Nonstructural Component Simulator University at Buffalo - MCEER ...

University at Buffalo 

Department of Civil, Structural and Environmental Engineering 

By: Andre Filiatrault, Gilberto Mosqueda, Andrei Reinhorn, 

Mark Pitman, Scot Weinreber & Rodrigo Retamales 

e-mail: rr62@buffalo.edu 

Nonstructural Component Simulator 


Preliminary Report Seismic Performance Assessment of a 

Full-Scale Hospital Emergency Room 

Introduction 

This preliminary report presents a description of the experimental tests performed to demonstrate the 

actual testing capabilities of the Nonstructural Component Simulator (UB-NCS), the latest testing 

equipment upgrade at the Structural Engineering and Earthquake Simulation Laboratory (SEESL) at the 

University at Buffalo. A detailed description of the UB-NCS and its capabilities is out of scope of this 

report and it can be found in references [1] and [2]. 

Section 1 of this report presents a general overview of the project, describing its main objectives and 

scopes. Section 2 presents a description of the testing specimen. Section 3 describes the testing protocol 

used for the seismic performance assessment of the emergency room. Section 4 summarizes the tests 

program conducted and Section 5 presents some preliminary experimental results. Finally, Section 6 

presents the preliminary conclusions of this research. 

1. Objectives and scopes 

The primary objectives of the test series are: 

• Evaluate the actual capabilities and limitations of the UB-NCS in assessing the seismic performance 

of full scale nonstructural components and systems. 

• Evaluate individual seismic performance and seismic interaction among nonstructural components 

and medical equipment typically found in a health care facility’s emergency room. 

• Assess the efficiency of a new testing protocol specifically developed to assess seismic interactions 

between acceleration and displacement sensitive nonstructural components and systems. 

Among the secondary objectives of the demonstration project are: 

• Evaluate the fidelity of the UB-NCS for imposing the desired full-scale floor motions on full-scale 

specimens. 

• Evaluate the efficiency of the testing protocol in evaluating the seismic performance of nonstructural 

components, systems and contents that may be sensitive to the accelerations and/or displacements 

imposed by the motion of buildings during strong earthquakes. 

The scopes of the present work include: 

• Taking into account that one of the main objectives of this test series is to demonstrate the actual 

testing capabilities of the UB-NCS, some constructive details have been disregarded. In particular, 

some details currently used in field for the seismic protection of several of the components included 

in this demonstration project have been intentionally omitted. 

• The seismic performance of the testing specimen has been assessed considering the floor motions 

obtained from a suitable testing protocol and the simulated seismic response of a four story building. 

UB-NCS Preliminary Results ER Tests - 2/4/2008 Page 1 of 33






Recorded building floor motions are not directly used in the test series but their properties are implicit 

in the properties of the testing protocol considered. 

2. General description of mock Emergency Room (ER) 

The test series considers the evaluation of the seismic performance of a full scale mock hospital’s 

emergency room replica under full scale floor motions. The room is enclosed by steel studded gypsum 

partition walls constructed following standard hospital construction techniques. The nonstructural 

partition wall model is based on a similar specimen tested by Restrepo and Lang [3, 4] at University of 

California-San Diego, using the pseudo-static loading protocol developed by Krawinkler et al. [5]. Slight 

modifications were made to the original design in order to fit the specimen within the UB-NCS [6]. The 

walls were constructed between concrete slabs attached to the UB-NCS as shown in Figure 1. The room is 

approximately 4.4 m (14’-6”) in length, 3.2 m (10’-7”) in width, and 3.85 m (12’-6”) in height, Figure 2. 

Figure 3 shows the elevations with dimensions of the partition walls. 

14'-5 1 2 " 

6'-0 15 16 " 71 2 " 3'-8" 2'-1" 2' 

2'-4" 

4'-7 5 8 " 1'-4" 4'-7 5 8 " 

Utility Cutout 

(The top of the square 

utility cutout is 12' - 0" from 

the floor) 

5/8" Gypsum board 

2'-0 1 2 " 

1'-6" 3'-9 5 8 " 5'-35 8 " 

10'-7 1 4 " 

10'-11 1 2 " 1'-6" 2' 

Figure 1. Photograph of UB-NCS with 

concrete slabs 

Figure 2. Geometry of specimen for the UB-NCS 

demonstration project 

Figure 4 and Figure 5 show isometric views of the steel stud framing system and the finished specimen, 

respectively. The simulated emergency room was furnished with medical equipment critical for 

inpatients’ life support following a seismic event. Medical equipment found in other hospital’s critical 

services [7] was also considered for demonstration purposes. The nonstructural components and systems, 

along with the medical equipment included in the demonstration project, are summarized in Table 1. 

For fragility assessment purposes, the hospital room was subjected to increasing levels of shaking 

including design and maximum considered earthquake level motions. The floor displacement histories 

considered in testing were obtained from the simulated response of an existing medical facility located in 

the San Fernando Valley, in Southern California, and from a testing protocol specially developed to take 

full advantage of the UB-NCS testing capabilities and to impose the seismic demands expected at the 

upper levels of multistory buildings. 


8' 






14'-6 1 8 " 14'-11 4 " 

12'-7 1 4 " 

12'-7 1 4 " 

Ceiling Level 

4 8 16 16 10'-3" 

11'-0 3 " 3'-53 " 2'-113 " 1'-87 " 

Exterior View North Wall 

Interior View North Wall 

Scale: - 

Scale: - 

14'-6 1 8 " 14'-61 8 " 

Ceiling Level 

12'-7 1 4 " 12'-71 4 " 

9'-4" 3'-3 1 4 " 9'-4" 3'-31 4 " 

4'-1 1 16 " 3'-8" 6'-91 8 " 

5'-8 11 16 " 79 16 " 3'-8" 1'-113 16 " 2'-113 16 " 

Exterior View South Wall 

Scale: - 

Interior View South Wall 

Scale: - 

7 1 4 " 

12'-7 1 4 " 12'-71 4 " 

1'-4" 

2'-11 1 4 " 1'-4" 6'-51 4 " 

Ceiling Level 

10'-8" 

10'-8 1 2 " 9'-103 4 " 

Exterior View East Wall 

Interior View East Wall 

Scale: - 

Scale: - 

10'-8 1 2 " 9'-103 4 " 

Ceiling Level 

12'-7 1 4 " 12'-71 4 " 

9'-4" 3'-3 1 4 " 9'-4" 3'-31 4 " 

1'-8 7 16 " 8'-95 8 " 4'-113 8 " 3'-53 8 " 1'-6" 

Exterior View West Wall 

Scale: - 

Interior View West Wall 

Scale: - 

Figure 3. Elevations partition walls 







Figure 4. Isometric view steel framing 

Figure 5. Isometric view finished specimen 

Table 1. Nonstructural components and medical equipment included in specimen 

Nonstructural components and equipment 

1 Steel stud framed gypsum partition walls 

2 Suspended ceiling and suspended light 

3 Flooring 

4 Sprinkler system 

5 Medical gas pipes with in-wall outlets 

6 Wall mounted patient monitors 

7 Freestanding poles with IV infusion pumps 

8 Ceiling mounted surgical lamp 

9 Operating room video equipment rack on casters 

10 Gurney 

11 Cart on casters 

12 Dummy 

Detailed descriptions of the most important items listed in Table 1 are presented in following subsections. 

Figure 6 shows a general layout of the medical equipment included in the experiment. 







1'-6" 

4' 

13 

20 

Monitor 4 

13x16x7 in 

@ h=60 in 

1'-6 1 2 " 

15 15 

22 

Monitor 3 

15x18x18in 

@ h=52 in 

22 

Monitor 2 

15x18x18in 

@ h=52 in 

28 

78 

Gurney 

(h=40 in) 

Pole 2Ø26 

h=72in 

4' 

15 

28 

Pole 1 

h=72in 

Ø26 

22 

Monitor 1 

15x18x18in 

@ h=52 in 

Cabinet 

26x28x76 

26 

5' 

2.1. Partition walls 

Figure 6. General layout for equipment included in demonstration project 

The steel studs used are model SSMA 362S125-43 (18 gauges in thickness) with a typical spacing of 40 

cm (16”) and the slotted tracks are model SSMA 362T125-43 (18 gauges in thickness). The tracks were 

connected to the platform slabs using standard power driven 25 mm (1”) fasteners, spaced at 30 cm (1’). 

Two fasteners were considered at the ends and intersections of walls. A gun Ramset model SA-270 and 

shots Ramset .27 caliber were used in construction. Studs were screwed to tracks and studs using standard 

Phillips self-drilling #8 screws. Gypsum wallboard panels (4’x8’) with a thickness of 15 mm (5/8”) were 

screwed to the studs and finished with metal corner beads (1-1/4”), drywall paper joint tape (2-1/16” in 

width), mud and water-based semi-gloss paint. The construction of the steel stud framing followed 

specifications in the Steel Stud Manufacturers Association manual [8]. Gypsum panels were attached to 

the steel stud framing using standard Phillips self-drilling #6 screws. Typical spacing for screws were 12 

and 8 inches for internal and perimeter screws, respectively. Drywall panel joints were offset on opposite 

faces of partition walls. A gap of approximately 1/8” is left in the connection between top ends of studs 

and top tracks. Figure 7 shows typical stud arrangements at wall intersections. 

Figure 8 and Figure 9 show pictures of the typical connection of studs to tracks and studs to studs at wall 

intersections. Figure 10 shows a general view of the frame around the door fenestration. Figure 11 shows 

the detail of the door frame connection to the main steel stud frame. Figure 12 shows the detail of the 

connection of the top track to the concrete slab. In Figure 12 it is seen the typical distribution of fasteners 

along the track. Figure 13 shows a general view of the final steel stud framing system. Figure 14 and 







Figure 15 show general views for the installed gypsum drywall panels in south, and east and north walls, 

respectively. 

Self Drill Screw #6 

at 12 in o.c. (Typ) 



Steel Stud SSMA 

362S125-43 (Typ) 



5/8" Gypsum Board (Typ) 

Figure 7. Stud arrangement at wall intersections 

Figure 8. Detail wall intersection 

Figure 9. Detail stud to track connection 

Figure 10. General view door opening frame 

Figure 11. Detail door opening frame 







Figure 12. Detail top track connection 

Figure 13. General view steel stud framing 

Figure 14. General view gypsum drywall panels in 

south wall 

Figure 15. General view gypsum drywall panels in 

east and north walls 

2.2. Suspended ceiling 

A suspended ceiling system was considered for finishing the interior of the hospital emergency room. The 

ceiling was placed at a height 9’ over the level of the finished floor. The main beams and cross tees are 

Armstrong model Prelude XL 15/16” Fire Resistant, exposed tee. The wall moldings used are 2”x2” angle 

sections. The main runners were installed in the east-west direction at spacing of 4’ on center. 4’ cross 

runners were installed in the north-south direction at spacing of 2’ on center, whereas 2’ cross runners 

were installed in the east-west direction at spacing of 2’ on center. The tiles used were 24"x24" Fine 

Fissured with Angled Tegular edge profile. 

Figure 16 shows the layout of the ceiling grid installed. Steel hanger wires #12 in gage were used to 

support the main beams at the points indicated in Figure 16. The spacing between wire hangers is 4’ in 

both directions. A light was installed directly supported on the ceiling grid, Figure 16. 

Figure 17 shows a detail of the connection between main beams and 4’ cross tees. Figure 17 also shows a 

detail of the wire used to suspend the ceiling grid and partially the light attached to the grid. Figure 18 

shows a detail of the wall molding angle profiles. Figure 19 shows a general view of the installed ceiling 

while Figure 20 shows the detail of the ceiling around one of the UB-NCS columns. 







1'-10 1 8 " 2' 2' 2' 2' 1'-10 7 8 " 

11 3 8 " 2' 2' 2' 2' 11 3 8 " 

Main beam 

4' Cross runner (Typ) 

4' Cross runner 

Opening for light 

2x2" angles attached to 

gypsum partition walls (Typ) 

Main beam 

2' Cross runner (Typ) 

#12 Suspension 

wire (Typ) 

Figure 16. Layout ceiling grid 

Figure 17. Detail main runner and grid hanger 

Figure 18. Detail ceiling grid and wall molding 

Figure 19. General view ceiling 

Figure 20. Detail ceiling around UB-NCS column 







2.3. Sprinkler system 

A fire extinguishing system, composed of vertical and horizontal schedule 40 pipe runs, ½” in diameter, 

was considered for the specimen test. Figure 21 and Figure 22 present a plan view and an elevation 

showing the layout of the pipe runs. The rising portion of the pipe runs was attached to the UB-NCS 

concrete slabs using a combination of flanges and pipe clamps. The horizontal sprinkler pipe run was 

attached to the partition walls using a combination of flanges and pipe clamps as shown in Figure 23 and 

Figure 24, and to the top UB-NCS concrete slab using, additionally, 3/8” all threaded rod hangers 7” in 

length (Figure 25). Figure 26 show a general view of the fire suppression pipe run. A Standard Spray 

Pendant sprinkler head model Rasco F1 with glass bulb type LPC-VdS (rated for response at 155°F) was 

considered to interact with the suspended ceiling system (Figure 27). During testing the fire extinguishing 

system was connected to a hydrant providing typical working pressure. 

2'-2" 

10" 1'-4" 

Sprinkler head 

4'-10" 

4'-9" 

9" 3' 1' 

2" 

9" 

Vertical rod hanger 

Horizontal Run. D=1/2" 

5'-7" 

2'-1" 

Clamp to gypsum partition wall 

Vertical Rise. D=1/2" 

Figure 21. Plan view sprinkler pipe runs 

2.4. Medical gas 

A medical gas system was considered in the testing specimen. The piping layout is shown in Figure 28. 

The copper pipes used were ½” in diameter. The vertical portion of the pipes was attached to the UB-NCS 

concrete slab, as shown in Figure 23. The horizontal portion of the runs were hanged from the top slab 

using trapezoidal hangers whose all threaded 3/8” in diameter rods (L=35 in) were inserted into the top 

concrete slab using standard drop-in devices. Figure 29 shows a general view of the trapezoidal hanger. 

Figure 30 and Figure 31 show details of the clamp used to connect the medical pipe to the trapeze and 

details of the drop-in device used to attach the threaded rod to the concrete slab, respectively. Figure 32 

shows a general view of the medical pipes within the partition walls and the corresponding in-wall 

mounted outlets. The outlets were installed using standard hospital construction techniques (Figure 33). 

Bracing systems were not considered for the trapezoidal hanger. In order to reproduce normal ER 







operation conditions, the medical pipes were pressurized for testing. The free space between the 

horizontal medical gas run and the sprinkler vertical run is 6 inches (Figure 29). 

2'-1" 2'-7" 2'-2" 

2'-8 1 2 " 71 2 " 


Ceiling Level 

Sprinkler head 

9'-4" 


Attachment to concrete slab 

7'-10" 

Connection to waterline 

4'-10" 

Figure 22. Elevation sprinkler pipe runs 







Figure 23. Detail attachment vertical pipe to UB- 

NCS concrete slab 

Figure 24. Detail sprinkler run crossing partition 

wall 

Figure 25. Detail vertical rod hanger 

Figure 26. General view of 

sprinkler pipe runs 

Figure 27. Detail sprinkler 

head 

Trapezoidal hanger 


4'-31 8 " 

Trapezoidal hanger 

105 16 " 

2'-015 16 " 

2'-8 3 8 " 

4'-1 7 8 " 


Figure 28. Layout medical gas pipes 







Figure 29. Medical pipe run trapezoidal hanger 

Figure 30. Detail clamp connecting pipe to hanger 

Figure 31. Connection hanger to 

top concrete slab 

Figure 32. In-wall outlets and gas 

pipes 

Figure 33. Detail in-wall outlets 

mounting 

Additional components (Figure 34 through Figure 36) included in the testing specimen will be further 

described in the final report. 

Figure 34. Dummy sitting on gurney, 

poles with IV pumps, video rack, cart 

and monitor. 

Figure 35. Medical gas 

piping, outlets and 

monitor. 

Figure 36. Video rack. 







3. Description testing protocol 

Testing protocols currently used for the seismic performance assessment of nonstructural components and 

equipment (FEMA 2006, ICC-ES 2007) focus either on displacement or acceleration sensitive 

components, through racking or shake table protocols. However, some nonstructural systems may be 

sensitive to both displacement and accelerations. Further, nonstructural systems typically found in 

buildings may be composed of components that individually may be either acceleration or displacement 

sensitive, but when combined with other systems may become sensitive to both accelerations and 

interstory drifts. In hospitals, for example, acceleration sensitive patient monitors are typically attached to 

displacement sensitive partition walls. The seismic performance of individual nonstructural components 

and the assessment of interactions between components can be evaluated through a testing protocol taking 

full advantage of the UB-NCS capabilities. To this end, an innovative dynamic testing protocol applicable 

for both experimental seismic fragility assessment and seismic qualification of acceleration and/or 

displacement sensitive nonstructural systems is proposed. The testing protocol was mainly developed for 

use with the UB-NCS; nevertheless, the methodology can be used for experimental seismic fragility 

research on acceleration sensitive components performed using conventional shake table simulators. 

Previous research developed by Wilcoski et al. (1997) and Krawinkler et al. (2000) constitutes the basis 

for the development of this testing protocol. 

3.1 Structural model 

The protocol used in this experiment consists of a pair of displacement histories for the bottom and top 

levels of the UB-NCS that simultaneously match: (i) a target mean 84 th % Floor Response Spectrum 

(FRS), and (ii) a target mean 84 th % Generalized Interstory Drift (GID), both specified at a given 

normalized building height h/H, where h is the height above grade where the nonstructural component is 

located, and H is the total height of the building. 

The continuous beam model shown in Figure 37 was considered to model a generic multistory building 

and to describe the seismic load path traveling from the ground to nonstructural components. The building 

model consists of a continuous elastic cantilever beam combining a flexural and a shear beams connected 

throughout an infinite number of axially rigid links distributed along height. The shear and flexural beams 

undergo the same deformations, allowing for modeling of generic buildings whose seismic resistant 

systems are composed by either shear walls, moment resistant frames or a combination of both. This 

model has been extensively studied by other researchers (e.g. Iwan 1997, Chopra and Chintanapakdee 

2001, Kim and Collins 2002, Miranda and Akkar 2006), and has shown promising results in simulating 

the seismic responses observed in multistory buildings during real ground motions (e.g. Reinoso and 

Miranda 2005, Taghavi and Miranda 2006). The parameterα accounts for the relative stiffness of shear 

and flexural beams. A limiting value α = 0 allows for replicating the deformation pattern associated to 

theoretical pure flexural buildings, while α = ∞ represents a pure shear system. 

The input ground motion for the building is characterized by a power spectral density function 

PSDF, Su g 

( ω) 

, which is a frequency domain representation of the ground motion energy content. The 

PSDF can be obtained for: (i) a Deterministic Local Seismic Hazard (DLSH) using the specific barrier 

model (Papageorgiou and Aki 1983, Halldorsson 2005) or (ii) a Probabilistic Local Seismic Hazard 

(PLSH) compatible PSDF (Gupta and Trifunac 1998). Input/output PSDF relations for the combined 

primary/secondary system were derived using principles of stochastic processes (Soong and Grigoriu 

S T x,ω , the 

1993). The PSDF for the absolute accelerations along the height of the building model, ( ) 

u 







absolute acceleration of a SDOF secondary system located at a height h, S T ( h, ω , ωs 

) 

us 

interstory drifts, S ( x, ω ), are expressed as follows: 

θ 

, and the generalized 

Nm 

⎧ 

2 

⎪Su ( ω) 2S ( ) ( ) ( )( ) 

g 

u 

ω γ 

g 

nϕn x ⎡1 Hn ω ωn 2iζnωω 

⎤ 

 

− 

∑ ⎣ 

+ + 

n ⎦ 

+ 

n= 

1 

⎪ 

Nm 

Nm 

⎪ 

2 

S T ( x, ω) 

= 

u ( ) ( ) ( ) ( 

u ⎨S ω γ )( ) 

g 

nγmϕn x ϕm x ⎡1 2Hn ω ωn 2iζnωω 

⎤ 

∑∑ 

⎣ 

− + 

n ⎦ 

+ 

 

⎪ 

n= 1 m= 

1 

⎪ Nm 

Nm 

⎪∑∑γ nγmϕn ϕm ωnωm ⎣ωnωm ζζ 

n m 

ωωζ 

m n 

ωnζ m ⎦ u 

ω 

g n 

ω 

m 

ω 

⎪⎩ n= 1 m= 

1 

T 

us 

* 

( x) ( x) ⎡ + 4 + 2i ( − ) ⎤S 

( ) H ( ) H ( ) 

2 2 2 4 * 

( ω ω ) = T ( ω)( ζ ωω + ω ) ( ω) ( ω) 

S h, , S h, 4 H H 

s u 

s s s s s 

N N 

dϕ 

dϕ 

Sθ 

x, S x x H H 

dx dx 

m m 

n 

m 

* 

( ω) = 

u 

( ω) ( ) ( ) ( ) ( ) 

g 

∑∑γnγm n 

ω 

m 

ω 

n= 1 m= 

1 

(1) 

(2) 

(3) 

x 

GA 

EI 

ζ p , T p, ζ s , T s 

1 GA 

α = 

H EI 

H 

k 

c 

s 

s 

m 

s 

u (h,t) s 

h 

u(x,t) 

.. 

u g(t) 

Figure 37. Building model 

2 2 

In Equations 1, 2 and 3, ( ) ( ) 1 

H 

ω = ω − ω + 2iζ ωω − 

, ωs 

and 

s s s s 

ζ 

s 

denote frequency response function, 

2 2 

H ω = ω − ω + 2iζ ωω − 

, 

natural frequency and damping ratio for the secondary system, respectively; ( ) ( ) 1 

n n n n 

ω 

n 

, ζ 

n 

, γ 

n 

, and ϕ n 

denote frequency response function, natural frequency, damping ratio, modal 

participation factor and modal shape for the n th vibration mode of the primary system, respectively. 

i= − 1 denotes the imaginary unit and N m is the number of significant modes considered in the analysis. 

Principles of random vibration theory were used to estimate the mean peak seismic demands expected on 

nonstructural components (Cartwright and Longuet-Higgins 1956). 

3.2 Estimation of seismic demands on nonstructural components 

In developing and calibrating the testing protocol, a PLSH with a probability of exceedance of 10% in 50 

years for a region of high seismicity such as Northridge, California, was considered. The corresponding 

uniform hazard spectrum, obtained from USGS, is shown in Figure 38. In order to compute a PLSH 







compatible PSDF to serve as input for the continuous beam model, the ground response spectrum was 

fitted using: 

S 

a 

( T) 

A 

o 

= 

T 

1 + 

0.1T 

⎛ T ⎞ 

1+ 4.5⎜ ⎟ 

⎝ T 

o ⎠ 

q 

⎛ T ⎞ 

T 1 + ⎜ ⎟ 

T 

R 

o 

+ ⎝ o ⎠ 

p 

(4) 

where the constants p, q and R are estimated using best fit techniques. A o denotes the peak ground 

acceleration at the site and T o =0.2 s. The best fit curve is also shown in Figure 38. The ground response 

spectrum compatible PSDF, Su g 

( ω) 

, is then obtained as shown in Figure 39, where T d corresponds to the 

duration of the stochastic process, estimated according to Papageorgiou and Aki (1983). 

In computing the demands to be imposed by the protocol, buildings with deformation patterns defined by 

parameters α = 0, 5 and 10 were considered. Primary systems with fundamental periods T p in the range 

0.1 to 5 sec, and secondary systems with natural periods T s in the range 0 to 5 sec, were considered. The 

damping ratio for primary (all modes) and secondary systems is assumed equal to 5%. N m =10 modes are 

considered in the analysis. Figure 40 shows a three dimensional FRS, obtained using Equations 1 and 2 

for the case α = 5 at building roof level. Similar 3D FRS’s are obtained for the other building heights and 

α values. The data in Figure 40 and that obtained for other α’s have been statistically processed to obtain 

the expected floor demands that will be applied by the general testing protocol, independent of building 

deformation pattern α and primary system fundamental period T p . In order to do so, the mean (over α) 

84 th % (over T p ) FRS’s were computed as functions of normalized building height. The results of this 

statistical analysis are shown in Figure 41a. In order to smooth the mean 84 th % FRS’s shown in Figure 

41a, a function FRSFactor 

( h H ) is used to extrapolate the ground response spectrum shown in Figure 38 to 

FRS h H is calculated as the quotient between 

obtain the FRS’s along building height. The function ( ) 

the peak value of the FRS’s along the height of the building and the peak spectral amplitude of the ground 

FRS h H . 

response spectrum (Figure 41b). Figure 41b also shows the best fit curve (Equation 5) for ( ) 

Factor 

Finally, the extrapolated mean 84 th % FRS to be matched by the testing protocol, shown in Figure 41c, is 

calculated using Equation 6. 

Factor 

Spectral Acceleration S a (g) 

1.6 

1.4 

1.2 

1 

0.8 

0.6 

0.4 

0.2 

USGS Hazard Consistent Ground Response Spectrum 

USGS Data 

Best Fit USGS Data 

Interpolating Function 

p 

⎛ 

T 

⎞ 

1 + 4.5 ⎜ ⎟ 

A 

o 

T 

o 

S 

a ( T 

) 

= 

⎝ ⎠ 

q 

T 

1 

+ 

⎛ 

T 

⎞ 

T 

1 

+ ⎜ ⎟ 

0.1T 

o 

+ ⎝T 

o 

⎠ 

R 

0 

0 0.5 1 1.5 2 2.5 

Period T (sec) 

Figure 38. Design level ground 

response spectrum 

PSD (cm 2 /s 3 ) 

250 

200 

150 

100 

50 

USGS Hazard Consistent PSD Ground Acceleration 

⎛ 

⎞ 

2 ⎜ 

⎟ 

6ζSa 

( ω) 

Tdω γ 

S u ( ωζ , ) 

⎜ ⎛ ⎞ 

2ln 

⎟ 

= 

g 

2 2 4 ⎜ 

πω ( 12ζ π ζ 3) 

π 

⎟+ 

+ + ⎜ ⎝ ⎠ Tdω 

⎟ 

⎛ ⎞ 

⎜ 

2ln⎜ 

π 

⎟ 

⎝ ⎠ 

⎟ 

⎝ 

⎠ 

0 

0 5 10 15 20 25 30 35 40 45 50 

Frequency f (Hz) 

Figure 39. Ground response 

spectrum compatible PSDF 

−2 

Absolute Acceleration (g) 

15 

10 

5 

0 

5 

4 

3 

T p 

(s) 

3-D Floor Response Spectra at Roof Level, α=5 

2 

1 

0 

0 

1 

2 

T s 

(s) 

Figure 40. Example of 3D-FRS 

for α =5, roof level (h/H=1) 

3 

4 

5 

14 

12 

10 

8 

6 

4 

2 

Using Equation 3, the PSDF for generalized interstory drifts (GID) were calculated. Principles of 

stochastic processes were used to compute the Generalized Drift Spectra (GDS) shown in Figure 42. To 







make the testing protocol independent of deformation patterns and dynamic properties of primary system, 

the mean (over α ) 84 th % (over T p ) GID along building height were determined. Figure 43 shows the 

mean 84 th % GID and its best fit along building height (Equation 7). 

2 3 

⎛ h ⎞ h ⎛ h ⎞ ⎛ h ⎞ 

FRSFactor 

⎜ 1 10 19.4 12.4 

H 

⎟= + − + 

H 

⎜ 

H 

⎟ ⎜ 

H 

⎟ 

⎝ ⎠ ⎝ ⎠ ⎝ ⎠ 

⎛ h ⎞ ⎛ h ⎞ 

FRS ⎜T s 

, ⎟= 

FRSFactor ⎜ ⎟Sa ( Ts 

) 

⎝ H ⎠ ⎝ H ⎠ 

2 0.55 

δ ⎛ h 1 h 6 h h 

⎜ ⎞ sin 7 1.9 % 

H 

⎟= ⎛ 4 

⎜ ⎞ − ⎛ ⎞ + 

⎛ ⎞ 

H 

⎟ 

5 

⎜ 

H 

⎟ ⎜ 

H 

⎟ 

⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ 

(5) 

(6) 

(7) 

FRS (g) 

6 

5 

4 

3 

2 

Mean 84 th Percentile FRSs along Building Height 

Ground 

h=0.1H 

h=0.2H 

h=0.3H 

h=0.4H 

h=0.5H 

h=0.6H 

h=0.7H 

h=0.8H 

h=0.9H 

Roof 

h/H 

1 

0.9 

0.8 

0.7 

0.6 

0.5 

0.4 

0.3 

Variation of Peak FRS / Peak S a 

Rates along Building Height 

FRS (g) 

6 

5 

4 

3 

2 

Extrapolated Mean 84 th Percentile FRSs along Building Height 

Ground 

h=0.1H 

h=0.2H 

h=0.3H 

h=0.4H 

h=0.5H 

h=0.6H 

h=0.7H 

h=0.8H 

h=0.9H 

Roof 

1 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 

Period Secondary System T s 

(sec) 

0.2 

0.1 

Data 

Best Fit 

0 

1 1.5 2 2.5 3 3.5 4 4.5 

Peak FRS / Peak S a 

Rate 

1 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 

Period Secondary System T s 

(sec) 

(a) (b) (c) 

Figure 41. (a) Mean 84 th percentile FRS along building height; (b) Floor response spectra 

extrapolation factor; and (c) Extrapolated (smoothened) mean 84 th percentile FRS along building 

height 

Drift δ (%) 

Drift δ (%) 

Drift δ (%) 

2.5 

2 

1.5 

1 

0.5 

2.5 

1.5 

0.5 

Generalized Drift Spectrum α=0 

0 

0 1 2 3 4 5 6 

Period Primary System T p 

(sec) 

2 

1 

2.5 

1.5 

0.5 


0 

0 1 2 3 4 5 6 


(sec) 

2 

1 


h=0.1H h=0.2H h=0.3H h=0.4H h=0.5H h=0.6H h=0.7H h=0.8H h=0.9H Roof 

0 

0 1 2 3 4 5 6 


(sec) 

Figure 42. GDS at several building heights 

Normalized Building Height h/H 

1 

0.7 

0.6 

0.5 

0.4 

0.3 

0.2 

0.1 

Variation of Mean 84 th % Generalized Drift along Building Height 

2 0.55 

0.9 

δ ⎛ h 1 h 6 h h 

⎜ ⎞ ⎟= sin ⎛ ⎜7 ⎞ ⎟− ⎛ ⎜ ⎞ ⎟ + 1.9 

⎛ ⎜ ⎞ 

⎟ 

⎝ H ⎠ 4 ⎝ H ⎠ 5⎝ H ⎠ ⎝ H ⎠ 

0.8 

Mean 84 th % 

Best Fit 

0 

0 0.2 0.4 0.6 0.8 1 1.2 1.4 

Generalized Drift δ (%) 

Figure 43. Mean 84 th % GID 

3.3 Generation of hazard consistent floor displacement protocol histories 

After computing the demands to be imposed by the testing protocol through the mean 84 th % FRS and the 

mean 84 th % GID, displacement histories for the bottom and top levels of the UB-NCS are generated. The 

key parameters for the protocol histories are the location of the component along the normalized building 

height and the target ground response spectrum. The frequency content targeted for the testing protocol 

covers the range between f min 

=0.2 and f max 

=5 Hz, which corresponds to the UB-NCS operating frequency 







range and the frequency content expected in the responses of multistory buildings. The instantaneous 

frequency f () t considered in the testing protocol is given by: 

() 

f t 

t−td 

⎛ 

td 

f ⎞ 

min 

= fmax 

⎜ ⎟ 

fmax 

⎝ 

⎠ 

−1 

(8) 

The duration of the testing protocol is 

reached and is given by: 

2t , where 

d 

t d 

is the time at which the minimum testing frequency is 

t 

1 

f 

= max 

d 

log 

(9) 

2 

Sr 

fmin 

where S denotes a constant sweep rate calibrated to induce the same number of “Rainflow” cycles 

r 

(ASTM 1997) on acceleration sensitive nonstructural components as it would be experienced during real 

floor motions. Equation 8 corresponds to an instantaneous testing frequency transitioning from high to 

low frequencies, then back to high frequencies. The final high frequency sweep is intended to capture the 

behavior of components that might be damaged initially by drifts and become sensitive to accelerations 

(e.g. partition walls acting as a cantilever after failure of top slab connection). The displacement protocol 

proposed for the UB-NCS bottom level, x Bottom 

, matches the mean 84 th % FRS expected at a given 

normalized height. The closed-form equation for the bottom level protocol is: 

h 

h 

x ⎛ β 

Bottom ⎜t, ⎞ ⎟= 

α f () t cos( ϕ() 

t ) w() 

t FRS 

⎛ ⎞ 

Factor ⎜ ⎟ 

⎝ H ⎠ ⎝ H ⎠ 

(10) 

where α =0.75 and β = -1.35 are calibration factors used to minimize the error in matching the ground 

wt is a 

response spectrum in the range of frequencies of interest; ( t) 

ϕ is the instantaneous phase; and ( ) 

sinusoidal windowing function used to smooth the ramp-up and ramp-down portions of the protocol. The 

interstory drift protocol is calibrated to impose a controlled number of rainflow cycles. The peak 

interstory drift reached during testing matches the mean 84 th % GID expected at a given normalized 

building height. The closed-form equation for the interstory drift protocol, Δ , is given by: 

2 

⎛t−t h 

d ⎞ 

⎛ ⎞ 

−⎜ ⎟ 

σ h 

⎝ ⎠ ⎛ ⎞ 

⎜ ⎟= 

NCS ⎜ ⎟ 

( ()) () 

Δ t, h e δ cos ϕ t w t 

⎝ H ⎠ ⎝ H ⎠ 

(11) 

where δ ( hH) 

is given in Equation 7; h NCS 

is the free interstory height of the UB-NCS; and 

−( ( t−t ) ) 2 

d σ 

e is a 

Gaussian-shaped modulating function in which σ is calibrated to control the amplitude of Rainflow 

cycles imposed by the interstory drift protocol. The closed-form equation for the top level displacement 

protocol, x 

Top 

, is given by: 

⎛ h ⎞ ⎛ 

x h ⎞ ⎛ h ⎞ 

Top ⎜t, ⎟= xBottom 

⎜t, ⎟+ 

Δ⎜t, 

⎟ 

⎝ H ⎠ ⎝ H ⎠ ⎝ H ⎠ 

(12) 

Figure 44 shows the proposed protocol histories for the bottom and top UB-NCS levels for a normalized 

building height h/H =1, corresponding to a generic building roof level. 







3.4 Calibration of induced/imposed vibration cycles 

In order to validate the proposed testing protocol and its main parameters including instantaneous 

frequency f ( t ) , the sweep rate S r 

, and the parameter σ controlling the interstory drift protocol’s 

enveloping shape, the number of cycles induced/imposed by the proposed floor motions has been 

assessed. The protocol is compared to floor motions recorded during past earthquakes at the upper levels 

of 6 instrumented buildings selected from the California Strong Motion Instrumentation Program 

(CSMIP) database. The buildings considered are listed in Table 2. Buildings with different number of 

stories, seismic resistant systems and natural periods, designed and built between 1960 and 1990 and 

affected by earthquakes with epicenters located at less than 15 km were considered. 

20 

Bottom Displacement History 

20 

Top Displacement History 

Disp (in) 

0 

Disp (in) 

0 

Veloc (in/s) 

-20 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Bottom Velocity History 

40 

20 

0 

-20 

-40 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Bottom Acceleration History 

1 

Veloc (in/s) 

-20 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Top Velocity History 

40 

20 

0 

-20 

-40 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Top Acceleration History 

1 

Acc (g) 

0 

Acc (g) 

0 

Inst Freq (Hz) 

-1 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Instant Frequency 

6 

4 

2 

0 

0 5 10 15 20 25 30 35 40 

Time (sec) 

(a) 


-1 

0 5 10 15 20 25 30 35 40 

Time (sec) 


4 

2 

0 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Figure 44. Testing protocol at hH=1 for (a) Bottom UB-NCS level; and (b) Top UB-NCS level 

(b) 

Table 2. CSMIP Instrumented buildings selected for cycle counting analysis and protocol calibration 

CSMIP 

Station 

ID 

City 

24464 North 

Hollywood 

Building 

Structural System 

Reinforced 

concrete columns 

and beams 

24514 Sylmar Concrete slab, 

metal deck, steel 

frames 

24629 Los 

Angeles 

Concrete slabs, 

steel frames and 

deck 

47459 Watsonville Concrete slabs 

and shear walls 

24322 Sherman Concrete slabs, 

Oaks beams, and 

columns 

24386 Van Nuys Concrete slabs, 

columns, 

spandrel beams 

Number 

of 

Stories 

Building 

Period 

(s) 

Type 

Design 

Date 

Site Geology 

20 2.56 Hotel 1967 Sandstone, 

shale 

Recorded 

Earthquakes 

in Station 

Whittier 

Northridge 

6 0.40 Hospital 1976 Alluvium Whittier 

Northridge 

Min. 

Distance 

to Source 

(km) 

Max. 

PGA 

(g) 

Max. 

PFA 

(g) 

15 0.30 0.65 

13 0.80 1.50 

54 6.20 Office 1988 Alluvium 

over 

sedimentary 

rock 

Northridge 32 0.13 0.18 

4 0.37 Commercial - Fill over Loma Prieta 17 0.58 1.20 

alluvium 

13 0.84 Commercial 1964 Alluvium Whittier 13 0.75 0.42 

Landers 

Northridge 

7 1.58 Hotel 1965 Alluvium Landers 7 0.45 0.58 

Big Bear 

Northridge 

In order to compute the number of cycles induced on acceleration sensitive nonstructural components, 

linear elastic analyses of SDOF systems were performed considering the bin of recorded floor 

accelerations and the proposed testing protocol for a normalized building height h/H=1. The “Rainflow” 

counting algorithm was used to simplify the response excursion amplitudes. A parameter 







Nλ 

= NCycles ACycle ≥ λ A , corresponding to the number of rainflow cycles 

Max 

N Cycles 

with amplitude A Cycle 

greater than λ % of the maximum rainflow amplitude A Max 

observed in the oscillator’s response, is 

considered to calibrate the testing protocol sweeping rate. In particular, two statistical parameters are 

calculated for acceleration sensitive components, N 10 and N 50 , representative of the total number of 

damaging excursions and the proportion of large excursions observed in the SDOF’s response, 

respectively. Figure 45 shows a comparison between the number of rainflow cycles induced by the testing 

protocol ( S r 

=12 oct/min) and the mean and 84 th % number of rainflow cycles induced by the set of 

recorded floor accelerations on SDOF with frequencies in the range considered. In Figure 45 it can be 

seen that for systems with frequencies lower than 1.5 Hz, the proposed testing protocol closely matches 

the mean number of large cycles N 50 induced by real floor motions. For SDOF’s with frequencies greater 

than 1.5 Hz, the testing protocol closely matches the 84 th % number of large cycles induced by real floor 

motions. Similarly, it is observed that the testing protocol closely matches the mean total number of 

damaging excursions N 10 for oscillators with natural frequencies lower than 3 Hz. For systems with 

frequencies greater than 3 Hz, the testing protocol slightly underestimates the mean total number of 

damaging excursions induced by real floor accelerations. 

The rainflow counting algorithm was also used to compute the number of cycles N λ 

observed in the 

interstory drift histories calculated from the displacement histories obtained after integration and baseline 

correction of the recorded floor acceleration histories. Rainflow algorithm was also applied on the 

proposed interstory drift protocol, considering several values for the parameterσ . Figure 46 shows the 

results of the calibration process for the testing protocol proposed for h/H=1. Dots in Figure 46 denote 

individually processed data. The optimal value for the parameterσ is 9.8 sec. In Figure 46 it is seen that 

the number of rainflow cycles imposed by the interstory drift protocol closely matches the 84 th % number 

of rainflow cycles imposed by real building interstory drift histories. It is highlighted that the combined 

use of the calibrated Gaussian window and the proposed instantaneous frequency function allows for 

closely matching the 84 th % number of cycles in the whole profile of rainflow cycle amplitudes, for 

which λ varies between 10 and 90. 

N 50 

Number of Rainflow Cycles Induced on Acceleration Sensitive Components 

30 

60 

Number of Cycles Imposed on Displacement Sensitive Components 

N 50 

20 

Mean 

10 

84 th 

Protocol 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 

Frequency Secondary System 

N 10 

Number of Rainflow Cycles Induced on Acceleration Sensitive Components 

150 

N λ 

50 

40 

30 

20 

Protocol 

h/H=1 

Mean Floor 

Motions 

84 th % Floor 

Motions 

100 

N 10 

50 

10 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 

Frequency Secondary System 

Figure 45. N 10 and N 50 number of rainflow cycles 

induced on acceleration sensitive components 

0 

0 10 20 30 40 50 60 70 80 90 100 

λ 

Figure 46. Number of rainflow cycles imposed 

on displacement sensitive components 







4. Testing program 

4.1. Qualification protocol 

The testing specimen was subjected to the qualification protocol described in section 3 for a normalized 

building height h/H=1 (roof level). The non scaled floor motion histories used are shown in Figure 44. 

20 

Bottom Displacement History 

10 

Disp (in) 

0 

-10 

Velo (in/sec) 

-20 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Bottom Velocity History 

40 

20 

0 

-20 

-40 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Bottom Acceleration History 

1 

0.5 

Acc (g) 

0 

-0.5 

-1 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Top Displacement History 

20 

10 

Disp (in) 

0 

-10 

Velo (in/sec) 

-20 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Top Velocity History 

40 

20 

0 

-20 

-40 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Top Acceleration History 

1 

0.5 

Acc (g) 

0 

-0.5 


-1 

0 5 10 15 20 25 30 35 40 

Time (sec) 


5 

4 

3 

2 

1 

0 

0 5 10 15 20 25 30 35 40 

Time (sec) 

Figure 47. Testing protocol floor motion histories (scaled 100%) 







Figure 48a shows a comparison between bottom and top platform displacement histories while Figure 48b 

shows the interstory drift protocol history. Table 3 lists the tests series performed and the corresponding 

expected peak platform motion values. 

20 

15 

Displacement Protocol SH 10%/50yr 

Bottom level 

Top level 

1.5 

1 

Interstory Drift History 

10 

Displacement (in) 

5 

0 

-5 

-10 

Interstory Drift (in) 

0.5 

0 

-0.5 

-15 

h 

β 

h 

xp⎜ ⎛ t, f () t cos( () t ) w() 

t FRSFactor 

H 

⎟ ⎞ = α ϕ 

⎛ ⎜ ⎞ 

H 

⎟ 

⎝ ⎠ ⎝ ⎠ 

-20 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

-1 

⎛t−t d ⎞ 

⎛ h ⎞ 

−⎜ ⎟ h 

⎝ σ ⎠ 

⎛ ⎞ 

Δ⎜t, hNCSe cos( () t ) w() 

t 

H 

⎟= 

δ⎜ ϕ 

H 

⎟ 

⎝ ⎠ ⎝ ⎠ 

-1.5 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

Figure 48. 100% Testing Protocol Roof Level: (a) Comparison floor displacements histories; and 

(b) Interstory drift protocol history. 

Table 3. List of tests performed and envelope of peak floor motion. Protocol histories. 

Test 

Date Scaling Peak Displacements Peak Interstory Drift Peak Velocities Peak Accelerations 

Test Factor D Max Bot (in) D Max Top (in) d Max (in) δ Max (%) V Max Bot (in/s) V Max Top (in/s) A Max Bot (g) A Max Top (g) 

1 09-oct 10% 1,63 1,76 0,13 0,09% 3,1 3,3 0,07 0,08 

2 10-oct 25% 4,08 4,4 0,33 0,22% 7,6 8,2 0,18 0,19 

3 10-oct 50% 8,15 8,80 0,66 0,43% 15,3 16,3 0,37 0,39 

4 12-oct 100% 16,3 17,6 1,31 0,87% 30,5 32,6 0,73 0,77 

5 12-oct 150% 24,5 26,4 1,97 1,30% 45,8 48,9 1,10 1,16 

In Table 3, the 100% scaled protocol history is associated to a seismic hazard (SH) with a probability of 

exceedence (PE) of 10% in 50 years (design earthquake level), while the 150% scaled protocol histories 

impose demands associated to an earthquake with a PF of 2% in 50 years (maximum considered 

earthquake). 

4.2. Simulated building floor motions 

In order to validate the suitability of the testing protocol to impose earthquake compatible damage levels, 

the specimen was subjected to the floor motions obtained, for the same SH levels, from the simulated 

response of an existing four story steel framed medical facility located in the San Fernando Valley, 

California (Wanitkorkul and Filiatrault 2005). 

Figure 49 shows the testing histories for the simulated building response for a SH with a PE of 10% in 50 

years. Figure 49a illustrates the floor displacement histories while Figure 49b shows the interstory drift 

history. Figure 49c and Figure 49d show the floor velocity and acceleration histories, respectively. Figure 

49e presents a comparison of the Fourier transforms (FFT’s) for the simulated building floor motions and 

for the testing protocol (100%). Figure 49f shows a comparison of the floor response spectra (FRS’s) for 

the simulated building floor motions and for the testing protocol (100%). Figure 49g and Figure 49h show 

2 







comparisons of the number of “Rainflow” cycles induced and imposed on acceleration and displacement 

sensitive nonstructural components, respectively. 

10 

8 

6 

Floor Displacement Histories for SH with PE 10%/50yr 

Bottom 

Top 

1.5 

1 

Interstory Drift History for SH with PE 10%/50yr 

Floor Displacement (in) 

4 

2 

0 

-2 

-4 

-6 

-8 


0.5 

0 

-0.5 

-1 

-10 

0 5 10 15 20 25 

Time (sec) 

60 

40 

(a) 

Floor Velocity Histories for SH with PE 10%/50yr 

Bottom 

Top 

-1.5 

0 5 10 15 20 25 

Time (sec) 

1.5 

1 

(b) 

Floor Acceleration Histories for SH with PE 10%/50yr 

Bottom 

Top 

Floor Velocity (in/sec) 

20 

0 

-20 

Floor Acceleration (g) 

0.5 

0 

-0.5 

-40 

-1 

FFT 

FFT 

-60 

0 5 10 15 20 25 

Time (sec) 

2 

1 

(c) 

3 x 105 Comparison FFT UB-NCS Top Level. SH with PE 10%/50yr 

Protocol 

Hospital FM 

0 

0 1 2 3 4 5 6 

Frequency (Hz) 

3 x 105 Comparison FFT UB-NCS Bottom Level. SH with PE 10%/50yr 

2 

1 

Protocol 

Hospital FM 

0 

0 1 2 3 4 5 6 


(e) 

FRS (g) 

FRS (g) 

-1.5 

0 5 10 15 20 25 

Time (sec) 

8 

6 

4 

2 

(d) 

Comparison FRS UB-NCS Top Level. SH with PE 10%/50yr 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


Comparison FRS UB-NCS Bottom Level. SH with PE 10%/50yr 

8 

Protocol 

6 

Hospital FM 

4 

2 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


(f) 







N 10 

N o Cycles with A>0.1A Max 

Induced on Acc Sensitive Components. SH 10%/50yr 

80 

60 

40 

20 

0 

30 

Protocol 

Bottom Hospital FM 

Top Hospital FM 

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 

Frequency Secondary System (Hz) 



40 

N λ 

60 

50 

40 

30 

20 

N o Cycles Imposed on Disp Sensitive Components. SH 10%/50yr 

Protocol 

Hospital FM 

N 50 

20 

10 

10 

0 

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


0 

0 10 20 30 40 50 60 70 80 90 100 

λ 

(g) 

(h) 

Figure 49. Simulated building floor motions for an earthquake event with a PE of 10% in 50 yr: 

(a) Floor displacement histories; (b) Interstory drift history; (c) Floor velocity histories; (d) Floor 

acceleration histories; (e) Comparison FFT’s; (f) Comparison FRS’s; (g) Comparison of number 

of cycles induced on acceleration sensitive nonstructural components; and (h) Comparison of 

number of cycles imposed on displacement sensitive nonstructural components. 

Figure 50 shows the testing histories for the simulated building response for a SH with a PE of 2% in 50 

years. Figure 50a illustrates the floor displacement histories while Figure 50b shows the interstory drift 

history. Figure 50c and Figure 50d show the floor velocity and acceleration histories, respectively. Figure 

50e presents a comparison of the Fourier transforms (FFT’s) for the simulated building floor motions and 

for the testing protocol (150%). Figure 50f shows a comparison of the floor response spectra (FRS’s) for 

the simulated building floor motions and for the testing protocol (150%). Figure 50g and Figure 50h show 

comparisons of the number of “Rainflow” cycles induced and imposed on acceleration and displacement 

sensitive nonstructural components, respectively. 

15 

10 

Floor Displacement Histories for SH with PE 2%/50yr 

Bottom 

Top 

2 

1.5 

Interstory Drift History for SH with PE 2%/50yr 

1 

Floor Displacement (in) 

5 

0 

-5 


0.5 

0 

-0.5 

-1 

-10 

-1.5 

-15 

0 5 10 15 20 25 30 

Time (sec) 

(a) 

-2 

0 5 10 15 20 25 30 

Time (sec) 

(b) 







60 

40 

Floor Velocity Histories for SH with PE 2%/50yr 

Bottom 

Top 

1.5 

1 

Floor Acceleration Histories for SH with PE 2%/50yr 

Bottom 

Top 

Floor Velocity (in/sec) 

20 

0 

-20 

Floor Acceleration (g) 

0.5 

0 

-0.5 

-40 

-1 

FFT 

FFT 

N 10 

-60 

0 5 10 15 20 25 30 

Time (sec) 

3 

2 

1 

(c) 

4 x 105 Comparison FFT UB-NCS Top Level. SH with PE 2%/50yr 

Protocol 

Hospital FM 

0 

0 1 2 3 4 5 6 


4 x 105 Comparison FFT UB-NCS Bottom Level. SH with PE 2%/50yr 

3 

2 

1 

Protocol 

Hospital FM 

0 

0 1 2 3 4 5 6 


(e) 



80 

60 

40 

20 

0 

Protocol 

Bottom Hospital FM 

Top Hospital FM 

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 




40 

30 

FRS (g) 

FRS (g) 

N λ 

-1.5 

0 5 10 15 20 25 30 

Time (sec) 

10 

8 

6 

4 

2 

(d) 

Comparison FRS UB-NCS Top Level. SH with PE 2%/50yr 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


Comparison FRS UB-NCS Bottom Level. SH with PE 2%/50yr 

10 

8 

6 

4 

2 

Protocol 

Hospital FM 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


60 

50 

40 

30 

20 

(f) 

N o Cycles Imposed on Disp Sensitive Components. SH 2%/50yr 

Protocol 

Hospital FM 

N 50 

20 

10 

10 

0 

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


0 

0 10 20 30 40 50 60 70 80 90 100 

λ 

(g) 

(h) 

Figure 50. Simulated building floor motions for a SH with a PE of 2% in 50 yr: (a) 

Displacements; (b) Interstory drift; (c) Velocities; (d) Accelerations; (e) Comparison FFT’s; (f) 

Comparison FRS’s; (g) Comparison number of cycles induced on acceleration sensitive comp.; 

and (h) Comparison of number of cycles imposed on displacement sensitive components. 







Table 4 summarizes the peak demand parameters expected during testing 

Table 4. List of additional tests performed and envelope of peak floor motion. 

Simulated building floor motions. 

Test 

Test 

Scaling Peak Displacements Peak Interstory Drift Peak Velocities Peak Accelerations 

Description Factor D Max Bot (in) D Max Top (in) d Max (in) δ Max (%) V Max Bot (in/s) V Max Top (in/s) A Max Bot (g) A Max Top (g) 

6 Floor Motion PE 10%/50yr 50% 4,45 4,83 0,53 0,35% 20,5 25,5 0,35 0,53 




4.3. Quasi-static tests 

In order to assess the seismic performance of displacement sensitive components at larger drift levels, the 

interstory drift protocol history shown in Figure 48b was applied quasi-statically. Figure 51 shows the 

quasi-static testing protocol. 

1.5 

Interstory Drift Protocol Roof Level 100% 

1 


0.5 

0 

-0.5 

-1 

-1.5 

0 50 100 150 200 250 300 350 400 450 

Time (sec) 

Figure 51. Quasi-static interstory drift protocol (100%). 

The protocol shown in Figure 51 was applied by keeping in rest the bottom UB-NCS level actuators and 

imposing the protocol through the motion of the top UB-NCS level actuators. Table 5 summarizes the 

peak demand parameters expected during testing. 

Table 5. List of additional tests performed and envelope of peak floor motion. Quasi-static tests. 

Test 

Test 

Scaling Peak Displacements Peak Interstory Drift Peak Velocities Peak Accelerations 

Description Factor D Max Bot (in) D Max Top (in) d Max (in) δ Max (%) V Max Bot (in/s) V Max Top (in/s) A Max Bot (g) A Max Top (g) 

10 Drift Protocol 200% 2,62 1,74% 










4.4 Instrumentation setup 

Extensive instrumentation was considered for the platform, monitors and partition walls, as shown in 

Figure 52 through Figure 54 and Table 6. 

NEWVTE 

NEWVTEi 

NEWVTW 

NEWVTWi 

NWWVTE 

NWWVTEi 

NWWVTW 

NWWVTWi 

NEWLT 

NWWLT 

NEWD2 

NWWD2 

NEWD1 

NWWD1 

NEWLB 

NWWLB 

NEWVBE 

NEWVBEi 

SWWVTW 

SWWVTWi 

Exterior View North Wall 

NEWVBW 

NEWVBWi 

NWWVBE 

NWWVBEi 

NWWVBW 

NWWVBWi 

SEWVTE 

SEWVTEi 

SCWD1 

SCWLT 

SWWD1 

SCWD2 

SEWD1 

SWWD2 

SEWD2 

SWWLB 

SEWLB 

SWWVBW 

SWWVBWi 

SWWVBE 

SWWVBEi 

SEWVBW 

SEWVBWi 

Exterior View South Wall 

Figure 52. Instrumentation partition walls. 

SEWVBE 

SEWVBEi 







ACBLSE 

(ACTLSE) 

ACBTSE 

(ACTTSE) 

ACBTSW 

(ACTTSW) 

Reaction Wall 

ACBV 

(ACTV) 

ACBLNE 

(ACTLNE) 

Accelerometers in Platform 

Figure 53. Instrumentation platform. 

MXAVb 

MXANSb 

MXAEWb 

MXAVCh 

MXANSCh 

MXAEWCh 

Accelerometers in Monitor X 

Scale: - 

Figure 54. Instrumentation monitors. 







Table 6. Instrumentation list 

Channel ID Instrument 

Response Min Operation 

quantity Limits 

Orientation 

Location and comment 

Data acquired from Actuators 

1 Time Time 

2 ComActA Actuator Disp ±40in Actuator A command 

3 ComActB Actuator Disp ±40in Actuator B command 

4 ComActC Actuator Disp ±40in Actuator C command 

5 ComActD Actuator Disp ±40in Actuator D command 

6 DispActA Actuator Disp ±40in Actuator A displacement 

7 DispActB Actuator Disp ±40in Actuator B displacement 

8 DispActC Actuator Disp ±40in Actuator C displacement 

9 DispActD Actuator Disp ±40in Actuator D displacement 

10 ForceA Actuator Force ±50kip Actuator A 

11 ForceB Actuator Force ±50kip Actuator B 

12 ForceC Actuator Force ±50kip Actuator C 

13 ForceD Actuator Force ±50kip Actuator D 

Accelerometers in Platforms 

14 ACBLNE Accelerometer Acceleration ±10g EW Bottom Platform 

15 ACBLSE Accelerometer Acceleration ±10g EW Bottom Platform 

16 ACBTSW Accelerometer Acceleration ±10g NS Bottom Platform 

17 ACBTSE Accelerometer Acceleration ±10g NS Bottom Platform 

18 ACBV Accelerometer Acceleration ±10g V Bottom Platform 

19 ACTLNE Accelerometer Acceleration ±10g EW Top Platform 

20 ACTLSE Accelerometer Acceleration ±10g EW Top Platform 

21 ACTTSW Accelerometer Acceleration ±10g NS Top Platform 

22 ACTTSE Accelerometer Acceleration ±10g NS Top Platform 

23 ACTV Accelerometer Acceleration ±10g V Top Platform 

Instrumentation North East Wall 

24 NEWD1 String Pot Displacement ±6 in Diagonal Diagonal north east wall 

25 NEWD2 String Pot Displacement ±6 in Diagonal Diagonal north east wall 

26 NEWVTE Potentiometer Disp uplift ±1 in V North east wall top east end (stud to concrete) 

27 NEWVTEi Potentiometer Disp uplift ±1 in V North east wall top east end (stud to track) 

28 NEWVTW Potentiometer Disp uplift ±1 in V North east wall top west end (stud to concrete) 

29 NEWVTWi Potentiometer Disp uplift ±1 in V North east wall top west end (stud to track) 

30 NEWVBE Potentiometer Disp uplift ±1 in V North east wall bottom east end (stud to concrete) 

31 NEWVBEi Potentiometer Disp uplift ±1 in V North east wall bottom east end (stud to track) 

32 NEWVBW Potentiometer Disp uplift ±1 in V North east wall bottom west end (stud to concrete) 

33 NEWVBWi Potentiometer Disp uplift ±1 in V North east wall bottom west end (stud to track) 

34 NEWLT String Pot Displacement ±6 in EW North east wall relative motion top track 

35 NEWLB String Pot Displacement ±6 in EW North east wall relative motion bottom track 

Instrumentation North West Wall 

36 NWWD1 String Pot Displacement ±6 in Diagonal Diagonal north west wall 

37 NWWD2 String Pot Displacement ±6 in Diagonal Diagonal north west wall 

38 NWWVTE Potentiometer Disp uplift ±1 in V North west wall top east end (stud to concrete) 

39 NWWVTEi Potentiometer Disp uplift ±1 in V North west wall top east end (stud to track) 

40 NWWVTW Potentiometer Disp uplift ±1 in V North west wall top west end (stud to concrete) 

41 NWWVTWi Potentiometer Disp uplift ±1 in V North west wall top west end (stud to track) 

42 NWWVBE Potentiometer Disp uplift ±1 in V North west wall bottom east end (stud to concrete) 

43 NWWVBEi Potentiometer Disp uplift ±1 in V North west wall bottom east end (stud to track) 

44 NWWVBW Potentiometer Disp uplift ±1 in V North west wall bottom west end (stud to concrete) 

45 NWWVBWi Potentiometer Disp uplift ±1 in V North west wall bottom west end (stud to track) 

46 NWWLT String Pot Displacement ±6 in EW North west wall relative motion top track 

47 NWWLB String Pot Displacement ±6 in EW North west wall relative motion bottom track 

Instrumentation South West Wall 

48 SWWD1 String Pot Displacement ±6 in Diagonal Diagonal south west wall 

49 SWWD2 String Pot Displacement ±6 in Diagonal Diagonal south west wall 

50 SWWVTW Potentiometer Disp uplift ±1 in V South west wall top west end (stud to concrete) 

51 SWWVTWi Potentiometer Disp uplift ±1 in V South west wall top west end (stud to track) 

52 SWWVBW Potentiometer Disp uplift ±1 in V South west wall bottom west end (stud to concrete) 

53 SWWVBWi Potentiometer Disp uplift ±1 in V South west wall bottom west end (stud to track) 

54 SWWVBE Potentiometer Disp uplift ±1 in V South west wall bottom east end (stud to concrete) 

55 SWWVBEi Potentiometer Disp uplift ±1 in V South west wall bottom east end (stud to track) 

56 SWWLB String Pot Displacement ±6 in EW South west wall relative motion bottom track 

Instrumentation South East Wall 

57 SEWD1 String Pot Displacement ±6 in Diagonal Diagonal south east wall 

58 SEWD2 String Pot Displacement ±6 in Diagonal Diagonal south east wall 

59 SEWVTE Potentiometer Disp uplift ±1 in V South east wall top east end (stud to concrete) 

60 SEWVTEi Potentiometer Disp uplift ±1 in V South east wall top east end (stud to track) 

61 SEWVBW Potentiometer Disp uplift ±1 in V South east wall bottom west end (stud to concrete) 

62 SEWVBWi Potentiometer Disp uplift ±1 in V South east wall bottom west end (stud to track) 

63 SEWVBE Potentiometer Disp uplift ±1 in V South east wall bottom east end (stud to concrete) 

64 SEWVBEi Potentiometer Disp uplift ±1 in V South east wall bottom east end (stud to track) 

65 SEWLB String Pot Displacement ±6 in EW South east wall relative motion bottom track 

Instrumentation South Central Wall 

66 SCWD1 String Pot Displacement ±6 in Diagonal Diagonal south central wall 

67 SCWD2 String Pot Displacement ±6 in Diagonal Diagonal south central wall 

68 SCWLT String Pot Displacement ±6 in EW Sputh central wall relative motion top track 

Instrumentation Monitor X 

69 MXAVCh Accelerometer Acceleration ±10g V Chanel instrumentation 

70 MXANSCh Accelerometer Acceleration ±10g NS Chanel instrumentation 

71 MXAEWCh Accelerometer Acceleration ±10g EW Chanel instrumentation 

72 MXAVb Accelerometer Acceleration ±10g V Bottom equipment instrumentation 

73 MXANSb Accelerometer Acceleration ±10g NS Bottom equipment instrumentation 

74 MXAEWb Accelerometer Acceleration ±10g EW Bottom equipment instrumentation 







5. Preliminary results 

5.1 UB-NCS performance evaluation and basic results 

This section presents preliminary results and observations obtained from the tests performed using the 

testing protocol scaled to 100% (design earthquake level). In the figures it can be observed the fidelity of 

the testing equipment to impose the desired floor motions. 

20 

15 

Desired Versus Observed Displacement at Bottom Level 

Observed 

Desired 

20 

15 

Desired Versus Observed Displacement at Top Level 

Observed 

Desired 

10 

10 


5 

0 

-5 


5 

0 

-5 

-10 

-10 

-15 

-15 

-20 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

Figure 55. Comparison desired and observed floor 

displacements for bottom UB-NCS level. 

1.5 

1 

0.5 

Comparison Desired and Observed Interstory Drift 

Desired 

Observed 

-20 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 


displacements for top UB-NCS level. 


0.2 

0.1 

0 

-0.1 

Relative Displacement Actuators Top Level 

Drift (in) 

0 

-0.2 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

Relative Displacement Actuators Bottom Level 

0.2 

-0.5 

-1 

-1.5 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

Figure 57. Comparison desired and observed 

interstory drift. 


0.1 

0 

-0.1 

-0.2 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

Figure 58. Relative displacement between 

actuators located at a same platform level. 







Acceleration (g) 

1 

0.5 

0 

-0.5 

-1 

Acceleration Histories Bottom Level 

-1.5 

0 5 10 15 20 25 30 35 40 45 50 

Time (sec) 

Acceleration Histories Top Level 

2 

Force (kips) 

20 

15 

10 

5 

0 

-5 

Forces in Actuators 

Actuator A 

Actuator B 

Actuator C 

Actuator D 

Acceleration (g) 

1 

0 

-1 

-10 

-15 

-2 

0 5 10 15 20 25 30 35 40 45 50 

Time (sec) 


accelerations. 

FRS (g) 

9 

8 

7 

6 

5 

4 

3 

2 

1 

Comparison DRS and ORS Bottom Level 

Desired 

Observed 

Observed 

Limit 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


Figure 61. Comparison desired (DRS) and 

observed (ORS) floor response spectra for bottom 

UB-NCS level 

-20 

0 5 10 15 20 25 30 35 40 45 

Time (sec) 

Figure 60. Comparison force histories in actuators. 

FRS (g) 

9 

8 

7 

6 

5 

4 

3 

2 

1 

Comparison DRS and ORS Top Level 

Desired 

Observed 

Observed 

Limit 

0 

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 


Figure 62. Comparison desired (DRS) and 

observed (ORS) floor response spectra for top 

UB-NCS level 

5.2 Partition walls performance evaluation 

Figure 63 shows the ensemble of hysteresis loops for the tests performed using the proposed testing 

protocol and the quasi-static interstory drift protocol. The forces in partition walls shown in the hysteresis 

loops in Figure 63 were calculated by subtracting the inertial forces from the forces observed in the 

actuators (Figure 60). The inertial forces were estimated from the recorded acceleration histories and the 

estimated mass of the platform and specimen. The time lag existing between recorded accelerations and 

actuator forces at high frequencies makes inaccurate the computation of forces in the walls at high testing 

frequencies. However, the impact of this time lag in the overall shape of the hysteresis loop is minimum 

because the drift imposed at those frequencies is close to zero. 







30 

Ensemble Histeresis Loops Emergency Room 

20 

10 

Force (Kips) 

0 

10% Protocol 

25% Protocol 

-10 

50% Protocol 

100% Protocol 

150% Protocol 

-20 

200% Protocol (QS) 




-30 

-3 -2 -1 0 1 2 3 

Interstory Drift (%) 

Figure 63. Ensemble hysteresis loops. Tests: 1-5, 10-13. 

5.3 Test observations 

During the tests performed considering the testing protocol scaled to 25 and 50% only minor damage was 

observed. Some hairlines cracks along corner beads and paper joint tape, rising of screws along top and 

bottom tracks and an incipient diagonal crack at the door fenestration were observed. 

During the test for 100% of the testing protocol (design earthquake level), at which a peak floor 

acceleration of 0.77g and a peak interstory drift of 0.87% were reached, more extensive damage was 

observed. Extensive cracks were observed along corner beads and joints between gypsum panels. The 

number of raised screws increased. Two monitors attached to the partition walls perpendicular to the 

direction of loading broke off their mounting systems and the monitor shown in Figure 35 moved out of 

its supporting system’s retention clip (without falling), the large light in the surgical lamp shown in 

Figure 5 broke off its support after exhibiting excessive displacement and hitting the UB-NCS columns 

several times. One of the drop-in devices attaching one of the medical gas piping hangers to the top 

concrete slab was pulled out from the concrete. The gurney exhibited excessive motion after deactivation 

of the breaking system. The crash test dummy was thrown off the gurney. Severe impact was observed 

between the video cabinet and the patient monitor shown in Figure 34. 

During the test for 150% of the testing protocol (maximum considered earthquake level), an even more 

extensive level of damage was observed. Extensive residual crack openings (~1/8-1/4”) were observed 

along corner beads and joints between gypsum panels. Evidences of rocking of screws attaching gypsum 

panels to the steel stud frame were observed. The crack in the corner of the door opening propagated. The 

monitor shown in Figure 35 fell down at this intensity (it was repositioned following the 100% level test). 

Noticeable residual deformations were observed in the medical gas pipes. Once again, the gurney 

exhibited excessive motion after deactivation of the breaking system and the dummy was thrown off the 

gurney and almost out of the room. All medical supplies on the medical cart shown in Figure 34 fell 

down. 







No damage was observed in the sprinkler system during the test series. Comparable damage in 

acceleration sensitive components was observed following the tests performed using the simulated floor 

motions for the aforementioned medical facility. Damage in displacement sensitive components could not 

be compared due to the preexisting damage in the specimen. 

6. Conclusions 

An innovative testing apparatus and testing protocol have been developed for real-time experimental 

seismic qualification and fragility assessment of full-scale distributed nonstructural components. The 

testing frame provides the unique laboratory capabilities to replicate full-scale floor motions expected at 

the upper levels of multistory buildings, allowing for the simultaneous testing of displacement and/or 

acceleration sensitive building nonstructural components and systems and equipment. Further, the seismic 

interaction between components can be evaluated. The proposed testing protocol can be expressed in 

closed-form for consistent generation of motions compatible with both a target design basis floor 

spectrum for acceleration sensitive systems and a generalized interstory drift for displacement sensitive 

systems as expected for a specified normalized building height. The testing protocol has been calibrated 

to induce/impose on acceleration/displacement sensitive nonstructural systems a number of rainflow 

cycles compatible with the number of rainflow cycles induced/imposed by the floor motions recorded in 

buildings during real earthquakes. 

A full scale test of a hospital emergency room with architectural finishes and medical equipment was used 

to demonstrate the capabilities of the UB-NCS and the suitability of the proposed testing protocol for 

assessing the seismic performance of combined displacement-acceleration nonstructural systems. The 

experiments successfully verified the testing capabilities of the UB-NCS to reproduce, in a controlled 

environment, the full-scale floor motions generating damage levels compatible with the damage levels 

observed in multistory building during real seismic events. 







References (To be Updated!) 

1. Retamales, R., Mosqueda, G., Filiatrault, A., Reinhorn, A.M. Experimental study on the seismic 

behavior of nonstructural components subjected to full-scale floor motions. in Proceedings 8th 

US National conference on earthquake engineering. April 18-21, San Francisco, California. 

2006. 

2. Mosqueda, G., Retamales, R., Filiatrault, A., Reinhorn, A. Seismic behavior of nonstructural 

partition walls subjected to full scale floor motions. in 2005 ANCER Annual meeting. Asian- 

Pacific Network of Centers for Earthquake Engineering Research. 2005. Jeju, Korea. 

3. Restrepo, J.I., Lang, A.F., Interim report to PEER advisory board. Year 7: Performance 

evaluation of gypsum wallboard partitions. Results of test 1. 2005, Department of Structural 

Engineering. University of California, San Diego: San Diego. 

4. Lang, A.F., Restrepo, J.I., Seismic performance evaluation of gypsum wallboard partitions. 

Proceedings 8th US National conference on earthquake engineering. April 18-21, San Francisco, 

California, 2006. 

5. Krawinkler, H., Parisi, F., Ibarra, L., Ayoub, A., Medina, R., CUREE-Caltech Woodframe 

Project Publication W-02: Development of a testing protocol for wood frame structures. 2000, 

Consortium of Universities for Research in Earthquake Engineering. 

6. Keller, D., Mosqueda, G. Nonstructural Component Simulator: Specimen Design and 

Preliminary Study of Damage Mechanisms. in 2005 Earthquake Engineering Symposium for 

Young Researchers. 2005. Peppermill Hotel - Reno, Nevada. 

7. Myrtle, R.C., Masri, S.F., Nigbor, R.L., Caffrey, J.P., Classification and Prioritization of 

Essential Systems in Hospitals under Extreme Events. Earthquake Spectra, 2005. 21(3): p. 779- 

802. 

8. SSMA, Product Technical Information. ICBO ER-4943P. 2001, Steel Stud Manufacturers 

Association.

Nonstructural Component Simulator University at Buffalo - MCEER ...

Create successful ePaper yourself

Delete template?

Save as template?