Professional Documents
Culture Documents
TECHNICAL PAPER
INTRODUCTION
The use of high-strength concrete (HSC) in many special
structures (such as long bridges) is nowadays a rational
option to fulfil many requirements, such as strength,
durability, and economy. Even for normal structures (such
as buildings), the use of HSC can lead to competitive
economical solutions because the structural members are
smaller than normal-strength concrete (NSC) members. This
is mainly true for members with a high level of compression
(such as columns). In consequence, HSC reduces self-weight
and inertia. Those reductions constitute an important advantage
when structures are located in seismic regions.
Because of their economic advantages, HSC structures
were initially used without being sufficiently studied. Some
aspects of its mechanical behavior were assumed to be an
extension of NSC. Some of the code rules developed for
NSC must be fully studied to check their applicability to
HSC. Some of this work has already been carried out. As a
result, some codes already incorporate design rules for
concrete strengths higher than 50 MPa (7250 psi). The
Norwegian,1 the Canadian,2 the New Zealand,3 the European,4,5
and ACI6 codes are examples. Nevertheless, some codes4
still suggest that for concrete strengths above 50 MPa (7250 psi),
the rules have to be used with caution. In fact, some aspects
of the structural behavior of HSC members are either not
fully known or even completely unknown. This is the case of
beams under torsion. The new design rules for torsion may
need a large number of tests to confirm some theories.
Therefore, the ongoing research efforts need to be continued
to be able to correct future version of the codes.
In actual structures, torsion forces are usually combined
with moments, shear, and axial forces, but in some structures,
such as bridges, the torsion can become very important for
the design. Furthermore, the design procedures based on
force interactions need to know the behavior under pure
ACI Structural Journal/January-February 2009
39
RESEARCH SIGNIFICANCE
So far, only a very limited number of studies on HSC
beams under pure torsion have been carried out, among
which the publications by Rasmussen and Baker9,10 and
Wafa et al.11 are examples. These initial studies comprise
laboratory programs testing only small rectangular plain
beams. The authors did not find any study focused on HSC
hollow beams. As far as torsional ductility is concerned,
HSC beams are more problematic than NSC beams, and
hollow beams are more problematic than plain beams. The
authors are presenting a study on HSC hollow beams that
provides much needed information.
TEST PROGRAM
Test specimens
Sixteen rectangular hollow beams 5.90 m (232.28 in.) long
were tested on a ring that fixed one end and applied a torsion
force on the other end. The beams had a hollow square cross
section 0.60 m (23.62 in.) wide with walls 0.10 m (3.94 in.)
thick (Fig. 1). These dimensions were selected so that there
would be some similarity with the beams tested by Lampert
and Thurlimann.13 The ends of the test beams were designed
to be fixed to the heads of the testing equipment. The variables
of the experimental program were the compressive strength
of the concrete and the amount of steel reinforcement. The
concrete strength of cylinder specimens varied from 46.2 to
96.7 MPa (6699 to 14,022 psi). Three beam series were cast
and testedSeries A, B and Cclassified according to
concrete strength. Within each series, the limit percentages
of torsion steel reinforcement were the recommended
maximum and minimum values given by the Code (among
the investigated codes) that allowed the largest range for the
percentage of steel reinforcement.14 The purpose was to
use the largest range of reinforcement ratio allowed by all
the analyzed codes.
Table 1 presents a summary of the properties of the test
beams, including the average thickness of the walls of the
cross section (t), the identification of the torsion reinforcement,
the distances between centerlines of legs of the stirrups (x1
and y1), the area of longitudinal reinforcement (Asl) and of
Fig. 1Geometry and examples of test beams details. (Note: 1 in. = 2.54 cm; 1 m =
38.37 in.).
40
t,
cm
Longitudinal
reinforcement
Transversal
reinforcement, @s, cm
x1,
cm
y1,
cm
Asl,
cm2
Ast,
cm2
fc,
MPa
l,
%
t,
%
tot,
%
mb
A-48.4-0.37
9.8
48 + 166
6@9
53.7
54.7
6.53
0.28
48.4
0.18
0.19
0.37
0.96
A-47.3-0.76
10.7
412 + 1210
8@8
53.8
53.1
13.95
0.50
47.3
0.39
0.37
0.76
1.04
A-46.2-1.00
10.9
1612
10@9.5
54.0
53.5
18.10
0.79
46.2
0.50
0.49
1.00
1.02
A-54.8-1.31
10.4
416 + 2010
10@7
52.0
52.5
23.75
0.79
54.8
0.66
0.65
1.31
1.01
A-53.1-1.68
10.4
416 + 2012
12@8
52.8
52.8
30.66
1.13
53.1
0.85
0.83
1.68
1.03
B-75.6-0.30
10.1
206
6@11
53.9
54.4
5.65
0.28
75.6
0.16
0.14
0.30
1.11
B-69.8-0.80
10.8
412 + 208
8@7.5
53.3
53.4
14.58
0.50
69.8
0.41
0.40
0.80
1.02
B-77.8-1.33
10.9
416 + 2010
10@7
53.5
53.7
23.75
0.79
77.8
0.66
0.67
1.33
0.99
B-79.8-1.78
11.2
1616
12@7.5
52.3
53.6
32.17
1.13
79.8
0.89
0.89
1.78
1.01
B-76.4-2.20
11.6
2016
12@6
51.8
51.8
40.21
1.13
76.4
1.12
1.09
2.20
1.03
C-91.7-0.37
9.7
48 + 166
6@9
54.0
54.9
6.53
0.28
91.7
0.18
0.19
0.37
0.96
C-94.8-0.76
10.0
412 + 1210
8@8
53.2
53.3
13.95
0.50
94.8
0.39
0.37
0.76
1.04
C-91.6-1.29
10.3
416 + 2010
10@7
54.5
54.0
23.75
0.79
91.6
0.66
0.63
1.29
1.05
C-91.4-1.71
10.3
416 + 2012
12@8
54.6
54.5
30.66
1.13
91.4
0.85
0.86
1.71
0.99
C-96.7-2.07
10.4
420 + 1216
12@6.5
54.0
54.3
36.69
1.13
96.7
1.02
1.05
2.07
0.97
C-87.5-2.68
10.4
2416
12@5
53.3
52.9
48.25
1.13
87.5
1.34
1.34
2.68
1.00
Note: 1 in. = 2.54 cm; 1 in.2 = 6.45 cm2; and 1 MPa = 145 psi.
Series A
Series B
Thin aggregate, kg
205
164
Series C
83
Coarse aggregate, kg
914
908
766
718
734
780
360
375
530
Admixture, L
4.1
4.8
15
Silica fume, kg
41
60
Water, L
145
145
146
W/(C + SF)
0.40
0.35
0.25
Tcr ,
kNm
cr ,
degree/m
(GC)I,
kNm2
(GC)II,
kNm2
Tty,
kNm
A-48.4-0.37
104.08
0.071
84,070
3807
128.37
A-47.3-0.76
109.50
0.064
97,950
7559
239.36
1.08
247.06
1.17
254.77
1.66
1.13
2.40
2.12
A-46.2-1.00
113.27
0.057
113,315
8337
272.65
1.16
259.17
1.07
299.91
1.54
1.11
1.57
1.41
A-54.8-1.31
120.87
0.063
109,485
11,100
360.87
1.56
368.22
1.66
368.22
1.66
1.61
1.76
1.09
A-53.1-1.68
120.93
0.044
159,702
14,398
412.24
1.53
B-75.6-0.30
111.50
0.060
107,198
612
115.40
0.21
115.95
0.23
0.21
0.48
2.29
1.67
ly ,
Tr ,
ty,
y,
u,
degree/m Tly , kNm degree/m Tr , kNm degree/m degree/m degree/m
0.44
131.84
0.57
150.78
1.18
0.50
1.50
mq
2.99
B-69.8-0.80
116.72
0.044
151,621
6715
265.83
1.33
273.28
1.42
273.28
1.42
1.37
2.30
B-77.8-1.33
130.45
0.043
172,940
9790
355.85
1.45
B-79.8-1.78
142.93
0.061
134,251
16,590
437.85
1.24
B-76.4-2.20
146.26
0.066
126,968
15,698
456.19
1.24
C-91.7-0.37
117.31
0.038
177,073
4523
149.96
0.46
151.76
0.72
0.46
0.83
1.80
C-94.8-0.76
124.46
0.049
146,640
6924
244.78
1.07
246.55
1.10
266.14
1.44
1.09
1.66
1.54
C-91.6-1.29
131.93
0.064
118,473
9209
347.65
1.46
351.16
1.53
1.46
1.52
1.04
C-91.4-1.71
132.60
0.051
148,258
12,989
450.31
1.50
C-96.7-2.07
138.34
0.051
156,268
16,371
467.26
1.34
C-87.5-2.68
139.09
0.054
146,963
19,294
521.33
1.27
Note: = beams with brittle failure; 1 in.-kips = 0.113 kNm; 1 degree/in. = 39.37 degrees/m; 1 kNm2 = 348.42 in.2-kips.
EXPERIMENTAL RESULTS
Torque-versus-twist curves
Figure 5 shows the curves with torsion moment (torque) T
versus average angular deformation per unit length (twist)
m for all the series of tested beams. The average angular
deformation m was obtained by dividing the transversal
angles measured at Sections A-A by the distance between
Sections A-A and J-J sections (5.35 m [210.6 in.], as
shown in Fig. 2). Each individual T-m curve presents
some important points: cracking (), yielding of transversal
reinforcement (), and yielding of longitudinal reinforcement (). The yielding points were identified from the recorded
experimental data obtained through the strain gauges
ACI Structural Journal/January-February 2009
Table 4Equations to compute torsion strength and limits ratios of transversal reinforcement
ACI 318-8914
ACI 318-056
Tn = Tc + Ts
2
4h x y
T c = ------ ------- ( 2.4 f c )
x 3
At t x1 y1 fy
T s = -----------------------s
MC 904
2F Rtwi A ef
T n = T Rtwi = --------------------zi
z
F Rtwi = A Swi f yd cot g i ---i
s
Brittle failure:
2F Rcwi sin i A ef
T Rcwi = ---------------------------------zi
F Rcwi = f cd2 t i z i cos i
2A o A t f yv
- cot g
T n = ------------------s
Brittle failure:
2
8 f c 1.7A oh
T n = -----------------------------ph
Brittle failure:
T n = 5T c
bw s
A t, min = 0.375 f c ------f yv
25b w s
-------------f yv
bw s
A t, min = 25 ------fy
4T c s
A t, max = -----------------t x1 y1 fy
EC 25
At
T n = 2A o ----f
y cot g
s
T n = T Rd2 =
= 2A k ( f ywd A sw s )cot g
T Rd1
Brittle failure:
2
0.25f c 1.7A oh
T f = ---------------------------------ph
Brittle failure:
= 2vfcd tA k sincos
bw s
A t, min = 0.06 f c ------fy
f cd2 t i sen i s
A swi, max = ---------------------------f yd
8 f c A oh s
A t, max = -------------------------p h f yv cot g
CAN3-A23.3-042
0.25f c 1.7A oh s
A t, max = -----------------------------------2p h A o f y cotg
ACI 318-056
MC 904
CAN3-A23.3-042
Beam
Tr,exp,
kNm
Tr,calc,
kNm
T r, exp
-------------T r, calc
Tr,calc,
kNm
T r, exp
-------------T r, calc
A-48.4-0.37
150.78
123.18
1.224
101.62
1.484
91.52
1.648
106.68
1.414
109.65
1.375
A-47.3-0.76
254.77
192.74
1.318
212.18
1.197
191.00
1.330
221.30
1.148
234.59
1.083
A-46.2-1.00
299.91
240.36
1.248
284.94
1.053
245.36
1.222
292.64
1.025
324.24
0.925
A-54.8-1.31
368.22
292.63
1.258
298.13
1.235
356.66
1.032
411.02
0.896
415.82
0.886
0.825
Tr,calc,
kNm
T r, exp
-------------T r, calc
EC 25
Tr,calc,
kNm
T r, exp
-------------T r, calc
Tr,calc,
kNm
T r, exp
-------------T r, calc
A-53.1-1.68
412.24
334.76
1.231
302.90
1.361
427.25
0.965
507.02
0.813
499.93
B-75.6-0.30
115.95
127.53
0.909
78.30
1.481
66.91
1.733
81.41
1.424
82.24
1.410
B-69.8-0.80
273.28
217.06
1.259
223.26
1.224
199.79
1.368
232.74
1.174
249.07
1.097
B-77.8-1.33
355.85
321.77
1.106
383.61*
0.928
351.60
1.012
405.19
0.878
437.59
0.813
B-79.8-1.78
437.85
378.99
1.155
373.33*
1.173
438.67
0.998
523.14
0.837
535.20
0.818
B-76.4-2.20
456.19
434.04
1.051
343.05*
1.330
530.00
0.861
645.08
0.707
641.56
0.711
C-91.7-0.37
151.76
147.27
1.031
102.47
1.481
91.85
1.652
107.06
1.418
110.87
1.369
C-94.8-0.76
266.14
215.73
1.234
211.00
1.261
194.67
1.367
225.56
1.180
232.85
1.143
C-91.6-1.29
351.16
312.59
1.123
386.48
0.909
346.45
1.014
399.25
0.880
418.38
0.839
C-91.4-1.71
450.31
373.79
1.205
438.19*
1.028
428.80
1.050
508.86
0.885
533.62
0.844
1.060
515.84
0.906
593.58
0.787
647.16
0.722
1.318
676.89
0.770
835.53
0.624
870.91
0.599
C-96.7-2.07
467.26
437.67
1.068
440.87
C-87.5-2.68
521.33
464.46*
1.122
395.44*
Brittle failure due to insufficient concrete compressive strength expected in these beams.
Note: 1 in.-kips = 0.113 kNm.
Fig. 6Torsion strengths: experimental versus expected values. (Note: 1 MPa = 145 psi.)
Some of these conclusions can be easily observed in
Fig. 6, which give the Tr,exp /Tr,calc ratio as a function of
concrete strength fc. These figures also present the points for
a group of nine rectangular over-reinforced HSC plain
beams tested by Rasmussen and Baker9,10 (all the beams had
brittle failure by concrete crushing) with 3.0 m (118.11 in.)
between supports. The cross-sectional dimensions (0.16 x
0.275 m [6.30 x 10.83 in.]) and the reinforcement (l =
3.47% and t = 1.41% in the effective span) were constant
for all beams. Concrete strength was the only parameter that
varied (from 57.1 to 109.9 MPa [8280 to 15,936 psi]).
Figure 6 confirms the general conclusions drawn from
Table 5, corresponding to the hollow beams tested by the
authors. Such conclusions are also confirmed by the overreinforced plain beams tested by Rasmussen and Baker.9,10
As noted by Hsu,17 the skew-bending theory14 seems to be
better suited to predicting the ultimate torque, especially for
small plain beams. For this theory, Fig. 6 also shows that
the deviations from the predicted values decrease as the
concrete strength increases.
Figure 6 shows that the dispersion of the values for ACI
318-056 is slightly higher than for ACI 318-8914 and does
not depend on the concrete strength. Nevertheless, both
codes give safe predictions. For ACI 318-05,6 the higher
dispersion of values, in each beams series, occurs for the
beams that have the smallest and the highest reinforcement
ratios. Except for these beams, the predictions of this code
are comparable with those of ACI 318-89.14 The maximum
torque predicted for the over-reinforced beams tested by
Rasmussen and Baker9,10 is notably underestimated, with
values close to 1.9.
For MC 90,4 EC 2,5 and CAN-A23.3-04,2 Fig. 6 confirms
that, for beams with the highest reinforcement ratios and for
the highest concrete strengths, these codes are not safe and
the deviations from the actual maximum torque can be
considerable. For the beams with average-to-low reinforcement
ratios, these codes underestimate the maximum torque. For
CAN-A23.3-04,2 Fig. 6 shows that the deviations are similar
to those of EC 2.5 The deviations from the test values are
particularly apparent when applying MC 904 to beams with
low reinforcement ratios. The graphic representation of the
ACI Structural Journal/January-February 2009
ACI 318-056
MC 904
EC 25
CAN3-A23.3-042
Beam
A t, ef
--------- ,
s
cm2/m
A s, min
------------,
s
2
cm /m
A s, max
-------------- ,
s
cm2/m
A s, min
------------,
s
2
cm /m
A s, max
-------------- ,
s
cm2/m
A s, min
------------,
s
2
cm /m
A s, max
-------------- ,
s
cm2/m
A s, min
------------,
s
2
cm /m
A s, max
-------------- ,
s
cm2/m
A s, min
------------,
s
2
cm /m
A s, max
-------------- ,
s
cm2/m
A-48.4-0.37
3.11
1.62
14.02
2.04
9.53
1.37
15.04
0.86
17.52
0.75
16.80
A-47.3-0.76
6.25
1.48
14.31
1.84
8.72
1.49
14.99
0.84
17.53
0.80
16.34
A-46.2-1.00
8.32
1.44
13.87
1.77
8.68
1.53
15.54
0.83
17.52
0.81
15.84
A-54.8-1.31
11.29
1.44
15.21
1.93
8.88
1.56
15.98
0.86
18.79
0.86
17.99
A-53.1-1.68
14.13
1.54
15.61
2.02
9.02*
1.56
16.25
0.90
18.56
0.84
18.50
B-75.6-0.30
2.33
1.62
18.14
2.55
11.59
2.03
21.42
1.11
23.78
1.00
27.07
B-69.8-0.80
6.67
1.48
17.57
2.24
10.66
1.94
20.33
1.04
23.59
1.02
24.38
B-77.8-1.33
11.29
1.44
18.04
2.30
11.00*
2.18
20.92
1.08
24.61
1.09
26.71
B-79.8-1.78
15.07
1.54
20.40
2.48
11.18*
2.25
23.05
1.19
26.17
1.14
30.00
B-76.4-2.20
18.83
1.54
21.81
2.42
10.60*
2.33
23.81
1.21
26.40
1.16
29.84
C-91.7-0.37
3.11
1.62
18.86
2.81
13.23
2.23
22.36
1.17
26.03
1.06
31.43
C-94.8-0.76
6.25
1.48
19.10
2.61
12.27
2.31
22.05
1.12
25.77
1.12
30.79
C-91.6-1.29
10.53
1.44
18.04
2.50
11.74
2.36
20.68
1.10
24.42
1.12
29.56
C-91.4-1.71
14.13
1.54
18.98
2.65
12.42
2.37
22.52
1.17
25.71
1.13
31.48
13.30
2.49
23.80
1.21
28.00
1.17
33.59
11.34
2.28
22.76
1.07
24.73
1.11
28.24
C-96.7-2.07
C-87.5-2.68
17.38
22.60
1.54
19.93
1.43
18.40
2.73
2.41
*Brittle
CONCLUSIONS
The study reported herein is based on the results of 16 test
beams. Although this kind of experimental work is rather
expensive, it would be better to have more tests. Nevertheless,
the number of tests is sufficient to draw some conclusions.
The results clearly show that some codes are excessively
permissive and could lead to the acceptance of brittle beams
or unsafe values of the predicted maximum torque. Of all the
studied codes in use, ACI 318-056 is the only one that has
acceptable rules with respect to minimum and maximum
reinforcement of members under torsion. The other codes
need to be corrected, especially for the type of beams tested
in this research (HSC hollow beams under pure torsion). The
authors suggest that two main actions need to be taken, as
explained in the following.
First, all the undesirable beams that are at risk of brittle
failure should be banned. This is ensured by means of an
adequate limit value for both the maximum and the
minimum amount of torsion reinforcement, that is, for fc
40 MPa (5800 psi).
Second, the codes need to be more reliable in predicting
the maximum torque, and the deviations should diminish.
This is done by using a criterion better suited to predicting
the compressive failure of the concrete struts. This objective
is related to the limitation of the maximum amount of
transversal reinforcement. It should be noted that the need to
reduce the maximum value of the transversal reinforcement
has already been stated by Rasmussen and Baker10 for a
previous version of the European and Canadian codes.
Due to the limited number of available experimental
results of HSC beams, the authors find it a little premature to
propose new design rules. At this stage, they are only
indicating the direction that those new rules should take.
The experimental work on this subject must continue
with more tests to obtain the correct rules for both NSC
and HSC.
NOTATION
Ac
Asl
Ast
At,ef
At,max
At,min
fc
fy
(GC)I
(GC)II
s
T
Tcr
Tly
Tr,calc
Tr; Tr,exp
Tty
t
u
x; y
x1 ; y 1
y
48
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
cr
ly
m
Tr
ty
u
y
l
t
t,max
t,min
tot
=
=
=
=
=
=
=
=
=
=
=
=
=
REFERENCES
1. Norwegian Standards, Concrete Structures, Design Rules, NS3473,
Stockholm, Sweden, 1989, 78 pp.
2. Standards Council of Canada, Design of Concrete Structure for
Buildings (CAN3-A23.3-04), Canadian Standards Association, Mississanga,
Canada, Dec. 2004, 240 pp.
3. Standards Association of New Zealand, Concrete Structures NZS
3101Part 1: Design, 1995, 256 pp.
4. CEB-FIP, Model Code 1990, Comit Euro-International du Bton,
Lausanne, Switzerland, 1990, 461 pp.
5. CEN prEN 1992-1-1, Eurocode 2: Design of Concrete Structures
Part 1: General Rules and Rules for Buildings, Brussels, Belgium, Apr.
2002, 225 pp.
6. ACI Committee 318, Building Code Requirements for Structural
Concrete (ACI 318-05) and Commentary (318R-05), American Concrete
Institute, Farmington Hills, MI, 2005, 443 pp.
7. Priestley, M. J. N.; Seible, F.; and Wang, C. M., The Northridge
Earthquake of January 17, 1994Damage Analysis of Selected Freeway
Bridges, Report No. SSRP94/06, University of California, San Diego,
CA, Feb. 1994, 266 pp.
8. Ali, M. A., and White, R. N., Toward a Rational Approach for
Design of Minimum Torsion Reinforcement, ACI Structural Journal,
V. 96, No. 1, Jan.-Feb. 1999, pp. 40-45.
9. Rasmussen, L. J., and Baker, G., Torsion in Reinforced Normal- and
High-Strength Concrete BeamsPart 1: Experimental Test Series, ACI
Structural Journal, V. 92, No. 1, Jan.-Feb. 1995, pp. 56-62.
10. Rasmussen, L. J., and Baker, G., Torsion in Reinforced Normal- and
High-Strength Concrete BeamsPart 2: Theory and Design, ACI Structural
Journal, V. 92, No. 2, Mar.-Apr. 1995, pp. 146-156.
11. Wafa, F. F.; Shihata, S. A.; Ashour, S. A.; and Akhtaruzzaman, A. A.,
Prestressed High-Strength Concrete Beams under Torsion, Journal of the
Structural Engineering, ASCE, V. 121, No. 9, Sept. 1995, pp. 1280-1286.
12. Bernardo, L. F. A., Toro em Vigas em Caixo de Beto de Alta
Resistncia (Torsion in Reinforced High-Strength Concrete Hollow
Beams), PhD thesis, University of Coimbra, Portugal, 692 pp. (in Portuguese).
13. CEB, Torsion, Bulletin dInformation, No. 71, Lausanne, Switzerland,
Mar. 1969, 207 pp.
14. ACI Committee 318, Building Code Requirements for Reinforced
Concrete (ACI 318-89) and Commentary (318R-89), American Concrete
Institute, Farmington Hills, MI, 1989, 369 pp.
15. Shin, S.-W.; Kamara, M.; and Ghosh, S. K., Flexural Ductility,
Strength Prediction, and Hysteretic Behavior of Ultra-High-Strength
Concrete Members, High Strength Concrete, Second International
Symposium, SP-121, W. T. Hester, ed., American Concrete Institute,
Farmington Hills, MI, 1990, pp. 239-264.
16. Bernardo, L. F. A., and Lopes, S. M. R, Flexural Ductility of HighStrength Concrete Beams, Structural Concrete, V. 4, No. 3, 2003,
pp. 135-154.
17. Hsu, T. T. C., Torsion of Reinforced Concrete, Van Nostrand
Reinhold Co., 1984, 516 pp.
18. Mitchell, D., and Collins, M. P., Diagonal Compression Field
TheoryA Rational Model for Structural Concrete in Pure Torsion, ACI
JOURNAL, Proceedings V. 71, No. 8, Aug. 1974, pp. 396-408.