Documente Academic
Documente Profesional
Documente Cultură
ABSTRACT: This paper reports the development of a new design guideline for the stair slabs, supported on
the structural member spanning at right angle to the direction of the flight. The study is based on finite-element
analyses of two common types of stairs, dog-legged and open-well stairs. Results of the finite-element inves-
tigation have been compared with the experimental findings, ensuring reliability of the predicted response.
The stair slabs, usually supported on walls or beams at landing levels, derive significant rigidity from such
supporting arrangement. This results in an efficient distribution of moment along the length of the flight. This
distinctive feature of stair slab is not recognized by the leading codes of practice. In order to establish the
Downloaded from ascelibrary.org by KANSAS STATE UNIV LIBRARIES on 06/18/14. Copyright ASCE. For personal use only; all rights reserved.
characteristic behavior of stairs of the aforenamed types, a sensitivity study for the geometric parameters of
the stair slab has been carried out. Based on the findings of the present study, a new design rationale for stair
slab has been proposed. The guidelines presented here enable straightforward estimation of the design forces,
which are generally overestimated by 300-400% when determined by the conventional design practices.
INTRODUCTION
l
Owing to their inherent supporting arrangements, the stair AI A2 220
1\\1 1481
slab with supports on three sides at landing level have a very
efficient way of transferring load. The leading codes of prac-
I 1
'~ 'I I
~~~1481 n
tice do not provide proper appreciation to the actual behavior
of stair slab of this type.
---- - - - --ur.r.
.//
I A3
- _
Regarding the design of stairs, the American Concrete In- 4fol ·1- 2515 ~'I
250 1220 1220 250
stitute (ACI) Code (1989) does not provide any recommen- (a) 1101 (48) 1991 (48) llOl
dation for the restraining effect on the stair slab owing to its
support conditons. As a result, users of the code usually de-
sign a stair slab as a simple, one-way slab with span equal to
the horizontal distance between the supports (Winter and X AI A2 1220 X
(48)
Nilson 19(1). The British Code (CP 110, "Code" 1972), on L _
the other hand, provides some reduction in the effective span
of a stair slab. Reduction in the effective span as suggested
------- --\-_-~-~~--1--- , I ,
1220---.11
(48)
by the British Code, obviously results in some saving in the : I 1 A3 1220
, 1 1 1481
design. A limited experimental study (Saqib and Ahmad 1982) U_ .I - L +r,...........,JA
on a full-scale, single-flight stair suggested that the stair slab
does not respond in the same manner as do the simple one- 2$15
250 1220 1220250
way slabs. The study further revealed that a typical stair slab (b) 110) (48) 1991 1481 (10)
of same thickness as designed by the British Code requires
nearly half the reinforcement required by the conventional
American design practice (Winter and Nilson 1991). A single-
flight stair designed by the British Code was constructed and
tested. It was found that the stair sustained 133% of the load
corresponding to the ultimate strength without any sign of JOOl. [;fZz2zo/
failure. This fact indicated that even though the British Code 14 1 t
of Practice regarding the design of stairs is more liberal than
the ACI Code, further savings can be made. This was en- (c)
dorsed elsewhere (Ahmed 1983; Zahedi 1985).
FIG. 1. (a) Experimental Model and Typical Dimensions of Dog-
Legged Stair; (b) Typical Dimensions of Open-Well Stair; (c) Sec-
SCOPE tional Elevation X-x for Stair of (a) and (b) [mm (in.)]
Finite-element investigations of stair slabs, built monolith-
ically with structural members spanning at right angle to the legged and open-well type (Fig. 1) of stair have been studied.
direction of the flight have been carried out. The deflections The influences of the various geometric parameters on the
obtained by the finite-element method have been compared design forces have also been studied. Based on the findings
with the experimental results. The behaviors of both dog- of this study, a guideline for the estimation of design forces
is proposed. A comparison of the proposed and the conven-
'Asst. Prof., Dept. of Civ. Engrg.. Bangladesh Univ. of Engrg, and
Techno!.. Dhaka 1000. Bangladesh. tional design approaches has also been made.
'Assoc. Prof.. Dept. ofCiv, Engrg., Bangladesh Univ. of Engrg. and
Techno!.. Dhaka 1000. Bangladesh. EXPERIMENTAL VERSUS FINITE-ELEMENT MODEL
'Prof.. Dept. of Civ, Engrg.. Bangladesh Univ. of Engrg. and Tech-
no!., Dhaka 1000, Bangladesh. An experimental model for a stair slab was constructed
Note. Associate Editor: Arturo E. Schultz. Discussion open until [Fig. l(a)] and tested by Ahmed (1983). Test results of Ahmed's
December I. 1995. To extend the closing date one month, a written
request must be filed with the ASCE Manager of Journals. The manu-
(1983) study are compared with the results of the finite-ele-
script for this paper was submitted for review and possible publication ment investigation carried out by the present writers. For this
on September 15, 1993. This paper is part of the Journal of Structural purpose, the finite-element mesh of Fig. 2 has been used with
Engineering. Vo!. 121, No.7. July, 1995. ©ASCE, ISSN 0733-9445/951 thin shell linear element (stiff63) of ANSYS (Desalvo and
0IKl7-1051-1057/$2.1Kl + $.25 per page. Paper No, 6984. Swanson 1985). This element has four nodes with six degrees
JOURNAL OF STRUCTURAL ENGINEERING 1 JULY 1995/1051
~'~4
experimental results (Ahmed 1983) and the finite-element
Z ZZ ;il •
I
results of the present study. Fig. 3 presents this comparison
A of the load-deflection response. The overall deflection curve
IZ 20 ------.lJ obtained from the experimental study and that predicted by
1481
the present finite-element analysis are in good agreement,
INNER EDGE
x especially in the initial range of the load-deflection curve
where the material behavior is essentially elastic. However,
Z4 12ZO
148J
since the present finite-element analysis is based on a linearly
elastic material behavior, some difference is observed in the
Plan
postyielding loading range. It is to be noted that in construc-
z tion the stair slab for testing, concrete with a crushing strength
Downloaded from ascelibrary.org by KANSAS STATE UNIV LIBRARIES on 06/18/14. Copyright ASCE. For personal use only; all rights reserved.
of 17 MPa (2,500 psi) had been used. The present day tend-
r 15Z4
1601 Dimtnsions art In Mm linch J ency of using higher strength concrete 21 to 35 MPa (3,000
to 5,000 psi) is likely to produce a linear behavior up to a
load level higher than that observed in the experiment. Ar-
guably, this ensures that the predicted behavior correctly sim-
FIG. 2. Finite-Element Mesh for Stair of Fig. 1 ulates the actual response for the range of loads producing
elastic stresses.
Inch Since stresses were not measured during the experiment,
14 the analytical stresses could not be compared. However, the
experimental model sustained the design ultimate load with
12
a comfortable margin of safety. This ensures the reliability
10
of the analytical stress values.
E
.,a- S FINITE-ELEMENT ANALYSIS
.....
Z
... 6 Assumptions
..,
Cl
0 4 The study was carried out with the following assumptions.
--I
1. Double flight stairs having intermediate landing of the
dog-legged and open-well types are analyzed. Typical di-
mensions considered for the dog-legged stair and the open-
well stair are as shown in Fig. 1. For the purpose of analysis,
Central Deflection IlIIm) three consecutive flights have been considered. Symmetry
FIG. 3. Comparison of Load-Deflection Responses
conditions were applied at both the terminal landings (see
Fig. 4). The results of the intermediate flight (i.e., flight 6-
5-3-4 of Fig. 4) is considered for discussion in the subsequent
sections, since it was contemplated that this particular flight
would be free from any undue effect (if at all) of the boundary
conditions imposed at the terminating landings.
2. Waist slab and landing slab were assumed to have the
same thickness.
3. The material was assumed to be linearly elastic, ho-
mogeneous, and isotropic. A modulus of elasticity of 210 kN/
mm 2 and Poisson's ratio of 0.18 were assumed.
IQ I big I
4. Both live and dead loads were applied as gravity loads.
(a) A live load of 4.8 kN/m 2 , a dead load for IOU mm thick slab
(for a material unit weight of 23.5 kN/m') and a tloor finish
of I kN/m 2 were considered. Additional weight due to the
v Z
steps was also considered in the inclined waist portion of the
\Lx I'?777}l
I:'LLLLLJ
R~Tlxl'O
uIV) • 0 slab.
_ U\ZI'O 5. A bending moment, producing compression at the top
is considered positive. Bending in the longitudinal direction,
U(X)·O
which produces stresses in x-direction (Fig. 2) is termed as
M,. Similarly, moment-producing stresses in y-direction (Fig.
2) is designated as M,.
Finite-Element Mesh
Finite-element analyses of both dog-legged and open-well
type stairs have been carried out using the thin shell linear
element of reference (Desalvo and Swanson 1985). The ele-
FIG. 4. Illustration of: (a) Flights under Study 1-2-3-4-5-6-7-8; (b) ment mesh of Fig, 2 was used. The mesh comprised of 161
Boundary Conditions for a Three-Flight Analysis nodes in 128 elements was chosen as an optimum choice to
give the upper bound displacement (Ahmed 1989).
of freedom per node. Shear deformation is neglected in this
Bending Moment Distribution
thin shell element. The nodes along the walls (Fig. 1) at both
landings were restrained against vertical displacement. The The bending moment (M,) diagrams along the length of
midspan deflection along the inner edge of the stair slab shown the flight for stair slabs of typical dimensions as shown in Fig.
1052 I JOURNAL OF STRUCTURAL ENGINEERING I JULY 1995
U
C
0 0 is carried in the direction perpendicular to the length of the
Q>
ro fight at landing level. From Figs. 6 and 7, it can be seen that
the region closest to the kink is subjected to considerably high
-500 moment in the lateral direction (y), although this high mo-
ment is very localized.
-4 L-_-----l_ _-'-_--;----L.._-,.-:'-=_-=-==-_::-:!
o 1,000 2,000 3,000 4,000 5,000 6,000 Parametric Study
Distance in mm
The effects of varying the geometrical parameters on the
FIG. 5. Bending Moment M. for Dog-Legged and Open-Well Stair overall behavior of stair have also been studied. The param-
eters considered are (Fig. 4): (1) Length of the landing slab,
Inch a; (2) Horizontal projection of the waist slab (going), b; (3)
}
E 20 40 60 80 100
...... 1'2 2700 Height of the flight, h; (4) Width of the flight, w; and (5)
Z
E
~',
... X2 Opening c of an open-well stair [Fig. I(b)].
..>c:
8 - X3
( S•• 1800 In addition to the variation of the aforementioned geo-
E
Ql -0- X4
Fiq.21 C metrical parameters, three different distributions of live load
E
0 4 900 j were also considered. In the first case, full live load was
E I
~ applied on the waist slab and on the lower landing (of the
01
c
0 second flight), and the upper landing was loaded with dead
'i5
c:
Ql
load only; while the other two flights were loaded with full
CD
-4 -900 live load. Loading case 2 was composed of full live load on
0 127!5 2550 all panels. The third loading case was considered with full
Distance in mm live load on the waist slab portion of the second flight, while
the two landing slabs of the second flight were loaded with
FIG. 6. Bending Moment My at Different Locations of Landing Slab
for Dog-Legged Stair dead load only. As in the previous case, the first and third
flights were completely loaded with full live load. The bending
moments for these loading cases have been compared in Fig.
·E
E 8. Fig. 8 shows that the. influence of the position of live load
12 -o-XI 2700
Z
.><
considered is not very significant for the positive moments.
_X2
C /I ....lCS
[SI!E Fill. 2 1800~ For all the three load cases presented, the live load has been
Q>
E ~ present in the middle portion (waist slab), the location where
0
E
. ..... X4
900 Q the maximum positive moment occurs. The presence of live
OJ
c 0
load on the landing has no effect on the maximum positive
uc: 0
moment. The landing slab with or without live load has some
Q>
ro -4 -900 influence on the moments near the kink, although this effect
0 1000 2000 3000 4000
Distance in mm
is not significant (Fig. 8). For a dog-legged stair, parameter
c was taken as zero. Each of the parameters described before
FIG. 7. Bending Moment My at Different Locations of Landing Slab was varied independently keeping the remaining parameters
for Open-Well Stair constant. Results of these variations are shown in Tables I-
S. Effective span values shown in these tables were computed
on the basis that qb~/8 should be equal to the maximum
Inch
100 200 positive moment at midspan; where be is the effective span
6 ,--------:L-o-ad7c-a-se-1,..--:To"-a--:d-cas-e-:'"2-----:L-o-ad7ca-s--'eT'-------, (i.e. distance between the points of contraflexure), and q is
E ··,0--- ······A·.. ·
the total load per unit area on the waist slab duly magnified
E ~
z
.><
4
DL-
r ~ Load case 1
~Loadcase2
- 1000
on a horizontal projection.
The important results of a much wider study (Ahmed 1989)
~ 2
~ Load case 3 _. 500
E ~
a c: TABLE 1. Variation of Length of Landing a
; a b""'i;---++-------+.---...,.-:;;:~.---___l a ~
c: Maximum
U
~
c
-2 t a b 1= (2a + b) h w
positive Effective
moment span, be
~ x 100
- -500
(mm) (m) (m) (m) (m) (m) alb (kN'm/m) (m) (%)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
-4 a 1,000 2,000 3,000 4,000 5,000 6,000
100 1.22 2.52 4.95 1.52 1.22 0.49 4.34 I.H2 72.50
Distance in mm 100 1.37 2.52 5.26 1.52 1.22 0.55 4.11 1.77 70.50
100 1.52 2.52 5.56 1.52 1.22 0.61 4.06 1.76 70.20
FIG. 8. Inner-Edge Bending Moments for Different Load Cases
t a b 1= (2a + b) h w
positive Effective
moment span, be
~ x 100
6 Max +w moment
Max -va moment
(mm) (m) (m) (m) (m) (m) hlb (kN'm/m) (m) (%) 1,200
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) ~ 5
ilK) 1.22 2.52 4.95 1.45 1.20 0.58 4.19 1.79 71.22 ~ Ii ,.;:·::···:·:·
.......
·.········· ·0···· .., ,,".- 1,000
ilK) 1.22 2.52 4.95 1.52 1.20 0.61 4.34 1.82 72.50 E
~
ilK) 1.22 2.52 4.95 1.60 !.20 0.64 4.44 U4 73.35 ~ 4 c
fj
E - - w/b
OJ 800
c
'5 b/l
TABLE 3. Variation of Going b ~ 3
600
Downloaded from ascelibrary.org by KANSAS STATE UNIV LIBRARIES on 06/18/14. Copyright ASCE. For personal use only; all rights reserved.
Maximum
t
positive Effective
~ x 100 400
a b 1= (2a + b) h w moment span, be b
0.35 0.4 0.45 0.5 0.55 0.6 0.65
(mm) (m) (m) (m) (m) (m) bll (kN'm/m) (m) (%)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
wlb
ilK) 1.22 2.52 4.95 1.52 1.22 O.5! 4.34 1.82 72.50 FIG. 10. Results of Variation of Flight Width wand Going Width
ilK) 1.22 2.67 5.11 1.52 1.22 0.52 4.49 1.86 69.60 b
ilK) 1.22 2.82 5.26 1.52 1.22 0.54 4.86 1.93 68.40
Inch
8 r- ----"25 5,,0'----_""]
r
TABLE 4. Variation of Width of Flight w in Dog-Legged Stair
Maximum ...... - 1,500
6.--
positive Effective
~ x 100
t a b 1= (2a + b) h w moment span, be b
(mm) (m) (m) (m) (m) (m) wlb (kN'm/m) (m) (%) c=305mm - 1,000
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) " c=610 mm
ilK)
o
1.22 2.52 4.95 1.52 1.91 0.36 4.51 1.86 73.80
ilK) 1.22 252 4.95 1.52 1.22 0.48 5.06 1.97 78.17
ilK) 1.22 2.52 4.95 1.52 1.37 0.55 5.38 2.03 80.55
100 1.22 2.52 4.95 1.52 1.52 0.61 5.73 2.10 83.33
....
C
E
2 . 9 Q - f ' - - - - - - - - - - - - - - - - - - T 0.114
18.0 4000
u
CI 0.112 ~
!-
E
, 13.:5 o()o C. 30Smrn 3000
'0_ z .. C· 610mm
_E .e ~ .e
0.110 9.0 .... Ce 915mm 2000 ......
~~ " i
...> 2.75 0.108
.5
~
E 4.:5
-0- C- 1220mm
1000
.E
J
"'"
o '0
~
0
2.70 -h.......-r-r-..,-,.....,....,.-,---r-~r___T_._.,.._ll,__r'_1c_T_; 0.106 i.J
0.45 0.50 0.55 0.80 0.65 - 4.5
alb 0
-
..lII::
cQ)
E
0
9.0
4~0
..... w/ba 0.36
... ';/baO.48
... w/bao.8'
2000
1000
e
<:::
.~
1
along the inner edge of the flight.
The effective span b('. as seen from Fig. 5, can conserva-
tively be estimated to be equal to 70% and 80% of the going
(the horizontal projection ofthe inclined portion) respectively
E ~
for dog-legged (c = 0) and open-well stair (c = 1.22 m).
Cl
C 0 0 Parameters c and w appear to be the most important pa-
=ac rameters influencing the behavior of the open-well stair under
Q)
aJ -4.~0 -1000 consideration. It is interesting to note that the flexural be-
0 300 1200
havior at outer edge is insensitive of these parameters. In
Downloaded from ascelibrary.org by KANSAS STATE UNIV LIBRARIES on 06/18/14. Copyright ASCE. For personal use only; all rights reserved.
Distance in mm fact, the outer edge behaves somewhat like a restrained beam.
The effect of a restrained outer edge travels towards the inner
FIG. 13. Variation of Moment My at Line Z4 for Varying Parameter edge. Some relaxation of this restraining effect (at the inner
w for Open-Well Stair (C 1,220 mm) = edge) takes place for higher values of dimension c and w.
For c = w, the effective span at inner edge is of the order of
Inch 80% of the going (Fig. 5).
200 8,000
30
Inner and outer edges PROPOSAL FOR DESIGN GUIDELINE
6.000
E ~ Based on the preceding characterization of the behavior of
E 20
4.000 stair slab, the critical locations for moment, are identified as:
Z
-'"
c: Q)
10 2,000 c:
• Midspan of the flight, for a positive moment in the lon-
E '"
a
E 0 0 ~ gitudinal (x) direction.
• Kink location, for a negative moment in the longitudinal
~
Ol
c
'0
cQ) -10 -2.000 direction.
m " Fixed Ended • Landing slab, (a strip adjacent to the kink line of half
-4,000
-20 Inner end outer edges the width of the landing, a for moment in lateral direction
(y).
1,000 3,000 4,000 6,0004;,000
0 2,000
Distance in mm The guidelines for estimating the design moments for the
two different types of stair considered in this paper are pre-
FIG. 14. Flexural Behavior of Stair Slab for Different Arrange-
sented in the next section.
ments of Support
Dog-Legged Stair
in an increased longitudinal moment in the stair slab. The
reason for such behavior can be attributed to the fact that Midspan Positive Moment
the restraint provided by the landing slab becomes less ef-
Because of the sensitivity of the various geometrical pa-
fective as the well opening widens.
rameters, it is difficult to specify a general rule to quantify
For an open-well stair, the lateral moment My at landing
the maximum positive moment. However, it will be suffi-
does vary to a great extent depending on the weH opening c
as shown in Fig. 12. It is interesting to note that at the free ciently conservative (and still economical, compared to the
conventional practice as will be shown later) to assume an
edge of the landing (i.e. near kink line) the moment My
effective span equal to 90% of the going of the stair slab. In
increases as the distance (c) between the forward and the
other words, a moment formula can be assigned
backward flights decreases. The effect of variation of the
width of the flight (w) in an open-well type of stair on the (1)
moment M, at landing level is shown in Fig. 13. Increasing
the flight width increases the lateral moment My at landing where M" = ultimate moment; q = total load per unit area
only moderately near the kink line. However the variation on the waist slab, duly magnified on a horizontal plane; and
in M, is insignificant, away from the kink line. b = horizontal projection of the going of the stair.
The discussion in the present paper so far was restricted The preceding recommendation will be adequate for de-
to one particular supporting arrangement for the stair slab termining midspan positive moment for a dog-legged stair
(see Fig. 1). Behavior of the stair slab with two more sup- with the landing slab supported on the outer edge walls.
porting arrangements were also studied: (1) a stair slab simply
supported on walls at the end of both landings; and (2) a stair Negative Moment at Kink
slab completely fixed at wall at the end of both the landings. The negative moment at kink can be taken to be equal to
Results of these analyses are presented in Fig. 14. As men- the maximum positive moment specified earlier; i.e., a mo-
tioned earlier, the results presented in this figure also cor- ment value of qb 2110 is proposed for kink location.
respond to the second flight of a three-flight stair. The inner
edge and outer edge values match closely, despite the pres- Moment in Lateral Direction at Landing Level
ence of adjoining flights. The behaviors of these two cases
very closely resembles that of simple and fixed-ended beams, At the landing level of a dog-legged stair, positive moment
respectively. This implies that the stair slab does not require of considerable magnitude occurs. This moment is dominant
any special treatment because of is folded nature. only over a small region of the landing. Since the existence
The preceding findings reveal that stair slabs being sup- of this moment is concentrated over a small region, consid-
ported on landings, running at right angle to the direction of erable redistribution of stresses will result before failure takes
the flight. derive significant restraints from such supporting place. Hence, an approximate specification in ascertaining
JOURNAL OF STRUCTURAL ENGINEERING / JULY 1995/1055
OL == WIOO + WI + W, (7a)
where WHKl == self weight of 100 mm slab; WI == weight of
Midspan Positive Moment floor finish; and W, == self weight of steps, given here as
The midspan positive moment, for an open-well staIr IS DL == 2.35 + 1.0 + 1.8 kN/m 2 == 5.15 kN/m 2 (7h)
sensitive to the parameters c (the opening between the con-
secutive flights) and w (the width of the flight). However, an The total factored load on the horizontal projection is
effective span equal to the going of the stair would produce q == (1.7'4.8 + 1.4,5.15)'1.17 == 18.0 kN/m 2 (8)
a positive moment, which may be deemed satisfactory for the
common range of these parameters. Therefore, as in (1) Using (1), the ultimate moment is computed as follows:
where M" == maximum positive moment, q is the total load For a 20.7 MPa concrete and grade 300 steel, a slab thickness
on the waist slab duly magnified over a horizontal projection of 100 mm with a steel area of 660 mm2 /m is required. The
and b is the going of the stair. negative section is also controlled by the same requirement.
for this open-well stair, span is a tendency to follow the design rules specified for a more
general type of stair. This results in a gross overestimation
II = 2· w + c = 1.22·2 + 1.22 = 3.66 m (18) of the design forces. On the basis of a finite-element inves-
Using (4) the ultimate moment is computed as tigation, supported by experimental evidence, guidelines have
been developed for simple and straightforward design of stair
M" = Q,/i/l1 = 16.23 kN· mlm (19) slabs. Economic advantages resulting from this new devel-
which requires a steel area of 791 mm /m. 2 opment have also been indicated.
APPENDIX. REFERENCES
Comparison of Proposed Method with
Conventional Methods Ahmed, I. U. (1983). "Development of improved design procedure for
stair slabs," Master's thesis, Bangladesh Univ. of Engrg. and Techno!.,
The results of the design examples presented in the pre- Dhaka, Bangladesh.
ceding section are compared with the results based on anal- Ahmed, I. (1989). "A design rationale for stair slabs: based on finite
yses by the conventional practices in Table 6. From the Table element analysis," Master's thesis, Bangladesh Univ. of Engrg. and
Techno!., Dhaka, Bangladesh.
6, the cost-effectiveness of the proposed method is very clearly
"Code of practice for the structural use of concrete, CP 110: Part 1,
demonstrated. 1972." (1972). The Council for Codes of Practice, British Standard
Institution, London, U.K.
CONCLUSIONS Desalvo, G. J., and Swanson, J. A. (1985). "ANSYS-user manua!."
Swanson Analysis System Inc., Houston, Pa.
Behavioral studies on the two different types of stair slab Saqib, M. N., and Ahmad, S. (1982). "A closer look into the design of
commonly used in the residential and office buildings have stairs." 27th Annu. Convention of the [nst. of Engrs. (Bangladesh),
been made in this paper. The codes of practice of reference Dhaka.
(Winter and Nilson 1991) and (CP 110,1972) do not give Winter, G., and Nilson, A. H. (1991). "Design of concrete structures."
11th Ed., McGraw-Hili Book Co .. New York, N.Y.
proper appreciation to the distinctive feature of the stair slabs Zahedi, M. Z. (1985). "Improved design procedure for stair slabs."
where the landings are supported by walls in three sides. In Master's thesis, Bangladesh Univ. of Engrg. and Techno!., Dhaka,
absence of specific guidelines, the practicing engineers have Bangladesh.