Influence of the temperature under the pavement on deflection

measurements with the Falling Weight Deflectometer.

Diadji TALL1, Alassane DIENE1,2, Mamadou Salif DIALLO1,3

1
Laboratory of Sciences and Technology of Water and Environment (LaSTEE), Polytechnic School of
Thies PO Box A 10 Thies, Senegal.
2
Polytechnic School of Thies PO Box A 10 Thies, Electromechanical Engineering Department,
Senegal.
3
Efficiency and Energetic Systems Research Group, Alioune Diop University, Bambey, PO Box 30,
Senegal.

Corresponding author: diadjitall85@gmail.com

Abstract: The main focus of this paper is to study the mechanical behavior of the materials of
a flexible and bituminous pavement structure. The measurement of the deflection will be the
main indicator to estimate the condition of the pavement structure taking into account the
influence of the temperature of the bituminous materials such as the bituminous concrete.
This research work was based on an advanced documentary research of the types of
deteriorations and causes on the works conducted by the Senegalese Road Network Manager
on the temperature survey works under pavement.
Because of the strong dependence of the bituminous material on temperature and frequency,
this study has allowed to simulate the integration of a temperature gradient in the pavement and
its influence on the deflection measurement. Using a comprehensive thermal model, the
temperature gradient in the pavement is reconstructed by measuring the ambient temperature,
air velocity and sunshine.
Keywords: deflection, temperature, deformation, gradient, mechanics.

1 Introduction
In order to preserve the road heritage, the Government of Senegal has adopted an optimal
management strategy for the development and maintenance of its classified road infrastructure,
which is currently estimated at about 16,495 km, based on a three-year planning and
programming of interventions [1].
The Road Data Bank (RDB) has opted, in accordance with the strategy of the Government of
Senegal, for a long-term monitoring of the infrastructure. This monitoring is based on the
regular updating of the data bank for a rational programming of maintenance works [1]. It is in
this perspective that periodic campaigns of deflection measurements and counting of heavy
vehicles are carried out in the national territory to measure the mechanical behavior of
bituminous pavement materials to define the road maintenance technique to adopt following
observed degradations [2].
Temperature gradients exist in the layers of bituminous materials, which can lead to a variation
of modulus in these layers. With the improvement of models describing the behavior of
pavement materials and thus the viscoelasticity of bituminous materials, temperature has
become a significant parameter in behavior studies and diagnosis of structures under real

1
conditions. Temperature in inverse pavement structure analysis programs have often been
integrated as an empirical correlation formula [3].
This paper aims at improving the interpretation of deflection measurements with the falling
weight deflectometer (FWD) on a flexible pavement structure. The operation of the devices
(FWD, Deflectograph) has shown that their measurement assumptions lead to measurement
bias [4]. Temperature and frequency have an important influence on the level of deflection,
which, if not quantified, generate a strong variation in deflection. Hence the need to find a
temperature correction factor to adjust the measured deflections [5].
2 Materials and methods
2.1 Determination of the characteristic deflection
The falling weight deflectometer (FWD) test uses a weight that falls on a spring system
mounted on a plate. Typically the applied stress can last for a period of 20 milliseconds. The
principle of the FWD measurement is to reproduce the pavement loading caused by a passing
heavy vehicle and measure the pavement response. The FWD simulates the passage of a heavy
vehicle at a speed of about 70 km/h and records the induced deformation [6].

Figure 1 : Deflection measurement with FWD

A deflection measurement campaign with the falling mass deflectometer (FWD) was conducted
in 2020 on the Mbour-Joal departmental road (both directions) named D13100 on (384x2)
points measured at the center of the FWD, at a distance of 0 meter from the eight other
geophones. The structure pavement D13100 is flexible [7] :
Bituminous concrete wearing layer (AC): 5 cm,
Base layer in crude laterite: 23 cm
Foundation layer in crude laterite: 32.5 cm

2
 deflection from the FWD.
180
160
deflection d1 (10-2mm) 140
120
100
80
60
40
20
0
0 5000 10000 15000 20000 25000 30000
distance (m)

Figure 2 : deflection measurements on the FWD

The method of cumulative sums (Cusum) is used to identify the homogeneous zones. This
technique of study of the evolution of the homogeneity of the results during a serie of tests
allows the division of the studied zone into homogeneous zones with regard to the studied
parameter [8]. In our case this method is applied to determine the characteristic deflection (dc)
in each homogeneous zone.
Using geophones, the FWD measures the maximum deflection at different impact points
(usually every 50 m or 100 m) of a road (section) to be studied. Based on these maximum
deflections (Figure 3), the road is subdivided into homogeneous zones. These are road sections
with very similar maximum deflection values and/or with a statistical behaviour « monotone »
[9].
𝐶𝑢𝑠𝑢𝑚(𝑃0) = 0
𝐶𝑢𝑠𝑢𝑚(𝑃𝑛) = 𝐶𝑢𝑠𝑢𝑚(𝑃𝑛 − 1) + 𝑑(𝑃𝑛) − 𝑑𝑚 with dm the average deflection.
We obtain for the maximum deflection d1 at 0 meter:

3
 Cusum of d1
500
distance (m)
0
0 5000 10000 15000 20000 25000 30000
-500
Cusum

-1000

-1500

-2000

-2500

-3000

CUSUM d1 MBOUR-JOAL CUSUM d1 JOAL-MBOUR

Figure 3 : Cumulative sum curve Cusum

Thus (06) homogeneous zones are found on both directions of the D13100 section. The
characteristic deflections are determined by the following expression on each homogeneous
zone.
𝑑𝑐 = 𝑑𝑚 + 1.3𝜎
2.2 Thermal modelling
During the deflection measurement campaign on the D13100, the pavement surface
temperatures are measured with the FWD machine on all (384x2) points.
The internal temperatures under pavements are not measured, temperatures values are
calculated using the under-pavement heat equation and the under-pavement temperature
measurements taken during the 2012-2013 survey.
The two sprues N°1 and N°2 are respectively 27 and 28 cm thick of bituminous materials (5 to
6 cm of AC and 23 to 22 cm of CB) at the Dakar site, Highway A1. The diagram below shows
the position of the 8 sounders on the basis of depth [10] .

4
 Figure 4 : sprues N°1

Figure 5 : sprues N°2

The black ball and the weather probe are respectively at 1.64 and 1.45 m above the ground as
shown in the diagram below [10]:

5
 Figure 6 : Air temperature measurement
The mechanical behavior of bituminous concrete is largely dependent on temperature,
expanding with increasing temperature and contracting with decreasing temperature, the results
being interpreted as thermal cracking. At high temperatures bituminous concrete is relatively
soft and susceptible to rutting (permanent deformation), in accordance with repeated traffic
loads. For these research purposes, measurements were made that predicted the surface layer
temperature in situ and a model was adopted regarding the anticipation of the temperature
variation of bituminous concrete over time [11].
A predictive model allowing to extrapolate the temperature under pavement is established. This
model will be used to obtain temperatures at different depths (from 0 to 36 cm) on the D13100.
Expression of the temperature
The heat equation in ground is :
𝜕𝑇 𝜕2𝑇
= 𝛼(𝜕𝑧 2 + ∆𝑟 𝑇) (2.1)
𝜕𝑡

We suppose ∆𝑟 = 0
We assume that the pavement surface temperature is variable [12].
𝑇 = 𝑇𝑖 + 𝑇𝑚𝑜𝑦. cos[w(t − t0)] (2.2)
𝑇𝑖 : Initial temperature inside the pavement
𝑇𝑚𝑜𝑦 : Average daytime temperature (𝑑egree).
w : Angular frequency [𝑤 = 2𝜋/24] (rad/hour)
𝑡0 : Time of maximum temperature during the day (solar midday). (𝑡0 = 12.5 ℎ).
With the change of variable we have

6
𝜃 = 𝑇 − 𝑇𝑖 = 𝑇𝑚𝑜𝑦. cos[w(t − t0)] (2.3)
Equation number (2.1) gives:
∂2 𝜃 ρi Ci ∂𝜃
= (2.4)
∂z2 λ i 𝜕𝑡

The boundary conditions are as follows :

• Z=0 𝜃 = 𝑇𝑚𝑜𝑦. cos[w(t − t0)]
• Z= 𝜹 𝜃 =0
• t= 0 𝜃 =0
After solving equation number (2.4) by the method of separating of variables [13]:
𝜃 = 𝑍(𝑧). 𝑌(𝑡)
We obtain the following expression:
𝑤
𝑇𝑚𝑜𝑦.𝑐𝑜𝑠[𝑤(𝑡 − 𝑡0)] (−√ 𝑧) 𝑤 𝑤
2𝛼
𝜃(𝑧, 𝑡) = .𝑒 . [cos(𝑤𝑡). cos(−√2𝛼 𝑧) − sin(𝑤𝑡). sin(−√2𝛼 𝑧) +
cos(𝑤𝑡)

𝑤
𝑖 (cos(𝑤𝑡). sin(−√2𝛼 𝑧) +
𝑤
sin(𝑤𝑡). cos(−√2𝛼 𝑧))] (2.5)

𝛼 : The diffusivity of the material as [m2 /s]

Here the temperature is more important on the real part of the complex solution
𝑤
𝑇𝑚𝑜𝑦.𝑐𝑜𝑠[𝑤(𝑡 − 𝑡0)] (−√ 𝑧) 𝑤
2𝛼
𝜃(𝑧, 𝑡) = .𝑒 . [cos(𝑤𝑡). cos(−√2𝛼 𝑧) +
cos(𝑤𝑡)

𝑤
sin(𝑤𝑡). sin(−√2𝛼 𝑧)] (2.6)

Hence the expression of the temperature T is given by :

𝑤
𝑇𝑚𝑜𝑦.𝑐𝑜𝑠[𝑤(𝑡 − 𝑡0)] (−√ 𝑧) 𝑤
2𝛼
𝑇(𝑧, 𝑡) = 𝑇𝑖 + .𝑒 . cos(𝑤𝑡 − √2𝛼 . 𝑧) (2.7 )
cos(𝑤𝑡)

This model will be validated by the measurements of the temperature sounders realized during
the campaign of the manager of the road network in 2012-2013 on the Kaolack road and the A1
Dakar-Diamniadio highway and to correct the measurements of deflections at the equivalent
temperature of 34°C applied in the dimensioning of the pavements [14].
𝒅𝟏(𝟑𝟒) = 𝑭𝒄 ∗ 𝒅𝟏(𝑻𝒄) with Fc temperature correction factor and Tc temperature under
pavement [5].

3 Results
3.1 Thermal model validation
7
The heat equation is verified at different times from 9:00 am to 4:00 pm with the maximum
temperature measurements taken during the sounding campaign of temperature measurement
of core well N°1 in well 1, see table 1.
Table 1 : Maximum temperature measured over the period 10/07/2012-09/07/2013_Site
Dakar Autoroute

Zsound (m) 0 0.01 0.03 0.05 0.09 0.14 0.2 0.27 0.36
T°max(C) 59 58.5 53.7 51.2 47.7 43.5 40.9 39.3 38.6
The expression of the temperature to be verified with the measurements of the table 1 [5].
𝑤
𝑇𝑚𝑜𝑦. 𝑐𝑜𝑠[𝑤(𝑡 − 𝑡0)] (−√2𝛼 𝑤
𝑧)
𝑇(𝑧, 𝑡) = 𝑇𝑖 + .𝑒 . cos(𝑤𝑡 − √ . 𝑧)
cos(𝑤𝑡) 2𝛼
The temperature profile depending on the depth is given below at different times of the day.

Figure 7: The temperature profile

3.2 Influence of the temperature in the deflection measurements at the center point of
d1

8
Table 2 : Characteristic deflection
Characteristic
Homogeneous Temperature
deflection
zone ( z=0.03,t)
d1(10-3mm)
1 31.2863633 84.24
2 33.9356427 68.99
3 37.9554004 109.31
4 52.2791583 122.77
5 64.6425261 126.89
6 55.2409469 133.08

Deflection from FWD

140
120
Defection (10-2 mm)

100
80 y = -0,0742x2 + 8,656x - 122,32
R² = 0,8501
60
40
20
0
30 35 40 45 50 55 60 65 70
Temperature (°C) mi-depth 3 cm)

Série1 Poly. (Série1)

Figure 8 : Temperature-deflection relationship

The design temperature of the pavement, called the equivalent temperature in Senegal, is 34
°C [14].
The deflections measured at different temperatures under the pavement will be adjusted to 34°C
using a correction factor (Fc) obtained with the tendency on the D13100 axis of Mbour-Joal.
𝒅𝟏 = −𝟎. 𝟎𝟕𝟒𝟐𝑻𝒄𝟐 + 𝟖. 𝟔𝟓𝟔𝑻𝒄 − 𝟏𝟐𝟐. 𝟑𝟐 (3.1)
With Tc the temperature under the pavement at mid-depth of the bituminous concrete surface
layer.

9
 Correction Factor
3

2,5
Correction Factor
2

1,5

1
y = -0,0016x2 + 0,1814x - 2,5642
0,5 R² = 0,8501

0
30 35 40 45 50 55 60 65 70
Temperature (°C)

Figure 9 : Correction factor

The deflection correction factor d1 found is given by the relation :

𝑭𝒄 = −𝟎. 𝟎𝟎𝟏𝟔𝑻𝒄𝟐 + 𝟎. 𝟏𝟖𝟏𝟒𝑻𝒄 − 𝟐. 𝟓𝟔𝟒𝟐 (3.2)
Thus the deflection adjusted to the equivalent temperature of 34°C obtained after
measurement with the FWD is given by the following formula :
𝒅𝟏(𝟑𝟒°) = 𝒅𝟏(𝑻𝒄)[−𝟎. 𝟎𝟎𝟏𝟔𝑻𝒄𝟐 + 𝟎. 𝟏𝟖𝟏𝟒𝑻𝒄 − 𝟐. 𝟓𝟔𝟒𝟐] (3.3)
4 Discussions
The temperature fluctuates with the deflection, hence the need to correct the deflection
measurements obtained by considering the temperature of the bituminous concrete material in
the case of the flexible structure. Thus, it is recommended not to make deflection measurements
at temperatures higher than 34°C.
To compensate for these temperature effects, temperature correction regulations have been
proposed in other countries as shown in the table below. They only allow to correct the
maximum deflection. But these temperature corrections are only used on pavements with a
bituminous structure.
Table 3 : Deflection correction with other countries
French method Spanish method Belgium method
𝑑𝑇 200𝑑𝑇 𝑑20 = 𝑘(𝑇)𝑑𝑇
𝑑15 = 𝑑20 =
𝑇 − 15 3𝑇 + 140
1+𝐾∗
15 𝑘(𝑇) = 0.0003𝑇 2 − 0.026𝑇 + 1.4

dT : deflection at temperature T (surface or middle layer)

d15 : deflecion at 15 °C
d20 : deflexion at 20 °C
k : coefficient depending on the type of structure
10
A simultaneous campaign of temperature survey under pavement and deflection with the falling
mass deflectometer on the national road system will allow the administrator of the road system
to validate the tendency obtained on the Mbour-Joal departmental road.
Conclusion
The study of the characteristic deflection as a function of temperature under pavement with
bitumen has been presented. The following evolution has been established:
𝒅𝟏 = −𝟎. 𝟎𝟕𝟒𝟐𝑻𝒄𝟐 + 𝟖. 𝟔𝟓𝟔𝑻𝒄 − 𝟏𝟐𝟐. 𝟑𝟐
From the equivalent temperature under pavement of 34°C used in the design of pavements in
Senegal, a temperature correction factor is used to correct the deflection measurements obtained
with the FWD:
𝑭𝒄 = −𝟎. 𝟎𝟎𝟏𝟔𝑻𝒄𝟐 + 𝟎. 𝟏𝟖𝟏𝟒𝑻𝒄 − 𝟐. 𝟓𝟔𝟒𝟐
to be extended to semi-rigid, thick bituminous and composite structures. The finalization of this
work will contribute to the elaboration of the Senegalese pavement strengthening guide.

11
References

[1] SENELABO et AGEROUTE , «Organisation et exécution d'une campagne de sondage

sur les routes revêtues du réseau routier classé»Rapport technique,2012.

[2] CEREEQ,TECHNOSOL,ACI et AGEROUTE, «Campagne de mesure de déflexions sur

le réseau routier classé» Rapport technique,2019.

[3] Benoit Picoux et Rémi Tautou, «Temperature Influence in a Dynamic Viscoelastic

Modeling of a Pavement Structure» Journal of Transportation Engineering, 2017.

[4] Vinciane Le Boursicaud, "Nouvelles utilisations des mesures de bassins de déflexion

pour caractériser l'état structurel des chaussées" Thèse de doctorat , 2018.

[5] Dar-Hao Chen et al, «Temperature Correction on Falling Weight Deflectometer

Measurements» Journal of the Transportation Research Board, 2000.

[6] INFRALAB,"Information sur le FWD, Rapport de recherche, 2001.

[7] EGIS,SENELABO,IFSTTAR,AGEROUTE, "Catalogue de structures de chaussées

neuves et Guide de dimensionnement des chaussées au Sénégal", Rapport
technique,2015.

[8] STAC, "Guide D'auscultation Des Chaussees Souples Aeronautiques au HWD", 2013.

[9] CRR, «Mesure des caractéristiques structurelles des chaussées,» Rapport

technque,2019.

[10] IFSTTAR,EGIS,SENELABO, «Installation des sondes de température à Dakar et

Kaolack» Rapport technique,2012.

[11] Paul Teodor MARC, "Conception et réalisation de structures routières à haute

performances", Thèse de Doctorat,2011.

[12] B. Yousef, "Modèlisation d'un échangeur air/sol dans le but de renforcement des
locaux",Mémoire de Master, 2012.

[13] Alassane DIENE et al «Etude du transfert de chaleur à travers une chaussée en régime
dynamique fréquentiel: l'influence des paramètres extrinséques,» International Journal
of Innovation and Applied Studies , 2017.

[14] IFSTTAR,TECHNOSOL,ACI,AGEROUTE, «Caractérisation mécanique des

matériaux; prise en compte des conditions climatiques sénégalaises,» 2014.

12
13