Revision Eurocode 8
Revision Eurocode 8
Revision Eurocode 8
P. Bisch1
Chairman of CEN/TC250/SC8
1
EGIS Industries, Montreuil, France, [email protected]
Abstract: The objective of this article is to present the main developments of the new generation of Eurocode
8, in terms of evolution of concepts and new subjects covered. The general programme for the 2nd generation
of Eurocodes and the positioning of Eurocode 8 in this program are presented, Part of this program was to
improve consistency between all parts of Eurocodes, an important subject for Eurocode 8, which refers to
many of them. The most important thing in this respect was to align the definitions with those of EN1990. It
was necessary to reorganise the six standards constituting Eurocode 8 to implement a more compact structure
avoiding repetitions. The evolutions of the main concepts (e.g., ductility classes) used are presented, and
some of the new subjects.
1. Introduction
Eurocode 8 (EN1998), dedicated to the design of civil engineering structures in seismic regions, has been
used in Europe since 2004. It was preceded by a pre-standard intended for experimentation, but which did not
have the character of a European standard. It belongs to a series of 58 standards brought together in
Eurocodes, constituting the first generation of European standards for the design of structures.
The Eurocodes were developed within CEN, at the instigation of the European Commission, to provide
common standards for the design of works within the European Free Trade Association, for the proper
functioning of the market.
They aim to provide rules allowing compliance with the structure resistance directives, including fire resistance.
Their final objective is therefore the safety of European citizens, which translates into a probability of failure,
in the socio-economic context of Europe. In the end, these are rules applied by an industry, Construction, so
the rules must be practical and economical. CEN procedures guarantee the participation of all Member States
who so wish through the various Technical Committees.
For Eurocodes (EC) as for all European standards, a periodic survey process allows, for standards in use, to
collect comments resulting from their use. This is the Systematic Review procedure. Furthermore, regarding
the Eurocodes, a need was felt to complete the topics covered to better standardise the types of construction
actually used in Europe. These findings therefore led to a proposal to the Commission for a development
program to move from the first to the second generation of Eurocodes. In addition, for Eurocode 8, it was
necessary to modernise the rules used to take into account the rapid development of knowledge in the field of
Earthquake Engineering.
WCEE2024 Philippe Bisch
3. General guidelines
In its Mandate, the Commission laid down essential rules to be followed by all Eurocode writers:
• The reduction of national parameters (NDP): the NDPs are the choices concerning safety, for example
the partial factors, which is legitimate since this remains the responsibility of the Members; these can,
under certain conditions, be verification procedures, when they are equivalent in terms of compliance
with the principles. But to ultimately achieve total convergence of rules, the Commission has asked to
reduce the number of NDPs, by concentrating them on safety choices, and therefore avoiding “technical”
NDPs.
• The “ease of use” of the Eurocodes, affecting different aspects of use of the texts, from the formulation
of the rules to the circulation between standards. This last aspect is not easy to resolve since the set of
standards covered by the Eurocodes is 58 in the first generation.
Reducing NDP implies that a consensus can emerge on technical choices, to be able to reduce them to safety
choices. At the same time, it was decided not to develop alternative methods with the same field of application.
But we can develop, for example, a general method and a simplified method having a more restricted field of
application. At the current stage, SC8 has achieved the required objective, since, between the two generations,
the number of NDPs has been approximately halved, despite the increase in scope.
2
WCEE2024 Philippe Bisch
It is true that the precise definition by TC250 of the verbs used in the requirements made it possible to clarify
their interpretation and to clearly identify the possibilities for national flexibility:
• “shall” corresponds to a strict obligation, a “principle” in the first generation. It must be respected in all
situations.
• “should “indicates a strong recommendation, which is an obligation in most cases, but which may
exceptionally be subject to deviation. Indeed, the Eurocodes cannot cover all situations, despite their
broad scope. In addition, it may happen that in a country a particular situation, for example the
production of a construction product, does not allow full compliance with a requirement. Such an
alternative is in principle given in the NA, but it can also be decided in a project, for example for a special
structure. In all cases, reasoned justification is required for not applying the rule given by this verb.
• “may “describe a rule that can be used to comply with the Eurocode but is not exclusive. This is generally
a possible rule for applying stricter requirements.
No other wording such as “it is recommended to” is permitted. In Eurocode 8, a lot of work has been done to
identify the three levels of choice and many abusive “shall” have been eliminated. Most rules are indicated by
"should" but whenever it seemed possible without betraying the concepts of the code, permission was used to
provide a "standard" solution to the designer.
The risk associated with seismic situations is all the higher as the hazard is stronger. This is why it is not
necessary to comply with Eurocode 8 in areas of very low seismicity. In addition, it is allowed to develop
simplified rules for certain types of structures in areas of low seismicity, provided that they respect the
principles; it is up to the NA to decline them. This gives great latitude for adaptation in large areas of Europe.
Regarding ease of use, it is clear that this could not concern the overall volume of texts, given the increase in
scope, even if a reduction effort could be made at the level of each individual subject. In fact, this objective
was to be achieved through several means:
• The formulation of the rules, already clarified by the strict use of verbs.
• Consistency of rules between Eurocodes. We must be able to move from one to the other without
contradiction or discontinuity. This obligation was particularly important for Eurocode 8, the seismic
situation being by nature particular and requiring physical behaviour different from current situations,
therefore requiring specific verifications.
• Circulation between the different parts of the same Eurocode, but also between Eurocodes, in particular
through the clarity of cross-references.
• The elimination of alternative procedures.
3
WCEE2024 Philippe Bisch
4
WCEE2024 Philippe Bisch
But it turns out that Eurocode 8 includes from its origin a part 3 devoted to the assessment and retrofitting of
existing buildings. Given the specificity of the seismic situation, the concepts on which this part is based may
be different from those recommended in EN1990-2. It was therefore necessary to ensure that bridges existed
between the two approaches so that the designer could carry out all the necessary verifications in a seismic
zone, for seismic or non-seismic design situations. However, there is room for progress to improve continuity
between the two standards on this subject, which can be done in the future.
5. The reorganisation
Although there was a solid foundation developed in the pre-standard, the first generation was developed with
the desire to provide an effective solution for engineers for each type of structure. But there was no general
framework fixed in advance in terms of organisation of the texts. This resulted in redundancies, or even (rarely)
provisions that could appear contradictory. However, one of the rules for developing the 2nd generation of
Eurocodes is to avoid repetition. From the start of the work, it was therefore necessary to rethink the
organisation of the six standards constituting Eurocode 8. In particular, the definition of the seismic action and
the general rules of design and analysis are the same for all types of structures.
EN1998-1-1 EN1998-5
GENERAL
Seismic action Geotechnics
RULES
& general
rules
DESIGN OF
NEW EN1998-1-2 EN1998-2 EN1998-4&6
STRUCTURES Buildings Bridges Other
structures
ASSESSMENT
OF EXISTING EN1998-3
STRUCTURES Buildings & bridges
It was therefore decided to cut the old part 1 which covered seismic action and the rules for buildings into two
parts and to keep all the general rules in part 1-1. The organisation put in place therefore has three levels
(Figure 1):
• Parts 1-1 (general rules) and 5 (geotechnics) which are called by the other parts constitute the first level.
• The second level covers the different types of new structures.
• The third level covers existing structures and refers to the other levels.
Furthermore, the previous parts 4 and 6 found themselves deprived of all the general rules which were
duplicated there, and it was no longer justified by their volume to make them separate parts, in particular the
old part 6 covering the masts and chimneys. So, all these types of constructions were grouped together in part
4. Finally, the number of standards constituting Eurocode 8 remained at 6.
5
WCEE2024 Philippe Bisch
TC250 decided to give, whenever possible, the same structure to all material Eurocode parts. This alignment
effort was made in Eurocode 8, in particular for the first chapters, whose titles are the same as for the other
Eurocodes. The first are “standard”: foreword, introductions, scope, normative references, definitions, and list
of symbols. Chapter 4 gives the Basis of Design. The following chapters deviate from the standard structure,
considering the specificity of the contents in Eurocode 8 parts.
Chapter 5 deals with seismic action and its representation.
Chapter 6 contains all the general rules for modelling, analysis, and verification. In particular, it deals with anti-
seismic devices covered by the EN15129 standard, no longer only the total isolation described in the first
generation.
Finally, chapter 7 was introduced to give the elastic and ultimate deformation criteria allowing calculations to
be carried out in displacement-based analysis for the usual configurations.
Several normative or informative annexes supplement the main text.
6
WCEE2024 Philippe Bisch
remains maintained, which makes it possible to avoid distortions of competition by inhomogeneous safety
choice biases.
A completely elastic design remains possible, as the use of ductility may be problematic or superfluous for
certain types of structures. In this case, the force method and associated verifications must be used, the action
effects being calculated with the elastic spectrum. Additional conditions are given in the other parts, depending
on the type of structure and material.
As in the first generation, a distinction is made between the primary elements, which participate in bracing,
and the secondary elements, which only have a (vertical) loading function. Most design rules concern primary
elements, but secondary elements must be verified to continue to perform their loading function.
For new structures, verification of the SD LS is mandatory, NC resistance being implicit and ensured by the
code. The behavioural rule for the design is to ensure the ductility of the primary structure, in particular by
avoiding brittle fractures and unstable mechanisms. This is necessary to cover exceedances of the design
seismic action level while avoiding “cliff effects”. This is done by capacity design rules, which are detailed in
the different parts. To ensure this ductile behaviour, three ductility levels are proposed, and the structures must
belong to one of these classes (except purely elastic design).
8. Seismic action
Mandate M/515 scope for Eurocode 8 includes a component for the seismic action: it was requested to
standardise seismic zoning in Europe. A first task was to make a description of the zones by homogenising
the names and the intervals covered by the seismic areas, parameterised by the spectral acceleration at the
plateau. A second step in this direction is the introduction of hazard maps as informative annexes, on a small
scale, for a return period of 475 years. These maps are only information giving a good idea of the distribution
of earthquakes in Europe, but the definition of precise national maps remains the responsibility of the Members.
Furthermore, for the association between the Limit States and the Consequence Classes, on the one hand,
and the associated seismic levels, on the other hand, Members were given the possibility of choosing return
periods of the design earthquake, or performance factors, the two choices being presented in the form of
tables. This alternative had already been used by certain Member States, although it was not described in the
1st generation.
For the description of seismic action, the representation by elastic spectra is preferred, but series of equivalent
accelerograms can also be used, in particular to make non-linear time history response analysis.
The elastic spectra of the first generation were calibrated in relation to a single parameter, the ground
acceleration ag, but this quantity is not really representative for the analysis of structures, none being infinitely
stiff. This is why two parameters were introduced instead: the acceleration on the plateau of the spectrum,
useful for stiff structures, and that at 1 s, for flexible structures. This makes it possible to better refine the
spectra and the definition has sufficient flexibility to adapt to a particular seismicity. In addition, to easily allow
analysis by a displacement-based method, the elastic spectra are also given in displacements.
Finally, the formulation of the damping correction factor has been modified to allow a better approximation
when using dissipative systems.
A more precise description of the seismic action is the subject of another presentation.
9. Ductility classes
The definition of ductility classes and their number have since the beginning been a subject of debate within
SC8. It was considered during the conversion to improve the ductility L in order to keep only two classes by
removing the H. Finally, SC8 decided to keep three classes. Several arguments led to a fairly thorough review
of the way classes are defined and constructed:
• DC H is often too complicated and ultimately little used, especially in concrete, so there is no point in
keeping it as is.
• There is no real conceptual difference between M and H; the rules are simpler and less restrictive in M
because the ductility demand is lower.
7
WCEE2024 Philippe Bisch
• The clauses were developed without a very clear relationship with the level of seismicity, except the
certainty of properly covering the high seismicity. However, most of the European territory has low or
moderate seismicity where simpler rules may suffice.
• The 1st generation recommends using ductility L only in low seismicity, but on the one hand this limit is
an NDP and, on the other hand, certain countries with areas of high seismicity have not limited its use.
Furthermore, the term “recommends” can no longer be used.
Therefore, the ductility classes have been redefined such that the two ductile classes are better adapted to
the level of seismicity:
• DC1 is an extension of the DCL, where the limitations (by the seismic action index) depend on the type
of structure and the material. The behaviour is quasi-elastic, only the overstrength is considered. Control
of drifts between floors is nevertheless required. With this extension of the limits, a large number of
structures can be built in DC1 in moderate seismicity areas.
• DC2 is built to cover most common structures, in moderate areas. It is not the DCM and must therefore
be considered new. The local overstrength (in deformation) is imposed by design criteria of capacity
and plastic deformation. The overall behaviour is controlled by limiting drifts and 2nd order effects, but
no criterion such as "weak beams - strong columns" is imposed. The wording of the rules is simplified
as much as possible. On the other hand, the use of DC2 is limited for certain types of structures.
• DC3 adds to the DC2 the control of the global plastic mechanism, with the strong column criterion. The
rules are between DCM and DCH, depending on the case. There are no limitations on the use of DC3.
All application conditions are given in the different parts depending on the type of bracing and the material.
8
WCEE2024 Philippe Bisch
deformation d considered in the plastic domain. This coefficient is taken equal to 0,5, unless otherwise provided
for certain types of structure. The SD state therefore corresponds locally to the middle of the plastic domain,
which is judged to comply with the definition given above.
Global criteria on the pushover curve, therefore in displacement, may also be used, particularly for masonry.
On the basis of existing comparative calculations, it was possible to establish correlations between the two
methods making it possible to evaluate the behaviour factors and to compare the two methods, in order to
best ensure their equivalence (Figure 2).
V
Ve = qS qR qDVd = q Vd
V2 = qS qRVd
DL SD NC
V1 = qSVd
OP
V d = Fb Progressive yielding phasis
Quasi-elastic phasis
dd d1 dy d2 du dtop
du = q dd = qD d2 = qD qR d1
Figure 2. Comparison between calculation methods (forces and displacements) and positioning of LS
To allow the pushover analysis, it is necessary to have rules making it possible to determine the local
deformations at the elastic limit and ultimate. These rules do not depend on the type of structure, only on the
material and the structural element. This is why they are considered general rules; they are the subject of
Chapter 7, which covers concrete, steel and composite structures and timber. The rules for masonry are given
in parts 1-2 and 3, masonry being developed only for buildings.
9
WCEE2024 Philippe Bisch
• Partial factors for resistance gRd determined on a unified basis from a single fractile of the resistance
distribution (NDP).
• Spatial model of seismic action.
• Structures equipped with more general anti-seismic devices than isolators.
• Systems with energy dissipation.
• Soil-structure interaction.
• Auxiliary elements and floor spectra.
• Flat slabs (buildings).
• Frames with filling and claddings (buildings).
• Aluminium structures in part 1-2 (buildings).
• Bridges in Part 3 (existing structures).
• Timber structures in parts 2 (bridges) and 3 (assessment of existing structures).
10