COMPORTAMENTO SISMICO DI EDIFICI BASSI IN LEGNO
 
 

Gli edifici a uno o due piani, realizzati con una struttura lignea leggera (platform frame),

 

 

Low-rise, light-framed wood buildings have performed

adequately in earthquakes provided they

acted as a unit, had adequate shear walls, and

were reasonably symmetric in plan and elevation.

This paper is the third literature review related to

the performance of wood structures in earthquakes.

The first [1] reviewed literature through

1984; it was based mainly on observations from

the 1964 Alaska earthquake (8.6 Richter) and the

1971 San Fernando earthquake (6.6 Richter). The

second [2] reviewed literature through 1988; it had

observations from two additional earthquakes -- the

1983 Coalinga (6.2 Richter) and the 1984 Halls

Valley (6.5 Richter). There has been much more

interest in research on wood buildings in recent

years as a result of observations from these earthquakes.

Additionally, the 1989 Loma Prieta earthquake

(on national TV, 7.0 Richter) has increased

interest and support for seismic research. This report

summarizes recent observations and research

on component and building response and discusses

current design philosophy.

PAST PERFORMANCE

Damage to wood buildings resulting from earthquakes

prior to 1988 has been documented in the

previous literature surveys [1,2]. In summary, the

primary cause of overall residential damage was

inadequate lateral support, particularly near large

openings such as garage or patio doors. The 1964

Alaska earthquake did limited damage to houses

with simple rectangular configurations, continuous

floors, and small door and window openings.

These types of buildings had symmetric box-like

lateral resisting systems. On the other hand. the

split level homes with large garage openings at

ground level were particularly susceptible to damage.

Their nonsymmetry made them vulnerable to

torsional as well as lateral motion. Residences

subjected to the Coalinga and Halls Valley earthquakes

failed due to inadequate lateral support

provided by short wood stud walls in the substructure

(cripple walls). There are limited observations

of commercial and industrial buildings during these

earthquakes; most literature describes residential

construction.

The American Plywood Association [3-5] and National

Institute of Standards Technology [6] reported

on damage from the 1989 Loma Prieta

earthquake. This earthquake resulted in 64

deaths, more than 3,700 injuries and $10 billion in

economic losses. Canadian literature [7,8] also reported

on the Loma Prieta in relation to what the

implications are for Vancouver, British Columbia

and for the Canadian building codes.

COMPONENT AND BUILDING RESPONSE

Floor and roof diaphragms and vertical shear walls

constitute the lateral load-resisting elements of timber

structures. Research prior to 1988 has been

reported [1,2] which related to the strength and stiffness

of horizontal and vertical diaphragms and

their connections.

Past research on shear wall strength and stiffness

was mainly experimental studies of plywoodsheathed

walls subject to a static load. Recent research

has included other types of sheathing, dynamically

loaded walls, and analytical methods.

Recent research on plywood-sheathed shear walls

includes input of the El Centro and Taft earthquakes

into both the analysis and experimental

verification of the walls' performance [9]. Dynamic

testing was also performed to determine the hysteretic

damping and period of vibration of a plywood

shear wall [10]. The load-deflection

characteristics of a nailed and glued plywood wall

San Fernando residences that were two-story and with openings were determined in Japan [11].

USDA, Forest Service, Forest Products Laboratory, One Gifford Pinchot Drive, Madison, WI 53705-2398 1

3

Experimental load-deflection characteristics were

also determined for gypsum- [12]. plasterboard-

[13], particleboard- [14], and waferboard-sheathed

walls [15].

The conclusions reached from the aforementioned

studies indicate any panel material will function for

shear walls provided the nail heads don’t pull

through the sheathing; however, the shear capacity

of gypsum and plasterboard is lower than plywood

or particleboard for the same nail spacing. Further,

the strength and stiffness of the shear wall is

more strongly related to the nail spacing than to the

sheathing material.

Several studies developed analysis techniques to

predict the dynamic characteristics of shear walls. Modeling recommendations are summarized.

Less research has been conducted on horizontal

diaphragms than on shear walls. A finite element

analysis was used to predict displacements for a

16 foot by 48 foot modeled diaphragm [18]. Both

finite element analysis and experimental verification

were done on three roof systems with gypsum

as the diaphragm material [19]. An overview paper

discusses horizontal diaphragms and gives guidance

on analysis, design, and details [20].

The characteristics of a whole building differ from

that of its components. There have been a number

of attempts to model an entire wood building subjected

to lateral loads. The nonlinear behavior of

connections and racking walls requires a nonlinear

analysis. Moody [21] presented a nonlinear structural

analysis model and compared it to experimental

results from Japan.

The Japanese have been especially active in both

analytical modeling and experimental verification.

Traditional Japanese post and beam buildings

have been analyzed to determine lateral displacements

using nonlinear finite element models [22]

and experimental studies [23]. Glulam frame structures

[24] and historic structures [25] have also

been analyzed and tested. The traditional post and

beam structures have included both diagonal bracing

and gypsum shear walls to increase racking resistance

[26].

US-type, light-framed buildings are also being

adopted in Japanese construction. One study determined

the increase in racking resistance to a

sheathed house by including the sheathing applied

to wall spaces above or below window and door

openings [27].

The Japanese are also studying multistory buildings.

One study was to load a two-story timber

building experimentally to understand the ductility

of the connection system combining tension bolts

and steel plates [28]; this system was compared

also to a nailed gusset system. Another study

found experimentally the distribution of shear

forces from a full-scale test of a three-story

sheathed building [29].

Dynamic characteristics have also been determined

experimentally from full-scale buildings. The

natural frequency and displacements were found

for a two-story post and beam building with diagonal

bracing [30]; the damping values and displacements

were found for a three-story glulam timber

building with shear walls of light-weight concrete

and cemented wooden chipboard.

Some research is also underway in Yugoslavia. A

general seismic response analysis was done using

the second order differential equations of motion

[32]. A full-scale, one-story truss frame house was

tested and the results for lateral displacements

were compared to a nonlinear finite element analysis.

DESIGN ASPECTS

Low-rise wood buildings are designed based on experience,

empirical methods, and theory for idealized

materials. Examples of buildings are given

based on the experience of 35 centuries from ancient

Greece [34]. A review of design requirements

are summarized for the Eurocode, and

codes of Canada, Yugoslavia, and Japan [35].

Several researchers have performed structural

analysis of wood buildings subjected to different

earthquake accelerograms of ground motions.

One study determined the response analysis of

one-, two-, and three-story dwellings subject to five

earthquake ground motions [36]. The results are

plotted in diagrams of maximum deflection vs

strength for the dwellings. Another study analyzed

the effect of different earthquake accelerograms on

the nonlinear loaddisplacement response for a

single-story portal frame building [37]. The effects

of differing moment-rotation in the joints were also

studied.

Energy dissipation and ductility in wood structures

have been studied by several researchers. One

study describes structural wood systems that can

develop ductility and gives recommendations for

details with nailplates, toothplates, or bolted connections

[38]. A seismic design procedure is recommended

to ensure that inelastic deformation is

confined to those components having ductile capacity.

Another study presents an example of how

to dissipate energy in connections of shear walls

constructed of screwed laminated veneer lumber

Post-earthquake repair requirements are outlined

in a Japanese manual to restore wood buildings

 

SUMMARY

This paper reviews literature since 1988 related to

the performance of wood structures in earthquakes.

The Loma Prieta earthquake, 1989, has

been the most significant and visual seismic occurrence

in the last several years.

Most recent research has focused on two areas.

The first relates to the strength and stiffness of

shear walls. Earlier research concentrated on

plywood-sheathed shear walls. Recent research

has studied other sheathing materials. The major

result is that most materials perform adequately in

shear walls; the strength and stiffness are more

strongly related to the nail spacing than to the

sheathing material. The second area relates to

analyzing and testing entire buildings. A number of

these studies have been completed; however,

there is not yet enough generality of results to impact

code design requirements.