109S76
ACI STRUCTURAL JOURNAL TECHNICAL PAPER
ACI Structural Journal, V. 109, No. 6, NovemberDecember 2012.
MS No. S2011017.R1 received February 16, 2011, and reviewed under Institute
publication policies. Copyright 2012, American Concrete Institute. All rights
reserved, including the making of copies unless permission is obtained from the
copyright proprietors. Pertinent discussion including authors closure, if any, will be
published in the SeptemberOctober 2013 ACI Structural Journal if the discussion is
received by May 1, 2013.
ACI Structural Journal/NovemberDecember 2012 867
Defection Control of Concrete Slabs Longitudinally
Reinforced with ASTM A1035/A1035M07 Steel
by Admasu S. Desalegne and Adam S. Lubell
The ACI 31808 design code for reinforced concrete construction
provides both an implicit check of slab defection control based on
minimum member thickness and a direct computation method for
defection. Similar provisions are given in the ACI ITG6 design
guide (ACI ITG6R10) for members reinforced with highperfor
mance ASTM A1035/A1035M07 steel. This paper reports an
analytical study that compared the maximum spandepth ratios
from the implicit defection provisions with corresponding ratios
determined from direct defection calculations. Emphasis was
placed on defection control at the serviceability limit state (SLS)
for oneway slabs longitudinally reinforced with ASTM A1035/
A1035M07 steel where the nominal steel stress at the ultimate
limit state (ULS) ranged from 60 to 120 ksi (414 to 828 MPa). The
results indicate that the maximum spandepth ratio should decrease
as the span length increases, as the design load increases, as the
concrete strength decreases, or as the maximum permissible defec
tion decreases. The maximum spandepth ratio can be increased
as the longitudinal reinforcement ratio is increased beyond that
required to satisfy the fexural demand. These relationships with
the maximum spandepth ratio were all nonlinear in nature and
were of similar shape for all nominal reinforcement stress magni
tudes considered. Furthermore, these relationships were similar
when ULS fexural design was completed using either the simplifed
or general fexural design models provided in the ACI ITG6R10
guidelines. The study recommends that direct defection calcula
tions should be used for the design of all slabs and proposes graph
ical design aids for use in initial thickness selection.
Keywords: cracking; defection; highperformance reinforcement; oneway
slabs; reinforced concrete; stiffness.
INTRODUCTION
Reinforced concrete fexural members must have accept
able defections at the serviceability limit state (SLS) while
providing adequate strength at the ultimate limit state (ULS).
The maximum SLS deformations of structural members,
including the effects of incremental defection from
sustained loads, should be appropriate for their intended use
and minimize signifcant damage to nonstructural elements.
The longitudinal tensile reinforcement ratio r for a oneway
spanning slab is usually based on the fexural strength require
ments at ULS. Use of higher or lowerstrength reinforce
ment will change the required r and, hence, the reinforce
ment stresses and corresponding member curvatures at the
SLS condition. Thus, it is generally believed that minimum
slab thickness must change as a function of the nominal
reinforcement design stress f
D
to maintain adequate defec
tion control. The reinforcement stress at SLS is commonly
approximated as 0.67f
y
for traditional steel reinforcement
grades (refer to ACI 31808, Section 10.6.4).
1
However, the
ratio between the stress at SLS and f
D
used for ULS design of
ASTM A1035/A1035M07
2
steel can differ from the 0.67f
y
approximation, especially if the ULS design considers the
nonlinear region of the ASTM A1035/A1035M07 steel
stressstrain response.
3
Defections of reinforced concrete members depend on
many factors, including the degree of cracking, the time
dependent characteristics of the concrete, the mechanical
properties of the reinforcement, and the support and loading
conditions.
4
ACI 31808
1
provides two methods to satisfy
defection control requirements for reinforced concrete
members. The ACI ITG6 design guide (ITG6R10),
3
which
provides modifcations to ACI 31808 provisions for use
with ASTM A1035/A1035M07 Grade 100 (690 MPa)
steel, adopts the same twoapproach method of defection
control. In the frst approach, an implicit evaluation of
member defection is used, whereby a member with suff
ciently large overall depth h is deemed to comply (DTC)
with the defection requirements. As given in ACI 31808,
Table 9.5(a) (reproduced herein as Table 1), the minimum
member thickness h for span L is based on member type
(for example, slab or beam) and support configuration
(for example, simplespan or continuous). Footnote b)
of Table 1 gives an adjustment coeffcient to increase h as
the reinforcement yield strength f
y
increases above 60 ksi
(414 MPa). In the second approach, the member defection
is directly calculated using an effective moment of inertia
I
e
to account for the variation in stiffness along the member
length due to cracking. The calculated defections are
then compared to established defection limits. Due to the
simplicity of the DTC approach for defection control, this
method is usually preferred over direct defection calcula
tions for member size selection in design practice. Thus, it is
important that the DTC approach yields members that also
satisfy the defection control criterion under the direct calcu
lation method while still promoting structural economy.
Several previous studies have proposed different maximum
L/h relationships for defection control to replace those in
Table 1. Grossman
5
used computer simulations to develop
a simple expression for the minimum thickness of oneway
members based on the maximum permitted defection, the
longitudinal reinforcement ratio, and the loading. Gardner
and Zhang
6
used a layered, nonlinear fnite element model
and approximated the required increase in the maximum
L/h ratio as inversely proportional to the cube root of the
service momenttoultimate moment ratio M
a
/M
u
. They
also identifed that the limiting L/h ratio increases as the
concrete strength f
c
increases and as the fexural tension
868 ACI Structural Journal/NovemberDecember 2012
ACI member Admasu S. Desalegne is a PhD Student in structural engineering at
the University of Alberta, Edmonton, AB, Canada. He received his BSc and MSc in
structural engineering from Addis Ababa University, Addis Ababa, Ethiopia. His
research interests include analysis and design of concrete structures reinforced with
highperformance materials.
ACI member Adam S. Lubell is an Assistant Professor of Civil Engineering at the
University of Alberta. He received his PhD from the University of Toronto, Toronto,
ON, Canada. He is past Secretary of ACI Task Group ITG6, High Strength Steel
Reinforcement, and is a member of ACI Committee 440, FiberReinforced Polymer
Reinforcement; 544, FiberReinforced Concrete; and Joint ACIASCE Committee 445,
Shear and Torsion. His research interests include the design and rehabilitation of
reinforced and prestressed concrete structures and the development of structural
detailing guidelines to allow the use of highperformance materials.
or compression reinforcement ratios (r and r) increase.
Scanlon and Choi
7
showed that the minimum slab thickness
can be reduced as L decreases and as the live load decreases.
Gardner
8
compared maximum L/h relationships from the
literature and from several codes of practice and recom
mended maximum L/h ratios that decrease as r decreases, as
f
c
decreases, and as the ratio of maximum sustained moment
to ultimate moment capacity increases. Choi et al.
9
used a
Monte Carlo simulation to calibrate a proposed simpli
fed expression for maximum L/h. Among the parameters
included were span length, load intensity, support condi
tions, concrete strength, and steel strength.
Tang and Lubell
10
used a holistic approach to consider
member thickness and the corresponding reinforcement,
which would simultaneously satisfy the fexure, shear, and
defection requirements of oneway spanning members,
and which could form the basis of graphical design aids for
selecting member thickness. The maximum L/h ratios from
these plots were then compared with the corresponding L/h
ratios derived from the implicit defection control provisions
(that is, the DTC approach from Table 1). The study used
requirements from CSA A23.304,
11
which are similar to
those in ACI 31808. The results showed that the maximum
L/h ratios should decrease as the span length L increases,
as the design load w increases, or as the cracking moment
M
cr
decreases. The study also demonstrated that the ULS
design strength of longitudinal reinforcement f
D
did not have
a signifcant effect on the minimum h required to satisfy the
defection criterion in contrast to the assumed relationship
from Footnote b) of Table 1. Bischoff and Scanlon
12
later
developed simple expressions for the maximum L/h ratio for
oneway slabs and beams that had a similar shape to those
from Tang and Lubell
10
by considering the parameters of
reinforcement ratio, cracking moment, specifed defection
limits, compressive strength of concrete, and yield strength
of steel. The use of these expressions to check for adequate
defection control, however, requires prior knowledge of r,
which is typically unknown until the member thickness h
is selected. Thus, the Bischoff and Scanlon
12
expressions
cannot be easily used for optimized member size selection.
Furthermore, the structure of the relationships prevents their
consistent use when the ULS design considers the nonlinear
stressstrain response of highperformance reinforcement,
such as ASTM A1035/A1035M07 steel.
The study by Tang and Lubell
10
used a linear elastic
perfectly plastic stressstrain model for the steel reinforce
ment and only considered steel design strengths f
y
up to 80 ksi
(552 MPa). This paper forms an extension to the Tang and
Lubell
10
approach by specifcally considering oneway
concrete slabs longitudinally reinforced with ASTM A1035/
A1035M07 Grade 100 (690 MPa) steel. Higher nominal
design strengths f
D
of up to 120 ksi (828 MPa) were used
in the analytical calculations, including consideration
of the nonlinear stressstrain response of ASTM A1035/
A1035M07 steel in some cases.
RESEARCH SIGNIFICANCE
Oneway slabs reinforced with highstrength steel typi
cally have low longitudinal reinforcement ratios; however,
there has not been previous work to systematically establish
whether the maximum L/h ratios specifed by the DTC defec
tion control method of ACI 31808 are also appropriate for
lightly reinforced members with higherstrength reinforce
ment. In the case of slabs reinforced with ASTM A1035/
A1035M07 steel, it is also important to evaluate whether
these DTC provisions can be applied to members where the
fexural design uses the nonlinear portion of the steel stress
strain response. A comparison was completed between
maximum L/h values determined from a holistic design
approach considering fexure, shear, and direct defection
calculations with L/h ratios derived from the DTC provi
sions. The aim was to establish an appropriate method for
selection of minimum member thickness with due regard for
infuences that arise both from the overall member confgu
ration and from the different ULS analysis techniques given
in ACI ITG6R10.
3
REINFORCEMENT PROPERTIES
ASTM A1035/A1035M07 steel has a different metal
lurgy and microstructure than conventional reinforcing steel
commonly used in most new construction.
3
These changes
result in ASTM A1035/A1035M07 steel with an effec
tive yield strength signifcantly higher than conventional
ASTM A615/A615M06
13
(60 or 75 ksi [414 or 518 MPa])
or ASTM A706/A706M08
14
(60 ksi [414 MPa]) reinforcing
steel while also being less susceptible to corrosion.
15,16
To
account for the mechanical properties of ASTM A1035/
A1035M07 Grade 100 (690 MPa) steel, ACI Innovation
Table 1Minimum thickness of nonprestressed beams or oneway slabs unless defections are
calculated (adapted from ACI 31808, Table 9.5(a)
1
)
Minimum thickness h
Simply supported One end continuous Both ends continuous Cantilever
Member Members not supporting or attached to partitions or other construction likely to be damaged by large defections
Solid oneway slabs L/20 L/24 L/28 L/10
Beams or ribbed oneway slabs L/16 L/18.5 L/21 L/8
Notes: Values given shall be used directly for members with normalweight concrete and Grade 60 (414 MPa) reinforcement. For other conditions, values shall be modifed as follows:
a) For lightweight concrete having equilibrium density w
c
in the range of 90 to 115 lb/ft
3
(1440 to 1840 kg/m
3
), values shall be multiplied by (1.65 0.005w
c
) but not less than 1.09.
b) For f
y
other than 60 ksi (414 MPa), values shall be multiplied by (0.4 + f
y
/100,000).
ACI Structural Journal/NovemberDecember 2012 869
Task Group 6
3
developed representative analytical stress
strain curves that are suitable for use in design and were
adopted in this study
29, 000 0.0024
0.43
170 for 0.0024 0.02 (ksi)
0.0019
150 0.02 0.06
s s
s s
s
s
f
= <
<
(1a)
200, 000 0.0024
2.96
1170 for 0.0024 0.02 (MPa)
0.0019
1040 0.02 0.06
s s
s s
s
s
f
= <
<
(1b)
DESIGN PROCEDURES FOR ONEWAY SLABS
This analytical study evaluated the defection control of
oneway simply supported slabs longitudinally reinforced
with ASTM A1035/A1035M07 steel in buildingtype struc
tures. All slabs studied were subjected to uniformly distrib
uted foor loading. Loading consisted of the member self
weight w
DL
, superimposed dead loads w
SDL
to account for
mechanical systems and architectural fnishes, and typical
live loads w
LL
defned in ASCE/SEI 705
17
for different
building occupancy conditions. Load and resistance factors
from ACI 31808, with specifed modifcations from
ACI ITG6R10, were used. While all reinforcing steel consid
ered was ASTM A1035/A1035M07 Grade 100 (690 MPa),
designs corresponding to different nominal steel stress
values f
D
of 60, 100, and 120 ksi (414, 690, and 828 MPa)
at the ULS condition were developed, consistent with the
ACI ITG6R10 provisions. This allowed the nonlinear
stressstrain response of the steel at ULS to be directly eval
uated for its corresponding infuence on defection control
at the SLS. The range of parameters and member confgura
tions studied are provided in Table 2.
Overview of typical design sequence
Slabs must be designed to have adequate strength in
fexure and shear at the ULS condition. The defections of
these members must also be controlled to within accept
able limits for their intended use at SLS. The typical design
sequence used in practice to achieve these objectives is given
by the fowchart in Fig. 1 and briefy described. Unique
design aspects within each step, as they pertain to differ
ences between ACI ITG6R10 and ACI 31808 provisions
to account for the use of ASTM A1035/A1035M07 steel,
are provided in the following sections.
Initially, a member thickness h must be selected that is
expected to satisfy the ULS and SLS design criteria (Step 1).
Due to its simplicity, the initial selection of h is typically
made using the DTC defection control provisions (that is,
Table 1) with the modifcation specifed in the table foot
note for f
y
60 ksi (414 MPa). However, if h is directly
determined in later steps from the governing case of explicit
defection calculations or strength requirements, this modi
fcation is not directly applicable. Next, for the member size
selected, the longitudinal reinforcement quantity A
s
is deter
mined to satisfy the fexural strength requirement at ULS
(Step 2). The shear capacity of the slab is evaluated at ULS
and compared against the loading demand (Step 3). If the
member satisfes the requirements of the DTC defection
control method, including adjustment for f
y
, no further check
of defection is required (Step 4a). Alternatively, direct
Table 2Design parameters considered
Parameter U.S. customary units Metric units
Concrete strength f
c
5 and 10 ksi 34.5 and 69 MPa
Nominal steel design
strength f
D
60 to 120 ksi 414 to 828 MPa
Live load intensity w
LL
50 and 100 lb/ft
2
2.4 and 4.8 kPa
Superimposed dead load
intensity w
SDL
20 lb/ft
2
1 kPa
Span length L 10 to 32 ft 3 to 10 m
Slab thickness h 3 to 22 in. 75 to 550 mm
Fig. 1Design procedure for optimized slab thickness
according to defection control.
870 ACI Structural Journal/NovemberDecember 2012
defection calculations at SLS are completed and compared
against the appropriate limits (Step 4b). The defection of a
member satisfying Step 4a could also be evaluated at Step 4b
so as to allow optimization of h. To satisfy the strength or
defection criteria at Steps 2, 3, or 4, the slab thickness h
at Step 1 can be adjusted. To emphasize the relationships
between the various design parameters, this study reports
results for optimized values of h that just satisfy the most
stringent criterion from Steps 2, 3, or 4b, whereas industry
practice will use practical incremental thicknesses for slabs.
DESIGN AT ULS
Flexural design of slabs with ASTM A1035/
A1035M07 steel
As noted previously, the stressstrain relationship for
ASTM A1035/A1035M07 steel given by Eq. (1) does not
include a distinct yield point or yield plateau. Flexural design
of members longitudinally reinforced with ASTM A1035/
A1035M07 steel must account for the curvilinear stress
strain response. ACI ITG6R10 permits the use of two
different fexural analysis models that are based on different
assumptions for the reinforcement stressstrain response and
corresponding momentcurvature response of the concrete
section. Both fexural analysis models are used in this study
to highlight their respective infuences on the maximum L/h
ratios that provide adequate defection control.
In the frst fexural model, herein termed the Mast model,
18
a
simplifed elasticplastic representation for the stressstrain
response of ASTM A1035/A1035M07 steel is used with
an effective yield stress of f
y
= 100 ksi (690 MPa). Flexural
capacity at ULS is calculated with the ACI 31808 rectan
gular stress block provisions and with an assumed concrete
strain at the extreme compression fber of e
cu
= 0.003.
The ACI ITG6R10 partial safety factor for fexure f
f,1
,
according to this approach, is given by 0.65 (f
f,1
= 0.45
+ 50e
s
) 0.9, where e
s
is the reinforcement strain at ULS.
Owing to its simplicity in the required calculations, the Mast
method
18
is expected to be the more widely used fexural
design approach in industry practice. This study used the
Mast model
18
to consider members with nominal steel design
strength values f
D
of 60 and 100 ksi (414 and 690 MPa).
In the second fexural model provided in ACI ITG6R10,
herein termed the Appendix B model, the full nonlinear
stressstrain relationship for ASTM A1035/A1035M07 steel
according to Eq. (1) is used. The concrete strain at the
extreme compression fber is taken as e
cu
= 0.003 at ULS
and the fexural response of the section can be solved using
a strain compatibility approach based on engineering beam
theory. Due to the nonlinear stressstrain relationship for the
reinforcement, an iterative calculation approach is typically
required. A provision of adequate fexural capacity can be
easily checked when the reinforcement quantity is known.
For the selection of an optimized quantity of reinforce
ment for a given fexural demand M
u
, it is typically easiest
to establish a target maximum nominal stress level f
D
in
the reinforcement. ACI ITG6R10 gives a partial safety
factor for fexure f
f,2
for members designed according to the
Appendix B approach from 0.65 (f
f,2
= 0.23 + 100e
s
)
0.9. This study used the Appendix B model to consider slabs
with f
D
values of 100 and 120 ksi (690 and 828 MPa).
To compare these two fexural design approaches,
Fig. 2 plots the relationship between the nominal moment M
n
and curvature F for the 12 x 12 in. (305 x 305 mm) cross
section defned in the fgure. Note that the full M
n
F response
in Fig. 2 was prepared using a variable e
cu
value
19
and will
have minor variance from M
n
F values calculated with
constant e
cu
at ULS for the two fexural models described
previously. The Mast and Appendix B models both
predict a similar M
n
F response prior to cracking and for
values of M
n
up to approximately 40 kipft/ft (178 kNm/
m). This point corresponds to the proportional limit of
ASTM A1035/A1035M07 steel from Eq. (1). Beyond this
point, the Mast model
18
gives a slightly stiffer response due
to the use of an effective f
y
larger than the proportional limit
stress; however, the maximum calculated fexural capacity
quickly plateaus. The Appendix B model gives increasing
fexural capacity at a decreasing rate as the curvature F (that
is, slab defection) increases. Points have been marked on the
plot to correspond to nominal reinforcement stress magni
tudes f
D
of 60, 100, and 120 ksi (414, 690, and 828 MPa).
It is important to note that for each of these conditions, the
corresponding point representing the SLS condition will have
steel stresses below the proportional limit when typical load
and resistance factors are applied. Thus, for the case shown,
it is possible to consider the SLS design requirements based
on the elastic methods used for traditional reinforcing steels
that exhibit welldefned yield plateaus.
Structural slabs designed according to ACI 31808 must have
a minimum quantity of longitudinal reinforcement in the span
direction that satisfes the shrinkage and temperature reinforce
ment provisions of Section 7.12.2. ACI ITG6R10 requires
a designer to follow these provisions and notes that for
ASTM A1035/A1035M07 Grade 100 (690 MPa) steel, the
corresponding minimum gross reinforcement ratio is 0.14%.
This requirement is easily satisfed for thin slabs with reason
able bar spacing, as the maximum spacing is limited to the
smaller of 18 in. (457 mm) or 5h. According to ACI ITG6R
10, Section 4.9.4, to provide crack control at a reasonable bar
spacing for members with increased cover, it is necessary to
limit the steel stress at the service load to less than 67 ksi
(460 MPa). Section 4.2 of ACI ITG6R10 also suggests
limiting the maximum strain in the reinforcement at ULS to
0.015 (that is, 144 ksi [994 MPa]) to avoid excessive cracking
of members.
Fig. 2Momentcurvature relationship for fexurally
cracked slab with ASTM A1035/A1035M07 steel.
ACI Structural Journal/NovemberDecember 2012 871
Shear design of slabs with ASTM A1035/
A1035M07 steel
The shear capacity of reinforced concrete slabs that do not
contain stirrups are infuenced by many design parameters,
including the concrete strength f
c
, the effective depth d,
and the longitudinal reinforcement confguration.
2023
Note
that these same parameters will also infuence the fexural
design and the overall member defection. With regard to
the use of ASTM A1035/A1035M07 steel as the longitu
dinal reinforcement in slabs, the higher nominal strength
f
D
compared to traditional Grade 60 (414 MPa) reinforcing
steel allows slabs with a lower reinforcement ratio r and
higher steel stress f
s
to still satisfy fexural strength require
ments. These slabs, however, will exhibit larger diagonal
crack widths at the ULS condition, which will impact the
shear strength.
24
With additional modifcations to account for the possible
nonlinear stressstrain response of ASTM A1035/A1035M07
steel at ULS, Desalegne and Lubell
25
showed that the shear
model proposed by Hoult et al.
26
can be used to predict the
shear capacity of slabs reinforced with this steel. The Hoult et
al.
26
model enhances the modifed compression feld theory
24

based CSA A23.304
11
shear model to better account for the
infuence on oneway shear capacity from large longitudinal
reinforcement strains, with the shear capacity at the critical
section given as
0.7
0.3 616.6
(lb)
39.37 0.5 (1000 0.15)
n c w v
ze x
V f b d
s
_ _
,
+ + + , ,
(2a)
V
s
f b d
n
x ze
c w v
+ +
_
,
_
,
0 3
0 5 1000 0 15
1300
1000
0 7
.
. ( . )
.
(N)
(2b)
where parameter e
x
represents the effective axial strain at
midheight and is derived from the reinforcement stress at the
critical section
25
; the shear depth d
v
is taken as 0.9d; and the
effective crack spacing parameter s
ze
can be taken as 0.9d for
the concrete with 3/4 in. (19 mm) aggregate assumed in this
study. While a simplifed version of this shear capacity method
that is compatible with the simplifed Mast fexural model
assumptions is included in the ACI ITG6R10 guide,
3,25
the
general version (that is, Eq. (2)) is used in this study due to
the use of the Appendix B flexural method in some cases.
ACI ITG6R10 adopts the same partial safety factor for shear
f
sh
= 0.75, as given in ACI 31808.
DEFLECTION OF REINFORCED
CONCRETE SLABS
ACI 31808 DTC with defection limits
Table 9.5(a) in ACI 31808 (Table 1 in this paper)
provides minimum thickness values for members that
are DTC with defection requirements for members not
supporting or attached to partitions or other construction
likely to be damaged by large defections. By this defection
control specifcation, members sized using this technique
would be expected to limit the total incremental longterm
defection after installation of nonstructural items to D
inc
L/240.
1
Values for minimum thickness h are provided
as functions of the span length L, based on the member
type and the support condition. For members conforming to
Table 1 and the defection limit specifcation noted previ
ously, the fexural stiffness does not need to be directly
determined because ACI 31808 and ACI ITG6R10 do
not require direct checks of the predicted defection for
these members. In the case of lightly reinforced members,
however, ACI ITG6R10 recommends making direct
defection calculations.
Direct defection calculations for slabs
Deformations of slender, oneway spanning reinforced
concrete slabs without shear reinforcement are assumed to
be consistent with the wellknown hypothesis that plane
sections before bending remain plane after bending. The
defection is determined by considering the corresponding
curvatures along the member length. Thus, the instantaneous
defection of a member subjected to uniform transverse
loading can be computed with the wellknown relationship
4
5
384
i
c e
wL
K
E I
=
(3)
where D
i
is the instantaneous defection; w is the uniform
transverse loading considered; L is the span length; E
c
is the
secant modulus of elasticity of concrete taken as 57,000f
c
psi (4735f
c
MPa); I
e
is the effective moment of inertia of
the transformed cross section; and K is a coeffcient based
on the boundary conditions (1.0 for simple span; 0.416 for
fxedpin; 0.2 for fxedfxed). Timedependent infuences
on defection must also be considered. According to the
ACI 31808 provisions, the incremental longterm defection
D
inc
resulting from creep and shrinkage of fexural members
can be determined by multiplying the immediate defection
caused by the sustained load by the factor l
D
1 50
=
+
(4)
where r is the compression reinforcement ratio taken at
midspan for simple and continuous spans, and at the support
location for cantilevers; x is the timedependent factor for the
sustained loads taken equal to 1.0, 1.2, 1.4, and 2.0 for loads
sustained for 3, 6, 12, or more than 60 months, respectively.
ACI 31808 defection provisions limit the imme
diate defection D
i
from live loads to D
max,imm
= L/180 or
L/360 for roofs or foors, respectively, when not supporting
or attached to nonstructural items likely to be damaged by
large deflections. Defection limits of D
max,inc
= L/240 and
L/480 are used for the portion of defection that occurs after
attachment of nonstructural elements (sum of the incremental
longterm defection due to all sustained loads D
inc
and the
immediate defection D
i
due to any additional transient live
load) if they are not likely, or are likely, to be damaged by
large defections, respectively. According to Gardner,
8
there
is general agreement that this total longterm defection after
installation of nonstructural items (that is, D
max,inc
) is typically
the more critical case compared to the immediate transient
live load defection limit D
max,imm
. While both criteria were
checked in this study, the D
max,inc
criterion was confrmed
to be the D
max
governing case for all maximum L/h ratios
presented in this study.
872 ACI Structural Journal/NovemberDecember 2012
Calculating fexural stiffness
As part of the defection calculation for Eq. (3), an evalua
tion of an appropriate moment of inertia for the cross section
is required to address the variable cracked nature along the
member length. Bischoff
27
developed a formulation for
effective moment of inertia I
e
that gives estimates of member
defection that are in better agreement with test results than
those using the I
e
formulation in ACI 31808 developed by
Branson.
28
The Bischoff
27
I
e
model given by Eq. (5) was
adopted by ACI ITG6R10 and was used in this study
2
1 1
cr
e g
cr cr
g a
I
I I
I M
I M
=
(5)
where M
a
is the maximum characteristic moment under the
load being considered, taken herein as the maximum service
moment (full dead load + full live load) as a simplifed tech
nique to consider infuences from earlyage loading during
construction
29
; M
cr
is the cracking moment; I
g
is the moment
of inertia of the gross section about the centroidal axis,
neglecting the reinforcement; and I
cr
is the cracked moment
of inertia of a singly reinforced section, given by
3 2
1
( ) ( )
3
cr w s
I b kd n A d kd = + (6)
where the modular ratio n = E
s
/E
c
; b
w
is the member width; d
is the effective depth of the reinforcement from the compres
sion face;
2
2 ( ) k n n n = r+ r  r ;
and A
s
is the area of fexural tension reinforcement, with the
reinforcement ratio evaluated as r = (A
s
/(b
w
d)).
According to ACI 31808, the cracking moment for
normalweight concrete, M
cr
, is related to the modulus of
rupture f
r
= 7.5f
c
psi (0.623f
c
MPa) and the gross section
properties through the expression
cr r g
cr
t
f I
M
y
= (7)
where g
cr
is a coeffcient adopted in this study to account
for a reduced cracking moment due to restrained shrinkage
and is taken as 0.67, as per the recommendation of Bischoff
and Scanlon.
30
PARAMETRIC INFLUENCES ON DEFLECTION
For the DTC defection control technique (Table 1), the
corresponding L/h ratios are constant for each member type
and support condition. Only the infuence of reinforce
ment yield strength f
y
is given additional consideration
through Footnote b) of Table 1. However, the ULS design
methods for fexure and shear (Steps 2 and 3 in Fig. 1) and
for detailed defection computations at SLS (Step 4b in
Fig. 1) will be infuenced by various parameters that are
applicable to each particular design case. In general, these
can be classifed as: 1) parameters infuencing the service
moment magnitude and its fraction relative to the ulti
mate moment; and 2) parameters infuencing the propor
tion of the member that will be cracked in fexure. For this
study using ASTM A1035/A1035M07 steel reinforce
ment, the nominal design strength f
D
and corresponding
fexural design method represents a third classifcation. A
systematic evaluation of the infuence on defection from
the main parameters in these three primary classifcations
was completed. The limiting L/h ratios were developed for
each case by considering the total defection D as the incre
mental defection D
inc
from 28 days (assumed time of appli
cation of sustained live load and superimposed dead load)
until 60 months (that is, x = 2) combined with the immediate
defection D
i
of the transient live load fraction. The limiting
L/h ratios represent the case where D = D
max.
A sustained live
load fraction of g
LL
= 70% was used for all analysis reported
in this study, but the infuence of this parameter was found
to be relatively minor within the typical range of 40 to 70%
applicable for many structures.
10
Factors affecting SLS and ULS moments
The defection of a oneway slab is a function of the magni
tude of the cracking moment M
cr
, the service moment M
a
, and
the corresponding ultimate moment M
u
, as shown in Eq. (3)
and (5). The moments M
a
and M
u
are related to the span
length L, the support confguration, and the applied loading
w. Because M
u
is used for determining the amount of longi
tudinal reinforcement, a higher M
u
will result in an increased
r for a constant slab thickness h, which thereby increases
I
cr
. If the slab thickness is allowed to adjust, however, the
ratio of dead load to live load will change. Because both
ACI 31808 and ASCE/SEI 705 use basic load factors of
1.2 and 1.6 applied to dead load and live load, respectively,
the ratio M
a
/M
u
will change as h changes. In addition, as the
superimposed dead load w
SDL
for items such as architectural
fnishes increases as a fraction of the total load, the M
a
/M
u
ratio will also change. In both cases, this change in ratio will
affect I
e
and, hence, the defections at SLS.
The relationship between the maximum L/h ratio and
member span L according to the direct defection calcula
tion method was determined for different live load inten
sities w
LL
and concrete strengths using the f
D
values and
corresponding fexural design methods identifed previously
(refer to Fig. 3 and 4). Figure 5 illustrates the variation in
maximum L/h as the superimposed dead load w
SDL
changes
for a slab with a span of L = 20 ft (6.1 m). Superimposed
dead load w
SDL
refers to dead load other than the selfweight
of the member. It is observed that the maximum L/h ratio
for adequate defection control will decrease as L increases
for all values of f
D
. Furthermore, the maximum L/h ratio
decreases as the applied live load w
LL
increases or as w
SDL
increases, while the other parameters are kept constant. By
comparing Fig. 3 and 4, it is also observed that increasing
the concrete strength f
c
increases the maximum allowable
L/h ratio for given values of w
LL
and f
D
. It is observed that
defection control of lightly loaded slabs is more sensitive to
the span length and the superimposed dead load because the
slopes of the maximum L/htoL and L/htow
SDL
relation
ships decrease as the live load intensity increases. For the
cases considered, thinner slabs can be used for shorter spans
or for lighter loading conditions than the corresponding
minimum thickness determined from the DTC defection
provisions of ACI 31808.
ACI Structural Journal/NovemberDecember 2012 873
Fig. 3Infuence of span length on maximum spandepth ratio for normalstrength concrete. (Note: 1 ksi =
6.89 MPa; 1 lb/ft
2
= 0.048 kPa.)
Fig. 4Infuence of span length on maximum spandepth ratio for highstrength concrete. (Note: 1 ksi =
6.89 MPa; 1 lb/ft
2
= 0.048 kPa.)
Factors affecting degree of concrete cracking
According to Eq. (7), the cracking moment M
cr
is directly
related to the modulus of rupture f
r
and, hence, f
c
. Thus,
as the concrete strength increases, M
cr
will also increase,
thereby increasing I
e
(refer to Eq. (5) and (7)) and allowing
thinner sections for a given span length L. This is observed
by comparing Fig. 3(a) and 4(a) or Fig. 3(b) and 4(b). The
relationship between f
c
and maximum L/h can be observed
from Fig. 6 for typical residential foor loading and a span
of L = 20 ft (6.1 m). It is observed that slabs satisfying the
DTC approach will typically also satisfy the defection
requirements from direct defection calculations for prac
tical concrete strengths when designed using higherstrength
steel. However, for L = 20 ft (6.1 m) slabs designed using
Grade 60 (414 MPa) steel and concrete strengths lower than
approximately 5 ksi (35 MPa), the DTC approach underesti
mates the required member thickness in comparison to direct
defection calculations.
Providing excess reinforcement
It is common practice that the A
s
provided in a slab exceeds
that required by the ULS criteria due to practical consider
ations, including the use of convenient bar spacing. Further
more, criterion for minimum longitudinal reinforcement
quantities may exceed that required by the fexural demands.
From Eq. (6), it is observed that the provided A
s
will impact
the cracked moment of inertia I
cr
and the corresponding
defection. Figure 7 depicts the relationship between the
874 ACI Structural Journal/NovemberDecember 2012
maximum L/h ratio and the area of steel provided normalized
by the area of steel required for the fexural demand (A
sp
/A
sr
).
As A
sp
increases beyond the fexural capacity requirement
A
sr
, the maximum L/h ratio increases almost linearly for
the case of L = 20 ft (6.1 m) and w
LL
= 50 lb/ft
2
(2.4 kPa).
Similar relationships occur if w
LL
is increased, except that, as
expected, the maximum L/h ratio is lower for higher values
of w
LL
. In general, providing excess longitudinal reinforce
ment within practical limits will increase the member stiff
ness and decrease the SLS defection of oneway slabs,
allowing a minor reduction in the required thickness h.
Infuence of defection limit
Figure 8 shows the variation of the maximum L/h ratio
for different limits of maximum midspan defection D
max
for
a span of L = 20 ft (6.1 m). To facilitate comparisons to
typical design code defection requirements, the common
D
max
limits of L/240, L/360, and L/480 are also indicated.
The fgure shows that the maximum L/h ratio increases
as the permitted D
max
increases. The maximum L/h values
diverge for different values of f
D
at large values of D
max
,
resulting in the need for thicker slabs for higher f
D
(that is,
Grade 100 and 120 [690 and 828 MPa] steel). However,
for the typical design code limits of D
max
smaller than
L/240, the difference in required h for all f
D
values consid
ered was small. It is also noted from Fig. 8 that slabs with
Grade 60 (414 MPa) steel sized according to the DTC
method may not satisfy common defection control require
ments compared to those sized using direct defection calcu
lations, especially for the case of higher live load intensi
ties. According to Ramsay et al.,
31
deflection predictions
can have an error of 20% for common ratios of M
cr
/M
a
.
If a target maximum deflection 20 or 30% smaller than
a typical design code limit was desired to accommodate
this error range, Fig. 8 suggests that the required change
in the maximum L/h ratio for a slab would be minimal in
comparison to the discrepancy between DTC and direct
deflection calculations.
Infuence from fexural design method
As discussed previously, two fexural analysis models
from ACI ITG6R10 were used in this study: the Mast
method and the Appendix B method. Due to the different
reinforcement stressstrain models in these methods, the
required r to satisfy the ULS fexural strength requirements
can differ for the same member geometry and applied load.
Changes in r will have corresponding impacts on the defec
tion calculations and could alter the maximum L/h ratios for
adequate defection control.
As observed in Fig. 3 through 8, the shapes of the respec
tive L/h curves are nearly the same, regardless of the f
D
value
or fexural design method. Some divergence among the
Fig. 5Infuence of superimposed dead load intensity on
maximum spandepth ratio. (Note: 1 ksi = 6.89 MPa; 1 lb/ft
2
= 0.048 kPa.)
Fig. 6Infuence of concrete strength on maximum span
depth ratio. (Note: 1 ksi = 6.89 MPa; 1 lb/ft
2
= 0.048 kPa.)
Fig. 7Infuence of providing excess reinforcement
compared to ULS requirement on maximum spandepth
ratio. (Note: 1 ksi = 6.89 MPa; 1 lb/ft
2
= 0.048 kPa.)
ACI Structural Journal/NovemberDecember 2012 875
curves within each plot occurs when other design parameters
are varied, but in general, the choices of f
D
resulted in offsets
to the curves with higher f
D
values, resulting in smaller L/h
limits for a given member confguration. Furthermore, the
L/h limits for higher f
D
(100 and 120 ksi [690 and 828 MPa])
had negligible sensitivity to the fexural design method, as
the reinforcement confguration was typically controlled by
minimum reinforcement requirements for the cases studied.
This demonstrates that member design at ULS using the
nonlinear response of the ASTM A1035/A1035M07 steel
should not have defection control provisions that result in a
disproportionate impact on the maximum permitted L/h ratio.
A case study was used to further examine the infuence
of the selected fexural design method on the holistic design
of slabs with minimum thickness h. An L/240 incremental
defection limit was used and both direct defection calcula
tions and the DTC approach were considered. Results are
presented for a typical residentialtype foor with a live load
of w
LL
= 50 lb/ft
2
(2.4 kPa) and superimposed dead load
w
SDL
= 20 lb/ft
2
(1.0 kPa), but similar trends in the results
are found for other loading cases, such as offce occupancy
loads, where the live load intensity is larger. Slab widths
were taken as 39.4 in. (1.0 m) and the ASTM A1035/
A1035M07 longitudinal reinforcement quantities corre
sponded to A
s,min
and 2A
s,min
. The Mast fexural method was
used for f
D
= 100 ksi (690 MPa), and the Appendix B method
was used for f
D
= 120 and 144 ksi (828 and 994 MPa). As
shown in Table 3, the same minimum h results, regard
Fig. 8Infuence of permissible defection limit on maximum spandepth ratio. (Note: 1 ksi = 6.89 MPa;
1 lb/ft
2
= 0.048 kPa.)
Table 3Design example data and results summary
Flexural design method
Direct defection calculations DTC
Mast Appendix B Appendix B Mast
f
D
, ksi (MPa) 100 (690) 120 (828)
*
144 (994)
100 (690)
f
c
= 5 ksi (35 MPa)
L = 20 ft (6 m)
A
s
= A
s,min
h, in. (mm) 12.3 (313) 12.3 (313) 12.3 (313) 16.5 (420)
M
r
, kipft (kNm) 56 (76) 67 (91) 79 (108) 77 (105)
A
s
, in.
2
(mm
2
) 0.68 (438) 0.68 (438) 0.68 (438) 0.91 (588)
Dh, % 0 0 34
DM
r
, % 20 42 38
DA
s
, % 0 0 34
f
c
= 5 ksi (35 MPa)
L = 20 ft (6 m)
A
s
= 2A
s,min
h, in. (mm) 11.3 (286) 11.3 (286) 11.3 (286) 16.5 (420)
M
r
, kipft (kNm) 90 (123) 108 (147) 128 (174) 140 (190)
A
s
, in.
2
(mm
2
) 1.24 (802) 1.24 (802) 1.24 (802) 1.82 (1176)
Dh, % 0 0 47
DM
r
, % 20 41 54
DA
s
, % 0 0 47
*
Yield stress assumed as 0.2% offset value.