00434 Iztok Sustersic

8/13/2019 00434 Iztok Sustersic

1/8

SEISMIC STRENGTHENING OF EXISTING BUILDINGS WITH

CROSS LAMINATED TIMBER PANELS

Iztok Sustersic1, Bruno Dujic

2

ABSTRACT: This paper deals with the issue of seismic retrofit and energy sanation of existing older buildings. A

possibility for solving both problems at once by applying a new outer shell made of cross laminated timber (crosslam or

XL) plates and effective aerogel insulation is presented. A seismic strengthening case study is presented on a 3 story

reinforced concrete frame building. Thermal insulation properties of the insulated panels are presented.

KEYWORDS:timber, seismic retrofit, energy

1 INTRODUCTION 123A majority of the buildings on seismically active areas

built before the nineteen sixties or seventies have two

major problems first they are seismically unsafe

because of the lack of seismic design codes at the time

they were built. And second, they have a high energyconsumption because of lack of insulation, proper details

etc. The proposed retrofit system deals with bothproblems at once; a new outer cross laminated timber

wall stabilises a building against horizontal shear forces

that are caused by earthquakes on one hand the timber

panels have a low mass and therefore dont contributemuch to seismic forces, but are very stiff on the other

hand [1] and provide high shear resistance. In addition

the new outer wall if combined with an effective

insulation provides a very good thermal insulation of

the building a combination of a 95 mm cross laminated

timber plate and 60 mm thermal Aerogel Spaceloftinsulation gives a U factor of 0,19 W/m2K [2]. Timber

panels also store CO2. A 170 mm outer shell (including

a facade) could therefore provide sufficient buildingthermal insulation and strengthen a reasonably sized

building against earthquakes. The new outer shell couldbe integrated with windows, doors and a facade already

in the manufacturing plant. Than the panels could be

transported to sight and rapidly attached to the building

(with proper detailing, subjected to earthquake

demands). The system is still in development but so far

seems to be most suitable for buildings up to 4 or 5floors, even floor plans, structures with stiff floor

1Iztok Sustersic, CBD d.o.o. , Lopata 19G, 3000 Celje,

Slovenia. e-mail: [email protected]. Bruno Duji, CBD d.o.o. , Lopata 19G, 3000 Celje,

Slovenia. e-mail: [email protected]

membranes and access to all outer walls. Another

positive aspect of the outer shell is that there are no

harsh interventions to a building and that people dont

have to move out during the construction phase (unlike

when using most of conventional methods for seismicretrofit). All together it makes a unique system that

solves two major problems in older buildings on

seismically active areas and therefore prolongs the

lifespan of constructions, contributing to sustainability.In the following chapters the system is presented more in

detail. Crosslam timber panels [1, 3, 4] are presented aswell as the Aerogel Spaceloft material [2]l. The

background of seismic analysis and the performance

based design N2 method [5] used for the evaluation of

the seismic resistance of the case structure are discussed.

Thermal properties of the outer crosslam shell combined

with Spaceloft are presented as well as a comparisonwith conventional insulation.

2 MATERIAL CHARACTERISTICS2.1 CROSS LAMINATED TIMBER PANELSCrosslam timber panels have been developed in Austria

in the late 1990s. The panels are glued together from

several (min. 3 and up to 9 layers for standard setups)

layers of spruce boards where each layer (Fig 1) runs

perpendicular to the neighbouring two (or occasionally

the most outer two layers run parallel to achieve greater

strength and stiffness in one direction).

Due to the high stiffness, strength and in-plane stability,

crosslam quickly gained popularity among architects due

to the possibilities the system offered in construction

design.


2/8

Figure 1:A 5layer crosslam panel

The first publications regarding the panels mechanical

properties and proposed methods for calculation were

published in the beginning of this century. The first

seismic tests were conducted in 2005 [1] and the firstmulti-storey (3-story) crosslam building (Fig 2) was

tested in full scale on a shaking table in Japan in 2006

[3] and another 7-story building in 2007. Unfortunately

to date crosslam is not yet covered with Europes current

standards and codes of practice [6, 7] for timber andseismic design.

2.2 AEROGEL SPACELOFTSpaceloft is a product of American Aspen Aerogels. It is

classified as one of the so called nano insulations. It is

basically a combination of a polymer fabric and the

silica aerogel. With the term aerogel we generally defy amaterial that is formed directly from a fluid gel in a

process that replaces the liquid in a gel with air. The

polymerisation of molecules in the solution forms

scattered nano-particles. Under the influence of catalysts

the particles form chains that create a net of nano pores.

In such state the liquid can no longer disperse though out

the structure. It is removed in a process of super critical

drying. The whole process is applied directly to the

polymer fabric and dried in the end.

Figure 2:Aerogel Spaceloft

Table 1:Aerogel Spaceloft technical characteristics

Margin Value Unit

Thickness (basic) 0,01 m

Density 150 kg/m3

Thermal conductivity 0,014 W/m K

Heat capacity 1046 J/kg K

Water vapour diffusion 4,51 Ng/Pa s m

Fire classification C -

Compression strength 70 kPa

Dynamic stiffness 23,7 MN/m3

Table 2: Comparison of basic technical characteristics ofinsulations

Aerogel

Density kg/m3 150

Thermal conductivity W/m K 0,014Heat capacity J/kg K 1,046

Water vapour diffusion Ng/Pa s m 4,8

Polistyren Rock wool

Density 15 200

Thermal conductivity 0,041 0,041

Heat capacity 1,26 0,84

Water vapour diffusion 25 4

3 SEISMIC RETROFIT CASE STUDY3.1 SPEAR STRUCTUREIn the paper a case study of seismic retrofit shall be

performed on a three-storey plan-asymmetric structure

(Fig. 3). The structure was conceived as a representative

of typical non-earthquake-resistant older constructions in

Southern European countries. It was designed for

vertical loads only, with the construction practice and

materials commonly used in Southern Europe in the

early 70s. This structure was pseudo-dynamically tested

at full-scale and analysed within the scope of the

European project SPEAR

Figure 3: The SPEAR structure [8]


3/8

The typical reinforcement in columns and beams is

shown in Figure 5. For the analysis in the paper the

structure was initially modelled without any additional

retrofitting systems. An elastic modal spectral analysis

and more important a nonlinear static (pushover)

analysis were performed. The latter serves as a basis for

the application of the N2 method a performance baseddesign method with which structural damage can be

assessed for different types of earthquake intensity. The

N2 method is discussed more in detail in the next

section.

Figure 4: Floor plan and cross section of the SPEARstructure analysed

Figure 5: Typical reinforcement in means and columns

3.2 SEISMIC ANALYSISCurrent codes of practice suggest two different

approaches for design of ductile structures in

earthquake-prone regions [9]. The first approach, well-

known and widely used [7], is referred to as the Force-Based Design (FBD) method since it mainly focuses on

designing the strength of the structure. The objective is

the evaluation of the behaviour factor q, which is

employed to transform the elastic demand spectrum into

an inelastic design spectrum. In this way a non-linear

structure can be designed using a linear-elastic static or

dynamic (modal response spectrum) analysis under

seismic action, with the structural ductility only

implicitly considered when evaluating the behaviour

factor q. The second approach, which explicitly refers to

the structural ductility in addition to the strength, is

based on a Non-linear Static Analysis (NSA) procedure.

The purpose of this approach is the evaluation of theactual structural response mainly in terms of ductility

demand and, hence, ultimate displacement induced in the

structure by the earthquake ground motion [4].

The NSA procedure is more complex than the FBD,

however it allows the designer to take into account the

actual dissipative behaviour of the structure.Furthermore, it can be used for Performance-Based

Design (PBD), where the design is achieved for different

performance levels such as no damage, limited structural

damage, important structural damage without collapse,

etc. Each level is generally linked to the structural

displacement by defining a damage index (i.e. for r. c.

beams and columns) and by assigning a limit value forevery performance level.

The NSA procedures are generally based on the

evaluation of the push-over curve, which represents the

response of the structure under a lateral loading

distribution schematising the seismic action. A numberof different methods have been proposed, including the

modified version of the Fajfars N2 method [5], which

has been adopted by the Eurocode 8 [7]. The aim of such

methods is the evaluation of the seismic displacement,

which is linked to the damage control of the structureand has to be kept below some reference values. The N2

method considers a performance point defined in terms

of both strength and displacement, where the structuralcapacity is compared with the demand of the seismic

ground motion. The base shear force and the top

displacement of a Multi-Degree-of-Freedom (MDOF)


4/8

system are first computed by means of a non-linear

Push-Over Analysis (POA) and then converted

respectively to the spectral acceleration and

displacement of an equivalent Single-Degree-Of-

Freedom (SDOF) system. The demand of the seismic

ground motion is represented through the response

spectrum in terms of pseudo-acceleration anddisplacement. Such an inelastic spectrum depends upon

the cyclic behaviour of the SDOF system and the

characteristics of the ground motion (peak ground

acceleration and shape), and can be obtained from the

elastic spectrum using suitable reduction factors. The N2

method was found to provide the best approximation

among various NSA methods for SDOF systems with

different hysteretic models and for MDOF systems [10],

however the N2 was never developed for the design of

timber buildings with specific hysteretic behaviour.

Therefore it must be noted that the results derived in this

study could be non-conservative, because hysteresis

loops with pinching, slip and strength degradation(typical for connections in timber structures discussed

more in detail in the following sections) dissipate less

energy than bilinear plastic loops with the same ductility.

Figure 6: Relation between R T

Nevertheless, it should be also pointed out that for both

analysed retrofitted setups, the SDOF systems equivalent

to the multi-storey building have periods longer than Tc

which is usually the value from where the reduction

factor (R) and ductility factor () are considered to be

the same (Fig. 6), regardless the type of hysteresis loop.The results of these analyses should therefore be

considered as a preliminary study aimed to investigate

the effect of different retrofit panel setups on the seismicresistance of the case reinforced concrete frame building.

3.2.1 Modelling of crosslam panelsCurrently Eurocode 8 [7] does not provide extensive

seismic design guidelines for timber buildings. Crosslam

structures are not even included in it as well as in the

Eurocode 5 [6]. There are, however, some papers dealing

with modelling of crosslam [1, 4, 11, 12].

The complex panel layout can be modelled using anorthotropic, homogenised orthotropic or homogenised

isotropic material, depending on the possibilities offered

by the FEM software.

Figure 7: Proposed [4] reduction coefficients formodelling crosslam wall panels

The proposed [4] homogenised-orthotropic-plane stress-

reduced cross section method, which is based on the

reduction of a multilayer to a single layer section using

the coefficients k3 and k4 in Figure 7 to modify the

stiffnesses and strengths, is precise enough for the needsof seismic modelling, where building behaviour mostly

depends on the behaviour of connections.By assuming a plane stress state, only two moduli of

elasticity (E0 and E90), one shear modulus (G12) and

one poissons coefficient (12) need to be defined. The

thickness of the panels (finite elements) remains thesame as does the shear modulus. If the adjacent boards

of individual layers are not glued along their thickness

(i.e. for KLH panels), a 10% reduction in the shear

modulus is suggested. The aforementioned method was

used to defy the crosslam panels used in this study. 120

mm (layers 40-40-40) plates were taken into account.

3.2.2 Reinforced concrete plastic hinge definitonIn the FE model used for the analysis, the plastic hingesthat form at the ends of beams and columns are set in

accordance with Eurocode 8, part 3, that deals with the

assessment and retrofit of buildings. The hinges in theFE model in SAP2000 software [13] and their plastic

rotation capacity is set in accordance with the following

expression:

0,0145 (0,25) ,(,)

, ,

,

25

(1,275) (1)

The parameters in the expression mean the following: el

for primary seismic elements is 1,8 is the normalised

axial force in an element, is the mechanical

reinforcement ratio of the tension and compression

longitudinal reinforcement, fcand fyare the compression

strength of concrete and the tension strength of

reinforcement, Lvis taken as half of an elements length,

h is the width of an element, is the confinement

effectiveness factor, sx ratio of transverse steel

parallel to the direction x of loading, fyw is the

tension strength of the shear reinforcement and d theratio of the diagonal reinforcement. An additional

reduction factor of 0,375 is used due to the use of

smooth reinforcement bars.


5/8

Figure 8: Definition of limit states in relplastic hinges

The plastic hinges are defined with a bil

rotation-bending moment. The crac

sections and a drop in load capacitcollapse (NC) state is neglected. The li

following: DL (damage limitation) pre

of reinforcement or the maximal elastic

cross section. At this point the plastic

remains 0. SD (significant damage) is

of the full plastic rotation of a cross(near collapse presents a full plastic r

section. A value of 3 um,pldefies the T

state, though in our study it is merely

only consider states up to NC.

3.2.3

Characteristics of connectionsconcrete frame and crosslam p

The connections are modelled on the

experimental response of a BMF 105 a

ten 60 mm long 4 mm diameter nails.

the response of the 3rd cycle is taken a

slip-force relationship in a non-linear l

in the FE model for connecting the R

retrofitting XL panel.

Figure 9: Calibration of the non-linear Fbackbone curve of the 3rd cycle of theresults of BMF105 brackets with 60 mmto shear force.

tion to bending in

inearised relation

ing in the RC

y after the nearmit states are the

sents the yielding

capacity of a r. c.

rotation (um,pl )

efined with 75%

section. The NCtation in a cross

C (total collapse)

theoretical as we

between theanels

slipshear force

ngle bracket with

The backbone of

the input for the

ink element used

C frame with the

EM link on thexperimentalnails subjected

Figure 10:A BMF105 angularpresumed XL panel r.c. plate cinside of the building is on the

The reason for picking theaccount the accumulated da

angular bracket. It is visible f

1st cycle yields about 30% hi

to the 3rd cycle. The connecti

the same response is model

horizontal direction. The cas

structure being retrofitted (thethe possibility of accessing

inside since the structure has

a case of a real residential co

probably have to be changed,

it from the outside.

3.2.4 Seismic resistance of tWith the use of the N2 metho

building in the X directions

The basic SPEAR structure,

an earthquake with peak grou

0,2 g has been retrofitted in toption was to try strengtheni

panels. A BMF 105 bracket

panel to the main structure at

That resulted in a not partic

the behaviour of the panels

the short leverages to the co

deformations. As a resultacceleration is raised by rough

hand the long panels with l

connectors resulted in an

allowable ground acceleratioeasier to assemble as the pan

doors as well as the facade

production plant and hence n

needed on the building.

buildings floor height is ov

transport can be necessary to

bracket (left) and theonnection (right). Theright side.

3rd cycle is to take intomage that occurs in the

om Figure 9, that i. e. the

gher peak strength oppose

on is simplified and hence

led for both vertical and

e study presumes that the

SPEAR construction) hasthe connections from the

o outer walls or infills. In

struction, the detail would

in order to allow access to

he SPEAR structure

d seismic resistance of the

is assessed.

hich could itself withstand

d acceleration of less than

o different ways. The firstg it with shorter crosslam

as used to attach the outer

every 30 cm of the panel.

larly stiff structure, since

as mostly in bending and

nections caused extensive

the allowable groundly 21 percent. On the other

nger leverages and more

almost 90% increase in

. Such a system is alsols allow the windows and

to be assembled in the

mayor additional work is

owever if the existing

r 2,95 m, an exceptional

et the panels on site.


6/8

Figure 11: The basic SPEAR structureboth cases of the seismically retrofittedmiddle the option with short panels andoption with long panels.

n the left andtructure. In the

on the right the

Figure 22:AD format elastic sspectra and capacity diagramsdifferent retrofitted constructiosystems written next to capaci

Table 3: Comparison of maxiacceleration with basic and ret

Basic structure

Retrofitted with short pane

Retrofitted with long panel

4 THERMAL INSULOF A CROSSLAM4.1 CROSSLAM PANELS

SPACELOFT

In Table 4 a comparison of d

attached to a crosslam wall i

thick XL wall panel was in

polistyren, 18 cm thick rock w

Spaceloft. Mass of a wall se

(U), phase shift (t), temperat

surface temperature (Tn,surf)

Table 4: Comparison [2] of difwall insulation

m [kg]

Polistyren 18 cm 65

Rock wool 18 cm 98

Spaceloft 18 cm 89

[-]

Polistyren 18 cm 117

Rock wool 18 cm 199

Spaceloft 18 cm 1340

The results show that when co

with insulation just on the out

of Spaceloft is primarily its s

rather small mass, the pha

ectra (for 0,3 g), inelasticfor the basic and the twos. Period values of SDOFy curves.

um allowable groundrofitted structures

Maximum

allowable

ground

acceleration [g]

0,197

ls 0,239

s 0,374

ON PROPERTIESACKET

INSULATED WITH

fferent types of insulation

s presented [2]. A 95 mm

sulated with 18 cm thick

ool and 6- and 18 cm thick

up (m), heat conductivity

re damping () and inner

re compared.

erent types of a crosslam

U [W/mK] t [h]

0,19 6,69

0,19 12,37

0,07 17,81

Tn,surf [C]

19,27

19,27

19,72

mparing a basic wall setup

r side, the main advantage

mall thickness. Due to its

se shift and temperature


7/8

damping are not as good as with conventional insulation.

However it still performs better than polistyren, though

the latter has an advantage in even lower mass (an

important issue for seismic retrofit).

But a choice between one of the insulations when using a

crosslam outer jacket would significantly be influenced

by the primary structure being retrofitted. If a heavymasonry structure is being considered, phase shift is not

a really a problem due to the thermal capacity of the

basic structure. In such a case one strives for a highly

insulating and thin outer shell (not to push the windows

even deeper into the building), which is exactly what a

Spaceloft setup can provide.

4.2 CROSSLAM PANELS COMPARED WITHOTHER TYPES OF WALL SETUPS

INSULATED WITH SPACELOFT

Zrim [2] has optimised each of the outer wall type setups

(brick, concrete, timber), depending on the position and

thickness of the outer and/or inner insulation. Thepreferred setups for each of the basic wall materials arepresented and evaluated in the following table. Mass

(m), heat conductivity (U), phase shift (t), temperature

damping () and inner surface temperature (Tn,surf) are

compared in Table 5.

Table 5: Comparison [2] of optimised wall setups fordifferent basic wall materials

m [kg] U [W/mK]

Masonry wall 418 0,15

Crosslam wall 83 0,09

Reinforced concrete wall 776 0,13

t [h] [-] Tn,surf [C]

14,68 2840 19,40

14,09 1644 19,64

12,04 4628 19,48

The crosslam setup was generally pointed out as the

most desirable, since it performed best on 3 out of 5

compared criteria. It came out second on phase shift and

finished last on temperature damping.

The crosslams low mass is extremely favourable forseismic retrofit, so we dont unnecessarily add more

mass to the structure being retrofitted and hence enlarge

the seismic forces.

5 CONCLUSIONSIn the paper we have presented a new system for seismic

retrofit and energy sanation of existing buildings. As far

as the preliminary studies show it is most advisable to

use longer (if possible) crosslam panels instead of

shorter segments. If shorter segments are used, it would

be advisable to join adjacent panels together on the

vertical sides as well, though this option has not yet been

explicitly analysed. As the crosslam jacket does not

influence the structures ductility, just its stiffness and

strength, it would be advisable to use stronger instead of

more ductile connections. The ductility of the basic

structure is still limited with the capacity of the primary

construction system. There is of course a possibility to

neglect the contribution of the latter and assume that all

load is transferred to the outer crosslam shell though

that option has not been investigated yet. The crosslam

panels can be upgraded with various insulation types.

The most efficient in terms of thickness is the AerogelSpaceloft. The small thickness is desirable when dealing

with existing structures with thicker walls. In that case

the temperature damping (which is a drawback of the

light Spaceloft system) does not present a problem any

more. Due to its light weight that does not contribute

much to seismic forces, polistyren presents an interesting

(and cheaper) option as well.

ACKNOWLEDGEMENTThe research support provided to both authors by the EU

through the European Social Fund 'Investing in your

future' is gratefully acknowledged. The kind cooperationof Mr. Zrim is also acknowledged.

REFERENCES[1] Dujic B, Klobcar S, Zarnic R., Influence of

openings on shear capacity of wooden walls,

Research report, 2005, University of Ljubljana and

CBD Contemporary Building Design Ltd, Slovenia.[2] Zrim G., Applicability test of new nanotechnology

thermal insulation spaceloft into building

constructions, graduation thesis. University of

Ljubljana, Faculty for civil and geodetic

engineering, 2009, pp. 112..

[3] Ceccoti A, New technologies for construction ofmedium-rise buildings in seismic regions: the

XLAM case, IABSE Structural Engineering

International, Special Edition on Tall Timber

Buildings 2008, 18(2), pp. 156-165.

[4] Blass HJ, Fellmoser P., Design of solid woodpanels with cross layers, 8th World Conference on

Timber Engineering, WCTE 2004, Finland; pp.543-548.

[5] Fajfar P, A nonlinear analysis method forperformance-based seismic design, Earthquake

Spectra 2000; 16(3):573-592.

[6] European Committee for Standardization (CEN),Eurocode 5 Design of timber structures Part 1-

1: General rules and rules for building,; 2004,

Brussels, Belgium.

[7] European Committee for Standardization (CEN),Eurocode 8Design of structures for earthquake

resistance. Part 1: General rules, seismic actions andrules for buildings, prEN 1998-1, Brussels, 2003.

[8] Rozman, M., Fajfar, P. 2009. Seismic response of aRC frame building designed according to old and

modern practices. Bull Earthquake Eng 7: 779-799.


8/8

[9] Chopra AK, Dynamics of structuresTheory andapplications to earthquake engineering, Upper

Saddle River, NJ: Prentice Hall; 1995.

[10]Fragiacomo M, Amadio C, Rajgelj S. Evaluation ofthe structural response under seismic actions using

non-linear static methods. Earthquake Engineering

& Structural Dynamics 2006; 35(12):1511-1531.[11]Jbstl R. A, Bogensperger TH, Schickhofer G. In-

plane Shear Strength of Cross laminated Timber.

2008. Meeting 41 of the Working Commission

W18-Timber Structures, CIB; 2008, St. Andrews

(Canada).

[12]Moosbrugger T, Guggenberger W, Bogensperger T.Cross-Laminated Timber Wall Segments under

homogeneous Shear with and without Openings.

2006. WCTE 2006 - 9th World Conference on

Timber Engineering - Portland, OR, USA.

[13]Computers & Structures Inc., SAP2000 -Integrated Finite Element Analysis and Design of

Structures. Computers & Structures Inc., Berkeley,California, USA, 2000.

00434 Iztok Sustersic

Documents