ENGINEERED LUMBER: LVL AND SOLID WOOD REINFORCED WITH NATURAL FIBRES

ENGINEERED LUMBER: LVL AND REINFORCED WITH NATURAL FIBRES SOLID WOOD Emanuela Speranzini1, Simone Tralascia2, ABSTRACT: This work presents the fir...
Author: Joel Simpson
1 downloads 1 Views 386KB Size
ENGINEERED LUMBER: LVL AND REINFORCED WITH NATURAL FIBRES

SOLID

WOOD

Emanuela Speranzini1, Simone Tralascia2,

ABSTRACT: This work presents the first results of an experiment on elements made of LVL wood and elements in solid wood reinforced with FRP (Fibre Reinforced Plastic) in natural fibres of basalt, flax and hemp. After a careful study of the properties of the materials to be used, two series of LVL and solid wood elements were prepared reinforced with the three types of natural fibres and were then subjected to bending tests on four points. The results of the elements that had been reinforced in this way were compared with those obtained by bending tests on elements with no reinforcement or reinforced with traditional glass or carbon fibres. The natural fibres behaved well as regards flax and hemp fibres. KEYWORDS: Solid wood, LVL, engineered lumber, FRP, natural fibres

1 INTRODUCTION 123 This research aims to strengthen the properties of wood by joining the wooden elements with FRP of natural fibres to create products that fulfil the philosophy of current sustainable projects. For this purpose an experiment on elements made of LVL (Laminated Veneer Lumber) and solid wood reinforced with the natural fibres of basalt, flax and hemp is currently being run. Many experiments have been carried out in the past on the reinforcement of wooden structures using composite materials, such as the case of CFRP sheets on life-size or restricted size test samples and these were subjected to bending tests in order to evaluate the increase in stiffness and strength [1,2 and 3]. The first stage of this experiment used test samples made of LVL and solid wood reinforced with one-way bands of natural fibres (basalt, falx and hemp) subjected to bending tests. The materials used were tested to characterize their mechanical properties. Special attention was paid to the choice of resin for which traction shear tests were carried out on the bonding together with adhesion tests for pool-out in order to test the quality of the adhesion. The results obtained from the samples with natural fibres were compared with those obtained from samples

without fibre or reinforced with traditional glass or carbon fibres.

2 MATERIALS USED TO PREPARE THE ELEMENTS 2.1 WOOD Solid wood of natural fir tree and LVL with the same tree oil were used to prepare the test samples (Table 1). The LVL used is a product that consists of “continual”, very thin (no more than 6 mm), overlapping layers of wood, joined together to form a cross section half of which consists of adhesives capable of guaranteeing strength and durability over time. The layers are usually all placed with the grain facing in the same direction. Some LVL wood products could, however, be made with the grain facing orthogonally compared to the majority of layers that make up the “parcel” (usually in a 1:5 ratio). Table 1: Physical and mechanical properties of the solid wood and LVL LVL LVL Wood (1:5) (1:1) Density [Kg/ m3] Bending strength [N/mm2] 2

Compressive strength [N/mm ]

430

510

510

67

36

50

38

26

35

5.1

1.3

2.3

14000

10500

13800

1

Emanuela Speranzini, Faculty of Civil Engineering, Perugia University, via G. Duranti n. 93, 06125 Perugia, Italy. Email: [email protected] 2 Simone Tralascia, Faculty of Civil Engineering, Perugia University, via G. Duranti n. 93, 06125 Perugia, Italy. Email: [email protected]

2

Shear strength [N/mm ] Modulus of elasticity [N/mm2]

2.2 RESIN The careful choice of resin and adequate preparation of the adhesion surfaces are necessary requisites to maintain the integrity of the element made of wood and FRP. One of the factors that most influences the duration of the bonding is the absorption of humidity. In FRP materials, absorption is significantly lower compared to that of wood: this causes dissimilar materials to behave differently. Therefore, when the FRP-wooden interface is in a working position and subjected to changing environmental and hygrometric conditions, induced stress is created along the interface. As time passes, these actions can weaken the adhesive interface and in some cases may cause the failure of this composite material. The resin used was chosen on the basis of both research studies previously carried out on glued wood and as a result of the tests carried out on three types of resin (named A, B and C for commercial reasons): fluid epoxy resin “A”, bi-component thixotropic epoxy resin “B” and adhesive epoxy mortar “C” (Table 2).

3 TESTS ON THE CONSTITUENT MATERIALS 3.1 TESTS ON THE RESINS 3.1.1 Traction shear tests for bonding The aim of the test was to define the quality of the bonding for resins used in wooden structures. Each of the 30 samples was obtained by bonding together two non-treated, wooden elements (100 mm long, 10 mm wide and 50 mm deep) using the resin to be tested. The overlapping area was 10 mm long. The wood samples for bonding were first acclimatised under standard environmental conditions in order to guarantee uniform results. The test was carried out on thin lines of resins approximately one millimetre thick. The test sample was subjected to an increasing traction force until the joist broke (Figure 1). The break traction force represents the shear strength of the resin.

Table 2: Physical and mechanical properties of the resin Resin A

Resin B

Resin C

Density [g/cm3]

1.8

1.10

1.10

Viscosity [poise]

8-10

4-6

6-8

Mixing ratio resin/hardenr

100/50

100/50

100/50

Hardening time

7 giorni

7 giorni

7 giorni

>50

>95

>38

>30

>30

>23

Compression strength [N/mm2] 2

Tensile strength [N/mm ]

2.3 FIBRES The fibres used in the experiment to make the reinforcement are: glass fibres, carbon fibres, basal fibres, flax fibres and hemp fibres. Their technical specifications taken from the manufacturers’ technical charts are given in Table 3. Table 3: Physical and mechanical properties of the fibres EStress Strain Density N moduls 2 [N/mm ] [%] [g/cm2] yarrns/cm [N/mm2] Carbon

4800

240000

2,00

1,78

4,00

Glass

2000

73000

3,50

2,50

2,50

Basalt

2800

89000

3,15

2,8

3,3

Hemp

1000

40000

2,5

1,5

7

Flax

1500

50000

3,0

1,5

9

Figure 1: Traction shear tests for bonding

The test samples subjected to traction behaved differently; three types of break were identified; a cohesive break, an adhesive break and a break in the connection joist. The cohesive break occurred within the material of the joist. This type of break was found on the majority of test samples and reflected the ideal conditions for resin application. The adhesive break occurred between the adhesive and the adherend when the interface was weaker than the adherend and this was observed in very few cases. The fracture/break in the joist only occurred in one case. The maximum value of adhesion was reached by the epoxy resins A and B (Table 4). The C-type resin reached breaking point at lower load values compared to the other types used. Resin A became the final choice not only because of its greater transparency but it also enables a better bonding between wood and fibre, so that the fibre has greater wettability and consequently this benefits adhesion to the wooden surface.

Resin A

1895,38

92,652

0,02156

along a 20 mm length, since in the tests with a 30 mm long bonding strip the tangential stress was minimal after the initial 20 mm. The results obtained were corrected according to the aforementioned observations and proved to be less than the maximum strength the system could actually develop.

Resin B

1353,89

66,873

0,01817

3.2 TESTS ON THE FIBRES

Resin C

1220,66

57,682

0,01737

The mechanical properties of the fibres were verified by means of a series of tensile tests in compliance with the norm ASTM D 3039 [4]. The tests were carried out to control shifting at a speed of 2 mm/min using an Instron 4505-type dynamometer (Figure 3).

Table 4: Test results of the 30 samples Maximum Maximum Mean strain load stress 2 [N] [N/mm ]

3.1.2 Adhesion test for pool-out The adhesion tests for pool-out enabled the adhesion force to be evaluated according to the length of the bonding. Test samples were prepared, each of which consisted of two wood elements to which different types of fibre had been glued according to the different 10 mm, 20 mm and 30 mm lengths of anchorage.. The “A”type resin was used following the results obtained from the traction shear tests for bonding. The segments were then subjected to an adhesion test performed at a constant strain rate of 2 mm/min (Figure 2).

Figure 3: Tensile tests: a)good break, b)and c) unacceptable

Figure 2: Adhesion test for pool-out

After the tests the transmissible force was observed to demonstrate good anchorage for all the types of fibres under examination. By combining the results of the tests it was observed that only the first centimetres of anchorage transmit appreciable strain during the elastic stage and in the light of the tests a 10 mm long anchorage is sufficient to guarantee that the system will function correctly. Table 5: Test results of pool-out test Maximum Tangential Stress [N/mm2] l=10mm

l=20mm

l=30mm

Carbon fibres

70,75

152,6

---

Glass fibres

115,46

120,29

117,78

Basalt fibres

75,42

148,67

143,97

Hemp fibres

75,47

82,34

113,38

Flax fibres

75, 45

88,32

115, 78

Table 5 shows the average values obtained for the maximum tangential stress of adhesion, which was measured by assuming a uniform distribution of stress

The test sample consisted of a total of 50 samples made with an epoxy polymeric “A” type matrix and one-way strips of the various types of fibres. All the test samples had the same fraction ratio in weight. The composites break behaviour was particularly complex and depended on various factors; for these reasons the break criteria chosen for “maximum stress” for the comparison refers to a macro-mechanical scale based on the assumption of homogeneity and of a linear behaviour up to the point of collapse, as confirmed by the experiment (Table 6). Table 6: Results of the tensile tests on the fibres Maximum stress [N/mm2] Carbon fibres

478,93

Glass fibres

142,08

Basalt fibres

244,6411

Hemp fibres

36,3076

Flax fibres

25,4048

4 BENDING TESTS ON THE REINFORCED ELEMENTS In order to evaluate the flexural behaviour of the elements reinforced with natural fibres, a series of bending tests were carried out on four points in compliance with the norm EN 14374, the layout of which is given in Figure 4.

4.1 TEST SAMPLES MADE OF SOLID WOOD

Figure 4: Test layout

Tests were carried out on solid wood elements both with and without reinforcement, elements made of grain oriented LVL (ratio of 1:5) and test samples of LVL with the grain facing in the same direction (ratio of 1:1) without reinforcement either edgewise or flatwise. The wood samples prepared for the tests were all the same size (40x50x1000 mm). Special care was taken over the positioning of each sample which rested on cylindrical rollers capable of providing sliding support. In order to reduce local crushing of the wood, appropriate divisions were inserted not only between the test sample and the load pressure pads, but also between the test sample and the supports. A hydraulic jack placed inside the metal frame, connected to the pump via a hydraulic circuit was used to apply the load up to breaking point. LVDT-type (Linear Variable Differential Transformer) displacement transducers were used to obtain movement. The values measured during the test enabled the flexural strength fm to be calculated according to the equation:

fm 

a  Fu 2 W

In the flexural analysis of the non-reinforced wooden element it was observed how the way in which it breaks depends essentially on: a) the ratio between the ultimate tensile strength and compression strength; b) the nonlinear behaviour of wood under compression at its final limit state; c) the volume of material subjected to the tensile test, a parameter that is directly proportional to the probability of the effects of localised defects. In general, it can be said that the most common collapse in solid wood occurs when the limit strength value is reached in the tension area (Table 7), when plasticisation is present in the compressed area. Table 7: Mean value of the bending test results on unreinforced solid wood

samples

Mean value of maximum load [KN]

Mean value of flexural resistance [N/mm2]

Mean value of E-moduls [N/mm2]

6,12

58,03

9722,65

As an example, Figure 5 shows the diagram which represents the behaviour of the non-reinforced test sample. The upper curve represents the behaviour of a test sample without any defects: the break occurs in the tension area, where a longitudinal lesion splits at rightangles to the grain. The other curve represents materials with defects: the inclined fracture follows the grain and starts from the node in the central part of the test sample.

(1)

where: Fu is the final load (N) a is the distance between the point of application of the load and the nearest support (mm); W is the section modulus (mm3), determined according to its actual size. A test with the same geometry and load as that set by the norm was carried out to establish the modulus of flexural elasticity was established in order to determine flexural strength. The movement w was measured in the middle of the beam using high precision equipment. The modulus of elasticity was calculated using the following equation:

E m ,g

3 l 3  F2  F1   3a   a    3        bh  w2  w1   4l   l  

Figure 5: Diagram of load-deflection of a solid wood sample with and without defects

(2)

where:

l (mm), is the distance between the supports; F2  F1 (N), is an increase in load on the rectinear stretch of the load-movement curve;

w2  w1 (mm), corresponding to

is

the

F2  F1 .

increase

in

movement

In the case of reinforced wood, the insertion of FRP aims to obtain elements with higher resistance. Thus, the ratio between the ultimate tensile strength and compression strength increases so as to exploit the compressed flap in order to enable plasticisation of the material and ensure greater ductility of the collapse (Figure 6). Thanks to the progressive movement of the neutral axis towards the tension area, a greater tensile strength is reached as the influence of the defects is reduced.

From the diagram in Figure 6 the reinforced test samples can be seen to behave very differently to the nonreinforced sample: i.e. they show a much higher break load and greater ductility due to the plasticisation of the cross-section. The ultimate load is maximum in the case of reinforcement with carbon fibre (approximately 42.32%) and minimum in the case of reinforcement with basalt, glass and hemp fibre (24.00%) which show the same values on average (Table 8).

Figura 6: Load-deflection diagram of the samples with and without reinforcement

were even higher than those given by carbon fibres, probably as a result of a better adhesion by the flax and hemp fibres to the wood. 4.2 LVL 4.2.1. LVL with the grain facing in the same direction in a ratio of 1:5 As regards LVL, since the collapse is directly connected to the quality of the wood used, the checks of the various stages of manufacture make its behaviour qualitatively different to solid wood. The bending fracture tests show how difficult it is for the compressed edge to reach the state of plasticisation. LVL wood shows a tendency to be have a far more fragile breaking point than solid wood. (Figure 7). The results of the tests carried out on reinforced samples show a modest increase in bending strength compared to the elements without reinforcement. The elastic moduli do not undergo any substantial variations, neither do the deformations differ greatly in the various cases. Sometimes the beams broke cleanly into two parts. In a part of the samples, however, the break occurred in the area under maximum bending and shear stress. In these cases the shear tension in the beam almost reached the values of shear strength supplied by the manufacturer. The reinforcement was, therefore, unable to contribute to high values of bending stress.

solid wood

As regards the natural fibres, the samples behaved well when reinforced with flax fibre (35.35%), despite the presence of a much lower characteristic break strength compared with that of carbon. Table 8: Mean values of the bending test results of the solid wood samples Mean value Mean value Mean value of of flexural of maximun resistance E-Moduls load [N] [N/mm2] [N/mm2] 8067,18 72,00 Unreinforced 14226,79 Carbon Reinforcement Glass Reinforcement Basalt Reinforcement Hemp Reinforcement Flax Reinforcement

11481,31

103,33

16270,29

10055,38

90,49

15610,88

9942,49

89,49

15112,34

10004,32

90,04

16588,77

10919,09

98,27

17666,21

As regards the modulus of elasticity, a modest increase was identified in the samples that were reinforced with fibres compared to those without reinforcement. The result obtained with flax and hemp fibres was very interesting. These samples showed strength values that

Figura 7: Diagram of load-deflection of some samples without reinforcement

LVL

4.2.2. LVL with the grain facing in the same direction in a ratio of 1:1 Bending tests with two different layouts were carried out on LVL samples with the grain facing in the same direction (“package” in a ratio of 1:1) without reinforcement. The load was applied to the plane of the element (case a: edgewise bending layout) or at right angles to it (case b: flatwise bending layout). The results of the experiment are shown in Table 9. . All the samples showed a linear elastic behaviour up to the breaking point, after which none of the tested elements broke into two pieces due to the mutual weave between the veneers and the action of the adhesive substance.

Table 9: Summary of the results of the bending tests on the test samples in both edgewise and flatwise layout Mean value of maximun load [N]

Mean value of bending resistance [N/mm2]

Mean value of E-moduls [N/mm2]

Edgewise layout

7220

68,58

16287,77

Flatwise layout

5520

69,03

13932,31

The way in which the edgewise layout broke was similar to that of the solid wood test samples (Figure 8).

Figure 8: Edgewise layout

In the flatwise layout all the cracks alternated between the single layers, and long cracks, which were often localised on the interface of single layers, ran parallel in the direction of the position of the layers themselves (Figure 9).

and rigidity far superior to those of non-reinforced wood. The comparison between reinforced and non-reinforced wood confirmed one of the fundamental aspects behind the idea of reinforcement, i.e. to obtain a more homogeneous, final product once the influence of physiological defects in the wood on the structural behaviour had been reduced. The global behaviour of the element obtained in this way gave positive results even in terms of ductility. On the contrary to solid wood which, on bending, presents a fragile break in the taut area, the reinforced beams broke in the taut area with greater ductility due to the plasticization of the composite cross-section. For all types of natural reinforcements applied a considerable increase in strength and the elasticity modulus were obtained and even the movements registered were weaker compared with those of the beams in solid wood without reinforcement. More modest advantages were obtained by applying reinforcements to LVL wood. It was impossible to exploit the reinforcement to its best advantage due to the weak, shear strength of the material. A detailed investigation into LVL will shortly be restarted, as it had been stopped due to the impossibility to source other material; test samples with reinforcement for LVL with the grain in a sole direction will also be tested. This first stage will be followed by an experiment on life-size test samples to confirm the results obtained.

ACKNOWLEDGEMENT Figure 9: Flatwise layout

The strength and elastic modulus values of the LVL test sample were much higher than those expected for solid wood and GLULAM. A very important result, that confirms the characteristic of LVL wood, is that the break in the solid wood samples began from the defects, particularly from the knots, whereas the break in LVL wood samples occurred after the fibres had resisted traction in the taut area, which is typical of a bending test and a sign, therefore, of the tendency of the LVL wood to be a flawless material. As regards solid wood, the deformations which occurred at break point in the LVL samples tested in an edgewise layout are noticeably less severe than those in solid wood and this result was appreciable for the wooden structures where it is sometimes difficult to reduce the extent of the deformations.

5 CONCLUSIONS This work has enabled to test the mechanical behaviour of elements made of wood reinforced with natural fibre composite materials to be tested. The use of natural fibre composite materials enabled wooden beams to be obtained with properties of strength

Our thanks goes to the Region of Umbria for their financial support of this research project. We would also like to thank Fida s.r.l. of Perugia for the supply of the fibres and KIMIA for the supply of the resins.

REFERENCES [1] Borri A., Corradi M., Speranzini E.,: Travi in legno rinforzate con barre o con tessuti in fibra di carbonio, (Wood beams reinforced with bars or with carbon fibres sheets), Italian, L’Edilizia, No 4, 2001. [2] Borri a., Corradi M., Grazini A., A method for flexural reinforcement of old wood beams with CFRP materials, Journal of Composite part B 36/2, pp. 143-153, 2005. [3] Plevris N., Triantaflou T.C., FRP Reinforced Wood as Structural Material”, Journal of Material in Civil Engineering, ASCE, Vol. 4, No 3, 8, 1992. [4] ASTM D3039: Standard test method for tensile properties of polymer matrix composite materials. [5] UNI EN 302-1: Adhesive for wood structures. Tests. - Adesivi per strutture in legno. Metodi di prova. Parte 1: Determinazione della resistenza del giunto al taglio a trazione longitudinale [6] UNI 9595: Adhesive, Traction shear tests. - Adesivi. Determinazione della rapidità di presa a freddo di adesivi per legno mediante prove di taglio per trazione”

Suggest Documents