Production and Technology of Cross Laminated Timber (CLT):

A state-of-the-art Report

REINHARD BRANDNER
Univ.-Assistant, Deputy Director of the Institute 1) | Area Manager 2)
Institute of Timber Engineering and Wood Technology, Graz University of Technology 1)
Competence Centre holz.bau forschungs gmbh 2)
Graz, Austria

Summary
Cross laminated timber (CLT) has been developed into a worldwide well-known and versatile
building material. Progressive rates in production volume and distribution can be observed
currently. In fact CLT opens new horizons in timber engineering thanks to its laminar structure
which makes it well suited for use in constructions of that type which had been subject to mineral
building materials like concrete and masonry so far.
After a short introduction this contribution aims on demonstrating current production processes
used for rigidly composed CLT. In chapter 2 the process steps are individually described and
essential requirements as well as pros and cons of various production techniques are discussed.
Latest results of R & D and developments as well as innovations in production technology are
presented. In chapter 3 test and monitoring procedures in frame of the internal quality assurance,
known as factory production control (FPC), are presented. Thereby diverse regulations as
anchored in technical approvals for CLT as well as in the CLT product standard prEN 16351 [1]
are discussed. Additionally, some technological aspects of the product CLT together with a
comparison of geometrical and production relevant parameters of current technical approvals in
Europe are provided in chapter 4.
In the final discussion the main content of production and technology is presented in a condensed
way. An outlook in regard to current and future developments as well as concerning the ongoing
establishment of the solid construction technique with CLT is given. The product CLT comprises
an enormous potential for timber engineering, but also for the overall society. Standardisation and
further innovations in production, prefabrication, joining technique, building physics and building
construction makes it possible that timber engineering achieves worldwide success.

1. Introduction
Cross laminated timber (CLT) constitutes a plate-like engineered timber product which is
optimised for bearing loads in and out of plane. CLT is composed of an uneven number of layers
(in general three, five, seven or even more), each consisting of side-by-side placed boards (or
beams), which are crosswise arranged to each other normally under an angle of 90° and quasi
rigidly connected by adhesive bonding. Due to the continuous bonding and, consequently, the
quasi rigid composite action between the single layers, a very compact and versatile product
arises. As a consequence, the produced dimensions allow its application in form of large-sized
walls, floor elements and other large-sized load-bearing plane-like- as well as linear structural
components. In this way, modular dimensions, as known from light-weight wooden
constructions (e.g. frame system), can be neglected, due to the fact that window and door
openings can be freely placed. This product has opened new dimensions in timber engineering
and allows architects and engineers to design and realise monolithic buildings. This is now
possible in a manner and dimension which was subject to reinforced concrete, brick or other
mineral based building materials. Hence, this product opens up new vistas concerning a new
building technique, the so-called “solid timber construction technique with cross laminated
timber”, which makes it possible to design and construct with timber in so far unknown
dimensions and scales.
The first ideas and development date back roughly two decades. Motivated by a missing market
for the side-boards from sawmilling at that time a solid and in regard to swelling and shrinkage
in plane direction locked engineered timber product was developed. This locking effect caused
by crosswise arrangement of the single layers can be seen as an analogy to the single wood fibre
(tracheid) or to a composite of cells. In this way, every wooden cell constitutes a composite of
several cell layers winding around the cell lumen in varying crosswise fibre angle and, on their
part and in dependency on their function, shows a specific orientation of the cellulose fibres, the
primary constituent (total share of 50 ÷ 60 %) in (clear) wood and (structural) timber.
Meanwhile, the advantages of this specific orientation between the layers in regard to the load-
displacement and failure behaviour of the wooden cell composite but also in analogy with
artificial fibre composites have been well described (e.g. [2]; [3]; [4]; [5]). In the broader sense
CLT can be also seen as a synergetic product or as further development of historical timber
construction techniques of logs or staves, respectively, with their origins in Central and Northern
Europe. The combination of both principles to a composite with rigidly bonded crosswise layers
constitutes the substantial innovative part of the new solid timber construction technique in CLT
(see [6]).
The advantages of CLT as large-sized and panel-like solid timber construction element for the
building sector are in particular obvious because of its outstanding degree of pre-fabrication, the
dry and clean construction technique and the short erection times on site (e.g. roughly one to two
days per family house). The high dimensional stability underlines the fitting accuracy with
lowest tolerances, as already well-known for timber constructions in general. The opportunity to
transfer the loads two-dimensionally together with its low self-weight, which are both of
particular importance and predestine CLT for reconstruction and upgrading of existing buildings
(e.g. from Wilhelminian time), but also for resisting exceptional loadings (e.g. earthquakes), are
further decisive advantages of this product. In contrast to the light-weight timber structures (e.g.
framing, post and beam system), the merits of a clear separation of load-bearing from insulation
& installation layers, the low air permeability, the distinctive specific storage capacity for
humidity and temperature, the independence of modular dimensions in arranging window and
door openings as well as in fastening of furniture have to be outlined as well. The low mass, the
stiffness and the bearing capacity of this structural element against in plane and out of plane
stresses can be regarded as powerful arguments for its utilisation in multi-storey residential and
office buildings, in schools, single family houses, halls and the conversion and upgrading of
existing buildings and constructions, but also in wide-span structures like bridges. In particular
for wide-span structures rib floors or box beams, as a composite of CLT with linear timber
products, like (finger jointed) construction timber, duo or trio beams or glued laminated timber
(GLT, glulam), or constructions by means of folded panels are highly advantageous. Not at least
because of its versatile applicability, dynamic processes in development and establishment of
production capacities with growth rates of 15 ÷ 20 % per year have been observed (see Fig. 1).
These developments have been first realised in Austria and Germany with a current production
volume of roughly 500,000 m³/a (2012) and a share of two-thirds of the total worldwide
production volume sole in Austria. Worldwide activities in R&D as well as processes for
erecting (small & medium) production sites are ongoing and observable.
Although CLT seems to become a mass product on the first view, in reality selling is different
from products like GLT and, consequently a production of “standard” CLT elements in stock is
unimaginable. In fact, production and selling of CLT conditions a horizontal diversification at
the producing industry in terms of incorporating or integrating an engineering department which
itself acquires projects and provides technical support for customers (e.g. architects, civil
engineers, carpenters and builders). In that sense the production of CLT has to be on commission
with batch sizes of ≥ 1. Thus, the processes of cutting and joining have to be directly embedded
in the overall production process.
phase 1: phase 2:
niche product Í ideas | patents | prototypes pilot projects | market launch Î mass product

1st CLT multi-storey b. | AT

1st CLT multi-storey b. | DE
1st CLT residential b. | CH

1st nat. techn. approvals

first ETA for CLT

(KLH | AT, MDH | DE)

(2012: ≥10#)

1.0 MILL m³
2006
67 % AT

≈ 315 TSD
≈ 50 TSD
≈ 25 TSD

≈ 260 TSD
1998
1993

1995

25 % DE

≈ 215 TSD

0.6
8 % rest

WCTE’10 | ICSA’10 | promo_legno

laminated timber plates” PhD [G.

“6.GraHFT”, TUG | AT, Sofie, | IT

seminars | workshops, TUG | AT

structure of boards & beams

engineering”, CLT as web in

[G. Dröge | K.-H. Stoy | DE]

„Rigidly and flexibly jointed

PhD [T. Moosbrugger | AT]

at COST E5 | Venice | IT
„Basics of modern timber

R&D-programme at TUG
„branding of CLT timber”
structure”, SAH-Meeting
„CLT as slab- and plate-
„board-layered plates” –
„Timber roof structures“

“39.SAH-Meeting” | CH
PhD, [N. Lischke | DE]

stressed deck-bearing

PhD [A. Gülzow | CH]

multi-layer composite

PhD [P. Mestek | DE]

PhD [A. Jakobs | DE]
PhD [A. Scholz | DE]
[E. Cziesielski | DE]

PhD [R. Bosl | DE]

solid-web-girders

Schickhofer | AT]

FORECAST
[A. Steurer]

2010
2008

2015
2002
1981

1989

1994

2006
2003
2000

2004

2007
1974

1985

2005

2011
2009

2013

Fig. 1: Development of CLT – timeline ([7]; adapted)

For a maximum and reliable exploitation of CLT’s potential and its worldwide distribution it is
required to standardise CLT as much as possible and to extend its production, the design process,
handling and joining technology. In that sense international regulations in the frame of standards
which comprise all five areas “production & quality assurance”, “testing and evaluation”,
“design & verification”, “construction & assembling” and “joining technique” have to be
intended; this can be seen as the basis for a reliable distribution and handling of this product, its
technology and thus, forms the established solid timber construction technique. Of course,
current regulations are primary subject to the individual producers and given in technical
approvals, in Europe enforced by the DIBt (Deutsches Institut für Bautechnik, Berlin, Germany),
the OIB (Österreichisches Institut für Bautechnik, Vienna, Austria) or as European Technical
Approvals (ETAs), with reference to national or European standards. The developments in
standardisation within the last years in Europe, but also in Canada, United States and China
allow expectations in regard to a dynamical constitution and a foreseeable launch of required
standards.
This contribution aims on the topics “production & quality assurance” and on accompanied data
regarding main technological characteristics of CLT. It is intended to give an overview of current
productions with focus on the industrial scale as currently mainly established in Europe.

2. Production and processing of CLT: overview and step-by-step

The production process of CLT is in most steps largely comparable with the one of glulam. The
relevant steps for producing CLT are shown in Fig. 2.

(KILN DRIED) BASE MATERIAL (e.g. BOARDS)

ƒ strength (stiffness) grading

ƒ trimming
ƒ finger jointing

(FINGER JOINTED) LAMELLAS

ƒ (four-side) planing
ƒ edge bonding
ƒ (four-side) planing ENGINEERED
SINGLE-LAYER PANELS
ƒ surface bonding WOOD PRODUCTS

ƒ surface bonding

CROSS LAMINATED TIMBER (CLT)

Fig. 2: CLT production process: overview

Basically the production of CLT can be divided into the following steps: (1) strength or stiffness
grading of already (kiln) dried boards, (2) cutting out of local growth characteristics which do
not meet the requirements of the strength class and finger jointing of the residual board segments
to endless lamellas, (3) division and cutting of lamellas for later use in longitudinal and
transverse layers of CLT, (4 optional) adhesive bonding of lamellas to single-layer panels, (5)
assembling and adhesive bonding of lamellas or single-layer panels to CLT, and (6) cutting and
joining to structural elements (customizing). Within the next sections all relevant steps are
treated step-by-step and process parameters are discussed.

2.1 Characteristics of raw material and grading process

Normally, CLT is composed of boards with thickness tB = (12 ÷ 45) mm (cf. prEN 16351 [1]);
following current technical approvals a range of even (4 ÷ 80) mm, which comprises veneer and
beams, is given (cf. Tab. 5). In the context of standardisation and here focusing on construction
tenders, a general agreement on standard CLT layer thicknesses with tB = (20, 30, 40) mm could
be achieved in Austria. Further standardisation, in particular with regard to layups of CLT
elements optimised for stresses out of plane (e.g. for floors and roofs) and in plane (e.g. for
walls), is highly recommended. There is no upper limit for the board width but due to rolling
shear stresses in-between the CLT-layers a minimum width of wB ≥ 4 · tB is recommended and
thus anchored in current technical approvals. If this requirement is not kept or the distance
between relieves dR is too short (dR < 4 · tB), then a reduced resistance in rolling shear has to be
considered. Following prEN 16351 [1], the board width is regulated within (40 ÷ 300) mm. In
accordance to structural timber, the reference board width is proposed with wB,ref = 150 mm, as
given in EN 338 [8] and EN 384 [9]. In general, only boards of prismatic cross section are used
for CLT. In some cases profiling of the longitudinal edge may be meaningful, e.g. by tongue-
and-groove or special types of clearance profiles (e.g. [10]). In this course, the emersion of
adhesive is widely prevented and the possibility to create shadow gaps, which come up as
consequence of swelling and shrinkage, is included. Furthermore, a higher stability in top layers
composed of profiled single boards during pressing is given. Special emphasis has to be put on
the assurance that all laminations of the same layer in CLT are of equal thickness. This is to
ensure that during surface pressing all zones in CLT are exposed to the same transverse pressure
and, consequently, fulfil the requirements concerning maximum gap widths between the boards
of different layers. This has to be done in dependency of the used adhesive system, e.g. for one-
component polyurethane adhesive the bond line thickness has to be within (0.1 ÷ 0.3) mm.
Currently, mainly softwood species are used for CLT. The main species is Norway spruce (Picea
abies) in assortment with a small amount of White fir (Abies alba). Furthermore, softwood
species like Scots pine (Pinus sylvestris), European larch (Larix decidua), Douglas fir
(Pseudotsuga menziesii) and Swiss stone pine (Pinus cembra) are made use of, whereby the last
mentioned species are primary for CLT of high appearance quality and thus are used for the top
layers. The worldwide distribution of CLT and of the solid timber construction technique also
maintains the application of other species, e.g. of Maritime pine (Pinus pinaster) in
Sardinia / Italy. The utilisation of hardwood is of course also possible and has already been
made, e.g. within the project “massive_living” (three-storey building) in Brucknerstrasse, Graz /
Austria where one of in total 22 flats was realized completely with wall elements of CLT
composed of silver birch (Betula pendula). Further possible species are poplar (Populus spp.),
ash (Fraxinus excelsior) and others which are of economic interest, available in adequate amount
in sawmilling qualities and provide a minimum in mechanical properties as required for CLT as
structural load-bearing element. On one hand, hardwood species allow to set courses in
appearance. On the other hand, the systematic use influences and optimises especially the
mechanical potential of CLT. An increase in bending stiffness by means of in shear more rigid
transverse layers or top layers of higher bending stiffness (e.g. birch, ash, black locust) is
obvious as well as the improvement of the rolling shear resistance of CLT by means of species
like birch or poplar for the transverse layers, without increasing the CLT thickness or even by
reducing.

Fig. 3: Five-layered CLT element (schematically): homogeneous, symmetrical layup (left);

(multiple) double outer layers (middle); base material with relieves (right)
As required and perhaps in dependency of later additional functions of CLT (e.g. air tightness,
higher resistance against rolling shear, acoustic or haptic requirements) it is possible to substitute
single layers by laminar engineered timber products, e.g. laminated veneer lumber (LVL),
oriented strand board (OSB), plywood or single- or multi-layer solid wood panels. The
suitability of the substitute has to be verified, in particular if such a layer is explicitly taken into
account in load transfer.
The basic material for CLT is in general technical (kiln) dried and conditioned to a moisture
content of u = (12 ± 2) %. Following current technical approvals, a range of even
(8 ÷ 12) % ± 2 % can be found. Next, the material is visually or mechanically graded to strength
(stiffness) classes e.g. according to EN 14081-1 [11] or DIN 4074-1 [12]. It is of utmost
importance that within a single layer all boards have to be of the same grade. In other cases the
grade of the single layer shall be taken as equal to the lowest grade of the used boards. Currently,
the strength class system of EN 338 [8] is applied in Europe, although this system has been
developed for structural timber which is mainly stressed edgewise in bending. Common strength
classes are C24 for a homogeneous layup and, if combined with C16 / C18, for the transverse
layers. In fact, the bearing capacity of CLT stressed out of plane in bending is primary governed
by the resistance of the top layers in tension parallel to grain. The same situation, a demand-
oriented grading, is valid for glulam and, thus, the grading of boards according to requirements
in tension parallel to grain is recommended.
According to the glulam standard prEN 14080 [13] the grading in T-classes based on the
characteristic 5 %-quantile of strength and characteristic mean of E-modulus in tension parallel
to grain is suggested. Following the current technical approvals in Europe the product CLT is
primary composed of C24 according to EN 338 [8]. Of course many technical approvals allow a
specified amount (i.e. 10 % or 30 % per board layer) of boards dedicated to the next lower
strength class without considering the mechanical properties of CLT. It is assumed that this
regulation can be traced back to a rough interpretation of tolerances for visual grading given in
DIN 4074-1 [12] (section 6.3.2), in which a deviation of ≤ 10 % from grading criteria within
≤ 10 % of the material volume is allowed. There is a strong dissent concerning the statements
before the conclusion claiming that generally it would be permissible to mix boards of different
grades within one layer ([6]). Although not common today, in some cases, e.g. in case of a CLT
floor plate, which is primary stressed in bending, grading the board material rather in stiffness, in
conjunction with compliance of minimum requirements of strength, than in strength can be more
constructive. In fact CLT elements used in this way have to be rather designed on the subject of
serviceability (according the deflection or in cases of longer spans according the vibrations) than
in ultimate limit state. This is because of the transverse layers, which make CLT more flexible in
shear. Consequently, the optimisation of stiffness is an economical valuable target. Based on
several research works which addressed the homogenisation effect as a consequence of the
common action of boards in a (quasi) rigid composite (e.g. in glulam), it is well known that the
dispersion in strength properties of system products, like GLT or CLT, in comparison to that of
their base material is significantly reduced. The homogenisation is even more noticeable and
increases with increasing dispersion in strength properties of the base material (board; elements)
(see e.g. [14]; [15]; [16]; [5]). Due to the fact that the selectivity in visual grading is usually
lower in comparison to mechanical adjustment, a higher dispersion in strength properties of
boards is given in general. If these boards (elements) are joined to CLT (system), higher
homogenisation effects can be expected (e.g. in resistance against bending stresses out of plane
and against tension and compression stresses in plane). Consequently, stiffness grading
combined with a method assuring the compliance with minimum requirements of strength (e.g.
by exclusion of specific growth characteristics or by stressing of each board with a predefined
proof load) can be a very constructive and economic grading concept.
The layup of CLT is in general symmetrical. In cases in which additional layers are applied and
rigidly connected (e.g. acoustic panels) in order to keep dimensional stability in some
circumstances, it can be advisable to apply a counteracting layer which is in regard to swelling
and shrinkage equivalent to the additional layer(s). Normally CLT is composed homogeneously
of layers of equal strength properties. A combined but symmetric composition of layers with
different strength classes is possible, but requires special consideration in calculating the
mechanical properties of this CLT, e.g. by taking into account the rigid composite theory. Some
compositions of CLT which are optimised for example for stresses out of plane feature two or
more parallel layers than combined top-layers, applied to optimise the performance of CLT’s
bending stiffness by increasing the moment of inertia (see Fig. 3, middle).

2.2 Production of finger jointed lamellas

Based on longitudinal incremental results as output of strength grading local (discrete) growth
characteristics which do not meet the requirements of the strength class are cut out and the
remain board segments are joined again by means of finger joints (FJs). Thus finger joints
provide an economical approach for joining board segments longitudinally of which discrete
sections with a possible negative impact on the target strength distribution of the board sample
were removed selectively with the aim that the main part of the original board remains.
The finger joint itself constitutes a self-centring profile representing a folded scarf joint. Thanks
to the optimised profile, finger joints enable a simple, fast and form-fit connection between
elements by maximising the bond surface and minimising longitudinal losses of board material.
For CLT lamellas, profiles which have been already optimised and approved for the production
of glulam are used with a finger length of lFJ = (15, 20) mm. For an overview of these profiles
and of main geometric parameters see Tab. 1. Of course it is also possible to joint whole CLT
elements by means of large finger joints (LFJs) with finger lengths of lFJ ≥ 45 mm; this is
discussed in more detail in section 2.6.

Tab. 1: Finger joint profiles, geometric measures and loss in cross section
lFJ p bt bn lt α v(bn) lt lFJ
[mm] [mm] [mm] [mm] [mm] [°] [%]
15 3.8 0.42 0.52 0.5 5.6 13.6%
bn
bt

20 5.0 0.50 0.60 0.5 5.7 12.0%

20 6.2 1.00 1.11 0.5 6.0 17.8% α
p

50 1) 12.0 2.00 2.48 3.0 4.6 20.7%

lFJ … finger length; p … pitch; bt … tip width; bn … base width; lt … tip gap; α … flank angle; v(bn) … loss in cross section
1)
recommended profile for large finger joints (see EN 387 [17])

The position of finger joints can be edgewise (fingers visible on the side face; as common in
glulam) or flatwise (fingers visible on the narrow face) (see Fig. 4). The advantage of flatwise
finger joints is primarily in regard to a higher optical quality as no fingers are visible on the
plane surface of CLT. Additional advantages are given in regard to building physics, e.g.
airtightness.
The glued finger joint constitutes a quasi-brittle longitudinal joint between board segments
which are composed to endless lamellas. In cases where these lamellas are stressed in tension
parallel to grain, these stresses have to be primarily transferred by shear within the joint and
between the flanks. These shear stresses are optimal for bonded joints in general. Due to the loss
in cross section and the specific stress situation, finger joints have to be positioned within the
clear wood zone of boards, e.g. in a zone free of knots and apparent local or global grain
deviation. In doing so degrees of utilisation of the finger joint even higher than the ones of the
board segments (adherends) without joints are possible, although the cross section at the finger
tips is reduced up to (12 ÷ 18) % (see Tab. 1). The shear stresses at the flanks occur in interaction
with stresses perpendicular to grain. These stresses perpendicular to grain are minimised by
reducing the angle α. According to [18], the optimum angle would be α = 4° while a significant
reduction in strength can be already observed at α > 5.7°. Furthermore, due to stress
concentrations at the finger gap a ratio of lt / bt > 1.00 or at least of > 1.50 is proposed. More
details and further discussion as well as a literature survey can be found e.g. in [18], [19], [20],
[21], [22] and [5].

Fig. 4: Edgewise finger joint (left); flatwise finger joint (right)

The production requirements of finger joints are regulated in EN 385 [23] and prEN 15497 [24].
Some technical approvals for CLT also provide regulations in reference to DIN 1052 [25]. The
main requirements are: (1) finger joints have to be placed within the clear wood zone of
structural timber elements, i.e. free of local (e.g. knots) and global grain deviations and reaction
wood; (2) the suitability of the used adhesive system has to be assured (e.g. by technical
approvals, EN 301 [26] (type I) or EN 15425 [27]); (3) the technical requirements of using an
adhesive system (moisture content, temperature, applied quantity, possibilities of application,
bonding pressure, hold time, etc.), as regulated by the relevant standards as well as by the
adhesive producer, have to be met. In particular the bonding pressure and the applied quantity of
adhesive have to be adjusted to the timber species. Whereas a pressure which is too low may
lead to an insufficient wetting of adhesive on all flange surfaces and also to an adhesion which is
too low between the adherends during the transport of the lamellas, a too high pressure may
provoke unduly splitting of the adherends at the finger base or even of the whole jointed board
segments. Thus the longitudinal bonding pressure has to be adapted according to the parameters
(i) finger joint profile, (ii) timber species, (iii) moisture content of the adherends, and (iv) cross-
section dimension of the adherends. In general it can be said that the same regulations as already
given for finger joint connections occur in solid timber and glulam.
To assure an optimum performance of the finger joints, i.e. the suitability of the adhesive system
used to produce strong, stiff and durable joints, it is required to adjust the adhesive system to the
requirements of the structure and to the timber species. To minimise stress concentrations within
finger joints the application of adhesives, which show comparable elastic and shear properties as
the adherends, is recommended. In general and as mentioned before only adhesive systems
which are permitted to use in load-bearing timber structures, e.g. according EN 301 [26],
EN 15425 [27] or according technical approvals, are allowed. Currently mainly melamine-urea-
formaldehyde (MUF) and one-component polyurethane adhesives (1K-PUR) are used. Both
adhesives have an (nearly) uncoloured bond line, which in case of 1K-PUR is generally more
flexible but also more vulnerable against higher temperatures (T < 60°), if not modified
adequately. Both adhesives are also resistant against exposition to sunlight and humidity and also
against hydrolysis. The advantages of MUF can be seen in its higher resistance against high
temperatures (e.g. in case of fire) and its gap-filling and penetrating properties. Furthermore, the
curing process can be accelerated by increasing the temperature or by means of high-frequency
technology. Its disadvantages are the emission of formaldehyde, its limited storage stability (1K-
systems) and the strict mixing ratio of resin and hardener (2K-systems). In contrast, 1K-PUR can
be easily adapted to the individual production requirements, in particular in regard to its
reactivity and curing time. Polyurethanes are also free of formaldehyde and provide some
amount of internal pressure during bonding. Due to the increasing formation of gas cavities with
increasing bond line thickness, the tolerances in thickness of boards within the same CLT layer
have to be strictly kept.
In the course of shifting the focus to strength, it can be said that in general the performance of
finger joints stressed in tension parallel to grain primarily depends on the performance of the
joined timber elements, the adhesive and the quality of production. Thus the resistance of finger
joints is governed by the one of the weakest element. In timber engineering the resistance of
adhesive bonding has to be at least as strong as the one of the joined timber elements.
Consequently, a single finger joint connection can be reduced to a serial system of N = 2 joined
timber elements, in the context of strength constraint to the weaker one. Of course, although the
quality of joined timber elements is essential, the complexity of production parameters
influencing the joint performance individually is subject to each producer and production line
(e.g. bonding pressure, adhesive, moisture content, temperature, vibrations immediately after
pressing, curing time) making it impossible to model and rule objectively explicit requirements
of the finger joint performance. Consequently and in line with current regulations on finger joint
tensile strength for glulam according to EN 1194 [28], it is recommended to regulate minimum
requirements in relation to the joined timber elements and thus in dependency on the strength
class of each CLT layer (e.g. [29]; [5]).
In order to define minimum requirements of the tension strength parallel to grain of finger joints
the median strength and, thus, the failure probability of a series of n finger joints and that of a
series of m = (n + 1) board segments, which together comprise a lamella at reference length, are
kept equally, with ft,0,B,50,m = ft,0,FJ,50,n. In congruence with EN 1194 [28], a reference length
lL,ref = 2,000 mm of the lamella is applied. Examinations conducted on Central European glulam
lamellas showed an expectable distance between finger joints of E[dFJ] = [2.0 ÷ 2.5] m ([30]),
which is in good agreement with lL,ref. Thus on average and to be on the “safe” side, at least one
finger joint per reference lamella (n ≥ 1 = m – 1) has to be considered. The minimum
requirement of the 5 %-quantile of finger joint tension strength ft,0,FJ,05 depending on that of
board tension strength ft,0,B,05 at lB,ref = lL,ref = 2,000 mm can be formulated as
f t,0,FJ,05 ≥ ζ 05 ⋅ f t,0,B,05 . (1)

Based on extensive test experiences a coefficient of variation CV[ft,0,B] = (30 ± 10) % for board
tension strength parallel to grain can be expected. This range can be further divided into a sub-
range of CV[ft,0,B] = (35 ± 5) % in case of visually or mechanically graded boards in only two
(three) classes (including the class of reject), and into a sub-range of CV[ft,0,B] = (25 ± 5) % if
the boards are mechanically graded in more than two (three) classes (cf. [31]; [29]; [5]). In
regard to the tension strength of finger joints a range of CV[ft,0,FJ] = (15 ± 5) % is expectable (cf.
[29]; [5]). Based on an extensive data analysis and in congruence with EN 385 [23] and
prEN 15497 [24] the two-parametric lognormal distribution 2pLND is taken as representative
distribution for ft,0,B and ft,0,FJ. For reasons of simplicity both properties are modelled as being
independent of each other. Thus a very simple model approach can be formulated. Tab. 2
provides the minimum requirements of the finger joint tension strength based on the expected
ranges of CV[ft,0,B], CV[ft,0,FJ] = 15 % and n ≤ 2 (cf. also [5]). In this way, a very simple approach
of high practical relevance is given.

Tab. 2: Minimum requirements of finger joint tension strength parallel to grain

f t,0,FJ,05 ≥ ζ 05 ⋅ f t,0,B,05 ζ05 = 1.40 for CV[ft,0,B] = (35 ± 5) %

ζ05 = 1.20 for CV[ft,0,B] = (25 ± 5) %

Regulations for continuous internal as well as semi-annual external quality assurance can be
found in the technical approvals as well as e.g. in prEN 16351 [1], the European standard for
CLT. Further details on quality assurance procedures are discussed later in chapter 3.

2.3 Production of single layers (optional)

In general, the producers of CLT aim on reducing the width of gaps. This is achieved in respect
to building physic aspects (in particular in regard to fire design, airborne sound and airtightness)
but also in regard to joining technique, in particular considering pin-shaped fasteners like nails,
screws or dowels. A further reason can be found in aesthetics if the plane surface of CLT is left
visible in final use.
As a consequence, some CLT production lines create single-layer panels, which are further
cross-wise surface bonded to CLT as an intermediate step. These solid panels are applied to the
whole CLT or only to specific layers, e.g. to the top. In doing so, gaps can be completely
eliminated or at least reduced to gaps between the panels. A further advantage occurs in final
surface pressing to CLT. Since the surface of these panels is already smooth, equalised and of
course more precise in thickness than CLT composed of single boards, a lower surface bonding
pressure is required. Depending on the thickness of the single-layer panels and the used timber
species, both are relevant parameters to the expectable plate stiffness. Thus they are pertinent to
the resistance against pressing. A pressure as reachable in vacuum presses or by bracket, nail or
screw pressing can be already sufficient for an adequate bond quality. The suitability of each
pressing procedure, in particular in connection with the used adhesive system and the allowed
tolerances in glue-line thickness, has to be clarified and assured. For further details see
section 2.5. A further advantage of single-layer panels is the defined edge bonding between the
lamellas in contrast to an incomplete and undefined edge bonding which sporadically occurs
during edge and surface pressing of single boards to CLT. In particular in case of large-sized
CLT elements edge pressing is usually limited to layers with orientation in direction of
production. Therefore, these layers are overlapping of the transverse layers.
As already mentioned, a defined edge bonding has advantages in building physics, e.g. in regard
to fire, airborne sound and airtightness. In fact, numerous technical approvals for CLT allow a
ratio of wB / tB smaller than four, if the boards or laminations are edge bonded. Nevertheless, due
to internal stresses caused by swelling and shrinking as consequence of climatic variations the
occurrence of checks on the surface of CLT as well as within the CLT layers cannot be avoided.
Thus the advantages of edge bonding on building physics and the mechanical potential of CLT
can only be charged to a certain limit. As consequence of the quasi-rigid connection between the
boards, an irregular pattern of checks can be expected because of the fact that longitudinal
checks seldom occur in the bond line. In contrast, in CLT composed of layers of boards without
or only undefined edge bonding the swelling and shrinkage takes place between the boards and
results in a more regular pattern of gaps (see Fig. 5).

Fig. 5: Checks due to swelling and shrinkage in CLT with edge bonded top layers (left) and
without edge bonding (right)
During the production of single-layer panels the suitability of the adhesive system used for edge
bonding has to be assured in the frame of an internal and external quality control. Frequently
used and suitable adhesive systems are e.g. aminoplast-adhesives, according to EN 301 [26]
(type I; melamine-formaldehyde, MF and melamine-urea-formaldehyde, MUF), and one-
component polyurethane adhesives (1K-PUR), according to EN 15425 [27]. For more
information on adhesives and quality control see chapter 3.
With regard to producing single-layer wood panels as intermediate product in CLT production
there can be principally differentiated between three approaches:

2.3.1 Approach 1: single-layer panels gained by edge bonding of boards or lamellas

The used board material has to be adjusted to strength (or stiffness), e.g. according to EN 14081-
1 [11] or DIN 4074-1 [12]. Thus the same requirements as given in section 2.1 and 2.2 can be
applied. The four-sided planned boards of specific length, with or without finger joints, are
continuously joined to endless plates by edge bonding. These single-layer panels are
subsequently equalised and formatted according to the dimensions required for longitudinal and
transverse layers of CLT. Since these layers are subsequently composed to CLT, there are
normally no specific mechanical requirements of the edge bonding. So far the requirement of
wP ≥ 4 · tB, with wP as the width of the panel or the expectable distance between checks due to
swelling and shrinkage or distance between relieves, is fulfilled. This limit has been introduced
to prevent early failures in rolling shear and has been anchored in numerous technical approvals
for CLT.

2.3.2 Approach 2: single-layer panels according to EN 13986 [32]

The requirements according to EN 13986 [32] or of appropriate technical approvals are applied.
Subsequently it has to be ensured that the physical / mechanical properties, as required for
composing single layers to CLT of a specific strength class, are met. It has to be mentioned that
the mechanical properties of single-layer panels according to EN 13986 [32] are based on plate-
stripes. The strength grading as mentioned before in section 2.3.1 is not applied. Thus the
suitability and quality of the panels for the production of CLT has to be assured by the
implementation of an adequate internal and external quality assurance. EN 13986 [32] gives also
specific requirements of the shear strength of the edge bonding, which are regulated depending
on later service class.

2.3.3 Approach 3: single-layer panels gained by axial splitting of glulam

In this approach single-layer panels are gained by axial splitting of glulam which is normally
done by band resaws. In general only homogeneously composed glulam shall be used. The
strength grading performed on finger jointed lamellas of the glulam beam (e.g. according to
EN 14081-1 [11] or DIN 4074-1 [12]) is of course not transferable to the panels. On the subject
of the second approach in section 2.3.2 again an adequate internal and external quality assurance
procedure has to be set up to approve the adequacy of the panels for CLT production.
In a further step, independent of the approach and before pressing to CLT the single-layer panels
are in general planned or sanded. To guarantee that the whole surface area of CLT is under equal
and uniformly distributed pressure smoothness and the prevention of a minor tolerance in
average thickness (e.g. according to [33] of ± 0.15 mm) has to be maintained.

2.4 Application of adhesive for surface bonding

In general the guidelines and requirements of the adhesive manufactures have to be followed. It
has to be remarked that some parameters, like bonding pressure, quantity of applied adhesive,
moisture content of adherends and others are based on experiences made so far with glulam. In
the meantime some adhesive producers have adapted their regulations for CLT. In particular the
parameters bonding pressure and applied quantity are of relevance (cf. e.g. section 2.5.1). In
comparison to GLT the possibility that the applied adhesive is pressed out from the bond lines of
the plate-like product CLT is significantly reduced. This aspect requires consideration in
regulating the pressing conditions. This is also confirmed in section 2.5.1 referencing [34]. In
this work very good and sufficient bond qualities have already been reached by applying the
minimum of by adhesive producers proposed quantity of adhesive per square meter.
The application of the adhesive to surface bonding is normally carried out mechanically and
without contact (i) on single lamellas in a continuous through-feed device or (ii) on CLT layers
already pre-positioned in a positioning or press bed. A line-wise discrete application of adhesive
is preferred.

2.5 Composing and pressing of boards or single layers to CLT

2.5.1 Some general comments and recommendations for bonding pressure

In general CLT can be composed of flexible connected boards, lamellas or layers, e.g. by joining
them crosswise by means of ring-shank nails (e.g. [35]), hardwood dowels (e.g. [36]) or
hardwood screws (e.g. [37]). Of course, more generally CLT is composed of quasi-rigidly
connected lamellas or layers, i.e. by surface bonding. This contribution focuses on quasi-rigidly
composed CLT elements.
Depending on the pressure device the following differentiation can be made:
(1) surface bonding by means of hydraulic press equipment;
(2) surface bonding by means of vacuum press equipment;
 (3) surface bonding utilising the pressure of screws, brackets or nails.
Depending on the device, bonding pressures of (0.10 ÷ 1.00) N/mm² and even higher can be
provided by (1) a hydraulic equipment, whereas vacuum presses (2) and pressing with screws,
brackets or nails (3) attain bonding pressures in the range of (0.05 ÷ 0.10) N/mm² and
(0.01 ÷ 0.20) N/mm², respectively (cf. e.g. [38]).
Of course regulations of an ideal surface bonding pressure for CLT are still missing. Thus a more
general discussion for further clarification is provided. Generally, in cases of suitable surface
condition of flatness and of only negligible deviations in thickness of the adherends within and
between the single layers for CLT production theoretically no bonding pressure is required. In
practice there is always some roughness on the surface, warp and twist of the adherends and
deviations in the thickness of the adherends. Consequently, a minimum bonding pressure is
required and has to be regulated at best in dependency on the main parameters. The demand on a
minimum bonding pressure can be also made due to the adhesive application system. For
example, in case of line-wise application a specific minimum pressure securing a complete
wetting of the surface is required. Thus the main parameters defining the requirements of a
minimum bonding pressure are specified to ensure (i) a complete wetting and (ii) a defined
permitted bond line thickness. In regard to (ii) types of adhesive differ first in terms of close
contact adhesives and gap-filling adhesives and secondly in means of swelling adhesives (e.g.
polyurethane adhesives) and shrinking adhesives (e.g. aminoplast- and phenoplast-adhesives).
For ensuring a defined bond line thickness the two parameters applied, quantity and bonding
pressure, have to be borne in mind. According to [38] and due to swelling and shrinking, the
minimum bonding pressure in case of phenol or melamine based adhesives is in general in the
range of (1.40 ÷ 2.00) N/mm² (!), whereas for polyurethanes (0.01 ÷ 0.10) N/mm² should be
sufficient. Following the regulations for the production of glulam in EN 386 [39] and
prEN 14080 [13] a surface pressure of 0.60 N/mm² for tB ≤ 35 mm and (0.80 ÷ 1.00) N/mm² for
35 mm < tB ≤ 45 mm thick lamellas without relieves is given. The regulations depending on the
lamella thickness take account of the resistance of lamellas against deformation, i.e. against
longitudinal and transverse bending deformation and torsion.
Besides the minimum also the maximum allowable bonding pressure requires regulation. It is
known that a too high pressure causes damage of the adherends’ surfaces ([40]), e.g. by crushing
the cell structure. Consequently a reduced adhesive penetration and shear resistance ([41]) can
be observed in combination with an increasing possibility of exorbitant squeezing out of
adhesive from the bond line, which causes insufficient bonding. Thus the bonding pressure has
to be also regulated in dependency of the timber species. This is of vital importance on the one
hand due to the expectable resistance against torsional and bending deformations and on the
other hand due to the differences in the stress-strain curves in compression perpendicular to
grain and hereby in particular in regard to the proportional limit of the linear-elastic course.
According to EN 408 [42], on the basis of compression perpendicular to grain tests
accomplished on Norway spruce (Picea abies) following statistics can be found:
fc,90,mean = (3.0 ÷ 3.4) N/mm², fc,90,05 = (2.2 ÷ 2.5) N/mm² and a coefficient of variation
CV[fc,90] = (22 ÷ 28) % ([43]). The proportional limit can be estimated to be at 30 % of fc,90
(according to [44] roughly at 50 %). Thus, to widely (≥ 95 %) prevent plastic deformations on
the surfaces of adherends, the upper limit of bonding pressure can be estimated with
1.10 N/mm². In [44] it was found that the cell structure in Norway spruce starts to be damaged
already at 0.60 N/mm² and 1.00 N/mm², respectively, in cases of horizontal and vertical annual
growth rings, while a decrease in shear strength was already noticed at a pressure of
≥ 0.40 N/mm² and ≥ 1.0 N/mm². Following [45] in Norway spruce the internal pressure has to be
limited to ≤ 1.0 N/mm². Thus the externally applied pressure needs to be adapted to the adhesive
system, no matter whether it swells or shrinks, and to the swelling pressure of the timber itself as
consequence of surface wetting. Of course, as the annual ring pattern for products like glulam or
CLT is not restricted in case of Norway spruce or comparable timber species, the surface
bonding pressure has to be limited to ≤ (0.4 ÷ 0.6) N/mm².
To summarise in brief, the required surface bonding pressure can be defined as function of (i) the
adhesive system, (ii) the timber species, (iii) the geometry of the adherends in regard to
roughness and flatness of the surface and allowed tolerances in thickness, (iv) the adhesive
application system, and (v) the applied quantity of adhesive. The applied quantity itself depends
on the roughness of the adherend’s surface and consequently on the timber species (for example:
ringporous vs. diffuse porous hardwoods).
For clarifying the consequences of the interacting parameters bonding pressure and applied
quantity on the CLT production a comprehensive research project was conducted (cf. [34]). In
the context of this project two types of 1K-PUR adhesives, three bonding pressures of (0.1, 0.3,
0.6) N/mm² and various applied quantities, which were defined to secure a complete wetting in
unidirectional surface bonding in dependency of bonding pressure, were investigated.
Additionally, the effect of cyclic climatic variations (20 °C / 90 % rel. humidity and 30 °C / 40 %
rel. humidity; quantities of cycles: 0, 10, 21, 25) on the properties of bonding was also analysed.
This research was performed on under specific conditions in laboratory produced three-layered
CLT elements of Norway spruce of strength class C16 and C24 according to EN 338 [8]. The
surface bonding properties were investigated by means of rolling shear tests on whole CLT
elements in bending according to EN 408 [42], block (rolling) shear tests on the single glue line
according to EN 392 [46] and delamination tests according to EN 391 [47], approach B. To
summarise, although the applied adhesive quantities were generally on the lower limit of
producer’s recommendations, the investigated bonding pressures were found to be sufficient to
realise adequate bond qualities. So far the dispersion in thickness between boards of the same
CLT layer was kept low. The relevance of this demand can be easily demonstrated by a simple
calculation example: With Hook’s law for linear elastic material and with the parameters for C18
according to EN 338 [8] with Ec,90,mean = 380 N/mm² it can be shown that with regard to
compression a 40 mm thick board by 0.10 mm an average surface pressure of 0.95 N/mm² is
already required. In contrast to this, parameters like warp or twist of the board material showed
nearly no or at least negligible effects on surface bonding. This is because of the relatively low
E-modules longitudinal and transverse in bending and the G-module of timber in torsion. Of
course, a positive relationship between bonding pressure and shear strength was observed in
cases where the applied adhesive quantity was beneath the recommended quantity and thus was
too low or the deviations in thickness were too high.
Hence, it can be stated that also for CLT composed of single boards a low pressure can be
sufficient, as far as a very strict and small tolerance in the thickness of lamellas within one CLT
layer is ensured. Therefore the parameter board thickness should be controlled in frame of the
internal quality assurance procedures. Following the calculation example mentioned before and
in regard to the thickness range of CLT lamellas, it is highly advisable to make use of a tolerance
of ≤ (± 0.10) mm, although this demand is stricter than that meanwhile partly anchored in
technical approvals for CLT (cf. e.g. [33], [48], [49]). This strict tolerance is also argued
concerning the requirements of the allowed bond line thickness of e.g. (0.1 ÷ 0.3) mm in case of
1K-PUR.

2.5.2 CLT production by means of hydraulic press equipment

By means of hydraulic press equipment it is possible to provide nearly every desired surface
pressure. Of course current systems provide an upper limit of 0.8 (1.5) N/mm², which makes
them flexible enough to extend the productions also to thicker and thus stiffer single-layer
elements and also to hardwoods. In contrast to vacuum facilities or nail, bracket or screw
pressing, it is possible to provide specific edge pressure, e.g. solely on the top, transverse or on
individual layers for assuring homogeneously closed gaps. Of course hydraulic press systems are
normally restricted to produce even CLT elements without curvature or of other shapes. As a
consequence of parallel pressing surfaces, it is also not possible to balance local unevenness or
deviations in thickness as it is for example provided in a vacuum press.

Fig. 6: Placing and aligning of the single layers (left); positioning of layers and application of
adhesive (right) (© Minda-Industrieanlagen GmbH / DE)

Fig. 7: Hydraulic surface and edge pressing device (left); unloading of ready produced CLT
(right) (© Minda-Industrieanlagen GmbH / DE)
A great advantage of hydraulic facilities is their flexibility in regard to automation of process
steps before, during and after pressing. This comprises the positioning and alignment of single
boards or layers, the application of adhesive, the conveying into and out of the press, the
application of specific edge pressure and the pressing itself, in which differentiation has to be
made also in regard to the adhesive system and the curing process (cold, hot or with high
frequency). Depending on the production volume and market orientation modular processing
units with different degrees in automation are provided by some press producers. For example
Minda-Industrieanlagen GmbH / DE offers a press system (cf. Fig. 6 and Fig. 7) with stages of
expansion semi-mechanically equipped with three press cycles per shift up to twelve press
cycles per shift in a fully mechanical processing. Consequently, it is possible to move step-by-
step into the CLT market.
Depending on the CLT production, further differentiation can be made in press facilities for
small CLT elements, which are further connected by large finger joints, and large-sized CLT
elements with a dimension of up to l / w / t = ≤ 18.0 m / ≤ 3.5 m / ≤ 400 mm. In dependency on
the required production volume and the adhesive properties, single or multiple CLT elements can
be produced in one press cycle. Also based on the production volume and the degree of
automation, it may be meaningful to adapt the adhesive system to allow for example one press
cycle every 40 minutes.
Further differentiation can be made in the production of CLT elements with or without edge
bonding and with or without door and window openings. In the last case the adhesive application
system has to be adapted to omit these openings. In regard to the press system itself further
differentiation is possible in (i) fixed press facilities and moving CLT elements (e.g. Minda-
Industrieanlagen GmbH / DE, Springer Maschinenfabrik AG / AT, Kallfass GmbH / DE) and in
(ii) horizontally displaceable press facilities as e.g. constructed by Fr. Leiße & Söhne GmbH &
Co. KG / DE. The productivity of current press systems allows a CLT production volume of
25,000 m³ per year and shift.
A comparable press facility is provided by Ledinek Engineering d.o.o. / SI. Here a pneumatic
press system is combined with tie bars.

2.5.3 CLT production by means of vacuum press equipment

Another possibility to produce CLT as rigid composite provided by adhesive bonding is to press
single boards or layers by vacuum (cf. Fig. 8). In doing so a pressure of (0.05 ÷ 0.10) N/mm² can
be reached. Thus specific and strict requirements of the surface quality, in particular on evenness
and minor tolerances in thickness, have to be met to assure an adequate bond line quality.
Also limits in warp and twist have to be considered. On the subject of reducing the stiffness
against bending and torsion deformations, relieves are made longitudinal to the grain and,
according to prEN 16351 [1], not deeper than 90 % of the lamella thickness and not wider than
4 mm in width. Due to the fact that, as already mentioned, also in this case the compliance of the
ratio dR / tB ≥ 4 has to be kept or reduced, rolling shear strength of CLT needs to be considered.
Because of the limited bonding pressure also a limitation in processable layer thickness or timber
species can occur. In regard to the requirements of adherends’ surface, evenness and thickness
tolerances usage of single-layer panels can be advantageous.

Fig. 8: Vacuum press in combination with an adhesive application system (© woodtec

Fankhauser GmbH / CH)
Overall vacuum press facilities allow an economical surface bonding and are well suited for
small and median CLT productions with a production capacity of (2,000 ÷ 5,000) m³ per year
and shift. The processing itself is in general semi-mechanical. The boards or layers are placed
manually whereas adhesive application is operated semi-mechanical. It is also possible to
compress the top layers by means of lateral pressing bars before the air in the airtight synthetic
rubber foil is evacuated and the CLT is compressed homogeneously. This homogeneously
distributed pressure in principle enables to produce also curved or general shaped CLT elements
and offers also the production of composite elements like box-beams or rib floors. Also local
thickness deviations can be balanced to some amount and sufficient bond qualities achieved.

2.5.4 CLT production by means of bracket, nail or screw pressing

Another alternative approach for producing CLT as rigid composite is to provide the bonding
pressure discretely by nails, screws or brackets. The herein commonly achieved pressure
amounts to (0.01 ÷ 0.20) N/mm² and thus is comparable with vacuum press facilities.
Consequently, the same requirements of the base material used for composing CLT can be
applied, cf. section 2.5.3. The advantage of this approach is given by the minimum efforts and
investments necessary to produce CLT. Of course, in comparison to a flexible composed CLT,
e.g. in which the layers are solely connected by nails, screws or brackets, knowledge and
experience about adhesive bonding is required to assure a proper production.
To prevent damage of tools which are used later in cutting and joining processes it is advisable
to use aluminium instead of steel fasteners. Following e.g. [50] aluminium brackets can be found
as being appropriate. To achieve a sufficient and widely homogeneously distributed bonding
pressure it is required to regulate the spacings between the discretely placed fasteners. Therefore
the rules for screw pressings in DIN 1052 [25] can be used as basis. According to this standard,
only self-tapping full-threaded wood screws with a nominal diameter d ≥ 4 mm are allowed. The
maximum area allocable per screw is A ≤ 15,000 mm² and the maximum spacing amounts to
150 mm. The thickness of structural timber used for each layer is restricted to t ≤ 35 mm,
whereas the use of engineered timber products according to DIN 1052 [25], section 14.1 (4) up
to t ≤ 50 mm is allowed.

2.6 Large finger joints between CLT elements

An alternative production process of CLT involves producing small CLT elements in single or
multi-layer cycle presses first and afterwards joining them to larger CLT elements by means of
large finger joints. However, it is also possible to use large finger joints as connections in order
to join already large-sized CLT elements or cut-outs from door or window openings. These large
finger joints comprise the whole cross section of joined elements. The advantage of producing
small elements, e.g. as done in [48], is given by the small-scaled press and the much smaller
forces as well as in the handling of the elements before and after pressing. A common and
standardised profile for large finger joints is defined with lFJ = 50 mm, as presented in Tab. 1.
The production requirements are based on the experiences gained with glulam and are given in
reference to EN 387 [17] and section 3.5. As consequence of the joint, the bending strength of
CLT elements at that position has to be reduced on the basis of a 25 % reduction of the
characteristic 5 %-quantile bending strength of the base material (see e.g. [48]). Of course, in
case of adequate planning processes this does not or only negligibly restrict the design process.
2.7 Finish of standard CLT elements
After pressing, standard CLT elements are trimmed on their edges. The surface of the elements
after pressing is treated differently, without further processing by planning or sanding.
Depending on later use, also the application of additional non load-bearing layers like OSB,
acoustic panels, gypsum plaster boards or three-layered solid wood panels is possible (cf. also
section 2.1). These additional layers are primarily connected by surface bonding.

2.8 Cutting and joining: customising

Cutting and joining of CLT elements immediately after production and finishing constitute
essential and logical process steps in an order-related, small (single) batch production. It is the
aim to continue the precision achieved in production also in cutting and joining. Approved
devices are portal machines which operate as multiple processing centres (e.g. of Hans
Hundegger Maschinenbau GmbH / DE). After the CLT element has been aligned accurately,
these machines accomplish all relevant processes for dimensioning and further joining, like
trimming, cutting, milling (e.g. for connection technique, stepped rabbet or profiling of edges for
later joining of e.g. ceiling elements, for installation channels, etc.), drilling (on both surfaces
and all edges up to 2 m in depth from one side) on both surfaces (top and bottom) and all four
edges (cf. Fig. 9), including marking and labelling. The tools (moulding cutters, saws, chain
saws, etc.) provided in tool magazines are available on time. Thus large-sized CLT elements in
thickness up to 350 mm, in length up to 16 m and in width up to 4.3 m can be readily processed
to wall, floor and ceiling elements. Another advantage of these processing centres is the
possibility to encase the device for minimising emissions of noise, dust and chips whereby dust
and chips as by-products can be collected in concentrated form.
Depending on the CLT production volume and on the market orientation, for securing a
continuously running production it can be meaningful to operate more than one processing
centres parallel. Therefore process centres in various dimensions and configurations are
available and systems are provided, which allow step-by-step adaptation on production volume
and market demands. In dependency on the required flexibility, three- to five-axis machining
centres are available. Of course not only the processing centres, but also the combination with
software packages for optimising the layout and thus the degree of utilisation of CLT elements
together with a well-operating process planning office create an economical and powerful
customising centre and add value to the product CLT. Meanwhile also customising centres
without own CLT production have been established in combination or cooperation with
carpentries or assemblers.

Fig. 9: Machining centre: moulder (left), chain saw (middle), saw (right) (© Hans Hundegger
Maschinenbau GmbH / DE)
In regard to the assembling on site it is essential to optimise the logistics and to load of the
elements after cutting and joining e.g. on trucks in inverse order to that required later on site.

3. Factory production control (FPC) – internal quality assurance

The aim of this chapter is to present the main internal test and monitoring procedures as required
and regulated within technical approvals for CLT as well as within prEN 16351 [1] as far as the
factory production control (FPC) is concerned. The following sections individually give
(minimum) requirements for each quality criteria whereby the requirements of prEN 16351 [1]
are treated first. Consequently, additional processes for quality assurance and monitoring as
partly given in technical approvals of Germany and the European Technical Approvals are
briefly presented.
Complementary to FPC an external quality control by an independent accredited institution is
generally required semi-annually. Thereby the conformity of production and monitoring
processes according to the underlying guidelines, approvals and standards is assessed. These
institutions are also responsible for initial type testing and the determination of some process
parameters (e.g. declared strength values for (large) finger joints, etc.). These external test and
evaluation procedures are not part of this chapter.
Within the frame of FPC it is also required to establish an internal guidance procedure for quality
control. This procedure should provide regulations and responsibilities for testing and
monitoring of production processes and in particular of actions in cases in which test results are
conspicuous or do not meet the requirements.

3.1 Control of climatic conditions during production

To ensure an orderly bonding the requirements given by the adhesive manufacturer, i.e. in regard
to temperature of adherends and surrounding, the relative humidity and moisture content of the
adherends, the applied adhesive quantity, the time schedule, the bond pressure, etc. have to be
met. The prEN 16351 [1] recommends some of these parameters as general minimum conditions
for the production of CLT, e.g.
ƒ during bonding: T ≥ 15°C and rel. humidity (40 ÷ 75) %;
ƒ during curing: T ≥ 18°C and rel. humidity ≥ 30 %;
ƒ moisture content of adherends u = (6 ÷ 15) % (≤ 18 % in case of preservative treatment);
ƒ maximum difference in moisture content between two parallel layers Δu ≤ 5 %.

3.2 Delivery control of adhesives

According to prEN 16351 [1] it is required to control every dispatch of adhesive in regard to
quality and suitability for the production of CLT or for a specific process step (e.g. finger joints,
edge and surface bonding). Additionally, the adhesive system used for large finger joints has to
be controlled in every shift it applies. The adhesive systems which are principally allowed for
use in CLT production according to prEN 16351 [1] are:
ƒ phenoplast- and aminoplast-adhesives according to EN 301 [26], type I or according to
technical approvals which certify the appropriateness of the adhesive system for load
bearing timber structures and in particular for CLT for use in service class one or two;
these adhesives (primary MUF) are principally applied to bonding of (large) finger joints
as well as to edge and surface bonding; if used for (large) finger joints the minimum
 holding time for longitudinal pressure and the mixing ratio of synthetic resin and
hardening agent have to be monitored in addition; for large finger joints the applicability
is additionally limited to adhesive systems which are certified for bond line thicknesses
up to 1 mm;
ƒ one-component polyurethane adhesives (1K-PUR) according to EN 15425 [27] or
according to technical approvals which certify the appropriateness of the adhesive system
for load bearing timber structures and in particular for CLT for use in service class one or
two; this type of adhesive is in principal suitable for bonding of (large) finger joints as
well as edge and surface bonding;
ƒ emulsion-polymer-isocyanate adhesive (EPI) as far as the requirements are met as given
in EN 15425 [27] or within a technical approval which certifies the appropriateness of
the adhesive system for load bearing timber structures and in particular for CLT in
service class one or two; this type of adhesive is in principle allowed for bonding of
finger joints as well as for edge and surface bonding, but according to prEN 16351 [1]
not for large finger joints.

3.3 Delivery control of the base material used for load-bearing purposes (solid timber /
single-layer wood panels)
According to prEN 16351 [1], CLT can be produced of structural timber adjusted according to
EN 14081-1 [11] and / or of engineered timber products (e.g. single-layer panels) which met the
requirements of EN 13986 [32] or EN 14374 [51]. For structural timber so far only softwood
species are considered. In regard to the single layers it is allowed that ≥ 90 % of the board
material is of the declared strength class, e.g. according to EN 338 [8], whereas up to ≤ 10 % of
the boards can be of a strength class with a maximum deviation from the declared strength
values of 35 %.

3.4 Minimum FPC requirements of finger joints

The requirements concerning the production of finger joints in prEN 16351 [1] follow in
principle the ones of EN 385 [23] or prEN 15497 [24]. In the context of FPC the fulfilment of
minimum requirements of finger joint strength can be tested in tension parallel to grain or
bending. Following prEN 16351 [1], similar regulations to the glulam product standard EN 1194
[28] can be found, for example
f t,0,FJ,k ≥ 5 + f t,0,B,k ; (2)

f m,FJ,k ≥ 8 + 1.4 ⋅ f t,0,B,k . (3)

Testing comprises at least three specimens per shift and production line of the highest produced
strength class or strength profile and per adhesive. The test can be performed flatwise in four-
point bending or in tension parallel to grain, both in reference to EN 408 [42]. Deviating from
this standard the maximum (failure) load has to be reached within (60 ± 15) s. Furthermore, in
case of bending tests the test span can be reduced to lspan ≥ 15 · tL and in tension to a free test
length of lfree ≥ 3 · wL, respectively, with tL and wL as thickness and width of the laminations. It
has to be ensured that at least five of the last 100 test values are below the declared characteristic
5 %-quantile of the finger joint strength fFJ,dc,k and that within the last 15 tests none of the tests
was below fFJ,15, with fFJ,15 = k15 · fFJ,dc,k and k15 as parameter which considers the dispersion in
strength (restricted to CV[fFJ] ≥ 10 %) and the sample size assuming a lognormal distributed
strength.
Following the German technical approvals for CLT, in general testing of at least two specimens
per shift is required. FPC in regard to finger joint strength can be also done by bending and
tension tests, the last one with a minimum test length of lfree ≥ 200 mm. The requirements of the
bending strength fm,FJ,k are regulated in reference to DIN 1052 [25], annex H, Table H.1 or
DIN 68140-1 [52] (cf. Tab. 3).

Tab. 3: Minimum requirements of fm,FJ,k

strength class acc. to fm,FJ,k [N/mm²] acc. to grading class acc. to fm,FJ,k [N/mm²] acc. to
EN 338 [8] DIN 1052 [25] DIN 4074-1 [12] DIN 68140-1 [52]
C16 ≥ 25 S7 / MS7 -
C24 ≥ 30 S10 / MS10 ≥ 30
C30 ≥ 35 S13 ≥ 35
C35 ≥ 40 MS13 ≥ 40
C40 ≥ 45 MS17 ≥ 45

For example in [53], the minimum requirement of tension strength is regulated by 70 % of fm,FJ,k,
according to DIN 1052 [25]:
f t,0,FJ,k ≥ 0.7 ⋅ f m,FJ,k . (4)

Of course, in regard to the arguments in section 2.2 it is recommended to regulate the minimum
requirements of the finger joint strength based on tension tests parallel to grain according to the
formulations in section 2.2, Tab. 2 and, thus, in dependency on the stochastics of the material.

3.5 Minimum FPC requirements of large finger joints

According FPC in prEN 16351 [1], the production requirements of bond line thickness and tip
gap of the finger joint have to be controlled concerning at least one specimen per shift. The
maximum allowed bond line thickness is 0.5 mm for phenoplast- and aminoplast-adhesives and
0.3 mm for 1K-PUR. The relative tip gap has to be within e = (0.02 ÷ 0.10), with e = lt / lFJ. The
characteristic 5 %-quantile of the bending strength of large finger joints fm,LFJ,k, determined by
means of four-point bending tests by large finger joints connected full-size CLT-elements
according to EN 408 [42], has to be at least as high as the declared value fm,LFJ,dc,k.
FPC requirements concerning large finger joints in German technical approvals for CLT are
frequently referred to EN 387 [17]. Again this standard imposes requirements of the geometry,
the bond line thickness (in general ≤ 0.5 mm) and the gap tip lt = (1 ÷ 6) mm. The compliance
has to be controlled concerning at least one cylindrical specimen (diameter 25 mm) per shift or
at least regarding one per ten produced joints taken from the centre of the joint. If all test results
within a period of at least three months fulfil the requirements, the sampling may be reduced to
one per 30 produced joints, but at least one per shift. The bending strength of large finger joints
has to be determined on full-size jointed CLT elements according to EN 408 [42] and EN 386
[39]. For example in [48] the minimum required strength is defined as share of the bending
strength of the board material, for example
f m,LFJ,k ≥ 0.75 ⋅ f m,B,k . (5)
3.6 Minimum FPC requirements of edge bonding
According to prEN 16351 [1], the resistance of edge bonding has to be controlled by means of
block shear tests. Per shift at least one specimen comprising the whole width of a single-layer
panel has to be taken and at least two bond-lines need to be tested according to EN 392 [46].
Before testing the compliance of the bond line, the thickness with allowed values has to be
checked. The minimum requirements of shear strength are regulated in relation to the share of
fibre and wood on the fractured surface, see Tab. 4.

Tab. 4: Requirements of edge bonds according to prEN 16351 [1]

average value single value
fv [N/mm²] 6.0 8.0 ≥ 11.0 4.0 ÷ 6.0 6.0 ≥ 10.0
FF [%] 1) ≥ 90 % ≥ 72 % ≥ 45 % 100 % ≥ 74 % ≥ 20 %
1)
share of fractured surface covered by fibres (share of wood failure)

The shear strength fv has to be calculated as

Fu
fv = k ⋅ , with A = b ⋅ t and k = 0.78 + 0.0044 ⋅ t , (6)
A
with Fu as the ultimate failure load, A as shear area and k as thickness correction factor.

3.7 Minimum FPC requirements of surface bonding

3.7.1 Delamination according to prEN 16351 [1]

For controlling the adhesion or the resistance against fractures in the bond line, specimens of
defined geometry have to be exposed to a specific series of climatic conditions and afterwards
the delamination of their bond lines has to be determined. Therefore at least one specimen per
20 m³ produced CLT (or in case of positive results over a time period of at least three month at
least one per 40 m³) comprising the whole depth of CLT, in width ≥ (75 ± 5) mm and in length
large enough for a surface of A ≥ 10,000 mm² or a cylindrical specimen with a diameter of
≥ (95 ± 5) mm has to be taken. After determining mass and measurement of the length of all
bond lines visible on the end-grains the specimen has to be exposed to the following conditions:
ƒ completely submerged and surrounded by water of T = (10 ÷ 20) °C;
▫ exposition to a vacuum of (70 ÷ 85) kPa (absolute pressure (15 ÷ 30) kPa) for
30 min;
▫ exposition to a pressure of (500 ÷ 600) kPa (absolute pressure (600 ÷ 700) kPa)
for 120 min;
ƒ drying in a chamber at T = (65 ÷ 75) °C, (8 ÷ 10)% rel. humidity and air velocity of
(2 ÷ 3) m/s till (100 ÷ 110)% of the mass before testing is reached; this should be
possible within (10 ÷ 15) h;
Afterwards the share of delamination has to be determined on all bond lines. The maximum
share of delamination per single bond line and per specimen has to be calculated as
 lmax,delam
Delam max = 100 ⋅ [%], (7)
lglue line

and the overall share per each specimen is determined as

ltot,delam
Delam tot = 100 ⋅ [%]. (8)
ltot,glue line

The allowed shares of delamination are Delammax ≤ 40 % and Delamtot ≤ 10 %. In case that one
or both criteria are exceeded, each bond surface has to be split and the share on surface fractured
in wood or covered by fibres has to be determined. Per each bond surface a minimum share of
wood and fibre failure of 50 %, and on average of all bond surfaces per specimen a minimum
share of 70 % (maximum average delamination of 30 %) have to be kept otherwise the test has
failed.

3.7.2 Delamination according to DIN 53255 [54] / DIN 68705-4 [55] and alternative test
methods
FPC requirements of delamination in German technical approvals for CLT are in general referred
to DIN 53255 [54]. Therein a method for testing the quality and resistance of surface bonding in
cross laminated wood and timber products is provided. It examines the local dissociation of each
individual bond line by means of a special designed dissociation tool. As stated in prEN 16351
[1], a minimum average share of wood and fibre failure on all bond surfaces per specimen of
70 % is required. Before testing, each specimen has to be exposed to a cycle of specific climatic
conditions according to DIN 68705-4 [55], specification for BST 100. In doing so it is differed
between a cold water test (24 h completely submerged at T = (20 ± 2) °C) and a hot water test
(4 h completely submerged in boiling water, followed by (16 ÷ 20) h storage in a climate
chamber at T = (60 ± 2) °C, 4 h completely submerged in boiling water and (2 ÷ 3) h cooling
down completely submerged in water at T = (20 ± 5) °C).
Alternatively, some approvals allow block shear tests according to DIN 52187 [56] on at least 10
specimens per working day. The average shear strength of the last ten tests shall met
fv,mean ≥ 1.5 N/mm² and the characteristic 5 %-quantile of the last 100 tests fv,k ≥ 1.25 N/mm², but
shall not value below 1.00 N/mm².
A further alternative is to perform shear tests according to EN 789 [57], annex C on at least one
specimen per working day and thickness range of produced CLT.
Some approvals also allow delamination tests according to EN 391 [47], approach B instead of
the delamination test according to DIN 53255 [54]. The climatic conditions as well as the limits
are equal to prEN 16351 [1] (cf. section 3.7.1). Tests which exceed the limit Delamtot ≤ 10 %
have to be exposed to a second cycle of equal climatic conditions and to a new limit of
Delamtot ≤ 15 %. If this limit is also exceeded, the specimen has to be tested according to
DIN 53255 [54]. The required minimum average share of wood and fibre failure on all bond
surfaces per specimen is 70 %.

3.7.3 Excursus: experiences made with delamination

In reference to section 2.5.1 and to the research project reported in [34], some results and
experiences made in regard to delamination are presented. Fig. 10 gives an overview of the
results gained by testing CLT specimens in rolling shear according to EN 408 [42] (lspan = 12 · t)
and in delamination according to EN 391 [47], approach B. The results of delamination comprise
the maximum delamination per bond line (Delammax) and the maximum delaminated bond
surface per specimen Adelam,max. The presented results embrace nine sub-series per each test: three
variations in surface pressure (SP; (0.1, 0.3, 0.6) N/mm²) and three variations in number of
climatic cycles (CC; 0, 10, 25) to which the specimens were exposed before testing; The climate
varied weekly between 20 °C / 90 % rel. humidity and 30 °C / 40 % rel. humidity. Hence, one
climate cycle took two weeks and caused a variation in moisture content of (12 ÷ 17) %. Per
each sub-series at least five specimens were tested concerning rolling shear and then regarding
delamination.

SP = 0.1 N/mm² SP = 0.3 N/mm² SP = 0.6 N/mm²

2.5
fr,12 [N/mm²]

2.0

1.5

fr,12,k of CLT composed of boards without relieves

1.0
1.0
0.8
Delammax [--]

0.6
limit of Delammax acc. to prEN 16351
0.4
0.2
0.0
1.0
0.8
Adelam,max [--]

0.6 limit Adelam,max single bond line acc. to prEN 16351

0.4
0.2
0.0
0 10 25 0 10 25 0 10 25
CC [--] CC [--] CC [--]

Fig. 10: Results (excerpt) of CLT tested in rolling shear (fr,12; above) and delamination
(Delammax; middle and Adelam,max; below) in dependency on surface pressure (SP) and
number of climatic cycles (CC) together with FPC limits according to prEN 16351 [1]
([34])
To summarise the results briefly it can be said that all specimens tested in rolling shear delivered
strength results on the save side. The limits according to prEN 16351 [1] for the two criteria
examined in regard to delamination, Delammax and Adelam,max, were also met except of one sub-
series where the limit of Delammax was exceeded. Of course also these specimens passed the test
afterwards in examining Adelam,max. Although a slight increase of CC = 0 to CC = 10 and a
reduction in fr,12 at CC = 25 can be observed, no relation was found neither between CC and
Delammax, nor between CC and Adelam,max in delamination. In some sub-series a coherent course
of Delammax and Adelam,max is given. A relation between fr,12 and properties of delamination cannot
be confirmed. Focusing on delamination the dispersion in Adelam,max is much smaller than in
Delammax. This indicates a higher stability in the results of Adelam,max. Although it is not possible
on the basis of these tests to relate the delamination results to realistic expositions of structures
and to define limits for Delammax and Adelam,max, experiences gained during testing suggest that
the bond line quality is indicated by the combined judgement of delaminated bond lines and
identified bond surfaces.

3.7.4 Delamination: conclusions

To summarise, the approach presented in section 3.7.1 examining the delamination of surface
bonding in combination with splitting of the surfaces by a metal wedge is preferred. The main
reason for this is because it is believed to provide a higher degree in comparability and
reproducibility than the alternative methods in section 3.7.2. Of course a fundamental and
quantifiable relation between the exposition to extreme climatic test conditions and the practical
relevance, in particular with respect to the established service class system, has not been
available yet. As it is the case for glulam the limits in delamination for CLT are based on
empirical analyses developed and manifested by experiences made so far. Due to the cross
laminated structure of CLT and the higher internal stresses in the context of swelling and
shrinkage it was necessary to adapt the limits which had been until than anchored for glulam.

3.8 FPC requirements of rolling shear strength of CLT

Following the FPC regulations in German technical approvals for CLT the testing of one
specimen per working day in rolling shear by means of a four-point bending test according to
EN 408 [42], but with reduced span of lspan ≥ 15 · t is required.

3.9 FPC requirements of maximum bond-line thickness according to prEN 16351 [1]
Following prEN 16351 [1] the maximum allowed bond line thickness for aminoplast- and
phenoplast-adhesives is ta ≤ 0.6 mm and ta ≤ 0.3 mm, respectively, for common and separate
application of resin and hardener. In case of 1K-PUR the limit is ta ≤ 0.3 mm.

3.10 Additional FPC requirements

“Additional requirements” according FPC comprise regulations of (i) the used timber species,
(ii) the durability of the base material(s), (iii) criteria for preservative treated base material(s),
(iv) criteria for classifying the resistance of CLT if exposed to fire, (v) the dimensions, geometry
and assembly, and (vi) the compliance with release limits on formaldehyde and other harmful
agents. Further information can be found in the technical approvals as well as in prEN 16351 [1].

4. Main geometrical and technological parameters of CLT

Within this chapter the main geometrical parameters relevant for production of CLT are
presented. An overview of the regulation of these parameters according to current technical
approvals for CLT in Europe is provided in Tab. 5. To summarise, producers aim on reducing the
regular gap width between boards or single-layer panels within one CLT layer. Following the
regulations gaps of ≤ 2 mm and ≤ 4 mm between boards in top and core layers, respectively, are
common presently. The approvals are widely restricted to softwoods, whereby Norway spruce
(Picea abies) is definitely preferred. The strength classes of the basic material are dominated by
C16 and C24 according to EN 338 [8]. In this regard the quality of the basic material within each
CLT layer is judged under relaxed regulations. Common dimensions of CLT are up to 18 m (or
even 30 m) in length, up to 4 (4.8) m in width and seldom above (300 ÷ 400) mm in thickness.
Although hydraulic press facilities dominate the production of CLT, in volume no tendency can
be observed concerning the production parameters “single-layer panels” and “edge bonding”.
Also the requirements of FPC can be said to be diverse.
The rates of swelling and shrinkage of CLT of Norway spruce (Picea abies) in and out of plane
as well as differential swelling and shrinkage rates that apply so far to the moisture content of
CLT kept within (6 ÷ 22) % are:
ƒ for both directions in plane: 0.02 % per each percent change in moisture content;
ƒ for the direction out of plane: 0.24 % per each percent change in moisture content.

Tab. 5: Overview of some geometrical and technical characteristics of European CLT

producers
dimension BM (SP) 1)

max gap width [mm]

l [m] / w [m] / t [mm]

FPC test procedures

further information

further information
technical approval

single-layer panels

adhesive system 3)

surface pressing
w [mm] / t [mm]

timber species 2)
dimension CLT

(Y/N/possible)

(Y/N/possible)
strength class

edge bonding

for CLT 4)
bonding
≤18/≤4/60÷400 SW EN301 RS; BS;
[50] 250÷1200/15÷45 2(4) Y N pressure by
≥3 layers ≥C16 1K-PUR FJ
brackets
≤30/≤4.8/≤300 80÷220/10÷33 SW RS; FJ;
[58] 6 N - EN301 -
≥3 layers TL w/t≥4 ≥C16 DL (D)
≤30/≤4.8/30÷30080÷220/10÷33 SW EN301
[59] 6 N N - -
≥3 layers TL w/t≥4 ≥C16 EN15425
Tl 60÷150/5.75
[60] -/-/19÷42
5) Cl SW - Y - - - D; B(t)
3 layers
19÷150/7.5÷30.5
Tl
[61] -/-/16÷57 SW EN301
5) 80÷140/5.5÷13.2 - Y - - D; B(t)
3 layers ≥C16 MF
Cl 80÷140/5÷31.6
[62] ≤3(18)/≤1.25/ 60÷240/12÷40 ≥C24
6) >> - - 1K-PUR - -
60÷300 TL w/t≥2.4 ≥GL24
SW
[63] ≤3(18)/≤1.25/ 60÷240/12÷40 D; FJ;
6) ≥C24 >> EN301 -
60÷300 TL w/t≥2.4 LFJ; BS
≥GL24
SW EN301
≤18/≤3/36÷280 70÷280/12÷40
[64] C16- 2(4) N N EN15425 - -
3÷13 layers TL w/t≥4
C35 MUF
≤18/≤3/36÷280 70÷220/12÷40 SW EN301 RS or Sh;
[65] 4 - - -
3÷7 layers TL w/t≥4 ≥C16 MUF D; FJ
 ≤5/≤1.25(24)/ EN301
80÷250/18÷45 SW TL 2
[66] 60÷350 Y LL Y SP: MUF - -
TL w/t≥4 C16/C24 LL 0
≥3 layers 1K-PUR
large
large elements large elements elements
≤22/≤3.5/51÷215 100÷200/17÷43 EN301 vacuum
SW RS or Sh;
[48] system format system format 4 Y Y/N SP: MUF 80÷90 kPa
C16/C24 D; FJ; LFJ
≤5/≤1.25(24)/ 80÷250/18÷45 1K-PUR system
54÷350 TL w/t≥4 format
hydraulic
LL 80÷260/15÷45
TL 80÷260/15÷40 pneumatic
≤18/≤3.5/60÷400 LL 3 EN301
[67] TL w/t≥4 S,P,F,L Y/N N 0.5÷0.8 -
3÷11 layers TL 6 MUF
solid wood panels MPa
-/15÷45
40÷300/14÷45 EN301
≤16.5/≤3/42÷350 TL w/t≥4 S,P,F pos. EN15425
[68] 2(4) pos. - -
3÷20 layers solid wood panels ≥C16 TL Y SP: EPI
(TL) 250÷1,600/- 1K-PUR
40÷300/14÷45
EN301
≤16.5/≤3/42÷500 TL w/t≥4 SW pos. RS; DL, D
[53] 2(4) pos. SP: EPI -
3÷27 layers solid wood panels ≥C16 TL Y or BS; FJ
1K-PUR
(TL) 250÷1,600/-
≤16.5/≤2.95/
80÷240/10÷40 S EN301 hydraulic
[69] 57÷250 3(6) - - -
TL w/t≥4 ≥C16 1K-PUR ≥0.6 MPa
3÷9 layers
≤16.5/≤3/57÷500 80÷240/10÷40 S EN301
[70] 3(6) - - - RS; D; FJ
3÷27 layers TL w/t≥4 ≥C16 1K-PUR
S,P,D
≤16/≤3.2/50÷300 80÷200/18÷40 EN301 RS or Sh;
[10] Tl C24 6 - - -
≥3 layers TL w/t≥4 1K-PUR D; FJ
Cl C16
≤6/≤3.25/≤345 SW ring shank
[35] 140÷260/23 - - - EN301 FJ
≥3 layers ≥C16 nails
LL 80÷240/20÷80 S,P,F,L,
≤18/≤3/60÷300 LL 3 EN301 hydraulic
[71] TL 80÷240/20÷40 D - - -
3÷9 layers TL 6 EN15425 0.6 MPa
TL w/t≥4 ≥C16
3L: Tl
91÷190/4.5÷12
3L -/-/13÷49
[72] Cl 44÷150/4÷25 D;
5) 5L -/-/27÷42 ≥C16 - - TL Y approval -
5L: Tl B(t)
3 or 5 layers
117÷190/4.5÷8.5
Cl 44÷150/5÷9
≤20/≤4/45÷280 40÷300/15÷40 SW vacuum RS or Sh;
[49] 2(4) pos. pos. EN301
3÷7 layers TL w/t≥4 ≥C16 80÷90 kPa D; FJ
SW
≤20/≤2.5/60÷200 EN301 vacuum
[33] 80÷160/20÷40 Tl ≥C24 - - - RS; D; FJ
≥3 layers 1K-PUR 80÷90 kPa
Cl ≥C16
≤10/≤3/≤400 hardwood
SW
[36] orientation 90°, ≥100/24÷60 10 - - - dowels -
≥C16
45° or 0° d=20 mm
≤15.5/≤3.45/ S,F
60÷300/9÷30 LL D or BS;
[73] 27÷210 C16÷C3 2(4) - EN301 -
TL w/t≥4 pos. FJ
3÷7 layers 0
 ≤20/≤4/57÷280 80÷200/19÷45 S or sim. LL EN301
[74] 3 pos. - -
3 or 5 layers TL w/t≥4 ≥C16 pos. MUF
1)
BM … base material; SP … single-layer panel; TL … transverse layers; LL … longitudinal layers; Tl … top layer; Cl …
core layers; w … width; t … thickness
2)
SW … softwood species; S … Norway spruce; P … Scots pine; F … White fir; L … European larch; D … Douglas fir;
sim. … similar timber species
strength class according to EN 338 [8] (or EN 1194 [28], prEN 14080 [13])
3)
data of technical approvals complemented by manufacture’s data (product leaflet, reports, etc.);
adhesives according EN 301 [26] only of type I
4)
RS … rolling shear of CLT; BS … block shear CLT; FJ … finger joint; DL … delamination CLT; D … delamination
(dispartment at glue line) according to DIN 53255 [54]; B(t) … transverse third-point bending; Sh … (rolling) shear test
5)
3- or 5-layers wood panels for load bearing purposes
6)
dissolved cross laminated timber products for load bearing purposes

5. Discussion and conclusion

This paper provides an overview of current production techniques for cross laminated timber
(CLT). The focus is on industrial production lines, although also productions for small and
median scaled producers are concerned. The work aims on CLT as rigid composite, composed of
cross-wise arranged and surface bonded layers of boards or single-layer panels.
To summarise in brief it can be said that currently and in regard to the production volume of CLT
the hydraulic press facilities are dominating. Their further gain in market share, in particular
concerning large productions with automated lines is expected and supported by the meanwhile
offered modular press and production systems. Of course no clear trend regarding the production
parameter “edge bonding” has been observed yet. In fact, the gain by edge bonding concerning
building physics and connection technique (e.g. in case of pin-shaped fasteners) has to be
questioned since the occurrence of checks, as a consequence of swelling and shrinkage due to the
exposition of CLT to cyclic climatic conditions, cannot be avoided. Thus producers aim on
reducing the gap widths between boards or single-layer panels within the CLT layers. Latest
developments in press technology allow the application of lateral pressure individually on all CLT
layers. This enables to produce CLT elements with gap widths of zero, but without edge bonding.
Furthermore there is a trend that machine manufacturers offer their facilities together with CLT
production licences (e.g. woodtec Fankhauser GmbH / CH with [33]; Hans Hundegger
Maschinenbau GmbH / DE with [35]).
Further distribution of CLT results in the indispensability of standardised regulations not only in
Europe, but worldwide. First important steps in this regard are in process, e.g. the product
standard prEN 16351 [1] for CLT and the efforts in harmonising the lamella thickness with
tL = (20, 30, 40) mm. Further steps, in particular concerning the design procedures, the detailing
(building physics; leading details; structural engineering) and joining technique are required.
The product CLT provides not only timber engineering, but also the whole building sector with
new possibilities and offers new horizons. Currently the potential of CLT is seen in multi-storey
timber constructions for office and residential buildings and thus may initiate a Renaissance of
timber engineering in our cities. To improve its economics in particular in competition to mineral
building materials like reinforced steel, masonry and steel structures, the development of CLT
building systems and, consequently, the establishment of the solid timber construction technique
with CLT are seen as the next mile stone (see also the four-year research project “focus_sts” at
the competence centre holz.bau forschungs gmbh / AT). Therefore it is a must to take care of the
peculiarities of timber as building material, in particular its vulnerability in context with moisture.
As a consequence of the establishment of a building system, a vertical extension of CLT
production lines by assembling stations is expected. Within these stations whole wall and ceiling
elements including not only windows, doors and installations, but also the finale facade system
with insulation, as known from current production lines for light-weight timber constructions, can
be readily processed. Another possibility is to prefabricate plug-and-play facade modules. Parallel
to this, the increase of establishing engineering offices directly at, or rather in close cooperation
with, CLT producers is prognosticated.
Thereby the solid timber construction technique with CLT is not regarded as competitor against
the existing timber building sector with focus on linear timber elements. Building technique with
CLT has already shown to open and extend the possibilities to realise structures in timber. In fact,
meanwhile CLT is in direct competition with mineral based solid building materials, like
reinforced concrete and masonry. Further extension of this position is expected. This is in
particular enforced by the circumstance that the product CLT has only minor requirements
regarding the mechanical potential of the basic material. So far minimum principles as mentioned
in this contribution are applied. Thus also local timber species can be sustainably utilised and
value can be gained regionally, which may lead to the worldwide establishment of CLT.
Consequently, the creation of further small and median scaled production lines and companies as
well as of some big players worldwide all operating in the field of CLT is expected.
Acknowledgement: This work was accomplished and financed by Graz University of
Technology, Institute of Timber Engineering and Wood Technology. The support has to be
thankfully acknowledged. In addition, the supports provided by companies as producers for
press facilities as well as cutting and joining facilities have to be thankfully acknowledged as
well.

6. References
[1] prEN 16351:2011-11, “Timber structures – Cross laminated timber – Requirements”
[2] Schickhofer G., “Starrer und nachgiebiger Verbund bei geschichteten, flächenhaften
Holzstrukturen”, 1994, Dissertation, Institute of Steel, Timber and Shell Structures, Graz
University of Technology
[3] Salmén L., “Micromechanical understanding of the cell-wall structure”, 2004, C. R.
Biologies, 327:873-880
[4] Wathén R., “Studies on fiber strength and its effect on paper properties”, 2006,
Dissertation, KLC communications, No. 11, University of Technology, Helsinki, Finland,
ISSN 1457-6252, p. 97
[5] Brandner R., Stochastic System Actions and Effects in Engineered Timber Products and
Structures, Dissertation, Institute of Timber Engineering and Wood Technology, Graz
University of Technology, 2012, p. 467.
[6] Augustin M., Blaß H.J., Bogensperger T., Ebner H., Ferk H., Fontana M., Frangi A.,
Hamm P., Jöbstl R.A., Moosbrugger T., Richter A., Schickhofer G., Thiel A.B., Traetta G.
and Uibel T., BSPhandbuch: Holz-Massivbauweise in Brettsperrholz – Nachweise auf
Basis des neuen europäischen Normenkonzepts, Verlag der Technischen Universität Graz,
2010, ISBN 978-3-85125-109-8
[7] Schickhofer G., “Cross Laminated Timber (CLT) in Europe – from Conception to
 Implementation”, 2010, presentation, University of British Columbia, Department of Wood
Science, Vancouver, Canada
[8] EN 338:2009-10, “Structural timber – Strength classes”
[9] EN 384:2010-04, “Structural timber – Determination of characteristic values of
mechanical properties and density”
[10] Z-9.1-721 “ED-BSP (Brettsperrholz) aus Fichte, Kiefer oder Douglasie”, EUGEN
DECKER – Holzindustrie KG, Deutsches Institut für Bautechnik (DIBt), 2008, valid until
25.09.2013
[11] EN 14081-1:2005-11, “Timber structures – Strength graded structural timber with
rectangular cross section - Part 1: General requirements”
[12] DIN 4074-1:2012-06, “Strength grading of wood – Part 1: Coniferous sawn timber”
[13] prEN 14080:2012-02, “Timber structures – Glued laminated timber and glued solid timber
– Requirements”
[14] Daniels H. E., “The Statistical Theory of the Strength of Bundles of Threads. I”, 1945,
Proceedings of the Royal Society A, 183:405-435
[15] Sexsmith R. G., Boyle P. D., Rovner B. and Abbott R. A., “Load sharing in vertically
laminated, post-tensioned bridge decking”, 1979, Technical Report No. 6, Forintek Canada
Corp., Western Forest Products Laboratory, Vancouver, British Columbia, ISSN 0708-
6172, p. 18
[16] Bakht B. and Jaeger L. G., “Load sharing in timber bridge design”, 1991, Canadian
Journal of Civil Engineering, 18:312-319
[17] EN 387:2001-10, “Glued laminated timber – Large finger joints – Performance require-
ments and minimum production requirements”
[18] Aicher S. and Klöck W., “Spannungsberechnungen zur Optimierung von Keilzinken-
profilen für Brettschichtholz-Lamellen”, Bauen mit Holz, 1990, 92(5):356-358, 360-362
[19] Colling F. and Ehlbeck J., “Tragfähigkeit von Keilzinkenverbindungen im Holzleimbau”,
Bauen mit Holz, 1992, 94(7):586-593
[20] Radovic B. and Rohlfing H., “Über die Festigkeit von Keilzinkenverbindungen mit
unterschiedlichem Verschwächungsgrad”, Bauen mit Holz, 1993, 3:196-201
[21] Groom L.H. and Leichti R.J., “Effect of adhesive stiffness and thickness on stress
distributions in structural finger joints”, Journal of Adhesion, 1994, 44:69-83
[22] Smardzewski J., “Distribution of stresses in finger joints”, Wood Science and Technology,
1996, 30:477-489
[23] EN 385:2001-10, “Finger jointed structural timber – Performance requirements and mini-
mum production requirements”
[24] prEN 15497:2011-08, “Structural finger jointed solid timber – Performance requirements
and minimum production requirements”
[25] DIN 1052:2008-12, “Design of timber structures – General rules and rules for buildings”
[26] EN 301:2006-06, “Adhesives, phenolic and aminoplastic, for load-bearing timber struc-
 tures – Classification and performance requirements”
[27] EN 15425:2008-02, “Adhesives – One component polyurethane adhesives for load bearing
timber structures – Classification and performance requirements”
[28] EN 1194:1999-09, “Timber structures – Glued laminated timber – Strength classes and
determination of characteristic values”
[29] Brandner R. and Schickhofer G., “Glued laminated timber in bending: thoughts,
experiments, models and verification”, 11th World Conference on Timber Engineering
(WCTE), Riva del Garda, Italy, 2010, p. 11
[30] Colling F., Tragfähigkeit von Biegeträgern aus Brettschichtholz in Abhängigkeit von den
festigkeitsrelevanten Einflussgrößen. Dissertation, Universität Fridericiana Karlsruhe,
1990, p. 205
[31] Brandner R. and Schickhofer G., “Glued laminated timber in bending: new aspects
concerning modelling”, Wood Science and Technology, 2008, 42(5):401-425
[32] EN 13986:2002-06, “Wood-based panels for use in construction – Characteristics,
evaluation of conformity and marking”
[33] Z-9.1-576 “Woodtec Massivholzplatten”, Woodtec Fankhauser GmbH, Deutsches Institut
für Bautechnik (DIBt), 2010, valid until 31.05.2015
[34] Brandner R. and Schickhofer G., “SSTC 1.1.2-5 clt_panel_pressdruck: Untersuchungen
betreffend der Definition eines für die Produktion von Brettsperrholz optimalen
Pressdruckes”, 2012, Research Report, Institute of Timber Engineering and Wood
Technology, Graz University of Technology, Competence Centre holz.bau forschungs
gmbh, p. 76
[35] Z-9.1-602 “MHM-Wandelemente (Massiv-Holz-Mauer-Wandelemente”, MHM
Entwicklungs GmbH, Deutsches Institut für Bautechnik (DIBt), 2009, valid until
30.06.2010
[36] Z-9.1-574 “THOMA-Holz 100 System”, Ing. Erwin Thoma Holz GmbH, Deutsches Institut
für Bautechnik (DIBt), 2008, valid until 30.06.2013
[37] ETA-11/0338, “Elements jointed with "wooden screws" made of beech – Solid wood slab
element – element of dowel jointed timber boards to be used as a structural element in
buildings”, Rombach Bauholz+Abbund GmbH, Deutsches Institut für Bautechnik (DIBt),
2011, valid until 17.10.2016
[38] Kairi M., “Glued / Screwed Joints / Screw Glued Wooden Structures”, Chapter 4.4, In:
Johansson C. J., Pizzi T. and Leemput M. V. eds., COST Action E13 “Wood Adhesion and
Glued Products”, Working Group 2: Glued Wood Products: State of the Art Report, 2002,
2nd Edition, ISBN 92-894-4892-X
[39] EN 386:2001-10, “Glued laminated timber – Performance requirements and minimum
production requirements”
[40] Hoyle R. and Woeste F., “Wood Technology in the Design of Structures”, 1989, Fifth
Edition, Iowa State University Press, p. 380
[41] Kiviluoto J. and Muilu J., “Liimasauman mikroskopia”, 1988, VTT, Tiedotteita 561
[42] EN 408:2010-07, “Timber structures – Structural timber and glued laminated timber –
Determination of some physical and mechanical properties”
[43] Brandner R., Schickhofer G., Ruli A. and Halili Y., “Non-KInd B_S_H – Leistungspotential
von Brettschichtholz – Beanspruchung auf Längsdruck und Querdruck: Zusammen-
fassende Darstellung”, 2006, Research Report, Institute of Timber Engineering and Wood
Technology, Graz University of Technology, Competence Centre holz.bau forschungs
gmbh, p. 106
[44] Wassipaul F., “Einfluß des Preßdruckes auf die Festigkeit der Leimverbindung bei
Brettschichtträgern”, 1982, Ingenieurholzbau in Forschung und Praxis, Bruderverlag,
Karlsruhe
[45] Baumann H. and Marian J. E., “Der Verleimungs-Pressdruck als Funktion physikalischer
Faktoren”, 1961, Holz als Roh- und Werkstoff, 19(11):441-4xx
[46] EN 392:1995-01, “Glued laminated timber – Shear test of glue lines”
[47] EN 391:2001-10, “Glued laminated timber – Delamination test of glue lines”
[48] Z-9.1-534 “Binderholz Brettsperrholz BBS”, Binderholz Bausysteme GmbH, Deutsches
Institut für Bautechnik (DIBt), 2012, valid until 06.12.2014
[49] Z-9.1-793 “Brettsperrholzelemente STEPHAN – FLEXcross”, Stephan Holzbau GmbH,
Deutsches Institut für Bautechnik (DIBt), 2012, valid until 14.06.2016
[50] Z-9.1-680 “HMS – Element”, Haas FERTIGBAU GmbH & HMS Bausysteme GmbH,
Deutsches Institut für Bautechnik (DIBt), 2007, valid until 31.01.2012
[51] EN 14374:2004-11, “Timber structures – Structural laminated veneer lumber
Requirements”
[52] DIN 68140-1:1998-02, “Wood finger jointing – Part 1: Finger jointing of softwood for
load-bearing structures”
[53] Z-9.1-559 “CLT – Cross Laminated Timber”, Stora Enso Wood Products Oy Ltd,
Deutsches Institut für Bautechnik (DIBt), 2012, valid until 13.01.2017
[54] DIN 53255:1964-06, “Testing of wood adhesives and glued wood joints; shear and
delamination tests for determining the failing strength of plywood bonds (plywood and
coreboard)”
[55] DIN 68705-4:1981-12, “Plywood; core-plywood for building purposes”
[56] DIN 52187:1979-05, “Testing of wood; determination of ultimate shearing stress parallel
to grain”
[57] EN 789:2004-10, “Timber structures – Test methods – Determination of mechanical
properties of wood based panels”
[58] Z-9.1-501 “MERK Dickholz ® (MDH)”, Finnforest Merk GmbH, Deutsches Institut für
Bautechnik (DIBt), 2009, valid until 31.01.2014
[59] ETA-10/0241 “LenoTec – Solid wood slab elements to be used as structural elements in
buildings”, Finnforest Merk GmbH, Deutsches Institut für Bautechnik (DIBt), 2010, valid
until 12.08.2015
[60] Z-9.1-640 “Massivholzplatten – Layer-plus-static”, Rettenmeier Holding AG, Deutsches
Institut für Bautechnik (DIBt), 2006, valid until 30.09.2011
[61] Z-9.1-209 “Dreischichtplatten aus Nadelholz der Fa. Schwörer Haus KG”, SchwörerHaus
KG, Deutsches Institut für Bautechnik (DIBt), 2011, valid until 01.09.2016
[62] ETA-05/0211 “Lignotrend block panel elements: Timber units for walls, roofs and
ceilings”, LIGNOTREND AG, Deutsches Institut für Bautechnik (DIBt), 2010, valid until
29.11.2015
[63] Z-9.1-555 “LIGNOTREND-Elemente”, LIGNOTREND AG, Deutsches Institut für
Bautechnik (DIBt), 2008, valid until 25.06.2013
[64] ETA-09/0036 “MM – CLT: Cross Laminated Timber (CLT) – Solid wood slab elements to
be used as structural elements in buildings”, Mayr-Melnhof Kaufmann Gaishorn GmbH,
Österreichisches Institut für Bautechnik (OIB), 2009, valid until 22.03.2014
[65] Z-9.1-638 “Brettsperrholz: M1 BSP crossplan”, Mayr-Melnhof Kaufmann Gaishorn
GmbH, Deutsches Institut für Bautechnik (DIBt), 2011, valid until 01.01.2016
[66] ETA-06/0009 “MM – CLT: Binder Brettsperrholz BBS: Multilayered timber elements for
walls, ceilings, roofs and special construction components”, Binderholz Bausysteme
GmbH, Deutsches Institut für Bautechnik (DIBt), 2011, valid until 20.12.2016
[67] ETA-11/0189 “Derix Cross Laminated Timber: Solid wood slab element to be used as a
structural element in buildings”, W. u. J. Derix GmbH & Co., Deutsches Institut für
Bautechnik (DIBt), 2011, valid until 10.06.2016
[68] ETA-08/0271 “CLT – Cross Laminated Timber: Solid wood slab elements to be used as
structural element in buildings”, Stora Enso Wood Products Oy Ltd, Deutsches Institut für
Bautechnik (DIBt), 2011, valid until 27.04.2014
[69] ETA-06/0138 “KLH solid wood slabs: Solid wood slab element to be used as structural
elements in buildings”, KLH Massivholz GmbH, Österreichisches Institut für Bautechnik
(OIB), 2006, valid until 26.07.2011
[70] Z-9.1-482 “KLH-Kreuzlagenholz”, KLH Massivholz GmbH, Deutsches Institut für
Bautechnik (DIBt), 2010, valid until 17.11.2015
[71] ETA-11/0210 “Merkle BSP: Solid wood slab elements to be used as structural element in
buildings”, Merkle Holz GmbH, Deutsches Institut für Bautechnik (DIBt), 2011, valid
until 05.07.2016
[72] Z-9.1-465 “NORDPAN Drei- und Fünfschichtplatten”, NORDPAN SPA AG, Deutsches
Institut für Bautechnik (DIBt), 2010, valid until 14.03.2015
[73] Z-9.1-809 “Heißerer Brettsperrholz”, Gebrüder Heißerer, Deutsches Institut für
Bautechnik (DIBt), 2010, valid until 24.05.2015
[74] ETA-12/0281, “NORITEC X-LAM: Cross Laminated Timber (CL T) – Solid wood slab
elements to be used as structural elements in buildings”, NORITEC Holzindustrie GmbH,
Österreichisches Institut für Bautechnik (OIB), 2012, valid until 19.08.2017