See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228841056

MECHANICAL BEHAVIOUR OF STRAW CONSTRUCTION FOLLOWING THE

GREB TECHNIQUE.

Article · October 2009

CITATIONS READS

0 1,454

3 authors, including:

Laurent Arnaud Frédéric Sallet

Ecole Nationale Supérieure d'Arts et Métiers Ecole Nationale des Travaux Publics de l'Etat
52 PUBLICATIONS   925 CITATIONS    5 PUBLICATIONS   18 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Frédéric Sallet on 26 May 2014.

The user has requested enhancement of the downloaded file.

Proceedings of the 11th International Conference on
Non-conventional Materials and Technologies (NOCMAT 2009)
6-9 September 2009, Bath, UK

MECHANICAL BEHAVIOUR OF STRAW CONSTRUCTION

FOLLOWING THE GREB TECHNIQUE.

Laurent Arnaud 1, Céline La Rosa, Fréderic Sallet 2

1, 2
Université de Lyon, Lyon, F-69003, France ; Ecole Nationale des Travaux Publics de
l'Etat, CNRS, URA 1652, Département Génie Civil et Bâtiment, 3, rue Maurice Audin,
Vaulx-en-Velin, F-69120, France

Abstract : Building construction with straw bales is of interest since straw is available
everywhere, easily renewable and recyclable, and it also results in a very low
environmental impact. This paper focuses on the mechanical properties of walls built
following the GREB technique.
The GREB technique was developed in Canada. It combines various components : straw
bales are used as filling material in a timber load bearing structure reinforced with metallic
ties between each layer of straw bales. The straw bales are then rendered in between the
wood posts. This complex structure is not easy to be mechanically optimized.
The paper presents an experimental approach to understanding and then optimizing the
mechanical behaviour of a full size wall based on straw bale construction. Characterization
tests have been performed for each component. A novel mechanism allowing the
assessment of the bi-dimensional behaviour of walls has been developed at ENTPE
laboratory. Walls (2.40m x 1.70m) are loaded under both compressive and shear stress.
The main interest is that these compressive and shear stress can be applied
independently.
Alongside this a numerical simulation based on finite element software is produced. This
study identifies the mechanical influence of each component and assesses the elasto-
plastic behaviour of the beam - post assembly.

Keywords : Straw bales, GREB technique, mechanic, bi-axial test, numerical simulation.

1 Introduction
Sustainable construction nowadays requires the use of environmental friendly
materials, high levels of performance (especially thermal performance) and low cost of
construction. That is the reason why people are looking differently at the natural building
materials. The use of straw as building material enters this category.
Straw construction presents some advantages such as renewability with low energy
input, recyclability, CO2 storage, and local availability that implies short distribution cycle
and consequently very low transportation cost.
Nevertheless, attention needs to be paid to the design of buildings using such natural
materials. Intrinsic properties are not well-known. This is the same for the construction
system : new natural materials need specific construction systems.

1
Professor, [email protected]
2
Technician, [email protected]
 Figure 1 : The GREB technique : 1- wooden frame structure, 2- straw bales 3-
reinforcement with screws, nails and iron strap, 4- lightweight mortar
(Brossamain and Thevard 2006).

This paper deals with the straw-bale construction following the GREB technique. This
technique was developed in Canada and it is now used particularly by self-builders of
houses in France.
As shown figure 1, the technique combines a wooden bearing structure with an infill of
straw bales reinforced by nails, screws and metal straps. Finally the straw is rendered in
between the wooden structure. The experimental programme and the numerical approach
is presented below.

2 GREB technique : experimental programme

2.1 A composite material

Building construction based on straw is nowadays ignored and under valued in part
due to the fascination for modern materials such as concrete steel or bricks. Nevertheless,
in the particular context of the “Grenelle de l’Environnement” and the need for more
sustainable materials, natural materials are becoming more popular. So does straw bale
construction. Studies have shown that such construction could have a significant influence
on thermal performance of houses (Goodhew et al. 2003). There are very few references
for straw construction in France. Many examples can be found in UK, USA and Canada.
Various techniques exist for straw construction depending whether the straw bales play
structural role or not (Grelat 2004). This paper focuses on the GREB technique that has
been developed in Canada. The principle of the GREB technique is to associate a
structural wooden structure with straw bales using nails and metal straps and finally, a
lightweight mortar is cast in-between the wooden structure against the straw bales (cf.
figure 1).
Various parameters need to be considered in the design of such structures since many
materials and many kinds of connections between these materials are involved.
First, the load bearing structure is composed by similar beams assembled between
each other using screws (figure 2).
Some experiments have determined the compressive strengths and the flexural
resistance of timber samples for two species (douglas fir and pine). Classically, wood
exhibits an anisotropic elasto-plastic behaviour. An example of experimental results is
presented figure 2 and table 1 synthesizes the mean values of the mechanical parameters

2
measured.
Then, as it is relatively difficult to assess the mechanical resistance of these
assemblies, real structures have been tested in order to measure their initial mechanical
behaviour (elastic modulus) and their compressive and shear strengths (see figure 3). The
objective is also to determinate the resistance of the screwed assembly.

Figure 2 : Strength - strain curves for compressive strength tests on wood samples
(Douglas fir and pine species).

Table 1 : Mechanical characteristics of the materials used for straw-bale construction.

ρ E Rc ε ν Rt
3
(kg/m ) (MPa) (MPa) (-) (MPa)
Pine 494 2750 23.4 0.015 0.48 25.1
Douglas fir 606 2560 >0.015 23.0
Straw Bale 129 0.75 - > 0.1
Mortar 28 days 0.72 0.011 0.15
Mortar 36 days 1230 940 0.77 0.012 0.16 1.39
Metal strap 2700 70000 0.33
(Biblio)

Figure 3 : The GREB technique is based on a wooden bearing structure built with the
same beam section, the different beams are assembled with screws.

3
 Figure 4 : Four point bending test on wood sample, this test allows the determination of
the tensile strength of wood species.

Straw bales were tested under compressive load at the Magdeburg University
(Danielewicz and Reinschmidt 2007). The material is highly compressible and it exhibits
an elasto-plastic behaviour with non-linear secant modulus : it can be strained with an
increasing modulus without breaking abruptly.

Figure 5 : Compressive strength tests on straw bale sample (Danielewicz and

Reinschmidt 2007).

Then, the lightweight mortar is used, it composed of sand (0/4), lime based binder
(Calcitic Lime) and cement (CEM II B LL 32,5), and sawdust. The use of sawdust is
intended to reducine density and increase the vapour diffusion capacity throughout the
wall. The mortar is characterized by a low density, a reduced compressive strength and a
ductile behaviour (Table 1).

2.2 Wall tests

To understand the mechanical behaviour of the straw wall, life-size walls (2.40 m long x
1.60m in high) were built at the laboratory. These samples were tested first under normal
distributed loads, then by combining normal and shear loads as shown in figure 6. Cyclic
loads were applied to distinguish elastic and plastic strain, they were followed in stages by
a constant load in order to assess the creep behaviour of the wall prototypes (cf. figure 7).

4
 The maximum load is calculated to reach the maximum value for a two-storey house,
following the Eurocode 1 and 2 (La Rosa 2008). The maximum load is evaluated as 33.18
kN. For the shear load related to the wind load, the side wall is subjected to a distributed
load of 2 kN/m (Eurocode 1), the transverse load is then evaluated to 4.8 kN.

Figure 6 : Bi-axial test device, principle and photography of a tested sample, life-size wall
built following the GREB technique.

Figure 7 : Bi-axial test device, experimental programme for load versus time (left side),
photography of a wall sample loaded (right side).

Strains in the three space directions are followed up by means of 14 displacement

sensors (LVDT sensors). The measurements are recorded and stored for future analysis.

5
 Figure 8 : Axial displacement (x in mm) for various LVDT sensors as function of applied
axial load (y in kN) : bulk modulus of the life-size wall under normal loads.

These tests lead to the following conclusions.

The wall samples under compressive and shear loads show an elasto-plastic
behaviour. For low levels of normal loads, a purely elastic behaviour is seen, it
corresponds to the reaction of the load bearing structure. Under increasing normal
distributed loads, we observed increasing strains with a plastic component and significant
differences for strains on each face of the wall and at the centre (figure 8). These
differences can be explained due to the non-symmetrical assembly of the wooden beams
in the load bearing structure. For the numerical simulations, a bulk modulus is fixed at 34.5
kN.
Under a normal load of 38 kN, for increasing shear load a linear strain is observed and
the shear bulk modulus is evaluated experimentally to 30.7 kN/mm.

3 Numerical modelling
Numerical modelling aims at improving the design of straw-bale construction.
Nevertheless, the modelling needs a good understanding of how this composite wall works
mechanically. The calculations were performed using Comsol® Software.
In a first approach, we consider an isotropic linear elastic behaviour of the structure.
The wooden beams assembly are assumed to be perfect, that means no differential
displacement (rotation) is allowed. The anisotropy of straw bale is not considered. We
assume the displacement continuity between mortar and wood structure. The boundary
conditions are summarized figure 9. They correspond to the experimental maximum loads
applied.
The comparison between the experimental results and the modelling shows very
consistent values for the bulk moduli under small normal loads and also for small
associated shear loads.
Nevertheless, significant discrepancies are noted for the displacement in particular at
the angles of the load bearing structure. The assumption of perfect screwed assembly is
no longer valid since plastic strains are measured under higher loads. The render coat in

6
between the structure against the straw bales permite the reduction of the plastic strain of
these assemblies. The objective of the on-going work is now to propose a realistic
modelling for the elasto-plastic behaviour of the screwed wooden beam assembly.
The modelling shows clearly the role of the metal strap in the reduction of the buckling
of the wooden posts.

Figure 9 : Numerical modelling, boundary conditions considered

Figure 10 : Numerical modelling, displacement in direction 2 for the complete wall (left
side) and strain analysis of the angle timber assembly in detail (right side).

5 Concluding remarks
This paper deals with the mechanical behaviour of straw bale construction. Following
the GREB technique, the constructive system is based on a timber load bearing structure
composed of screwed timber beams of same section, straw bales used as filling material,
reinforcement of the structure between each layer of straw bales with metallic straps, then
render in between the timber post and straw bales. This complex structure is not easy to
be mechanically optimized.

7
 This paper presents a comparison between experimental measurements on life-size
samples of wall and numerical modelling based on finite element software.
These combined approaches enables to determine first the main mechanical
parameters for all the components (wood, straw bale, mortar) involved in the straw bale
construction. Then with the bi-axial strength test device, the elasto-plastic behaviour of
such composite structures is specified.
The bulk moduli have been experimentally deduced and they are compared to the
numerical modelling achieved. This study underlines the fundamental role of the wooden
beam assemblies in the load bearing structure (not yet well-defined).
Subsequent work will aim at measuring the thermal performances of straw-bale
construction including the moisture transfers in order to optimize a straw-bale constructive
system with a multicriteria approach.

6 Acknowledgments
Building with straw bales is nowadays diffused and encouraged through the association
“Approche Paille”. The authors want to warmly thank the association and in particular J.B.
Thevard for their collaborations.

7 References
BROSSAMAIN V. THEVARD JB, 2006, Construire son habitation en paille selon la
technique du G.R.E.B, Edition V.BROSSAMAIN, Mars 2006
DANIELEWICZ I., REINSCHMIDT J., 2007, Electromechanic tests with big bales at the
University of Applied Science Magdeburg-Stendal, Internal report, 2007
GOODHEW S., GRIFFITHS R., WOLLEY T., 2003, An investigation of the moisture
content in the wall of a straw-bale building. Building and environment, 17, 2003
GRELAT Alain, 2004, Utilisation de la paille en parois de maisons individuelles à ossature
bois, Extrait du rapport final, Tome 2, Expérimentations en laboratoire, Edité par le
Centre d’Expertise du Bâtiment et des Travaux Publics, Juillet 2004
LA ROSA C., 2008, Contribution aux mesures mécaniques sur la construction paille
suivant la technique GREB, TFE Ecole Nationale des Travaux Publics de l’Etat, Vaulx-
en-Velin FRANCE.

View publication stats