BS8110
BS8110
Synopsis
The two codes are compared in the context of design of primary structural
elements and information is given on the availability of design aids to assist
the practitioner in becoming familiar with and using the new code.
Keywords
Introduction
The ENV version of Eurocode 2 Part 1 General Rules and Rules for Buildings
has been around for some years as a draft for development. A National
Application Document (NAD) was prepared to be used in conjunction with the
ENV and together with the main document was published by BSI in 19921.
The intention was that this document would be trialled on real structures and it
is referred to in Approved Document A of the Building Regulations. However
use of the existing ENV is believed to have been very limited.
The EN versions of Part 1 and 1.2 of the code dealing with fire have been
through several draft revisions and are now being finalised1. The target date
for publication of these Parts of the code is Summer 2003.
There will be National Annexes to accompany each part of the code, which
will include values for what are called Nationally Determined Parameters. In a
similar way to the NAD the intention is that the code will be used with the
appropriate National Annex in each member state.
Grades of concrete
EC2 allows benefits to be derived from using high strength concretes, which
BS8110 does not. Concrete strengths are referred to by cylinder strengths,
which are typically 10-20% less than the corresponding cube strengths. The
maximum characteristic cylinder strength fck permitted is 90N/mm2, which
corresponds to a characteristic cube strength of 105N/mm2.
Part 1 of EC2 specifically does not cover this and a separate standard (termed
an Execution Standard) has been prepared. This is currently in ENV form and
a national document based on the existing National Structural Concrete
Specification2 is in preparation and is expected to be available towards the
end of 2003.
One issue, which does need to be considered as part of the main code, is the
tolerance on cover. Cover to meet durability and bond requirements is
specified as a minimum value with a tolerance of up to 10mm to be added on
top. This is in contrast to BS8110 where cover is specified as a nominal value
and a tolerance of 5mm accepted. In situations where good quality control is
exercised there is scope for reducing the tolerance.
This paper does not consider this topic, and in making comparisons between
BS8110 and EC2 assumes that covers and dimensions of members are
largely unaffected by the changed design process.
1
The full references are EN1992-1-1 Eurocode 2: Design of concrete structures - Part 1:
General rules and rules for buildings and EN 1992-1-2 Eurocode 2: Design of concrete
structures Part 1.2: General rules – structural fire design
In Part 2 of EC2 there is a prescriptive method encompassing simplified
approaches based on covers and member dimensions, which is broadly
similar to the approach taken in BS8110. There are however also more
sophisticated performance based methods that can be used and the
information is much more extensive than in Part 2 of BS8110.
In making comparisons between BS8110 and EC2 this paper assumes that
covers are largely unaffected by the changed design process. Simplified
guidance to enable engineers to specify appropriate grades of concrete for
particular exposure conditions with appropriate covers is in course of
preparation and will be included within the National Annex to the code.
The concept of an explicitly defined design life and the recognition of the need
to take additional measures if this design life is required to be significantly
exceeded must be seen as a positive step forward.
As with BS8110 EC2 uses a basic material partial safety factor γm for concrete
of 1.5. Several years ago the material partial safety factor for reinforcing steel
in BS8110 was reduced from 1.15 to 1.05. EC2 uses a value of 1.15 although
this is subject to a National Annex. This is unlikely to have any practical
impact however as steel intended to meet the existing yield strength of
460N/mm2 assumed by BS8110 is likely to be able to meet the 500N/mm2
assumption made by EC2, so that the design yield strength fyd will be virtually
identical.
In due course these will be given by EC1. The comparisons made in this
paper in general consider only the resistance side of the equation although
some mention is made of the partial load factors to be used. It is worth noting
that a value of 25kN/m3 is taken for the density of normal weight concrete as
opposed to the currently assumed value of 23.6 kN/m3.
The combined impact of the partial load factors in conjunction with values for
basic design loads and other items such as column load reduction factors and
the assessment of slenderness in column members has been considered as
part of a separate small-scale study in relation to a whole building design. The
building studied was a typical RC framed flat slab structure.
The conclusions from this study were that, at least for the particular building
studied, the overall impact of using EC2 instead of BS8110 was minimal.
Several options are given for the type of stress-strain relationship that may be
assumed for concrete. In many cases the designer is likely to opt for the
simple rectangular stress block.
The stress block used in EC2 is compared with that in BS8110 below.
The following basic equations may be derived for the design of elements in
flexure.
M α cc xmaxηλ λx
K= K '= d − max
bd 2 f ck d γc
2
2
d γc
z= 1 + 1 − 2 (min K , K ') ≤ 0.95d
2 ηα cc
M 2 = bd 2 f ck (K − K ') ≥ 0
M2
As2 =
f sc (d − d 2 )
M −M2 f
As = + As 2 sc
f yd z f yd
In these equations xmax is the maximum neutral axis depth permissible before
compression steel is to be provided. This in turn depends on the amount of
redistribution assumed. The effect of redistribution is dependent on the
concrete strength with one set of values up to and including C50 and a
differing set of values for higher concrete strengths. The values are subject to
a National Annex, and the UK recommended values for strengths up to C50
lead to the following equation:
xmax = (δ - 0.4)d
where for example δ =1.0 means no redistribution and δ = 0.8 means 20%
redistribution. This is basically the same equation as in BS8110. It may also
be considered advisable to set some upper limit on xmax regardless of the
amount of negative redistribution (i.e. redistributed M being greater than
elastic M).
Span/depth ratios
In both BS8110 and EC2 the allowable span/depth ratio depends on concrete
strength and tension and compression reinforcement ratios. The attached
flowcharts show how the permissible span/depth ratio is arrived at in each
case.
The study showed that EC2 tended to be more conservative at low concrete
strengths. However EC2 permits much higher span/depth ratios for cantilevers
where a low reinforcement percentage is used, even restricting the maximum
enhancement in steel area. In practice however, economic rather than
minimum permissible depths will generally be used, and these are very similar
in both codes.
Shear
When checking normal shear, EC2 is the same as BS8110 in that there is a
shear stress below which only minimum shear reinforcement need be
provided. In EC2 as in BS8110 this shear stress depends on concrete
strength, effective depth and tension steel ratio.
The recommended design shear stress of the concrete alone for comparison
with the values of vc given in Table 3.8 of BS8110 is:
0.18
k (100 ρ l f ck )
1 3
v Rd ,c = 3 ≥ 0.035k 2
f ck
γc
Where k = 1 + √(200/d) ≤ 2
ρl = As /(bd) ≤ 0.02
The value 0.18/γc and the expression for the minimum concrete shear stress
are subject to the National Annex.
Choosing a value of 0.12 for 0.18/γc in the above equation, and the expression
for concrete shear stress as given above, BS8110 generally allows a higher
shear stress before shear steel is required. Because of the minimum shear
stress that can be carried, EC2 can however allow higher shear stresses for
low reinforcement percentages and this effect is accentuated the higher the
strength of the concrete.
EC2 differs from BS8110 in that above the limit at which the concrete alone
has sufficient capacity, the designed shear steel to be provided is determined
ignoring the contribution from the concrete. The design method used is known
as the variable strut inclination method and is based on a truss model.
For members not subjected to axial forces the required area of shear steel
needing to be provided in the form of links at a distance d from the support
face is given by:
The designer should choose an appropriate angle θ (the angle between the
assumed concrete compression strut and the main tension chord) to use in
the model. The limits on θ are between 22 and 45 degrees such that cot θ is
greater than or equal to1 but less than or equal to 2.5.
0.45 ν fcd
where ν = 0.6(1-fck/250)
fcd = αccfck/γc
ν = 0.6 up to C60
ν = 0.9 - fck/200 > 0.5 for grades above C60
The two codes have been compared choosing values of αcc = 0.85 and γc =1.5
and ignoring the increase allowable for ν if the stress in the shear steel is
restricted. EC2 will allow a smaller maximum shear capacity at low strengths
but a higher capacity at higher strengths principally arising from the cut off of 5
N/mm2 in BS8110.
The increase in the allowable shear stress becomes quite significant when
increased values of ν are permitted even ignoring the cut-off in BS8110 as
illustrated in Figure 1.
Figure 1: Comparison of maximum permissible shear stresses
14
12
10
Shear stress (N/mm )
2
8
EC2
EC2 shear steel stress restricted
BS8110 (40N/mm2 limit removed)
6
0
0 10 20 30 40 50 60 70 80 90 100
fck
For a given required shear capacity the amount of shear steel to be provided
when designing to EC2 is dependent on cot θ which should be maximised as
stated above. In practice the following inequality needs to be satisfied:
ω + ω2 −4
1 ≤ cot θ = ≤ 2 .5
2
0.9b w dνf cd
where ω = cot θ + tan θ =
V Ed
Indirectly the concrete strength can therefore influence the amount of shear
steel provided if cot θ needs to be less than 2.5 to satisfy the criterion on
maximum shear capacity.
The two codes can in general be expected to give similar results in terms of
the number and spacing of links to be provided.
40
35
30
N/bh (N/mm )
2
25 BS8110 4T32
20
15 EC2 4T32
10
5
0
0 2 4 6 8
2 2
M/bh (N/mm )
.
Dealing with slenderness
In EC2 the allowable slenderness ratio λ is calculated from l0/i where i is the
radius of gyration of the uncracked cross section. For a rectangular section
ignoring the reinforcement this simplifies to λ =3.464 l0/h where l0 is the
effective length. The slenderness should be checked in both directions.
Where the column is slender when designing to EC2 and using the nominal
curvature method which it is probably the most straightforward, the final
design moment is increased by the additional moment to account for second
order effects. Once this adjustment has been made the N-M interaction charts
may be used as before. The same approach is used for BS8110 except that
the second order moments will be calculated differently.
Biaxial bending
EC2 states that a separate design may initially be carried out in each principal
direction. Imperfections need be taken into account only in the direction where
they will have the most unfavourable effect.
λy / λx ≤ 2 and λx / λy ≤ 2
and (ey/h)/(ex/b) ≤ 0.2 or (ex/b)/(ey/h) ≤ 0.2
ex and ey are the effective total eccentricities including second order effects.
These are beyond the scope of this paper. However it is hoped to include
some guidance on this in a future paper underpinning the provisions within the
National Annex for the code.
Flat slabs
EC2 Part 1 now has an Informative Annex dealing with flat slabs which was
noticeably absent from the ENV version. The widths of column and middle
strips are the same as in BS8110. The percentages of moments carried by
these strips are given as ranges but the BS8110 values fall within these
ranges and hence may still be used.
The other major issue when designing flat slabs is dealing with punching
shear. The code provisions in EC2 dealing with this topic have recently been
revised and it is believed worthwhile to revisit the implications of these. Initial
indications are that EC2 is marginally more economic, mainly because the link
arrangements are more efficient. Detailing of links should also be easier.
The complete set of possible load combinations and load cases is obtained
from EN1990 Basis of Structural Design. In practice these can be simplified
greatly for the design of everyday building structures.
For slabs the UK National Annex is currently permitting the all spans loaded
condition to be considered sufficient subject to the restrictions as currently
imposed in BS8110.
A major difference between the two codes is the partial safety factor
appropriate to the dead load for unloaded spans.
EC2 BS 8110
Loaded spans: γG = 1.35, γQ = 1.5 γG = 1.4, γQ = 1.6
Unloaded spans: γG = 1.35 γG = 1.0
Strictly speaking the above table relates only to the design of loaded spans.
The design of unloaded spans should theoretically be considered separately
taking γG = 1.0 on all spans, but in practice this is very rarely likely to prove
the governing load case.
Detailing issues
It is believed that spacing rules may lead to more and smaller bars, unless
crack widths are checked.
1. The advent of EC2 as for the other Eurocodes will have a big impact on
the design of all types of structures. There will be a learning curve
associated with gaining familiarity and using the new code.
3. In general EC2, used in conjunction with the National Annex, is not wildly
different from BS8110 in terms of the design approach. It gives similar
answers and offers scope for more economic structures.
4. There will be an opportunity for comment on the values proposed for the
Nationally Determined Parameters to be included within the National
Annex, before it is published.
5. Overall EC2 is less prescriptive and its scope is more extensive than
BS8110 for example in permitting higher concrete strengths. In this sense
the new code will permit designs not currently permitted in the UK, and
thus give designers the opportunity to derive benefit from the considerable
advances in concrete technology over recent years. The authors believe
that, after an initial acclimatisation period, EC2 will be generally regarded
as a very good code and a step in the right direction.
Acknowledgements
The authors would like to acknowledge the funding for this work provided by
the ODPM and the BCA. The paper is endorsed by the Concrete Industry
Eurocode 2 Group (CIEG) referred to in Reference 4.
References