You are on page 1of 9

Transportation Research Record 1755 ■ 69

Paper No. 01-0106

Quality Assessment During Road


Formation and Foundation Construction
Use of Falling-Weight Deflectometer and Light Drop Weight
Moshe Livneh and Yair Goldberg

Mechanistic empirical-based specifications, which focus on the mechan- of inherent problems, not the least of which is the use of a criterion,
ical properties of materials, facilitate quantitative evaluations of alter- usually dry density, that does not reflect the engineering properties of
native construction practices and materials such as reclaimed materials. soils in roadway- or runway-service conditions (2). Also, Fleming
Thus, to represent these mechanical properties, quality-control and reached conclusions of the same nature (1). The results of his two full-
-assurance testing would be expected to include stiffness along with den- scale trafficking tests show the inadequacy of density as a predictor or
sity measurements. In two case studies, a falling-weight deflectometer safeguard of performance. Thus, it seems that an additional criterion
(FWD) was used during the construction of two major interchanges in is required in order to ensure adequate performance characteristics.
Israel. FWD testing was found to be very useful in identifying local spots There is no doubt that, with ongoing developments, combined
with poor performance, thus enabling execution of the required remedy end-product and mechanistic-empirical-based specifications will
actions. The two studies also indicated that, for a given feature, both a govern quality-control and -assurance testing in Israel. Together
representative central deflection at the 95 percent reliability level and a with in-situ density tests, these will include formation and founda-
coefficient of variability of measured deflections may serve as potential tion stiffness testing. Obviously, the conventional falling-weight
quality-control indicators. The possible use of the German light drop- deflectometer (FWD) is classified as a suitable device for stiffness
weight (LDW) device for measuring the mechanical properties of the measurements but is sometimes considered unnecessarily sophisti-
formation of flexible pavements was also examined. Based on the cost cated for formation and foundation testing; furthermore, it is not
per test, rate of testing, and quality of the data, the German LDW can without limitations on weaker substrates in regard to both transducer
serve as a cost-effective testing device for quality control and assurance range limits and portability (3). Thus, the German light drop-weight
during subgrade and capping-layer compaction. (LDW) device, being lightweight, portable, and simple to apply for
repeated testing, should be considered as providing a potentially
valuable way forward for in-situ stiffness measurements.
The degree of compliance with current common specifications and the In light of the above, this study will concentrate on the possible use
quality of performance are two different elements and are not always of the conventional FWD device and the German LDW device for
directly related to each other. Thus, a performance-based specification measuring the mechanical properties of the compacted layers creat-
covering pavement performance is required to control the long-term ing the formation and foundation of asphalt pavements. Formation
functional and structural performance. With the coming transition here means the partial structure of the pavement that contains the
from empirical to mechanistic-empirical pavement-design proce- natural and compacted subgrade and the capping layers laid on this
dures, it would be advantageous to move toward more mechanistic- subgrade; foundation means the partial structure containing the for-
based specifications, which focus on the mechanical properties of mation and the granular subbase layers laid on the formation. The
materials. This move is desirable because it facilitates quantitative objectives of this study are as follows:
evaluations of alternative construction practices and materials, such
1. To summarize the deflection and stiffness target values avail-
as reclaimed materials, both of which have beneficial cost and envi-
able in the technical literature, and to evaluate the various statistical
ronmental implications (1). Thus, quality-control and -assurance
parameters for serving as potential quality-control indicators.
testing would be expected to include stiffness along with density mea-
2. To examine the use of the FWD device for evaluating the
surements. This requirement also stems from the fact that the possi- construction quality of layers, as well as identifying weak spots by
ble natural variation in site materials along any tested strip may running routine FWD tests at two construction sites in Israel.
require several reference densities, the obtaining of which is usually 3. To compare the conventional FWD deformation moduli with
beyond any practical scope of work. Thus, routine in-situ density tests their comparative LDW deformation moduli by conducting in-situ
cannot alone ensure the compliance of the layers constructed with the comparative tests.
compaction requirements. These conclusions were also drawn by
Pinard, who stated that end-result specifications suffer from a number
DEFLECTION AND STIFFNESS TARGET VALUES
M. Livneh, Transportation Research Institute, Technion–Israel Institute of
Technology, Transportation Research Institute, Technion City, Haifa 32000,
An important step in advancing quality-control procedures and the
Israel. Y. Goldberg, T&M Technology and Management Ltd., 1 Hayesod Street, assurance of formation and foundation construction is to compare
Tel Aviv 68167, Israel. measured deflections with target deflections (4). The latter are a
70 Paper No. 01-0106 Transportation Research Record 1755

function of layer thickness and the nature of the material. Multilayer layers. The results of these measurements will be used to locate sec-
analyses are commonly used to predict pavement deflections. These tions having irregular performance, on which additional routine tests
analyses, which assume static loading and linear elastic material (gradation, plasticity, moisture content, degree of compaction, and
properties, led Syed and Scullion to calculate the required FWD tar- the like) will be conducted to detect the reasons for their deviation.
get deflection in order to monitor the performance of pavement recy- Following this detection, all necessary work will be carried out to
cling projects (5). The FWD loading cycle is, however, a dynamic raise the deviant sections to the required standards.
one with a duration of 30 to 40 msec. Furthermore, pavement In addition to the target deflections, other agencies specify target
materials and subgrades are not linear elastic materials, and their moduli. For example, Fleming et al. suggest a stiffness value of
response to static loads is different than it is to dynamic loads. 50 to 60 MPa, measured on the completed formation, and 100 MPa
Therefore, the use of multilayer analyses to predict target deflections on the completed foundation (8). In Germany, a minimum static plate-
is not recommended. bearing modulus, termed a deformation modulus (EV2), of 45 MPa
To overcome the above-mentioned limitations, Zaghloul and is required at the top of the completed formation. EV2 is calculated
Saeed suggested an empirical approach in which trial sections are from the amount of deflection under the second loading of the
constructed and FWD tests performed at points that show accept- plate (9) (see Figure 1). The thickness of the subbase necessary
able density levels, in order to determine the required target deflec- above the formation level is determined by the amount of frost
tions (6). In the quality-control and -assurance procedure, long resistance required. A surface modulus (EV2) requirement using
pavement sections are divided into homogeneous segments, and the plate-bearing test is also mandatory at the top of the subbase:
FWD tests are conducted on each layer of the segments at inter- 120 MPa for light traffic and 150 MPa for heavy traffic. In the same
vals of 10 to 20 m. Statistical tests are conducted on measured manner, Chaddock and Brown suggest a value of 80 MPa, as mea-
deflections in order to evaluate the construction quality as well as sured by the FWD, with a 450-mm plate and 200-kPa contact stress,
to identify weak points. Zhou et al. additionally suggested that the as an acceptable formation-stiffness cap (10). Finally, the values
coefficient of variation (COV) of the measured deflection be cal- suggested by Nunn et al. are an LDW stiffness value of 30 MPa or
culated for each homogeneous segment in order to evaluate the an FWD stiffness of 40 MPa, measured on top of the completed for-
uniformity of construction (7). mation; and 50 MPa or 65 MPa, respectively, measured on the top
Along these same lines, the Israeli Public Works Department con- of the completed foundation (11).
sidered including the following clauses in its mandatory construc-
tion specifications: the subgrade or capping-layer compaction will
be measured for deflections with an FWD device in accordance with LOCAL EXAMPLES OF FWD USE
ASTM D4694-96. The test will be conducted by means of three
weight drops in each test station with a device equipped with a plate Recently, an FWD device was implemented at two construction sites
of 450-mm diameter having an impact of 22 kN, on the basis of in Israel (A and B) for quality-control and -assurance operations (12).
which deflection results will be normalized. For each carriage way, This FWD testing was conducted on each layer before the construc-
the tests will be carried out at every design cross section (maximum tion of the next layer. The surfaces of the subgrade, the subbase, the
spacing to be 20 m). The representative deflection for each test sta- base course, and the asphalt layers were measured for central deflec-
tion will be the average of the three weight drops. For each lot, every tions with this device. It was equipped with a 450-mm-diameter plate
representative deflection should be less than 0.5 mm, and the COV and had an impact force of 22 kN for all layer measurements except
less than 40 percent for a natural subgrade or a subgrade in cut and the asphalt layers and an impact force of 55 kN for the latter mea-
less than 0.4 mm and 30 percent, respectively, for fill and capping surements. A lower impact load was adopted for the subgrade and the

FIGURE 1 An example of a plate-load test according to the German testing procedure (9).
Livneh and Goldberg Paper No. 01-0106 71

FIGURE 2 FWD measurements of central deflection (normalized to 22 kN) taken along the surface
of the subgrade at Site A (Section 45) (12).

granular layers because the actual pressure exerted on these layers was found to be almost four times that of the other two (stable) seg-
was lower than that exerted on the surface of the finished pavement ments (0.4 mm versus 1.5 mm). Thus, investigations were carried
structure. Thus, in order to eliminate the nonlinearity effects of the out to uncover the possible reasons for the significant difference
materials on the measured deflections as much as possible, it was between the stable segments and the weak segment. A low-strength
essential to apply a lower impact load. The longitudinal spacing in material was revealed to characterize the subgrade of the weak seg-
these measurements was 20 m for each of the two carriage ways ment: marl versus chalk. For this reason, a 600-mm thickness of the
under construction. marl subgrade was replaced by an approved granular material.
Figure 2 displays the variation of the central deflection measure- Similarly, Figure 3 displays the variation of the central deflection
ments (normalized to 22 kN) that were performed on the finished measurements (normalized to 22 kN) that were carried out on the
subgrade at Site A. A brittle chalk formed the subgrade of this site, compacted subbase laid on the sandy clay subgrade at Site B. This fig-
with a design California bearing ratio (CBR) value of 5 percent. Fig- ure discloses, again, a weak segment along the tested section (Stations
ure 2 discloses a weak segment along the tested section (Stations 4510–4511). For the weak segment, a high mean deflection was
4518–4521). The mean deflection measured in the weak segment recorded: it was more than three times that of the two other (stable)

FIGURE 3 FWD measurements of central deflection (normalized to 22 kN) taken along the
surface of the subbase at Site B (Section 44) (12).
72 Paper No. 01-0106 Transportation Research Record 1755

TABLE 1 Calculated Target Deflection Values for Designed Pavement Structure at Site B

segments (0.2 mm versus 0.7 mm). Following additional in-site inves- as the difference between the deflection calculated for the surface of
tigations, the sandy clay subgrade in the weak segment was found a given layer and that calculated for the surface of its overlying layer.
to be overwetted in the compaction process: 17 percent versus the Figure 4 shows the representative values of the measured central
required 10 percent. Thus, remedial action was taken; this involved deflections obtained at the 95 percent reliability level (mean value
removing the subbase material, reprocessing the underlying wetted plus 1.96 times the standard deviation) for the various sections at
subgrade, and moving back and reprocessing the subbase material. Site B. Figure 5 shows their accompanying restraining deflection val-
In addition to its ability to identify weak points along given sec- ues. The two figures should be compared with the values in Table 1.
tions, the FWD central deflection measurements were also used to Doing so will show that, in terms of restraining deflection, the sub-
examine the sections’ compliance with target deflections. These base in the third and seventh sections did not comply with the
were calculated for the construction sites with the assumption that required values of Table 1, and the subbase of the first section
their pavement structures both follow elastic linear laws and possess revealed the best bearing properties. The routine nature of the
moduli values that comply with those assigned in the extended CBR tests, however, did not enable such a distinction. Thus, the calcu-
pavement-design procedure as described by Uzan (13). ELSYM5 lated target values and their derived restraining deflection values
software was used to make the calculations (14). The results of these cannot be regarded as the absolute required values in all cases; the
calculations are shown in Table 1, which includes, for each layer, the reasons mentioned in the previous section provide the physical
restraining deflection value, a new term introduced to express the con- explanation for this. In the same manner, the restraining deflection
tribution of each layer to decreasing the deflection value. It is defined value of the asphaltic layer was found to be much higher than that

FIGURE 4 Representative values of measured central deflections at the 95 percent reliability


level (mean plus 1.96 times the standard deviation) for the various sections at Site B (12).
Livneh and Goldberg Paper No. 01-0106 73

FIGURE 5 Representative values of the restraining measured central deflection for the various
sections at Site B (12).

defined in Table 1, where no additional evidence could be pro- Finally, values for the coefficient of variation were employed to
vided to support this apparent superiority of the asphaltic layer. As control production uniformity. Figure 6 displays these coefficient of
for the base-course material, which was found to possess very low variation values of the measured central deflections along eight sec-
and even negative values of restraining deflection, additional evi- tions of Site A. Figure 7 displays the same values along eight sections
dence was provided to support its resultant inferiority. The out- at Site B. These two figures indicate that the variation in the level of
come of all these findings is that, first, target deflection values deflection in the subgrade may considerably affect the coefficient of
should not be derived from theoretical calculations but from con- variation of each successive layer of the site. This may be learned from
ducting deflection measurements in representative trial sections, a comparison of Sections 3 and 7 at Site A, which possess low varia-
and second, the use of restraining-deflection lower-limiting values tion values, with Sections 4, 5, and 8 at the same site, which possess
should be encouraged. comparatively high variation values; or a comparison of Section 6

FIGURE 6 Coefficient of variation of FWD central deflection measured on each pavement layer at
Site A (12).
74 Paper No. 01-0106 Transportation Research Record 1755

FIGURE 7 Coefficient of variation of FWD central deflection measured on each pavement layer
at Site B (12).

at Site B, which has low variation values, with Section 1 at the same tine test operation, the deflection peaks of the loading plate caused
site, which has high variation values. In addition, Figures 6 and 7 by three load pulses are continuously measured in mm and stored.
show that the coefficient of variation decreased for almost every sec- After the start button is pushed again, the display shows, in addition
tion as the measurement moved from the subgrade surface toward the to these three deflection peaks, the mean value of the three defec-
asphalt surface. tions (δC) and the dynamic deflection modulus (EV1) in MPa. The
expression for calculating this deformation modulus is equal to that
required for calculating the surface modulus of a layered medium
LDW DEVICE possessing a uniform Poisson ratio of µ. The expression assumes a
constant load:
Because the FWD and DCP devices are fully described in the techni-
2 × (1 − µ 2 ) × p × 150
cal literature (15, 16), this section is limited to the description of the
LDW device. The LDW tester, ZFG02, was developed in Germany, EV1 = (1)
where field-stiffness testing is usually required during construction δC
operations. As described later in this section, this tester serves as an
alternative to a static plate-bearing test. Although a standard correc- where p is the uniform distributed contact pressure (MPa) of the plate.
tion between the two is given (Equation 2), it is recommended that one As already mentioned, the German code for the design of flexible
derive a site-specific relationship. pavement structures specifies the required technical instructions for
The tester consists of a loading device to produce a defined load performing in-situ plate-bearing tests. According to this code, other
pulse, a loading plate, and the electronic deflection-data device ZSG test methods—such as the present LDW tests—can substitute for
02 with an integrated data memory and an RS 232 interface. The these plate-bearing tests. In these other tests, the EV1 value is assumed
loading device is a 10-kg falling weight, which falls freely onto a set to be equal to the deformation modulus obtained from the regular
of steel springs and produces a load pulse with a peak of 7.07 kN plate-bearing test during the first load cycle (9) (see Figure 1). For the
and a pulse width of 18 ms (Figure 8). In order to guarantee its safe second cycle of the regular plate-bearing test, the deformation modu-
operation, the falling weight is supplied with a transportation lock lus, EV2 (9), as shown in Figure 1, can be calculated from this EV1
pin and a guide rod with a stabilizer. To maintain the necessary drop value, measured by the LDW device:
height for every measurement, a mechanism is mounted on the top
of the rod to stabilize and then release the falling weight. 300
EV 2 = 600 × ln (2)
The average pressure amplitude beneath the loading plate, which 300 − EV1
has a diameter of 300 mm, is 0.1 MPa. Two carry grips and a casing
with a ball to center the load are mounted on the top of the loading
plate. An accelerometer is placed inside the casing for the center- CORRELATION OF FWD AND LDW
deflection measurement.
The electronic deflection-data device ZSG 02 is a small digital At this juncture, the question arises whether a complete equivalency
hand-held data collector with an automatic triggering mechanism exists between these LDW EV2 values and the FWD ES
and continuous data storage. It can be run on rechargeable batteries, values. In order to evaluate the LDW EV1 values, test pits were exca-
on an AC-DC adapter, from a car’s electrical supply. During the rou- vated at two locations, one containing a uniform CH (i.e., heavy
Livneh and Goldberg Paper No. 01-0106 75

m
CBR =  S 
E
(5)
 k 

where k and m are empirical parameters varying from source to


source.
According to Livneh, Equation 5 yields almost identical values
to those obtained from the well-known U.K. Transportation
Research Laboratory (TRL) equation, where ES = 17.6 × CBR0.64,
provided that k = 15.0 and m = 1.41 (17). Figure 9 also includes
the German practice relationship (18) between CBR values and
EV2 values. This relationship, which in general terms is below the
curve of Equation 5, supports the present findings for the sandy
material.
When the FWD tests are used to obtain EV2, it is suggested that
the backcalculated ES values be reduced in the following manner for
clayey and sandy materials:

300
EV2 = 600 × ln (6)
300 − α × ES

where α is equal to 0.40 and 0.27 for clayey and sandy materials,
respectively. Equation 6 shows that the German EV2 is about
85 percent of the FWD backcalculated value (ES) for clayey material,
whereas it is only about 55 percent of the value for sandy materials.
Fleming (1, 3) reported that his extensive field-stiffness mea-
surements on live construction sites showed a relatively consistent
correlation of LDW EV1 stiffness to FWD ES stiffness of 0.6. In
another study, Fleming et al. demonstrated that this correlative ratio
reduces to about 0.5 (8); according to their data, in other words, the
FIGURE 8 LDW testing before loosening of the sliding hammer. α value of Equation 6 is about 1.5 times the value obtained in this
study. This means that the German EV2 equals about 110 percent of
the FWD backcalculated value, in contrast to the 55 to 85 percent
observed in this study.
clay) and the other a uniform SP (i.e., poor graded sand) material. The discrepancy in these results requires the use of the most criti-
For this study, comparative LDW and DCP tests were carried out on cal values in order to obtain the proper safety-level factor. Thus, the
staggered surfaces arranged at depths of about every half-meter. German code calling for a minimum static plate-bearing modulus
The results obtained in these tests are summarized in Figure 9, (second cycle of loading, Figure 1) of 45 MPa, measured on the top
which contains four calculated functions for EV2 versus CBR. In of the capping layers, is equivalent to about 80 MPa FWD deforma-
this figure, the clayey-stratum function was derived from a com- tion modulus and about 35 MPa LDW deformation modulus of EV1.
bination of Equation 2 and the regression equation as derived from These values are more stringent than those suggested by Nunn et al.,
the in-situ test results. The expression for the clayey stratum is especially for values assigned to conventional FWD stiffness (11).
shown in Equation 3:
CASE FOR MECHANICAL
300 PERFORMANCE MEASURES
EV 2 = 600 × ln (3)
300 − 6.019 × CBR (1 1.41)
A review of current and intended practice in Israel and elsewhere
Similarly, the sandy-stratum equation was derived from a combina- leads to the conclusion that the next development step in the con-
tion of Equation 2 and the regression equation as derived from the struction of road formations and foundations will move toward
in-situ test results. The expression for the sandy stratum is shown in performance-based specifications. The adoption of performance-
Equation 4: based specifications will have two important effects: first, it will
allow the use of materials currently outside the specifications if they
300 can be shown to produce the necessary in-situ performance, and
EV 2 = 600 × ln ( 4)
300 − 4.0354 × CBR (1 1.41) second, it will shift responsibility for the quality of the foundation
construction to the contractor.
In addition, Figure 9 shows that the EV2 values given in Equations To adopt an end-product specification, suitable tests must be
3 and 4 are considerably smaller than the regular ES values drawn up to assess the performance of the material in situ. The
obtained by conventional FWD tests using Equation 5 with ideal tests should include both the shear strength and the elastic
k = 15.0 and m = 1.41. Moreover, the LDW EV1 measured in sandy stiffness of the material tested (19) and its resistance to permanent
material was also found to be smaller for the same CBR value (in deformation (11). For monitoring shear strength, Siekmeir et al.
percent): suggest the use of the DCP along with some other tests for in-situ
76 Paper No. 01-0106 Transportation Research Record 1755

FIGURE 9 Calculated LDW E V2 versus in-situ DCP CBR values according to Equation 3; Equation 4;
Equation 5 for k  15.0 and m  1.41; and the German code according to Esveld (18).

subgrade and granular-base characterization (19). On the other potentially valuable instrument for stiffness measurements in situ. The
hand, there is currently no satisfactory in-situ method of directly FWD is also suitable, but unnecessarily sophisticated, for formation
assessing the foundation’s resistance to permanent deformation. It and foundation testing; furthermore, it is not without limitations on
is therefore proposed that the state of compaction and elastic stiff- weaker substrates in regard to both transducer range and portability.
ness should be measured both at the formation level and on the
completed foundation in order to ensure consistent construction CONCLUSION
quality.
The two Israeli case studies presented in this paper demonstrate the In light of the above, this study concentrated on the possible use of
potential of FWD as a means of compliance testing during pavement the German light drop-weight device for measuring the mechanical
construction, in particular for assessing the pavement formation and properties of the pavement formation. The main findings derived in
foundation. Localized weak-point areas that are due to the poor qual- this connection were as follows:
ity of materials, poor workmanship, and construction defects were
easily identified through FWD testing. 1. Compared with conventional FWD moduli, the LDW moduli
These two studies also indicated that, for a given feature, both a (EV1) are of a smaller magnitude: 0.3 to 0.4 times the conventional
representative central deflection at the 95 percent reliability level FWD moduli in this study and 0.5 to 0.6 times these moduli in other
(i.e., mean value plus 1.96 times the standard deviation) and a studies.
coefficient of variability of the measured deflections may serve as
2. The adjusted LDW moduli to a second cycle of loading (EV2),
potential quality-control indicators. The nonlinear elastic and stress-
when compared again with the conventional FWD moduli, are still
dependent behavior of the granular and subgrade materials make it
of a smaller magnitude: 0.55 to 0.85 times the conventional FWD
necessary, however, to define target values for the measured central
moduli in this study, which is similar to the German relationship, but
deflections; this can be better done by conducting calibration tests in
1.1 times these moduli in other studies.
representative trial sections rather than calculating these values from
an elastic multilayered system. 3. This discrepancy in moduli results requires the use of the most
The findings of the above studies supported the use of target critical values in order to obtain the proper safety-level factor. Thus, if
restraining-deflection values that characterize the contribution of the German code for a minimum static plate-bearing modulus (second
each layer to decreasing the deflection value. A restraining deflec- cycle of loading) is to be followed (i.e., 45 MPa measured on the for-
tion is expressed by the difference between a measured deflection at mation of the pavement), an 80-MPa FWD deformation modulus or a
the surface of a given layer and that measured at the surface of its 35-MPa LDW deformation modulus for EV1 should also be followed.
overlying layer.
Finally, based on the cost per test, rate of testing, and quality of data, Finally, it should be emphasized that mechanical performance mea-
the German LDW can serve as a cost-effective testing device for qual- sures must always be employed along with density testing for soils
ity control and assurance during subgrade and capping-layer com- because soil compacted dry of optimum (high negative pure pres-
paction. In other words, the German LDW, being lightweight, sures) may display high moduli upon compaction that decline upon
portable, and simple to apply for repeated testing, could provide a subsequent saturation (when negative pure pressures decline).
Livneh and Goldberg Paper No. 01-0106 77

ACKNOWLEDGMENTS Proc., 5th International Conference on Bearing Capacity of Roads and


Airfields, Vol. 3, Trondheim, Norway, 1998.
The authors wish to thank the Israel Airports Authority, the Israel 9. Baugrund, Versuche and Versuchsgeräte, Plattendruckversuch. DIN 18
134, Deutsches Institut fur Normung, June 1990.
Ports and Railroads Authority, and Cross Israel Highway Ltd. from
10. Chaddock, B. C. J., and A. J. Brown. In Situ Tests for Road Founda-
whose projects the results presented in this paper were obtained. tion Assessment. Proc., UNBAR4 Symposium, Nottingham, United
The paper was prepared with the assistance of Arieh Aines, graphics Kingdom, 1995.
editor, to whom thanks are also due. 11. Nunn, M. E., A. Brown, D. Weston, and J. C. Nicholls. Design of
Long Life Flexible Pavements for Heavy Traffic. TRL Report 250,
Transport Research Laboratory, Crowthorne, Bershire, United Kingdom,
REFERENCES 1997.
12. Goldberg, Y. Nondestructive Tests for Quality Control of Pavement
1. Fleming, P. R. Recycled Bituminous Planings as Unbound Granular Construction. Master’s thesis. Technion–Israel Institute of Technology,
Materials for Road Foundations in the U.K. Proc., 5th International Con- Haifa, 1998 (in Hebrew with English abstract).
ference on Bearing Capacity of Roads and Airfields, Vol. 3, Trondheim, 13. Uzan, J. A Pavement Design and Rehabilitation System. In Trans-
Norway, 1998. portation Research Record 1539, TRB, National Research Council,
2. Pinard, M. I. Innovative Compaction Techniques for Improving the Washington, D.C., 1996, pp. 110–115.
Bearing Capacity of Roads and Airfields. Proc., 5th International Con- 14. Alborn, G. Elastic Layer System with Normal Loads. Institute for Trans-
ference on Bearing Capacity of Roads and Airfields, Vol. 3, Trondheim, portation and Traffic Engineering, University of California, Berkeley,
Norway, 1998. 1972.
3. Fleming, P. R. Small-Scale Dynamic Devices for the Measurement of 15. Haas, R., W. R. Hudson, and J. Zaniewski. Modern Pavement Man-
Elastic Stiffness Modulus on Pavement Foundation. In Nondestructive agement. Krieger Publishing, Malabar, Fla., 1994.
Testing of Pavements and Backcalculation of Moduli, Vol. 3, STP 1375 16. Livneh, M., A. N. Livneh, and I. Ishai. Israeli Experience with the
(S. D. Tayabji and E. O. Lukanen, eds.), ASTM, West Conshohocken, Regular and Extended Dynamic Cone Penetrometer for Pavement and
Pa., 2000. Subsoil-Strength Evaluation. In Nondestructive Testing of Pavements
4. Zaghloul, S. M. Development of Performance Based Quality Specifi- and Backcalculation of Moduli, Vol. 3, STP 1375 (S. D. Tayabji and
cations for Asphalt Pavements. Proc., 5th RILEM Symposium on E. O. Lukanen, eds.), ASTM, West Conshohocken, Pa., 2000.
Mechanical Tests for Bituminous Materials, Lyon, France, 1997. 17. Livneh, M. Small-Scale Dynamic Devices for Subgrade and Granu-
5. Syed, I., and T. Scullion. In-Place Engineering Properties of Recycled lar Layers Characterization. Proc., 3rd Transportation Engineering
and Stabilized Pavement Layers. Proc., 5th International Conference Specialty Conference, The Canadian Society for Civil Engineering,
on Bearing Capacity of Roads and Airfields, Vol. 3, Trondheim, London, Ontario, Canada, 2000.
Norway, 1998. 18. Esveld, C. Modern Railway Track. MRT-Production, Duisburg,
6. Zaghloul, S. M., and N. S. Saeed. The Use of the Falling Weight Germany, 1989.
Deflectometer in Asphalt Pavement Quality Control, Quality Manage- 19. Siekmeir, J. A., D. Young, and D. Beberg. Comparison of the Dynamic
ment of Hot-Mix Asphalt. STP 1299 (D. S. Decker, ed.), ASTM, West Cone Penetrometer with Other Tests During Subgrade and Granular
Conshohocken, Pa., 1996. Base Characterization in Minnesota. In Nondestructive Testing of
7. Zhou, H., R. R. Rada, and G. H. Elkins. Investigation of Backcalculated Pavements and Backcalculation of Moduli, Vol. 3, STP 1375
Moduli Using Deflections Obtained at Various Locations in a Pavement (S. D. Tayabji and E. O. Lukanen, eds.), ASTM, West Conshohocken,
Structure. In Transportation Research Record 1570, TRB, National Pa., 2000.
Research Council, Washington, D.C., 1997, pp. 96–107.
8. Fleming, P. R., C. D. F. Rogers, and M. W. Frost. Performance Pa-
rameters and Target Values for Construction of UK Road Foundations. Publication of this paper sponsored by Committee on Soil and Rock Properties.

You might also like