compute the dielectric constant value, which
could be correlated with other techniques, such as PaveScan or directly from GPR horn
antennas. When asphalt is compacted, the amount of low-dielectricity air in the material is
reduced, resulting in higher volumetric proportions of high-dielectricity bitumen and rock, and
hence higher asphalt dielectricity values (Saarenketo, 1997). The dielectric constant of the
HMA layer is around 4–8, although the dielectric constants of the base and subgrade layers are
often greater than 8. It means that the EM waves will reflect at the interface of two layers,
providing the physical basis for pavement layer identification (Li et al., 2013).
The dielectric characteristics of a pavement material can be calculated directly from
GPR data (for example, by analysing reflected wave amplitudes from air-coupled horn

10
antennas) or indirectly by correlation with other data (such as calibration with core samples)
(Evans et al., 2008).
Table 1 Range of dielectric constant value of various materials (Porubiaková & Komačka,
2015)
Composition of material Dielectric constant value Velocity {mm.ns-1}
Air 1 299
Asphalt material 4-10 90-160
Concrete 5-9 100-130
Aggregate 6-18 70-120
Water 81 33

2.3.1.1 Calculation of dielectric according to interval time and known thickness

The dielectric constant may be calculated by estimating the velocity of a GPR signal
and the two-way travel time (TWT) within a material. Therefore, the velocity is linked to the
dielectric as shown in Eqs. (2) and (3) (Evans & Frost, 2011; Loizos & Plati, 2007; Maser &
Scullion, 1992). This method is called time-of-flight (TOF) (Hoegh et al., 2019).
𝑡
𝑑=𝑣× (2)
2
d - layer thickness [mm]
v - electromagnetic wave velocity [mm.ns-1]
t - two-ways travel time of reflected signal [ns]
𝑐
𝑣= (3)
√ 𝜀𝑟
c - speed of light 299-millimetre per [nanosecond]
√𝜀𝑟 - the dielectric constant of the layer

2.3.1.2 Calculation of the dielectric from the surface reflection amplitude

Calculations of reflection amplitudes from electrical interfaces, such as the air-

pavement contact, are used to determine the dielectric value of pavement. The dielectric
constant is calculated from amplitudes by comparing the reflection amplitudes from an asphalt
surface and a metal plate surface (100 % reflection). A physical equation based on the surface
reflection method was developed and is extensively used to accurately measure the dielectric
constant of the HMA surface layer εHMA, as indicated by Eq. (4) (Diamanti & Redman, 2012;
Hoegh et al., 2019; Leng & Al-Qadi, 2014; Kenneth R Maser & Scullion, 1992).

11
 2
1+(𝐴0 ⁄𝐴𝑝 )
𝜀𝑟1 = [1−(𝐴 ] (4)
0 ⁄𝐴𝑝 )
εr1 - the dielectric value of the surface course layer
𝐴0 - the amplitude of the reflection from the surface course layer
𝐴p - the amplitude of the reflection from a metal plate

The dielectric of the second asphalt layer can be calculated as the following equation
using the value of 𝜀𝑟 computed in Eq. (4) (Al-Qadi & Lahouar, 2005; Li et al., 2013).
2
𝐴0,1 2 𝐴
1−( ) +( 0,2 )
𝐴𝑝 𝐴𝑝
𝜀𝑟2 = 𝜀𝑟1 [ 𝐴0,1 2 𝐴0,2
] (5)
1−( ) −( )
𝐴𝑝 𝐴𝑝

εr2 - the dielectric value of the second binder course layer (layer 2)
𝐴0,1 - the amplitude of the reflection from the second binder course layer (layer 2)

Following the same manner as in Eqs. (4) and (5), the dielectric constant for the nth
layer is then calculated iteratively as shown in Eq. (6) (Al-Qadi & Lahouar, 2005; Li et al.,
2013).
2
𝐴0,1 2 𝐴0,𝑖+1 𝐴0𝑛
1−( ) +∑𝑛−2
𝑖=1 𝛾𝑖 𝐴𝑝 + 𝐴𝑝
𝐴𝑝
𝜀𝑛 = 𝜀𝑛−1 [ 𝐴
2 𝐴0,𝑖+1 𝐴0𝑛
] (6)
1−( 0 ) +∑𝑛−2
𝑖=1 𝛾𝑖 𝐴𝑝 − 𝐴𝑝
𝐴𝑝

γi - the coefficient is the reflection co-efficient at the interface between the two
following layers, n represents the specific layer.

Figure 5 shows an example of an EM reflection from a layered system consisting of

homogeneous and lossless materials (Leng & Al-Qadi, 2014). Consider the d: depth of each
pavement layer. Dielectric permittivity is often computed using reference materials' back-
calculation technique when employing ground-coupled systems. On the other hand, the surface
reflection approach is the most utilized for air-coupled antenna systems (Rasol et al., 2022).
The relationship between the amplitude of the GPR signal reflection and the number of roller
passes could be utilized to monitor the asphalt mixture compaction process and identify the
optimal number of roller passes. The compaction process reduces the amount of air, which has
the lowest dielectric constant of all substances. It, therefore, raises the bulk dielectric constant
of asphaltic material, which is a combination of air, asphalt, and aggregate (Leng et al., 2012).

12
 Fig. 5: Reflections from layer interfaces on pavements (Leng & Al-Qadi, 2014)

2.3.2 Estimation of the layer thickness using various methods

Establishing the thicknesses of flexible and stiff pavement layers is critical for
pavement evaluation and offers valuable data to pavement management systems (PMS).
Accurate estimates of pavement layer thicknesses, for instance, are required for overlay design,
quality control and assurance, and structural capacity estimation of existing pavements to
anticipate their remaining service life (Al-Qadi & Lahouar, 2005). Because it is vital to check
the thickness of the pavement in all areas along the road, a continuous assessment of the
pavement layer is required, which may be accomplished using the measurement of GPR
measurement (Rasol et al., 2022). Air-coupled antennas are strongly recommended (Al-Qadi
& Lahouar, 2004) because they can capture layer thickness data at small spatial intervals and
high speeds (Li et al., 2013).

2.3.2.1 Calculation of layer thickness

The GPR technology employed in this study works by transmitting an EM pulse

through an antenna to the pavement surface and then recording the reflected pulses from the
inside surfaces where there is a contrast in dielectric properties 𝜀𝑟 as shown in Fig. 6 (Al-Qadi
& Lahouar, 2005). The time difference observed between the reflected pulses (i.e., t1 or t2)
may be utilized to calculate the thickness of the surveyed layer in combination with its

13
dielectric characteristics. According to (Cao et al., 2011), because the wave travels back and
forth inside the pavement structure, the layer thickness h is calculated as the product of the
transit time ti and the signal velocity v within the layer.

Fig. 6: The GPR reflections from pavement layers (Al-Qadi & Lahouar, 2005)

From Eq. (2) and (3) mentioned by (Evans & Frost, 2011; Loizos & Plati, 2007; Maser
& Scullion, 1992) in (Section 2.3.1.2), the thickness equation (Eq. 7) will be:
𝑑𝑖= 𝑐×𝑡𝑖 (7)
2√𝜀𝑟,𝑖

Where, 𝑑𝑖 is the ith layer's thickness of the ith layer. 𝑡𝑖 is the EM wave's two-way travel
time of the EM wave through the ith layer (as illustrated in Fig. 6), c is the speed of light in
space (c = 108 m/s or 299 mm/n. s), and 𝜀𝑟.𝑖 is the ith layer's dielectric constant.

2.3.3 PaveScan RDM

The GSSI PaveScan (RDM) is a ground-penetrating radar (GPR) instrument designed

to measure asphalt mixtures' surface dielectric. Stakeholders benefit from the capacity to
complement, and maybe eventually replace, most field coring operations for acceptance in
various ways risks (Wilson, 2017): 1) Large regions can be evaluated rapidly and with little
impact on traffic. 2) Longitudinal dielectric may be evaluated for compliance with the

14
specification. 3) Longitudinal dielectric may be assessed for specification compliance. 4) The
PaveScan RDM ensures life and quality by providing precise real-time measurements.
As indicated in Fig. 7, the air-coupled GPR (GSSI PaveScan system) was configured
with a 2 GHz air-launched GSSI Model 42600 antennas that may function as a transmitter and
a receiver of signals, and a portable computer linked to the cart handles for data collecting
(GSSI, 2019). Hence, this evaluation equipment can detect the dielectric during application
without causing any damage and expose any irregularities that may emerge throughout the
paving process (Joshaghani & Zollinger, 2021). The surface reflection method calculates the
dielectric constant; hence, the dielectric values are affected by the arrival time and amplitude.
The software is calibrated in air and against a metal plate before data collecting (the metal plate
reflects 100 % of the incident waves). Accordingly, the bulk dielectric constant of the surface
is calculated by comparing the amplitude of the reflection from the air with that of the incident
amplitude (represented by the reflection from a metal plate) (Steiner et al., 2020). As the cart
moves across the pavement surface, the software can instantly determine the dielectric constant
for the asphalt surface (Joshaghani & Zollinger, 2021). The three-sensor system has a coverage
range of up to 6 feet across the lane. In conjunction with the optional GPS, the wheel encoder
can properly track the PaveScan's progress on the road.

Fig. 7: GSSI PaveScan RDM Components on Cserje Sor Road

15
2.4 Nuclear density gauge: Density property

Nuclear gauges are among the most precise techniques for determining the density of
soils, aggregates, and other materials (Nuclear Gauges - Humboldt Mfg. Co., n.d.). The nuclear
density gauge is significantly faster, portable, non-destructive, and less expensive than the core
sample approach; however, it has the following drawbacks: (1) usage of a radioactive source
requiring high adjustment and specific training of expert operators, (2) obtaining a permit,
renovating equipment, training technicians, and sophisticated operation (Ziari et al., 2010). The
thicker the pavement, the more likely a gamma-ray will be diverted toward the detector. As a
result, the detector count is proportional to the pavement density. During the study, the nuclear
gauge The CPN MC-3 PORTAPROBE used works by generating radiation (InstroTek, n.d.):
1). The Cesium-137 source delivers gamma radiation into the test sample. Some gamma
radiation will pass through the material and be recorded by the Geiger-Mueller detectors
housed within the CPN MC-3. A low-density material will yield a high count per test period.
A high-density substance will produce a low count during the same period because it absorbs
more gamma radiation (InstroTek, n.d.). Nuclear density gauges are commonly used in two
modes: A. Direct transmission. B) Backscatter, which is used in this study (Fig. 8) (Pavement
Interactive, 2002).

Fig. 8: Nuclear density gauge components and two operation modes (Pavement Interactive,
2002)

16
 Chapter 3 Material and methods

3.1 Study area

Two newly paved roads were chosen in 2021 in Szeged city, Hungary (Fig. 9).

(Cserje sor road) (Vásár road)

Channel 4
(Left lane)

(Left lane)
Channel 3
Channel 2
Channel 1
(Right lane)

Channel 4

Channel 3
Channel 2
Channel 1

(Right lane)
Fig. 9: Study area location map of Vásár and Cserja sor

17
 The two roads, Vásár and Cserja sor, are approximately 234 m long. Both streets are 6
m wide and divided into the right lane (R) and the left lane (L). Both streets were built with the
same asphalt composition. From the bottom up, three layers of asphalt were laid on the sub-
base section of the road. Below the sub-base is the road base, and below the base is the subgrade
layer.
GPR applications have been done to assess the pavement properties, for example,
thickness, dielectric constant, and amplitude analysis during pavement compaction. As
preparation of the site before starting the measurement, aluminium foils with an approximate
length and width of 45 cm and 25 cm, respectively, represent a perfect reflector and aim to
distinguish between the interfaces of each layer (Fig. 10). These aluminium foils were put
every 20 m under the base layer and asphalt layers in the right and left lanes of the road.

A B

C D

Fig. 10: Positioning aluminium foils during asphalt construction. A) Measurements of the
coordinates of the foil using RTK GPS; B) After road base construction; C) Before first
asphalt layer; and D) After first layer construction.

18
3.2 Methods and data collection and processing

Two pavement roads were investigated to demonstrate the applicability of various

techniques for pavement condition assessment. Figure 11 shows how the objectives are linked
to the methods used.

Fig. 11: The research workflow of the connection between objectives and methods

19
3.2.1 Acquisition of data from GPR air-launched antennas

The interpretation of the ground-coupled GPR data was compatible with the findings
of the air-launched GPR (Khamzin et al., 2017). In terms of the GPR system, a simplified
schematic diagram of a GPR system is shown in Fig. 12.

GPS System

SIR 30 unit

DMI (Inside the vehicle)

1 GHz 2 GHz 1 GHz 2 GHz

horn horn horn horn

antenna antenna antenna antenna

Fig. 12: A standard GPR component (SIR 30 unit), GPS and antennas arrangement

The survey was carried out with a GSSI SIR-30 system outfitted with two air-launched
horn antennas with a centre frequency of 1 GHz, and its Model is 42000S, and the others have
a centre frequency of 2 GHz and Model is 41000S. As the pavement depth varies, two centre
frequency antennas were used to cover both the resolution and penetration depth. The SIR 30
is a high-performance multi-channel radar control device for the next generation. A GPS
antenna was mounted on the truck's roof, and encoders or a mechanical distance measuring
instrument (DMI) device was mounted on the survey vehicle's rear wheel. The antenna
arrangement from left to right was 1, 2, 1, and 2 GHz depending on the centre frequency, and

20
each antenna functions as a transmitter and a receiver simultaneously. To adjust the GPS
position with GPR horn antennas, inline and crossline offsets were manually determined at the
start of the survey or edited during GPR data processing using Radan version 7 software from
GSSI. The distance between the GPS antenna and the first 1GHz GPR horn antenna on the left
side of the antenna arrays is 3.3 metres, which is the inline offset. The crossline for each antenna
was measured from the first 1GHz GPR antenna to the other horn antennas (Fig. 13).

Fig. 13: Distance between a GPS antenna and each of the horn antennas

Furthermore, air-launched antennae were mounted to the back of the survey vehicle,
and the antennas were placed 50 cm above the pavement. This height was selected to provide
the best image of the pavement overlay. The effective investigative depth of up to 91 cm under
perfect conditions (as specified by the manufacturer) (GSSI, 2015).

3.2.1.1 Horn antenna distance mode

After the completion of each road layer, the study team used SIR-30 with four horn
antennas to conduct GPR scanning. The left and right lanes were measured four times forward
and backwards, and a validation measurement was performed for both lanes. The initial
measurement was taken following the completion of the base road. The second measurement
was taken after the first asphalt layer was built. The third measurement was taken after the
construction of the second asphalt layer. The fourth measurement was taken after the laying of
the third asphalt layer. All measurements were taken between 19th August and 29th October.
Air-launched GPR studies were conducted in dry weather. The survey vehicle’s average speed

21
was 16 km/h. The GPR data acquisition procedure using a horn antenna and SIR 30 is
summarized in Fig. 14.

Fig. 14: Summary of procedures of using GPR with horn antenna and Sir 30 on site

The recommended parameters for 1GHz and 2GHz antennas are summarised in Table
2.

Table 2 Parameters of data collection used for 1GHz and 2GHz horn antennas (GSSI, 2018a)
Input paramerters SIR 30
Sample/scan 2048
Sample rate 2048
Scan/second 340
Time window • 1GHz antenna: 30 nanoseconds
• 2GHz antenna: 20 nanoseconds
Scan/unit 50
Filter Custom filter on
Gain Auto gain

22
 The GPR system was calibrated by collecting GPR data across a metal plate positioned
below the antennae, and 100–200 scans were acquired (Saarenketo, 2006). Static metal plate
reflections were acquired before data collection to record the reference GPR signal at the
various heights that the antennas may encounter throughout data capture and the related
reflection amplitude values (Khamzin et al., 2017). The gathered data were used to generate an
air-launched antennae calibration file, which was then applied to the acquired GPR data to
adjust for the antennae bouncing wheel during the collection of data and estimate the velocity
of the GPR signal through the pavement (GSSI, 2015a). In this investigation, an air-launched
GPR system was used to examine the condition of asphaltic pavement sections (Fig. 15). To
restrict interpretations of the air-launched GPR, core samples were collected in certain places
(Khamzin et al., 2017).

A B

Fig. 15: Field survey of the air-coupled horn antenna measurement illustrating photos. A)
GPR measurements after constructing the base of the road; B) GPR measurements after
constructing the first asphalt layer.

3.2.1.2 Horn antenna time mode

After asphalt compaction using roller passes, the analysis was aimed to detect changes
in the amplitude of radar waves reflecting from the surface. In general, the density, dielectric
constant, and, in parallel, the amplitude would increase as the number of cylinder passages
increased. As mentioned (Saarenketo, 1997), after pavement compaction, solid asphalts create
a three-phase dispersion system in which the bitumen and stone frame volumes remain constant
while the free voids content decreases after compaction. The dielectric constant of asphalt
increases as the void decreases with each compaction roller. As a result, the difference between

23
the asphalt and the air's dielectric constant rises, causing the amplitude of the radar signal
reflected off the asphalt's surface to rise throughout compaction steadily.
Therefore, road sections were chosen for repeated GPR measurements in time mode
before and after each compaction roller. Three locations were chosen to test the influence of
compaction on amplitude and dielectric permittivity changes. Consequently, two locations
were on Cserje sor street, while one was on Vásár street.

A B

Small roller pass

First binder course

Air-launched horn antennas

C D

Cserje sor jobb

sáv 0+320 km+m

Fig. 16: The location of GPR data that was collected during compaction, with two sections
chosen from the Cserje sor (C & D) road and one from the Vásár (B) road.

24
 Figure 16 displays an air-coupled GPR system made by Geophysical Survey Systems,
Inc., which was used to gather data of the first binder course from the construction site to
investigate the link between the number of compaction passes and the GPR observations during
construction (Leng et al., 2012). The test was carried out after the first, second and third asphalt
layers were constructed.

During asphalt construction, the following time modes were used:

• Before the first compaction roller pass.
• After the first small compaction roller pass.
• After the first vibration roller pass.
• After the second vibration roller pass.
• After the third vibration roller pass.
• After the fourth vibration roller pass.
• After the first large (normal) roller pass.

3.2.2 GPR ground coupled antennas data acquisition

A ground-coupled high (1 GHz, Model 5101) frequency antenna was used in contact
with the asphalt surface to validate the final interpretations. The interpretation of the ground-
coupled GPR data was compatible with the findings of the air-launched GPR (Khamzin et al.,
2017). This antenna was equipped with SIR 3000 data acquisition system manufactured by
GSSI. The parameters of data collection used for the 1GHz antenna were based on the
manufacturer recommendations shown in Table 3. The 1 GHz frequency can investigate up to
100 cm deep and has a lower sample rate than higher frequency antennas (GSSI, 2018a).

Table 3 Parameters of data collection used for 1GHz ground antenna (GSSI, 2018a)
Input parameters SIR 3000
Sample/scan 512
Sample rate 1024
Scan/second 340
Time window 30 nanoseconds
Scan/unit 40

25
 A month later, after road paving, the ground-coupled 1GHz data was collected. 4
longitudinal profiles were measured for both roads (2 profiles for each road) and 12 transverse
profiles on the Vásár road (Fig. 17). The transverse profile started from the left lane of the road
to the right lane with a distance interval of 20 m for each profile.

Fig. 17: Distributions of the measured profile by the 1GHz ground coupled antenna for Vásár
and Cserje sor roads

The survey wheel was calibrated at 20 metres. GSSI recommends that the horizontal
distance and measurement were correct and accurate references for the markings on the data.
After laying the third asphalt layer, GPR measurements were taken using a 1GHz ground
antenna (Fig. 18) to correlate with and confirm the findings with the data from the horn antenna.

26
 Cserje sor Street

Control unit (SIR3000)

Controller box

Vásár Street
1GHz antenna

Fig. 18: GPR system component and data collection of the 1GHz antenna along Vásár road

3.2.3 GPR data processing

GSSI RADAN software version 7 was used to process the GPR data (GSSI, 2015b).
The raw data's processing steps were done to remove noise and improve the signals. Figure 19
demonstrates the flow of processing steps for the air-launched (horn) and ground-coupled
antennas.
The SIR30 GPR system measures the processing steps of horn antennas as mentioned
in the antenna manual (GSSI, 2018a). Firstly, the metal calibration file, which was done before
the acquisition, entered RADAN software, and the metal plate calibration was made. Secondly,
the specific raw file inputted into the software and the inline and crossline values were edited
manually. Thirdly, the reflection picking step was performed. This procedure reduces
undesired signal noise to improve reflections from the pavement layers using the Horn
Calibration.CZT file that was previously prepared. Fourthly, FIR filters have a finite-duration
impulse response. The FIR filter includes 1) The triangular Filter was used. This filter is a
weighted moving average filter that emphasises the filter's centre more than the ends of the
filter. 2) Horizontal and vertical filters: Horizontal represents stacking. It merges the
neighbouring chosen radar scans and delivers a single scan through a simple running average

27
to stack the data. Concerning the vertical filters, the vertical Low Pass Filter removes high-
frequency noise that is frequently detected in somewhat deep radar profiles. It will reject
frequencies that fall below a certain threshold; for the 1GHz antenna, we used 2000 MHz, and
for 2 GHz were used 4000 MHz. Vertical High Pass Filter Reject frequencies that exceed a
predefined threshold, for 1GHZ antenna, we used 250 MHz, and for 2 GHz, we used 500 MHz.
Finally, the range gain step was applied. Range gain leads to changing the gain curve at
different locations and using different ways. The Exponential Gain function was selected
between gain points to add an exponential gain curve. The gain curve is then multiplied by the
scans. The gain was set at 5-7 points with exponential values determined to depend on the case.
Consequently, the processed file is ready for further interactive processing categories;
the GPR data were then split into single-channel recordings for specific processing (four
channels). This analysis, for instance, layer picking to get the thickness, amplitude, and
dielectric values of each layer as CSV output for further analysis; these will have prevailed in
the result chapter.
Some principles were considered during the analysis approaches of the processed
profile. When the lower layer has a higher dielectric value than the top layer, positive (white)
reflections arise, and negative (black) reflections appear when the lower layer has a lower
dielectric value than the upper layer (Saarenketo, 2006). The calibration of a signal technique
is used in this module to measure the flow velocity of the GPR wave through the asphalt
pavement by measuring significant layer interface amplitudes from paved road data. This
calibration allows for the automatic computation of pavement thickness, base thickness, and
other properties without ground truth (core sample). The dielectric permittivity (typically about
5) used to analyse each data set was estimated empirically based on the obtained core control.
The average dielectric permittivity value based on available core control data is 5. The GPR
wave reflections from the various layer interfaces inside the pavement material were visually
recognised. The amplitudes of the layer reflection and apparent thicknesses were determined
(Khamzin et al., 2017). The asphalt dielectric constant was calculated using the time interval
between positive amplitude peaks, showing the thickness of the asphalt layers and the
amplitude values (Porubiaková & Komačka, 2015).

28
Fig. 19: GPR data processing steps for horn and ground antennas

29
 Turning into the processing steps of the ground-coupled antenna measured by the
SIR3000 GPR system. To begin with, it may be required to change the vertical position of the
entire profile in the data window on occasion, "Time Zero Correction". This is because it sets
the top of the scan to a near approximation of the ground surface. Secondly, a background
removal filter was used. Background Removal eliminates horizontal noise bands. Actual
horizontal reflectors often create these bands, but they can also be caused by low-frequency
noise, such as the ringing of the antenna. Thirdly, the FIR filter and, fourthly, range gain were
discussed in the case of the horn antenna. As a result, the processed file becomes ready for
further analysis, which was discussed in the next chapter.

3.2.4 Acquisition of PaveScan RDM 2.0 data

Achieving adequate dielectric values is critical to avoiding early pavement failures. The
GSSI PaveScan RDM monitors dielectric value after each asphalt layer has been built to
identify poor homogeneity and density fluctuations. The dielectric constant values were
measured by driving the GPR van along each test lane of the two roads. The summary of
PaveScan procedures before and during data collection is shown in Fig. 20.
In the study area, the following procedures were performed to assess the dielectricity
of the pavement surface using the surface reflection approach. Firstly, assemble, power on, and
configure the system appropriately. Regarding a three-sensor extended cart configuration, the
left sensor represents serial number 298, the central sensor represents serial number 307, and
the right sensor represents serial number 306. Warming up the sensors will take around 10
minutes, giving enough time to transport the device to the collecting site and add markers along
the pavement collection area to make data gathering easier. Secondly, the system must then be
calibrated by raising all three sensors to a 45-degree angle, gathering an air calibration file
within 10 seconds for each sensor, and positioning the metal plate and an HDPE verification
plate, which is a special plastic material that has a fixed dielectric constant of 2.35 beneath the
sensor(s), with the aid of the laser light, to be calibrated within 5-10 seconds. The purpose of
PaveScan calibration (quality assurance) with these plates ensures that the device will measure
their dielectric accurately before starting with data collection. Following that, the user inputs
the essential data about the gathering on the tablet and aligns the sensors at the start point. It is
now time to start gathering data (Fig. 21).

30
 Fig. 20: PaveScan RDM Asphalt Density Assessment Tool from GSSI (Joshaghani &
Zollinger, 2021)

Thirdly, The PaveScan RDM, via its sensors, transmits and receives high-frequency
electromagnetic waves using RADAR (radio detection and ranging) technology. This signal
hits the asphalt's surface and is reflected as a high-frequency electromagnetic reflection back
to the receiver sensor. The PaveScan RDM sensors detect signals at a frequency of 2.5 GHz.
Fourthly, Variations in asphalt dielectric and density produce variable reflections detected by
the system. The amplitude and timing related to the return of the reflected signals are recorded
by an integrated computer and used to determine the asphalt dielectric values of the pavement.
Finally, while going to the finish point, note any dielectric changes, and mark them for future
reference before storing the data and restarting the collecting procedure (ASPHALTPRO,
2021; GSSI, 2018b; Ziari et al., 2010).

31
 A B

C D

Fig. 21: PaveScan calibration. A) Loosen this thumbscrew, raise the sensor arm, and then
retighten before measuring airwave; B) Sensors positioned for airwave measurement; C)
Metal plate calibration; D) HDPE plate calibration.

As mentioned in the manual of the PaveScan device, the data collection of PaveScan in
the study area compromised of two data collection types (GSSI, 2018b) (Fig. 22),
❖ Field measurement: Using all the associated antennas, collects dielectric
measurements in a continuous file.
❖ Core measurements: consist of two modes, 1) Time mode: Measures dielectric
properties using a sensor centred over a core point. It takes approximately 10 seconds to
complete the measurement. The dielectric shows below the pushed red button at the end of the
test. 2) Distance mode: Measures dielectric properties over a 15 cm distance. The user places
the chosen antenna on one side of the core location, hits the red button on the screen of the

32
device corresponding to the antenna at the core location, and then gently slides the antenna
across the core surface area. Following the traversal of 15 cm, the measuring is automatically
completed, and the core location dielectric appears underneath the pressed red button.

A B

Fig. 22: PaveScan data collection. A) During the file measurement; B) preparing for the
distance measurement.

3.2.5 Acquisition of nuclear gauge data

The density measurement during compaction is studied in this research using a nuclear
density gauge (Kesharwani, 2018). A nuclear density device is a microprocessor-based gadget
that instantly computes values and corrects measurements (Hausman & Buttlar, 2002). A
nuclear density gauge is used to determine the density of bituminous mixtures. It provides the
Inspector and Contractor with a way to collect rapid, precise, nondestructive, and in-place
measurements of pavement layer densities. The instrument measures the density of the
substance to be evaluated using a tiny radioactive source, such as a gamma-ray (Timm et al.,
2013). To avoid the negative health effects of gamma radiation, a person must complete a
Nuclear Safety course (Liao et al., 2006). Before the measurement of the nuclear density
gauge, calibration and standardisation should be done. During the study, the nuclear gauge The
CPN MC-3 PORTAPROBE used, The CPN MC-3 PORTAPROBE® (Fig. 23), works by
generating radiation from two safety-sealed radioactive sources: Cesium-137 and Americium-
241.

33
 A B

Fig. 23: The CPN MC-3 PORTAPROBE® data collection. A) and B) Filed density
measurements during asphalt compaction.

A nuclear density gauge uses gamma radiation to determine the density of in-place hot
mix asphalt (HMA) (Nuclear Gauges - Humboldt Mfg. Co., n.d.). The CPN MC-3
PORTAPROBE® may save up to 200 tests in record numbers that the operator selects. The
data may be examined later on the display screen or transmitted to a computer or printer.
Asphalt pavement compaction in terms of density is measured by backscatter measurement
(InstroTek, n.d.).
Using the Nuclear Density Gauge as data acquisition to measure the asphalt density
through these steps: 1) Turn on the CPN MC-3 gauge to enable the device to warm up before
proceeding with the testing. This step should be done while the testing location is being
prepared. 2) The nuclear device must be positioned flat against the compacted layer to get
accurate readings. 3) Before installing the gadget, the chosen location should be levelled. All
these procedures were performed by a trained specialist.

3.2.5 Coring extraction and data collection

Fauchard et al. (2001), Loizos & Plati (2007), Li et al. (2013), and Benedetto &
Scullion (2000) emphasized the significance of core control and appropriate GPR survey
equipment calibration and provided consistency of GPR-predicted pavement thicknesses as a

34
basis for practical highway survey needs. Asphalt pavement cores were taken from locations
with better conditions and from locations that appeared to be in worse condition. Core samples
were collected and recorded (core diameter, specimen length and a core id) (Fig. 24). Intrusive
data is useful and can calibrate non-intrusive investigations (R. Evans et al., 2007). Despite the
destructive nature of the core method, (Willett et al., 2006) showed how GPR could be used to
estimate asphalt pavement layer thicknesses with accessible ground truth data (i.e., core
control) so that interpretation of the GPR data could be restricted.

A B

Fig. 24: Core extraction in the field site. A) Coring with a water-cooled barrel; B) Extracted
cores from the asphalt layer.

Sixty-two cores were extracted from the study area using the device in Fig. 23. These
cores were distributed along the two roads, with thirty-six cores representing the Vásár road of
the first and the second layers and twenty-six cores representing the Cserje sor road of the first
and the second layers. The core diameter is 15 cm, and its thickness depends on its
corresponding layer. To calculate the dielectric constant using various approaches, such as
surface reflection and time of flight, and to compare the dielectric results from PaveScan and
core samples (Hoegh et al., 2019; Maser & Scullion, 1992). In the beginning, the cores were
extracted from the pavements in service and dried and transported to the laboratory.
Measurements were done to these dry cores through SIR 3000 GPR system after placing the
core sample above the metal plate as an indicator of the base reflection because it represents a
100 % reflector. After that, the core specimen was surrounded by crushed andesite until its
surface to lessen wave diffraction and ensure that the GHz antenna was placed on a flat surface,
and data collection started (Fig. 25).

35
 A B

Fig 25: Measurements on the core sample in the dry condition in the lab. A) Position of the
core sample in a plastic container, surrounded by andesite; B) During data collection on the
core sample in the laboratory.

The weight of each selected item was measured before it was submerged in the water
box to determine the water content. Samples were then immersed in a water-filled box for 48
hours after testing in the dry state to prepare for testing in the wet condition (Fig. 26). To
eliminate any possible effects on the results, it was crucial to assess all core samples at the
same temperature. Each core was removed from the box after being immersed and tested in
this wet state as soon as the excess water was allowed to drain out at room temperature and its
weight had been measured. Hence, eater content could be measured by the difference between
weight in wet and dry conditions divided by dry conditions. The cores were tested so that the
dry and wet conditions could be compared. The parameters used in SIR 3000 GPR system were
attached with a 1GHz antenna as an input, emphasised in Table 4.
After the measurement in both dry and wet conditions, the raw files were entered into
the RADAN processing software as a batch file to measure the dielectric constant through
various approaches. These approaches, for instance, Time of flight of positive and negative
bands and through the surface reflection method. This study's results concerning the dielectric
constant were analysed and correlated in the case of dry and wet conditions. The water content
was the difference between the weight of the core samples in dry and after submerging the
sample in the water box.

36
 A B

C D

Fig. 26: Measurements of the core sample in the wet condition in the lab. A) Submerge the
core sample in the water box and surrounded by andesite; B) the sample weight after drying it
from surface water; C) the initiation of the SIR 3000 with the 1GHz antenna; D) During data
collection on the core sample in the laboratory.

Table 4 Data collection parameters used for 1GHz ground coupled antenna on core samples
(GSSI, 2018a)
Mode of measurement Time
Dielectric 4.59
Scan / unit 2/cm
Position Manual, Offset=28.88, Surface (%) = 3

37
 Chapter 4 Results and discussion

4.1 Thickness analysis

4.1.1 Thickness analysis using the air-launched horn of the Vásár road

Starting with the right lane, the thickness of each single asphalt layer after the pavement
is emphasised, estimated, and measured by the air-launched horn antenna using 2 GHz
equipped with SIR 30 for each layer from top to bottom with the aid of the interface (base or
bottom) of each layer which is represented by horizontal-coloured lines. Figure 27 shows the
thickness of each paved (picked) layer from top to bottom. The horizontal axis represents the
distance (m); it starts from 0 with 2 m as the main interval. While the vertical axis represents
the depth (cm), it starts from zero with 25 cm as a major interval. In Figure 27, each profile
(A, B, C, and D) includes a coloured line representing an interface (bottom of each layer). This
interface represents the base of the upper layer and the surface of the lower layer; consequently,
every two interfaces represent one layer with its thickness. Each layer is defined in each profile
with the specific colour of its interfaces. The yellow circle or oval shape represents the
reflection from aluminium foils. Looking at Fig. 27., the interfaces picked for each profile have
almost the same pattern. Some changes may be because of a shift in the vehicle movement
during measurement. In profile D, the lower interface of the binder course at a horizontal
distance between 12 m and 16 m looks a bit different. This difference may occur because of
the irregular surface of the second base layer (Finely crushed limestone).
Notably, the interfaces in the case of the binder course and the surface course seem to
be almost flat because of evenness during the pavement. However, other layers showed rough
or uneven interfaces, especially the first base layer (Coarsely crushed limestone) and the sub-
grade layer. As highlighted in Table 5, the mean thickness of each layer and profile was
measured by the RADAN software directly using Eq. 7 (See Section 2.3.2.1). The maximum
difference and its corresponding maximum percentage are calculated and show that the
maximum difference is 0.6 in the case of the sub-base and first base layers. The minimum
difference is 0.2 for the second binder course. This slight difference may be due to the shift
that happened during the measurement of each profile.

38
 Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Interfaces between layers Sub-grade layer (Natural soil)

First binder course

Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Second binder course

First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Aluminum foil reflection Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)

Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Fig. 27: The thickness measured by 2 GHz air-launched horn antenna of every picked layer
of Vásár road. A) After placing the second base layer; B) After placement of first binder
course; C) After placement of second.

39
 As mentioned in Willett et al. (2006), ground truth data (i.e., core control) may be used
to restrict and validate the interpretation of GPR measurements. Hence, during the GPR data
processing, some core samples were chosen with their coordinates to know the exact horizontal
location on the GPR profile using RADAN software. After the exact location of the core sample
was known, their thickness was input for the second and first binder courses, and the velocity
method was converted into the core data (Table 6). Hence, the thickness of the binder course
layers could be estimated. Then the thickness could be compared to the core sample thickness,
as presented in Table 7.

Table 5 The mean thickness measured by the air-coupled horn antenna of each profile
corresponding to each layer
Layers Mean thickness (cm) Maximum Max.
Difference (cm) Percentage of
A B C D
difference %
Surface course 4.8
Second binder course 6.2 6 0.2 3.27
First binder course 8 8.2 8.3 0.3 3.68
Second base layer 5.55 5.3 5.4 5.5 0.25 4.6
First base layer 13.5 13.4 14 13.7 0.6 4.37
Sub-base 22.9 23.5 23.2 23 0.6 2.58
A) After placing the second base layer. B) After placement of the first binder course. C) After
placement of the second binder course. D) After placing the surface course layer.

Table 6 The exact location (horizontal distance) of the core samples on the measured GPR
profiles were analyzed as ground truth data of the second and first binder courses
corresponding to core samples
Vásár Road
Layer Lane Core sample Thickness Distance (Position) of core samples on the GPR
(cm) profile (m)
Second binder Right 5.1 39
course
Left 6.1 49.3
Left 6 185.3
First binder Left 8 12.2
course Right 8 57.5
Left 7.8 79.3
Left 8.1 161.4
Right 8.4 132.6

40
 Table 7 The mean thicknesses of the GPR data (using ground truth data) of the second and
first binder courses correspond to the selected core samples
The thickness of the Thickness from the GPR horn
Layer Lane % Of difference
selected core samples using ground truth (2GHz)
Second Right 5.1 5.3 3.92
binder Left 6.1 6.3 3.22
course Left 6 5.8 3.38
Left 8 7.8 2.53

First Right 8 7.9 1.26

binder Left 7.8 8 2.53
course Right 8.1 8.5 4.82
Left 8.4 8.8 4.65

Table 7 summarises and discusses, as an example, the main findings of the comparison
of the thicknesses of GPR data from channel 2 of the horn antenna (2 GHz) of the second and
first binder course using core thickness data as ground truth and the core samples thickness,
including the percentage of difference. The percentage of the difference between the
thicknesses of the binder course layers and the core samples ranges from 1.26 % to 4.82 %,
which is consistent with the previous literature findings of Al-Qadi & Lahouar (2005), which
indicated that thickness errors measured by GPR were 6.7 during hot mix asphalt (HMA).
Overall, the highest percentage of difference occurred in the first binder course, which ranges
from 4.65 % to 4.82 % in both road lanes. This great difference may be because the base of the
first binder course is in contact with the second base layer, which is finely crushed limestone
that might cause a rough surface with different thicknesses in certain positions, while the lowest
percentage of difference was in the case of first binder course of 1.26 % of the right lane.
However, concerning the second binder course, it showed an approximately fixed percentage
of difference along both lanes, ranging from 3.22 % to 3.92 %.
Turning into the analysis of the depth of the complete profile of the Vásár road along
with its horizontal distance of the road through the measurement by SIR 30 GPR system
outfitted with 2 GHz horn antenna. This analysis aims to detect the percentage of error
(difference) of the GPR method and to notice the matching and correlation between the two
channels (2 and 4) of the 2 GHz central frequency antenna.
As shown in Fig. 28, the comparison between the depths of all the pavement layers
measured with channel two and channel four (2 GHz) and the percentage of error. Figure 28

41
(A and B) shows a strong correlation between channels 2 and 4 and almost the same pattern of
interfaces between layers. The interfaces of the surface and binder course layers are almost flat
and become uneven or rough downward. There are few differences in pattern; for example,
from the horizontal distance of 116 to 120 m, this may be because the distance between the two
channels is approximately 1.12 m, as mentioned earlier in Fig. 13 (Section 3.2.1). Talking
about (C), the results demonstrate a good correlation between the thicknesses obtained from
the two channels. The percentage of the difference between them ranges from 1.75 % to 4.47
%, which is consistent with the findings of Infrasense (2006), which suggest that thickness
errors measured by GPR are (2-5%) are generally associated with newly constructed
pavements.

42
Fig. 28: Comparison and correlation of the depth of the picked layer measured by air-coupled
horn antenna measurements of Vásár road (2GHz). A) and B) Comparison of all pavement
layer depths measured with channels 2 and 4 (2 GHz); C) Thickness and percentage
difference correlation.

Turning into the left lane of the Vásár road revealed a good correlation of layers after
thickness picking, and the interfaces of each layer were almost shown the same pattern.
Moreover, the interfaces of the surface and binder course layers seemed to be flat. However,
the layer started to show some irregularities because of the rough surface, especially for the
interfaces of the first base course and the sub-base and the sub-grade layers. The maximum
difference was 0.8 cm in the case of the sub-base, with a percentage of difference of 3.55 %
and first base layers 2.23 %, and the minimum difference was 0.1 cm in the case of the second
binder course with the percentage of difference of 1.6 %. Moreover, it showed a superior
matching when comparing the channels 2 and 4 with minor differences. (See Appendix A,
Figs. A1, and A2).

4.1.2 Thickness analysis using air-launched horn antenna of the Cserje sor road

Concerning the right lane of Cserje sor road, the thickness of every single asphalt layer
after the pavement has been estimated by the same GPR system using a 2 GHz horn antenna,
done in the Vásár road, and highlighted for every layer. The interfaces for each layer were
picked in a coloured form to notice each layer for comparison purposes. This comparison shows
that the picked interfaces of each profile have almost the same pattern with a minor difference.
The largest difference is 1 cm in the case of the sub-base, with a percentage difference of 4.44
%; this difference may be because of the attenuation of signals while travelling downward. In
comparison, the minimum difference is 0.1 cm in the case of the second binder course, with a

43
percentage difference of 1.26 %. Furthermore, when comparing the depths along with the
horizontal distance of the road and the resultant pattern of channels 2 and 4, there were slight
variations; however, it demonstrated greater matching. (See Appendix A, Figs. A3 and A4).

Turning into the left lane of Cserje sor road, comparing each profile following thickness
picking interfaces reveals a great match, with nearly identical patterns. The highest difference
was 0.9 cm in the case of the sub-base, with a percentage difference of 4 %; this discrepancy
could be due to signal attenuation as it travelled downward or maybe a vehicle's shift during
the measurement of profiles. In the second binder course instance, the minimum difference is
0.1 cm, with a 1.6 % percentage difference. Furthermore, it reveals improved matching when
comparing channels 2 and 4 with minor differences; the difference at the beginning of the
profile may be because of the attenuation of the signal or the distance between channels 2 and
4, which is 1.12 m. (See Appendix A, Figs. A5 and A6).

4.1.3 Comparison of the thicknesses of various antennas of the Vásár road

Taking the left lane as an example, Fig. 29 shows the thickness of each layer from top
to bottom with the aid of the interface (base or bottom) of each layer which is represented by
horizontal-coloured lines. The thickness data was measured by the SIR 30 equipped with a 1
GHz horn antenna and SIR 3000 GPR system outfitted with a 1 GHz ground-coupled antenna.
The horizontal axis represents the distance (m); it starts from 0 with 2 m as the major interval
in the horn antenna and 2.5 in the case of the ground antenna. While the vertical axis represents
the depth (cm), it starts from zero, with 25 cm as a major interval for both antennas. Figure 29
shows a strong matching of the picked interfaces with minor differences, especially in the sub-
base and the sub-grade layer, especially in profile A. This difference may be because of the
signal strength in the case of the ground antenna more than in the case of the horn antenna,
which results in deeper penetration with suitable signal energy. Notably, the interfaces between
the binder course and the surface course appear nearly flat; however, the roughness increases
downward. The shallow layers are more distinguishable in the case of horn antenna; however,
deeper layers are more recognized in-ground antenna. Furthermore, every single layer was
presented in each profile and the reflection of aluminium foils was emphasized as yellow
circles. It was noticeable that the flatness of the surfaces of the layers appeared in the shallow

44
layer interfaces of the surface and binder layers, while the roughness of the layers appeared in
the deeper layers, for instance, Sub-base and sub-grade interfaces.

Surface course
Second binder course
First binder course

Base coarse (Finely crushed limestone)

Base coarse (Coarsely crushed limestone)
Aluminum foil reflection
Sub-base Layer (Compacted sand) Aluminum foils reflection

Sub-grade layer (Natural soil)

Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Aluminum foils reflection
Sub-grade layer (Natural soil)
Fig. 29: The thickness of piking of every layer of the air-coupled horn antenna and ground
coupled antenna measurements (1 GHz) of Vásár road. A) Depth profile obtained using a 1
GHz horn antenna. B) Ground antenna depth profile measured at 1 GHz.

Figure 30 (A and B) compares the depths of all pavement layers measured with ground
antennas and horn antennas (1 GHz). This comparison compares the paved layers, notices any
differences in each layer's depths and interface pattern, and correlates between each layer.
Figure 30 shows a strong correlation and almost the same pattern of interfaces between layers.
The surface and binder course layers are almost flat and become uneven downward. There were
a few differences in the pattern, for example, from horizontal shift during measurement and the
attenuation of the signal during travel beneath the surface. Turning into (C), the results
demonstrate a strong correlation between the thickness obtained from the two channels. The
difference between them ranges from 1.82 % to 3.72 %, and these findings are in line with
those found in the literature, as mentioned before. The maximum difference is 0.84 cm in the
case of the sub-base, with a percentage difference of 3.72, and the minimum difference is 0.1
cm in the case of the second binder course, with a percentage difference of 2.2%.

45
 Fig. 30: Comparison and correlation of depth of the horn and ground coupled antenna
measurements (1 GHz) and percentage difference of Vásár road. A) and B) Comparison
between depths (cm) of all the pavement layers measured with the horn and ground antenna,
respectively (1 GHz). C) Correlation of the thicknesses with the percentage of difference (%).

46
 Interestingly, the results reveal that the percentage difference in the case of binder
course layers, which were measured by a 1 GHz antenna, is greater than in the case of
measurements made by a 2 GHz antenna. This increase in the percentage of difference may be
because the higher the frequency, the greater the resolution but, the shallower. The binder
course is at a shallow depth, so it was detected well in the case of the 2 GHz antenna. Figure
31 indicates the transverse measurement on the Vásár road as an example. Measurements were
made every 20 m and started from the forward direction, as presented in Fig.17 (Section 3.2.2).
The following profile shows the measurement of the beginning, middle, and end.

6 6 6
5 5 5
4 4
4
3 3
3

2 2
2

1
1 1

Sub-grade
Sub-grade Sub-grade

6 6 6
5 5 5
4 4
4
3 3
3
2
2
2
1 1
1

Sub-grade Sub-grade
Sub-grade

Fig. 31: Transverse measurement along Vásár road using SIR 3000 with 1 GHz ground
coupled antenna in certain parts of Vásár road. (A) and (B) measured at the beginning of
the road. (C) and (D) measured in the middle of the road. (E) and (F) measured at the end of
the road.
Note: Concerning the pavement layers from top to bottom, 6: represents the surface course. 5: represents
the second binder course. 4: represents the first binder course. 3: represents a second base. 2: represents
the first base. 1: represents sub-base. The sub-grade layer at the down most.

47
 These profiles can indicate the efficiency of GPR in assessing and highlighting the
thickness and interfaces of each layer accurately almost in the same manner as in longitudinal
profiles as mentioned earlier. Hence, the road can be assessed precisely in longitudinal and
transverse configurations for better results.

4.1.4 Comparison of the thicknesses of various antennas of the Cserje sor road

Taking the left lane as an example of longitudinal measurement, it can be noticed a

strong matching in the pattern of the picked interfaces with small differences, particularly in
deeper value but less than in the case of Vásár road. One of the most important findings relates
to signal penetration strength. The ground antenna shows a strong signal, especially in deeper
penetration rather than in shallower depth, translating to better resolution in deeper depth other
than shallower depth. In contrast, the horn antenna shows a good resolution in shallower depth,
for example, binder and surface (wearing) course layer. The signal weakens downward, for
example, sub-base and sub-grade layer, due to attenuation of the signal. Surprisingly, the
incoming reflection from aluminium foils is obvious in the case of the ground antenna and
repeated every 20 m interval; this may be because of a strong impulse signal. The thickness
profile of the ground antenna contains coherent noise more than the horn antenna (See
appendix A, Fig.A7).

The comparison between depths values of ground and horn antennas (see Appendix A,
Figs. A8, A and B) shows a strong correlation, particularly in shallow layers, with minor
irregularities in sub-base and subgrade layers. Figure (A8, C), the maximum difference
percentage was in the sub-grade layer, which is 3.08 %, and the lowest is in the first binder
course, which is 1.23 %. Furthermore, Figure (C) reveals that the percentage of difference in
the case of the Cserje sor road showed lower values than in the Vásár road. Overall, the
thickness results in the case of horn and ground antennas show great improvement compared
to many previous studies.

48
4.2 Dielectric analysis

4.2.1 Heat map of the dielectric constant

The dielectric was measured using the SIR 30 GPR system with 1GHz and 2GHz
antennas along both lanes of the road to scan the dielectric of the entire road. The dielectric
heat map aims to analyse the change in amplitude, and dielectric values along the road can help
interpret road quality. The data was collected for the whole road using four antennas. The first
and third channels used 1 GHz antennas, while the second and fourth used 2 GHz antennas
(Fig. 32). The dielectric constant was obtained through RADAN software. After that, the
horizontal distance, the intervals from Fig. 32 for each antenna (1 GHz and 2 GHz), and the
dielectric value for each channel were put in Surfer software, and Kriging interpolation was
used to prepare the dielectric heat map. A 3D cube is made as follows, each layer from top to
bottom. Figs. 33, and 34 show the dielectric heat map for both lanes of the Vásár and Cserje
sor roads using 1 GHz, which represents by A and 2 GHz antennas, which represent by B. Each
layer's description is placed under its heat map.

Fig. 32: The arrangement of channels of (1 and 2 GHz) antennas on the

measured road's right and left lanes, as well as the lateral distance (intervals)
between them. A) Channels 2 and 4; B) Channels 1 and 3.

Regarding Vásár road, Fig. 33 provided the dielectric permittivity map for the right and
the left lane. It can be noted that the overall values of the dielectric permittivity of the 2 GHz
antenna were higher than those of the 1 GHz antenna. For the shallower layers, which have
almost the same components, for example, binder and surface (wearing) course, it can be
noticed that the dielectric constant in the case of surface course is greater than those of binder
course; this might be because of the difference in particle size that influences its density.

49
 According to particle size, the binder course particle size was small and well-sorted, so
it contained bituminous material rather than air. Hence, during the compaction process, the
surface course was well compacted or high density, leading to high amplitude and dielectric
constant. In contrast, in the case of the binder course, between the larger particle size, it
contained more air with a lower dielectric constant value. Therefore, during compaction, the
binder course was poorly compacted, leading to a lower density and a lower dielectric constant.
The base course layers witness a higher dielectric constant because limestone material contains
a higher dielectric. The sub-grade layer is compacted sand with a lower dielectric constant.
Overall, the dielectric constant is in the range of each dielectric permittivity for each layer, as
mentioned in (Porubiaková & Komačka, 2015) Table 1, which interprets the good quality of
the road. Expectedly, the dielectric constant values of the binder and surface course in the case
of a 2 GHz antenna in the centre of the road are higher than those of the road's edges. This may
be because of the difficulty of the roller passes to compact the edges with the same quality as
in the centre. Turning into Cserje sor road (Fig. 34), it followed almost the same pattern as
Vásár road; the surface course shows higher dielectric values than the binder course, especially
in A. In addition, in the case of the base course, especially the first base course made of
limestone has the greatest dielectric values, and the subgrade compacted sand has the lowest
dielectric. Overall, the dielectric values of a 2 GHz antenna are much greater than in the case
of a 1 GHz antenna. Table 8 summarizes the mean (average) dialectic permittivity εr of 1 and
2 GHz antennas of all layers along the Vásár and Cserje sor roads. Indeed, using different
antennas produces different amplitudes, hence different dielectric permittivity.

Table 8 Comparison of the percentage difference between the Vásár and Cserje sor roads' 1
and 2 GHz antennas for all layers
Vásár road mean εr % Difference Cserje sor road mean εr % Difference
Layers between 1 between 1
1 GHz 2 GHz and 2 GHz 1 GHz 2 GHz and 2 GHz

Surface course 5.4 5.8 5.3 6.1 6.2 1.6

Second binder course 5.1 5.4 3.77 5.2 5.46 3
First binder course 5.2 4.9 3.9 5.9 5.35 1.83
Second base course 5.4 5.9 5.2 5.5 5.7 3.57
First base course 6.6 7 3.88 6.68 7.2 4.98
Sub-base 4.5 4.7 4.3 4.55 4.6 1.09

50
 εr of the surface course at a depth ~ 0 to 5 cm
εr of the surface course at a depth ~ 0 to 5 cm

εr of the second binder course at a depth ~ 11 to 12 cm εr of the second binder course at a depth ~ 11 to 12 cm

εr of the first binder course at a depth ~ 19 to 22 cm εr of the first binder course at a depth ~ 19 to 22 cm

εr of the second base course at a depth ~ 25 to 32 cm εr of the second base course at a depth ~ 25 to 32 cm

εr of the first base course at a depth ~ 35 to 45 cm εr the first base course at a depth ~ 25 to 32 cm

εr of the sub-grade layer at a depth ~ 55 to 65 cm εr of the sub-grade layer at a depth ~ 55 to 65 cm

Fig. 33: Horizontal heat maps of dielectric permittivity of different antenna along Vásár road. A)
Dielectric maps using 2 GHz antennas; B) Dielectric maps using 1 GHz antennas.

51
 εr of the surface course at a depth ~ 0 to 5 cm
εr of the surface course at a depth ~ 0 to 5 cm

εr of the second binder course at a depth ~ 10 to 12 cm εr of the second binder course at a depth ~ 10 to 12 cm

εr of the first binder course at a depth ~ 20 to 22 cm

εr of the first binder course at a depth ~ 20 to 22 cm

εr of the second base course at a depth ~ 25 to 32 cm

εr of the second base course at a depth ~ 25 to 32 cm

εr of the first base course at a depth ~ 36 to 45 cm εr of the first base course at a depth ~ 36 to 45 cm

εr of the sub-grade layer at a depth ~ 55 to 65 cm εr of the sub-grade layer at a depth ~ 55 to 65 cm

Fig. 34: Horizontal heat maps of dielectric permittivity of different antenna along Cserje sor road. A)
Dielectric maps using 2 GHz antennas; B) Dielectric maps using 1 GHz antennas.

52
 The aim is to reveal the percentage difference between the two antennas during the
same area and condition measurement. It can be noted that the range of percentage of difference
of Vásár road between the 1 and 2 GHz antenna is from 3.77 % to 5.3 %, respectively and the
lowest percentage in case of second binder course and the greatest percentage in case of the
second base course. Regarding Cserje sor road, the percentage of difference range between the
1 and 2 GHz antenna is from 1.09 % to 4.98 %, respectively. Overall, the percentage of
difference in the case of Cserje sor road is lower than that of Vásár road.

4.2.2 Dielectric constant comparison of various methods on Vásár road

Starting with the measurements of the right lane, a 2 GHz horn antenna and PaveScan
were used to measure the surface dielectric constant. This comparison between the two methods
aims to correlate and compare the percentage error between them on each layer of the binder
course and surface layer. From Fig. 35, it can be noted that there is good matching with some
differences between the two methods in determining the dielectric permittivity along the road.
This difference might be because the two methods' measurement line is not the same but still
gives a great matching. Table 9 present the percentage of the difference between the two
methods for dielectric determination. The best matching (lowest percentage of error, 4.11 %)
is in the case of the surface course, and the lowest percentage error is in the case of the first
binder course, 1.7 %.

Turning into the measurements of the left lane, again, it has a good correlation between
both methods with minor differences (See Appendix B, Fig. B1). The percentage of the
difference between the two methods is presented, showing that the greatest percentage
difference is in the case of the first binder course, 3.93 %, and the lowest percentage difference
is in the case of the surface course, 1.59 %, which consider the best matching (See Appendix
B, Table B1). Overall, the dielectric measured by the PaveScan is slightly higher than the horn
antenna.

53
Table 9 The percentage difference between the two methods for dielectric determination
Horn antenna (2 GHz) PaveScan % Difference
Surface course 5.9 5.8 1.7
Second binder course 5.7 5.56 2.48
First binder course 5.48 5.71 4.11

Fig. 35: Dielectric comparison of air-coupled horn antenna (2 GHz) and PaveScan in the
Vásár road, right lane. A) Surface course; B) Second binder course; C) First binder course.

54
4.2.3 Dielectric constant comparison of various methods Cserje sor road

Starting with the left lane, Fig. 36 shows a good match with minor discrepancies
between the two approaches for estimating the dielectric permittivity along the road. The
percentage difference between the two dielectric determination methods is shown in Table 10.
Notably, the percentage of difference in the case of Cserje sor is greater than in the case of
Vásár road. The percentage of the difference between the two methods shows that the greatest
percentage difference is in the case of the first binder course, 4.4 %, and the lowest percentage
difference is in the case of the surface course, 2.9 %, which is considered the best matching.

Table 10 The percentage difference between the two methods for dielectric determination
Horn antenna (2 GHz) PaveScan Percentage of difference
Surface course 5.2 5.31 2.9
Second binder course 5.51 5.7 3.38
First binder course 5.55 5.8 4.4

Turning into the measurement of the right lane indicates a great correlation, especially
in the surface course, with some differences during going deeper (See Appendix B, Fig. B2).
The data regarding the percentage difference is presented and shows that the best matching is
in the case of the surface (wearing) course with a percentage difference of 0.35 %; however,
the greatest percentage difference is in the case of the first binder course with 4.84% (See
Appendix B, Table B2). Overall, as with Vásár road, the dielectric permittivity values
measured by PaveScan in the case of Cserje sor road are slightly greater than the horn antenna
measurement.

55
 Fig. 36: Dielectric comparison of air-coupled horn antenna (2 GHz) and Pavescan in the
Cserje sor road, left lane. A) Surface course; B) Second binder course; C) First binder course.

4.3 Evaluation of the compactness of asphalt layers using various antennas

This analysis aims to monitor the amplitude, dielectric permittivity, and density change
during the compaction process of the Vásár road, as an example. The amplitude of the reflected
signal from the GPR system using 1 and 2 GHz antennas increases as compaction increases
due to the density increase until the specific point at which it becomes almost constant. Fig. 37
shows that the amplitude of the reflected signals increased as the increasing roller passed during
the compaction of layers. For example, For A, B, and C, the amplitude continues increasing till
the number of passes is 6. So, after the number of roller passes is 6, there would be any
significant change in the amplitude of the reflected signal. For D, there is a direct relationship
between the amplitude and dielectric permittivity; hence, it takes the same manner as
amplitude. For E, the density values measured by the nuclear gauge continue increasing until
the number of roller passes is 6. That was a great correlation with the amplitude analysis from
GPR data, which indicates a good correlation between the GPR and Nuclear gauge methods.

56
The number of roller pass 6, at which the amplitude, dielectric, and density become almost
fixed, could be taken as a reference for other studies, and the results are consistent with the
findings of Leng et al. (2012).

57
 Fig. 37: The change in amplitude, dielectric and density change during the compaction
process in each layer. A) Represent the surface course layer; B) Second binder course; C)
First binder course; D) Measurements of dielectric permittivity of first binder course; E)
Measurements of Nuclear gauge density of second binder course.

4.4 Coring data analysis

Multiple methods were used to determine the dielectric permittivity εr of core samples
in dry and wet conditions (after submerging in a water box). Those approaches to the core
sample include the determination of the dielectric permittivity of the dry samples by positive
and negative surface reflection (SR) methods and positive and negative time of flight methods
(TOF). Those two approaches were discussed in (Section 2.3.1) and measured by SIR 3000
GPR system and GHz ground antenna. The aim was to compare the dielectric values of the
horn antenna with that from PaveScan. Moreover, compare the dielectric of horn antenna with
that of Positive and negative surface reflection (SR) and positive and negative time of flight or

58
time intervals. Consequently, better correlation and matching could be noticed from these
techniques.

As shown in Fig. 38, there were comparisons between the dielectric permittivity εr of
multiple approaches represented by A, B, C, D, and E. in addition; there was a comparison
between horn antenna dielectric and the bulk density of the core samples represented by F.

Starting with A, it provides a quantitative analysis of the results of this comparison.

Notably, there is a strong matching between the two methods. Turning into B, it shows a high-
level correlation between the two methods. There is a relatively good correlation between the
two methods concerning C, D, and E. Finally, F highlights the strong direct relationship
between dielectric from horn antenna and bulk density of the core sample. This argument is
consistent with the findings of (Sebesta et al., 2012), which showed the same direct relationship
between the two approaches. Overall, the comparison between the dielectric permittivity
between the horn antenna (2 GHz) as a field measurement and the PaveScan as a lab
measurement shows the best fit; however, D presents the lowest correlation between all of the
other comparisons.

Turning into the wet measurements, six-core samples were selected, three samples for
the first binder course and three samples for the second binder course of Vásár and Cserje sor
road. This method aims to study the effect of the water on the positive SR dielectric permittivity
of core samples, the measurement was done after taking off the core samples from a water box
with surface drying (2 seconds).

Figure 39 reveals a relationship between the dielectric permittivity in the dry and wet
conditions of six different core samples along the two roads to notice the resultant measurement
of dielectric in the wet condition and compare it with the dielectric permittivity of the dry
condition of each core sample. Each diagram shows a clear trend of increasing the dielectric
permittivity on each increasing period. For instance, in A, it can notice that the dielectric
increase after 22 hours of submerging in water from 6.3 (a dry measurement which is fixed
during the comparison) to 6.7 (Wet measurement), and after two days, the dielectric increase
to 7.3.

59
Fig. 38: Comparison of the dielectric and density of multiple methods in dry condition

60
 Fig. 39: The measurement of the core samples in dry and wet conditions (after submerging
the core samples in a water box). A), B), and C) are measurements from the second binder
course; D), E), and F) are measurements from the first binder course.

However, in the period of 96 and 168 hours after submerging in water, the value of the
dielectric permittivity tends to be unchanged in their values; this may be because there are no
extra voids to be filled with extra water during this period. For B, C, D, E, and F, it shows the
same manner as in A's case, and the measurement of the dielectric permittivity in wet conditions
tends to be constant during the same period (from 96 – 168 hours). The water content of each
core sample also shows an increase in the weight till the period (from 96-168 hours) that shows
almost no difference in weight (See appendix C, Fig. C1). The effect of water content on
increasing the amplitude of the core sample is consistent with the findings of Evans (2010).

61
 Chapter 5 Conclusion
The pavement condition assessment is an essential step in paving construction to
confirm that the pavement was constructed correctly, which is important for transportation
infrastructure management. It is imperative to maintain the quality of the transportation
infrastructure so that road users can commute comfortably, safely, and cost-effectively. This
study aims to delineate the thickness and interfaces of each paved layer, estimate the dielectric
permittivity of the pavement layers and the core samples, and monitor the amplitude change
during the pavement compaction process. For comparison purposes, multiple approaches
measured all these research points, for instance, GPR, PaveScan, and nuclear density gauge.

The thickness analysis shows a great fit and correlation in the case of the air-launched
horn coupled antenna and ground coupled antennas with minor differences. Concerning the
case of air-launched horn antenna, the percentage difference for the 2 GHz antenna depth
measurement during layer pavement ranges from 1.6 to 4.82% on Vásár road and 1.26 to 4%
on Cserje sor road. The percentage of different ranges between 1 GHz air-launched horn and
ground antennas ranges from 1.82 to 3.72% in Vásár and 1.23 to 3.08% in Cserje sor road.
According to the results of the comparison of the GPR data of the air-launched horn antenna
(2GHz) using the core thickness data as ground truth and the core samples as thickness, there
is a difference in thickness of 1.26 to 4.82 % between the binder course layers and the core
samples. Overall, horn antennas, especially the 2 GHz antenna, which is recommended in this
study, show high resolution in the shallower part, and the 1 GHz ground antenna shows the
highest signal strength that is enough for deeper penetration. Hence, the thickness data from
air-launched antennas can replace that of ground antennas, especially if a quick acquisition is
needed to reduce time, cost, and risk from the passing vehicles. It is possible to measure the
thickness of the road layers using GPR data (especially using the 2GHz horn antenna) instead
of coring to reduce the destructive methods involved in coring.

The amplitude analysis demonstrates that the overall values of dielectric permittivity
from the 2 GHz antenna were higher than those of the 1 GHz antenna. In addition, the dielectric
constant in the case of a surface course is greater than that of a binder course. Moreover, the
base course layers have a higher dielectric constant because of the limestone material that

62
contains a higher dielectric constant than asphalt materials and the sub-grade layer, which is
compacted sand with a lower dielectric constant. It can be noted that the range of the percentage
difference of Vásár road between the 1 and 2 GHz antennas is from 3.77 % to 5.3 %,
respectively, while the percentage difference of Cserje sor road ranges between the 1 and 2
GHz antennas is from 1.09 % to 4.98 %, respectively. The findings from the comparison of the
dielectric permittivity using various methods (2 GHz air-launched horn antenna and the
PaveScan) show a great fit and match between them, and the percentage difference between
them ranges from 1.59 to 3.93 % for Vásár road and 0.35 to 4.84 % for Cserje sor road.
Moreover, the results of amplitude analysis during the compaction process show that the
amplitude of the reflected signal from the GPR system and the dielectric permittivity using 1
and 2 GHz antennas increases as compaction increases due to the density increase until the
specific point at which it becomes almost constant (6 roller passes). Consequently, the air-
launched horn antennas (especially 2 GHz) can be more reliable than the dielectric permittivity
data from the PaveScan if a quick acquisition is needed to reduce time, cost, and risk from the
passing vehicles.

The coring data analysis was studied for comparison purposes. Multiple methods were
used to determine the dielectric permittivity εr of core samples in dry and wet conditions (after
submerging in a water box). For the dry measurements, the comparison between the dielectric
permittivity of the PaveScan and the horn antenna (2 GHz) again shows a strong fit. After that,
other comparisons show a good correlation, for example, comparing the dielectric permittivity
of air-launched horn antenna with the dielectric from the positive surface reflection method,
the negative surface reflection method, and positive and negative time-of-flight methods. There
is a strong direct relationship between the dielectric from the horn antenna and the bulk density
of the core sample. For the wet measurement, quantitative analysis of the periodic wet
measurements shows an increase in the measured dielectric permittivity of the core samples
after submerging the core samples in water until the period of (69–168 hours). Overall, the
water content can affect (increase the values) the amplitude and the dielectric permittivity
measurements until the saturation stage, and the saturation stage for the core samples is reached
approximately three days after submerging of the core sample in water.

63
 References

Al-Qadi, I. L., & Lahouar, S. (2004). Use of GPR for thickness measurement and quality
control of flexible pavements. Asphalt Paving Technology, AAPT, 501–528.
https://www.academia.edu/download/63479217/04-GPR_220200531-64946-
1ahyh1r.pdf
Al-Qadi, I. L., & Lahouar, S. (2005). Measuring layer thicknesses with GPR - Theory to
practice. Construction and Building Materials, 19(10), 763–772.
https://doi.org/10.1016/j.conbuildmat.2005.06.005
Anderson, N., & Ismail, A. (2006). A generalized protocol for selecting appropriate
geophysical techniques. Geology and Geophysics, 19.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.461.5335&rep=rep1&type=
pdf
ASPHALTPRO. (2021). GSSI’s PaveScan RDM Asphalt Density Assessment Tool.
AsphaltPro Magazine. https://theasphaltpro.com/articles/gssis-pavescan-rdm-asphalt-
density-assessment-tool/
Barnes, C. L., & Trottier, J.-F. (2004). Effectiveness of Ground Penetrating Radar in Predicting
Deck Repair Quantities. Journal of Infrastructure Systems, 10(2), 69–76.
https://doi.org/10.1061/(asce)1076-0342(2004)10:2(69)
Bell, C. A., Hicks, R. G., & Wilson, J. E. (1984). EFFECT OF PERCENT COMPACTION
ON ASPHALT MIXTURE LIFE. ASTM Special Technical Publication, 107–130.
https://doi.org/10.1520/stp32504s
Benedetto, A., & Pensa, S. (2007). Indirect diagnosis of pavement structural damages using
surface GPR reflection techniques. Journal of Applied Geophysics, 62(2), 107–123.
https://doi.org/10.1016/j.jappgeo.2006.09.001
Benedetto, A., Tosti, F., Ortuani, B., Giudici, M., & Mele, M. (2015). Mapping the spatial
variation of soil moisture at the large scale using GPR for pavement applications. Near
Surface Geophysics, 13(3), 269–278. https://doi.org/10.3997/1873-0604.2015006
Bhushan Mahajan. (2020). Road Construction Layers (Road Pavement Layers). Civiconcepts.
https://civiconcepts.com/blog/what-is-pavement-types-of-pavement-road-
construction-layers

64
Bryan Wilson, Stephen Sebesta, T. S. (2018). Rolling Density Meter to Ensure Long Term
Performance of Flexible Pavement. https://dokumen.tips/documents/evaluation-of-the-
rolling-density-meter-for-rapid-continuous-measurement-of.html
Cao, Q., & Al-Qadi, I. L. (2021). Effect of Moisture Content on Calculated Dielectric
Properties of Asphalt Concrete Pavements from Ground-Penetrating Radar
Measurements. Remote Sensing, 14(1), 34. https://doi.org/10.3390/rs14010034
Cao, Y., Labuz, J., & Guzina, B. (2011). Evaluation of pavement system based on ground-
penetrating radar full-waveform simulation. Transportation Research Record, 2227,
71–78. https://doi.org/10.3141/2227-08
Chen, D. H., & Scullion, T. (2008). Detecting subsurface voids using Ground-coupled
Penetrating Radar. Geotechnical Testing Journal, 31(3), 217–224.
https://doi.org/10.1520/gtj101397
Crispino, M., Rampini, R., & Pozzi, M. (2007). An experimental analysis of the effects of
compaction on asphalt pavement macrotexture. Advanced Characterisation of
Pavement and Soil Engineering Materials - Proceedings of the International
Conference on Advanced Characterisation of Pavement and Soil Engineering
Materials, 1, 205–217. https://re.public.polimi.it/handle/11311/265589
Diamanti, N., & Redman, D. (2012). Field observations and numerical models of GPR response
from vertical pavement cracks. Journal of Applied Geophysics, 81, 106–116.
https://doi.org/10.1016/j.jappgeo.2011.09.006
Evans, R. (2015). Sustainable assessment of structures and materials using ground penetrating
radar (GPR). Contemporary Trends in the Regenerative and Sustainable Built
Environment:Technical and Managerial Aspects, 77–87.
Evans, R. D. (2010). Optimising Ground Penetrating Radar (GPR) to assess pavements
[Loughborough University]. https://www.semanticscholar.org/paper/Optimising-
Ground-Penetrating-Radar-(GPR)-to-assess-
Evans/e5e851e5e5fe867653f3a7b80ac7957700bc54f8
Evans, R. D., & Frost, M. F. (2011). The effect of moisture on ground penetrating radar (GPR)
data from asphalt road pavements. International Conference Bituminous Mixtures and
Pavements, 5, 1–10.

65
Evans, R., Frost, M., Stonecliffe-Jones, M., & Dixon, N. (2006). Ground-penetrating radar
investigations for urban roads. Proceedings of the Institution of Civil Engineers:
Municipal Engineer, 159(2), 105–111. https://doi.org/10.1680/muen.2006.159.2.105
Evans, R., Frost, M., Stonecliffe-Jones, M., & Dixon, N. (2007). Assessment of in situ
dielectric constant of pavement materials. Transportation Research Record, 2037, 128–
135. https://doi.org/10.3141/2037-12
Evans, R., Frost, M., Stonecliffe-Jones, M., & Dixon, N. (2008). A review of pavement
assessment using ground penetrating radar ( GPR ). 12th International Conference on
Ground Penetrating Radar, June 16th-19th 2008, Birmingham, UK, 2–11.
https://dspace.lboro.ac.uk/2134/3590
Fauchard, C., Côte, P., Le Brusq, E., Guillanton, E., Dauvignac, J. Y., & Pichot, C. (2001).
Step-frequency radar applied on thin road layers. Journal of Applied Geophysics, 47(3–
4), 317–325. https://doi.org/10.1016/S0926-9851(01)00075-1
Fernandes, F. M., Fernandes, A., & Pais, J. (2017). Assessment of the density and moisture
content of asphalt mixtures of road pavements. Construction and Building Materials,
154, 1216–1225. https://doi.org/10.1016/j.conbuildmat.2017.06.119
GSSI. (2015). Ground Penetrating Radar Equipment Applications - GSSI. Online.
https://www.geophysical.com/products/roadscan
GSSI. (2015). RADAN | GPR Software by GSSI | Underground Mapping Software. Online.
https://www.geophysical.com/software
GSSI. (2018). Ground Penetrating Radar Antennas | GSSI. Online.
https://www.geophysical.com/antennas
GSSI. (2019). PaveScan RDM System | Ground Penetrating Radar Equipment | GSSI. GSSI
Wesbsite. https://www.geophysical.com/products/pavescan-rdm
Hall, K. T., Correa, C. E., Carpenter, S. H., & Elliott, R. P. (2001). Rehabilitation strategies for
highway pavements. Final Report, NCHRP Project, 35(May), 1–38.
Hausman, J. J., & Buttlar, W. G. (2002). Analysis of TransTech Model 300 pavement quality
indicator: Laboratory and field studies for determining asphalt pavement density.
Transportation Research Record, 1813, 191–200. https://doi.org/10.3141/1813-23
Hoegh, K., & Dai, S. (2017). Asphalt Pavement Compaction Assessment Using Ground
Penetrating Radar-Arrays. Congress on Technical Advancement 2017: Construction

66
 and Forensic Engineering - Papers from Sessions of the 1st Congress on Technical
Advancement, 118–126. https://doi.org/10.1061/9780784481035.011
Hoegh, K., Roberts, R., Dai, S., & Teshale, E. Z. (2019). Toward core-free pavement
compaction evaluation: An innovative method relating asphalt permittivity to density.
Geosciences (Switzerland), 9(7), 280. https://doi.org/10.3390/geosciences9070280
Infrasense, I. (2006). Feasibility of Using Ground Penetrating Radar ( GPR ) for Pavements ,
Utilities , and Bridges. In The United States Department of Transportation, Federal
Highway Administration. (Issue August). https://trid.trb.org/view/1225393
InstroTek, I. (n.d.). MC-3 EliteTM | InstroTek, Inc. Retrieved April 10, 2022, from
https://www.instrotek.com/products/mc-3-elite
Ismail, M. A. (2017). Nuclear gauge application in road industry. IOP Conference Series:
Materials Science and Engineering, 271(1), 012087. https://doi.org/10.1088/1757-
899X/271/1/012087
Joshaghani, A., & Zollinger, D. G. (2021). Evaluating the quality of curing applications on
concrete pavements with ground-penetrating radar. Transportation Research Record,
2675(8), 106–120. https://doi.org/10.1177/0361198121996361
Kesharwani, S. (2018). A Review Paper on Comparative Study of Density of Bituminous Layer
by Various Methods. International Journal for Research in Applied Science and
Engineering Technology, 6(7), 546–549. https://doi.org/10.22214/ijraset.2018.7081
Khamzin, A. K., Varnavina, A. V., Torgashov, E. V., Anderson, N. L., & Sneed, L. H. (2017).
Utilization of air-launched ground penetrating radar (GPR) for pavement condition
assessment. Construction and Building Materials, 141, 130–139.
https://doi.org/10.1016/j.conbuildmat.2017.02.105
Laurens, S., Balayssac, J. P., Rhazi, J., Klysz, G., & Arliguie, G. (2005). Non-destructive
evaluation of concrete moisture by GPR: Experimental study and direct modeling.
Materials and Structures, 38(9), 827–832. https://doi.org/10.1007/bf02481655
Leng, Z., & Al-Qadi, I. L. (2014). An innovative method for measuring pavement dielectric
constant using the extended CMP method with two air-coupled GPR systems. NDT &
E International, 66, 90–98. https://doi.org/10.1016/J.NDTEINT.2014.05.002
Leng, Z., Al-Qadi, I. L., Shangguan, P., & Son, S. (2012). Field application of ground-
penetrating radar for measurement of asphalt mixture density: Case study of Illinois

67
 route 72 overlay. Transportation Research Record, 2304, 133–141.
https://doi.org/10.3141/2304-15
Li, J., Walubita, L. F., Simate, G. S., Alvarez, A. E., & Liu, W. (2015). Use of ground-
penetrating radar for construction monitoring and evaluation of perpetual pavements.
Natural Hazards, 75(1), 141–161. https://doi.org/10.1007/s11069-014-1314-1
Li, J., Walubita, L. F., Simate, G. S., Alvarez, A. E., & Liu, W. (2013). Use of ground-
penetrating radar for construction monitoring and evaluation of perpetual pavements.
IEEE Region 10 Annual International Conference, Proceedings/TENCON, 2–5.
https://doi.org/10.1109/TENCON.2013.6718938
Li, Q., Xia, Z., Shao, J., Chen, J., Liu, X., & Fang, G. (2013). Non-destructive survey of
pavement layer thicknesses with ground penetrating radar. IEEE Region 10 Annual
International Conference, Proceedings/TENCON.
https://doi.org/10.1109/TENCON.2013.6718938
Liao, Y., Sargand, S., & Kim, S. (2006). Non-nuclear density gauge comparative study for
OC/OA in HMA construction. Proceedings of the 2006 Airfield and Highway Pavement
Specialty Conference, 2006, 365–376. https://doi.org/10.1061/40838(191)31
Linden, R. N., Mahoney, J. P., & Jackson, N. C. (1989). Effect of compaction on asphalt
concrete performance. Transportation Research Record, 1217, 20–28.
Loizos, A., & Plati, C. (2007). Accuracy of pavement thicknesses estimation using different
ground penetrating radar analysis approaches. NDT and E International, 40(2), 147–
157. https://doi.org/10.1016/j.ndteint.2006.09.001
Mallick, R. B., & El-Korchi, T. (2011). Pavement engineering: Principles and practice. In
Pavement Engineering: Principles and Practice (Third Edit). Boca Raton, Fla. : CRC
P. https://discovery.hw.ac.uk/primo-
explore/fulldisplay/44hwa_alma2130291100003206/44HWA_V1
Malvar, L. J., & Cline, G. D. (2002). Void Detection Under Airfield Pavements.
128.173.204.63, 15. http://128.173.204.63/pavement/fwd/26.pdf
Maser, K. R., & Scullion, T. (1992). Automated Pavement Subsurface Profiling Using Radar:
Case Studies of Four Experimental Field Sites. Transport Research Record, 1344, 148–
154. https://trid.trb.org/view/370805

68
Maser, K. R., Holland, T. J., Roberts, R., & Popovics, J. (2006). NDE methods for quality
assurance of new pavement thickness. International Journal of Pavement Engineering,
7(1), 1–10. https://doi.org/10.1080/10298430500501985
Maser, K., & Carmichael, A. (2015). Ground Penetrating Radar Evaluation of New Pavement
Density Ground Penetrating Radar Evaluation of New Pavement Density Paving
Project - SR 539 in Lynden , WA. February.
http://www.wsdot.wa.gov/research/reports/fullreports/839.1.pdf
Maser, Kenneth R.; Scullion, T. (1991). Automated detection of pavement layer thickness and
subsurface moisture using Ground Penetrating Radar. TRB Paper.
https://scholar.google.com/scholar_lookup?title=Automated Detection of Pavement
Layer Thicknesses and Subsurface Moisture Using Ground Penetrating
Radar&author=K.R. Maser&publication_year=1990
Nuclear Gauges - Humboldt Mfg. Co. (n.d.). Retrieved April 4, 2022, from
https://www.humboldtmfg.com/nuclear-gauges.html
Nuclear Gauges | US EPA. (n.d.). Retrieved April 4, 2022, from
https://www.epa.gov/radtown/nuclear-gauges
Olhoeft, G. R. (2003). Electromagnetic field and material properties in ground penetrating
radar. Proceedings of the 2nd International Workshop on Advanced Ground
Penetrating Radar, 144–147. https://doi.org/10.1109/AGPR.2003.1207309
Pais, J. C., Figueiras, H., Pereira, P., & Kaloush, K. (2019). The pavements cost due to traffic
overloads. International Journal of Pavement Engineering, 20(12), 1463–1473.
https://doi.org/10.1080/10298436.2018.1435876
Pavement Interactive. (2002). Nuclear Density Gauge - Pavement Interactive.
https://pavementinteractive.org/reference-desk/construction/compaction/nuclear-
density-gauge/
Persico, R. (2014). Introduction to Ground Penetrating Radar: Inverse Scattering and Data
Processing. In Introduction to Ground Penetrating Radar: Inverse Scattering and Data
Processing (Vol. 9781118305). https://doi.org/10.1002/9781118835647
Pitoňák, M., & Filipovsky, J. (2016). GPR Application - Non-destructive Technology for
Verification of Thicknesses of Newly Paved Roads in Slovakia. Procedia Engineering,
153, 537–549. https://doi.org/10.1016/j.proeng.2016.08.184

69
Porubiaková, A., & Komačka, J. (2015). A comparison of dielectric constants of various
asphalts calculated from time intervals and amplitudes. Procedia Engineering,
111(TFoCE), 660–665. https://doi.org/10.1016/j.proeng.2015.07.129
Rasol, M. A. R. (2021). Development of new GPR methodologies for soil and cement concrete
pavement assessment [Universitat Politècnica de Catalunya]. In TDX (Tesis Doctorals
en Xarxa). http://www.tesisenred.net/handle/10803/671504
Rasol, M., Pais, J. C., Pérez-Gracia, V., Solla, M., Fernandes, F. M., Fontul, S., Ayala-Cabrera,
D., Schmidt, F., & Assadollahi, H. (2022). GPR monitoring for road transport
infrastructure: A systematic review and machine learning insights. Construction and
Building Materials, 324, 126686. https://doi.org/10.1016/j.conbuildmat.2022.126686
Saarenketo, T. (1997). Using ground-penetrating radar and dielectric probe measurements in
pavement density quality control. Transportation Research Record, 1575, 34–41.
https://doi.org/10.3141/1575-05
Saarenketo, T. (2006). Electrical properties of road materials and subgrade soils and the use
of Ground Penetrating Radar in traffic infrastructure surveys.
Saarenketo, T., & Scullion, T. (1994). Ground Penetrating Radar Applications on Roads and
Highways.
Sebesta, S., Saarenketo, T., & Scullion, T. (2012). Using Infrared and High-Speed Ground-
Penetrating Radar for Uniformity Measurements on New HMA Layers. In The National
Academies Press. Transportation Research Board. https://doi.org/10.17226/22769
Solla, M., Pérez-Gracia, V., & Fontul, S. (2021). Review of GPR application on transport
infrastructures: Troubleshooting and best practices. Remote Sensing, 13(4), 1–5.
https://doi.org/10.3390/rs13040672
Steiner, T., Hoegh, K., Teshale, E. Z., & Dai, S. (2020). Method for Assessment of Modeling
Quality for Asphalt Dielectric Constant to Density Calibration. Journal of
Transportation Engineering, Part B: Pavements, 146(3), 04020054.
https://doi.org/10.1061/jpeodx.0000210
Timm, A., Wen, H., Sharma, S., & Rose, M. (2013). Evaluation of Non-nuclear Density
Gauges for Measuring In-place Density of Hot Mix Asphalt.
https://www.semanticscholar.org/paper/Evaluation-of-Non-nuclear-Density-Gauges-
for-of-Hot-Timm-Wen/718a957619e8802fc93f48f6b81442177857d933

70
Tosti, F., Bianchini Ciampoli, L., D’Amico, F., Alani, A. M., & Benedetto, A. (2018). An
experimental-based model for the assessment of the mechanical properties of road
pavements using ground-penetrating radar. Construction and Building Materials, 165,
966–974. https://doi.org/10.1016/j.conbuildmat.2018.01.179
Utsi, E. C. (2017). Ground penetrating radar: Theory and practice. In Ground Penetrating
Radar: Theory and Practice. Elsevier.
Willett, D. A., Mahboub, K. C., & Rister, B. (2006). Accuracy of ground-penetrating radar for
pavement-layer thickness analysis. Journal of Transportation Engineering, 132(1), 96–
103. https://doi.org/10.1061/(ASCE)0733-947X(2006)132:1(96)
Ziari, H., Behbahani, H., Izadi, A., & Divandary, H. (2010). Field evaluation of nondestructive
tests in measuring the pavement layers density. Journal of Applied Sciences, 10(23),
3091–3097. https://doi.org/10.3923/jas.2010.3091.3097

71
 Acknowledgement

Words cannot express my gratitude to my supervisor, Dr György Sipos, for his

continuous support during my MSc study and his patience, motivation, enthusiasm, and
immense knowledge. I benefited greatly from his guidance during the research and writing of
this thesis.

Many thanks to Roden Ltd for providing the opportunity to use their GPR instruments.
And special thanks to Dr Csaba Tóth for providing the density data.

My sincere thanks also go to Diaa Sheishah, Boglárka Runa, and Olivér Balogh for
their help during the field measurements. During the data processing process, my colleague
Diaa helped me to a great extent.

I want to express my gratitude to the Cultural Affairs and Missions Sector—Central

Administration of Missions, representing the Egyptian government, and Stipendium
Hungaricum, which represents the Hungarian government, for their financial support and
nomination, which gave them financial support and nominated me for the opportunity to study
at the University of Szeged. I am very grateful to my institute (NRIAG) and to my teachers,
especially Professor Gad El-Qady, President of NRIAG, for their support and for the
opportunity to apply for this scholarship.

I also thank all the teachers at the Science and Informatics faculty at Szeged University
who contributed to my knowledge and experience.

I am deeply indebted to my family, especially my wife; her belief in me has kept my

spirits and motivation high. My wife, my warm and heartfelt thanks go to you for the
tremendous support and hope you have given me.

Lastly, I want to thank God for helping me overcome all the challenges.

72
 Appendixes

Appendix A

Base coarse (Finely crushed limestone)

Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

First binder course

Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Core sample location

Aluminum foil reflection Sub-grade layer (Natural soil)

Second binder course

First binder course
Base coarse (Finely crushed limestone)

Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)
Core location

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Fig. A1: Thickness piking of every single layer of Vásár road left lane. A) After placing the
second base layer; B) After placement of the first binder course; C) After placement of the
second binder course; D) After placing the surface course layer.

73
Fig. A2: Depth and thickness comparison and correlation of Vásár road left lane. A) and B)
Comparison between the depths of all the pavement layers measured with channel 2 and
channel 4 (2 GHz); C) Correlation of the thickness with the percentage of difference.

74
 Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

First binder course

Base coarse (Finely crushed limestone)

Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Second binder course

First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-grade layer (Natural soil)

Fig. A3: Thickness piking of every single layer of Cserje sor road right lane. A) After placing
the second base layer; B) After placement of the first binder course; C) After placement of the
second binder course; D) After placing the surface course layer.

75
 Fig. A4: Depth and thickness comparison and correlation of Cserje sor road right lane. A)
and B) Comparison between the depths of all the pavement layers measured with channel 2
and channel 4 (2 GHz); C) Correlation of the thickness with the percentage of difference.

76
 Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-base Layer (Compacted sand)

First binder course

Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-base Layer (Compacted sand)

Second binder course

First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)
Sub-base Layer (Compacted sand)

Sub-base Layer (Compacted sand)

Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)
Sub-base Layer (Compacted sand)

Sub-base Layer (Compacted sand)

Fig. A5: Thickness piking of every single layer of Cserje sor road left lane. A) After placing
the second base layer; B) After placement of the first binder course; C) After placement of
the second binder course; D) After placing the surface course layer.

77
Fig. A6: Depth and thickness comparison and correlation of Cserje sor road left lane. A) and
B) Comparison between the depths of all the pavement layers measured with channel 2 and
channel 4 (2 GHz); C) Correlation of the thickness with the percentage of difference.

78
 Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)
Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Sub-base Layer (Compacted sand)

Surface course
Second binder course
First binder course
Base coarse (Finely crushed limestone)

Base coarse (Coarsely crushed limestone)

Sub-base Layer (Compacted sand)

Aluminum foil reflection

Sub-base Layer (Compacted sand)
Fig. A7: The thickness of piking of every layer. A) Depth profile measured with a 1 GHz
horn antenna; B) Depth profile measured by 1 GHz ground antenna.

79
 Fig. A8: Comparison and correlation of depth and correlation. A) and B) Comparison
between depths (cm) of all the pavement layers measured with the horn and ground antenna,
respectively (1 GHz); C) Correlation of the thicknesses with the percentage of difference (%).

80
Appendix B

Fig. B1: Dielectric comparison of horn antenna (2 GHz) and PaveScan in the Vásár road, left
lane. A) Surface course; B) Second binder course; C) First binder course.

Table B1 The percentage difference between the two methods for dielectric determination
Horn antenna (2 GHz) PaveScan Percentage of difference
Surface course 5.69 5.6 1.59
Second binder course 5.55 5.7 2.66
First binder course 5.49 5.71 3.93

81
 .

Fig. B2: Dielectric comparison of horn antenna (2 GHz) and PaveScan in the Cserje sor road,
right lane. A) Surface course. B) Second binder course. C) First binder course

Table B2 The percentage difference between the two methods for dielectric determination
Horn antenna (2 GHz) PaveScan Percentage of difference
Surface course 5.7 5.72 0.35
Second binder course 5.59 5.74 2.65
First binder course 5.52 5.24 4.84

82
Appendix C

Table C1 Dry and wet weight measurements of a core sample from the second and the first
binder course
Layers from which core samples Weight of the core samples (kg)
extracted
Dry weight Wet weight
Core from the second binder course 2433 2439
(A) 2440
2440
2441
Core from the second binder course 3436 3440
(B) 3442
3442
3442
Core from the second binder course 2837 2843
(C) 2844
2843
2843
Core from the first binder course 3446 3451
(D) 3452
3453
3453
Core from the first binder course 2179 2189
(E) 2192
2192
2192
Core from the first binder course 3557 3563
(F) 3564
3565
3565

83
 Statement

I declare that the results shown in this thesis work are the product of my research activity. I
have only used those materials and references introduced in the appropriate sections of my
thesis. I hereby give my consent to the University of Szeged to deposit my thesis in the publicly
available section of the university library.

Place, Date Signature

Szeged, 2022

84

View publication stats