Sei sulla pagina 1di 52

19/07/2017

Earthquake-resistant
design
Francisco López Almansa

Objectives

 Earthquake-resistant design of buildings and


bridges

Earthquake-resistant design. Francesc López Almansa. Barcelona 2

1
19/07/2017

Index
Effects of seismic inputs on buildings 4
Effects of seismic inputs on bridges 13
Conceptual seismic design of buildings 14
Estimation of fundamental period 36
Types of seismic analyses 43
Single-mode linear static analysis 45
Multi-mode linear static analysis 55
Nonlinear static analysis 59
Performance-based design 74
Nonlinear dynamic (time-history) analysis 91
Nonstructural components 98
Bibliography 101
Internet Sites 103

Earthquake-resistant design. Francesc López Almansa. Barcelona 3

Effects of seismic inputs on buildings (1)


 Effect of horizontal components is more severe; vertical analysis is only necessary
for long span-length, long cantilever, interrupted columns and prestressed concrete
 xg: ground displacement (input, excitation)
 y: absolute displacement
 x: relative displacement with respect to the base
 Relative displacement between adjoining stories is termed “interstory drift”
 Relative displacement and interstory drift report about structural damage
 Absolute acceleration report about human comfort and non-structural damage
 Relative displacement and absolute acceleration cannot be minimized
simultaneously
 If buildings are designed very stiff, x  0 and : structural damage is
minimized but non-structural damage is not
 If buildings are designed very flexible, x   xg and 0: non-structural damage
is minimized but structural damage is not

x x x x
xg y xg y
Stiff xg
Normal Flexible

xg xg xg
Earthquake-resistant design. Francesc López Almansa. Barcelona 4

2
19/07/2017

Effects of seismic inputs on buildings (2)


 Effects of gravity forces on moment-resisting frames
 Common design criteria:
– Alike beams. Beam sections depend only on load and span-length
– Columns are strongest in lowest floors
– In lowest floors, central columns are strongest
– In highest floors, alike columns in each floor

Def. N V M

Earthquake-resistant design. Francesc López Almansa. Barcelona 5

Effects of seismic inputs on buildings (3)


 Effects of lateral forces on unbraced moment-resisting frames
 Common design criteria:
– Strongest beams and columns in lowest floors
– Alike columns in each floor
 Moment/shear inversion on beams: this effect is higher in bottom
levels, outer joints and inner frames
 Moment inversion precedes shear inversion
 Shear inversion can lead to tensioned columns and even to uplift

Def. N V M

Earthquake-resistant design. Francesc López Almansa. Barcelona 6

3
19/07/2017

Effects of seismic inputs on buildings (4)


 Signs of moments in joints without moment inversion:

 Signs of moments in joints with moment inversion:

Earthquake-resistant design. Francesc López Almansa. Barcelona 7

Effects of seismic inputs on buildings (5)


 Lateral behavior of moment-resisting frames is
“awful” and “un-natural”, with huge interstory drift
 Bending of beams and columns is an un-natural
behavior; axial behavior is much better since
strength and stiffness are higher
 Strategy of avoiding bending, leads to the concept of No bracing
bracing
 For building frames, there are two major types of
braces: diagonal and chevron
 In both cases, interstory drift causes only axial
tension/compression in braces
 Pushing forces are distributed between main frame
and bracing: since braces are much stiffer, take most Diagonal bracing
of the force
 In buckling analysis of a braced building, non-sway
behavior is commonly assumed if interstory drift is
reduced more than 80%
 Diagonal braces perform slightly better than chevron
ones (they are more horizontal), but chevron are
frequently preferred for architectural reasons
Chevron bracing
Earthquake-resistant design. Francesc López Almansa. Barcelona 8

4
19/07/2017

Effects of seismic inputs on buildings (6)


 Gravity loads have no effect on braces (even their own
weight)
 With this aim, chevron braces are connected to the top
beam once most of the load is present
 Separate (?) behavior: frame for gravity loads and
bracing for lateral actions (wind and seismic) No bracing
 Braces are always steel-based, but bracing can be used
both for concrete and steel buildings
 In diagonal bracing, if the compressed brace buckles
(bends), the remaining tensioned one can still resist
 In chevron bracing, if the compressed brace buckles,
the remaining tensioned one can not resist since
additional bending is generated in beam (undesired Diagonal bracing
behavior)

Chevron bracing
Earthquake-resistant design. Francesc López Almansa. Barcelona 9

Effects of seismic inputs on buildings (7)


 In multi-story
buildings, separation
between main frame
and bracings is not
complete, since
braces transmit axial
forces to columns
 Braces in the same
bay: huge over-
compression in a
given column

Earthquake-resistant design. Francesc López Almansa. Barcelona 10

5
19/07/2017

Effects of seismic inputs on buildings (8)


 If braces are distributed
among several bays (mainly
in lower stories), axial over-
compression is distributed
among several columns
 Although this solution is
better than the one in the
previous slide, braces
transmit important shear
forces to columns; this can
generate brittle failure

Earthquake-resistant design. Francesc López Almansa. Barcelona 11

Effects of seismic inputs on buildings (9)


 In concrete buildings, structural walls (also termed “shear
walls”) can play a similar role than braces, although
without the localized shear forces transmitted to columns Shear
 “Shear buildings”: their lateral behavior is shear-like behavior
 “Shear-wall buildings”: their lateral behavior is moment-
like
 Both behaviors are quantitatively and qualitatively
different
 Walls are better tan frames, but openings are necessary:
intermediate solutions are preferred
Flexural
behavior

Earthquake-resistant design. Francesc López Almansa. Barcelona 12

6
19/07/2017

Effects of seismic inputs on bridges (1)


 In seismic areas, commonly bridge deck is
continuous (hyperstatic) and is rigidly connected
to the supporting members (piles and abutments)
 Deck big inertia forces are transmitted to piles
(piers) and abutments
 Broadly speaking, bridges are equivalent to
single-story buildings (piles and abutments:
columns; deck: slab; no infill walls)
 Spatial variation of seismic input is relevant in
long bridges
 Vertical input is not more relevant than in
buildings (except in long-span bridges)
 Strong deck / weak pile
 Pounding between adjoining segments of deck is
relevant

Earthquake-resistant design. Francesc López Almansa. Barcelona 13

Conceptual seismic design of buildings (1)


1. Lateral resistance and rigidity
2. Plan (mechanical) symmetry
3. Regularity (mechanical and geometrical) along height
4. Compact plan configuration
5. Lightweight (mainly in top stories)
6. Torsion strength and stiffness
7. In-plane rigidity of slabs (rigid diaphragm effect)
8. Ductility
9. Damping
10. Structural redundancy
11. Strong column-weak beam
12. No “short columns”
13. Subjection and strengthening of non-structural elements
14. Tying of footings and pile caps
15. No long cantilevers, no interrupted columns
16. Simple structural behavior
17. Separation to adjoining buildings
Earthquake-resistant design. Francesc López Almansa. Barcelona 14

7
19/07/2017

Conceptual seismic design of buildings. Lateral resistance (2)


 Buildings should be provided with bidirectional lateral strength and rigidity (i.e., in
two directions); strength is useful to resist the seismic forces, and rigidity ensures
that there are no excessive relative displacements, thus preventing P- effects and
collision with adjoining buildings
 In framed structures, lateral strength and rigidity are provided both by columns and
beams; however, weak columns cannot be compensated with over-resistant beams
 In framed structures, lateral strength and rigidity rely on the rigidity of the
connections; therefore, the seismic behavior of precast concrete buildings is doubtful
 Braces or structural walls are preferred to frames, but are less ductile

Flexural
behavior

Earthquake-resistant design. Francesc López Almansa. Barcelona 15

Conceptual seismic design of buildings. Plan symmetry (3)


Non-structural
cladding masonry
wall ●R ●G

R●G

Asymmetric!

 Symmetric / asymmetric buildings: the centers of mass G and rigidity R are


approximately coincident / clearly eccentric in each floor
 Symmetric buildings provide better performance
 Centers of mass G of each story refer to the supported (above) weight
Earthquake-resistant design. Francesc López Almansa. Barcelona 16

8
19/07/2017

Conceptual seismic design of buildings. Uniformity along height (4)

 Vertical
regularity
ensures that
damage is
uniformly
distributed
along height
 Regular does
not mean
alike!

Earthquake-resistant design. Francesc López Almansa. Barcelona 17

Conceptual seismic design of buildings. Uniformity along height (5)


 Non-uniform buildings

Weak
first floor

Earthquake-resistant design. Francesc López Almansa. Barcelona 18

9
19/07/2017

Conceptual seismic design of buildings. Uniformity along height (6)


 Damage concentrated in one story

Earthquake-resistant design. Francesc López Almansa. Barcelona 19

Conceptual seismic design of buildings. Compact plan configuration (7)


 Plan configuration close to square is preferred
 Longer plan size should not exceed shorter one by more
than about four times
 Seismic joints can be used; are similar to expansion joints
but should accept wider motions in two directions
 Can be used for bridges and for buildings

Earthquake-resistant design. Francesc López Almansa. Barcelona 20

10
19/07/2017

Conceptual seismic design of buildings. Lightweight (8)

 Since seismic forces are NO


proportional to mass,
lightweight buildings are less
affected by earthquakes
 Conclusions:
 Timber better than any other
material
 Steel better than concrete
 Prefabricated concrete awful
 Heavy masses are more
dangerous in top part of the
buildings:
 No roof swimming pools!
(very romantic but hazardous) YES
 No “big-top” buildings

Earthquake-resistant design. Francesc López Almansa. Barcelona 21

Conceptual seismic design of buildings. Torsion strength and stiffness (9)

 Even symmetric buildings can experience torsion (twist) motion,


because of both accidental eccentricities and torsion excitation 
 Therefore, stiffening elements (braces or structural walls) should
be located as separated as possible 
 Centrally-located staircases (typical of tall buildings) do not
provide usually enough torsion strength and stiffness 






Earthquake-resistant design. Francesc López Almansa. Barcelona 22

11
19/07/2017

Conceptual seismic design of buildings. Rigid diaphragm (10)


 Slabs and roofs should be infinitely rigid in its own plane; this guarantees joint
cooperation of all vertical resisting members (columns, braces, walls) and
provides a simple and regular structural behavior
 Light steel roofs do not posses this quality unless they contain horizontal trusses
(in-plane bracing)
 Any building slab with a reasonably compact plan configuration behaves as a
rigid diaphragm
 In-plane rigidity of timber slabs is controversial
 ASCE 7-10 (12.10.1 Diaphragm Design) proposes design forces for diaphragm
design

Earthquake-resistant design. Francesc López Almansa. Barcelona 23

Conceptual seismic design of buildings. Ductility (11)


 Ductility (resilience, tenacity) is the capacity of further resistance after the onset
of damage (end of linear behavior)
 Ductility can be measured in terms of force, displacement or product of both
(energy)
 Since earthquakes are indirect actions (imposed displacements), force ductility is
only of little interest
 Ductility can be defined at sectional, member or structural level
 Since earthquakes are highly unpredictable, ductility is a convenient quality

Brittle

Ductile
Brittle Ductile
Earthquake-resistant design. Francesc López Almansa. Barcelona 24

12
19/07/2017

Conceptual seismic design of buildings. Damping (12)


 Damping is always beneficial
 Damping is spread along the building; therefore, it is difficult to provide
damping
 A convenient strategy is to install energy dissipators; this will discussed in the
corresponding part of the course
 In base isolated buildings, it is easier to provide damping

Earthquake-resistant design. Francesc López Almansa. Barcelona 25

Conceptual seismic design of buildings. Structural redundancy (13)

 Statically redundant structures with high degree of


hyperstaticity are preferred
 Redundant members provide additional safety

Earthquake-resistant design. Francesc López Almansa. Barcelona 26

13
19/07/2017

Conceptual seismic design of buildings. Strong column-weak beam (14)


 In rigid beam-column connections, failure of beams should precede (by a given
factor) failure of columns
 In other words, beams protect columns as “structural fuses”
 This condition is based on the assumption that columns are more crucial, to
structural integrity, than beams
 Conclusion: if beams are strengthened, columns should be also strengthened by,
at least, the same factor

Earthquake-resistant design. Francesc López Almansa. Barcelona 27

Conceptual seismic design of buildings. No “short columns” (15)

 Short columns have less capacity to absorb interstory drift than normal ones

Earthquake-resistant design. Francesc López Almansa. Barcelona 28

14
19/07/2017

Conceptual seismic design of buildings. Non-structural elements (16)

 Non-structural elements (appliances, appendages, antennae, etc.) should be rigidly


connected to the main structure as to avoid detaching, falling and overturning
 Infill masonry walls can be either detached from the main structure or fixed to it,
as to provide stiffness and strength; some design codes (e.g. New Zealand)
consider the cooperation of masonry walls

Earthquake-resistant design. Francesc López Almansa. Barcelona 29

Conceptual seismic design of buildings. Tying of foundation (17)


 Because of the spatial variation of the seismic action, footings and pile caps
should be tied together
 Foundation slabs usually fulfill this condition
 Bottom basement pavement can be considered for this purpose
 Usually, seismic design codes indicate the design values of axial forces

xg x’g

Earthquake-resistant design. Francesc López Almansa. Barcelona 30

15
19/07/2017

Conceptual … buildings. No long cantilevers, no interrupted columns (18)

 At risk because of vertical inputs

Earthquake-resistant design. Francesc López Almansa. Barcelona 31

Conceptual seismic design of buildings. Simplicity (19)


 Complex structures are difficult to comprehend
 There can be hidden failure modes
 Commonly, software codes are not helpful

Earthquake-resistant design. Francesc López Almansa. Barcelona 32

16
19/07/2017

Conceptual seismic design of buildings. Separation (20)


 Pounding between colliding buildings
(or between several parts of a building
separated by joints) is dangerous
 Risk is higher is slabs are unaligned
 Solutions: separation (gap) or bumpers
 Separation should be equal to the sum of
both displacements
 Displacements must be calculated from
nonlinear analysis without accounting
for ductility
 Torsion should be considered

Earthquake-resistant design. Francesc López Almansa. Barcelona 33

Conceptual seismic design of buildings. Wrong? (21)


Rem Koolhaas building in Beijing (CCTV Headquarters)

Earthquake-resistant design. Francesc López Almansa. Barcelona 34

17
19/07/2017

Conceptual seismic design of buildings. Exercise (No. 8) (22)


Provide sketches or pictures (better) of actual buildings fulfilling the following
requirements. You can either browse on the Internet, take your own pictures or draw
or use sketches.

1. Buildings with low lateral strength (in one or two directions)


2. Buildings with plan asymmetry
3. Buildings with soft first story
4. Buildings with interrupted columns
5. Adjacent pounding buildings with aligned/unaligned slabs
6. Light steel roofs with and without in-plane bracing
7. Buildings with complex structural system but adequate seismic-resistant
configuration
8. Slabs with dubious rigid diaphragm effect
9. Masonry infill walls separated from the main structure

Earthquake-resistant design. Francesc López Almansa. Barcelona 35

Estimation of fundamental period (1)


 Seismic design depends on many building
characteristics, but mainly on fundamental period
(in any direction), since it influences spectral
ordinates 10
 Therefore, design codes include empirical 9
expressions providing preliminary estimations of 8
the fundamental period 7
 Fundamental period depends on mass and stiffness 6
parameters: (K – 2M) = 0; stiff and light 5
buildings have short periods while flexible and 4
heavy buildings have long periods 3
 Fundamental period depends mostly on building 2
height, being little correlated to horizontal size 1
 The most simplified criterion is TF = N / 10 s (N: z
number of floors)
 This criterion can be only applied to modern and x

regular concrete or steel frame buildings


designed for seismic regions y

 In wall or braced buildings, period is shorter TFx  1 s; TFy < 1 s


 In buildings designed for non-seismic regions,
period is longer
 In tall buildings, this expression yields too long
periods
Earthquake-resistant design. Francesc López Almansa. Barcelona 36

18
19/07/2017

Estimation of fundamental period (2)


 Design spectrum by EC-8

“Single- Low-rise High-rise


Mid-rise
story” buildings buildings
buildings
buildings
Short–span, Long–span,
low–rise high–rise
bridges bridges
Earthquake-resistant design. Francesc López Almansa. Barcelona 37

Estimation of fundamental period (3)


 EC-8. Art. 4.3.3.2.2.(3)
 T1 = Ct H ¾; H: height (m) (H  40 m)
 Ct = 0.085 in steel frames 10
 Ct = 0.075 in concrete frames and eccentrically- 9
braced steel frames 8
 Ct = 0.050 elsewhere; or 7
 Ct = 0.075 / Ac1/2; Ac =  [Ai (0.2 + (lwi / H))2] 6
(concrete or masonry walls) 5
4
 T1 = 2 d 1/2; d is the top floor lateral
3
displacement (m) under “horizontal gravity 2
forces” 1
 This result arises from assuming that the z

building behaves as an equivalent SDOF (1st x


mode) with the period of the first mode and all
the mass of the building: y
TFx = 0.085 (31)¾ = 1.12 s
TFy = 0.050 (31)¾ = 0.66 s
2 2 2 2

Earthquake-resistant design. Francesc López Almansa. Barcelona 38

19
19/07/2017

Estimation of fundamental period (4)


 ASCE 7-10, 12.8.2.1
 Ta = Ct hnx; hn: height (m)
 Ct = 0.0724, x = 0.8 in steel frames 10
 Ct = 0.0466, x = 0.9 in concrete frames 9
 Ct = 0.0731, x = 0.75 in steel eccentrically-braced 8
frames or with buckling-restrained braces 7
 Ct = 0.0488, x = 0.75 elsewhere 6
5
 TF = N / 10 s (N  12, h  3 m) 4
 For masonry or concrete shear wall structures: 3
2
1
z

y
TFx = 0.0724 (31)0.8 = 1.13 s
TFy = 0.0488 (31)0.75 = 0.64 s

Earthquake-resistant design. Francesc López Almansa. Barcelona 39

Estimation of fundamental period (5)


 For estimating the higher mode periods: Ti = TF / (2 i  1) (expression from
NCSE-02)
 Infill walls. Former EC-8 (part 1.3, art. 2.9.4) proposed empirical expressions for
reducing TF because of stiffening effect of non-structural infill walls (provided
they are not detached from the main structure):
T’1 = (T1b + T1i) / 2
 T’1: effective period (to be used for design); T1b/T1i: fundamental period
without/with infill walls
 Two expressions are provided: 0.065

0.080

1
16 0.075
¾

 Aw: area of walls per story, G: shear modulus of walls, H/B: building height/length
(m), W: building weight, n: number of stories
 EC-6 states that G = 0.4 E; E = 1000 fk = 1000 K fb fm (MPa)
 fb and fm are brick and mortar strength, coefficient K ranges between 0.25 and 0.8,
depending on the type of brick units and mortar, for ordinary situations  = 0.7, 
= 0.3
Earthquake-resistant design. Francesc López Almansa. Barcelona 40

20
19/07/2017

Estimation of fundamental period (6)


 Once the building is modelled with the usual software codes, the fundamental
period is obtained from eigenvalue (or Ritz vectors) analysis
 The obtained values should be similar to the preliminary estimations (15% at
most); this can be used for checking modelling errors and for calibrating the
model
 For severe inputs, damage progresses along the input duration and the
fundamental period of the buildings elongates significantly
 This circumstance is commonly taken into consideration when generating the design
spectra
 This is also relevant when evaluating the performance of damaged buildings under
expected aftershocks
Sd(T)

B C
agS0/q

D
agS A

T
TA = 0 TB TC TD

Earthquake-resistant design. Francesc López Almansa. Barcelona 41

Estimation of fundamental period (7)


 Once the building is modelled with a software code, periods can be approximately
obtained by the Rayleigh-Ritz method
 Eigenvalue problem:   . Since modal vectors are orthogonal to
 
mass and stiffness matrices (    0, i  j): ω
 
 ∑ 
; fi is the vector that contains the forces that generate i
  ∑ 
shape deformation (fi = K i)
 This expression can be used to estimate any natural frequency (without knowing
the stiffness matrix), provided that we “guess” the modal shape; for the first mode
(i = 1), we can assume that fi correspond to the equivalent static forces
 Example:
0 0
m 0 0
k   3
m 0 0 1 ω 0.463
k 2 0 1 2   14
m
k 2 3
0
 The error is not very relevant (actually, ω 0.445 / )
 Excess error!
Earthquake-resistant design. Francesc López Almansa. Barcelona 42

21
19/07/2017

Types of seismic analyses (1)


 Linear static analysis (after absolute acceleration response spectra)
– Single-mode
– Multi-mode
 Nonlinear static analysis (pushover)
– Ordinary pushover analysis
– Multimodal pushover analysis
– Adaptive pushover analysis
 Nonlinear dynamic analysis
– Ordinary dynamic analysis
– Incremental Dynamic Analysis (IDA)
 Energy-based formulations
 Commercial software codes: ETABS, GSA, MIDAS, PERFORM-3D, RISA,
ROBOT, SAP, SOFISTIK, STAAD, STRAND7, TEKLA
 Advanced commercial software codes: ABAQUS, ANSYS, DIANA, NASTRAN
 These lists are non comprehensive
 Scientific software codes: OPENSEES, SEISMOSOFT (several programs),
ZEUSNL, RUAUMOKO, IDARC

Earthquake-resistant design. Francesc López Almansa. Barcelona 43

Types of seismic analyses (2)


 Types of seismic analysis according to EC-8:

Earthquake-resistant design. Francesc López Almansa. Barcelona 44

22
19/07/2017

Single-mode linear static analysis (1)


 The dynamic effect of a given W S  ag S a ζ, TF 
horizontal ground motion in each V
R
direction is represented by static
P SU Z C
equivalent forces Fk determined to V
generate displacements equal to the R
maximum dynamic ones (“actual”) m z
Fk  V N k k
 V: base shear
 W: building weight 
k 1
mk z k
 S: soil coefficient Fk
 : importance factor
 ag: seismic acceleration m φ m φ
Fk  V N k 1k  V Tk 1k
 Sa(,TF): spectral ordinate φ1 M r
 R: response modification 
k 1
mk φ1k
(reduction) factor
 Grey coefficients have been already
described N
V   Fk
Earthquake-resistant design. Francesc López Almansa. Barcelona k 1 45

Single-mode linear static analysis. Building weight (2)


 Building weight (“seismic weight”) corresponds to the level of W S  ag S a ζ, TF 
occupation that is expected when earthquake arises V
R
 European regulations consider permanent (G), variable (Q) and
accidental (A) actions; obviously, earthquakes are accidental
 In America G/Q are D/L (dead/live)
 In Europe there are safety factors for actions (G, Q, A) and for
strengths of materials (M, c, s); the  coefficient that is employed
in America (playing the same role) is not considered in Europe
 In Europe, under normal conditions in ULS: G = 1.35, Q = 1.5, A
= 1, M = 1.05 (higher values for connections), c = 1.5, s = 1.15
 In Europe, there are combination coefficients (for variable actions):
0, 1, 2 (0 ≥ 1 ≥ 2)
 Combination coefficients () depend on the type of action (people
for housing, people on a pedestrian bridge, snow, wind, etc.)
 For ordinary occupation buildings: 0 = 0.7, 1 = 0.5, 2 = 0.3
(except in congregation areas)
 Seismic building weight corresponds to the combination G + E Q
 E = 2  (art. 4.2.4 EC-8) 0.5    1
 Seismic combinations: G + E Q + EX + 0.3 EY; G + E Q + 0.3
EX + EY. This applies to Europe, America and everywhere

Earthquake-resistant design. Francesc López Almansa. Barcelona 46

23
19/07/2017

Single-mode linear static analysis. Importance factor (3)


 : importance factor depending on the use of building or bridge (also for non-
structural elements)
 EC-8 classifies buildings in four importance classes: I (minor), II (ordinary), III
(crowded), IV (essential); I = 0.8, II = 1, III = 1.2, IV = 1.4
 EC-8 classifies bridges in three importance classes: I (not critical), II (ordinary
road and railway), III (crowded and essential); I = 0.85, II = 1, III = 1.3
 ASCE 7-10 considers four risk categories (I, II, III, IV); importance factors (Ie)
are 1, 1, 1.25, 1.5, respectively
 Risk Category and seismicity determine the Seismic Design Category (A, B, C,
D)
 There are also categories E and F (highest)
 The importance factor can be considered linked to the return period of the seismic
action to be considered W S  ag S a ζ, TF 
V
R

Earthquake-resistant design. Francesc López Almansa. Barcelona 47

Single-mode linear static analysis. Response modification factor (4)

 If the base shear was obtained as V = W S  ag Sa, design forces


would be enormous
 Our aim is that building is not damaged under a severe shake (475
years return period); unfortunately, this objective is overambitious
 Being more realistic, we need to accept some degree of damage
 Then, simulating the actual behavior of damaged building would be
cumbersome
 Solution: to divide the design forces by a factor ( 1) and recover
the original design objective (linear analysis!) W S  ag S a ζ, TF 
V
 R: response modification (reduction) factor R
 R is commonly obtained after easy empirical criteria in terms of the
type of structure (even its material), plan symmetry, vertical
uniformity and overall quality (ductility)
 Using R = 1 corresponds to not accept any damage (ambitious
objective); the greater R, the higher the accepted damage
 Commonly, R ranges between 3 (low ductility) and 8 (very high
ductility)
 EC-8 states that q  1.5 (even for highly non-ductile structures)
Earthquake-resistant design. Francesc López Almansa. Barcelona 48

24
19/07/2017

Single-mode linear static analysis. Response modification factor (5)


 We should keep in mind that the use of R is merely a “design strategy”; in other
words, it is not real, and we cannot “believe our own lie”
 For instance, lateral displacements and other similar calculations should be performed
“without R” (i.e. R = 1); see art. EC-8 art. 4.3.4
 EC-8 considers “behavior factor” q; there are three levels of quality (“ductility
classes”): L (“low”, DC L), M (“medium”, DC M) y H (“high”, DC H)
 In the American regulation, they correspond to ordinary, intermediate and special
structures, respectively
 Important warning: expression V = W S  ag Sa / R should not be understood literally
(it is not found like that in any design code)
 For instance, if the design spectra have an initial growing branch, the value of the
initial ordinate is usually kept (except in the EC-8, see eqn. 3.13)
 This is equivalent to say that R is not actually constant

Sd(T)

B C
agS0/q T0 corresponds to TB in
D the EC-8
agS A

T
TA = 0 TB TC TD
Earthquake-resistant design. Francesc López Almansa. Barcelona 49

Single-mode linear static analysis. EC-8 and ASCE 7-10 (6)


 EC-8 considers two spectra: Se (q =1) and Sd (q >1) W S  ag S a ζ, TF 
V
 EC-8 (art. 4.3.3.2.2): Fb = Sd(T1) m  R
 Fb is the base shear force (V)
 Sd(T1): ordinate of the design spectrum; it includes the
influence of S, , ag, Sa and q (it plays the same role than R)
 Correction factor :  = 0.85 if T1  2 Tc and N > 2;
otherwise,  = 1
  accounts for the fact that in buildings with at least three
stories, the effective modal mass of the fundamental mode is
smaller, on average by 15%, than the total building mass
 Apparently, the effect represented by  is over-simplified
(small change of period can generate huge effect on base
shear)
 Be aware that Sd(0) = 2/3 ag S (instead of ag S)
 ASCE 7-10: V = Cs W
 Cs: seismic response coefficient; it includes the influence of
S, , ag, Sa and R

Earthquake-resistant design. Francesc López Almansa. Barcelona 50

25
19/07/2017

Single-mode linear static analysis. Vertical distribution (7)


 The base shear force has to be distributed W S  ag S a ζ, TF 
among all the stories: ∑ V
R
 EC-8: ∑ ; si is the
displacement of i-th story in the
mk φ1k mk φ1k
fundamental mode shape Fk  V V
N
φ1T M r
m

 ASCE 7-10: ; hi is the k φ1k

k 1
“cumulated” height of i-th story, and
exponent k = 1 (straight configuration) if
TF ≤ 0.5 s or k = 2 if TF ≥ 2.5 s (parabola).
Fk
For periods in between 0.5 and 2.5, linear
interpolation for k is suggested
 Some codes (e.g. New Zealand, Chile)
state that the force in the top floor should
be bigger; this accounts, in a simplified
way, for the higher-mode effects

N
V   Fk
Earthquake-resistant design. Francesc López Almansa. Barcelona k 1 51

Single-mode linear static analysis. Horizontal distribution (8)


 Single-mode method can be used only for buildings with plan symmetry, therefore,
static forces are applied in the centroid of each story
 However, accidental torsion effects must be taken into consideration:
 If the building is represented by a 3-D model, 5% accidental eccentricity between centers
of gravity and rigidity must be considered (EC-8, ASCE 7-10). This operation is
performed inside the software
 If the building is represented by 2-D models, uniform distribution among equal-rigidity
frames is corrected with factor 1 + 1.2 x / Le; outer frames receive 60% more (art.
4.3.3.2.4(2) EC-8). This operation is performed outside the software
 If most of the members providing torsion rigidity (walls and bracings) are
concentrated in façades, these criteria are conservative; in the opposite case, they are
under-conservative

Earthquake-resistant design. Francesc López Almansa. Barcelona 52

26
19/07/2017

Single-mode linear static analysis. P- effects (9)


 If the top floor displacements are important, second order analysis
might be necessary (P- effect)
 General criterion: P- analysis is necessary if, in the columns, the 2nd
order moments exceed the 1st order ones more than 10% (as to classify
structures sway or non-sway)
 Following verifications should be carried out for each story
(commonly, first and second are the most critical)

 ASCE 7-10 art. 12.8.7: θ 0.1. Where, Px: unfactored

gravity force, : interstory drift, Ie: importance factor, Vx: seismic shear
force, hsx: story height, Cd: coefficient similar to R (Table 12.2-1)
 The objective of Cd is to neutralize the effect of R
 Remarkably,  is to be obtained “without R” (e.g. with Cd)

 EC-8 art. 4.4.2.2(2): θ 0.1

 EC-8 art. 4.4.2.2(3): if 0.l <  ≤ 0.2, second-order effects may
approximately be taken into account by multiplying the relevant
seismic action effects by 1 / ( 1  )
 EC-8 art. 4.4.2.2(4): coefficient  shall not exceed 0.3
 Vtot and dr should be “q free”: ds = qd de (art. 4.3.4, qd = q)

Earthquake-resistant design. Francesc López Almansa. Barcelona 53

Single-mode linear static analysis. Exercise (No. 9) (10)


 Exercise (No. 9)
 Obtain the equivalent static
forces Fk at each story according
to EC-8
 Assume that the forces are
proportional to the floor mass
and height
 4-story building with square plan
layout 10  10 m2. Story height
is 3 m (H = 12 m). Seismic
acceleration is ag = 0.4 g
 Moment-resisting concrete
frames in both directions
 Soil B; damping factor  = 0.05; S (T)
G = 12 kN/m2; Q = 3 kN/m2; E
d

= 0.5 (floors); E = 0 (roof); q =


4 (both directions) a S /q g 0
B C

 Type 1 spectrum
 Hint: easy exercise! (be careful
D
aS g A
with units) T
TA = 0 TB TC TD

Earthquake-resistant design. Francesc López Almansa. Barcelona 54

27
19/07/2017

Multi-mode linear static analysis. SRSS (1)


Wi *  mi*g m1   mr  0.9 mT
* *
 r: number of modes to be included in the
analysis (in each direction)
Wi * S  ag S a ζ i , Ti 
 Usual criterion: the sum of the equivalent modal Vi 
weights covers 90% of the total weight R
 Wi*: equivalent modal weight of i-th mode mk φ ik m φ
 Sa(i,Ti): spectral ordinate for i-th mode Fik  Vi N  Vi Tk ik
φi M r
 Quadratic combination (SRSS)
 Commonly, higher modes (have small base k 1
mk φ ik 
shear) influence mostly top stories
S
 SRSS can be used only if modal periods are well
separated (Ti / Tj ≤ 0.90) T 2
Fik  Wk i S a ζ i , Ti  φ ik
(φ i M r ) R
mi* 
Vik  FiN    Fi k φ iT M φ i Vk  V    V
2
1k
2
rk
10 10 10
9 9 9
8 8 8
7 7 7
6 6

=
6
5
4
+ 5
4
5
4
3 3 3
z
2 2 2
1
Fk  Vk 1  Vk
1 x 1
y

Earthquake-resistant design. Francesc López Almansa. Barcelona


55

Multi-mode linear static analysis. CQC (2)


 If modal periods are not well separated (ratio > r
0.90), the possibility of having almost Vk    ij VikVjk
simultaneous maxima is higher, and SSRS might i, j1
become unconservative
 CQC (Complete Quadratic Combination) 3
criterion results of incorporating “cross terms”  Ti   Ti  2
into the combination 8ζ 1  
2
 T  T 
 ij = ji  ij   j  j 

 If Ti = Tj, ij = 1 (therefore, ii = 1)  T   2 2 2

 Since ij  0, VjkVik must be positive too 1   i    4 ζ 2 Ti 1  Ti 


 T   Tj  Tj 
 This criterion is just necessary in irregular   j 
buildings
 Another version (CQC-4) has been proposed
1
10 10 10
0.8
9 9 9
 = 0.05
8 8 8
7
0.6 7
7
6 6

=
6

+
0.4
ij

5 5 5
4 4 0.2 4
3 3 3
z
2 2 0 2
1 x 1 0.51 0.6 0.7 0.8 0.9 1
y
Ti / Tj

Earthquake-resistant design. Francesc López Almansa. Barcelona 56

28
19/07/2017

Multi-mode linear static analysis. EC-8 and ASCE 7-10 (3)


 EC-8 (4.3.3.3 Modal response spectrum analysis): apart from the usual criterion, any
mode with more than 5% modal participation factor (m*i / m) cannot be excluded
 If this requirement cannot be satisfied (e.g. in buildings with a significant
contribution from torsional modes), the minimum number (k) of modes to be included
should satisfy k  3 N ½ (N: number of floors) and Tk  0.2 s
 ASCE 7-10 (12.9 Modal response spectrum analysis): where the combined response
for the modal base shear (Vt) is less than 85% of the calculated base shear (V) using
the equivalent lateral force procedure, the forces shall be multiplied by 0.85 V / Vt (Vt
is the base shear from the required modal combination). Therefore, the reduction will
not be lower than 0.85
 Where the calculated fundamental period T exceeds Cu Ta in a given direction, Cu Ta
shall be used instead of T in that direction; Cu: coefficient for upper limit on
calculated period (Table 12.8-1). This applies also for Equivalent Lateral Force
Procedure

Earthquake-resistant design. Francesc López Almansa. Barcelona 57

Multi-mode linear static analysis. Asymmetric buildings (4)


 Analysis of asymmetric buildings for excitation in x direction
 Each mode contains components of horizontal displacement in x and y
directions, and torsion rotations 
 The sum of the modal masses should exceed 90% of the total mass in
each direction (x, y and )
 Displacement vector corresponding to a generic mode i:

 Ti 
2  xi  T
 i φ i q   y  Γ i  φTi M r r  (1,...,1, 0
T
q i  S a Ti , ζ  ,...,0, 0,...,0)
 i   
 2π i
φ 
φi M φi N N N

 i  φ xi 
 
S S φ i   φ yi 
Fx ik  Wk i S a Ti , ζ i  φ x ik Fy ik  Wk i S a Ti , ζ i  φ y ik
R R φ 
 i 
S
Tik  I k i S a Ti , ζ i  g φ φ ik
 Fx ik: Force in x direction, mode i, story k
R  Fy ik: Force in y direction, mode i, story k
 Tik: Torque, mode i, story k
a2  b2
Wk  mk g I k  mEarthquake-resistant
k design. Francesc López Almansa. Barcelona 58
12

29
19/07/2017

Nonlinear static analysis. Definition (1)


 Pushover analysis of a construction (building or bridge) consists
of investigating its behavior under constant gravity loads and
growing lateral forces
 Vertical forces correspond to seismic excitation (G + E Q)
 Horizontal forces grow keeping the same vertical variation pattern
 Common patterns: uniform, triangular or modal (1st mode shape)

V/W
Capacity
Fk Vu / W curve

Vy / W
/H
V   Fk
 Subindexes y and u account for yielding and ultimate, respectively
 Damage and displacement are clearly correlated, force and damage not so clearly
Earthquake-resistant design. Francesc López Almansa. Barcelona 59

Nonlinear static analysis. Plastic hinges (2)


 Nonlinear behavior of structural
elements is commonly
represented by plastic hinges,
usually located at their ends
 There are plastic hinges of V, N,
T, M, y “N + M”
 Nonlinear behavior of plastic
hinges is commonly described by
± moment-curvature (or rotation)
or force-displacement laws
Failure

Residual
Plastification strength

Earthquake-resistant design. Francesc López Almansa. Barcelona 60

30
19/07/2017

Nonlinear static analysis. Plastic hinges (3)


 Design codes (e.g. FEMA 356) propose moment-curvature (or force-displacement)
laws for the most common situations
 The initial branches of these laws are obtained from theoretical (local) analysis and
the plastic branches are selected as the envelopes of the experimental hysteresis loops
 These laws are implemented in the major software codes
 For unusual or new members, there are no available experiments; therefore, tests and
theoretical analysis are required (in this order)

Earthquake-resistant design. Francesc López Almansa. Barcelona 61

Nonlinear static analysis. Plastic hinges (4)


 Typical moment-curvature law
 Q / Qy: M / My; : rotation ( / l: curvature, where l is the hinge length)
 AB: linear elastic branch; BC: plastic branch; CDE: residual branches (to be
determined after testing)
 Before yielding (B), rotation is zero; however, some rotation (y) is assumed for
point B
 For beams: y = My L / 6 E I; for columns: y = My L / 6 E I (1  N / Ny). This the
“chord rotation”
 Noticeably, the initial flexibility of the moment-rotation law (AB branch) should
not be considered in the global analysis; fully rigid connection must be considered
instead

M / My My
My

Beam

 / y
Earthquake-resistant design. Francesc López Almansa. Barcelona 62

31
19/07/2017

Nonlinear static analysis. Plastic hinges (5)


 Concentrated Plasticity. Plastic hinges can be either zero-length or a fixed
length can be assigned to them; there are a number of simple empirical
expressions
 Distributed Plasticity. Fiber Models fcm = fck + 8 (MPa)
 Average values of material parameters instead of characteristic ones
 No safety factors!

Earthquake-resistant design. Francesc López Almansa. Barcelona 63

Nonlinear static analysis. Initial cracking (6)


 In reinforced concrete, members are commonly cracked form the very
beginning, i.e. for zero pushing forces
 FEMA 356 indicates percentages of reduction of stiffness for beams,
columns, slabs and walls

 Meaning of subindexes; c: concrete, g: gross, w: web, s: steel.


Earthquake-resistant design. Francesc López Almansa. Barcelona 64

32
19/07/2017

Nonlinear static analysis. Case (7)

 “Strong column-weak beam” condition is basically fulfilled


 Hinges in the bottom section of the bottom columns might be not realistic
 Beams should not be uniform along height
 There is moment inversion in the bottom stories
 Since the behavior is nonlinear, the bending moment laws of the 1st floor
columns are different than foreseen by linear analysis

Earthquake-resistant design. Francesc López Almansa. Barcelona 65

Nonlinear static analysis. N-M (8)


 N-M interaction diagram for reinforced concrete

Earthquake-resistant design. Francesc López Almansa. Barcelona 66

33
19/07/2017

Nonlinear static analysis. Example (9)


 Single-bay single-story moment-resistant steel frame
 Gravity load 30 kN/m (beam); steel S275 (fy = 275 MPa); beam IPE 450 (Wpl = 1702
cm3, My = 468.05 kNm); columns HEB 300 (Wpl = 1869 cm3, My = 513.96 kNm)
 The structural behavior is linear, except for flexural plastic hinges at member ends
 No reduction in yielding moment in columns due to interaction with axial force
 In the top joints, hinges will appear earlier in beam than in columns
 Simplified hinge behavior: infinite initial stiffness and horizontal plastic branch (no
strain hardening)
 First-order analysis; also members are assumed to be infinitely rigid in axial direction

Earthquake-resistant design. Francesc López Almansa. Barcelona 67

Nonlinear static analysis. Example (10)


 Bending moments laws under gravity and pushing loads
 First two plastic hinges appear almost simultaneously at the right top and bottom
joints

Earthquake-resistant design. Francesc López Almansa. Barcelona 68

34
19/07/2017

Nonlinear static analysis. Example (11)


 Combined bending moments law under gravity and pushing loads
 Horizontal pushing force and displacement are 283.7 kN and 45.48 mm,
respectively

Earthquake-resistant design. Francesc López Almansa. Barcelona 69

Nonlinear static analysis. Example (12)


 Bending moments law under pushing load; to be added to the previous law
 Third plastic hinge appears at the left bottom joint
 Horizontal pushing force and displacement are 54.95 kN and 21.9 mm,
respectively

Earthquake-resistant design. Francesc López Almansa. Barcelona 70

35
19/07/2017

Nonlinear static analysis. Example (13)


 Bending moments law under pushing load; to be added to the previous law
 Fourth plastic hinge appears at the left top joint
 Horizontal pushing force and displacement are 57.36 kN and 113.2 mm

Earthquake-resistant design. Francesc López Almansa. Barcelona 71

Nonlinear static analysis. Example (14)


 Capacity curve
400
Force (kN)

300

200
and?
ASCE 7-10 does not
100 consider pushover as a
valid design strategy
0
0 50 100 150
Displacement (mm)
Earthquake-resistant design. Francesc López Almansa. Barcelona 72

36
19/07/2017

Nonlinear static analysis. R factor (15)


 First utility of pushover analysis: Equal-displacement approach
 Blue point: design force (theoretical onset of yielding)
 Purple point: design force (actual onset of yielding)
 Red point: collapse force (no stiffness)
 Capacity curves provide estimations of response modification factor (R or q)
 These calculated values of R will be higher than default ones

Fe Fe Fu Fe
R   Ω FEMA-356
Fy Fu Fy Fu

αe αe αu
q  EC-8
α1 α u α1

Earthquake-resistant design. Francesc López Almansa. Barcelona 73

Performance-based design. Definition (1)


 PBD (Performance-Based Design)
 American regulations (FEMA 356) propose four Performance Levels: FO
(“Fully Operational”), IO (“Immediate Occupancy”), LS (“Life Safety”)
and CP (“Collapse Prevention”)
 European regulations (EC-8 Part 3) propose three Limit States: DL
(“Damage Limitation”), SD (“Significant Damage”) y NC (“Near
Collapse”)
 DL  IO, SD  LS, NC  CP
 Performance Levels and Limit States have been proposed for retrofit but
can be also used for new construction

Earthquake-resistant design. Francesc López Almansa. Barcelona 74

37
19/07/2017

Performance-based design. FEMA 356 (2)


 FEMA 356 2000 (“ ... seismic rehabilitation of buildings“)
 FO (“Fully Operational”): no damage at all.
 IO (“Immediate Occupancy”): only very limited structural damage, the
construction remains safe to occupy, the structure essentially retains the
pre-earthquake design strength and stiffness, the risk of life-threatening
injury is very low, and there is no permanent drift.
 LS (“Life Safety”): damage to structural components but retains a
margin against partial or total collapse, the risk of life-threatening injury
is low, and it should be possible to repair the structure.
 CP (“Collapse Prevention”): damage to structural components such that
the structure continues to support gravity loads but retains no margin
against collapse. Structural damage potentially includes significant
degradation in the stiffness and strength of the lateral-force-resisting
system, large permanent lateral deformation, and (to a more limited
extent) degradation in vertical-load-carrying capacity. The structure may
not be technically practical to repair.

Earthquake-resistant design. Francesc López Almansa. Barcelona 75

Performance-based design. EC-8 Part 3 (3)

Earthquake-resistant design. Francesc López Almansa. Barcelona 76

38
19/07/2017

Performance-based design. PBD vs T (4)


 In EC-8, correspondence between Limit
States and return periods: DL  225 years,
SD  475 years, and NC  2475 years
 In FEMA 356, three objectives are stated:
Basic Safety, Enhanced and Limited
 Basic Safety Objective cares only for LS
(475 years) and CP (2475 years)
 Enhanced Objectives care for LS (475
72
years), CP (2475 years) and FO and IO
(72, 225 or 475 years); also FO, IO or LS 225
alones (2475 years) 475
 Limited Objectives care for LS (475 years)
2475
or CP (2475 years); LS (72 or 225 years)
and CP (72, 225 or 475 years)

Earthquake-resistant design. Francesc López Almansa. Barcelona 77

Performance-based design. PP (TD) (5)


 Each “performance point” (or “target drift”) represents the effect on the
building or bridge (in terms of force and displacement) of a ground motion whose
severity corresponds to the return period associated with the limit state

LS

IO

CP

Performance
points

Earthquake-resistant design. Francesc López Almansa. Barcelona 78

39
19/07/2017

Performance-based design. Acceptability (6)


 Research projects
HAZUS (in America) SD
MD ED
and RISK-UE (in HD
Europe) have
calibrated damage in ND
terms of yielding (y)
and ultimate (u) levels LS CP

 ND: No Damage; SD: IO


Slight Damage; MD:
Moderate Damage;
ED: Extensive
Damage; HD: Heavy
Damage

Earthquake-resistant design. Francesc López Almansa. Barcelona 79

Performance-based design. Sa Sd (7)


 Absolute acceleration and relative displacement response spectra

Sv(,T) = Sa(,T) (T / 2) Sd(,T) = Sa(,T) (T / 2)2


Sd

TB TC TD
Sa T

2.50

TB TC TD
80 T
Earthquake-resistant design. Francesc López Almansa. Barcelona

40
19/07/2017

Performance-based design. ADRS (8)


 Demand is presented as spectrum
 Instead of Sa(T) and Sd(T) vs. T, Sa(T) vs. Sd(T) (ADRS, Acceleration
Displacement Response Spectrum)

Sa(T) T < TB T = TB T = TC T > TC

2.50
 < 0.05
 = 0.05
 > 0.05
1 T = TD
Sd(,T) = Sa(,T) (T / 2)2
Sd(T)
Sd(TB;0.05) Sd(TC;0.05) Sd(TD;0.05)

Earthquake-resistant design. Francesc López Almansa. Barcelona 81

Performance-based design. ATC-40 (9)


 Intersections between capacity curve and demand spectra are the “performance
points” (or “target drifts”)
 Conversion factors between both charts:
V /W V /W
Sa  
φ
T
1Mr 2

/ rT M r
m1* / mT
V/W φ M φ1
T
1
Capacity
 
S d  curve 
φ1T M r 1 Iterative procedure:
φ1T M φ1
φ1N  1
/H
Sa(T)
 = 0.05
 < 0.05  = 0.08
 = 0.05  = 0.10
 > 0.05

Sd(T)

Earthquake-resistant design. Francesc López Almansa. Barcelona 82

41
19/07/2017

Performance-based design. ATC-40 (10)


Procedure A (8.2.2.1.2)
1. Trial performance point (api, dpi)
2. Equal-energy bilinear approximation providing the yield
point (ay, dy)
3. Effective damping ratio corresponding to trial point (%):
β 5 
4. Coefficient  is described in Table 8-1 (0.33   1)
5. If yield and trial points are coincident, eff = 5
6. Reduce the demand spectrum with factors SRA and SRV
. . . .
7. SR .
; SR .
8. If eff = 5, SRA = SRV = 1
9. Minimum values of SRA and SRV apply (Table 8-2)
10. If the new intersection point does not fit point (api, dpi)
with 5% tolerance, return to step 1 selecting the
intersection as new trial point
11. Repeat until convergence

Earthquake-resistant design. Francesc López Almansa. Barcelona 83

Performance-based design. ATC-40 (11)


Procedure A (8.2.2.1.2)
 If trial and yield points are aligned with the origin, eff = 5
 The higher the angle between both straight lines, the higher the value of eff
 In saw tooth capacity curves, the bilinear approximation is generated for the
(degraded) branch the trial point belongs to

Earthquake-resistant design. Francesc López Almansa. Barcelona 84

42
19/07/2017

Performance-based design. N2 (12)


 N2 method (EC-8) φT M r

 M ; N = 1 φT M φ
 m∗ T M r r
 F* = m* g
 Fb = F* ; dn = d* 
 A: collapse point
∗ ∗ ∗
  ½ ∗ ∗
 ∗ is the area under the
capacity curve
 ∗ 2 π ∗/ ∗

 Fy* = k* dy*
Earthquake-resistant design. Francesc López Almansa. Barcelona 85

Performance-based design. N2 (13)


 N2 method (EC-8)
Demanding

∗ ∗
 d ADRS

 ∗ (short periods range):
If F ∗  ∗ ∗ (elastic) Capacity
∗ ∗ curve
d d
If F ∗ ∗ ∗
(anelastic)

d
d∗ 1 1 ∗  d∗
∗ ∗
Demanding
F∗
ADRS
 ∗ 
(long and medium Capacity
periods): curve
d∗ d∗

 d 3 d∗

Earthquake-resistant design. Francesc López Almansa. Barcelona 86

43
19/07/2017

Performance-based design. N2 (14)


 N2 method (EC-8)
 If d∗  d∗ (the
considered input does
not generate collapse),
an iterative procedure
might be used:
 In the bilinear capacity
curve d∗ is replaced by
d∗ ; d∗ is accordingly
modified

Earthquake-resistant design. Francesc López Almansa. Barcelona 87

Performance-based design. Examples (15)


 Wide-beam buildings in Spain

Earthquake-resistant design. Francesc López Almansa. Barcelona 88

44
19/07/2017

Performance-based design. Examples (16)


 Thin-wall beam buildings in Peru

Earthquake-resistant design. Francesc López Almansa. Barcelona 89

Performance-based design. Examples (17)


 Steel buildings in Colombia

Earthquake-resistant design. Francesc López Almansa. Barcelona 90

45
19/07/2017

Nonlinear dynamic analysis. Definition (1)


 Seismic design based on dynamic (time-history) analysis consists in representing
expected seismic inputs by accelerograms
 Since actual behavior of structures under severe inputs (i.e. corresponding to large
return periods) is highly nonlinear, dynamic analysis must be also nonlinear
 This formulation is the most natural one; also, apparently, is more accurate than
using spectra
 However, there are two major problems:
 There is a big variety of expected inputs (amplitude, frequency content, pulses,
duration, etc.)
 Nonlinear dynamic analyses can be extremely costly, in terms of computational effort
 Design codes define the number and characteristics of the accelerograms to be
employed
 Since actual seismic excitations are 3-D (or even 6-D, if we account for rotational
components of ground motion), dynamic analysis should consider, at least, the
joint excitation of both horizontal components of input ground motion
 For this purpose, both accelerograms should be exactly synchronized
 The same accelerogram may not be used simultaneously along both horizontal
directions (EC-8 3.2.3.1.1(2))
Earthquake-resistant design. Francesc López Almansa. Barcelona 91

Nonlinear dynamic analysis. Accelerograms (2)


 Accelerograms can be either recorded (historical), simulated (from the known
seismogenic mechanism) or artificial (fitting design spectra)
 Simulated accelerograms are (almost always) unfeasible
 EC-8 3.2.3.1.3. Requirements for the artificial accelerograms: (i) stationary phase
at least 10 s, (ii) minimum 3 accelerograms, (iii) the mean of the zero-period
spectral ordinates (Sa(0)) should not be smaller than ag S, and (iv) in the range 0.2
T1  2 T1, no value should be under 90% of design elastic response spectrum
 ASCE 7-10 16.1.3. Similar than (iv) from EC-8, but with 1.5 T1 instead of 2 T1
 This range accounts for the influence of higher modes (range 0.2 T1  T1) and the
period elongation during the shake (ranges T1  2 T1 or T1  1.5 T1)
 Nowadays, nonlinear dynamic analysis is not yet widely employed, despite the
powerful computational tools that are commonly available
 A good spectrum-based analysis is better than a poor dynamic analysis
 Dynamic analysis is only justified: (i) in research, (ii) for highly complex or very
important structures, (iii) when required by design codes (e.g. in base isolation),
and (iv) when the results from simplified analyses (i.e. spectra-based) are not
acceptable

Earthquake-resistant design. Francesc López Almansa. Barcelona 92

46
19/07/2017

Nonlinear dynamic analysis. Computational issues (3)


 Nonlinear equation of motion:
 Solution of this equation (and of the linear one) is known as “time-history analysis”
 Same models than for pushover analysis can be used
 Response should be obtained, in discrete time, by a step-by-step procedure; commonly,
time step (discretization period or sampling period: t) is constant along the whole
duration of analysis
 If structural behavior is linear (Q(x) = K x), response for next instant is obtained
directly assuming an interpolation criterion in the interval; each criterion leads to a
different calculation algorithm
 If structural behavior is nonlinear (Q(x) = Kt x, where is the Kt tangent stiffness
matrix), stiffness at the beginning and at the end of interval is different and, therefore,
response for next instant cannot be obtained directly (even assuming an interpolation
criterion in the interval) and calculation must be iterative
 For both linear and nonlinear analysis, the most employed algorithm is Newmark
 Selection of time step is a crucial issue; usual criteria for linear analysis do not apply
for nonlinear analysis; significantly shorter time steps are required
 The faster the excitation (high frequency) and the stiffer the structure (short period), the
shorter the required time step
 The only valid rule for time step selection is to start with a coarse time discretization
(t = 0.01 s) and then refine it (lowering t) until obtaining “time convergence”; values
below 0.001 s are extremely frequent
Earthquake-resistant design. Francesc López Almansa. Barcelona 93

Nonlinear dynamic analysis. Computational issues (4)


0.01 (s) 0.001 (s) 0.0005 (s)
3

EXAMPLES OF
Storey

2
SENSITIVITY TO t

1
0 10 20 30 40
Energy (kJ) Hysteretic
energy and
4000 pounding force
Pounding Force (kN)

dt = 0.01 s in a colliding
dt = 0.001 s building
dt = 0.0005 s
2000

Time (s)

0
8 8.05 8.1 8.15 8.2 8.25
Earthquake-resistant design. Francesc López Almansa. Barcelona 94

47
19/07/2017

Nonlinear dynamic analysis. IDA (5)


 IDA “Incremental Dynamic Analysis” is a kind of dynamic pushover analysis; a
given structure undergoes inputs scaled with different (growing) factors
 There are two types of IDA analysis: for a single input or for several inputs
 IDA Curves are similar to capacity curves; vertical axis refers to any parameter
characterizing input severity, and horizontal axis contains top floor displacement (or
similar magnitudes)
 In each dynamic analysis, the structure “is reset” (e.g. is undamaged)

Multi-input

Single-input

Earthquake-resistant design. Francesc López Almansa. Barcelona 95

Nonlinear dynamic analysis. Structural resurrection (6)

Earthquake-resistant design. Francesc López Almansa. Barcelona 96

48
19/07/2017

Nonlinear dynamic analysis. Example (7)

100 NO WALLS
LOW WALL DENSITY
HIGH WALL DENSITY
50
Displacement (mm)

0
0 2 4 6 8 10
Time (s)
‐50
Permanent
displacement!
‐100

Earthquake-resistant design. Francesc López Almansa. Barcelona 97

Nonstructural components (1)


 Non structural elements
(appendages) can be
antennae, façades, cladding
panels, tanks, machinery,
furniture, appliances, etc.
 Those elements must be
seismically designed
 The seismic input of the
device depends on its H
position in the building; the
higher the worst (the
seismic excitation is
progressively amplified z
along the building height)
 Therefore, the roof is the
most critical (dangerous)
location

Earthquake-resistant design. Francesc López Almansa. Barcelona 98

49
19/07/2017

Nonstructural components. EC-8 (2)


 Article 4.3.5 (EC-8 1) specifies the seismic analysis of non structural elements
 Fa = Sa Wa a / qa; Fa: seismic design force; Wa: element seismic weight; a = 1 or
1.5; qa = 1 or 2
 Sa: floor spectra;  = ag / g; S: soil coefficient; z: height of the point the element
is attached; H: building height; Sa ≥  S; T1 / Ta: fundamental period of building /
element
 The higher the point where the element is installed, the greater the seismic
excitation; if z = H (roof appendage):
Sa a) = Sa(2 T1  Ta); Sa(0) =  S 2.5; Sa(T1) =  S 5.5; Sa(4.32 T1) = 0; Sa() =   S
(T
0.5

5
  Floor spectrum
 3 1  z 
4

  H  3
Sa  α S   0.5
Sa
2
 1  1  Ta 
2

  T1  
1

0
0 1 2 3 Ta / T1 4

Earthquake-resistant design. Francesc López Almansa. Barcelona 99

Nonstructural components. ASCE 7-10 (3)


 Chapter 13 (ASCE 7-10) specifies the seismic analysis of non structural elements
 A list of nonstructural components which are exempt from verification is included
.
 1 2 ; 0.3 1.6

 Fp: seismic design force; ap: component amplification factor (ranges between 1
and 2.5, plays the role of spectral ordinate); Wp: component operating weight; Rp:
component response modification factor (1 to 12); Ip: importance factor

 Alternative: ; 0.3 1.6

 Ax: torsional amplification factor; ai: acceleration at level i


 In addition, the component shall be designed for a concurrent vertical force ±0.2

 Where the weight of a nonstructural component is greater than 25% of the
building seismic weight, it shall be classified as a non-building structure and
shall be designed (15.3.2) jointly with the building

Earthquake-resistant design. Francesc López Almansa. Barcelona 100

50
19/07/2017

Bibliography. Books
 Akiyama H. Earthquake-Resistant Design for Buildings. Tokio University Press 1988.
 Ambrose J.E., Vergun D. Diseño simplificado de edificios para cargas de viento y sismo. Limusa
1986.
 Bazán E., Meli R. Diseño sísmico de edificios. Limusa 2002.
 Bozorgnia Y., Bertero V.V. Earthquake Engineering: from Engineering Seismology to
Performance-Base Engineering. CRC Press 2004.
 Bozzo L.M., Barbat A.H. Diseño sismorresistente de edificios. Ed. Reverté 2000.
 Chandrasekaran S. et al. Seismic Design Aids for Nonlinear Analysis of Reinforced
Concrete Structures. CRC Press 2010.
 Datta T.K. Seismic Analysis of Structures. J. Wiley 2010.
 Dowrick D.J. Earthquake Resistant Design for Engineers and Architects. J. Wiley 1977.
 Fajfar P., Krawinkler H. Seismic Design Methodologies for the Next Generation of Codes.
Balkema 1997.
 García L.E. Dinámica Estructural Aplicada al Diseño Sísmico. Universidad de Los Andes
(Bogotá) 1998.
 Naeim F. The Seismic Design Handbook. Van Nostrand Reinhold 2002.
 Newmark N.M., Rosenblueth E. Fundamentos de ingeniería sísmica. Diana 1978.
 Paulay T., Priestley M.J.N. Seismic Design of Reinforced Concrete and Masonry Buildings.
John Wiley 1992.
 Priestley M.J.N., Seible F., Calvi G.M. Seismic Design and Retrofit of Bridges. John Wiley
1996.
 Priestley M.J.N., Calvi G.M., M.J. Kowalski. Displacement-Based Seismic Design of
Structures. IUSS Press 2007.
 Rosenblueth E. Design of Earthquake Resistant Structures. Pentech Press 1980.
 Wakabayashi M. Earthquake Resistant Design for Buildings. McGraw-Hill 1986.

Earthquake-resistant design. Francesc López Almansa. Barcelona 101

Bibliography. Codes
 NCSE-02. Norma de Construcción Sismorresistente: Parte General y Edificación.
Ministerio de Fomento 2002.
 NCSP-07. Norma de construcción sismorresistente. Ministerio de Fomento 2007.
 Seismic Provisions for Structural Steel Buildings. AISC (American Institute on Steel
Construction) 2005.
 Prequalified Connections for Special and Intermediate Steel Moment Frames for Seismic
Applications. AISC (American Institute on Steel Construction) 2005.
 FEMA 356. Pre-standard and commentary for the seismic rehabilitation of buildings.
Federal Emergency Management Agency 2000.
 ACI 318-11. Building Code Requirements for Structural Concrete. ACI (American
Concrete Institute ) 2011.
 ASCE/SEI 7-10. Minimum Design Loads for Buildings and Other Structures. ASCE
(American Society of Civil Engineers) 2010.
 Fardis M.N., Carvalho E., Elnashai A., Faccioli, Pinto Plumier A. Designers’ Guide to
Eurocode 8. Thomas Telford 2005.

Earthquake-resistant design. Francesc López Almansa. Barcelona 102

51
19/07/2017

Internet Sites
 http://www.asce.org/
 http://www.concrete.org/
 http://www.aisc.org/
 https://www.atcouncil.org/
 http://www.fema.gov/
 http://peer.berkeley.edu/
 http://earthquake.usgs.gov/
 http://mceer.buffalo.edu/
 http://eurocodes.org.ua/
 http://www.roseschool.it/
 http://mae.cee.illinois.edu/software_and_tools/zeus_nl.html
 http://www.civil.canterbury.ac.nz/eq/eqeng.shtml

Earthquake-resistant design. Francesc López Almansa. Barcelona 103

52

Potrebbero piacerti anche