materials

Article
Comparison of Calculation Methods of Elastic Bonding: Limits
of the Gamma Method Using an Example of a Wood–Concrete
Composite Floor with Single Loads
Christian Huber 1 and Karl Deix 2, *

1 Camillo Sitte Bautechnikum, HTL Wien 3, A-1030 Vienna, Austria; [email protected]

2 Institute of Material Technology, Building Physics and Building Ecology, University of Technology Vienna,
A-1040 Vienna, Austria
* Correspondence: [email protected]; Tel.: +43-1-58801-21519

Abstract: Various methods are available for the calculation of timber–concrete composite floors. The
gamma method, which is important in construction practice, as well as the differential equation
method, are based on the simplified assumption of a continuous bond between wood and concrete.
This makes it possible to analytically calculate the internally statically indeterminate partial section
sizes and deformation sizes, analogous to the force size method. In this paper, two typical load
situations of concentrated loads (central and off-centre) were analytically and numerically evaluated
and compared using the above-mentioned methods (gamma and differential equation), with a



discrete method for the case of a timber beam reinforced with a concrete slab using screws as


fasteners. The calculation results show significant deviations, which speak for the application of
Citation: Huber, C.; Deix, K.
discrete methods in certain load situations and thus limit the usability of the gamma method under
Comparison of Calculation Methods
certain conditions. For the problem of deflection determination, which is not dealt with in the
of Elastic Bonding: Limits of the
literature for the discrete method, a numerical method is described in the present work, which was
Gamma Method Using an Example of
a Wood–Concrete Composite Floor
first developed and presented by the first author.
with Single Loads. Materials 2021, 14,
7211. https://doi.org/10.3390/ Keywords: timber–concrete composite floors; static; discrete method
ma14237211

Academic Editors: Hélder David da

Silva Craveiro and Rui António 1. Introduction
Duarte Simões Wood is becoming increasingly important in multi-story residential and office con-
struction as a natural and sustainable building material. In addition to the ecological
Received: 11 October 2021
advantages, the favourable indoor climate and natural appearance of wood surfaces on
Accepted: 18 November 2021
walls and ceilings are also in the foreground. Here, the wood–concrete composite construc-
Published: 26 November 2021
tion method, which makes optimum use of the mechanical properties of both building
materials, has proved to be particularly forward-looking. A large part of the applications of
Publisher’s Note: MDPI stays neutral
this method is carried out as reinforcement systems for existing wooden ceilings, especially
with regard to jurisdictional claims in
in connection with the overall building stiffness under earthquake loads. For sustainability
published maps and institutional affil-
reasons, wood–concrete composite floors are also increasingly being used in new buildings,
iations.
especially with the use of CLT (cross-laminated timber) panels.
The lower wooden element, in the form of beams or slabs, mainly carries the bending
tensile forces, while the upper continuous concrete slab mainly carries the bending com-
pressive forces. The necessary shear connection is made with various elements, such as
Copyright: © 2021 by the authors.
bolts, screws, shear collars (cleats), etc. These form a compliant, or elastic, bond between
Licensee MDPI, Basel, Switzerland.
the two elements, which leads to a load-bearing behaviour that lies between no- (loose)
This article is an open access article
and rigid-bond behaviour.
distributed under the terms and
The following is a brief summary of the development of wood–concrete composite
conditions of the Creative Commons
floors based on the literature. The first wood–concrete composite floors were developed
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
about 100 years ago, as indicated by Holschemacher et al. in [1], Yeoh et al. in [2] and Grosse
4.0/).
et al. in [3]. Design formulas were developed or adopted from Kolbitsch et al. [4] and

Materials 2021, 14, 7211. https://doi.org/10.3390/ma14237211 https://www.mdpi.com/journal/materials

Materials 2021, 14, 7211 2 of 15

modified from the “doweled beam theory”. A comparison of the calculation methods was
made by Grosse et al. in [5]. The long-term behaviour was modified by Kuhlmann et al. [6]
using the gamma method. It was also treated by Grosse et al. [7], Avak et al. [8] and
Gerold et al. [9]. Schmidt et al. compared the gamma method with the finite element
method in [10] and gave design proposals with graded fastener spacing in [11]. In [12],
Rautenstrauch et al. showed a practical design using the framework model. The behaviour
of CLT–concrete composite floors with the extended gamma method and the finite element
method was investigated by Forsberg et al. in [13], based on the work of Wallner et al. [14].
Essential for load-bearing behaviour are the connecting elements between the wood
and concrete. In [15], fully threaded screws were investigated by Heller, and in [16], dowel
bars were discussed by Schröter et al. Bonded connections were treated by Schäfers et al.
in [17]. Numerical modelling was performed for cleats by Grosse et al. [18], and appropriate
models and failure criteria were applied by Schönborn et al., who gave design rules for
shear collars in [19]. In one of the newest papers [20], Woschitz et al. described bending
tests with CLT and prefabricated concrete plates and compared the calculation methods.
In [21], the state-of-the-art of timber–concrete composite structures from cost (European
Cooperation in Science and Technology) is described. This documentation forms the basis
for the new Eurocode.
For the simplified calculation of wood–concrete composite floors, the strictly spatial
system is reduced to an elastically coupled flexural beam system, for which several calcula-
tion methods are now available. The differences between the methods result, on the one
hand, from the different methods used to generate the static model, and on the other hand,
from system-specific conditions (in particular the types of action) that lead to different
governing equations.
In this paper, the gamma method according to Möhler [22] and Heimeshoff [23],
which is established in practice and relatively easy to calculate, and which is prescribed by
Eurocode 5 [24], is compared to and evaluated with the more stringent differential equation
method according to Natter and Hoeft [25], as well as the discrete method according to
Stüssi [26,27] and Huber [28], on the basis of a representative design situation for selected
load cases. This is especially relevant, since the gamma method is known to provide exact,
or satisfactorily accurate, results only for symmetrical sinusoidal load situations with an
approximately parabolic moment curve.
In reality, when composite screws are used, there is rather a punctual-shear coupling
between concrete and wood. The discrete method takes into account the beam sections
resulting from the splitting during modelling. It determines the sectional force quantities
on a numerical basis, following the finite element method, which is very close to reality.
With a relatively dense distribution of the fasteners, all three of the above methods pro-
vide satisfactorily accurate results under uniform load on the single-span beam. However,
in the case of single loads on the system or eccentric load location, significant deviations
from the valid gamma method can occur. In the case of point load application, which is
frequently encountered in construction practice, such as for columns and walls, no evident
data are available regarding the accuracy of the gamma method.

2. Static System
Since the corresponding values diverge with increasing deviations between the real
system and the calculation model, a relatively high single-force application with densely
distributed screw spacing was taken as a basis for the calculations, and the uniform load
occurring due to self-weight effects was not taken into account.
A typical case (Figure 1), in practice, was calculated with the reinforcement of wooden
beams by means of a concrete slab. Figures 2–4 show the cross-section dimensions, the
load situation and the characteristic values; Table 1 contains the characteristic values.
 Materials 2021, 14, x FOR PEER REVIEW 3 of 16
Materials 2021,
Materials 14, 14,
2021, x FOR PEER
x FOR REVIEW
PEER REVIEW 3 of3 16
of 16
Materials 2021, 14, 7211 3 of 15
Materials 2021, 14, x FOR PEER REVIEW 3 of 16

Figure 1. Wood–concrete composite floors—scheme.

Figure 1. Wood–concrete composite floors—scheme.
Figure 1. Wood–concrete
Figure Wood–concretecomposite
compositefloors—scheme.
Figure 1.1.Wood–concrete composite floors—scheme.
floors—scheme.

Figure 2. Cross-section dimensions.

Figure 2. Cross-section dimensions.
Figure 2. Cross-section dimensions.
Figure 2. Cross-section dimensions.
Figure 2. Cross-section dimensions.

Figure
Figure 3.
3. Load
Load situations.
situations.
Figure 3. Load situations.
Figure 3. Load situations.
Figure 3. Load situations.

Figure 4.
Figure Idealised cross-section
4. Idealised cross-section with
with static
static parameters.
parameters.
Figure 4. Idealised cross-section with static parameters.
Figure 4. Idealised cross-section with static parameters.

Figure 4. Idealised cross-section with static parameters.

Materials 2021, 14, 7211 4 of 15

Table 1. Characteristic values.

Plate Beam
Concrete C20/25 Wood C22
Elastic modulus E1 3000 kN/cm2 Elastic modulus E2 1000 kN/cm2
Thickness h1 6 cm Beam high h2 22 cm
Contributing plate width b1 91 cm Beam width b2 16 cm
Area A1 546 cm2 Area A2 352 cm2
Moment of Inertia I1 1638 cm4 Moment of Inertia I2 14,197 cm4
Centre of mass distance e 16.4 cm Thickness of the formwork s 2.4 cm

For the “SFS composite screw crossed” fastener, the following displacement moduli
were applied according to approval [29]:
Kser = 166 KN/cm for the calculation of the serviceability limit state.
Ku = 2/3 . . . Kser = 111 KN/cm for the calculation of the ultimate limit state.
The screws, arranged crosswise, were anchored with a length of 100 mm in the wood
and 45 mm in the concrete. The bolt heads were located in the concrete in the area of the
steel reinforcement. The displacement modulus was a compliance coefficient analogous
to the modulus of elasticity, where it corresponded to the proportionality factor in the
point and to the linear load-displacement law of the fastener in the contact joint direction.
In the differential equation method, the joint stiffness (K) is also important. This was
calculated as the quotient of the displacement modulus K (Kser for the deformations and
Ku for the forces) divided by the fastener spacing e’ (displacement modulus notionally
smeared over the bolt spacing). The fastener spacing in the longitudinal direction of the
beam was e’ = 11.1 cm, which resulted from 45 sections for the span of 500 cm.

3. Theory and Calculation

3.1. Calculation of Force Quantities with the Discrete Method
In the method first presented by Stüssi [26,27], the fasteners were represented on the
idealised structural model as point-acting single dowels with linear elastic deformation
laws (transverse to the joint). Likewise, linear material laws and the validity of the Bernoulli
hypothesis were assigned to the concrete belt bar and the timber beam for the individual
cross-sections (but not for the total cross-section). Thus, there was an internal highly
statically indeterminate structure, with the excess dowel forces acting horizontally in
the connection joint (Figure 5). For further consideration, the system was split up and,
according to the superposition principle, superimposed on the composite-free girder—
each loaded with the dowel forces (acting in the gravity line of the contact joint) and the
external load.
Li corresponds to the dowel-force resultant, calculated field by field from the support.
The relationship between the dowel-force resultant and the section-normal force is shown
in Figure 4. In the belt, for equilibrium reasons, Ni = −Li , while in the web Ni = Li , applies.
Using the main bending equation and the material law, length changes in the contact joint
due to the affected dowel-force resultants and partial moments (respectively due to the
external load and due to the dowel forces) can be calculated.

3.1.1. Partial Moments Due to the External Load

Due to the relatively simple conditions at the unbonded beam, the partial moments
due to the external load can be expressed by the total moment M. Following Figure 4, the
absence of partial-normal forces means that (N1 = N2 = 0) is valid for the unbonded beam
for this load case:
M = M1 + M2 (1)
Due to the equal deflections of belt and web (effects due to the twisting of the cross-
section were not taken into account), whereby for any beam location w1 = w2 , after
 Materials 2021, 14, 7211 5 of 15

differentiating twice and neglecting the shear deformation, the geometric compatibility
condition follows this equation:

M1 M2
= (2)
E1 · I1 E2 · I2

Materials 2021, 14, x FOR PEER REVIEW Thus, two equations are available for the two unknowns (M1 and M2 ), and the task can5 of 16
be solved mathematically and unambiguously if the total moment M is known, whereby the
partial moments can be expressed by the total moment. The total moment M0m, averaged
over the section, was used to determine the elongation changes.

Figure 5. Exposed composite beam with cutting forces.

Figure 5. Exposed composite beam with cutting forces.
3.1.2. Partial Moments According to Dowel Forces
LThe
i corresponds to thecurve
partial moment dowel-force resultant,
due to the calculated
dowel forces fieldin
is constant bysections.
field from the support.
The relationship between the dowel-force resultant and the section-normal force is shown
in3.1.3.
Figure Compatibility Condition
4. In the belt, for equilibrium reasons, Ni = −Li, while in the web Ni = Li, applies.
Using The fulfilment
the main of the
bending constraint
equation andcondition, according
the material to Figure
law, length 6 andinEquation
changes (3), joint
the contact
leads directly to Equation (4), which can be set up for each field.
due to the affected dowel-force resultants and partial moments (respectively due to the
externalIn this
loadcase:
and due to the dowel forces) can be calculated.
ei ’ + ∆ei ’1u + vi = ei ’ + ∆ei ’2o + vi−1 ⇒ ∆ei ’ 1u + vi = ∆ei ’ 2o + vi−1 (3)
3.1.1. Partial Moments Due to the External Load
ei ’: Due
Centre todistance of dowels
the relatively in section
simple i in theatundeformed
conditions or unloaded
the unbonded beam, thestate.
partial moments
∆e ’ : Length change of the distance e ’ of the dowels
due to the external load can be expressed by the total moment M. Following
i 1u i at the height of the lower edge4, the
Figure
fibre of the upper beam 1 due to the forces and moments acting in this section (moments
absence of partial-normal forces means that (N1 = N2 = 0) is valid for the unbonded beam
are to be averaged).
for this load case:
vi : Deformation of dowel i in the direction of the contact joint due to dowel force Fi.
∆ei ’2o : Change in the length of the distance ei ’ of the dowels at the level of the upper (1)
M= M + M
edge fibre of the lower beam 2 due to forces and moments acting in this section.
Due to the equal deflections of belt and web (effects due to the twisting of the cross-
section were not taken into account), whereby for any beam location w1 = w2, after differ-
entiating twice and neglecting the shear deformation, the geometric compatibility condi-
tion follows this equation:
M M (2)
=
E ∙ I E ∙ I
Materials 2021, 14, 7211 6 of 15

vi−1 : Deformation of the dowel i-1 in the direction of the contact joint due to the dowel
force Fi−1 .
Materials 2021, 14, x FOR PEER
" REVIEW !# 6 of 16
0 h21 h22 1 1 M0im ·K·e0 ·(h1 + h2 + 2·s)
− Li−1 + Li · 2 + K·ei · + + + − Li+1 = (4)
4·E1 ·I1 4·E2 ·I2 E1 · A1 E2 · A2 2·(E1 ·I1 + E2 ·I2 )
3.1.3. Compatibility Condition
where M0im is the mean total moment in section i. With 45 sections, this results in a
systemTheoffulfilment
equations ofwith
the constraint condition,
45 equations and 45according
unknowntodowel-force
Figure 6 andresultants
Equation(L
(3),) or
i
leads directly to Equation
partial-normal forces (Ni ). (4), which can be set up for each field.

Figure 6. Continuity condition in the contact joint.

Figure 6. Continuity condition in the contact joint.
In matrix notation, it follows:
In this case:
    
c − 1 0 0 . . 0 0
i’1u + vi = ei’ + Δei’2o + vi−1 
0 0 L m
 −1 eci’ + Δe Δei 1u + vi = Δei 2o+ vi−1
1 1
 (3)
’ ’

 −1 0 . . 0 0 0 0   L2   m2 
 0 −1 c −1 . . 0
 Lor
  m 
ei’: Centre distance of dowels in section i0in the 0 undeformed
0  3 unloaded
  state.

3 
Δei
 0
’1u: Length0 change −1 ofcthe .distance . 0 ei’ of 0 the 0dowels 0 at the
 L4 
 height

 of 
 m 
the lower
4 edge
. . beam. 1 due . to .the. forces
. and . moments.  in. this
. acting  .(moments
    
fibre of the upper  section
= 

. . . . . . . . . . . .
  
are to beaveraged).
     
0 0 0 0 . . c − 1 0 0 L m
   
vi: Deformation
 of dowel i in the direction of the contact  joint42due to dowel

 42force
  Fi.
Δei − −
  
0 0 0 0 . . 1 c
’2o: Change in the length of the distance ei’ of the dowels
 1 0 L
 at43the
 
 level
 m
 of the
 43 upper
edge fibre of0 the lower
0 0beam02 due . to 0 −
. forces and c −1 acting
1 moments  Lin 44 this  m44 
 section.
0
vi−1: Deformation 0 0 the 0dowel
of . i-1 0 −1 of cthe contact
. in0the direction L45joint due to mthe
45 dowel
force Fi−1.
In the matrix, the diagonal value c corresponds to the constant bracket expression
on hthe left side h in Equation 1 (4). The 1 factor c is independentM ∙ K ∙ e of
∙ the
h +load,
h + and2 ∙ s depends on,
−L + L ∙ 2 + K ∙ e ∙ among other + things,+the displacement + − L = (4)
4∙E ∙I 4∙E ∙I E ∙ A E ∙ A modulus. The right 2 ∙ Ecolumn
∙ I + Evector
∙I of the equation
system (m1 to m45 ) also includes the load-case-dependent total moment curve (mi ) of the
where
bondlessM0im is the
basic mean total
system. Due moment in section
to the special problem i. With 45 sections,
definition, withthis
two results in a system
load cases and one
of equations with
displacement 45 equations
modulus each for and
the45load-bearing
unknown dowel-forcecapacity and resultants (Li) or partial-nor-
the deformations, a total of
mal
fourforces
equation (Ni).systems must be solved.
In
Thematrix notation,
numerical it follows: of the four possible combination cases lead to partial-
evaluations
normal forces in the joint. Figure 7 shows this for load-bearing capacity (the shape for the
calculation𝑐 of−1 0
the deformations 0 . is. similar).0 0These0 correspond
0 𝑚
𝐿to the dowel-force resultants,
−1 sign-weighted
accordingly 𝑐 −1 0 as. already . 0 described0 0 (belt:0 −, 𝐿web: +). In 𝑚 the case of load
⎛ ⎞ ⎛ ⎞ ⎛𝑚 ⎞
situation 0 −1 𝑐 −1 . . 0 0 0 0 ⎜𝐿 ⎟
⎜ A, the numerically calculated dowel-force resultants/partial-normal ⎟ 𝐿 ⎜ 𝑚 ⎟ forces of the
⎜
individual 0 0
beam −1
sections 𝑐 were. . 0
distributed 0 0
symmetrically 0 ⎟ ⎜ over ⎟the ⎜ . length,
beam ⎟ whereas
.
⎜ situation
in load . B,. they. were . .distributed
. . asymmetrically
. . ⎟ ⎜ .due ⎟ =to ⎜the different
⎟ partial
moment⎜ . distribution.
. . . . . . . . . ⎟⎜ . ⎟ ⎜ . ⎟
⎜0 0 0 0 . . 𝑐 −1 0 0 ⎟ ⎜𝐿 ⎟ ⎜𝑚 ⎟
⎜0 0 0 0 . . −1 𝑐 −1 0 ⎟ ⎜𝐿 ⎟ ⎜𝑚 ⎟
0 0 0 0 . . 0 −1 𝑐 −1 𝐿 𝑚
⎝0 0 0 0 . . 0 0 −1 𝑐 ⎠ ⎝𝐿 ⎠ ⎝𝑚 ⎠
 mal forces in the joint. Figure 7 shows this for load-bearing capacity (the shape for the
calculation of the deformations is similar). These correspond to the dowel-force resultants,
accordingly sign-weighted as already described (belt: −, web: +). In the case of load situa-
tion A, the numerically calculated dowel-force resultants/partial-normal forces of the in-
dividual beam sections were distributed symmetrically over the beam length, whereas in
Materials 2021, 14, 7211 7 of 15
load situation B, they were distributed asymmetrically due to the different partial moment
distribution.

Web partial-normal
Figure7.7.Web
Figure partial-normal forces
forces for
for the
the load-bearing
load-bearing capacity.
capacity.

Furthermore, the partial moments for the real system were obtained directly from this
Furthermore, the partial moments for the real system were obtained directly from
in accordance with Figure 5, with compound use of the following equations (Superposi-
this in accordance with Figure 5, with compound use of the following equations (Super-
tion principle):
position principle): M · E1 · I1

h1 + s

M1 = − Ni · (5)
E1 · I1 + E2 · I2 2
M⋅E ⋅I h + s
M = M · E2 · I2 −N  ⋅ h2 + s  (5)
M2 = E ⋅ I + E ⋅ I− Ni · 2 (6)
E1 · I1 + E2 · I2 2
where s is the thickness of the formwork according to Figure 2. The results are summarised
Mcentre
for the force application point (in the ⋅ E ⋅ I of the fieldhfor
+ load
s situation A and 0.94 m
M = −N ⋅ (6)
E ⋅for
distance from the right bearing point I + E ⋅situation
load I 2 Table 2.
B) in

Table 2.s Results

where according to
is the thickness ofthe
thediscrete
formworkmethod.
according to Figure 2. The results are summa-
rised for the force application point (in the centre of the field for load situation A and 0.94
Load Situation M1 M2 N σ1o σ2u
m distance from the right bearing point for load situation B) in Table 2.
A 565 kNcm 1622 kNcm 171.5 kN −1.35 kN/cm2 1.74 kN/cm2
B 436 kNcm 1255 kNcm 83.7 kN −0.95 kN/cm2 1.21 kN/cm2

3.2. Derivation of Deformations for the Discrete Method

In ref. [28], a proposal for the determination of deflections based on the “principle
of virtual forces” was developed for the discrete method. Here again, the “superposition
principle” was used as previously described.
According to the working principle, the auxiliary system (one-system) was again the
unbonded beam with a load of 1 at the point and in the direction where the deflection
was sought.
The respective moment curves in the belt and web were then to be superimposed accord-
ing to Equation (7), whereby the deflection components due to shear forces were neglected.

M· M
Z
w= · dx (7)
E·I
The following Figure 8 illustrates the procedure using the example of a single-span
beam with a point load in the centre of the span. For the total deflection, the individual
deflection components must be added.
 M⋅M
w= ⋅ dx (7)
E⋅I

The following Figure 8 illustrates the procedure using the example of a single-span
Materials 2021, 14, 7211 beam with a point load in the centre of the span. For the total deflection, the individual
8 of 15
deflection components must be added.

Materials 2021, 14, x FOR PEER REVIEW 9 of 16

Figure 8. Procedure
Figure 8. Procedure for
for determining
determining deflections
deflections with
with principle
principle of
of virtual
virtual forces.
forces.

The deflection at the point of force application (centre of the field for load situation A)
The deflection at the point of force application (centre of the field for load situation
resulted in 1.66 cm for load situation A and 0.69 cm for eccentric-load situation B, using the
A) resulted in 1.66 cm for load situation A and 0.69 cm for eccentric-load situation B, using
method described and including the numerically determined partial-normal force results
the method described and including the numerically determined partial-normal force re-
according to Figure 8.
sults according to Figure 8.

3.3. Calculation of Force Quantities and Deformations with the Differential Equation Method
The derivation and solution of the differential equations were carried out by Natterer
and Hoeft [25]. The basis for the derivation of the required additional governing equations
for the internally statically indeterminate problem (again assuming linear material laws
 Materials 2021, 14, 7211 9 of 15

3.3. Calculation of Force Quantities and Deformations with the Differential Equation Method
The derivation and solution of the differential equations were carried out by Natterer
and Hoeft [25]. The basis for the derivation of the required additional governing equations
for the internally statically indeterminate problem (again assuming linear material laws for
the composite partners, including the connecting means and the validity of the flatness of
the partial cross-sections after bending) was again the fulfilment of the constraint condition
in the contact joint of an infinitesimal beam element (dx) with a continuous connection
between concrete and wood. This leads to a differential equation system for the contact
joint displacement (u) and the beam deflection (w). In a publication [25] of March 1987, the
differential equation system was solved for the most common loading situations, leading
to continuous governing equations for the mechanical parameters.
Materials 2021, 14, x FOR PEER REVIEW In the following, the main equations and preliminary values for the calculation 10 ofof
16a
single-span beam with concentrated load are presented, where Φ indicates the location of
the concentrated load with respect to the span (Figure 9).

Figure 9. Load situation.

Figure 9. Load situation.
For the load position (single load at the centre of the field), Φ = 0.5, and the location
For the load
of the sought position (single
mechanical loadcan
quantities at the
be centre of the
expressed byfield), Φ = 0.5,
the related and
bar the location
length variable
ofζ the sought mechanical quantities can be expressed by the related bar length variable ζ
= 0.5.
= 0.5. In order to simplify the application of the equations of determination, the following
In order values
preliminary to simplify
mustthebe application
determined of the equations
(Figure of determination,
4 and Table the following
1 are valid) according to [25]:
preliminary values must be determined (Figure 4 and Table 1 are valid) according to [25]:
(E1 · A1 + E2 · A2 ) k · e2
 
2 2
λ = k ·E ⋅ A + E ⋅ A + k ⋅Ee2 · I2 · l (8)
E 1 · A1 · E 2 · A2 E1 · I1 +
λ = k⋅ + ⋅l (8)
E ⋅A ⋅E ⋅A E ⋅I +E ⋅I
a2 E1 · A1 · E2 · A2 · e2
= (9)
a1 − a2 (E1 · A1 E
+ E⋅2A· A2⋅ )E· (E⋅ 1A· I1⋅ +
e E2 · I2 )
= (9)
1−a E ⋅ A + E ⋅ A1 ⋅ E ⋅I +E ⋅I
a2 = (E1 ·A1 +E2 ·A2 )·(E1 ·I1 +E2 ·I2 )
(10)
E1 ·A1 ·E2 ·A12 ·e2
+1
a =
E ⋅ A + E ⋅ A E⋅ 1 E· A1⋅ ·IE2+· A
E2 ·⋅eI2 (10)
B = E1 · I1 + E2 · I2 + +1 (11)
E ⋅ A ⋅ E (E⋅ 1A· A⋅1 e+ E2 · A2 )
E1 ·I1
c= E ⋅A
⋅E ⋅A ⋅e (12)
B=E ⋅I +E E⋅1I· I1++ E2 · I2 (11)
E ⋅A +E ⋅A
E2 ·I2
d= (13)
E1 · I1E+ ∙EI2 · I2
c =
The normal force and the bendingEmoments
(12)
∙ I + E ∙ Iin:
result

a2 ∙ h[λ·(1 − Φ)]
 
1 sin
N = P·l· Φ)·
· (1 − d = λ
ζ − · · sin h ( λ · ζ ) (14)
(13)
e ∙ sin
∙ h(λ)

[λ·(1 − Φ)]
 
The normal force and the bending
2

moments 1 sin hin:
2 result
M1 = P·l·d· 1 − a ·(1 − Φ)·ζ + a · · ·sin h(λ·ζ) (15)
a 1 sinh λ ∙λ 1 − Φsin h(λ)
N = P ∙ l ∙ ∙ 1 − Φ ∙ ζ − ∙ ∙ sinh λ ∙ ζ (14)
e λ sinh λ h[λ·(1 − Φ)]

2 2 1 sin
M2 = P·l·c· 1 − a ·(1 − Φ)·ζ + a · · ·sin h(λ·ζ) (16)
λ 1 sinhsinλh∙(λ1) − Φ
M =P∙l∙d∙ 1−a ∙ 1−Φ ∙ζ+a ∙ ∙ ∙ sinh λ ∙ ζ (15)
λ sinh λ

1 sinh λ ∙ 1 − Φ (16)
M =P∙l∙c∙ 1−a ∙ 1−Φ ∙ζ+a ∙ ∙ ∙ sinh λ ∙ ζ
λ sinh λ
Materials 2021, 14, 7211 10 of 15

Table 3 summarises the results of the calculations.

Table 3. Results according to the differential equation method.

Load Situation M1 M2 N σ1o σ2u

A 565 kNcm 1632 kNcm 171.0 kN −1.35kN/cm2 1.75 kN/cm2
B 442 kNcm 1277 kNcm 82.0 kN −0.96 kN/cm2 1.22 kN/cm2

To illustrate and better classify the results, the fictitious case of a rigid bond between
wood and concrete was also calculated using this method, and the results are presented in
Table 4.

Table 4. Results for the rigid bond.

Load Situation M1 M2 N σ1o σ2u

A 256 kNcm 741 kNcm 244.1 kN −0.92kN/cm2 1.27 kN/cm2
B 158 kNcm 458 kNcm 149.2 kN −0.56 kN/cm2 0.78 kN/cm2

According to [25], the deflections were given by Equation (17) for load situation A
as 1.68 cm and for load situation B as 0.71 cm. In comparison, the deflections under the
assumption of a rigid composite would be calculated as 1.07 cm for load situation A and
0.40 cm for load situation B.

P · l3 1 sin h[λ·(1 − Φ)]

 2    
a 1 1 
2 2
w= · ·( 1 − Φ )· ζ − · · sin h ( λ · ζ ) + ·( 1 − Φ )· 2 · Φ − Φ − ζ ·ζ (17)
B 1 − a2 λ2 λ3 sin h(λ) 6

3.4. Gamma Method

Usually, the gamma method is used for composite structures. This is a simple cal-
culation method, which is also the standard method in Eurocode 5 [24]. The advantage
lies in compact and easy-to-use formulas, whereby the effective bending stiffness of the
entire beam can be used for deformation calculations. With the introduction of an effective
moment of inertia, the requirement that the curvature of the individual parts must corre-
spond to the curvature of the entire beam is again met for the composite-free case in an
extended sense.
M1 M2 M
= = (18)
E1 · I1 E2 · I2 Ev · Ieff
The comparative elastic modulus Ev can be free selected; mostly, Ev = E2 is selected.

M = M1 + M2 (19)

Thus, for the effective moment of inertia after some transformation

E1 E
Ieff = · I + 2 · I2 (20)
EV 1 EV

In the case of a rigid compound, the Steiner component also appears. In the compound-
less case (no Steiner part is effective), this corresponds to a weight of the Steiner part of 0.
All cases of the real compound can therefore be classified between these two cases (with
gamma as a weighting factor = 0 to 1).
The γ-value was obtained by comparing the terms of the gamma method with those
of the differential equation method. Compact terms were obtained only for sinusoidal
loads, but large and complicated terms were obtained for uniform loads and symmetrically
concentrated loads. According to [3], the following formulas result for the sinusoidal load:

π2 · E1 · A1 · e 0
f= (21)
l2 · K
Materials 2021, 14, 7211 11 of 15

1
γ= (22)
1+f
1 γ · E1 · A1 · (h1 + h2 + 2 · s)
a2 = · (23)
2 (γ · E1 · A1 + E2 · A2 )
1
a1 =· (h1 + h2 + 2 · s) − a2 (24)
2
where f is an auxiliary value. With the arbitrary comparative elasticity modulus Ev , the
effective moment of inertia can be calculated:
E1 E E E
Ieff = · I + 2 · I2 + γ · 1 · A1 · a1 2 + 2 · A2 · a2 2 (25)
EV 1 EV EV EV

For the total moment for load situation A:

40 · 5
M= = 50.0 kNm
4
and for load situation B:
40 · 4.056 · 0.944
M= = 30.631 kNm
5
The partial moments M1 and M2 , as well as the normal force, result in

M
M1 = ·E ·I (26)
EV · Ieff 1 1

M
M2 = · E2 · I2 (27)
EV · Ieff
M
N= · E 2 · A2 · a 2 (28)
EV · Ieff
This results in the values compiled in Table 5:

Table 5. Results according to the gamma method.

Load Situation M1 M2 N σ1o σ2u

A 444 kNcm 1282 kNcm 199.7 kN −1.18kN/cm2 1.56 kN/cm2
B 272 kNcm 785 kNcm 122.3 kN −0.72 kN/cm2 0.96 kN/cm2

The deflections can be determined with the conventional formulas of structural anal-
ysis, namely with the effective bending stiffness according to the gamma method in the
denominator. Load situation A for x = l/2 follows:

F· l3
w= (29)
48·Ev ·Ieff

and for the load situation B for x = 0.94

F· a2 ·b2
w= (30)
3·Ev ·Ieff ·l

This results in 1.65 cm for load situation A and 0.62 cm for load situation B.

4. Comparison of the Results

Figures 10–12 show a comparison of the bending moments M1 and M2 and the normal
forces N, and Figure 13 shows the stresses.
 This results in 1.65 cm for load situation A and 0.62 cm for load situation B.
This results in 1.65 cm for load situation A and 0.62 cm for load situation B.
4. Comparison of the Results
4. Comparison of the Results
Materials 2021, 14, 7211 Figures 10–12 show a comparison of the bending moments M1 and M2 and the 12 normal
of 15
forcesFigures
N, and10–12
Figureshow a comparison
13 shows of the bending moments M1 and M2 and the normal
the stresses.
forces N, and Figure 13 shows the stresses.

Figure
Figure10.
10.Moment
Moment M11,, calculated withdifferent
calculated with differentmethods.
methods.
Figure 10. Moment M1, calculated with different methods.

Materials 2021, 14, x FOR PEER REVIEW 14 of 16

Figure11.
Figure Moment M22,, calculated
11.Moment calculated with
withdifferent
differentmethods.
methods.
Figure 11. Moment M2, calculated with different methods.

Figure 12.Normal
Figure12. Normalforce
forceN,N,calculated
calculatedwith
withdifferent methods.
different methods.
Materials 2021, 14, 7211 13 of 15

Figure 12. Normal force N, calculated with different methods.

(a) (b)
Figure 13. Longitudinal
Figure 13. Longitudinal stresses
stresses for load for load situation
situation A (a)load
A (a) and and situation
load situation B (b).
B (b).

It should
It should first
first bebe notedthat,
noted that, under
under the
thegiven
givenconditions,
conditions, the the
differential equation
differential equation
method and the discrete method provide almost identical values for the selected load-
method and the discrete method provide almost identical values for the selected load-
bearing behaviour variables. This circumstance can be explained by the relatively dense
bearing behaviour variables. This circumstance can be explained by the relatively dense
arrangement of the bolt pairs, which creates an almost continuous bond between the wood
arrangement of the bolt
and the concrete. Thepairs, whichconstant
sectionally creates moments
an almostand continuous bondofbetween
normal forces the wood
the discrete
and method
the concrete. The sectionally constant moments and normal forces
deviate only slightly from the continuous lines of the differential equation method of the discrete
method deviate
at small only (here,
distances slightly from the
45 fields). continuous
In contrast lines which
to screws, of the are
differential equation
always arranged method
rela-
at small
tivelydistances (here,
closely, larger 45 fields).
distances In contrast
are present in the casetoofscrews,
cleats. Inwhich are always
[20], comparative arranged
calcu-
lationsclosely,
relatively with cleats,
largerwhich formed are
distances onlypresent
10 sections (andcase
in the single
of loads),
cleats. were carried
In [20], out,
comparative
and also with
calculations showed deviations
cleats, whichbetween
formedthe methods.
only 10 sections (and single loads), were carried out,
and also As showndeviations
showed in the comparative
between calculation, the gamma method underestimates more
the methods.
accurate methods with respect to stress determination for both load situations. For load
As shown in the comparative calculation, the gamma method underestimates more
situation A, the gamma method calculates a difference of about minus 13 per cent for σ1o
accurate methods with respect to stress determination for both load situations. For load
(based on the values calculated with the more stringent method) and minus 11 per cent
situation
for σ2uA, theload
. For gamma method
situation B, thecalculates
differencesawere
difference of about
even greater minus
(minus 13cent
25 per per for
cent
thefor σ1o
(based on the values calculated with the more
upper stress versus minus 21 per cent for the lower stress).stringent method) and minus 11 per cent
for σ2u . For
This can be explained primarily by a redistribution of the internal forces in the direc-for the
load situation B, the differences were even greater (minus 25 per cent
upper stress
tion of theversus minus 21which
rigid composite, per cent
wasfor thepronounced
more lower stress).
in load situation B, according to
the
Thisdata
canavailable here. Use
be explained of the gamma
primarily method leadsof
by a redistribution tothe
a reduction
internal of around
forces 22 per
in the direction
cent for load situation A (minus 39 per cent in load situation B) in
of the rigid composite, which was more pronounced in load situation B, according to the the case of the two
datapartial moments,
available here. compared
Use of the with the more
gamma stringent
method methods,
leads with a simultaneous
to a reduction of around in-22 per
crease in the partial-normal-force stress of 18 per cent (plus 49 per cent in load situation
cent for load situation A (minus 39 per cent in load situation B) in the case of the two
B). This allows the conclusion that, in the case of the central concentrated load and with
partial moments, compared with the more stringent methods, with a simultaneous increase
increasing deviation from this, the gamma method underestimates the bending stress and
in the
Materials 2021, 14, x FOR PEER REVIEW partial-normal-force stress of 18 per cent (plus 49 per cent in load15situation of 16 B).
This allows the conclusion that, in the case of the central concentrated load and with
increasing deviation from this, the gamma method underestimates the bending stress and
overestimates
overestimates the thepartial-normal-force
partial-normal-force stress.
stress. The values,
The values, therefore,
therefore, erroneously
erroneously approachapproach
the exactresults
the exact results assuming
assuming a rigid
a rigid composite.
composite. This
This was alsowas also ultimately
ultimately reflected inreflected
the de- in the
deflections
flections (in (in particular
particular load load situation
situation B, withB,minus
with 11
minus 11 per cent).
per cent).
The
Thedeflections,
deflections,shown
shown in Figure 14, are
in Figure 14, approximately the same
are approximately thefor all 3for
same calculation
all 3 calculation
models and
models andlarger
largerthan
thanwith
withthethe
rigid system.
rigid system.

Figure Deflectionw,w,
14.Deflection
Figure 14. calculated
calculated with
with different
different methods.
methods.

5. Conclusions
In summary, the gamma method is to be considered as a model-rate inaccurate
method for the computational determination of two-part wood–concrete hybrid beams
with common design features, especially for predominantly eccentrically located load sit-
Materials 2021, 14, 7211 14 of 15

5. Conclusions
In summary, the gamma method is to be considered as a model-rate inaccurate method
for the computational determination of two-part wood–concrete hybrid beams with com-
mon design features, especially for predominantly eccentrically located load situations
compared to the more stringent method.
The gamma method underestimates the stress and deformation quantities to be deter-
mined for both the ultimate limit state design and the serviceability check, and is therefore
on the unsafe side.
From a design point of view, the differences can, in any case, become decisive with
regard to an exact verification, and thus represent a design-relevant criterion under certain
conditions. Deviations can also occur with a larger spacing of the cleats—even with
uniform loads—as described in [20]. In view of these results, it is therefore recommended
to use the more stringent differential equation method for mathematical prediction in such
exceptional situations.
In summary, it is to be stated that the discrete method, which involves a high com-
putational effort, is a valuable methodology for the accurate simulation of conventional
wood–concrete composite beams, especially in the case of single loads.

Author Contributions: Conceptualisation, C.H. and K.D.; methodology, C.H.; software, C.H. and
K.D.; validation, C.H. and K.D., formal analysis, C.H.; writing—original draft preparation, C.H.;
writing—review and editing, K.D.; visualisation, K.D.; project administration K.D. All authors have
read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: Open Access Funding by TU Wien.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Holschemacher, K.; Selle, R.; Schmidt, J.; Kieslich, H. Wood-Concrete-Composite; Betonkalender: Berlin, Germany, 2013; pp. 241–287.
(In German)
2. Yeoh, D.; Fragiacomo, M.; de Franceschi, M.; Boon, K.H. State of the Art on Timber-Concrete Composite Structures: Literature
Review. J. Struct. Eng. 2011, 137, 1085–1095. [CrossRef]
3. Grosse, M.; Rautenstrauch, K. Numerical modelling of timber and connection elements used in timber-concrete-composite
constructions. Edinburgh 2004, 17, CIB-W18/37-7-15.
4. Kolbitsch, A.; Pauser, A.; Bölcskey, E.; Zajicek, P. Reinforcement of Existing Wooden Ceilings with Consideration of Subsequently Manu-
factured Wood-Concrete Composite Structures, Research Project F 1021—Part Wooden Ceiling Structures; Österreichische Gesellschaft
zur Erhaltung von Bauten: Wien, Austria, 1992. (In German)
5. Grosse, M.; Hartnack, R.; Lehmann, S.; Rautenstrauch, K. Modeling of discontinuously connected wood-concrete composite
slabs/Part 1: Short-term behavior. Bautechnik 2003, 80, 534–541. (In German) [CrossRef]
6. Kuhlmann, U.; Schänzlin, J.; Michelfelder, B. Calculation of wood-concrete composite floors. Beton-Und Stahlbetonbau 2004, 99,
262–270. (In German) [CrossRef]
7. Grosse, M.; Hartnack, R.; Rautenstrauch, K. Modeling of discontinuously connected wood-concrete composite slabs/Part 2:
Long-term behavior. Bautechnik 2003, 80, 693–701. (In German) [CrossRef]
8. Avak, R.; Glaser, R. Simplified method for predicting the long-term behavior of wood-concrete composite structures. Bautechnik
2005, 82, 200–211. (In German) [CrossRef]
9. Gerold, M.; Schänzlin, J.; Kuhlmann, U. Wood as an ideal composite partner for concrete as a building material. Bautechnik 2003,
80, 840–845. (In German) [CrossRef]
10. Schmidt, J.; Schneider, W.; Thiel, R. To design wood/concrete composite beams. Bautechnik 2003, 80, 302–309. (In German)
[CrossRef]
11. Schmidt, J.; Kaliske, M.; Schneider, W.; Thiele, R. Design proposal for wood/concrete composite beams considering graded
fastener spacings. Bautechnik 2004, 81, 172–179. [CrossRef]
Materials 2021, 14, 7211 15 of 15

12. Rautenstrauch, K.; Grosse, M.; Lehmann, S.; Hartnack, R. Practical dimensioning of wood-concrete composite floors, 6. Informationstag
des IKI; Bauhaus Universität Weimar: Weimar, Germany, 2003. (In German)
13. Forsberg, A.; Farbäck, F. Timber Concrete Composite Floors with Cross Laminated Timber—Structural Behaviour & Design, Division of
Structural Engineering, Faculty of Engineering; Report TVBK-20/5279; Lund University: Lund, Sweden, 2020.
14. Wallner-Novak, M.; Koppelhuber, J.; Pock, K. Cross-Laminated Timber Structural Design|Basic Design and Engineering Principles
According to Eurocode; Proholz: Vienna, Austria, 2014. (In German)
15. Heller, H. Wood-concrete composite floors (HBV). Bautechnik 2008, 85, 667–677. (In German) [CrossRef]
16. Schröter, H.; Simon, A.; Timmler, H.-G.; Rautenstrauch, K.; Raue, E. For the calculation of wood-concrete composite beams with
methods of mathematical optimization. Bautechnik 2010, 87, 474–481. (In German)
17. Schäfers, M.; Seim, W. Bonded composite components made of wood and high-strength and ultra-high-performance. Bautechnik
2011, 88. (In German)
18. Grosse, M.; Rautenstrauch, K.; Schlegel, R. Numerical modeling of wood and fasteners in wood-concrete composite structures.
Bautechnik 2005, 82, 355–364. (In German) [CrossRef]
19. Schönborn, F.; Flach, M.; Feix, J. Design rules and design information for shear collars in wood-concrete composite structures.
Beton-Und Stahlbetonbau 2011, 106, 385–393. (In German) [CrossRef]
20. Woschitz, R.; Deix, K.; Huber, C.; Kampitsch, T. Development of innovative wood-concrete composite slabs in prefabricated
construction method. Bautechnik 2021, 98, 12–22. (In German) [CrossRef]
21. Dias, A.; Schänzlin, J.; Dietsch, P. Design of Timber-Concrete Composite Structures; Shaker Verlag GmbH: Aachen, Germany, 2018.
[CrossRef]
22. Möhler, K. On the Load-Carrying Behavior of Flexural Beams and Compression Members with Composite Cross-Sections and Yielding
Connectors; TH Karlsruhe: Karlsruhe, Germany, 1956. (In German)
23. Heimeshoff, B. To calculate flexural beams made of yieldingly connected cross-section parts in engineered timber construction.
Holz Als Roh-Und Werkst. 1987, 45, 237–241. (In German) [CrossRef]
24. ÖNORM EN 1995-1-1, Eurocode 5. Designs of Timber Strutures, Part 1-1: General—Common Rules and Rules for Buildings; Austrian
Standards: Vienna, Austria, 2019.
25. Natterer, J.; Hoeft, M. The Load-Bearing Behavior of Wood-Concrete Composite Structures; Forschungsbericht CERS Nr. 1345;
EPFL/IBOIS: Lausanne, Switzerland, 1987. (In German)
26. Stüssi, F. About the doweled beam. Schweiz. Bauz. 1943, 122, 271–274. (In German)
27. Stüssi, F. Composite Solid Wall Beams; Abhandlung der Int. Ver. für Brückenbau und Hochbau: Zürich, Switzerland, 1947; pp.
249–268.
28. Huber, C. Document on the Design of Wood-Concrete Composite Floors with Selected Fasteners under Fire Exposure; Institut für Hochbau
und Technologie TU Wien: Vienna, Austria, 2010. (In German)
29. Zulassungsnummer Z-9.1-342. SFS-Compound Screws, VB-48-7,5x100 as a Fastener for the SFS Wood-Concrete Composite System;
Allgemeine Bauaufsichtliche Zulassung: Berlin, Germany, 2003. (In German)