See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/337916476

The first step towards the performance-based design of wind- excited buildings:
A critical examination of ductility-demand, hysteretic energy, and control of
damage accumulation

Conference Paper · September 2019

CITATION READS
1 228

3 authors:

Matiyas A. Bezabeh G. T. Bitsuamlak

Rowan Williams Davies & Irwin Inc. (RWDI) The University of Western Ontario
27 PUBLICATIONS   239 CITATIONS    172 PUBLICATIONS   2,085 CITATIONS   

SEE PROFILE SEE PROFILE

Solomon Tesfamariam
University of British Columbia - Okanagan
265 PUBLICATIONS   4,409 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Sustainable Cities and Society: Special issue on “Urban resiliency to extreme climate" View project

Force- and Performance-Based Seismic Design Guidelines for Timber-Steel Hybrid Structures View project

All content following this page was uploaded by Matiyas A. Bezabeh on 13 December 2019.

The user has requested enhancement of the downloaded file.

The 15th International Conference on Wind Engineering
Beijing, China; September 1-6, 2019

The first step towards the performance-based design of wind-

excited buildings: A critical examination of ductility-demand,
hysteretic energy, and control of damage accumulation
Matiyas Bezabeh a, c, Girma Bitsuamlak a, Solomon Tesfamariam c
a
Department of Civil and Environmental Engineering, Western University, Canada
c
School of Engineering, The University of British Columbia, Okanagan Campus, Canada

ABSTRACT: The lateral strength and stiffness requirements due to wind loads usually govern the de-
sign of tall buildings. The building codes in the U.S.A, Canada, and Europe recognize the first signifi-
cant yield point (i.e., linear-elastic design) as an ultimate limit state, which could result in conservative
and uneconomical design of the tall buildings. Thus, classical linear-elastic design arguments should
be re-examined with consideration of performance-based design approaches, innovative technologies,
and materials. In this paper, we critically examined the ductility-demand, the effect of wind duration,
hysteretic energy, and the rate of damage accumulation to introduce ductility-based design in wind
engineering. For this purpose, nonlinear time history analyses of bilinear and self-centering single-mass
oscillators subjected to artificially generated wind load time histories are conducted. Results of the
analyses reveal that, for the bilinear systems designed considering higher ductility-demand, damage
accumulation could trigger the failure of structural systems. Therefore, for bilinear systems, we pro-
posed a new ultimate limit state (controlled inelasticity-limit state) with explicit consideration of both
ductility-demand and rate of damage accumulation. On the other hand, the results of the self-centering
systems indicate their efficiency in controlling the damage accumulation under wind loads. Hence, with
the use of self-centering systems, economic structures can be designed for the ductility capacity of 5 or
more.

KEYWORDS: Performance-based wind design; Ductility-demand, Damage accumulation, Bilinear

systems, Self-centering systems; Strength reduction factor

1 INTRODUCTION

Tall buildings are prone to wind vulnerability due to complex bluff geometry, reduced weight, stiffness,
and damping properties. Consequently, the lateral strength and stiffness requirements due to wind load
usually govern the design of tall buildings. The building codes in the U.S.A, Canada, and Europe rec-
ognize the first significant yield point (i.e. linear-elastic design) as an ultimate limit state. The main
reasons behind this provision of the linear-elastic design of tall buildings are the damage accumulation
due to the longer duration of wind storms and unsymmetrical yielding (cyclic excursions to the plastic
range are usually in one direction due to the mean wind load). Despite most of the arguments behind
the linear-elastic design of tall buildings may not always strictly hold, the current design practices ig-
nore the plastic capacity of structural systems in the non-linear range. Therefore, classical linear-elastic
design arguments should be re-examined with consideration of performance-based engineering (PBE)
approaches, innovative technologies, and materials. Within literature, several self-centering systems
have been developed to control permanent set under dynamic loads by utilizing Shape Memory Alloys
(SMA) and post-tensioning technology (e.g. Priestley 1991, Christopoulos et al. 2002a). In wind engi-
neering, the studies towards the Performance-Based Wind Engineering (PBWE) can be categorized
into two broad research directions. The first group of studies focused on the response of simple yielding
structural systems subjected to turbulent wind load (e.g., Vickery 1970, Chen and Davenport 2000,
Hong 2004). Overall, these studies highlighted the importance of considering the damage accumulation
and ductility-capacity in wind design. However, to the best of our knowledge, limiting or eliminating
the wind-induced permanent set (damage accumulation) have not been researched yet. The second re-
search direction has focused on the development of a comprehensive PBWE framework, to name a
few, Petrini et al. 2009, Mohammadi 2016, Spence et al. 2016; Bezabeh et al. 2018b). Most of the
existing PBWE frameworks focused on extending the earthquake engineering methodology into wind
engineering. These formulations are essential in developing the roadmap towards complete PBWE.
However, they may suffer from the lack of accuracy due to the inherent differences between earthquake
and wind loads, primarily in load duration, frequency content, and damage mechanisms. In our opinion,
the first step in PBWE, therefore, should be understanding the response of structures in the non-linear
range and control damage accumulation. Thus, in this paper, we critically examined the ductility-de-
mand, the effect of wind duration, hysteretic energy, and the rate of damage accumulation using non-
linear time history analysis of Single-Degree-of-Freedom (SDOF) systems. We proposed the use of
self-centering systems and proved their capability in controlling the possible damage accumulation in
structural systems subjected to long-duration wind loads.

2 NONLINEAR DYNAMIC ANALYSIS OF SDOF SYSTEMS UNDER WIND LOAD

2.1 Hysteretic models and basic terminologies
In this study, we considered two idealized bilinear (Figure 1a) and self-centering (flag shaped, Figure
1b) hysteretic force-deformation models. The bilinear model represents the typical response of steel
moment frame structures. As shown in Figure 1, the bilinear and self-centering systems are defined by
initial stiffness (k), yield strength demand (Fy), post stiffness yielding ratio (α), and energy dissipation
capacity (β). For systems subjected to loads beyond the elastic strength demand, the ductility-demand,
hysteretic energy (HE), force modification factor (M), the rate of damage accumulation per crossing
(d), are the most important performance indicators. Ductility-demand (μ=xmax/xy) and force modifica-
tion factor (M = Fm/Fy) depends on the mean wind force (Fm), yield force (Fy), maximum displacement
(xmax) and yield displacement (xy).

Figure 1: Hysteretic force-deformation models: a) bilinear system; b) self-centering system; c) Compassion of the simulated
and target spectrum
2.2 The equation of motion and generation of artificial fluctuating wind speed time series
The equation of motion of the SDOF system under wind load is given in Equation 1. The right -hand
side of the equation is established using a quasi-steady approach (Davenport 1961).
1
𝑚𝑢̈ + 𝑐𝑢̇ + 𝑓𝑟 = 𝜌𝐶𝑑 𝐴𝑈 2 + 𝜌𝐶𝑑 𝐴𝑈𝑢(𝑡) (1)
2
where 𝑚𝑢̈ is the inertial force of the system, 𝑐𝑢̇ is the damping force, fr is the restoring force, ρ is air
density, A is exposed area, Cd is the drag coefficient, U is the mean wind speed, and u(t) is the fluctu-
ating wind speed. In this paper, u(t) is simulated by targeting the Davenport spectrum (Davenport 1961)
using the spectral representation method (Shinozuka 1972). Figure 1c shows a good agreement between
the target Davenport spectrum with the mean power spectra of three artificially generated wind time
data. The response of SDOF systems is obtained by solving Equation 1 using the Newmark-Beta
algorithm (Chopra 2000) by assuming stiffness proportional Rayleigh damping.
The 15th International Conference on Wind Engineering
Beijing, China; September 1-6, 2019

2.3 Response comparison between bilinear and self-centering SDOF systems

In this section, the nonlinear responses of bilinear and self-centering systems are compared. For the
dynamic analysis, the SDOF systems are characterized by fundamental frequency fn = 0.3Hz, α = 3%,
damping ratio ξ = 1%, and β = 0.6. To generate u(t) we assumed U = 30m/s, turbulence intensity (Iu)
= 18.75%, height the SDOF systems z = 10m, wind duration T = 600 sec. For brevity, we only discuss
the results for wind force reduction factors M = 0.3 and 0.6. It is worthwhile mentioning that M controls
the level of wind excitation with respect to the system capacity. Figure 2 compares the normalized
force-deformation curves. As expected, the ductility-demand (μ) increases with M. The hysteretic re-
sponses of the bilinear system (Figures 2a and 2c) dominated by elastic-unloading and elastic-reloading
where the system couldn’t form stable hysteretic loops. On the contrary, Figures (2b and 2d) show the
ability of self-centering systems in controlling the damage accumulation and forming more stable cyclic
loops with distinctive hysteretic energy (HE) dissipative region. Figure 3 provides additional evidence
on the capability of self-centering systems in controlling damage accumulation that is marked by neg-
ligible residual displacement at the end of the excitation.

a) b) c) d)
Figure 2: Normalized hysteretic responses: a) blinear system M = 0.3; b) self-centering system M = 0.3; c) blinear
system M = 0.6; d) self-centering system M = 0.6

a) b)
Figure 3: Variation of ductility-demand with time: a) M = 0.3; b) M = 0.6

However, Figure 3b depicts the unbounded growth of damage with time in bilinear systems resulting
in very large residual displacement that could either trigger collapse or need uneconomical repair after
the excitation. To investigate the effect of modeling parameters (ξ, β, α) and wind load characteristics
(U, T, Iu), we conducted a parametric study using 100 artificially generated wind load time histories
and nine levels of M. However, due to space limitation, we only present results of parametric study of
post-stiffness yielding ratio (α) and wind duration (T) with respect to ductility-demand. Figures 4a-4d
display the mean peak and residual ductility-demands with for α = 1%, 3%, 5%, 10%, 40%. As ex-
pected the increase in α reduces both the peak- and residual-ductility demands. Moreover, for higher
excitation levels (M > 0.35), the effect of α is more pronounced. Figures 4c and 4d show the lesser
influence of α on the ductility-demand of self-centering systems. Figures 4e and 4f display the relation-
ship between the wind duration and ductility-demands. For bilinear systems, both peak and residual-
demands are only sensitive to wind duration for excitation levels M between 0.2-0.6. Interestingly, self-
centering systems are less sensitive to wind duration (Figures 4g and 4h). This is due to the higher and
stable energy dissipation (HE) due to yielding and re-centering. Moreover, for all considered cases, the
residual ductility-demands of self-centering systems are less than one, which again shows their capa-
bility of controlling permanent displacements.
 Figure 4: Variation of peak- and residual-ductility demands with post-stiffness ratio and wind duration; bilinear systems (a,
b, e, f); self-centering systems (c, d, g, h)

In general, considering damage accumulation, it is shown that for bilinear systems designed considering
a ductility-capacity of 2 and α=5%. However, with the use of self-centering systems, structural systems
and connections can be designed for ductility of 5 or more.

3 CONCLUSIONS
In this paper, to introduce ductility-based design in wind engineering, we conducted nonlinear time
history analyses of SDOF bilinear and self-centering systems. Results indicated the capability of self-
centering systems in controlling the damage accumulation due to long-duration wind loads. On the
contrary, conventional bilinear systems subjected wind loads beyond their yield strength experienced
a significant amount of residual displacement that could trigger collapse or irreparable damage. Hence,
caution should be exercised during the design of bilinear systems targeting ductility capacity more than
2. Given the presented findings, with the use of self-centering systems, economic structures can be
designed for the ductility capacity of 5 or more. To validate the obtained results by including aeroelastic
effects, the authors are currently carrying out the world’s first nonlinear aeroelastic test of a tall building
using SMA springs as stiffness controllers.

4 REFERENCE
1. Bezabeh, MA., Bitsuamlak, GT., Popovski, M., & Tesfamariam, S. (2018). Probabilistic serviceability-performance assessment of tall mass-timber
buildings subjected to stochastic wind loads: Part II-structural reliability analysis, J.Wind Eng. Ind. Aerodyn., 181, 112-125.
2. Chen D, Davenport AG., Vulnerability of tall buildings in typhoons. InProc. Int. Conf. Adv. Struct. Dyn 2000.
3. Chopra, A.K., Dynamics of Structures, Prentice Hall, New Jersey, 2000.
4. Christopoulos C, Filiatrault A, Folz B., Seismic response of self-centering hysteretic SDOF systems. Earthquake Eng. Struct. Dyn. 2002 May;
31(5):1131-50.
5. Davenport AG., The spectrum of horizontal gustiness near the ground in high winds. Q. J. R. Meteorolog. Soc. 1961; 87(372).
6. Hong HP., Accumulation of wind induced damage on bilinear SDOF systems. Wind and Structures. 2004;7(3):145-58.
7. Mohammadi A., Wind Performance Based Design for High-Rise Buildings. Ph.D. Dissertation, FIU, U.S.A.
8. Petrini F, Ciampoli M, Augusti G., A probabilistic framework for performance-based wind engineering. In Fifth EACWE, Italy 2009.
9. Priestley MN. Overview of PRESSS research program. PCI journal. 1991 Jul; 36(4):50-7.
10. Spence SM, Chuang WC, Tabbuso P, Bernardini E, Kareem A, Palizzolo L, Pirrotta A., Performance-based engineering of wind- excited structures:
A general methodology. In Geotechnical and Structural Engineering Congress 2016 (pp. 1269-1282).
11. Shinozuka M., Monte Carlo solution of structural dynamics. Computers & Structures. 1972 Jan 1;2(5-6):855-74.
12. Vickery BJ. Wind action on simple yielding structures. Journal of the Engineering Mechanics Division. 1970 Apr; 96(2):107-20.

View publication stats