Beruflich Dokumente
Kultur Dokumente
Purdue e-Pubs
JTRP Technical Reports
1962
Recommended Citation
Yoder, E. J. Flexible Pavement Deflections-Methods of Analysis and Interpretation : Technical Paper.
Publication FHWA/IN/JHRP-62/04. Joint Highway Research Project, Indiana Department of
Transportation and Purdue University, West Lafayette, Indiana, 1962. doi: 10.5703/
1288284313602.
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact epubs@purdue.edu for
additional information.
Wm
f
E.J. YODER
PURDUE UNIVERSITY
LAFAYETTE INDIANA
Technical Paper
TO:
K. B. Woods, Director
Joint Highway Research Project
FROM:
6-20-6
File:
Project: C-36-52F
Respectfully submitted,
Attachment
Copies:
F. L.
J. R.
W. L.
W. H.
F. F.
F. S.
G. A.
Ashbaucher
Cooper
Dolch
Goetz
Havey
Hill
Leonards
J. F. Mclaughlin
R. D. Miles
R E. Mills
M. B. Scott
J. V. Smythe
J ! Waling
E. J. Yoder
http://www.archive.org/details/flexiblepavementOOyode
Technical Paper
Eldon J. Yoder
Research Engineer
Purdue University
Lafayette, Indiana
Eldon Jc Xodar
Joint Highway Research Project
Purdue University
Synopsis
Repeated
other factors*
The deflection of a flexible pavement is partly e3.a8tic on character,
Thus,
tions of load
caii
deflection measurements, including the work done at the WASHO and AASHO
Road Tests o
deflection at the pavement surface may not yield the desired reeults*
2C
4,
divided by height,)
5o
60
of elasticity can tolerate less deflection, for a given lead, than the
with low moduli*
by
Eldon Jo Todr
Research Engineer and Associate Prof awsor
of Highway Engineering
Purdue University
INTRODUCTION
Ttfhich
Large defies*
These include gross load, tire pressure, repetition of load, thickness end
Pavement distress may be influenced by a combination of the above fact:Pavement failure may result from excessive shear stresses, consolidati
differing points of
vie*r
Second,
theoretical conceptso
relatively long period of tins before new concepts can be proven outo
On
the other hand, theoretical equations are generally based upon simplified
bra
used in practice
The Inevitable result of the above has been that most design techni-
theoretical
rt
<>
>
In the ideal case the design engineer should relate pavement stress
to the ability of the paving materials to resist these stresses
However,
not only are the computed stresses under load open to question due to th
nonhonogen9i'ty of the problem, but in addition stresses are vers' difficult
Sine
load indicates to some extent its ability to resist the applied forces*
Engineers have for many years used the principle (at least in an
Ira-
direct sense) that deflection of a pavement under load indicates its load
carrying capacity,,
adopted by the Kansas State Highway Department and the Navy Department*
It
regarding the shape of the tire imprint on the pavement surface, the
vsith
deflection naasure-
Rapidly applied
Alsoj
flexible pavements are layered systems and -obey are not homogensouso
ifeverthelssSfl principles derived
elasticity have great use in pointing out to the engineer the interrela
tionso
data either by inference or by direct measurement relative to the stressstrain properties of a pavement structure*
Such
Pavement Distress
,
The complexity
tained by the various design methods as much as any other single point*
There is no exact definition in existence at the present tine which
states the ultimate desired performance of pavements.
widely in their concepts of acceptable performance*
Engineers vary
If one is -willing
Be-
of allowable deflections
are not standardised*
This in
tuzti has
to soma quostioric
important and in many cases override inferences that may be drawn from
total deflection
The movement of a pavement under load is determined in the final
This
used for determining stresses in the pavement structure and from this th
thickness of pavessant to keep the deflection
*ri.thin
a certain allowables
amount is determined*
tion Studies
and
Evalua
Pavement
Performance
Use in
!
i>imi
in
iuii
rir*' n naMMHMIMMMIWNMfl
n m m*
in in
u i.n u
i.
ii
ii
i|
in
Deflection msasurearants are extremely useful in evaluating the potential performance of existing pavements*
Use as a ResearchTool
Researchers have found measurement of deflection to be a very useful
toolo
METHODS OF ANALYSIS
ways,,
The
method of analysis should depend upon the intended usage of the data
In
some cases detailed data are not only unwieldy but are too complicated to
the maximum amount of information with the least amount of effort are
usedo
2o
5s
time,.
necessary to analyse the data in light of the depth below the pavement
surfac*o
of bending under load will crack more frequently than those with large
radii of bonding e
Use of unit deformation is subject to the restriction that deflection
is dependent upon depth below the surface as v*ell as type of sate rial*
on
airy
Pavement
deflsction has a definite relation to stresses and, for a given load, gives
Deflec-
Pavement
[in hi mmi Stiffness
wii iiiiim ii
11
Stiffness
It en-
<
According to conrtoa
These are
pavement is not truly rigid in the sense that it doesn f t deform under load*
but rather it is implied that its relative stiffness is much greater than
The classi-
arci
these two extremes where the pavement can ba termed semi-rigid o? samiflaxiblB
tirie
This
Elastic Behavior
proportional limit.
that the modulus of elasticity of practically all paving materials (particularly soils) is dependent on rate of loading*
On the other hand, most investigators have assumed elasticity in the conventional sense (1, 6, 35 )o
elastic manner,
of elasticity
Figure 1 illustrates the relationship between modulus
and strain of a loaded element,
strain in a direction
strain in the direction of the applied stress to tho
20
to evaluate*
Considering the special triaxial condition where #}_ *
^2 K
<:r
3 *k
A-|tf(l-%*)
(2)
0)
f-2ftt-2X)
For a Poisson's ratio of 05, the quantity
-^
=& is rfL*
V
This
This
to the corresponding unit strain for any value of stress below the propor>
elastic limits, since they do not have stress-strain curves which are
straight lines D
There appears to be
u
considerable Justification for this concept, especially when dealing with
<>
of strain can be standardized when dealing with measured values of pavement deflection..
paragraphs
ous individualej these articles will not be discussed but rather a bibliogrs&Xfs listing several pertinent references is given at the end of this
report*
sign requirements must be mst, (1) design against shear failure and (2)
dssign against consolidation of the
soil,,
which are moving, quality of paving materials, climate and many other
factorso
stress problamo
This
12
7)
-was
extended by Fox
(lj>)
Tabulated stresses
for the three layer problem have recently been presented by Joras (21)
Deflection equations for flexible pavements take the folloTJing general
form*
A-^F
where
(4)
& is deflection
p is contact pressure
a is radius of contact (circular)
subgrada
by Baker (l)o
Researchers are continually attempting to check theoretical stresses
and strains with the actual values which exist in flexible pavement systems c
As a rule the measured stress and deflection patterns follow the ideal (16,
20) but the exact numerical values cannot bo checked against the theore*
tical values
Hence, it becomes
WMNKMMM
13
and for the general case a large portion of the deflection takes place in
the subgradeo
statement can be made that deflection can be used as a guide for determining the ability of a pavement to insist stresses since deflection reflects,
Also,
Slab Action
The present trend in design of flexible pavements is towards building
stiffei* pavements e
In some
cases the base course materials are stabilised with cementing agents and
in addition bituminous pavements are thicker than they were several years
age
condition^
Ipy
These reactive
This, it
34
Certain assumptions axe made in the classical analysis of rigid pavements which should be recalled.-
pressure under the slab is proportional to deflection and that the reactive
pressures are vertical and vertical only*,
Jfext
5c
_J&_
12(1
(5)
^2)
Stiffness of a slab varies as the cube of the thickness and as the first
E^~
^-\/
22(1 ^2)k
\j 32(1
(6)
^f2)k
where
<*
rf m
nevertheless, practically
all pavements exhibit some degree of stiffness and the factors considered
15
to resist tensiono
surfaces have less stress (for a given deflection) than surfaces bent over
5)<>
In
the former case, considering a long and narrow rectangular footing, the
Like-
wise, heavier dowel bars are required over stiff subgrades than over yielding subgrades due to the sharply bent surface over the stiff subgrades
<,
a given deflection, and for a given subgrade, stress varies directly with
the rigidity of the slabo
Equation 6 is the well known radius of relative stiffness first defined by V.'estergaardo
of continuous contact between the slab and base and complete structural
integrity of the slab) probably are not completely fulfilled in prototype
pavements for a variety of reasons.
equi<=>
16
(a)
S-|
where P is the wheel load and
(9)
&
example of use)o
Insofar as pavement deflection is concerned, the important principle
equal to 500,000 psi and for various values of modulus of eubgrade reaction.,
The numbers on the curves indicate the pavement thickness used for the cal
culationso
Comparing the
txro
be noted that for a given deflection stress varies directly a3 the pave-
pavement increases the pavement ^takes up more and more of the load"
On
17
the other hand, as stiffness of the subgrade increases the subgrade itself
is stressed more
The important consideration shown in Figurs 6 is that permissible
invariably be smaller for stiff subgrade3 than for lesser stiff onsso
Likewise., allowable deflections must also be
No doubt
Neverth^
practical the external conditions which might affect the test values
Thus, even though for the general case these factors can be handled by
keeping them constant, the engineer 3hould know something of their rela-
Effect of Speed
Previous paragraphs of this paper have implied that the pavement
lished that flexible pavements are viscoelastic and that their deflection
fact that the load imposed on the pavement is dependent upon spaed of
loading*,
Quinn (26) and others have demonstrated that the load imposed
13
in its simplest form, the variation of contact load with speed for a smooth
hard wheel moving over a bump on a plane surfaee<.
Increase in load i3
due to acceleration in the vertical direction as the wheel oaves over the
bumpc
since the true load acting on the pavement is a function of the frequency
All
other things being equal the load between the wheel of a vehicle and the
pavement increases as speed of the vehicle increases
In contrast to this, theories of viscoelasticity indicate that ths
Various investigators
have assumed mathematical models which place these simple models in series
tiiae
above
de>
flections decrease but that as speed increases load imposed upon the pave-
19
Temperature
The effect of temperature on deflection i3 significant in some caseso
ture
(Figure 3)o
Failure of sons
flexible surfaces during the spring months, for example, results from
cracking much as is the case for rigid pavement s
to the advantage of the pavement structure during the spring melt period
relatively thin and vfoenever measurements were mad during the hot after-
betx-seen
20
wheel path*
-.Jlth
This
also accounts for low supporting capacity at the edge of the pavement | the
effect of edge condition on bearing capacity h&3 been established for many
years*
In particular,
it has been shown that the strength of soms soils under repeated loads is
over identical soils that have no stress history due to several factors*
These include densification, and rearrangement of the soil particles*
On
the other hand, other soils show reduced strength after repeated loadingo
Recent research ha3 indicated that previous stress history ha3 a pronounced
effect on the strength of base course materials*
Time, incidently, also has its effect on the properties of the asphaltie
surface*
Asphalt may become progressively harder with time and this in turn
some pavements*
21
Effect of Climate
Climate, as used herein, refers to seasonal variations in rainfall
A gen-
eral increase in the measured deflections under load is noted for all components of the pavement, particularly the plastic subgrades
It is not the intent here to discuss the matter of pavement support^
that a major portion of the distress to a pavement can occur under veiy
Pavement deflection
22
under spring time conditions are generally greater than those under other
conditions when the subgrade is dry and possesses greater rigidity,,
In addition* in situ
Thus, tdaan
The per-
centile value which should be used depends upon the variability of each
particular problem
determining from those the variability that exists at the location and an
of this paper to discuss statistical variation but it is necessary to recognize this factor in testing of this type
out that the relationship between stress and deflection is variable depending upon the properties of the pavement and underiying layers
It
has also been demonstrated that numerical values of deflection are dependent
23
it is
apparent that the true significance of making tests can be realized only
de=>
Hveem (20)
condition} however* if they crack they may revert to the flexIbXe condition,,
can be greater for thin surfaces than for relatively thick surfaces,,
It is believed that in general for corventional flexible pavements^
005
24
For the conventional pavement, a large portion of the gross deflection occurs in the subgradeo
of this reporto
layers*.
As stiffness of
However, numsrlcal
Figure 11
vdth depth and as a rule the major cumulative deflection occurs in the
subgradoo
transverse direction and explains why map cracking sometimes init iall y
25
over a waterbound macadam base which in turn rests upon a relatively coheoionless sand subbase
ness of the deflected surface over the granular subgrade compared to that
over the clay subgrade indicated that higher stress existed in the former
case than in the latter o
in Figure 14o
Typieallyv
trorko
However, the
It merely
requires that as the load is moved toward the probe, readings be taken at
26
Component Deflections
Relative deflection of a pavement layer decreases as depth increases.,
Therefora, to Bake use of component deflections it Is necessary to taka
Measurement of
002*=
The re-
sults of these measurements have in many cases pointed up weak layers that
exist in the pavement systemo
it is
with depths
bet\iraen
surface and
component deflection*
Component deflection can be used to estimate relative values of
It should be pointed
xvers
restrictive in that th3y do not take into account properties of the pave-
ment system
27
paving materials
the spring of the year, for example, to determine the suitability of one
ately apparent.,
tion informations
a vast amount of information on many pavements and can generally concentrate their efforts on detailed information from several carefully selected
pavement*
<,
to bring out that there is a need for conducting more research in which
stresses and strains are measured simultaneously
SUMMARY
factors
Theory as
2B
Likevdse, la
Relatively stiff pavements can tolerate less deflection under load than
can pavements of less stiffness.
loading conditions,.
behavior of pavements
fl
"
29
LIST OF REFERENCES
lo
2o
Eaker, Robert F a , "Pavement Deflections Using Photogrammetric Techniques," P roceedings, Highway Research Board, 1957
U*
5o
60
80
9o
10o
11
12
Davis, Mo Mo, N.
13 .
30
1U
15o
l6o
17/
13o
Haskin, B C and Ea Ho Lee, "Flexible Surfaces on Viscoelastic Subgrades*" Journal of the Engineering Mechanics Division. American
Society of Civil Engineers, 1959
19o
20o
21o
22o
23o
24o
Palmer,
25
26
Quinna,
27
Sebastyan, Go Yo, 'The Benkelman Beam Deflection as a Measure of Pavement Strength," Proceedings^ Canadian Good Roads Association* I96O0
28o
Seed, H Bo and McNeil, Robert L,, "Soil Deformations in Normal Compression and Repeated Loading Tests," Highway Research Board Bulletin
1U, 1956o
31
29p
Spangier, U C, "The Structural Design of Flexible Pavements, " Proocedinps* Highway Research Boards 19A2
30*
31o
Walker, Ro B, "Significance of Layer Deflections in Evaluating Fleasible Pavements," Joint Highway Research Proje ct, Purdue University,
Lafayette, Indiana, January 25, 1961
32
33o
Winnitoy, Wo E 9 , "The GGRA Benkelman Beam Procedure and Its Application in Testing Flexible Pavements," Proceeding s^ Canadian Good Roads
Association^, 1960 o
34e
35
TABLE lo
Subgrado
Type
(AASHO
Class o)
Subgra da
Subbase
Base
Fall
Spring
A-l
5.1
9o3
3.3
4.3
1.4
4.0
A-2
4.6
3*4
1.4
5.0
3.2
3.7
A-4
3.7
69
4.1
108
a9
5.3
A-6
4.1
14.4
3.2
5.0
1.9
84
Fall
rt
Sprim
Fall
Spring
TABUS 2
Thickness of
Pavement
Type of
Pavement
Max* Permissible
Deflection for Design. F-urposes
(Tentative)
&-in
0,012-in o
6in
&>012~in<>
A-in
Asphalt Concrete
3-in e
2-in.
o 025-in8
l-in
0,036-in
Surface Treatment
0050-in
l/2~in
017-in
020-in B
ff;
ORIGINAL SIZE
(BEFORE
FINAL SIZE
LOADING)
(
AFTER LOADING
<Ta
g
tK.
FIG.I
ELASTIC ELEMENT
f2
- f&*
LU
Q
<
en
A.
i.0
CD
z>
to
I>0
<^
GQ
CE
*
<o
co
A.
4"
CO
CO
UJ
cc
Ico
UJ
Q.
<
_]
T
UJ
_J
a.
LU
>
UJ
>
<
0.
CO
UJ
o
UJ
z
o
M
< ^
uj
q:
^
co
<
_i
X
CD
X 3
o O
ICC
UJ
X
CO
CM
e>
u.
I-
MEASURED
DEFLECTION
A,
SURFACE
BASE
SUBBASE
SUBGRADE
FIG. 3
DEFLECTION
PROFILES
I-
LU
lJ
>
<
Q.
Q
Or
(/)
LU
C/>
CO
LU
d
\-
LU
cr
Z)
(S)
(/)
LU
cr
CL
UJ
>
o
<
UJ
cr
LU
_l
Li_
LU
z
o
h-
o
UJ
_J
u.
LU
(3
Li_
Tension
(o.)
->%-
rTension Crack.
^Tension
Crack.
(b)
FIG. 5
P= 12,000 LBS.
PRESSURE
TIRE
70
PSI
w 350
a.
E
UJ
Ul
x
5
300
a:
uj
250
2,000,000 PSI
o
co
co
UJ
or
S
200
/ 8
500,000 PS
ico
Ixl
_l
150
w
z
UJ
UJ
CO
100
or
UJ
>
co
z
<
50
1
0.01
A
FIG.
1
0.02
1
0.03
1
0.04
1
0.05
0.06
RELATIONSHIP
IN.
0.35
180
DEFLECTION
V)
160
X
o
UJ
140
0.30
o
OS
o
C/5
UJ
_l
120
UJ
0.25
100
10
SPEED
40
30
20
IN
50
60
70
//////////
clx
X"2-
J
(C)
LU
Lu
CE
o
UJ
o
o
UJ
UJ
H
LU
cc
o
z
LU
o
o
UJ
cr
<
X
Q.
w
<
U.
Z)
<
or
UJ
Q_
UJ
h-
UJ
a:
=>
H
<
CC
LU
CL
2
Ui
t-
u.
o
UJ
u_
u_
UJ
-W
oo
CD
HONI d3d
SdlX
OlJ-Vd
N0I1031J3Q QV01
EDGE
PAVEMENT
^^s
PAVEMENT
CENTER LI NE-^^
V)
UI
io
5
?
'5
o
o
20 6
Ui
UJ
_i
UJ
uj
o
cr
25
X
5
oJ
_l
<
Q
<
r>
Q
cr
UJ
h-
1
'
z
o
o
UJ
-i
HUI
cr
UI
O
oJ
FIG. 9
EXAMPLE
REAR
OF
TRANSVERSE
AXLE OF A TRUCK,
PAVEMENT
THICKNESS
18"
DEFLECTION
5"
AC,
UNDER
TOTAL
FIG. 10
40
80
60
100
U.S. 41
U.S. 52
PURDUE AIRPORT
CALCULATED
a=
0.5'
2.0'
= 4.0'
28
32
FIG.
DEFLECTION
PATTERNS
ET AL
NOTES:
I.
2.
9,000
LB.
LB.
PER TIRE)
5"
ASPHALT CONCRETE
WATER BOUND MACADAM
5" SANDY SUBBASE
SUBGRADE A-6 AASHO CLASS
8"
3.
DISTANCE
FIG. 12
FROM
CENTERLINE
OF DUALS
FEET
TESTS
IS
AT
AVERAGE OF
16 LOCATIONS
0.024
0.020
CO
UJ
0.016
"""~
UJ
o
<
UJ
z
o
o
<
0.012
o
UJ
0.008
CO
z>
V)
QT
or
<
3
z
<
-I
0.004
or
co
CD
UJ
o
<
or
e>
<
-i
3
Z
<
or
o
Q
<
<s>
CO
CO
UJ
UJ
>-
<
_l
o
>1_1
co
CM
CO
<
<
<
UJ
or
<
o
m
3
or
PAVEMENT
SHOWING
LESS THAN
25 LINEAL FEET OF
CRACKING PER STATION
FIG.
13
DEFLECTION AND
or
or
<
_i
3
Z
<
or
CD
<
or
CO
CO
CO
3
to
to
0.00
o
<
or
<
_i
3
Z
<
or
e>
CM
V
<
_i
o
>-
to
CO
1
<
<
PAVEMENT SHOWING
MORE THAN
80 LINEAL FEET OF
CRACKING PER STATION
SUBGRADE
TYPE
PAVEMENT EDGE
PAVEMENT CENTERLINE
UJ
2
UJ
>
o
5
ll
O
z
o
Io
UJ
sr
NOTES:
I.
2.
3.
4.
SPLACEMENTS
FIG. 14
CONTOURS
OF
DEFLECTION
IN.
130
120
GRAVEL SUBGRADE
CL
O
o
o
100
in
=>
80
70
LU
60
>
I<
UJ
a:
SILTY SUBGRADE
90
BASE
50
y
SUBGRADE
40
30
20
SUBBASE
10
^i
*i
15