xt7vdn3zwd7k https://exploreuk.uky.edu/dips/xt7vdn3zwd7k/data/mets.xml  United States. Federal Highway Administration United States. Bureau of Public Roads United States. Federal Highway Administration. Offices of Research, Development, and Technology United States. Federal Highway Administration. Offices of Research and Development 1939 v.: ill., ports.; 30 cm. UK holds archival copy for ASERL Collaborative Federal Depository Program libraries. Call Number FW 2.7: 20/5 journals English Washington: U.S. Federal Highway Administration etc. Contact the Special Collections Research Center for information regarding rights and use of this collection. Works Progress Administration Transportation Publications Roads -- United States -- Periodicals Highway research -- United States -- Periodicals Public roads: a Journal of Highway Research July 1939 text Public roads: a Journal of Highway Research July 1939 1939 2019 true xt7vdn3zwd7k section xt7vdn3zwd7k ”I m 45,514«r.v~w~—~~y—~—:w”‘-‘rn*w».-.' “~'w~3:;:: " WET" . “”T'“ ”1 ““7““ ' "tr-:37"? 4*“ , 1, .
" F w v .n/Ir‘ ' . '

. “""II .1.an II""’""" I I “III l|""""' "III II III “'"III "' ”M
III - I IIIIIIII I II IIIIII I III III"

' IIIIIID

IIIIIIIIIIIIIII IIIII.. .mlIII .aIiII IIIIIIII... .....IIIIIIII "Int 11 lllll'll|., LAIIIIII .IIIIIIIIIIIIIII ..IIIIIII In... All” M If"

,, I
I A JOURNAL OF HIGHWAY RESEARCH _ ,
i ' 7,1
I FEDERAL WORKS AGENCY
I , PUBLIC ROADS ADMINISTRATION "
I . .- —: x I
I ~, Vy/OI—IIIQOI,L.NbJ35' ‘ I, v JULY 1939
V‘ ..r'ntI i
' 1‘». N ' I '3- ‘" "i » 1"" " ,, I
7%: fix 35:14 It: , ’- 1W“ 2‘ t- r . ;
‘ *3 5 . ‘-.. .. "" 1 « ‘f" ~ Ir'i¢.'.5~.~"‘ ‘ .- f f“ 2
I Mia, 5,4". as gs, ‘. » ,, “" " .. " 12‘ k v " f" “If.“ ' :
, f I" ~, ~ ~ V , » ‘
I Staff ~ :5 ~ ‘ W I”: ‘ I
. . . _. .1 f4" __ J .V .1 ,,,, .. .
.‘v y“.- 9'. “a, . ”a -' I ,..‘:‘—_ I
I A SECTION OF STATE ROUTE 96 IN NEW YORK ‘

I I"W sale by the Superintendent 0! Documents, Washington, D. C. - - — - - - - - - - - See page 2 of cover for prices
I

 ? . Fl . T i i -- T . I . . - _ ' ' . Ta; 3:41”?!
_____________—__.___—_—.____————-——-—-—‘———————‘—'—_—‘
_______________—____________—_____

Al
fiDfi >>>Afournalof»
. 1 l
' Hzgnway Research
I
I s s u e d b y th e
FEDERAL \X/ORKS AGENCY I
’ PUBLIC ROADS ADMINISTRATION D I
8
~ D. M. BEACH, Editor and tl
Volume 20, No. 5 JUIY l939 scatteI

‘ these I
The reports of research published in this magazine are necessarily qualified by the conditions of the tests from which the data are obtained. 313:5 (
Whenever it is deemed possible to do so, generalizations are drawn from the results of the tests; and, unless this is done, the conclusions 0:361?

formulated must be considered as specrfically pertinent only to described conditions. vali di t
—————————————_& value a
‘ In in
Single I
- cOnside
In This Issue . deallng
The ne
‘ Page have b(
Application of the Results of Research to the Structural Design of Concrete Pavements . 83 icrliaiifid f
. e I
. of this I
availab:
the Imp
1n reset
. . . . concretr
THE PUBLIC ROADS ADMINISTRATION . - - - - . WIllard Bulldlng, Washlngton, D. C. In .11
REGIONAL HEADQUARTERS - — - — ~ - - - - — - Federal Building, Civic Center, San Francisco, Calif. Elples 01
. 0 man)
v veys 13hr

matIcaI
DISTRICT OFFICES Several]
racy 6x1:
DISTRICT No. 1. Oregon. Washington. and Montana. DISTRICT No. 8. Alabama, Georgia, F lorida, Mississippi, and Tennessee. COIlCI‘ete
Post Office Building, Portland. Oreg. Post Office Building, Montgomery. Ala. From
DISTRICT N0. 2. California. Arizona, and Nevada. DISTRICT No. 9. Connecticut. Maine. Massachusetts. New Hampshire, New pave/met
Federal Building, Civic Center. San Francisco. Calif. Jersey. New York, Rhode Island. and Vermont. p-Oliffd 1)

o u' 'n an . . Cl
DISTRICT No. 3. Colorado. New Mexico. and Wyoming. 505 P St 0m“ B 'ld' 3' Alb 3" N Y Signey :1
254 New Custom House. Denver. Colo. DISTRICT No. l0. Delaware, Maryland. Ohio, Pennsylvania. and District the saijhe
DISTRICT No. 4. Minnesota, North Dakota. South Dakota, and Wisconsin. Of Columbia. Willard Building, Washington. D. C. fQI‘CeS pr‘
907 Post Office Building. St. Paul. Minn. nltude 0'
. . DISTRICTN . ll. Al k . '
DISTRICT NO. 5 Iowa" Kansas' Mlssoun' and NebraSka‘ o as aRoam 4l9. Federal and Territorial Building. Juneau, Alaska. StTOISS-es E
Masonic Temple Building, Nineteenth and Douglas St.. Omaha, Nebr. ISTRICT N 12 Id h d U h addltlon’
. . t .

DISTRICT No. 6. Arkansas, Louisiana, Oklahoma, and Texas. D o a 0 an a Federal Building, Ogden, Utah. :31!“ 28563;]
R°°’“ ”2'1““ 5““ °°“"h°“s°' Fm WW“ Te" DISTRICT No. 14. North Carolina, South Carolina, Virginia, and West of the di:
DISTRICT No. 7. Illinois. Indiana, Kentucky, and Michigan. Virginia. great dea'
South Chicago Post Office Building. Chicago. Ill. Montgomery Building. Spartanburg. S. C. ever, on 1'
——-————————-————-——-—————— assumptic

. . . . . . ”151% 9.
Because of the necessarily limited edition of this publication it is impossible to dIstrIbute It free to any person or InstItutIon other satlsfactor
than State and county officials actually engaged in planning or constructing public highways, instructors in highway engineering, Structu.
and periodicals upon an exchange basis. At the present time additions to the free mailing list can be made only as vacancies occur. use Of C01
. Those desiring to obtain PUBLIC ROADS can do so by sending $1 per year (foreign subscription $1.50), or ID cents per single copy, Stigfil’bar e '
to the Superintendent of Documents, United States Government Printing Office, Washington, D. C. 80-0511125)";
————————————————————————————— PTOVIde f0
CERTIFICATE: By direction of the Commissioner of Public Roads, the matter contained herein is published as administrative information m
and is required for the proper transaction of the public business. March 1939, B

PUBLIC ROADS.
- 157244-

 m ~41. ~»‘»‘*:-rs.~......;::. ' -- were» ~ : ~— :'~' >-.“-— ..
I SEARCH TO THE STRUCTURAL DESIGN
OF CONCRETE PAVEMENTS1
E .‘ Reported by E. F. KELLEY, Chief, Division of Tests, Public Roads Administration
i URING the past 20 years many studies have been impossible to make definite provision. In contrast to ' ,.
‘,‘. D made of the various factors that influence the this the current designs of concrete pavements are
structural performance of concrete pavement slabs generally such that the factor of safety, if any, is so _ '
and the numerous reports of these investigations are small as to be almost negligible. '
scattered through the technical literature. Most of The maximum combined stresses due to external
these reports, of necessity, are highly technical and the loads and to temperature in pavement slabs of the
mass ofdata presented and the detailed descriptions dimensions commonly used will very frequently be '
that are included, both as a matter of record and in found to be so close to the ultimate strength of the
order that the reader might have confidence in the concrete that there is little or no margin left to provide ,
, validity of the results, frequently tend to obscure the for unknown or unforseen conditions. In making this
value and importance of the conclusions. statement there is no intention to imply any general
‘ 4; In addition, individual reports frequently cover but a criticism of present practice since the present standards
single phase of a given subject and are useful only when of design have proven reasonably adequate. When
L censidered in connection with the available reports the need for the great mileage of existing pavements
‘; dealing with the remaining phases of the same subject. and the fact that structural failures of these pavements
The net result of this situation is that many facts that do not generally endanger human life are considered,
é have been well established by research are little appre— it seems probable that any significant increase in cost
ciated and too frequently are given scant consideration to provide a margin of safety comparable to that
in the practical design of pavements. It is the purpose provided in bridges, could not have been justified from '
,5; of this paper to bring together under one head and make the economic standpoint. However, it is important to
3,? available for the practical use of the designing engineer recognize that the low or negligible factor of safety that
g the important facts that have been developed thus far is provided in designing concrete pavements makes it .
. i in research work relating to the structural design of highly deSIrable to be somewhat conservative in assum-
i concrete pavements. ing design values for the different variables that must
- 2 In the field of bridges and buildings the basic prin— be considered.
:. 7:3. ciples of design have become so well established that, IMPACT REACTION DEPENDENT 0N FOUR VARIABLES
to many engineers the term “structural design” con— Wheel loads and impact.——Neglecting the unpredict-
‘:«-. veys_ the Idea Of a rather exact and accurate mathe— able forces caused by localized differential heaving or ,
If, matlcal procedure to be followed 1n proportiomng the subsidence of the subgrade soil, the external forces that
- ; several parts of a‘structure: No such presumed_accu- create stress in the pavement slab are produced by
5;} racy ex1sts 1n connectlon Wlth the structural des1gn Of vehicles. Naturally, the heavier vehicles are the more
.. ,3 concrete pavements. _ important.
l“ From the standpomt of stress analysis the concrete One of the earlier investigations (1)2 developed the
m pavement 1s a highly complex structure. It is sup- important fact that for heavy vehicles of the usual type,
l ported by scil whose physmal properties vary appre- that is, four- or six-wheel trucks or trailers, the critical
Y. elably at different .10031310113, at difierent pomts 1n the stress developed in a concrete pavement, when the axle
. same general lOCfithn, and even at different times at spacing is in excess of about 3 feet, is primarily a func-
m l the same p01nt. It is subjected to the action 0f external tion of the wheel load and not a function of the gross
.c. f forces produced by the wheels of vehicles and the mag- load on the vehicle or the axle spacing. By means of
mtude Of these forces and their effect on_pavement his theoretical analysis, Westergaard (2) subsequently
ka ; stresses are Influenced by a number of varlables. In arrived at the same conclusion and this has been
‘ i addltlon, it is constantly subJected to hlgh internal confirmed by later tests (3), This finding, which
he: stresses produced by changes In temperature and m01s- permits attention to be confined to wheel loads rather
"' ture. Much has been learned concerning the influence than gross loads, greatly simplifies a problem already
lest ; of the different variables on pavement stresses but a sufficiently complicated.
~ c great deal of additional research is still needed. How- The magnitude of the vertical force exerted on a
3' i ever, on_the basis of available information, reasonable pavement by the wheel of a moving vehicle may be
.— l assumptions of sufficient accuracy can be made to considered to be the sum of the static weight of the
3. Insure a pavement structure that will function In a loaded wheel and the additional impact or dynamic
,heri satlsfactory manner. force created by the movement of the wheel over the
lug; . Structural des1gn, in general, is distinguished by the irregularities that exist in the pavement surface. The
cur” use of conservative unit stresses which, for structural researches of the Bureau of Public Roads have demon-
)py,5: . steel, are well below the elastic limit and, for concrete, strated conclusively that the impact reaction of a ‘
, .3 well below the ultimate strength. This results in the moving wheel is sufficiently in excess of the static wheel
SO-Cillled factor of safety which is depended upon to load to make it an important factor in pavement design.
:3 WLdefm all the unknown conditions for which it is The impact reaction of a moving wheel depends upon
i Mxpape, presented a, the mm, "mm of W American Concrete Institute, four m'ajor variables—wheel load, vehlcle speed, t1re
l Ptiggib Egan Bec’i‘ause of its length. this report will 'be presented in two issues of —2—-——:- , _
. 5. he second installment will appeal m the August issue. Italic figurm in parentheses refer to the bibliography. p. 102.
J 157244—39——1 83
f2 . .
’\

 . , ...-....w—,;;._.

i 84 PUBLIC ROADS Vol.20,No.5 July
M_#———a—_—a———-——————-;——————

» equipment, and road roughness (4). Other variables INTENSITYOFIMPACT DECREASES ASFREQUENCYOFOCCURRENCE (fig
exert some influence but, in general, these four are the INCREASES ~ seeDf
important ones. An increase in Wheel load or pavement Another fact with respect to the effect of tire equip- on
roughness; a decrease in the cushioning qualities of the ment is that dual tires generally give somewhat higher exec
tires; and, within limits, an increase in vehicle speed; impact reactions than do single tires of the same type T

3 all result in increased impact reactions. and same load capacity. The difference is a vari- effe(

' The tests that have been made have amply demon— able which, from the practical standpoint, may safely mag

" strated the fact that the magnitude of the impact reac— be ignored since the increased stress in a concrete pave- tras
tion is a function of the wheel load. Also, these tests ment slab resulting from the greater impact effect of 31m,

‘, ' have brought out important facts, not previously dual tires may generally be expected to be more than degr

1 known, regarding the relation between wheel load and offset by the reduction in stress resulting from their roug
the impact reaction that it produces. In bridge design greater area of load application. For example, if it be test

j it is customary to express impact as a percentage of the assumed that a certain wheel load on dual high—pressure tire

1 static live load. Therefore it is important to observe tires produces an impact reaction of 10,000 pounds speei

i - that While the total impact reactions of the wheels of then the minimum reaction that may reasonably be It

3 motor vehicles increase with increase in Wheel load, the expected from the same load on a single high—pressure there

, percentage of impact, or the ratio of the dynamic incre- tire of comparable capacity would be of the order of nituc

,1 ment to the static load, actually decreases as the wheel 9,000 pounds. With reasonable assumptions as to 000m

; load is increased. This fact may be attributed largely area of tire contact and other variables the computed 000m

l to the relative effects of sprung and unsprung weights, stresses, by the original Westergaard analysis (2), for mag:
and t0 the relation between size of tire and its cushioning loads applied at the interior of a 6-inch slab, are inten

I properties. about 330 pounds per square inch for the 9,000-pound inten

. The force which the Wheel of a vehicle delivers to the load on the single tire and about 315 pounds per square inere;
road surface is made up of two component forces. One inch for the 10,000-pound load on the dual tires. equip
of these is caused by the unsprung weight on the wheel When a wheel runs over an obstruction, such as an 40 m

‘, (that is, the weight of the parts not supported by the inclined plane or a rectangular block, two types of . was fl

' springs), and the other is caused by the spring pressure vertical impact reactions are developed. One is caused reacti

‘ on the axle at the instant of impact. The part of the by shock as the wheel strikes the obstruction and the 80, a1

1‘ total impact reaction caused by the unsprung weight other is caused by the drop of the wheel from the ob- 1.65,

i is, in general, considerably greater than the part caused struction to the pavement. In the earlier investigations tude c

, by the sprung weight. However, the ratio of unsprung involving pneumatic tires operated over artificial less as

i weight to total Weight is not a constant but decreases obstructions at speeds up to about 55 miles per hour (5), impac

~i as the total or gross weight is increased. Also, as the it was found that the shock reactions increased approx- opera!

wheel load is increased the tire size is increased and imately in direct proportion to speed. It was also smoot

5 with it the ability of the tire to minimize the effect of found that drop reactions reached maximum values at and 1

i , surface irregularities. The result is that for a given relatively low speeds, of the order of 25 to 35 miles respec

5, condition of road roughness an increase in wheel load per hour, and that these were not exceeded by the shock It i:

i is not accompanied by a corresponding percentage reactions except at speeds of the order of 50 miles per tWeen

1 increase in the dynamic component of the impact hour. In a subsequent investigation (6) involving only from t:

l reaction. balloon tires, it was found that the use of artificial t0 sele

11 The magnitude of the impact'nforce is greatly de- obstructions resulted in maximum drop impacts at dynam

1 pendent on the type and condition of the tire equip— speeds of from 20 to 40 miles per hour and that these responi

ment. Solid, cushion, and pneumatic tires, in the were not exceeded by shock impacts at speeds up to f0! a Ir

; order named, produce impact reactions of decreasing 70 miles per hour. would

magnitude. The tests that developed this information From these tests with artificial obstructions it might necessa

were made at a time when rubber tires of the solid and be concluded that the effect of speed on impact reac- sufficiez

l cushion types were commonly used. Fortunately, tions is not important for speeds in excess of 40 miles frequer.

‘ these types are no longerin general use. The relatively per hour. However, such a conclusion would require mum ii

few solid tires that are now used must be operated at some modification as a result Of the tests (6) that have 011 an 2

5 such low speeds that, in comparison with the pneu— been made to determine impact reactions resulting from reasona

‘, matic tires used on high-speed trucks and busses, they the natural roughness of road surfaces. These tests Thei

i need be given 'no consideration from the standpoint were made at 28 locations Where the natural roughness any SH};

1 of impact. Therefore attention may be confined to was as severe as would permit the safe operation offl been d1:

I pneumatic tires. , heavy vehicle at high speed. In each of these 28 loci- been sti

j With respect to pneumatic tires it has been found (5) tions the shape of the curve of impact reaction versus mterpol

,1 that, other conditions being the same, the dynamic speed was different depending on the characteristics of 111 the _ri

increment of the impact reaction of high-pressure and the particular roughness condition. , as to gm

1 balloon tires is closely proportional to their inflation In some cases the maximum impacts were observet Present

pressures. Therefore, it follows that for a given wheel at relatively low speeds but in the majority of cases thl Clently :
load the impact reaction created by low-pressure impact reactions showed a general tendency to increasl fa,ClJOI‘S 1

. ‘ balloon tires is appreciably less than that caused by with increases in speed up to the maximum of 70 mile; Wlth du:

l high-pressure tires. From the standpoint of pavement per hour. However, this statement applies to indl- miles pi

i protection the balloon tire offers the additional im- vidual locations. When all the maximum impact reao degree 1

l portant advantage that it apphes the load to the pave— tions were plotted against speed it was found thfl? extreme]

1 - ment over a larger area Of contact, a condition that a general maximum was reached at about 50 miles 1391 “9811761

' results in a lower slab stress. This relation will be hour and that this remained constant up to 70 milt The p

I discussed in detail later. per hour, the maximum speed attained in the test 3:: 11,1311

1 W1

 . ~m'—»»—,.il’!~ ,2 g "r-‘*__.“"“"77 w__- “'~'~..,~wr- umvé‘mx;—-“W~ ——.A..~.w ‘ " e e" '7'5" . 3"“ _—“*~—~ .— .._"‘".,.. - ' \
$0.5 Ju1y1939 PUBLIC ROADS 85
2 ”Km
“CE (fig. 24, PUBLIC ROADS, Nov. 1932). Therefore, it interesting to observe that, with minor exceptions, the
‘ seems reasonable to conclude that the effect of speed order of rating would have been the same had they been
ip- on impact reaction may be neglected for speeds in rated for roughness by means of the impact—frequency
her excess of 50 miles per hour. curves. In other words, the roughness indicator gave
vpe Two investigations have been made to determine the a qualitative measure of the characteristics of the pave-
Mi- effect of conditions of general road roughness on the ment surface that determine the magnitude of impact.
?ely magnitude of impact reactions (6, 7). This is in con- However, while the roughness indicator is a useful in-
.ve- trast to the study of extreme conditions of roughness strument, it is not one of precision. As it has com— ,
2 of already described. In these tests, roads of various monly been used the motor vehicle on which it is
han degrees of roughness, as determined by the relative mounted becomes an integral part of the instrument
neir roughness indicator (8), were selected for study and the , and the results are reproducible only with the same car
)be test vehicles with different Wheel loads and different operated under the same conditions. Therefore, while
iure tire equipments were operated ovor them at various a given instrument mounted on a given car gives a
.nds speeds. qualitative measure of the relative roughness of differ-
‘ be It was found that, other conditions being the same, ent road surfaces, it is not possible to express these
sure there was a rather definite relation between the mag- results in absolute figures.

r of nitude of the impact reaction and the frequency of its
to occurrence. Of the great number of impacts that may TABLE 1-‘rfmpact factors and total impact-Toad reactions
ited occur on a given section of road, those of the greatest Speed—50 miles per hour.
for magnitude occur only a few times while those of lesser Frequency—100 per mile.
are intensity occur a greater number of times and the Condition of pavement surface—reasonably smooth-
.und intensity decreases as the frequency of occurrence a
uare increases. For example, in the tests with a motor bus Dual high-pressure Dual balloon tires
equipped with balloon tires and operated at a speed of “”5
s an 40 miles per hour over a very rough concrete road, it Staticwmelloadvpounds Total Total
58 of , was found that the impact factors (ratio of total impact 1333:: impact 133:3? impact
used reaction to static wheel load) for frequencies of 1, 40, _ ”3‘3“” mam”
l the 80, and 100 times per mile were approximately 2.20, Pounds Pounds
3 ob- 1.65, 1.55, and 1.50, respectively. However, the magni- 4000 2.05 8,200 1.70 6,800
tions tude of the impact factor for a given frequency becomes 3:833::t:::::::::::::::::::::::::::::: iii? 13:388 iifi $3133
fieial less as the roughness of the pavement decreases. The gggg fig 19283 {3; 13,235)
r (5), impact factors for the same vehicle as described abOVe, 0:000:22:iiIIZiIIIZIIIIZIIIIZZIZI 1341 12,’ 700 1.‘ 27 111400
prox- operatfid at the same speed of 40 miles per hourl over a 10000 1'36 131600 1- 24 121400
smoot concrete avement were a rox1mate 1.25 xx
631:: and 1.18 for freguencies bf 1 angp 100 perymile, The tests that form the bas1s for the data given in
miles respectively. . table 1 were made on pavements that appeared to repre-
;hock It is immediately apparent frOm this relation be- sent reasonable average conditions of surface roughness, »
,5 per tween frequency and magnitude of impact factors that, intermediate between extremely smooth .and extremely
. only from the standpoint of pavement design, it is necessary rough surfaces. A more precise definition cannot be
ificial to select some reasonable frequency and to compute given. on account 0f thls variable and the others that
ts at dynamic loads on the basis of the impact factor cor— affect the magnitude Of the impact reactions, the data
these responding to this frequency. Designing a pavement given in table 1 can be considered only as approxunate.
up to for a maximum load that may occur only once per mile They represent the best estimate thht can be made, on
would certainly be open to serious question and it is the basis Of eXIShmg data, Of the memum impact r eac-
[night necessary to select an impact force that occurs with tions, important With respect todes1gn, that can reason-
reac- sufficrent frequency to be of practical importance. A ably be expected to occur as the result 0f the normal
miles frequency of 100 per mile, corresponding to the maxi- operation of the heav1er_motor vehicles. . The digit in .
aquiie mum impact reaction that may be expected to occur the second decimal place In the figur es for impact factors
~.have on an average Of once every 50 feet, is suggested as a 18 Without Significance. _ It is included merely for the .
r from reasonable assumption. purpose of making the impact factors agree With the
, tests The existing data do not permit the evaluation, from total nnpact reactions which are given to the nearest
rhnesi any single series of tests, of all the variables that have hundred pounds.
ii of i been discussed. However,» some of the variables have IMPACT FACTOR USED SHOULD BE INDEPENDENT or POSITION or _
3 100,, been studied in each series of tests and it is possible, by _ POAD
versus interpolation and extrapolation, to combine the data AS W111 be ShOWIl later, m {i 00110113136 pavement Slab 9f
tics 0] 1n the reports that have been mentioned (4, 5, 6, 7) so uniform_thickness the magnitude of the critical stress is
as to give impact factors that are in agreement with our greatly influenced by the p0s1tion of the wheel. load;
served present knowledge of the subject and which are sufli- that is, whether it is near an edge, a corner, or 1n_the
3es till Clently accurate for purposes of design. Such impact center of the slab. Since the higher impact reactions
mew factors for a range of static loads on wheels equipped Will be produced at the_p01nts where the surfaceirregu-
) mile With dual high-pressure and balloon tires, a speed of 50 larities are greatest, it follows that higher impact .
) indi' miles per hour on a pavement having a reasonable reactions may be expected in the vicmity of transverse
t reat degree of smoothness (neither extremely rough nor Joints and'cracks than'm the interior of the slab. In
d thug extremely smooth), and a frequency of 100 per mile, View of this cons1deration Bradbury (.9) has suggested
les pg are~gIVen in table 1. that a higher allowance for impact be. made in the
) mile The pavements on which impact-frequency studies computation of stresses at transverse Jomts than in
e test Were made were rated with respect to degree of rough- other portions of the slab. However, n1 plain (non-
ness With the relative roughness indicator (8) and it is reinforced) pavements transverse open cracks are

 j 86 PUBLIC ROADS Vol.2o,No.5
1 quite likely to develop at random, except in very It is a well—known fact that stresses above the fatigue
short slabs, and thereby create a roughness condition limit cause progressive inelastic deformation and final

3 similar to that at formed joints. When this takes failure. However, the relation between intensity of

f place in a thickened-edge slab a condition of weakness is stress above the fatigue limit and the number of repe—

I created at the broken edge of the slab along the crack titions of this-stress that will cause failure is not well

° that makes it desirable to overdesign rather than und er— established even for rapid repetitions of stress. For less

I design the thickness of the pavement. frequent repetitions nothing is known concerning it.

1‘ Also when a truck wheel leaves the edge of the pave— On the majority of highways the heavier vehicles

‘ ment and then rolls back on the slab from a shoulder constitute a small percentage of the total traffic and

1: that frequently is not at the same elevation, an impact therefore the occurrence of maximum load stresses is

I reaction of considerable magnitude may be developed. relatively infrequent. It appears therefore that the

I These considerations lead to the conclusion that nice present practice of assuming the design stress to be

‘ distinctions with respect to the position of the load on approximately 50 percent of the ultimate strength 0'

I , the pavement are unwarranted, and that the same impact the concrete is a conservative one insofar as the stresses

II factor should be used irrespective of the position of the due to maximum Wheel loads are concerned. In view

I load. of the possibility that the fatigute liln'lih for filiese infrd'e-

I. quent repetitions 0 stress may e 11g er t an is in i-

.4: DESIGN STRESS EQUAL To£§§§§fffi$ ULTIMATE STRENGTH IS cated by available data, this practice may introduce

I} . _ . . some factor of safety of unknown magnitude.

:3 Fatigue lemzt' of Generate—Concrete, hke other However, the limitation of the design stress to 50

‘: structural materials, Will fall under repeated loads at percent of the ultimate strength is believed to be

I? unit stresses which. are much less than the. ultimate unduly conservative when the pavement slab is de-

, strength as determined by thBIS'fireSS at failure PTO‘ signed for the combined effect of stresses due to load

" duced by one application of static load. The stress at and those due to temperature warping since, as will

i which failure takes place under a very large number of be shown later the maximum combined stresses due

~‘II loadings is known as thefatigue limit or the endurance to load and teinperature occur only in the daytime

fII limit and, for concrete, it is expressed as a percentage during the spring and summer months. It is apparent,

I 0f the ultimate strength. , , , _ therefore, that the frequency of occurrence of maximum

:Ii Investigations 0f the fatigue hmlt 1h flexure under load stresses in combination with maximum tempera-

.II Stahc load (10’ 11’ 12) have Shown that concrete may ture stresses is considerably less than the frequency

{I} be subjected to an almost unlimited number of aPPh' of passage of the truck wheels that produce maximum

‘5' cations Of a stress. equal to about 55 percent 0f “5,3 11,1'51‘ load stresses. This is particularly true on those high-

II mate strength without danger 0f failure. A similar ways where the movement of heavy trucks is princi-

II study of the fatigue limit of concrete under impact pally at night.

II loads (13) gave Slmhar 119511th although the maXimum In attempting to establish safe unit stresses for use

‘!I number 0f load applications was ,OIhY about 83,000 as in the design of concrete pavement slabs several factors

II compared With the one or more millionthat are usually in addition to fatigue should be considered and these

II cons1dered des1rable 1T1 fatigue studies. From_th1s will be discussed later. It is sufficient here to point

II study lb was concluded that, Wlth respect '00 fatigue, out that the many uncertainties regarding the fatigue

I: tberehaVlol‘ 0f concrete may be assumed to be very characteristics of concrete render of doubtful value any

i? slmllar under both static and impact loads and that the refinements in the use of existing data.

I; same fatigue limit is applicable to both. _

Oh the baSIS Of these investigations It has become STATIC LOAD STRESSES MAY EXCEED IMPACT LOAD STRESSES

j rather general practice to assume about 50 percent of

I the ultimate flexural strength as a safe value of the Static stress versus impact stress—With respect to

j‘ working stress for use in designing pavements to resist the relative stress effects of static and impact loads,

I wheel loads. However, the fatigue limit of the order exhaustive tests by the Bureau of Public Roads (as

of 50 percent of the ultimate flexural strength of the yet unpublished) have shown that static and impact

1 concrete has been established by tests in which the load forces of the same magnitude, applied through rubber-

% applications were repeated at relatively short time tired truck wheels, produce approximately equal strains 1
intervals, as many as 40 per minute in tests in which the in concrete cantilever beams that are free to deflect. ]
I? loads were applied without shock. In contrast to this, The procedure followed in making these tests has been a
II und