202

Comparison of Prediction Equations for

Bridge Pier and Abutment Scour
J. STERLING JONES

ABSTRACT uation, they have been studied separately ând neeal

to be predicted inclependently in a design situation.
The tno conponents of prinary interest are pier
There are at Least 10 prediction equations and abutnent scour because they are direct conse-
for bridge pier scour, and designers are guences of bridge obstructions to waler flon and afe
often at a loss over which one to use. There therefore the primary responsibility of highway
are only three or four prediction eguations agencies. Aggradation and degradation are often the
for abutnent scour, but these have not been predoninant components of scour but usually âre
highly publicized. The pier scour equations site-specific phenornena associated with a streamrs
faLl into three basic categories: those of reaction to meander cutoffs, effective slope
the University of lowa, those of the changes, downstream rnining, reservoirs, and so on.
Colorado State University, and those based Àggradation ancl degradation are probabty best pre-
on foreign literature. The sinilarities dicted by a sedinent transport nodet such as the
anong the pier scour eguations and the range Corps of Engineers HEC-6 (617), the Chang nodel (7),
of ¿lata on whích they are based are shôvrn. or the Si¡nons-Li nodel Q). Contraction scour mây
FHI{A sponsored several studies during the occur naturally because of narrowing of the flood-
1970s aimed primarily at comparing field plain or may be bridge related because of the en-
data with the various equations to show croachment on the floodplain by e¡nbanknents. Abut-
which ones best predícted local scour for ment scour is a concentrated part of contraction
U.S. streams. These studies Írere sonewhat scour that can be accountecl for by empírically clis-
inconclusive because of the rnany inter- tributing the scour Ín a waterway opening.
related variables in the scour process, but
they do show which equations are conserva-
tive and whÍch are not. NCHRP SYNTHESIS ON SCOUR AT BRIDGE WATERWAYS

In 1969 a synthesis of available 1Íterature and

practices for dealing with scour at bridge watersrays
was made by the Eighway Research Board. More than
Scour is the term used to describe erosion phenonena 100 orqanizations, including highway agencies, to11
that involve unified flow patterns such âs those at road agencíes, consultants, railroad conpanies, and
bridge piers, abutments, and outlet structures. government agencies, were surveyecl.
Scour damage to highway structures has been esti- The synthesÍs report G) cited 12 bridge píer
mated to be as high as S20 rnillion per year G). scour prediction equations, but in the díscussion of
Years of research have been alevote¿l to resolving the the prediction equations it yras concludeil that it
problen of scour at brÍdges; yet in spite of the was (5, p. 14) "quite impossible to build a feeling
magnitude of Èhe damages attributed to scour, FHWA of confidence in any prediction methodn because of a
has never published an engineeríng circular bringing lack of field measurements with which to compare the
together the literature and giving guidance on hovt predictions. The only guidance given for selecting
to account for scour at bri¿lges. Lacking an en- the right prediction equation x¡as to check the back-
gineering circuLar on the subject of bridge scour, ground of each equation and examine the variables
there appears to be no better forum than â nationaL included ín each eguation.
conference among leading bridge engineers to reflect The synthesis report cited only one abutnent
the knowledge that has been gained on this pertinent scour equation. That \ras Laursenrs equation in an
topic. appendix written by Laursen hinself.
In a general sense, scour as defined by ASCE in
Manuals 43 12) and 54 (f) is the erosive action of
running water in streams that excavates and carries FHWA FOLLOW-UP STUDTES
away rnaterial fro¡n stream beds and banks. ltany
highway engineers rel-ate to this definition and FIIWA sponsored several studies in thê I970s after
identify alI stream erosion as scour. A preferred the synthesis aineil at inproving conficleñce in
NCHRP
definition restricts the term to vertical stream sone of the prediction nethods. The West Virginia
erosion 1415), thus distínguishing stream-bed ero- UnÍversity study (8) had Èhe objective of developing
sion from lateral stream nigration. instrumentation and collecting fíeld tlata for scour
For the purposes of this paper it is useful to around bridge piers. The researchers founct that it
look at the components of scour and to focus on vras less of a problem to dêvelop instrumentation
those cornponents that are primarily related to than to deploy it in a flooding envirorunent so that
bridges. The cotnponents are as follows: ít would be operational when needed.
Anderson (9) rearranged the equâtions to nake
I. Local pier scour, then as si¡nilar in format as possible so as to fa-
2. Contraction scour, cilitate an analytical cornparison. He reco¡nmended
3. tocal abutrnent scour, and that sotne large-scale laboratory studies be con-
4. ceneral aggradation and degradation. ducted to coreplement the fielcl tlata, especiglly for
extrapolating eguations beyond the range of the
Although â11 these co¡nponents may occur simultane- original tests. He h'as able to rearrange the pier
ously and are probably interrelated in a field sit- scour equations in terms of one br nore of three
Jones 203

dimensionless variables--flow depth and effective <l/b = 1.s4 (yolb)o'' Fo

" for F < F" (7b)
pier width, Froude nunber, and shear stress and
critical shear stress. In other words, these Èhree where Fc is the critical Froude nunber when sedi-
variables are the key factors that govern the pier ment transport is pending. [For 0 < (F - Fc)
scour process. Following Andersonrs reconmendatíon, < 0.20 use larger value fron both equãtions.l The
FllwÀ sponsored a study at the University of lowa procedure for conputing F. is as follows:
(p that had the objective of extending the range
of the key factors, especially the Froude numbert 1. Estimate the median diameter (dSO) for the
which includes flow velocity and clepth. bed naterial'
2. Determine r" fron Eigure 2.44 of AscE ManuåI
54 (3),
PIER SCOUR EQUATIONS 3. Cotnpute U*c = rc/Þ,
4. conpute 6 = lf.$e/U*. (assume v = 1.08 x l-0-5
The pier scour equations can be grouped into three ftz /secl ¡
basic categories for conparison. one category is 5. Compute d5O/6,
the group based on foreign research, primarily in 6. Select x from Figure 2.97 of. ASCE Manua1 54
Pakistan and India, where fine be¿l naterials are (3),
prevalent. A second category is patterned after the 7. Conpute y = 12.5 In(11.02yxld5g)l U*.¡ and
University of Iowa hypothesis that .depth of flow is 8. Compute Fc = Yc/ (gYclt/'.
more inportanÈ than velocity in sediment-transport-
ing pier scour. The third category is patterned
after Èhe work at Colgrado State University (CSU) Pier Scour Formulas Patterned After CSU Research
and includes velociÈy (expressed in a Reynolds num-
ber or Froude nurnber) as a pretlonínant term. shen r f1969 (17) I
The equations, identified by primary developer, óre
are shown in the following paragraphs. 4 =o.ooo73Reo (s)

where Re is the pier Reynolds number, Yb/v, and u is

Pier Scour Formulas Based on Foreign Research the kinematic viscosity.
The Reynolds nu¡nber is a viscosity parameter de-
Ahmad [Pakistan 1962 (I1) ] pendent on the water ternpèrature, but few designers
would be able to predict scour closely enough to ac-
d"/b = yolb (4.77F2t3 - Ð (1) count for water temperature, so Anderson reasoned
that he could use r.¡ = 1.2 x I0-5 and approximate
Bruesers lNetherlan¿ls 1964 (12) ] the exponent 0.619 by 0.66 to get
d"/b = 1 .a a) d"/b = AF2/3 (yolb)tlt (9)

Depending on when the exponent ís rounded, the

chitale Irndia 1962 (13) ]
coefficient A become.s 3.06 or 4.43. Ànderson got
d,/b =y¡lb (-5A9F2 + 6.65F -0.51) (3) 4.43 because he rounded the exPonenÈs before he
multiplied through by the constants. It is reason-
Inglis IIndia 1949 (14) ] able to use an internediate value of A = 3.4 and
make the Shen I equation identícal to the Shen II
d"/b =4.05 (yolb)3lo [F% -(yolb)] (4) equation that foIlows, because both eguations are
based on the sane data. Shen hinself consídered the
where equations to be equivalent.

ds = depth of scour rneasured from the nean Shen II t1969 (17) I

bed elevation, d*/b = 34 fvl(gb)\2tt (10)
Y¡r Y = aPProach flow dePtht
b = projected píer widthr which becomes directly
F = v/gY = Frou¿le number, and
v = velocity of aPProach flow. d"lb=3.4F213(yolu)t/' (11)

this will hereafter be referred to as the shen equa-

t ion.

csu tre75 (lg) l

Laursen sediment continuity equation t1958 (15) I
d"lb=2.2(yo/b)o 3sFo'43 02)
d/b=r/{s.st(r/1r.s)(d./yo)+111?0-1} (sa)
Even without considering the cornplexities of de-
Laursen lclear-water equation, not used in compari- bris and cohesive rnaterials, a designer is faced
sons that follow (tQ; unpublished data, 1977) I with nine equations to make a single conputation.
One might question why there are so many ilifferent
dlb = (rolrJh l5.s [(1/1 1 .s) (d"/ys ) + 11 t6 Ool"")% (sb) equations for the same predictions if èach re-
searcher was accurate in his work. Furthermore, one
NeiI Ifrom Laursenrs 1956 design curve (9)] might question whether the differenees in the egua-
tíons are as signíiicant as other environmental fac-
d/b = 1.s (y/b)o'3 (6) tors that could not be included in the model studies
used to develop the equatíons.
Jain I1979 (10) ] The equåtions are compared graphícally in Figures
I and 2. Because nost of the equations are in terms
d,lb=2.01r Jr";o'2s 1y7u;0's for (F -F")> 0.20 (7a) of y/b as vrell as the Froude nutnberr both figures
 204 Transportation Research Record 950

Local Pier Scour

ds/b Vs. F lor y/b - 2.0 b Fc - -33 lln Jain's Eq.)

12.5

€
!
i ro.o
.9
ô-
t
o
3 7.s

f
o
I¡
o
s.o

0.0
0.000 0.100 0.200 0.300 0.¿rc0 0.500 0.600 0.700 0.8ü)
F =v/{w
FIGURE I Graphical comparison of scour formulas (!) for variable Froude
numbers.

Local P¡er Scour

ds/b Vs y/b for F = 0.3

5.0

3
F ¿.0
@
ä
s
ä
o
t.o

Þ
o
go 2.0
¡6
!

1.0

0.0 L
0.(m 1.m0 2.0æ 3.(xrc 4.000 5.000 6.m0 7.m0 8.(m
y/b (Flow Oeplh/ P¡er Width)

FIGURE 2 Graphical comparison of scour formulas (!) for variable depth

ratios (y/b).

are needed to get a graphical conparison of the the design conditions should be useil. Most
eguations. Figure I is for an average depth rati.o designers do not hâve time to review the literature
(y/b) of. 2.0r Figure 2 is for an average Froude num- to deter¡nine the ¿lerivation conditions for aI1 of
ber of 0.3. The main difference in the eguations is these equations, so a summary of datå is shown in
not so ¡nuch in the data as in the way that the Figure 3 for equations fro¡n for- eign literature and
curves were fit to the data. All the equations are in Figure 4 for equaÈions fron U.S. Iiterature.
at least partly enpirical and most are reasonably Looking back at Figure 2 where the equations are
âccurate if applied within the range of the em-
pirical dâta. compared for variable values of. y/b, the Ahnad and
Chitale equations would not look so extrene if they
were not extended beyond the range of experinental
data (y/b approxinately 3.5).
SELECTING THE EQUATION FOR DESIGN ALI of the pier scour equations deríved in the
literature have been for noncohesive materials with
AccordÍng to the NCHRP synthesÍs report (5), there dSO ranging fron 0.17 to 1.5 mn. The Bruesers
are two approaches to selecting tþe most appropriate equation is based on l-Ímited dåta, but because of
equation or equations for design. First, the its sinplicity it serves as a good rule of thunb,
equations should be compared with field data to which is to anticipate pier scour around I.5 tines
deter- ¡nine which ones best duplicate fÍeld the projected pier width. The Neil equation, which
measurements. Second, lacking these data, the is based on the fuII range of Laursenrs data, re-
conditions under which the equations were derived duces to scour of 1.5 times the pier width when the
should be evaluated and the one that besÈ natches flow depth is equal to the pier width. Neil,s egua-
 205
Jones

Local P¡er Scour. Foreign lnvestigators

Exper¡menlal Lim¡ts

12.5
ds/b Vs F lor ylb = 2

lnveilrgator
3 F mm
3 ro.o
lYpe
Model 2.1 3.6 .19..49
d6n
.24
F"
.05..06
o Field
iI Bru6e6 Model 4.5.5.0 .09. .26 .20 .r4. .19
E Ch¡tale Model 0.9 3.4 .10..48 .16.1.51 .16..46
&
ô
7.5 lnqlis - Poona ) ) 17 39 ?

Legend:
o
Io s.o
Z¿¿or-.s lnd¡cates Range of Data on Curue

2.5

0.0
0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800

F =V/y'sy
FIGURE 3 Summary of data r¡sed to derive pier scour equations: foreign
literature.

Local P¡er Scour, American lnvest¡gators

Experimental Limits
ds/b Vs F for ylb = 2I Fc :.33 { in Jain's Eq.)

Range ol Data Ured in AnalYirs

I lnv6t¡gotor
d5¡ mm
! lvpe F

; Lau6en Mdel-Freld r.45 19 . .99 44 .2.25

o Nê¡l Fit Eq. ro Lâuænt Curue
iT
Model t&2 .5. r.5 25 .2.5 .29 . .63

Mo&l.F¡eld 8.1.2 .10 . .95 16. .68 19 . .26

ê
o
Shen

ô gme Dåla õ Sh6n

csu rl
f
o
o
Ø
!
Legend:øor *lndicstes ßangs ol Data
! on Curye.

0.(xþ 0.100 0.ã¡0 0.3æ 0.t100 0.5tx, u.but, u'lw I

F =v/ Viy

FIGURE 4 Summary of data used to deúve pier scour equations: U.S.

literatu¡e.

tion is just a règression fit to the ¿lesign curve Jaints equation is somewhat of a compronise be-
presented by Laursen in lowa Bulletin 4 (f9). Laur- tv¿een thosè of Laursen and Shen. It has a Frouale
sen later published his semíthêoretical equations nunbèr, but the tern has a relatively low exponent.
baseal on continuity of sedíment transport, although Jain introduced the threshold Froude number (Fc)
they also have an enpirical factor (Laursenrs r-val- as a \ray of accounting for relative sediment size.
ue) to tüake then fit expêritnental data. Neverthe- One criticisn of Jainrs equation is that it is dif-
less, Läursenrs equation probably has the best basis ficult to conpute Fc.
for extrapolation beyond its experimental base. The
nain criticísm of Laursenrs equation is that it does FIELD DATA
not include a Froude nunber (or velocity tern). A1-
though continuity may be satisfíe¿l without a veloc- for selecting one scour
The most convincing argurnent
ity tern' it see¡ns intuitively that velocity ¡¡ould equation rather than another is comparison of pre-
affect the strength of secondary currents around a dictions with field measurernents (9,29,2]-1. Unfor-
pier and therefore would be part of a prediction Èunately, field neasurenents are scarce, especially
equation.
Shenrs equation is an envelope curve that fits under flood conditions, and those rneasure¡nents that
the uppermost scour points for ã11 the available are available nust be carefully scrutinize¿l to iso-
¿lata [Figure 5 (12)]. This equatíon is appealing late one component of scour fro¡n another.
fron a design standpoint because it is intentionally Figure 7 (8) is a good example of some ôf the
on the consêrvative side. The CSU equation is a precautions thât need to be observed with field nea-
best fit to much of the same data lFigure 6 (f!) ]. surements. Tvro floods at the sane site are super-
 206 Transportation Research Recoral 950

Shen ll Eq. t0
I Svmbol Sourð Sdidît.i..

o",o = r.oþ-o¿)0" =..(Ë,,)'.( f A Ch.boñ &

€rFld'nrr 19,
Chsb.ñ å
-- E4ldhFr
O.E2 \
dSO. mm

I Fr.d
o.r" J ,"o
{91

0 o csu o.24
è1
Âà
\A
!
'a
L*- t,a,J-,-i"¡." 0¡3
¡ F

lll A + I
2.0 Escomes 2.2 lor R€cr. p¡å;
r lp o lr I I tl I r

ds 7^..94, 3g' 104 1o-3 1ù2 10-1 10-0

b
(s )i,
Y¡
(
Sour* , dso,..
I$ lshen,etat' 0.24(VA) FIGURE 6 CSU fit to pier scour data (lB).
a
t
o
r¡ Chabort and -
la Enæld¡nær '
0.26
0.52
oato O Knezevic3 O.2gs(dgo )

a O Maza & Sanchez 2 o-17

O Knezevic2 o.28s(dgo ) scour formulas. second, the entire cross section
o O No¡rl should be measured, not just one point, to even
¡ O zo ft. ¡lum speculate hosr much local scour occurred. Third, the
Piers: lCylindiø1, 2Rectangular pier footing, which was approximately 20 ft square¡
3Round No*d probably served as a scour arrester for the dlving
0.1 currents that were generated by the actuäI pier at
0.t 1.0 the hlgher stâge, whereas the footing was just belovt
pior Froudê Number. F - V the water surface at the lovrer stage an¿l tended to
generate rather than break up the aliving currents at
Ttb that stage. Data from a site like this give an in-
FIGURE 5 Shen's envelope fit to available data (U). tuitive feel for the nagnitude of scour but they
would be a weak argunent for valiilating the scour
preilíction equations.
inposed. The discharge of one was approxinately The fiel¿l data gathered fro¡n louisiana files by
three tirnes the discharge of the other. ?he stage Chang (20) are su¡nmarized in Figure B. Those sites
increased by 8 ft; the average velocity and Froude had good pier configuration and well-defined scour
nu¡nber at least doubled but there was apparently a ho1es, but for two reasons all the data were taken
reduction in local scour. How does one explain this at low Froude numbera. First, the LouÍsiana streams
discrepancy with the scour formulas? are 1ow-gradient streans and, second, the data were
First, this vras not a good site to choose for collected âs part of a routine survey of approxi-
such a comparison because the be¿l was already well ¡nately 90 sites by a hydrologic survey tean. The
belo¡{ the footing (or pile cap) i thus, the êffective surveys were rnade on a schedule and seldom coincided
pier was the battered pile group. These conditions with floods, which nay have generated sone higher
are not like any test conditions used to derive the local velocities.

Scale; V: l" = 20' H: 1" = /lQ'

O = 31,880 fi3/sec O = 10.440 ft3/sec = 1.65 lt/sec

W*" = 356 ft avg ::,] W*" = 312 lt avg Nr = 0065

A = 8813 ft2 A = 6322 Ír2

0 ='14"

Depth ot Local Scou¡ = 3.9 ft Depth of Local Scour = 7.05 ft

Stage Elev. = 58.0 ft, m.s.l.\
{for O = 31.880 cfs) \
Stage Elev. = 50.0 ft, m.s.l
{forO = 10,400cfs)

,-o"",r;=- .-\
lfor O = 16,49¡
"¡"¡
20

\O = 31.800 cfs

0 0

FIGURE 7 Cross section of Brazos River in Texas at bridge crossing (B).

 207
Jones

2.5
Arsumod F. = .05 lAvg. value for dso = o'ls --' Yo = 15 ftl lor Jåin's eq'

2.O

1.5 :Y? Notss:

Awvu oata (1980)
dso = ol to
0.16 mm

!b r.o
O Chang
La. Data 119791
droRanged From
.008 to .06 mm

Numerals Beside
Datâ Points
Are Froude
Numbers

1.0 r.5 2.O 2.5

Y016

FIGURE B Louisiana field measurements compared with prediction equations (2!)'

EFFECfS OF PIER SHAPE 2.75ds

lmaginary
The pier scour equations Presented in the preceding lnfl uence
o^
Zone --
Çiscussion are for rectangular piers. Although Wall --\
Abutment
Hole /
:
ùhere have been investigations l!!l of the effects Scour
of pier shape, a clesigner is not likely to be able
to take advantage of most of the streanlined shapes æ
because flow directions change as flooil levels Flow
change. The conservative approach for a designer is
to use the projected width of the píer in the
dírectíon of flow and to use the rectangular pier
equations excePt for rouncl piers, which scour about
90 percent as rnuch as rectangular piers.

APPLICATION OF PREDICTION EOUATIONS

Abutment
l- ¿------------*l

FIGURE 9 Typical scour at an abutment (lt).

A consideration that is ¡nore i¡nportant than fâctors
like pier shape and even ¡¡hich equation to use is
the manner in which a designer selects design Paran-
eters. None of the scour equations is based strictly ABUT¡{ENT SCOUR
on field neasurements in rrhich the cross section is
irregular and flo\d conclitions are nonuniform. llost Abutnent scour occurs when overbank f1o!r reenters
scour equations âre based on uniform, one-dinen- the nain channel and sets up large vortices in the
sional flow condítions. To use an eguatíon effec- briilge opening. Laursen reasoned that the continu-
tivety, a clesigner nust somehow visualize the field ity equation for sedinent flow needed to be satis-
condítions in a manner that resembles the test con- fied ancl conceptualized an approach abutment at the
ditions. The tendency is to use average depths and sl¡mnetrical half of a wide píer. typically vtater ín
velocities for an irregular cross section. A nore the nain channel is transporting secliment at capac-
reasonable approach nay be to use the dePth and ve- ity, and water in the overbank area is relatively
locity in a band of flow just upstream of the pier. free of sediment (so-calleil clear water). when these
velocÍty is harcler to predict this way, but ít ean flows ¡nix at the abutnent, there is a deficiency of
be assumed to follovt a logarithnic distribution. sedinent and this ileficiency is satisfíed with
The greatest discrepancy beteteen laboratory con- rnaterial fro¡n the abutnent scour hole. Laursen
ditions and field condítions is in the bed ¡nâterial. realizeil that the mixing occurred prinariLy in the
irrost of the laboratory tests htere run with uniform zone of flow near the bank and that it was not rea-
cohesionless soils. General practice in design is sonable to dilute the overbank flor¡ with the entire
to use thesè eguations as a conservative estimate if channel flow because the compute¿l scour woul-d tle-
a soll is considered erotlible. If a soil is con- crease directly with the width of the bridge open-
sidered nonerodible, scour is assu¡ned to be zero. ing. He defined an influence zone as shown in Fig-
The problem is the lack of something in between' but ure 9 and derived'the abutnent scour equations baseil
thaÈ is a problen that must renain unsolved until on no mixing outsi¿le the influence zone. The width
someone alevises a Plan to deal ¡¡ith the effects of of the influence zone is 2.?5 tirnes the depth of the
different soil ProPerties. abutrnent scour hole' which rneans that Laursenrs
 204 Transportâtion Research Record 950

implicit equâtions.
abutrnent scour eguations must be REF'ERENCES
For the, typical case where the flow in the in-
fluence zone (0I ) transports sediment ancl flow t. F.M.M. Chang. Statistical Sunmary of the Cause
fro¡n the overbank area (eg) is clear \rater, Laur- and cost of Bridge Failures. Report FIÍWA-RD-75-
senrs sedirnent continuity equation yields the fol- 97. FHWÀ, U.S. Department of Transportation,
lowing: Sept.1973.
2. Nomenclature for Hydraulics. Manuals and Re-
ports on Engineering practice 43. ASCE, New
(Qo/Qr) 2.7s(d,/y) = z.zs(4/v) t{tl') {a"lr)
{
* tl, tu - 1l (13) York, 1962.
3. Sedimentation Englneering. Manuals and Reports
The reconmended value for r in this case is 4.I. on Engineering Practice 54. ASCE, Ne\r york,
For the special sÍtuation in which both the flow 1975.
in the influence zone apil the overbank flow are 4. J.C. Brícê and J.C. Blodgett. Countermeasures
clear water, which could occur at relief bridges or for Hydraulic Problems at Bridges, Vol. l. Re-
erhere the abutnents are set back far enough on the port FEWA-RD-78-I62. FHWA, U.S. Department of
f,-loodplain, Laursenta eguation yields the followinc¡ Transportation, sept. I-978.
5. Scour at Bridge Waterways. NCHRP Synthesis of
y Highway Practice 5. HRB, National Research
e I = z.t s t¿.lrl ( { t(l /r) (4/v) + t17 r 6
¡ 1r o ¡ r - r) (14)
Council, Washington, D.C., 1970.
"¡h |
where
6. Scour and Deposition in Rivers and Reservoirs.
conputer Progran HEC-6. Hydrologic Engineering
Center, U.S. Army Corps of Engineers, Davis,
g = length 9q-tle approach enbankment, CaIif., l4arch 1977.
t s/ tç = v2 /L20d363y1/3, 7. An Evaluation of Flood Level prediction Using
r = 4.1 for 1o9r vetocities, and Conputer-Based Models of Alluvial Rivers. Con-
r = ll.5 for hÍgh velocities. nittee on Hy¿lrodyna¡nic Cornputer Uo¿lels for
Flood Insurance Studies, Natíonal Research
d" is È.he deepest part of the scour hole, which is Council, Washington, D.C., 1983.
assumed at the edge of the abutment. 8. G.R. Hopkíns, R.W. vãnce, and B. Kasraie. Scour
Laursenr s r-value essentíally distributes the Around Bridge piers. Report FHWA-RD-79-103.
scour in a triangular hole. The larger the r-value, FHWA, U.S. Departrnent of Transportation, 1980.
the larger the ratio of the deepest scour depth 9. A.G. Anderson. A Report on Scour at Bridge
(ds) to the average scour depth in the influence Water\rays: A Reviehr. Report FHWA-RD-75-99.
zone. FHI{A, U.S. Departrnent of Transportation, Nov.
cSU relatíonships for abutnent scour are as fôl- 197 4.
lows: I0. S.C. Jain and E.E. Fischer. Scour A,round Cir-
cular Bridge PÍers at High Froude Nunbers. Re-
40 Fo'33 port FHWA-RD-79-104. FHÍVA, U.S. Department of
4/y = 1.1 (r/y)o if Lly <2s (1 sa)
Transportation, April 1979.
Il-. Irl. Àhmad. Discussion of rscour at Bridge
d./y = 4Fo ¡: tf Lly > 25 (1 sb) Crossingsr" by 8.14. Laursen. Transactions of
the ASCE, Vol. 127, 1962, pp. L98-206.
There are no fÍe1d data to compare with the abut- 72. H.N.C. Bruesers. Scour Around DrfIling plat-
ment scour equations, but because there are only C\ro forms. Bulletin HydrauLic Research L964 and
equations to consider, it is reasonable to compute 1965, VoI.19r p.276. International Associa-
with both of them. A ¿lesigner would have to use the tion for Hydraulic Research, Delft, Nether-
equation that suíted his tendency to be more or less lands, 1965.
conservative. 13. S.V. Chitale. Discussion of ñScour at Bridge
Crossingsr" by E.M. Laursen. Transactions of
RECOITI¡/TENDATIONS
the ÀSCE, VoL. L27, 1962, pp. 191-196.
l-4. C.C. fnglis. The Behavior and Control of
There is still a need to docu¡nent field ilata for Rivers and Canals. Research Publ-ication 13,
both pier scour and abutment scour. Fielcl data Central Power, frrigation and Navigat.ion Re-
should be collected during floo¿ls and should as a port. Poona Research Station, India, 1949.
mini¡num include a full cross section at several 15. E.!1. Laursen. Scour at Bridge Crossings. 8u11.
flood stages. Data should be coltected by individ- 8. Iowa High$ay Research Board, Iovra City, Àug.
uals who are knowledgeable about how laboratory 1958r Transactions of the AscE, voI. I27, 1962.
tests are conducted and who can docu¡nent sufficient 16. E.l.l. taursen. An Analysis of Relief Bridge
infor¡nation to make va1Íd conparisons with predic- Scour. Journal of the Hydraulics Division of
t ions. the AscE, vol. 89, No. HY3, May 1969.
Although there are a large number of pier scour L7. H.W. Shen, V.R. Schnei¿lêr, an¿t S.S. Karaki.
equations, they can be narrowed down to three or Local Scour Àround Bridge piers. Journal of the
four without much loss in data used in alerivations. Hydraulics Division of the ASCE, Vol. 89, Nov.
The recommended eguationg are those by Laursen, 1969.
Jain, Shen, and CSU, Equations 5, 7, 11, and l-2,
respectively. 18. Highways' in the River Environment: Hyilraulic
There are only two equatíons (13 and 15) to con- and Environnental Design Considerations. FHWA
sider for âbutment scour if the main channel flow in training and Design ltânual. Civil Engineering
the zone next to the abutment (the influence zone) Department, Colorado State Universityt FHWA,
is transporting sedirnent. There is only one equa- U.S. Departnìent of Transportationr May L975.
tion {I4) for abutment scour for relief bridges anil 19. E.Itl. Laursen. Scour Around Bri¿lge Piers antl
large abutrnent setback where flow in that influence Abutrnents. BuIl. 4. Iowa Highway Research
zone is not transporting sediment. Board, Ior'ra City, üay 1956.
 209

20. F.F.M. Chang. Scour at Bridge Piers: Field 22. H.w. Shen and V.R. Schneider. Effect of Bridge
Data. Report FITWA-RD-79-105. ¡'HWA, U.S. Depârt- Pier Shape on Local Scour. Presented at ASCE
nent of Transportation, June 1979. National ltleeting on Transportation Engineering,
2L. E.M. Laursen. Model-ProtoÈype Conparison of Boston, Iuass., July 13-17, 1970.
Brídge-Pier Scour. HRB Proc., Vol. 34, 1955, h.tblîcatíon of this pøper sponsored by Committee on Hydrology, Hydraulícs
pp.188-193. and llater Swlity.

Riprap Stability Analysis

M.A. STEVENS, DARYL B. SIMONS' and E.V. RICHARDSON

ABSTRÀCT nethods of designÍng riprap or bank protection in

Hydraulic Engineering Circular 15. In the 1970s,
In the absence of wave and seepage forces, the problem of riprap design for side-slope protec-
the stability of rock riprap particles on a tion of banks in contact with flowing water at river
side slope is a functíon of the ¡nagnitude crossing encroachnents was investigated by colorado
and direction of the stream velocity next to State University for the Wyoning Statè Highway De-
the particles, thè angle of the síde slope, partment ín cooperatíon with FHwÀ, U.S. Departnent
and the characteristics of the rock, includ- of Transportation (l-6). As a result of this inves-
ing the geonetry, angularity, and density. tigation a rnethod of riprap design was develÕpe¿l
À rûethod of designing riprap was developed based on a functional relation between the forces
based on a functional relation among the noving the particle and those resisting these
variables. Rock particles on side slopès forces. The method defines a safety factor for the
tend to roll rather than to sIíde. There- rock riprap, which is defined as the ratio of the
fore, it is appropriate to consider the sta- mornents of the forces resisting rotation of a rock
bility in terms of monents about a point of particle and of the riprap to the nonents of forces
rotation. The functional relation has its tending to dislodge the particle. The critical con-
basis in the balance of noments about a dition is the flow for whích incipient ¡notion oc-
point of rotâtion at incipient motion be- curs. This critical condition has a safety factor
tween the forces trying to move the partielê of I. If the monents of the forces tending to dis-
and the forces resistÍng movenent. A safety lodge a particle are larger than the resisting rno-
factor was developed that is the ratío of nents, the safety factor is less than I, rocks are
the resisting moments to the moments tendÍng re¡noved from the riprap layer, and failure of the
to move the particle of riprap. This safety protection nay occur. $then the safety factor is
factor takes into consideration the side greater than 1, the riprap is safe frorn failure.
slope of thè bank being protectedi the size, The equations are developed fron theoretical con-
density, and angle of repose of the rockt siderations and existing empirical infor¡nation. The
and the lift and drag forces of thè follow- hydrodynanic lift U,t) and drag forces p) of the
ing water. The method is described, ex- fluid on the rock, the submerged weiqht and angLe of
ampl-es of its use are given, and it is com- repose of the rock, and the Shields criteríâ as ¡nod-
pared with other ¡nethods developed by the ified by cessler (!q) for incipient partlcle motion
Bureau of Public Roads, U.S. Arny Corps of are considered. The magnitude of the lift force is
Engineers, California Division of Highways, proportional to the rnagnitude of the drag force but
Bureau of Recla¡nationr the ASCE Task Com- is perpendicular Èo the drag force. This is impor-
nittee on Sedinentation¡ and others. taìt in analyzing particle stability (II). The
theoretical development, a design exatnple, and cal-
culation of the safety factor of several recomnended
design nrethods are presented.
Highway crossings or encroachnents of rivers usually
require so¡ne forn of protection for the encroaching
enbankrnents, bridge abutments, and adjacent river- THEORETICAL DEVELOPMENT
banks. Long approach embankments across the flood-
plain may need protection also. Usually this pro- In the absence of r+aves an¿l seepager the stabllity
tection is provided by rock riprap because of its of rock riprap particles on a side slope is a func-
low cost, flexibility, and ease of repair. The í¡n- tion of the ¡nagnitude antl direction of the strearn
portant factors in ilesigníng rock riprap are dura- velocity in the vícinity of the particles' Èhe angle
bility, sizer shape, angularityr angle of reposet of the side slope, and the châracteristics of the
and ilensity of the rocki side slope of the bank line rockr including the geometry, angularity, and ilen-
being protectedl; wave actíoni seepage forcesi and sity. In the following ¿leveloptnent of the safety
flow velocity (both rnagnitude and direction) close factor several flovr conditions are considered:
to the rock. The Bureau of Public Roads provided oblique flow on a side slope, horizontal flow on a