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

net/publication/266356498

The study of heavy load impact on road pavement

Article · January 2011

CITATION READS

1 919

1 author:

Atis Zariņš
Riga Technical University
19 PUBLICATIONS 26 CITATIONS

SEE PROFILE

All content following this page was uploaded by Atis Zariņš on 14 October 2016.

The user has requested enhancement of the downloaded file.

 ENVIRONMENTAL ENGINEERING ISSN 2029-7106 print / ISSN 2029-7092 online
th
ISBN 978-9955-28-829-9 (3 Volume)
The 8 International Conference ISBN 978-9955-28-827-5 (3 Volumes)
May 19–20, 2011, Vilnius, Lithuania
Selected papers http://enviro.vgtu.lt
© Vilnius Gediminas Technical University, 2011

THE STUDY OF HEAVY LOAD IMPACT ON ROAD PAVEMENT

Atis Zarins
1
Riga Technical University, Kaļķu str. 1, LV-1658 Riga, Latvia. E-mail: atis.zarins@rtu.lv

Abstract. Road transport efficiency is dependent on a number of factors. The most significant of them are: guaran-
teed capacity of cargo traffic and impact of this traffic load on road structures. Latvian road operation experience
show several problems associated with heavy goods transport and their impact on road pavements. A large proportion
of the paved road network has seen strong rutting, which the reason is cited - the overloaded heavy freight traffic.
To find the impact of heavy loads, and in particular - timber haulage transport, on road pavement deformations and
evaluate their prevention possibilities, as well as to investigate the efficiency of freight transport, for example by
increasing the authorized amount of the cargo, and establish conditions that will allow it to do, a study was carried
out.
The research methodology was based on dielectric constant measurements in a road structure using percometer.
Measurements were made in the various pavement structure layers in different depths, with different loading schemes
and in different moisture/temperature conditions. The obtained results confirmed the theoretically expected, that the
load effect on road structures was reduced by increasing the contact area of load, measured between tire and road
surface. Increasing of contact area was established by using the Central Tire Inflate system (CTI), which was installed
on the test car. The study led to the relatively cheap and simple method for determining pavement response, behavior
and for impact studies.

Keywords: road pavement, bearing capacity, dielectric permittivity, CTI, load impact.

1. Introduction conditions. Research on road operating parameters con-

sidering impact of frost heave, during which dielectric
When designing or evaluating operating parameters properties of the environment under study are analyzed is
of an existing road structure, methods and solutions based published by Saarenketo et al. (2000) and Saarenketo and
on theoretical considerations are commonly used. Thus a Aho (2005). There is pointed to the possibility of use of
specific situation has to be associated with certain fixed dielectric value under load as parameter and characteristic
parameters. But usually engineer has limited amount of of loading impact on the road body.
such standard parameters and situation may not necessar- This study will examine the feasibility of using di-
ily be appropriate to the relevant and well investigated electric conductivity parameters of the road materials, for
standard case. Therefore, in practice often have to make a the evaluation of road pavement response under full scale
direct study of the situation, which usually have to be loading in situ.
associated with sophisticated equipment and methods. It
is therefore important to put into practice research meth- 2. Dielectric properties of road material
ods, which require lower cost required for the study and
its implementing consequences. One such possibility is to Dielectric permittivity Er(ω) for soil or road material
fix and analyse the change of dielectric parameters of can be expressed with relationship:
road material with percometer.
The first indications of such a tool in research of
E (ω )
road construction are available from the 1995th when per- E r (ω ) = , (1)
cometer used in tube suction test by Saarenketo and Scul- E0
lion (1995) and Scullion and Saarenketo (1997).
There are several studies about dielectric properties where: E(ω) –absolute dielectric permittivity
of road materials and soils with regard to operational of material, which depends on frequency and is complex
characteristics: Saarenketo et al. (1998), Saarenketo et al. value, E0 – dielectric constant of material, or the
(2000). These studies were conducted under laboratory relative permittivity of a material for a frequency of zero.

1240
 Permittivity describes how much is the resistance of structure must be taken into account while evaluating
media to the electric field. Dielectric permittivity of soil measured dielectric data. The latter applies especially to
materials determined by intensity of various processes in the quantity of water solution, density and its chemical
material structure: 1) ionic and 2) dipolar polarisation and structure. In order to compare measurements made in
relaxation of water solution molecules and 3) atomic and different media, influence of all of the parameters must
4) electronic resonances in material media due to voltage be precisely identified and evaluated. However, two dif-
applied to. Character of each of those processes deter- ferent measurements taken at the same media can be as-
mines different measurement frequency to review them. sumed as comparable, even without knowing these pa-
Because for road body medium first two of mentioned rameters, with the assumption that they have been the
processes are more actual, the measurements can be car-
same in both comparable measurements. Electrical losses
ried out using frequencies that are common for them –
in soil results from the inherent heat loss, loss from proc-
about 2kHz for detecting of ionic and domain polarisation
ess of polarization of water molecules and the loss in
processes, and about 40-50 MHz for molecular level po-
larisation processes. However, possibly the whole spec- electrochemical processes between clay mineral particles.
trum of frequencies, or at least both mentioned together, This phenomenon is the result of a complex, frequency
can make an interest in soil material studies. dependent dielectric permittivity. The real part of the
Parameter measured with percometer is the real part dielectric permittivity for natural environments changes
of the complex relative dielectric permittivity value (1). in the range of 1 for air, up to 81 for water. Water is that
Measurement of Er for natural materials, including component of natural environment, including unbound
soils and granular building materials is relatively compli- road materials, which in great degree affect both the di-
cated. In order to avoid the impact of environmental het- electric permeability, as well as its mechanical properties,
erogeneity of conductivity parameters to measurements, for example - bearing capacity of pavement structure.
choice of right frequency of measurement and location of Note however that this effect also depends on other fac-
sensors has an importance. Also of importance is a cor- tors, for example: environmental moisture conditions,
rect understanding of the influence of water to the physi- etc.. Road structure can contain adsorption (hygroscopic)
cal processes, and of their structure and mechanism. water, capillary (matrix) water and free water. Dielectric
Conductivity of the natural media largely depends on its permeability of water within soil or road material de-
moisture. But much more on the electrochemical proper- pends on the total soil particle surface area available for
ties of water solution, i.e. salt content in water, as well as water molecules, as well as from the polarization of water
on electrochemical properties of media particles. It is molecules, as well as the density of the soil. Thus, change
possible to use empiric relationships between the conduc- of material conductivity and hence the dielectric permit-
tivity Er and the proportion of water in material, or mois- tivity resulting from load, or an compression, can be ex-
ture W. In generally relation Er = f (W) at different mois- plained with the releasing of colloidal particles and ions
ture levels and water electrical properties is non-linear,
from the clay particle surface and their suspension in the
and depend on water and soil particle interactions. Most
free and matrix water. An others factors causing this ef-
of typical materials have fixed inherent values of the Wv
fect is due to the 1) suction and 2) pumping phenomenon
and Er.
in capillary material, as a result of cyclic loading. Loss of
Main process, which can be analysed with respect to
dielectric capacity of soil and road material also can be
road material response to loading, is change of polarisa-
explained with suction, when free water from the upper
tion in water solution in interaction with soil particles and
layers is temporarily withdrawn to the lower layer in re-
voids during loading and consecutive volume change.
sult of the compression and subsequent release of the
Polarization reduces the electric field in dielectric media,
pavement structure.
because of the increasing of polarization charge density
on particle surface. Consequently, the surface parameters
of the media constituent material and their molecular

Fig 1. Load distribution schemes of test vehicles

1241
 Fig 2. Sensor configuration in test stations

3. Methodology and load combinations under research. The effect of us-

ing a reduced pressure in the tires was also considered.
The essence of the study is based on a comparison - The aim is to find out - whether it is possible to
the impact on the road structure from reference vehicle characterize the capacity and response of road structure
which consider conventional loads and dimensions, in using dielectric parameters during full loading test.
comparison with the impact of vehicle with dimensions The reference vehicle (40t, 6 axles) was determined
by following the appropriate legislative vehicle and Duration of measurement cycle was provided such
load parameters Three basic vehicle configurations with that it includes both - the loading, and recovering of the
different load distributions was used in tests (see Fig 1.). structure under test. Loading was applied with speed 15-
The standard six axle timber haulage vehicle was 25 km/h, so vehicle crosses test section of road in approx.
equipped with CTI system and load tests were carried out 5 seconds. Recovery time depends on the water and mate-
both: with a standard tire pressure (0.8MPa), and with a rial condition, and may be from 10 s to several minutes.
reduced pressure using a CTI system (0.35MPa) results Necessary length of data logging was fixed during test by
of which in graphs below are marked as “with CTI”. following the measurement.
Test equipment consists of percostation (3 stations All the test stations is located on functioning roads.
each equipped with 3-4 sensors), 4 channel data logger Fist one in section of asphalt paved road, third – on the
and portable weighing-machine. same road section with gravel pavement and second one
The research methodology is based on a road struc- on low volume forest road, with just renovated drainage.
ture dielectric constant measurements using percometer. From obtained records, the change of dielectric
Measurements were carried out in different layers of value ∆Er has been evaluated with purpose to express
pavement structure at various depths, and during different pavement response (see Fig3).
loads were applied. Measurements included the following
sessions: 1st - before freezing, (late November), when the
pavement structure has a maximum moisture due to au-
tumn rainfall, but coverage is not yet frozen (autumn
slush), 2nd after the surface thaw,(April) when the pave-
ment bottom has not yet been thawed, and the road struc-
ture is saturated with water (spring slush), 3rd dry struc-
ture (late May) and 4th - after summer rainy season (Au-
gust). Overall, 165 proper readings – response diagrams,
were obtained from the single sensor at the time of full
loading tests. Road pavement response study is based on
the consideration that dielectric permittivity change of an
unbound pavement material under the load correlates Fig 3. Sample of test record and parameters to be evaluate
with the amount of applied impact. So measurements of
Er in equal media conditions for different loads can be 4. Test results
used for impact evaluation. Therefore it can be assumed,
that induced parameter changes within a single measure- The overall results confirm hypothesis that dielectric
ment session (in equal conditions) are comparable. value change of an unbound pavement material under the
Assuming, that changes of dielectric parameters in load correlates with the amount of applied impact. This
road materials affecting from load impact, in greater ex- follows from graphs presented (see. Fig 4-6). Coefficient
tent depends on matrix water, in tests the measurements of correlation varies from 0.7 to 0.9, however amount of
was carried out using frequency – 2kHz for detecting Er. data evaluated can be treated as insufficient for statistical
Samples of Er have been taken using sampling rates 200 analysis, because of assumption, that only data from cer-
– 1000 Hz. tain station and certain measurement session can be com-
Stations of tested road structures were configured as pared in such way. So we were able to make no more
drawn in Fig 2. than only one cross for each load distribution scheme per
session.

1242
 Fig 7 shows pavement response differences in 1st
station (asphalt pavement). This comparison confirms
that elongated vehicle (24.5 m, 7 axles) with cargo load
52 t makes nearly equal impact, as the reference vehicle
with 40 t. In turn, using reduced tire pressure (52 t with
CTI), reduces impact almost twice. If weight increases
(60t), response to upper layers of pavement increases
proportionally (fig 4 to 6). However, as it can see, this
effect maintains only in top layers of pavement. On low-
est sensor the difference disappears (Fig 6 and 7).

Fig 7. Differences of Er for Station 1 (17.04.2010)

Test results show that 40t vehicle equipped with sin-

gle tires on asphalt pavement cause slightly stronger re-
sponse than reference vehicle (see Fig 8). In turn - situa-
tion on gravel road (Fig 9.) shows that single tire load
leaves significantly greater impact, than double tired ve-
Fig 4. Differences of Er for Station 3 (9.04.2010) hicles with the same cargo weight.

Fig 8. Differences of Er for Station 1 (07.10.2010)

Fig 5. Differences of Er for Station 3 (07.10.2010)

Fig 6. Differences of Er for Station 3 (27.05.2010) Fig 9. Differences of Er for Station 3 (26.11.2010)

1243
 tion, and obviously has a base frequency depending on
road material.
Other phenomenon fixed during tests is an opposite
response, when dielectric value drops down under the
load.
This was detected with lowest sensor of asphalt
pavement (see fig 12) while upper sensors show positive
response, and this reminds in all sessions. And it was
observed also in well drained upper part of forest road
(see Fig13).

Fig 10. Dielectric response Er in station 3. Vehicle 52 t

(26.11.2010) (vertical axes between sensors are shifted)

Fig 11. Oposite dielectric response Er in station 1

Fig 13. Oposite dielectric response Er in station 2

Fig 12. Dielectric response Er in station 3. Vehicle 24.5
m, 52 t (26.11.2010) (vertical axes between sensors are During tests the recovery time of dielectric value, or
shifted) time of media stabilization after loading induced re-
sponse, also was studied. Unlike to study of Saarenketo
An interesting phenomenon of dielectric value fluc- and Aho (2005) were recovery time is defined in different
tuation appears in records, which were done in early way, and was measured in ionic level spectrum part, in
spring and late autumn sessions (see Fig 10 and 11.). Re- current study it was defined as time while dielectric con-
cords done in other sessions didn’t show this. In Fig 10 ductivity value returns in initial position (see Fig 3), and
appears that loading initiates such waving. In turn, Fig was measured in 2kHz frequency. According to this
11shows that waving precedes vehicle movement and definition, the recovery time in most cases did not fixed,
appears even before vehicle passes test section. Moreover because of they exceeds of 10-15 min or more, or was not
– fluctuation has a distinct frequency. In this case it was tended to return in initial position at all. This can be ex-
about 5s. Time of full quieting lasts about 2 minutes. plained with unstable character of molecular level water
This means that there exists a particular soil determined movement in wet road material.
natural frequency of fluctuations. Also apparently that
fluctuation initiates by the definite axle load configura- 5. Conclusions
tion and other parameters (axle weight, speed, etc.). As it
is seen in records, fluctuation prevails in upper layers and Water movement and condition in road material and
quiets down at lowest part of pavement. This can be ex- soil can be characterised using dielectric permitivity
plained by the proximity of the upper sensor to the free measured within two frequecy intervals - that are com-
surface. It is obvious (Fig 11) that impact from fluctua- mon for them – about 2 kHz for detecting of ionic and
tion is close or equal with the impact from the applied domain polarisation processes, and about 40-50 MHz for
load and must be properly evaluated. molecular level polarisation processes. This study was
Fluctuation described can be explained with water carried out in lowest frequency interval. However results
pumping phenomenon which prevails in wet road condi- show that some response parameters, i.e. – recovery time,

1244
 can be researched with higher frequency, because of its Saarenketo, T.; Bell, S.; Berntsen, G.; Sundberg, S.; Vuontis-
more ionic level character. järvi, E. 2002. Roadex – Benchmarking Low Traffic Vol-
This study shows, that it is possible to characterize ume Road Condition Management in EU Northern Periph-
the capacity and response of road structure using dielec- ery Area. Proc. of BCRA 2002, Lisbon, Portugal,
Balkema. 69–77.
tric parameters of road material or subsoil during full
Saarenketo, T.; Kolisoja, P;, Luiro, K.; Maijala, P.; Vuorimies,
loading test. However it is clear that without additional
N. 2002. Percostation for Real-Time Monitoring of Mois-
researches and definng soil dielectric propertiis and water ture Variations, Frost depth and Spring Thaw Weakening.
regime in road structure, only relative analysis shall be Transportation, Research Board Meeting Proceedings
possible within one measurement station and session. 2002.
It was concluded, that reduced pressure in tires sig- Saarenketo, T.; Scullion, T.; Kolisoja, P. 1998. Moisture Sus-
nificantly reduces an impact on top layers of pavement. ceptibility and Electrical Properties of Base Course Ag-
However in depth of base layer (>30cm), this effect dis- gregates. Proc. of BCRA 1998.; July 6-8, Trondheim,
appears. In addition, it was fixed, that the impact reduc- Norway, Volume 3, 1401-1410.
ing effect in greater extent refers to unbounded pave- Scullion, T.; Saarenketo, T. 1995. Ground Penetrating Radar
ments. A similar effect also applies to single tires. Technique in Monitoring Defects in Roads and Highways.
In Proc. of the Symposium on the Application of Geo-
References physics to
Scullion, T.; Saarenketo, T. 1997. Using Suction and Dielectric
Engineering and Environmental Problems, SAGEEP, April 23– Measurements as Performance Indicators for Aggregate
26, 1995.; Orlando, Florida, 63–72 Base Materials. Transportation Research Record, vol.
Saarenketo, T.; Aho, S. 2005. Managing spring thaw weakening 1577.; 37-44.
on low volume roads, Roadex II project.; Available on the
Internet:

1245

View publication stats