s5 Ec8-Lisbon m Fardis

Embed Size (px)

Citation preview

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    1/28

    Dissemination of information for training Lisbon 10-11 February 2011 1

    Design and detailing ofsecon ary se sm c e emen s

    Provisions for concrete dia hra ms

    M.N. FardisUniversity of Patras (GR)

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    2/28

    Dissemination of information for training Lisbon 10-11 February 2011 2

    MASONRY-INFILLED FRAMES

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    3/28

    Overall effect of masonry infil ls on seismic performance

    Dissemination of information for training Lisbon 10-11 February 2011 3

    Field experience & numerical/experimental research do show

    that: masonry infills attached to the structural frame, in general have a

    beneficial effect on seismic erformance es eciall if the buildin

    structure has little engineered earthquake resistance.

    ,

    distributed infill panels: reduce, through their in-plane shear stiffness, storey drift demands &

    deformations in structural members

    increase, via their in-plane shear strength, storey lateral force resistance,

    contribute, throu h their h steresis, to the lobal ener dissi ation.

    In buildings designed for earthquake resistance, non-structural

    significant overstrength.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    4/28

    Position of EC8 on masonry infi lls

    Dissemination of information for training Lisbon 10-11 February 2011 4

    Eurocode 8 does not encourage designers to profit from

    the beneficial effects of masonry infills to reduce the

    seismic action effects for which the structure is designed.

    Eurocode 8 warns against the adverse effects of infills &

    .

    & the surrounding frame (by shear connectors, or other

    as a confined masonry building, not as a concrete

    structure with masonr infills.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    5/28

    Possible adverse effects of masonry infil ls

    Dissemination of information for training Lisbon 10-11 February 2011 5

    Infills that are too strong & stiff relative to the concrete

    structure itself

    may override its seismic design, including thee or s o e es gner e n en o o con ro

    inelastic response by spreading inelastic

    (e.g. when ground storey infills fail soft storey).

    Infills non-uniformly distributed in plan or in elevation:

    part of the structure.

    Adverse local effects on structural frame pre-emptive brittle failures.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    6/28

    Possible adverse effects of masonry infills (contd)

    Dissemination of information for training Lisbon 10-11 February 2011 6

    Best way to protect concrete building from adverse

    effects of irregular masonry infilling:

    shear walls sufficiently strong/stiff to overshadow anye ec s o e n ng.

    Shear walls that resist at least 50% of the seismic base

    -

    for waiving the special requirements for buildings with

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    7/28

    Possible adverse effects of masonry infills (contd)

    Dissemination of information for training Lisbon 10-11 February 2011 7

    Worst possible effect: Open ground storey soft-storey

    ~02 (c)

    infill strut

    1

    2-storey frame: Elements in infilled storey shielded from large moments &

    deformations. But round store columns are overloaded. See:

    (a) bending moments & deformation in frame w/o infills;(b), (c) bending moments & deformation in frame w/ stiff infills in 2nd storey.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    8/28

    Open ground storey

    Dissemination of information for training Lisbon 10-11 February 2011 8

    Collapse of ground storey due to reduction of

    infills:

    (a) Olive View Hospital, San Fernando, Ca, 1971;(b) Aegio (GR) 1995

    (b)

    Collapse of ground storey due to reduction of infills:(a) Olive View Hospital, San Fernando, Ca, 1971; (b) Aegio (GR) 1995

    a

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    9/28

    EC8 design for infi ll irregularity in elevation

    Dissemination of information for training Lisbon 10-11 February 2011 9

    Eurocode 8: Design columns of storey where infills are reduced

    re a ve o over y ng s orey, o rema n e as c n s n s orey a ove

    reach their ultimate force resistance:

    Deficit in infill shear stren th in a store is com ensated b an

    increase in resistance of the frames (vertical) members there:

    In DC H frame or frame-equivalent dual buildings, seismic M, V, N in

    multiplied by: qVV EdRw /1

    VRw: total reduction of resistance of masonry walls in storey

    concerned w.r.to storey above,

    Ed

    seismic members of the storey (storey design shear).

    If < 1.1: magnification of seismic action effects may be omitted.

    Although not required for DC M frame or frame-equivalent dual

    buildings, the above are (weakly) recommended for them as well.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    10/28

    Asymmetry of infills in plan

    Dissemination of information for training Lisbon 10-11 February 2011 10

    Asymmetric distribution of infills in plan torsional

    response to translational horizontal components of

    seismic action: em ers on e s e w e ewer n s ex e s e

    are subjected to larger deformation demands & fail first.

    The increase in global lateral strength & stiffness due to

    interstorey drift demands in plan:

    The maximum member deformation demands for planwiseirregular infilling do not exceed peak demands anywhere

    in plan, in a similar structure w/o infills.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    11/28

    EC8 design against infill planwise asymmetry

    Dissemination of information for training Lisbon 10-11 February 2011 11

    Eurocode 8: doubles accidental eccentricity (from 5 to 10%) in

    the analysis, if the infills are planwise irregular.

    Doubling of accidental eccentricity: is not enough for severely

    irregular arrangement of infills in plan , Need sensitivity analysis of the effect of stiffness & position of infills

    (disregarding one out of 3-4 infill panels per planar frame, especially on the

    ex e s es .

    In-plane modelling of infills. Simplest modelling of solid panel (without openings):

    Two diagonal struts.

    Effect of openings: Reduction factors?

    The above are required for DC H frame

    or rame-equ va en ua u ngs an

    (weakly) recommended for DC M ones.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    12/28

    Adverse local effects on structural frame

    Dissemination of information for training Lisbon 10-11 February 2011 12

    Shear failure of weak

    columns due to interaction

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    13/28

    EC8 design against local effect of strong infills

    Dissemination of information for training Lisbon 10-11 February 2011 13

    Shear loading of the column by the infill strut force:

    Eurocode 8 for all columns: verify in shear a length lc=winf/cos, near the top &

    e o om o e co umn over w c e agona s ru orce o n may e

    applied, for the smaller of the two design shear forces:

    The horizontal component of the infill strut force, taken equal to thehorizontal shear strength of the panel (shear strength of the bed joints times

    the horizontal cross-sectional area of panel); or

    Capacity design shear: 2 M (: design value of column moment resistance),

    / lc(: contact length) Width of the strut (e.g.):

    4.0inf

    cos

    .

    Hw n

    1

    4

    ncc

    ww

    HIE

    uroco e : rac on ~ o pane agona , bn cos

    Columns in contact with infill all along only one side: Full clear height = criticalregion

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    14/28

    Adverse local effects on structural frame (contd)

    Dissemination of information for training Lisbon 10-11 February 2011 14

    Shear failures of short (captive) columns

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    15/28

    EC8 design of squat captive columns

    Dissemination of information for training Lisbon 10-11 February 2011 15

    Capacity-design calculation of design shear force, w/: clear len th of the column, l = len th of the column not in contact to the

    infills &

    plastic hinging assumed to take place at the column section at the

    .

    Transverse reinforcement required to resist the design

    shear force is placed not just along the clear length of the

    column, lcl, but also into the column part which is incontact to the infills (over length equal to the column

    depth, hc, within plane of infill).

    Entire length of the column is taken as critical region, withs rrups e a e as n co umn cr ca reg ons.

    Use diagonal reinforcement over length of column not in

    , .

    times the column depth.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    16/28

    Dissemination of information for training Lisbon 10-11 February 2011 16

    SECONDARY SEISMIC ELEMENTS IN

    EC8

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    17/28

    Secondary seismic elements

    Dissemination of information for training Lisbon 10-11 February 2011 17

    Contribution of secondary seismic elements to resistance & stiffness forseismic actions is discounted in design (& in linear analysis model, too).

    The designer is free to assign elements to this class of elements,if:

    Their total contribution to lateral stiffness 15% of that of the other rimar seismic elements

    The buildings regularity classification does not change.

    Secondary seismic elements:

    not subject to the geometric etc. restrictions of EC8

    not ULS-designed for any seismic force demands,

    , .

    But: they are required to remain elastic under the deformations imposed by

    the design seismic action (: qtimes their deformations from an elasticana ys s w e con r u on o secon ary e emen s neg ec e :

    Requirement hard to meet.

    Therefore consider as secondar seismic onl those elements which

    cannot be made to meet EC8 rules (e.g., if they are outside EC8s scope:prestressed girders, flat-slab frames, etc.)

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    18/28

    Design procedure if some elements are Secondary seismic

    Dissemination of information for training Lisbon 10-11 February 2011 18

    1. Carry out linear analysis for the design seismic action using two models :

    Model SP: including the contribution of all elements (secondary orprimary seismic) to lateral stiffness;

    Model P: neglecting the contribution of secondary seismic elements

    (e.g., introduce appropriate hinges at their connections to primary elements,

    so that the secondary ones have stiffness only against gravity loads).

    .

    check that it is 115% at every storey.

    3. Estimate the deformations of secondary seismic elements under the

    design seismic action as qtimes their deformations from Model SP, times

    the P/SP-ratio of interstorey drifts in 2 above.

    secondary seismic element (50% of uncracked gross stiffness) find their

    internal forces and check that they are in the elastic domain.

    forces from Model SP, times q, times the P/SP-ratio of interstorey drifts from 2above.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    19/28

    7-storey wall building with f lat-slab frames taken as secondary

    seismicDissemination of information for training Lisbon 10-11 February 2011 19

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    20/28

    Contribution of secondary elements to lateral stiffness 15% of that

    of primary elementsDissemination of information for training Lisbon 10-11 February 2011 20

    - ,of the diaphragm and taking the flat slab as an effective beam w/ width of

    2.5m at the interior of the plan or 1.25m at the perimeter:

    Total contribution of flat slab frames and of the walls in their weakdirection to lateral stiffness: 13.9% of that of the walls in their strong

    rec on.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    21/28

    Deformation-induced seismic action effects in secondary seismic

    columnsDissemination of information for training Lisbon 10-11 February 2011 21

    Elastic M & V in the secondary columns from elastic analysis

    under design seismic action, multiplied by q and divided by

    the fraction of the base shear taken by the primary seismicelements, i.e., multiplied times:

    q - . = . :

    V=139kN,

    End moments: 240kNm and 127kNm at top & bottom.

    Maximum M in any interior column:

    m, a e groun s orey.

    (V=141kN, at the ground storey).

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    22/28

    Check of secondary seismic columns for the deformation-

    induced seismic momentsDissemination of information for training Lisbon 10-11 February 2011 22

    . . . - . .

    Top storey axial loadN=205kN, givingMRdc,n=346kNm>MEc,n.

    Ground storeyN=1435kN,MRdc,1=795kNm>>MEc,1.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    23/28

    Check of secondary seismic columns for the deformation-induced

    seismic shearsDissemination of information for training Lisbon 10-11 February 2011 23

    Max. tie spacing per EC2:Max. tie spacing per EC2:

    - w . bL c c - .

    8mm-dia. perimeter hoop and diamond-shaped ties mid-side vertical bars, @

    165mm centres:

    - w . bL c c - .

    8mm-dia. perimeter hoop and diamond-shaped ties mid-side vertical bars, @

    165mm centres:w= + x . x = . .

    Shear resistance for shear compression per EC2:VRd,max=0.3x(1-35/250)x0.6x0.9x0.565x(35000/1.5)sin2=1269kN>VEc,1,

    w= + x . x = . .

    Shear resistance for shear compression per EC2:VRd,max=0.3x(1-35/250)x0.6x0.9x0.565x(35000/1.5)sin2=1269kN>>VEc,1,

    if cot=2.5 Shear resistance due to the ties per EC2:V =b z f cot+N h-x/H , with neutral axis de th, x= d, at the

    if cot=2.5 Shear resistance due to the ties per EC2:V =b z f cot+N h-x/H , with neutral axis de th, x= d, at the,

    moment resistance of the column.

    - Top storey:

    - = + -

    ,

    moment resistance of the column.

    - Top storey:

    - = + -,s . . . . . . .0.084x0.565)/2.65 = 616.5kN>VEc,n.

    - Ground storey:

    ,s . . . . . . .0.084x0.565)/2.65 = 616.5kN>>VEc,n.

    - Ground storey:

    - Rd,s . . . . . . . -

    0.249x0.565)/2.65 =822.5kN>VEc,1.

    - Rd,s . . . . . . . -

    0.249x0.565)/2.65 =822.5kN>>VEc,1.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    24/28

    Concrete diaphragms

    Dissemination of information for training Lisbon 10-11 February 2011 24

    ULS verification of RC diaphragms in DCH buildings: For irregular geometry or divided shapes of diaphragm in

    plan, recesses or re-entrances;

    If irregular distribution of masses and/or stiffnesses (set-

    -

    In basements with walls only in part of the perimeter or onlyin part of the ground floor area;

    At the interface with core and walls in core or wall structural

    systems.

    o e suc ap ragms as eep eam or p ane russor strut-and-tie model, on elastic supports.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    25/28

    Strut-and-Tie model of diaphragm for check of top floor slab

    Dissemination of information for training Lisbon 10-11 February 2011 25

    Deep beam comprising:

    a Tension chord centred alonline 1 (width =lw of walls W1)

    b Semi-circular compressionchord connecting ends oftension chord, apex nearcentre of orthogonal wall W2on line 2;

    c ose y space tens on t esparallel to horiz. componentseismic action, running from

    opposite to line 2, to collectthe in-plane loadqE=1.728kN/m

    2 of top floordue to design seismic action

    and transfer it to compressionchord.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    26/28

    Verification of tension ties (contd)

    Dissemination of information for training Lisbon 10-11 February 2011 26

    Longest tension ties collect in-plane loadqE =1.728kN/m2 along the full

    , x. ,section through the flat slab normal to hor. direction X should havereinforcement area at least

    dq

    EL

    x/f

    yd=1.1x1.728x25/(0.5/1.15)=110mm2/m

    the moment due to the quasi-permanent floor gravity load,Mg+2q.

    (d=1.1: overstrength factor per EC8 for the design of diaphragms).

    The reinforcement of the flat slab has been dimensioned for ULS inbending for the flat slab moments under the factored gravity loads, Md.

    e surp us o re n orcemen area over an a ove w a s necessary orULS resistance under Md: As=max[As,min; Md/(zfyd)]-Mg+2q/(zfyd), where

    z=0.11m the internal lever arm, Mg+2q= (8.2/14.7)Md andAs,min them n mum re n orcemen area n e a s a per .

    Critical location forAs: whereMd is minimum.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    27/28

    Verif ication of tension ties

    Dissemination of information for training Lisbon 10-11 February 2011 27

    MinimumMd along longest tension ties:

    Sagging moment at mid-distance between W2 and 1st row of interior columns

    parallel to W2 (Section 1-1)

    = - = 2 < 2. . . . . . . .

    Increase reinforcement of flat slab within its middle strips between W2 and

    the 1

    st

    parallel row of interior columns, and between any rows of interiorco umns, o + x . . = mm m.

    Potentially critical: tension ties heading towards the edge column next to W2

    Section 2-2 :A = 1-8.2/14.7 x47.3/ 0.11x0.5/1.15 =438mm2/m >110mm2/m

    Between edge columns and 1st

    parallel row of interior columns (Section 3-3):As=(1-8.2/14.7)x37.85/(0.11x0.5/1.15)=350mm

    2/m >110mm2/m.

  • 7/21/2019 s5 Ec8-Lisbon m Fardis

    28/28

    Check of tension chord between supports of the deep beam by walls W1.

    Dissemination of information for training Lisbon 10-11 February 2011 28

    Tension force in chord from moment equilibrium between:

    ,

    uniform in-plane load of 1.728kN/m2 and force reactions to it at walls W1.

    Internal lever arm in dee beam zL /2 and force in tension chord:(qELxLy

    2/8)/(Lx/2) =qELy2/4.

    Required steel area:As,t-chord=dqELy2/(4fyd)=1.1x1.728x25

    2/(4x0.5/1.15)=2 2, . . . - .

    Minimum design moment along chord is in the middle strip, giving surplusAs=(1-8.2/14.7)x10.2/(0.11x0.5/1.15)=94.4mm

    2/m