GNGTS 2017 SeSSione 2.3

443

seIsmIC hAzArd mItIGAtIon of multI-storey buIldInGs VIA VIbrAtIon Control systems And enhAnCed GlAss CurtAIns C. Bedon, C. AmadioDepartment of Engineering and Architecture, University of Trieste, Italy

Introduction.. Glass facades are widely used in building structures, due to a series ofss facades are widely used in building structures, due to a series of facades are widely used in building structures, due to a series of aesthetic, thermal, lightening aspects. Wide transparent surfaces are realized in commercial,Wide transparent surfaces are realized in commercial,ide transparent surfaces are realized in commercial,realized in commercial,in commercial, residential but also strategic buildings (airports, museums, offices, etc.). From a structural pointbut also strategic buildings (airports, museums, offices, etc.). From a structural pointstrategic buildings (airports, museums, offices, etc.). From a structural point (airports, museums, offices, etc.). From a structural pointairports, museums, offices, etc.). From a structural point.). From a structural point. From a structural point From a structural pointFrom a structural point of view, however, these envelopes represent a critical component for buildings, due to the brittlethese envelopes represent a critical component for buildings, due to the brittleenvelopes represent a critical component for buildings, due to the brittlebrittlerittle behaviour and limited tensile resistance of glass, as well as to possible criticalities derivingpossible criticalities deriving from connections, detailing, etc., hence requiring specificconnections, detailing, etc., hence requiring specifics, detailing, etc., hence requiring specific detailing, etc., hence requiring specific, etc., hence requiring specific etc., hence requiring specific fail-safe design concepts (Haldimann et al., 2008; Feldmann et al., 2014). The estimation of the vulnerability and actual dynamicThe estimation of the vulnerability and actual dynamiche estimation of the vulnerability and actual dynamicand actual dynamic behaviour of glazing systems under exceptional loads (seismic events, explosions, fire, naturalof glazing systems under exceptional loads (seismic events, explosions, fire, naturalexceptional loads (seismic events, explosions, fire, naturalloads (seismic events, explosions, fire, natural (seismic events, explosions, fire, natural hazards, see Figs.1a to 1c - including their interaction with the building they belong - is currentlyincluding their interaction with the building they belong - is currentlyir interaction with the building they belong - is currently interaction with the building they belong - is currentlywith the building they belong - is currently the building they belong - is currently they belong - is currently currently an open topic, attracting the attention of several studies (Behropen topic, attracting the attention of several studies (Behr, attracting the attention of several studies (Behr attracting the attention of several studies (Behr (Behr et al., 1995; Zhang et al., 2013; Mackalicka et al., 2016; Behr, 2009; Masters et al., 2010; Bedon and Amadio, 2017a; Larcher et al., 2016). However, further efforts are still required. In this paper, the seismic performance. However, further efforts are still required. In this paper, the seismic performanceHowever, further efforts are still required. In this paper, the seismic performancefurther efforts are still required. In this paper, the seismic performanceefforts are still required. In this paper, the seismic performance. In this paper, the seismic performance In this paper, the seismic performanceIn this paper, the seismic performanceseismic performance of glass curtain walls is investigated. The feasibility and potential of special mechanicalf glass curtain walls is investigated. The feasibility and potential of special mechanical is investigated. The feasibility and potential of special mechanical. The feasibility and potential of special mechanical The feasibility and potential of special mechanicalhe feasibility and potential of special mechanical connectors interposed at the interface between a given multi-storey building and the enclosinginterposed at the interface between a given multi-storey building and the enclosingat the interface between a given multi-storey building and the enclosing curtain are numerically investigated in ABAQUS (2017). The final result, as shown, consists innumerically investigated in ABAQUS (2017). The final result, as shown, consists ininvestigated in ABAQUS (2017). The final result, as shown, consists inin ABAQUS (2017). The final result, as shown, consists in. The final result, as shown, consists in as shown, consists in consists in a full 3D assembly in which the facade works as a passive control system for the building, in the3D assembly in which the facade works as a passive control system for the building, in thein which the facade works as a passive control system for the building, in thes as a passive control system for the building, in the as a passive control system for the building, in thebuilding, in the, in the form of a distributed Tuned-Mass Damper (TMD), with enhanced global and local structurala distributed Tuned-Mass Damper (TMD), with enhanced global and local structuraldistributed Tuned-Mass Damper (TMD), with enhanced global and local structural, with enhanced global and local structural benefits (Bedon and Amadio, 2017)..

Fig. 1 - (a)-(b)-(c): damage in glass facades under seismic events or explosions, with (d) example of UGCW installation process, (e) detail of the typical glass-to-frame connection and (f) traditional rigid bracket.

444

GNGTS 2017 SeSSione 2.3

Design philosophy and regulations. Glass facades are usually designed as constructional components aimed to resist ordinary loads only (self-weight, wind), while enhanced curtains are properly calculated and generally over-dimensioned for special structures only. There, appropriate safety levels are mandatory, especially if exceptional loads are expected. Glass fragments represent in fact a critical issue for people, hence cracking of facade panes should be prevented. For the seismic design of glass curtains, general EU standards for buildings can be taken into account (i.e. EC8 (2004)). There, however, secondary components only are considered and no regulations account for the curtain wall typology, or for anchoring systems, materials, etc. The building as a whole is only required satisfy specific inter-storey drift values.

Distributed TMD concept. In this paper, unitized glass curtain walls (UCGWs) consisting of pre-assembled modular units, with insulating glass panels sealed to metal frames and fixed to the main building via rigid metal brackets (Figs.1d to 1f), are investigated under seismic events, with careful consideration for their role in the dynamic response of the buildings they belong. A novel design concept, consisting in a distributed TMD, is then numerically assessed. Based on such a design concept, glass curtains are actively involved in the dynamic performance of the enclosed building, with expected marked global and local benefits. The proposal takes advantage from the consolidated use of passive control systems in civil engineering applications, and specifically from the wide use in buildings of TMDs to wind and seismic loads effects (Lee et al., 2006; Hoang et al., 2006; Moon, 2016; Mohebbi and Joghataie, 2012), see Fig. 2a.

Preliminary design. Assuming that the single UGCW modular unit is connected to the primary structural system by means of four corner brackets, special mechanical connectors are proposed to replace traditional restraints, so that the feasibility and potential of the distributed TMD concept could be properly assessed. The typical control system can take the form of a viscoelastic (VE) solid damper, consisting of two metal plates and a rubber layer (thickness hd and surface area Ad), see Fig. 2b and Bedon and Amadio (2017a, 2017b). The base plate is bolted to the inter-storey floor of the primary structure, while the UGCW frame is rigidly connected to a sliding plate, so to enable possible crushing and rotations of the VE layer when the UGCW panels are subjected to external pressures. A preliminary estimation of the mechanical properties of such a device (stiffness Kd and damping term cd) can be carried out based on classical theories (Chopra, 2011), hence:

and (1) (2)with Mglass the mass of a single glass panel, ω1,glass the operating frequency of the UGCW module, ξ the damping coefficient of rubber and ccr its critical value. For design purposes, a high damping rubber was taken into account. The fundamental period T1,glass of the UGCW module, in addition, was set equal to the building period T1. Possible tearing of the VE layer was then prevented, through the parametric numerical investigation, by limiting its shear deformations:

in [mm] (3)

Case study building. An UGCW spanning from floor-to-floor was taken into account for the FE study (2.90 m×1.6 m the size of each module). The UGCW was assumed belonging to a 4-storey, steel frame building with residential destination (category of use ‘A’, based on EC1 2004), located in an earthquake-prone region of Italy, see Figs. 2c and 2d. Overall base dimensions of 10 m×20 m were taken into account, with 12 m the total height, 3×2 bays in the longitudinal and transversal directions. Inter-storey floors consisting of in-plane rigid, steel-concrete composite slabs were then considered. All the steel members, S275 grade (fy=275 MPa and fu=360 MPa the yielding and collapse stresses), were supposed designed in accordance with EC3 (2004) provisions, based on permanent and accidental loads at the Ultimate (ULS) and Service (SLS) Limit States.

Global dynamic effect of dissipative UGCWs. The overall dynamic effects due to UGCWs with VE fasteners was first considered. To this aim, eigenvalue FE numerical investigation were

GNGTS 2017 SeSSione 2.3

445

carried out in ABAQUS, on the full 3D steel frame inclusive of the glazing envelope (Fig. 2c). The numerical modelling of the examined structural system was carried out in accordance with past numerical efforts, as for example reported in (Bedon and Amadio, 2017a, 2017b), and generally consisted of computationally efficient but accurate FE models composed of beam and shell elements, as well as properly calibrated mechanical constraints able to reproduce the actual performance of a single constructional member, as well as its interaction with the other components.

For each VE device, Kd was then estimated with Eq. (1) as a function of T1=0.326 s, as numerically calculated for the 3D frame with a rigidly restrained UGCW. A non-dimensional magnifying coefficient RK was also defined, through a FE parametric study (with 1 ≤ RK ≤ 50), to assess the input mechanical features effects:

(4)By changing the RK reference value, no tangible variations were found for the fundamental

vibration shapes of the building (Fig. 3a). Interesting variations were observed in terms of vibration periods, for the examined configurations (Fig. 3b). A value RK≥50, in particular, proved to coincide with almost fully rigid brackets, hence with null benefits. RK values ≤2, conversely, gave evidence of local vibration modes of the UGCW, compared to the steel frame, hence suggesting to consider RK values in the range ≈2-30.

Seismic analyses. The dynamic performance of the case study building was then properly investigated. A set of 7 seismic records (two-component acceleration data, consistent with EC8 provisions for type ‘A’ soil, T1 topographic category, 0.35g PGA and 50 years of nominal life (ULS)) was imposed at the building base, through dynamic nonlinear analyses (see Fig. 3c).

In Fig. 3d, the top drift and the VE sliding (absolute maximum envelopes from the set of seismic records) are proposed, as a function of RK. Worth of interest is the effect of dissipative

Fig. 2 - a) Distributed TMD concept, with b) cross-sectional view of a VE solid damper. c) Case study building and d) reference FE model for the UGCW modular unit (ABAQUS).

446

GNGTS 2017 SeSSione 2.3

UGCWs, both in global and local terms. As far as Kd increases, cd further increases, see Eqs. (1) and (2). However, limited sliding only can occur in VE devices, with minimum damping contribution and hence benefits for the structural components. Based on Fig. 3d, an optimal balance of local and global performances for the examined case study was found to lie in the range of RK≈3-4 values.

In global terms (see Fig. 3e), markedly reduced top displacements of the 3D frame were noticed, in presence of VE connectors. At the local level - for properly designed VE devices - maximum stresses in the UGCW components can be also highly minimized. Even the imposed seismic records were found to do not lead glass panes to failure, the RK=4 configuration lead up to ≈-50% the stresses of the traditionally restrained UGCW. In addition, the flexibility of UGCW supports proved to beneficially affect the distribution of stresses in glass panels. Local peaks of stresses close to the glass supports and edges were in fact generally avoided, due to VE fasteners, see Fig. 3f, with obvious benefits for the overall dynamic performance of the full structural system. These global and local findings derive from the combined flexibility

Fig. 3 - Dynamic performance the 3D building with dissipative UGCWs (ABAQUS): a) fundamental mode and b) variation of vibration periods with RK; c) records for nonlinear dynamic seismic analyses, with d) measured drifts, e) maximum stresses in glass and f) related distribution (legend values in Pa).

GNGTS 2017 SeSSione 2.3

447

and damping properties of VE devices, since part of the incoming seismic input energy can be preliminary dissipated. As a result, the UGCW and the steel frame are potentially subjected to a reduced seismic impulse, compared to the same building with fully rigid brackets.

Conclusions. In this paper, the feasibility and potential of special mechanical joints interposed at the interface between a given multi-storey primary building and a traditional unitized glazing curtain wall (UGCW) have been investigated via accurate Finite-Element (FE) numerical models, under seismic events.

Based on properly designed fasteners able to introduce additional flexibility and damping capacities in the traditional building, the maximum effects and benefits of such connectors have been shortly emphasized via a case study of technical interest, both in terms of global performances as well as local and component behaviour for the building object of investigation, giving evidence of the potential of UGCWs acting as distributed TMDs for multi-storey buildings under hazards. It is hence expected, based on current research outcomes partly emphasized in the paper, that the same design concept could be further calibrated and optimized, as well as that related design criteria could be fully implemented towards the definition of practical tools for designers.ReferencesM Haldimann, A Luible, M Overend (2008). Structural use of glass. IABSE, Zurich, Switzerland, ISBN 978-3-85748-

119-2M Feldmann, R Kasper, B Abeln, P Cruz, J Belis, J Beyer, et al (2014). Guidance for European Structural design of

glass components - Support to the implementation, harmonization and further development of the Eurocodes. Report EUR 26439, EU JRC–IPSC, doi: 10.2788/5523

RA Behr, A Belarbi, AT Brown (1995). Seismic Performance of Architectural Glass in a Storefront Wall System. Earthquake Spectra. Vol. 11. No. 3. (1995): 367–391

X Zhang, H Hao, G Ma (2013). Parametric study of laminated glass window response to blast loads. Engineering Structures, 56: 1707-1717

K Machalická, M Charvátová, M Eliášová, P Kuklík (2016). The behaviour of fire resistant glass under fire. Structures and Architecture - Beyond their Limits, pp.991-997, ISBN 978-1-138-02651-3

RA Behr (2009). Architectural glass to resist seismic and extreme climatic events. Woodhead Publishing, ISBN 9781845693695

FJ Masters, KR Gurley, N Shah, G Fernandez (2010). The vulnerability of residential window glass to lightweight wind-borne debris. Engineering Structures, 32: 911-921

C Bedon, C Amadio (2017a). Passive control systems for the blast enhancement of glazing curtain walls under explosive loads. The Open Civil Engineering Journal, 11 (Suppl-1, M8): 396-419

M Larcher, M Arrigoni, C Bedon, A van Doormaal, C Haberacker, G Hüsken, O Millon, A Saarenheimo, G Solomos, L Thamie, G Valsamos, A Williams, A Stolz (2016). Design of Blast-Loaded Glazing Windows and Facades: A Review of Essential Requirements towards Standardization. Advances in Civil Engineering, http://dx.doi.org/10.1155/2016/2604232

ABAQUS(2017). Dassault Systèmes. ABAQUS v. 6.14, Providence, RIC Bedon, C Amadio (2017b). Enhancement of the seismic performance of multi-storey buildings by means of

dissipative glazing curtain walls. Engineering Structures, under reviewEN 1998-1-1 (2004). Eurocode 8 – Design of structures for earthquake resistance – Part 1: General rules, seismic

actions and rules for buildings. CEN, Brussels, BelgiumC-L Lee, Y-T Chen, L-L Chung, Y-P Wang (2006). Optimal design theories and applications of tuned mass dampers.

Engineering Structures, 28: 43–53N Hoang, Y Fujino, P Warnitchai (2008). Optimal tuned mass damper for seismic applications and practical design

formulas. Engineering Structures, 30(3): 707–715KS Moon (2016). Integrated damping system for tall buildings: tuned mass damper/double skin façade damping

interaction system. The Structural Design of Tall and Special Buildings, 25(5): 232-244M Mohebbi, A Joghataie (2012). Designing optimal tuned mass dampers for nonlinear frames by distributed genetic

algorithms. Structural Design of Tall and Special Buildings, 21(1): 57-76AK Chopra (2011). Dynamics of Structures - Theory and Applications to Earthquake Engineering, 4th ed., Prentice

HallEN 1991-1-1 (2004): Eurocode 1 – Actions on structures – Part 1-1: General Actions. CEN, Brussels, BelgiumEN 1993-1-1 (2004): Eurocode 3 – Design of steel structures – Part 1-1: General rules and rules for buildings. CEN,

Brussels, Belgium