The Impact of 500 MPA Reinforcement On The Ductility
The Impact of 500 MPA Reinforcement On The Ductility
The Impact of 500 MPA Reinforcement On The Ductility
Synopsis:
The relatively low uniform elongation of 500 MPa steel reinforcement has significant implications in the design and
behaviour of reinforced concrete structures, particularly with regard to ductility. Fracture of the tensile steel will be the
common failure mechanism in many under-reinforced beams and slabs and the usual assumptions made in ultimate
strength design and analysis are in question. The ability of structures to redistribute internal actions at the ultimate limit
state will be compromised. This paper addresses some of these design problems, their impact on current practice and
design methodologies, and some recent and relevant changes to AS3600.
Key Words: Ductility, high strength steel reinforcement, moment redistribution, reinforced concrete, ultimate
strength design, uniform elongation.
1. Introduction
Steel reinforcement with characteristic yield stress of 500MPa is now in the Australian market place and will soon be the
only conventional reinforcement available. This steel has higher strength but significantly lower ductility than 400 MPa
tempcore bars. A new Australian Standard for reinforcing steel (AS/NZS4671) has been introduced in 2001 with the
specification of 500 MPa reinforcement covered for the first time. It has replaced the existing Standards AS1302, AS1303
and AS1304. Also issued in 2001 is a revision of AS3600, incorporating Amendment 1 (issued in 1996) and Amendment 2
(issued for public comment in 1999), to accommodate the use of 500 MPa steel in the design of concrete structures.
The proposed new Standards classify 500 MPa reinforcement according to its ductility, Class L (low ductility) and Class N
(normal ductility). Class L steel includes cold worked wires and welded wire mesh. Class N steel includes hot rolled
deformed bar. The mechanical properties of the new steels, specified by characteristic values, are summarised in Table 1.
Re is the yield stress, with Rek.L and Rek.U being the lower and upper characteristic values; Rm is the tensile strength; and su
is the uniform elongation or the strain corresponding to maximum stress (just prior to the onset of necking).
AS1302, AS1303 and AS1304 did not specify any limits on su. For 400 MPa bars, with su in excess of 0.10, this did
not impose any difficulties. However, for some of the brittle welded wire meshes used in structures, with su an order of
magnitude less than this, the lack of attention to the ductility of reinforcing steel was a major concern. The alarmingly
small values of su of some commonly used meshes are now formally recognised and endorsed as Class L
reinforcement. The formal endorsement of brittle steel (500L) for use in concrete structures, albeit with certain
limitations, is a backward step. It is a great pity that Standards Australia has bowed to pressure to permit the use of this
inadequate reinforcing material.
The relatively low value of su of the new steel has very significant implications in the analysis and design of concrete
structures, particularly with regard to ductility. The failure mechanism of under-reinforced flexural members may
change, with ultimate curvatures often being governed by tensile steel fracture, rather than failure of the concrete in the
compressive zone. This applies not only when Class L steel is used, but also in some cases for Class N steel (when su is
0.05). The amount of moment redistribution permitted in AS3600 may no longer be appropriate and plastic design
techniques may no longer apply. Indeed, even elastic analysis techniques may not be applicable at the ultimate limit
states, when using Class L steels. Concrete structures are non-linear and inelastic, and must possess some ductility if the
actual distribution of internal actions is to redistribute towards the elastic distribution. This minimum ductility may not
be available for members containing Class L steel.
Some of these design implications have been addressed in the recent revision of the Australian Standard AS3600, with
varying degrees of success, but designers need to be fully aware of the potential problems associated with the use of the
new reinforcement. This paper outlines some of these problems, their impact on current practice and design
methodologies, and some relevant recent amendments to AS3600.
597
Concrete Institute of Australia (2001) Conference Proceedings
Mu (concrete failure)
2500
s = 0.015 s = 0.05
Moment
(kNm)
Mu (concrete failure)
1500
Ast = 3600 mm2
s = 0.015 s = 0.05
1000
Mu (concrete failure)
Ast = 1800 mm2
598
Concrete Institute of Australia (2001) Conference Proceedings
When using Class L steel, the curvature at failure of each section is governed by fracture of the steel and is almost
independent of Ast (curvature at failure actually increasing slightly as Ast increases).
Graphs of ultimate curvature ( u) versus Ast for the cross-section of Fig 1 are shown in Fig 3. The solid curve depicts u
when the extreme fibre concrete compressive strain governs failure, while the dashed curves show the situation when
Class L ( su = 0.015) and Class N ( su = 0.05) steels are employed. Where the dashed curves deviate from the solid
curve, failure of the tensile steel occurs prior to compressive failure of the concrete. The concept of under-reinforced
sections necessarily providing ductility is no longer valid. Tensile steel fracture will almost always be the failure mode
when using Class L steel and will often be the failure mode when using Class N steel (if su is at or near the minimum
specified value of 0.05).
It is difficult to agree with the claims of the manufacturers and promoters of 500 MPa steel that the acceptance and
endorsement of this relatively brittle steel is a progressive advancement for the reinforced concrete industry. Benefits
resulting from the increased strength are outweighed by the problems created by the reduced ductility.
200
150
Concrete failure ( o = 0.003)
100
st = 0.05
50
st = 0.015
0
0 1800 3600 5400 7200 9000
Ast (mm2)
Figure 3 Ultimate Curvature versus tensile steel area for section in Fig 1.
599
Concrete Institute of Australia (2001) Conference Proceedings
same as that assumed in design. With the introduction of 500 MPa steel (particularly Class L with su = 0.015), this is
no longer the case. When fracture of the steel becomes the failure mode of a critical section, the ability of the structure
to find the desired design load path may be lost. Consider the following examples.
3.1 Example 1
Consider an internal, 6 m, fixed-ended span of a lightly loaded continuous one-way slab of overall depth 200 mm. The
slab is subjected to a uniformly distributed dead load, w. The material properties are f'c = 40 MPa; f'cf = 3.8 MPa; Ec =
25000 MPa; Es = 200000 MPa; fsy = 500 MPa. The main tensile reinforcement in the bottom of the slab is 400 mm2/m
throughout the span at an effective depth of 170 mm and the main tensile reinforcement at the continuous supports is
400 mm2/m at a depth of 30 mm below the top surface of the slab. The critical sections at each support and at midspan
just comply with the minimum strength requirement of AS3600 (i.e. the ultimate moment capacity Mu exceeds 1.2 times
the cracking moment, Mcr).
For this slab, at each critical section, the moment required to produce an extreme fibre concrete tensile stress of 3.8
MPa on the uncracked section is Mcr = 26.1 kNm/m. If the load on the slab is gradually applied, the negative moment at
each support before cracking is wL2/12 and the positive moment at midspan is wL2/24. Cracking will first occur at the
supports when the moment reaches -26.1 kNm/m and, soon after, the moment corresponding to first yield of the tensile
steel at the supports (Msy = -32.0 kNm/m) will be reached. At this load level, the section at midspan is uncracked and
the bending moment diagram is shown as the dashed curve in Fig 4. Further increases of load will result in a very slight
increase in moment at the support accompanied by large rotation and an increase in moment at midspan. Cracking
occurs at midspan when the moment reaches Mcr and, eventually, the bottom steel at midspan will yield when the
moment reaches Msy. The moment diagram at this load level is shown as the solid curve in Fig 4. This load level will
only be reached if the section at the support has sufficient rotational capacity to allow the necessary change of curvature
at midspan. In this example, assuming a hinge length at each support equal to d = 170mm, the change of curvature
necessary at each support is s = 199 x 10-6 mm-1.
w
L = 6m
symm
-30
-20 first yield at support
-10
Moment 0
(kNm/m) +10
+20
+30
first yield at midspan
The moment-curvature relationship for the critical sections is plotted in Fig 5. If Class L steel is used the plot would
terminate at a curvature of 95.5 x 10-6 mm-1 (when the strain in the tensile steel reaches su = 0.015). If Class N steel is
used (with su = 0.05) the plot would terminate at a curvature of 308 x 10-6 mm-1. When the uniform elongation of the
steel is greater than about 0.07, the ultimate curvature is controlled by failure of the compressive concrete and equals
391 x 10-6 mm-1.
600
Concrete Institute of Australia (2001) Conference Proceedings
30
20 s
10 su = 0.015 su = 0.05
0
17.8 95.5 216 308 391
Curvature (x10-6 mm-1)
Figure 5 Moment -curvature relationship for critical sections of fixed ended slab.
It is evident from Fig 5, that when using Class L steels, the rotation at the supports necessary to develop the yield
moment at midspan cannot occur. Fracture of the steel at the supports will initiate a sudden failure of the span, probably
when cracking at midspan occurs. When Class N steel is used sufficient rotational capacity is available to develop Msy
at midspan, but insufficient rotational capacity is available to develop the ultimate moment Mu at midspan. The use of
Class L steel in this, or for that matter any, reinforced concrete structure is extremely unwise.
3.2 Example 2
Consider the two-span, one-way slab of Fig 6a subjected to a uniform load and supported by girders at A, B and C, as
shown. If an elastic analysis (based on gross section dimensions) is undertaken assuming rigid supports (ie. girders A,
B and C do not deflect), the bending moment diagram for the slab is shown as the solid line in Fig 6b, where MB = -
wL2/8. The slab load resisted by the girder at B is 3.33 times that resisted by the girders at A (and C). If the deflection
of the girder at B is vB greater than that at A and C, then the negative moment at B decreases by 3EIvB /L2 (where EI is
the flexural rigidity of the slab) and the bending moment diagram for the slab is shown as the dashed curve in Fig 6b.
Depending on the stiffness of the slab and the stiffness of the supporting girders, the value of the moment at B could
well be positive and the positive span moment of the slab may exceed wL2/8. For example, if L = 3000 mm, w = 20
kN/m2, and the slab is 150 mm thick (with E = 28000 MPa), the moment at B assuming vB = 0 (and assuming gross
cross-sectional properties) is MB = -wL2/8 = -22.5 kNm/m. If vB = 10 mm, then the change in moment at B resulting
from differential support deflection is 3EIvB /L2 = 26.25 kNm/m. The moment at B is therefore -22.5 + 26.25 =
+3.75kNm/m.
A B C
RA RB RC
L L
(a) One-way slab and beam system.
MB
Rigid supports
Flexible supports
If the slab cross-sections are ductile (and contain tensile steel with a su in excess of 0.1), it is quite reasonable to ignore
the differential deflection of the supporting girders and to design the slab for the solid bending moment diagram of Fig
6a. This is often what is done in practice. When the cross-section is non-ductile, such an approach is not safe. For the
example considered above, an enormous redistribution of moments is necessary if the solid elastic bending moment is
601
Concrete Institute of Australia (2001) Conference Proceedings
used for the design of the slab. It may not even be safely possible to adopt such an approach when using class N steel
with a uniform elongation as low as 0.05. The days of designers blissfully designing structures to satisfy equilibrium,
and relying on ductility, may be at an end, and should be at an end, if brittle 500 MPa steel reinforcement is used.
560
Concrete Institute of Australia (2001) Conference Proceedings
5. Conclusions
The relatively low uniform elongation of 500 MPa steel reinforcement (particularly Class L reinforcement) has
significant implications in the design of reinforced concrete structures, particularly with regard to ductility. The
necessary ability of structures to redistributed moments as the ultimate load is approached may be significantly
compromised; plastic design techniques may no longer apply; and the failure mechanisms of elements may change
significantly. It has been shown that for the design of flexural members, it is unwise to use Class L reinforcement in
any circumstances. Structures reinforced with Class L steel may not even have the ductility to justify the use of elastic
analysis.
Some of these design implications have been addressed in the recent revision of AS3600 (with varying degrees of
success), and the changes relating to the impact of 500 MPa steel have been discussed.
The introduction of 500MPa steel, with the low minimum uniform elongations currently specified, has been forced on
the industry and is a backward step. If the minimum uniform elongation of Class L and Class N steels were 0.03 and
0.10, respectively, (rather than 0.015 and 0.05), then the concerns related to ductility would disappear and the latest
revision of AS3600 with regard to ductility would be appropriate.
The disadvantages associated with the move to 500 MPa steel are many and far outweigh the benefits currently being
promoted by the steel reinforcement industry.
6. References
1. AS3600-2001, Australian Standard for Concrete Structures, Standards Australia, Sydney, (AS3600-1994, incorporating
Amendment 1 (issued in 1996) and Amendment 2 (issued for public comment in 1999)(2001).
2. AS/NZS4671 - 2001, Australian/New Zealand Standard - Steel Reinforcing Materials, Standards Australia and Standards
New Zealand, 2001.
561