Structures 71 (2025) 108017

Contents lists available at ScienceDirect

Structures
journal homepage: www.elsevier.com/locate/structures

Design and construction of a concrete shell by active bending of a flat plate

Johannes Berger a,* , Rupert Maleczek b, Mathias Bank Stigsen b
a
Department of Structural Engineering and Material Science, University of Innsbruck, Austria
b
Departement of Design | i.sd | Structure and Design, University of Innsbruck, Austria

A R T I C L E I N F O A B S T R A C T

Keywords: Complex formed concrete shells are widely used in building environments as thin, lightweight, and cost-effective
Shell solutions. To simplify the formwork required for such structures, this study explores the potential to form
Construction complex shells from flat, flexible concrete plates through a controlled uplifting process. Flexible concrete is
Active bending
characterised by a thin concrete layer with highly ductile reinforcement, allowing cracking under stress. This
Textile concrete
NL FEA
behaviour reduces stiffness and facilitates the formation of curved structures. To demonstrate the feasibility of
Experiments the method, a demonstrator was developed using paper models, CAD and non-linear finite element calculations.
This demonstrator, built from a plate measuring 7.0 × 7.0 × 0.04 m, reached a height of 2.0 m after deformation.
Findings indicate that this approach enables the construction of complex curved structures from flexible con­
crete. This research presents new perspectives for the application of flexible concrete in architecture and con­
struction, emphasising the potential for innovative, cost-effective, and resource-efficient building techniques.

1. Introduction printing and additive manufacturing [21–23], precast concrete [24,25]

or modular formwork.
Shells are impressive and elegant structures because they transfer This research addresses existing challenges by exploring an innova­
loads predominantly or only through membrane forces [1–3] when tive approach: the use of active bending to create thin concrete shells
appropriate designed. This load-bearing effect, known since the use of from flat plates. The primary goal of the study is to develop a more
arches [4,5], enables the support of large spans using very little material efficient and cost-effective technique for constructing curved concrete
[6–9]. This offers advantages in terms of weight, material consumption, structures by integrating two established methods: the fabrication of flat
waste reduction, and foundation. Although these structures have proven concrete sheets and their transformation into curved geometries through
to enable a wide variety in architectural design, as shown by the work of active bending. While producing thin concrete sheets with tailored
Candela, Nervi or Niemeyer, concrete shell roofs currently play a sub­ properties in a flat state is already standard practice, this research in­
ordinate role simply because of the costly formwork. Therefore, struc­ troduces the novel combination of active bending with concrete fabri­
tures are predominantly constructed as bending structures [10]. In light cation, aiming to provide a streamlined and economical solution for
of the current societal challenges in the construction sector, such as constructing curved shells. The main objectives of this work are to
resource scarcity and carbon dioxide emissions [11,12], shell structures identify and tackle the technical challenges associated with this
present themselves as attractive structural solutions. However, the approach, demonstrate its feasibility, and lay the groundwork for further
construction of shell structures typically requires a substantial amount advancements in concrete shell construction.
of labour. When concrete, which is highly suitable owing to its
compressive strength and castability, is used as a building material, the 2. Shell construction through active bending of flexible concrete
on-site construction of a shell necessitates the use of falsework and
formwork skin. These high expenses, and the limiting impact on the In shell construction by active bending [26–29], the unrolled surface
design space, significantly diminish the appeal of shell structures. of the shell is initially constructed in a planar form using flexible con­
Therefore, alternative methods have been developed for shell con­ crete and subsequently deformed into the desired shape, see Fig. 1. The
struction [13–15], such as pneumatic formwork [16], pneumatic lifting form-finding process plays a crucial role in this approach [30–34].
and shaping [17,18], KnitCrete [19], flexible formwork [20], 3D Flexible concrete is reinforced fine-grained concrete with the following

* Correspondence to: University of Innsbruck, Unit of Concrete Structures and Bridge Design, Technikerstraße 13, 6020 Innsbruck, Austria.
E-mail address: [email protected] (J. Berger).

https://doi.org/10.1016/j.istruc.2024.108017
Received 8 July 2024; Received in revised form 19 November 2024; Accepted 9 December 2024
Available online 13 December 2024
2352-0124/© 2024 The Author(s). Published by Elsevier Ltd on behalf of Institution of Structural Engineers. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
J. Berger et al. Structures 71 (2025) 108017

the plate, thus creating bending stresses on the structure. The bending
moment induces a curvature of the cross-section of the structures,
leading to a deformation of the plate towards the shell.
The large plastic deformations arising during the bending process are
accompanied by significant strains in the materials. These strains must
not lead to yielding or crushing, which is why safety measures, such as
strain limits for reinforcement, must be considered in the design.
Because it must be assumed that the building material is heterogeneous,
the deformation cannot be expected to occur exactly as planned;
therefore, deformation-correction measures must be implemented. This
can be achieved, for example, through local support and lifting [29] or
angle adjustments at the hinged supports, see Fig. 1.
For practical use, the actively bent shell must subsequently be
strengthened. This can be achieved, for example, through topping with
textile-reinforced concrete, ultra-high-performance concrete (UHPC) or
adding tension members. The actively bent shell can be used as an in­
tegral part of the shell where the stresses resulting from the deformation
process can be considered as pre-strains in the design.
For this type of shell construction, various forms and structural
configurations achievable through the bending process are explored.
During this conceptual phase, different shell geometries are considered
with a focus on aesthetics, functionality, and feasibility. The material
properties of flexible concrete and reinforcement strategies are exam­
ined to assess how they can support the intended shapes and loads. The
design space can be explored using scaled models made from heavy
paper, allowing for tactile feedback during physical deformation. These
initial studies are then digitally refined in a feedback loop, incorporating
simplified simulations with mesh relaxation methods, as directly
drawing the desired 3D shape is challenging.

3. Design and structural analysis of the shell

3.1. Project design

In the project design, the selected structure from the conceptual

phase is further developed into a constructible final geometry, informed
by the construction process. This includes readjusting the geometry,
determining precise material specifications such as the type of rein­
forcement (e.g., glass or carbon fibre), the concrete mixture (e.g., low
modulus of elasticity), and the exact process for lifting and deforming
the shell into its final shape. Structural calculations are also conducted
to ensure that the design meets safety and performance requirements,
considering factors like plastic deformations, strain limits, minimum
bending radii, and pre-strains generated during bending.
For the initial designs, computer-assisted parametric studies were
conducted using the CAD program Rhino in combination with Grass­
hopper and the plugins Kangaroo and Karamba3D, proving to be
effective tools in the design phases [37], see Fig. 2.
Following the initial models, which were characterised by numerous
rounded features, see Fig. 2, the geometry progressed to a quasi-
triangular shape in plan, with relaxation cuts arranged radially from
the actuation points, as shown in Fig. 3. The use of a paper model proved
to be particularly suitable for finding a satisfactory form. These cuts
Fig. 1. Principle of shell construction through active bending.
divided the plate into three connected segments, changing the boundary
conditions from a single continuous plate to a ring formed shell. As
properties:
observed with paper models, these cuts were necessary because the
• Reinforcement with high-elongation material behaviour εu ~ 10 –
closed surface of the flat plate was not compatible with the deformation
20 ‰ (e.g., glass or carbon fibre-reinforced plastics, steel cables, etc.).
requirements of the shell surface.
• Fine-grained concrete with a low modulus of elasticity (e.g., Ecm
Due to construction site and budget constraints, the plate dimensions
=16 GPa).
are limited to approximately 7 × 7 m, with only three supporting points,
The deformation process is performed such that the hardened con­
which also serve as points for actuation. For practical implementation,
crete plate is first lifted at key points (typically high points), and then the
an adaptation was made to the geometry. In addition to the relaxation
roller supports are shifted horizontally. Vertical lifting can be achieved
cuts, a triangular recess was created in the centre to serve as a workspace
using a lifting beam and horizontal shifting by tensioning devices [35,
for a person. This recess was necessary purely for technical reasons, as it
36], such as appropriately anchored coarse-threaded rods or tendons.
provided access for lifting the plate and installing steel components. The
The horizontal displacement generates a force that acts eccentrically on
dimensions of the final form are shown in Fig. 4.

2
J. Berger et al. Structures 71 (2025) 108017

Fig. 2. Images of the design process.

3.2. Structural analysis

The structural analysis (engineering design phase) focuses on the

detailed technical execution of the shell construction. Finite element
simulations, combined with structural analysis, are performed to predict
the behaviour of the concrete plate under bending stresses, accounting
for the nonlinear material behaviour of flexible concrete and the
occurrence of significant deformation. The use of Finite Element Models
(FEM) in the design of concrete shells ensures that deformation does not
lead to failure, thereby preventing yielding or crushing of the materials
during the bending process. FEM calculations proved essential for
Fig. 3. Paper models for form-finding. developing a thorough understanding of this innovative method. They
enabled time- and cost-efficient parameter studies, helped define and
optimise the demonstrator design, and ultimately allowed for

Fig. 4. Dimensions of the plate.

3
J. Berger et al. Structures 71 (2025) 108017

verification and evaluation of experimental results using the FEM static height d, as shown in Eq. (2).
model. ( √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ )
The results of the structural analysis were used to: 2 • (ρS1 + ρS2 • d2 /d)
xII = αs • (ρS1 + ρS2 ) • d • − 1 + 1 + (2)
αs • (ρS1 + ρS2 )2
• optimise the plate/shell geometry to ensure the bending process,
• adjust the plate thickness for appropriate stiffness distribution, with
• refine the reinforcement layout design based on plate geometry,
Es
• develop the lifting process, αs = (3)
Ec
• determine the required lifting and shifting forces,
• evaluate the stress levels and material utilisation, and As1
analyse the structural behaviour post-experiment. ρS1 = (4)
• b•d

This work results in a detailed prediction of the targeted shape, As2

ρS2 = (5)
specifications for reinforcement techniques and their precise locations, b•d
as well as a comprehensive erection strategy that facilitates the con­ To experimentally determine flexural stiffness, tests on small plate
struction of the shell structure. Models using 2D shell elements and 3D strips were conducted to identify materials with optimal deformability
solid elements with discrete embedded reinforcement were created for and ease of handling for construction purposes [37]. A four-point
the calculations, which were conducted using the software Abaqus [38]. bending test, as shown in Fig. 6, was performed on a plate strip with
test specimen dimensions of l/w/h = 1.2 m / 0.12 m / 0.04 m. In the
3.2.1. Linear finite element analysis (LIFEA) with 2D elements test, the concrete strength (fcm = 30 MPa and 90 MPa) and the type of
In the design process for constructing and erecting a demonstrator to reinforcement (AR glass fibre grid, carbon fibre grid, steel rope, and
create a concrete shell by actively bending a flat concrete plate, finite micro steel fibre) were varied. The results of the investigated parameters
element calculations were initially performed using 2D elements with are expressed as the ratio of deformation from the test to the deforma­
linear-elastic material behaviour. This approach minimised modelling tion obtained from linear elastic calculation (uTest/uLinear) and ranged
and computational efforts, allowing for rapid exploration of numerous from 15 to 42. This ratio serves as a simplified indicator of the reduction
parameters. They were conducted to gain preliminary insights into the in bending stiffness due to crack formation and was set conservatively to
deformation behaviour of the intended geometry and the necessary 1/10 of Ec. for the linear elastic calculations.
construction sequences. The concrete plate dimensions were as shown in In the finite element model, the boundary conditions were defined so
Fig. 4, except for the omission of the rounded ends of the relaxation cuts that the support points A-C allow horizontal translation, while the ver­
to simplify the model. The plate was modelled using 2D shell elements tical degrees of freedom are restrained. The loading was applied in two
(S4—4-node doubly curved general-purpose shell with finite membrane static steps. In the first step, lifting was performed at points hx,i&o
strains). A mesh size convergence analysis determined the optimal mesh through a vertical displacement U3. In the second step, the support
size, resulting in a global mesh size of 0.1 m. points A–C were displaced horizontally in the U1 and U2 directions,
For the plate, a constant thickness of 4 cm was selected and dis­ setting a perpendicular displacement vector along the edge to 70 cm.
cretised into seven integration points across the thickness. The concrete As a result, the concrete plate was successfully bent into the intended
was assigned a linear-elastic material behaviour. In order to account for shell shape. A doubly curved shell with a height of 2 m was achieved, as
the reduction in flexural stiffness due to cracking of the concrete, a shown in Fig. 7. The shell exhibits areas of positive and negative cur­
modified modulus of elasticity of Ec,mod = 3 GPa was used for the vature, with corresponding transitions in curvature. This is also reflected
modulus of elasticity of the concrete, with a Poisson’s ratio of ν = 0.2 in the distribution of positive and negative moments, as illustrated in
and a density of ρ = 2200 kg/m³ . For flexible concrete, the flexural Fig. 8. The horizontal force required for the deformation reached a
stiffness can be reduced between 1/10 to 1/40 due to cracking. These maximum value of 30 kN. This force is directly dependent on the stiff­
values were derived from analytical calculations and experimental tests ness of the structure, which in this case is influenced by the selected
on plate strips [37]. modulus of elasticity and dead weight of the plate/shell.
In the analytical calculation, the flexural stiffness of the reinforced
concrete is defined as the product of the modulus of elasticity E and the
effective moment of inertia Ieff, which is significantly reduced due to
cracking [35,39]. The moment of inertia for the doubly reinforced
rectangular cross-section under bending, see Fig. 5, is calculated ac­
cording to Eq. (1).

b • x3II
Ieff = + As1 • αs • (d − xII )2 + As2 • αs • (xII − d2 )2 (1)
12
The height of the compression zone xII depends only on the degree of
reinforcement, the elastic moduli of the materials, and the effective Fig. 6. Structural system of the small plate stripe test.

Fig. 5. Rectangular section with compression reinforcement in state 2 under pure bending.

4
J. Berger et al. Structures 71 (2025) 108017

Fig. 7. View of the deformation of a plate into a shell (2D LIFEA) - Vertical Deformation U3 in [m].

Fig. 8. Section Moments (SM) in [Nm] (Plot contours on undeformed shape).

3.2.2. Nonlinear finite element analysis (NLFEA) on a 3D model Nonlinear material behaviour was assigned to the concrete within
Three-dimensional nonlinear finite element analyses were conducted the control area. Elastic material behaviour was applied in the area of
to gain a deeper understanding of the deformation behaviour during the the supports as the thickness of the plate was widened, and a three-
active bending of a plate into a shell. To analyse the bending process, the dimensional reinforcement cage made of carbon fibre reinforcement
deformation patterns from the FEA calculations are compared with the was arranged, resulting in significantly increased stiffness.
experimental deformation pattern, obtained via a 3D laser scan, as
described in Section 4.2.2. The results from the nonlinear finite element 3.2.2.1. Reinforcement arrangement and properties. The plate reinforce­
analysis also provide a foundation for directly designing the reinforce­ ment consisted of a textile, bidirectionally reinforced grid made from
ment and concrete [40,41], examining the cracking behaviour [42], and media-resistant carbon fibre reinforced polymer (CFRP) and alkali-
analysing the lifting and shifting process. Additionally, these results resistant glass fibre reinforced polymer (GFRP), provided by the com­
informed the design of the supports and tension rods. pany "solidian•kelteks" [43]. The CFRP reinforcement used was the
The concrete plate, with dimensions shown in Fig. 4, was modelled Solid GRID Q71-CCE-51 [44], featuring a grid spacing of 51 mm in both
with solid elements (C3D8—8-node linear brick) using seven elements the longitudinal and transverse directions. The GFRP reinforcement was
over its height. A mesh size convergence analysis determined an optimal the Solid GRID Q121-RRE-38 [45], with a grid spacing of 38 mm in both
mesh size of 0.1 m in-plane and 6.6 mm in depth. The reinforcement directions.
strands were discretely embedded in the model as truss elements (2- The bottom layer of the reinforcement was covered across the entire
node linear 3D trusses). To streamline modelling, the actual reinforce­ surface with an AR-Glass Fibre RP Grid. In the upper layer, the basic
ment strands, spaced between 38 mm and 51 mm, were consolidated reinforcement in segment 1 consisted of an AR-Carbon Fibre RP Grid,
into trusses with a spacing of 0.10 m. and that in segments 2 and 3 included an AR-Glass Fibre RP Grid.

5
J. Berger et al. Structures 71 (2025) 108017

Additionally, galvanised steel cables were added to the apex areas of all Table 1
the segments, as shown in Fig. 9. The longitudinal reinforcement strands Reinforcement properties.
of the grids were aligned parallel to the edges of the respective segments; As E fm Arrangement
this resulted in the orientation of the strands of the grid in ξi and ηi di­ [mm2/m] [GPa] [MPa]
rection. Overlaps of the reinforcement grids were necessary in the areas AR- Glass Fibre 121 70 1.300 Segments 1 − 3: Bottom
of joints and where individual segments intersected, with an overlap reinforced plastic Bidirectional Segments 2 and 3: Top
length of 25 cm. The edge distance to the centre of gravity of the rein­ grid (ξi & ηi)
forcement was 10 mm, resulting in a concrete cover of 8 mm. The ar­ AR- Carbon Fibre 71 230 3.300 Segment 1: Top
reinforced plastic Bidirectional Loading Area: Top and
rangements and mechanical properties of the reinforcements are listed grid (ξi & ηi) Bottom
in Table 1. Galvanised Steel 30 80 1.900 Segments 1 − 3: Apex
Cable Longitudinal area, top (L=1 m at S1;
3.2.2.2. Concrete material behaviour [48]. The Concrete Damage Plas­ D= 3 mm 6 × 7 (ξi) L=1.4 m at S2,3)
FE
ticity Model (CDP) [46–49] is considered in ABAQUS for the non-linear
material behaviour of concrete. The CDP model is a continuum,
plasticity-based damage model for concrete that incorporates the
smeared crack model by using damage mechanics to represent the ef­ Table 2
fects of cracking in a continuum-based manner [50–52]. It assumes that Material properties of the concrete.
the two main failure mechanisms are the tensile cracking and Density Poisson´s Ratio E0 σcu σtu Gf
compressive crushing of the concrete material. The evolution of the 2200 kg/m³ 0.2 16 GPa 30 MPa 3 MPa 135 N/m
εplt and
yield (or failure) surface is controlled by two hardening variables, ̃
εplc , linked to failure mechanisms under tension and compression
̃
loading, respectively. These variables are referred to tensile and Table 3
compressive equivalent plastic strains, respectively [53]. Input parameters CDP model for ABAQUS.
Textile-reinforced concrete is a relatively new research area in structural Dilatation Angle Eccentricity fb0/fc0 κ Viscosity Parameter
concrete construction, and recent insights into nonlinear finite element [◦ ] [ -] [ -] [ -] [ -]
analysis are presented in references [53–59]. For the analyses carried 38 0.1 1.16 0.667 1E− 5
out, the associated damage variables and strains due to the tensile and
compressive loading of concrete are determined for concrete with ma­
terial strength properties as listed in Table 2. The required input pa­
σc k • η − η2
rameters for the analysis with the Concrete Damage-Plasticity (CDP) = (6)
fcm 1 + (k − 2) • η
model are chosen according to Simulia [38] and are listed in Table 3.
3.2.2.2.1. Concrete under compression stress - compressive crushing. with
According to EC [60,61], a stress-strain relationship described by Eq. 6 is
recommended for determining internal forces with non-linear methods εc
η= (7)
in the compression range. εc1

1
k = 1.05 • Ecm • |εc1 | • (8)
fcm
In the analysis with ABAQUS, this relationship is divided into a linear
and non-linear part. The linear range is described by the elastic modulus
Ecm and is limited by 40 % of the mean compressive strength fcm, as
shown in Fig. 10a. The damage variable for concrete under compressive
stress can take values from zero, representing the undamaged material,
to one, which represents the total loss of strength. Abaqus automatically
converts the inelastic strain values into plastic strain values using the
relationship:
dc σc
εplc = ̃εinc −
̃ (9)
(1 − dc ) E0
The damage variable for concrete under compressive stress, dc, as a
function of plastic strains, is given by Eq. (10). The concrete compressive
yield stress and compression damage behaviour used for the analysis are
shown in Fig. 10b.
σc (εc,pl )
dc = 1 − (10)
fcm
3.2.2.2.2. Concrete under tensile stress – cracking [48]. The
post-failure behaviour for direct straining is modelled with tension
stiffening, which allows the definition of the strain-softening behaviour
for cracked concrete [38]. This modelling approach also enables a
simple simulation of the interaction effects between reinforcement and
concrete. In reinforced concrete, specifying post-failure behaviour
generally involves providing the post failure stress as a function of the
εck
cracking strain, ̃t . Cracking strain is defined as the total strain minus
εck
the elastic strain corresponding to the undamaged material; that is, ̃t =
Fig. 9. Reinforcement arrangement of the top layer for the 3D NLFEA model.

6
J. Berger et al. Structures 71 (2025) 108017

Fig. 10. Response of concrete to uniaxial loading in compression.

εt − εel0t , as illustrated in Fig. 11a. Tension stiffening data are given in individual segments with maximum values of LE = 2.5 E-2. The stresses
terms of the cracking strain, ̃εck in the reinforcement were greatest in segment 1 at the top. This can be
t . When unloading data are available,
these are provided to ABAQUS in terms of the tensile damage curves, attributed to the use of carbon fibre reinforcement in segment 1, which
and ABAQUS automatically converts the cracking strain values to plastic has a significantly higher modulus of elasticity than the glass fibre
strain values using the relationship in Eq. (11). reinforcement used in segments 2 and 3. However, these reinforcement
stresses remained below the tensile strength limit.
dt σt Minimum principal logarithmic strain of the concrete on the un­
εplt = ̃εck
̃ t − (11)
(1 − dt ) E0 derside were particularly pronounced along the outer sides of the central
area of the individual segments with minimum values of LE = − 5 E-3.,
The damage variable for tensile stress dt is given by Eq. (12). For
see Fig. 14. The stresses in the reinforcement of the lower layer were
variable loading, the damage variable is modified using a factor
significantly lower than those in the upper layer, as the lower rein­
βd according to [62], as cracks partially close during unloading and
forcement layer is subjected to tension only in the load introduction
compressive stresses do not immediately develop. For the subsequent
area, where the segments intersect. In the remaining area, the lower
calculations, βd is assumed to be 0.5 (0 ≤ βd ≤ 1.0) based on the expe­
reinforcement is primarily under compression and was arranged mainly
rience of tests. The concrete tensile yield stress and tensile damage
for structural support. In the analysis, a maximum horizontal deforma­
behaviour used for the analysis are shown in Fig. 11b.
tion force of 25.3 kN was obtained.
σt The deformation patterns obtained revealed areas with both positive
dt = 1 − (12)
σt + E0 ∗ εck ∗ (1 − βd ) and negative curvatures. Positive curvatures occur in the apex region,
while negative curvatures occur in the load introduction area. The
3.2.2.3. Analysis steps and load application. The support points A-C were deformation patterns obtained from the finite element analysis calcu­
horizontally movable supports, which is why only the vertical lations were subsequently compared with those from the experiment.
displacement degree of freedom was constrained. In the analysis, the The load introduction area, where the plate thickness increases and the
load was applied in subsequent stages. In the first stage, the dead weight moment load is minimal (indicating little eccentricity), no cracking
was activated, and the plate was lifted 30 –40 cm at lifting points hx,i&o occurred, and only minimal curvature was observed. The area of the
(see Fig. 4). This was followed by horizontal deformation at the support load application can be considered as an inclined end tangent. The type
points A-C with a displacement of 70 cm. The deformed figure is shown of failure observed can be described as reaching the tensile strength of
in Fig. 12 and illustrate a significantly larger curvature in the apex area the reinforcement.
compared to the 2D LIFEA deformation result, see Fig. 7. Fig. 13 shows
the maximum principal logarithmic strain of the concrete at the top
surface and stresses in the reinforcement of the top layer. The largest
tensile strain on the upper side was observed in the outer vertex area of

Fig. 11. Response of concrete to uniaxial loading in tension.

7
J. Berger et al. Structures 71 (2025) 108017

Fig. 12. View of the deformation of a plate into a shell (3D NLFEA) - Vertical Deformation U3 in [m].

Fig. 13. Strains and stresses on the top surface resp. Layer.

4. Large-scale experiment to build a shell by active bending layer was positioned on spacers at a height of 28 mm. Specially manu­
factured concrete blocks, attached to the formwork and reinforced with
4.1. Construction adhesive, were used as spacers. It was necessary to glue the reinforce­
ment to the spacers to prevent it from floating due to the lower density of
A large-scale experiment, with the dimensions shown in Fig. 4, was the reinforcement compared to concrete. In the apex area of the upper
performed at the TVFA laboratory at the University of Innsbruck [63]. A reinforcement layer, galvanised steel cables with a diameter of 3 mm
formwork system was used in which formwork sheets were laid on a were added at distances of 10 cm to achieve a more uniform distribution
support grid of formwork beams, positioned approximately 20 cm above of curvature, thereby generating more height.
the ground. Care was taken in arranging the formwork beams to allow Pre-fabricated three-dimensional reinforcement cages were placed in
for the lifting of the concrete plate by raising individual beams on which the support areas. In the regions around the supports and segment in­
no form panels were placed. An increase in plate thickness was incor­ tersections, overlapping reinforcement led to multiple layers of rein­
porated in the load introduction area, achieved by creating recesses in forcement grids. In some areas, there were up to six layers at a single
the formwork. The hinged supports, consisting of steel components, reinforcement level; therefore, unnecessary reinforcement strands were
were attached to the concrete plate using bolts. removed from overlapping zones. The concrete used was a pumpable
Commercially available textile reinforcement grids were cut into polymer cement concrete (PCC), with a mix that included sand with a
pieces of various sizes and installed. The arrangement of the reinforce­ maximum grain size of 3.5 mm, a sand-to-cement ratio of 3.6:1, a water-
ment, as shown in Fig. 15a, was carried out as described in Section binder ratio of 0.6, and a polymer content of 10 % by mass of the
3.2.2.1 (see also Table 1). The bottom layer was placed on spacers at a cement. The component quantities are shown in Table 4. The strength of
height of 8 mm to ensure the required concrete cover [64], while the top concrete was specified in the NLFEA calculations (see Section 3.2.2.2).

8
J. Berger et al. Structures 71 (2025) 108017

Fig. 14. Strains and stresses on the bottom surface resp. Layer.

Fig. 15. Construction states.

wood panel, rigid foam, and a formwork beam. Due to this coupled
Table 4
spring system, the plate can be lifted evenly using multiple supports,
Material mix quantities per 1 m³ of concrete.
preventing any transverse bending.
Material Per 1 m³ Furthermore, the differential vertical displacement of the load
Sand Stone powder 151 kg application points enables the plate to accommodate variable defor­
0,09 − 0,63 mm 242 kg mation towards the shell shape. This multiple-point support reduces
0,63 − 0,80 mm 91 kg
stress on the plate due to punching and bending from its weight, creating
0,8 − 1,2 mm 182 kg
1,2 − 3,5 mm 848 kg a support concept similar to the tree-like support of a flat plate [65].
Cement (CEM II) 42,5 N 420 kg After completing the lifting process, the installed coarse-threaded
Water 252 kg rods (tension members) with a diameter of 20 mm were tensioned at
Polymer powder 42 kg the roller supports, see Fig. 17. The tension members were stabilised by a
star-shaped guide in the centre of the plate via a steel structure.
The hardening time of the concrete was 30 days, after which the Tensioning was carried out using a hydraulic hollow cylinder, with the
formwork was stripped, see Fig. 15b, and the concrete plate was lifted. procedures performed step-by-step to ensure uniform force application.
Subsequently, specially designed lifting beams, as shown in Fig. 16, For safety, the shell was supported at specific points underneath, and its
were used to achieve a uniform lift of the plate to a height of 0.75 m. shape was adjusted by local lifting.
This was necessary because lifting solely with the formwork beam The tensioning force, measured using a ring load cell, was 20 kN at a
would cause deformation of the formwork beam, resulting in the plate sagittal distance of 120 cm, which corresponded well with the NLFEA
being supported only at the side edges. This point support at the edges calculation. The frictional force of the roller support must also be
would cause the plate to deformation in the transverse direction under considered when applying the tensioning force.
its own weight, increasing stiffness and making the lifting and bending
process impossible. The lifting beam consists of a deformation-sensitive
system of load application points, a cantilever arm made of a multilayer

9
J. Berger et al. Structures 71 (2025) 108017

• Deformation measurement to determine the height

• Segment measurement to determine the curvature
• 3D Laser Scan

The following results were achieved:

• The maximum displacement of the roller support was 33 cm. Sup­

ports A and B required larger displacements than Support C.
• The maximum heights achieved in the experiment were h2;3,
o= 210 cm for segments 2 and 3, and h1;o= 110 cm for segment 1 at
the outer edge.
• The maximum inclination at the load introduction point at Support C
was 32◦ before failure.
• The smallest radius was determined to be slightly off-centre from the
apex at segment 2 and was R= 2 m.
• The deformation pattern was consistently somewhat asymmetric.

4.2.2. Geometric evaluation and accuracy analysis

Fig. 19a–d show a comparison of the deformation patterns from the
laser scans, 2D LIFEA, and 3D NLFEA for the sections defined in Fig. 20.
The FEA calculation results generally indicate a promising trend of the
deformation pattern; however, deviations of approximately ± 10 %
from the actual shape (laser scan) are observed in the geometry, as
Fig. 16. Lifting beam. shown in Fig. 19a, b.
The radius of curvature of the deformation profile, as shown in
Fig. 19a, b, d, was not constant along the length, with the greatest
curvature occurring at the apex, as expected. In the region within the
first 1.0–1.5 m from the support, a linear curve approximation is
applicable. In the analysed cross-sections, the smallest positive radius of
curvature was R = 4.8 m, and the smallest negative radius of curvature
was R = 9.0 m.
Sections X2-X2 and X3-X3, see Fig. 19c, illustrate the symmetry of the
deformation profile. The deformations of these two sections are not
identical, with height variations of 4 % observed between them. In the
Yi-Yi sections, which run perpendicular to the segments, no significant
curvature was detected. However, all sections exhibited a lateral incli­
nation with an outwards slope, as summarised in Table 5. The most
significant inclination was observed in section Y1-Y1, with a slope of
7.9◦ . Differences of 1.2◦ in transverse inclination were noted between
the calculation models and the actual shell. It is important to note that
this deformation analysis was not conducted at the point of maximum
deformation.
The results of the nonlinear finite element analysis (NL-FEA) align
closely with the experimental results. Due to cracking in the concrete,
locally larger curvatures were observed. Both the experiments and NL-
FEA calculations indicate that asymmetric deformation patterns, devi­
ating by several centimetres from the ideal shape, are likely. For prac­
tical applications, this suggests that geometric deviations in the form of
imperfections must be factored into calculations. Such asymmetries can
and should be corrected during construction through shape adjustments
Fig. 17. Support and actuation detail.
to achieve the desired geometry.

4.2. Interpretation of the results 4.2.3. Crack pattern

The failure of the shell is clearly identifiable, primarily due to the
4.2.1. Actively bent concrete shell pronounced local curvature in segments 2 and 3, which appeared
These investigations and experiments have demonstrated that this approximately 1 m from the arch apex toward support C. Additionally,
method can be used for more than just the construction of a simple arch bending shear cracks were observed on the front side, and a slight height
[35,36]; it can also enable the construction of a shell. This experiment is offset of the plate was noted on both the left and right sides of the critical
an essential step toward further development this method for imple­ crack. The crack pattern immediately preceding failure is shown in
mentation in the construction industry. The actively bent shell is shown Fig. 21. Failure occurred due to cross-sectional collapse along a crack
in Fig. 18a-c. To emphasise the slender nature of the shell, the edges are running nearly parallel to the edge of Support C, as illustrated in Figs. 21
coloured red. The shell achieved the desired deformation shape through and 22.
active bending. The parallel orientation of the fracture crack can be attributed to the
The following measurements were conducted to document the fact that, ultimately, tensioning was applied only on one side of support
shape: C, due to safety considerations. Spalling of the concrete compression
zone was observed, which was evident both acoustically and visually, as

10
J. Berger et al. Structures 71 (2025) 108017

Fig. 18. Actively bent concrete shell.

Fig. 19. Deformation analysis.

11
J. Berger et al. Structures 71 (2025) 108017

that the crack ran diagonally with an inclination of 25◦ in the primary
reinforcement direction, resulting in reduced bending resistance.

5. Discussion and outlook

The authors demonstrated that a novel, resource- and time-efficient

method for constructing shell structures by actively bending flat plates
made of flexible concrete is a reasonable and effective method. The
successful investigation was achieved through consistent and incre­
mental development.
The aim of fabricating concrete shells without traditional formwork
is rarely limited to active bending strategies. The necessary flexible
concrete solutions for our simplified fabrication approach are cutting-
edge, high-performance materials, aligning with the research’s re­
quirements and its sustainability goals. Unlike conventional methods,
this approach emphasises adaptability and reduced labour intensity,
marking a significant contribution to the field. This method expands the
conceptual and practical applications of reinforced concrete, which has
traditionally prioritised stiffness and minimal deformation. By
leveraging high-performance materials, the research introduces a novel
construction paradigm with applications in structural design and
architecture.

Fig. 20. Sectioning for deformation analysis.

Table 5
Surface inclination in the sections.
Laserscan NLFEA LIFEA

Section Y1-Y1 7.9◦ 7.8◦ 6.8◦

Section Y2-Y2 2.9◦ 3.1◦ 2.8◦
Section Y3-Y3 1.8◦ 3.0◦ 2.8◦

concrete debris fell along the entire length of the crack. Surprisingly, the
deformation continued to increase significantly. Due to the altered ki­
nematic situation, the bending or curvature of the arches became less
pronounced, and the quasi-straight shell halves were pushed upward
until the cross-section sheared, causing the shell to collapse to the
ground. This crack can be described as a hinge of the folding mechanism
of straight plates [66,67]. Closer analysis of the fracture pattern revealed Fig. 22. Mode of failure.

Fig. 21. Crack pattern.

12
J. Berger et al. Structures 71 (2025) 108017

Based on a list of potential candidates identified through a combi­ performed. The objective was to demonstrate that calculations ac­
nation of literature review and experimentation, a material choice for counting for large deformations and the nonlinear behaviour of flexible
the demonstrator was made based on material properties, availability, concrete are feasible. Initial insights were gained through simplified 2D
and budget. The successful iterative form-finding of the demonstrator shell element calculations with elastic material properties and modified
and its reinforcing, prefabricating, and forming process for this new stiffness, minimising computational effort. For a more accurate repre­
active bending application was successfully executed. While nonlinear sentation of the nonlinear behaviour of flexible concrete, further cal­
3D-NLFEA provided detailed insights, simpler 2D-LIFEA proved suffi­ culations were conducted using 3D volume elements with discretely
cient for predicting deformation shapes, striking a balance between embedded reinforcement. This approach provided realistic insights into
complexity and practical applicability. The practical construction of the deformation behaviour, required forces during erection, and stresses on
shell confirmed the hypotheses and underscored the approach’s poten­ both the reinforcement and concrete.
tial. Despite its success, the project also highlighted areas for improve­ In the experimental phase, a triangular plate of flexible concrete,
ment, emphasising the need for further refinement. approximately 7.0 × 7.0 × 0.04 m in size, was constructed. Through
active bending, the plate was transformed into a doubly curved shell
5.1. Outlook with a height of 2.0 m. Valuable insights into the active bending process
were documented. The resulting shell was measured using 3D laser
Although the study shows that the presented strategy is successful in scanning, and the deformations along defined sections were compared
resource efficiency, reduced costs, and the ease of creating unique forms, to FEA results. The best alignment of deformation patterns was achieved
the reduced geometric and formal complexity will be investigated with the nonlinear FEA calculations. Overall, the method proved suc­
further, with a focus on enlarging the spanning dimensions. The satis­ cessful, and further research into suitable forms, improvements in
fying but unexpectedly complex uplifting process necessitates a more deformation behaviour during bending, and optimised reinforcement
detailed investigation into the erection method and its required equip­ measures for future practical applications is planned.
ment in future research. In detail, the following topics will form part of
future investigations: CRediT authorship contribution statement

• Material Optimisation: Future efforts should focus on developing Mathias Bank Stigsen: Writing – review & editing, Visualization,
concrete mixes tailored for active bending, emphasising a low elastic Conceptualization. Johannes Berger: Writing – original draft, Valida­
modulus and high ductility. tion, Supervision, Resources, Project administration, Methodology,
• Structural Design: Refining plate geometry, reinforcement layout, Investigation, Funding acquisition, Formal analysis, Data curation,
and methods for load introduction has significant potential to Conceptualization. Rupert Maleczek: Writing – review & editing,
improve structural performance, efficiency, and ease of assembly. Conceptualization.
• Design Integration: Collaboration with architects and designers is
essential for integrating structural, aesthetic, and functional con­ Declaration of Competing Interest
siderations, such as insulation, waterproofing, and maintenance.
• Crack Pattern Management: Managing cracks resulting from active The authors declare that they have no competing financial interests
bending could involve pre-imprinting the anticipated crack pattern. or personal relationships that may have influenced the work reported in
This can be achieved by scoring the fresh concrete on its top surface this study.
or embedding strips or similar elements underneath.
• Strengthening: Practical applications of the actively bent shell
Acknowledgements
require further reinforcement. Potential methods include applying
textile-reinforced concrete, ultra-high-performance concrete
We express our gratitude to Röfix, Doka, and Felbermayr for
(UHPC), or the integration of tensile members.
providing materials and technical support. Recognition and apprecia­
• Feasibility for Larger Spans: To achieve practical relevance, the
tion are due to the students Philipp Heck and Anna Kleiser for their
method should accommodate spans of at least 10 m. Larger spans
assistance in construction and Prof. Thomas Weinold for surveying the
may improve the economic viability of the approach and address
shells.
challenges posed by low rise-to-span ratios.
• Fabrication: Addressing plate thickness and buoyancy effects due to
Data availability
fibre-reinforced plastic reinforcements is crucial. Prefabricating
segments in controlled environments and assembling them on-site
Data will be made available on request.
could enhance precision and scalability.

References
By addressing these factors, the method has the potential to revo­
lutionise shell construction, fostering advancements in sustainable [1] Antonio T, Pascual M. Shape and size optimisation of concrete shells. ISSN 0141-
design, structural efficiency, and architectural innovation. 0296 Eng Struct 2010;32(6):1650–8. https://doi.org/10.1016/j.
engstruct.2010.02.013.
[2] Ramm E. Shape finding of concrete shell roofs. J Int Assoc Shell Spat Struct 2004;
6. Conclusion 45:29–39.
[3] Bletzinger K-U, Ramm E. Structural optimization and form finding of light weight
Investigations were conducted on constructing shells from flexible structures. Comput Struct 2001;79(22_25):2053–62. https://doi.org/10.1016/
S0045-7949(01)00052-9.
concrete - a reinforced fine grained concrete with high ductility rein­ [4] Tang G. An overview of historical and contemporary concrete shells, their
forcement. Building on previous experience with active bending of construction and factors in their general disappearance. Int J Space Struct 2015;30:
singly curved shells [35,36], this approach was extended to enable the 1–12. https://doi.org/10.1260/0266-3511.30.1.1.
[5] Veenendaal D, West M, Block P. History and overview of fabric formwork: using
construction of doubly curved shells. As part of a prototype to demon­
fabrics for concrete casting. Struct Concr 2011;12:164–77. https://doi.org/
strate feasibility, initial form-finding processes were carried out using 10.1002/suco.201100014.
CAD and physical paper models, offering valuable insights into potential [6] Muttoni A, Lurati F, Fernández RM. Concrete shells - towards efficient structures:
shapes and construction techniques. construction of an ellipsoidal concrete shell in Switzerland. Struct Concr 2013;14:
43–50. https://doi.org/10.1002/suco.201200058.
To investigate the mechanical load-bearing behaviour, particularly [7] Fenu L, Congiu E, Marano G, Briseghella B. Shell-supported footbridges. Curved
concerning the erection process, finite element analyses (FEA) were Layer Struct 2020;7:199–214. https://doi.org/10.1515/cls-2020-0017.

13
J. Berger et al. Structures 71 (2025) 108017

[8] Fenu L, Madama G. A method of shaping R/C shells with heuristic algorithms and [35] Berger J. Effects of flexural stiffness on constraints of imposed deformations in
with reference to Musmeci’s work. Stud. E Ric. Politec. Di Milano Sc. Di Spec. reinforced concrete structures. Eng Struct 2022;272:114973. https://doi.org/
Costr. Cem. Armato 2004:199–238. 10.1016/j.engstruct.2022.114973.
[9] Fenu L, Madama G, Tattoni S. On the conceptual design of R/C footbridges with the [36] Berger J, Gericke O, Feix J, Sobek W. Experimental investigations on actively bent
deck supported by shells of minimal surface. Stud. E Ric. Politec. Di Milano Sc. Di concrete shells. Struct Concr 2020;21:2268–81. https://doi.org/10.1002/
Spec. Costr. Cem. Armato 2006;26:103–26. suco.202000045.
[10] Weißbach M. Steel and concrete sandwich arches - an application in bridge [37] Heck P. Construction of a Spatially Curved Structure from a Flat Concrete Slab
construction? Beton und Stahlbetonbau 2005;100:223–6. https://doi.org/ Using External Tendons, University of Innsbruck, Master’s Thesis, 2022.
10.1002/best.200590281. [38] Simulia. Abaqus Analysis User’s Guide, III. Providence, RI: Dassault Systemes;
[11] Hawkins W, Orr J, Ibell T, Shepherd P. A design methodology to reduce the 2018.
embodied carbon of concrete buildings using thin-shell floors. Eng Struct 2020; [39] ACI CODE-318-19(22). Building Code Requirements for Structural Concrete and
207:110195. https://doi.org/10.1016/j.engstruct.2020.110195. Commentary. American Concrete Institute; 2022.
[12] Concrete needs to lose its colossal carbon footprint. Nature, 597, 2021, Pages [40] Chaudhari S, Chakrabarti M. Modeling of concrete for nonlinear analysis using
593–594, https://doi.org/10.1038/d41586–021-02612–5. finite element code ABAQUS. Int J Comput Appl 2012;44(7). https://doi.org/
[13] Veenendaal D, Block P. Design process for prototype concrete shells using a hybrid 10.5120/6274-8437.
cable-net and fabric formwork. Eng Struct 2014;75:39–50. https://doi.org/ [41] Caycedo-Garcia MS, Siqueira GH, Vieira Junior LCM, Vizotto I. Evaluation of
10.1016/j.engstruct.2014.05.036. structural capacity of triangular and hexagonal reinforced concrete free-form
[14] Sharei E, Scholzen A, Hegger J, Chudoba R. Structural behavior of a lightweight, shells. Eng Struct 2019;188:519–37. https://doi.org/10.1016/j.
textile-reinforced concrete barrel vault shell. Compos Struct 2017;171:505–14. engstruct.2019.03.044.
https://doi.org/10.1016/j.compstruct.2017.03.069. [42] Mathern A, Yang J. A practical finite element modeling strategy to capture
[15] Hawkins W-J, Herrmann M, Ibell T-J, Kromoser B, Michaelski A, Orr J-J, et al. cracking and crushing behavior of reinforced concrete structures. Materials 2021;
Flexible formwork technologies – a state of the art review. Struct Concr 2016;17: 14:506. https://doi.org/10.3390/ma14030506.
911–35. https://doi.org/10.1002/suco.201600117. [43] Solidian GmbH, Sigmaringer Straße 150, 72458 Albstadt; https://www.solidian-
[16] Pugnale A, Bologna A. Architecture Beyond the Cupola – Inventions and designs of kelteks.com/.
Dante Bini. Mathematics and the Built Environment, 7. Birkhäuser -Springer [44] Technical Product Data Sheet Solidian GRID Q71-CCE-51; https://www.solidian-
Nature Switzerland AG; 2022. https://doi.org/10.1007/978-3-031-26735-2. kelteks.com/images/TDS/GRID%20carbon/v2407%20EN/solidian%20GRID%
[17] Kromoser B, Kollegger J. Efficient construction of concrete shells by pneumatic 20Q71-CCE-51%20Technical%20Product%20Dat%20Sheet%20v2407.pdf.
forming of hardened concrete: construction of a concrete shell bridge in Austria by [45] Technical Product Data Sheet Solidian GRID Q121-RRE-38; https://www.solidian-
inflation. Struct Concr 2020;21:4–14. https://doi.org/10.1002/suco.201900169. kelteks.com/images/TDS/GRID%20glass/solidian%20GRID%20Q121-RRE-38%
[18] Kromoser B, Kollegger J, Kari H, Gradenegger K, Ganster M. An innovative 20Technical%20Product%20Dat%20Sheet%20v2407.pdf.
concrete shell construction method in practice: fabrication of wildlife bridge AM2 [46] Lubliner J, Oliver J, Oller S, Oñate E. A plastic-damage model for concrete. Int J
with PFHC. Beton und Stahlbetonbau 2018;113:222–32. https://doi.org/10.1002/ Solids Struct 1989;25(3):299–326. https://doi.org/10.1016/0020-7683(89)90050-
best.201700069. 4.
[19] Popescu M, Rippmann M, Liew A, Reiter L, Flatt RJ, Van Mele T, et al. Structural [47] Hillerborg A, Modeer M, Petersson PE. Analysis of crack formation and crack
design, digital fabrication and construction of the cable-net and knitted formwork growth in concrete by means of fracture mechanics and finite elements. Cem Concr
of the KnitCandela concrete shell. Structures 2021;31:1287–99. https://doi.org/ Res 1976;6(6):773–82. https://doi.org/10.1016/0008-8846(76)90007-7.
10.1016/j.istruc.2020.02.013. [48] Lee J, Fenves G. Plastic-damage model for cyclic loading of concrete structures.
[20] Soto M, Pauletti R, Menghetti L. On the use of flexible formworks for concrete J Eng Mech 1998;124(8):892–900. https://doi.org/10.1061/(ASCE)0733-9399
structures. Rev IBRACON Estrut Mater 2024;17(5):e17506. https://doi.org/ (1998)124:8(892).
10.1590/S1983-41952024000500006. [49] Aymerich F. , Fenu L. , Loi G. FE Analysis of the Flexural Behavior of Cementitious
[21] Lin C, Mattei G, Cheibas I, Du C, Aejmelaeus-Lindström P, Gramazio F. 3D printing Composites Using the Concrete Damage Plasticity Model. In: di Prisco, M. ,
on inflatables. Archit, Struct Constr 2023;3:217–34. https://doi.org/10.1007/ Menegotto, M. (eds) Proceedings of Italian Concrete Days 2018. Lecture Notes in
s44150-023-00092-x. Civil Engineering, Vol 42. Springer, Cham. https://doi.org/10.1007/978–3-
[22] Hack N, Dörfler K, Walzer A, Wangler T, Mata-Falcón J, Kumar N, et al. Structural 030–23748-6_10.
stay-in-place formwork for robotic in situ fabrication of non-standard concrete [50] Hofstetter G, Mang H. Computational mechanics of reinforced concrete structures.
structures: a real scale architectural demonstrator. Autom Constr 2020;115: Vieweg Verlag; 1995.
103197. https://doi.org/10.1016/j.autcon.2020.103197. [51] Tian J, Wu X, Tan X, et al. Experimental study and analysis model of flexural
[23] Szabo A, Reiter L, Lloret-Fritschi E, Wangler T, Gramazio F, Kohler M, et al. ACDC: synergistic effect of reinforced concrete beams strengthened with ECC. Constr
the admixture controlled digital casting and its application to the thin folded Build Mater 2022;352:128987. https://doi.org/10.1016/j.
concrete structures. In: Bos F, Lucas S, Wolfs R, Salet T, editors. Second RILEM conbuildmat.2022.128987.
International Conference on Concrete and Digital Fabrication. DC 2020. RILEM [52] Tan X, Mahjoubi S, Zhang Q, et al. A framework for improving bridge resilience
Bookseries, 28. Cham: Springer; 2020. p. 956–66. https://doi.org/10.1007/978-3- and sustainability through optimizing high-performance fiber-reinforced
030-49916-7_93. cementitious composites. J Infrastruct Preserv Resil 2022;3:18. https://doi.org/
[24] Iori T. , Poretti S.Pier Luigi Nervi : His construction system for Shell and Spatial 10.1186/s43065-022-00067-0.
Structures, In: Obrebski J. , Tarczewski R. (eds), Proceedings of the International [53] Berger J, Haller D. Redistribution of internal forces from the span to the support.
Association for Shell and Spatial Structures (IASS) Symposium, 2013. Struct Concr 2022. https://doi.org/10.1002/suco.202200110.
[25] Nervi P. Precast concrete offers new possibilites for design of shell structures. ACI [54] Chudoba R, Sharei E, Scholzen A. A strain-hardening microplane damage model for
Struct J 1953;42(2):537–48. https://doi.org/10.14359/11835. thin-walled textile-reinforced concrete shells, calibration procedure, and
[26] Lienhard J, Gengnagel Ch, Knippers J, Alpermann H. Active bending, a review on experimental validation. Compos Struct 2016;152:913–28. https://doi.org/
structures where bending is used as a self-formation process. Int J Space Struct 10.1016/j.compstruct.2016.06.030.
2013;28:187–96. https://doi.org/10.1260/0266-3511.28.3-4.187. [55] Colombo I, Colombo M, di Prisco M, Pouyaei F. Analytical and numerical
[27] Kotelnikova-Weiler N, Douthe C, Hernández E, Baverel O, Gengnagel Ch, Caron J- prediction of the bending behaviour of textile reinforced concrete sandwich beams.
F. Materials for actively-bent structures. Int J Space Struct 2013;38:229–40. J Build Eng 2018;17:183–95. https://doi.org/10.1016/j.jobe.2018.02.012.
https://doi.org/10.1260/0266-3511.28.3-4.229. [56] Matheas J. Delamination Bruch in der Grenzfläche von Beton und Textilbeton mit
[28] Phocas MC, Alexandrou K. Numerical analysis and cable activation in hybrid Kontakt und Reibung. Mater Test 2007;49(10):554–9. https://doi.org/10.3139/
bending-active structures with multiple cables. Eng Struct 2018;174:561–72. 120.100821.
https://doi.org/10.1016/j.engstruct.2018.07.089. [57] Vakaliuk I, Scheerer S, Curbach M. Numerical analysis of textile reinforced
[29] Berger J, Gericke O, Feix J, Sobek W. Actively bent concrete shells. Struct Concr concrete shells: force interaction and failure types. CivilEng 2024;5:224–46.
2020;21:2282–92. https://doi.org/10.1002/suco.201900505. https://doi.org/10.3390/civileng5010012.
[30] Bessini J, Lázaro C, Monleón S. A form-finding method based on the geometrically [58] Vakaliuk I, Scheerer S, Curbach M. Numerical analysis of textile reinforced
exact rod model for bending-active structures. Eng Struct 2017;152:549–58. concrete shells: basic principles. Appl Sci 2024;14:2140. https://doi.org/10.3390/
https://doi.org/10.1016/j.engstruct.2017.09.045. app14052140.
[31] Miki M, Adiels E, Baker W, Mitchell T, Sehlström A, Williams CJK. Form-finding of [59] Djamai Z, Bahrar M, Salvatore F, Si Larbi A, El Mankibi. Textile reinforced concrete
shells containing both tension and compression using the Airy stress function. Int J multiscale mechanical modelling: application to TRC sandwich panels. Finite Elem
Space Struct 2022;37(4):261–82. https://doi.org/10.20944/ Anal Des 2017;135:22–35. https://doi.org/10.1016/j.finel.2017.07.003.
preprints202012.0355.v2. [60] ÖNORM EN 1992-1-1. Eurocode 2. Design of concrete structures, Part 1–1: General
[32] Li Q, Su Y, Wu Y, Borgart A, Rots JG. Form-finding of shell structures generated rules and rules for buildings.. Vienna: Austrian Standards; 2015.
from physical models. Int J Space Struct 2017;32(1):11–33. https://doi.org/ [61] ÖNORM B 1992–2. Eurocode 2. Design of concrete structures – Part 2: Concrete
10.1177/0266351117696577. bridges – Design and detailing rules – National specifications concerning ÖNORM
[33] Chiang Y-CH, Borgart A. A form-finding method for membrane shells with radial EN 1992–2, national comments and national supplements. Vienna: Austrian
basis functions. Eng Struct 2022;251:113514. https://doi.org/10.1016/j. Standards; 2019.
engstruct.2021.113514. [62] G. Hofstetter, FEM - Nichtlineare Festigkeitsanalysen, Universität Innsbruck,
[34] Bessini J, Lázaro C, Casanova J, Monleón S. Efficiency-based design of bending- Fakultät für Technische Wissenschaften LFU-UIBK, 2015.
active tied arches. Eng Struct 2019;200:109681. https://doi.org/10.1016/j. [63] Kleiser A. Studies on Shell Construction through Active Bending, University of
engstruct.2019.109681. Innsbruck, Master’s Thesis, 2023.

14
J. Berger et al. Structures 71 (2025) 108017

[64] Preinstorfer P, Kollegger J. New insights into the splitting failure of textile- [66] Maleczek R, Genevaux Ch. Design constraints for linear folded stripes. Int J Space
reinforced concrete. Compos Struct 2020;243:112203. https://doi.org/10.1016/j. Struct 2015;30(2):179–90. https://doi.org/10.1260/0266-3511.30.2.179.
compstruct.2020.112203. [67] Sharifmoghaddam K, Maleczek R, Nawratil G. Generalizing rigid-foldable tubular
[65] Rian I, Sassone M. Tree-inspired dendriforms and fractal-like branching structures structures of T-hedral type. Mech Res Commun 2023;132:104151. https://doi.org/
in architecture: A brief historical overview. Front Archit Res 2014;3(3):298–323. 10.1016/j.mechrescom.2023.104151.
https://doi.org/10.1016/j.foar.2014.03.006.

15