Background of the invention Man since time immemorial has used constructions and objects of bent wood. Since wood is a rigid material, it has to be softened before being formed and bent so that it does not split. Traditionally, this softening has been achieved using heat, or alterna- tively, using a combination of heat and moisture (for example using steam). Wood has also been softened by impregnating it with chemicals such as ammonia, poly- ethylene glycol and pyridine.

WO 99/20443 discloses a method for substantially increasing elasticity and bend- ability of diffuse-porous wood. According to this method, diffuse-porous wooden blanks are compressed isostatically with a pressure of at least 500 bar. Then, the dif- fuse-porous wooden blanks are glued together in such a way that the fibres of wooden blanks are orientated in parallel in the resulting composite wood material.

The modulus of elasticity can be reduced more than 10 times by that method.

One of the objectives of the technology of WO 99/20443 is to first increase bend- ability of a piece of a wooden material, bend said piece into a desired shape and fi- nally stabilising said bent piece into said desired shape. This is done by immersing the bent piece in water for 10 minutes. However, after this treatment impairs the fi- nal shaped wooden article is as hard as the diffuse-porous wood is after the isostatic compression step.

Summary of the invention Now, it has turned out that stabilised shaped products of a diffuse-porous wooden material can be made by a process comprising the steps of : a) supplying a piece of diffuse-porous wood which have been made elastic by isostatically pressing said specimens with a pressure of at least 300 bar; and b) fixing the piece obtained in step a) in a desired bent position; and c) heating to a temperature within the range from 90 °C to 225 °C until the piece has been stabilised in the desired bent position, where said heat treatment is carried out under such conditions that the net transport of water out from the piece is substan- tially reduced.

Definitions The term"isostatic pressing", which is used herein, relates to pressing with a pres- sure that equally great in all directions in space. Pressing wood with a pressure of this nature is described in WO 95/13908.

"Diffuse-porous wood"is wood in which the vessels are evenly distributed and are of approximately uniform size over the whole of the annual ring. Examples of trees having diffuse-porous wood are alder, aspen, birch, beech, maple, eucalyptus, Ca- nadian sugar maple, Betula pendula, Acer pseudoplantanus, Acer rubrum, Nyssa sylvatica, Liquidambar styraciflua, Populus balsamifera, Banksia prionotes and Banksia ilicifolia.

The term"wood specimen", or"wood piece", is used herein to signify a specimen of diffuse-porous wood. A"composite wood specimen", a"composite wood mate- rial"or a"composite wood piece"refers to a piece which consists of several smaller diffuse-porous wood pieces which have been glued together parallel to the direction of the fibres in the constituent pieces. In principle, most types of glue, which are

suitable for wood, can be used when producing composite wood specimens. Exam- ples that may be mentioned are cold-water glue, hot-melt glue, solvent-based glue, emulsion-based glue and polymerisation-based glue comprising one or two compo- nents. Use can be made, in particular, of glue containing polyvinyl acetate emul- sions, PVC, polystyrene, urea, melamine, melamine-formaldehyde, phenol and polyurethane. It is simple for a skilled person to select a suitable glue on basis of given conditions. A"flexible wood piece", or a"flexible composite wood material", respectively, refers to a piece of diffuse-porous wood that has been compressed isostatically, or a composite wood material with enhanced elasticity and bendability that has been produced in accordance with WO 99/20443.

Figures The invention will now be described with reference to the attached figures, in which: Figure 1A shows a disc that has been cut directly from a tree trunk. Figure 1B shows a horizontal cross-section of the disc. The annual rings are indicated. Figure 1C shows shaping of a flexible wood piece by fixing it in a desired position.

Figure 2 is referred to as a reference for describing how the elasticity and bendabil- ity is changed by the method of WO 99/20443; Figure 3 shows how composite wood specimens having a high degree of elasticity and bendability can be obtained by isostatically pressed diffuse-porous wood being sawn and glued in a specific pattern. The annual rings are fully indicated in this fig- ure; Figure 4 describes one way of how to form a composite wood specimen during sta- bilisation according to the present invention. A flexible composite wood specimen (2) is bent around a metal cylinder (1). Then, the flexible composite wood specimen

is fixed in this position by a bent metal piece (3) and a fastening means, such as a clamp (not shown). This arrangement (cylinder (1), wood specimen (2), and metal piece (3)) is then enclosed in a heating device (4); Figure 5 shows how to determine stiffness. A bent composite wood specimen 10 is attached to a slab 12 using two clamps 14,16 and a rod 18. A strip 20 is arranged in such a way that it touches both top ends 22,24 of the bent composite wood speci- men 10. When measuring the stiffness one end 24 is loaded with a certain force F.

The wood specimen 10 is the bent downwards and the distance 6 is a measurement of the stiffness.

Detailed description of the technology of WO 99/20443 As has already been mentioned above, the technology disclosed by WO 99/20443 is based on the unexpected discovery that the elasticity of a specimen of diffuse- porous wood is greatly increased after the wood has been isostatically pressed with a pressure of at least 500 bar. Without tying the invention to any particular theory, it is assumed that the increase in the elasticity after the isostatic pressing is due to the vessels or pores, which are quite large and uniformly distributed, in diffuse-porous wood, collapsing in an ordered structure. The strength of the fibres appears to be unchanged, as the force required to break the fibres is the same as for ordinary wood material. The increased elasticity does not therefore occur in all directions.

Figure 2 shows how the elasticity is altered in the diffuse-porous wood after isostatic pressing in accordance with the invention. Figure 2A shows a specimen of diffuse-porous wood in which the fibres are oriented from the surface ABCD to the surface EFGH. The annual rings are indicated in the surface ABCD. Figure 2B shows side DCGH of the wood specimen. Here, the fibres are therefore oriented from side DC to side GH. If a pressure is applied in the middle of the stretch DH, it is not possible to observe any increase in elasticity. Figure 2C shows side ABCD of the above-mentioned specimen. By contrast, if a pressure is applied in the middle of

stretch AD, it is possible to observe a distinct increase in elasticity. The result of this is shown in Figure 2D. By gluing diffuse-porous wood specimens together in paral- lel in the manner shown in Figure 3E, a composite wood material is obtained which possesses a very high degree of flexibility.

There are no restrictions with regard to the size of the wood specimen other than those, which relate to the size of the pressing device employed. However, it is par- ticularly advantageous to press disc-shaped wood specimens, and wood specimens having surface areas of more than 2 m2 can be pressed without difficulty as long as the size of the press permits this. Presses of the pressure cell type, which are de- scribed in SE-C-452 436, represent an example of a suitable pressing device, and the reader is referred to the above-cited WO 95/13908 with regard to the isostatic pressing of wood.

The wood specimen should have dried before the isostatic pressing takes place. It is advantageous if the moisture content has decreased to at most 50% of the content in the living wood. However, it is also possible to press moist wood isostatically if the liquid, which is pressed out can be taken care of, for example by means of absorp- tion, or conducted away from the pressing device. The technique of isostatically pressing moist wood is described in WO 97/02936.

The following preparation description discloses an example of how to manufacture a composite wood specimen having greatly increased elasticity according to WO 99/20443.

Preparation 1 A wood specimen in the form of a disc, having a diameter of 19.3 cm and a thick- ness of 1 cm, was sawn out of an aspen trunk. The disc was debarked and dried to a moisture content which was 48 % of the original (see fig. 1A and 1B). It was then pressed isostatically in a press of the pressure cell type (Flow pressure systems,

Vasteras, Sweden) in the manner described in Example 1 in WO 95/13908. The maximum pressure was 850 bar and the temperature was 33 °C. The total pressing time was 2 minutes.

The flexibility of the pressed disc was investigated. It was placed in a bowl having a maximum depth of 4 cm. The disc was so flexible that it was possible to adjust the shape of the disc to the shape of the bowl. The result is outlined in fig. 1C.

Preparation 2 A specimen of aspen having the dimensions 550 x 170 x 35 mm (Fig. 3A, the an- nual rings are indicated) and a moisture content which was 48% of that of the living tree was used as starting material. The specimen was pressed isostatically in the same manner as in Preparation 1. The maximum pressure was 1000 bar, the tem- perature 34°C and the pressing time 2 minutes. After pressing, the dimensions of the specimen were 438 x 136 x 22 mm. It was hand-planed all round to make it com- pletely smooth. The specimen was then sawn through along its length to give three specimens having the dimensions 146 x 136 x 22 mm. These specimens were in turn sawn into lamellae of approximately 20 mm in width, and the surfaces were levelled by hand-planing; the lamellae were then placed up against each other such that they lay in the same way as before sawing (Fig. 3B), and furthermore such that the three original specimens lay up against each other. Accordingly, 21 lamellae lay up against each other in the manner which is shown in Fig. 3C. A cold-water glue (Casco 3305, Casco, Sweden) was spread on the upper surface of all the lamellae apart from that furthest out to the right (Fig. 3D). All the lamellae were then turned a quarter revolution in the clockwise direction (Fig. 3E) and subsequently pressed against each other (Fig. 3F) using clamps; the glue was then allowed to dry. This re- sulted in a composite wood specimen (Fig. 3G) having the dimensions 146 x 410 x 22 mm. The specimen was crosscut at 15 mm along its length, resulting in a specimen having the dimensions 15 x 410 x 22 mm. This specimen was then

bent by hand until it was in the shape of a horseshoe having an internal diameter of 125 mm.

Detailed description of the present invention Accordingly, the present invention relates to a simple and effective method for fix- ing and stabilising the composite wood material disclosed in WO 99/20443. Ac- cording to this method a piece of elastic composite wooden material produced in ac- cordance with WO 99/20443 is bent into a desired shape. In order to choose a suit- able shape, it is important to take the maximum load of the composite wood mate- rial into consideration in order to avoid formation of cracks. As already mentioned, the elastic composite wood material is not equally elastic in all directions. That must also be taken into consideration when choosing a suitable shape. However, it is easy for the skilled person to determine suitable shapes after carrying out routine experi- ments.

The way the flexible composite wood material is fixed in a desired shape is not critical. Different devices, such as clamps, can be used in this respect. Another ex- ample is shown in the examples.

The fixed flexible composite wood material is then heated to a temperature within the range of 90-225 °C for a sufficient amount of time in order to stabilise the shape. If the process temperature is higher than 225 °C, there is a risk that the com- posite wood material will be damaged. If the temperature is lower than 90 °C the process time will be unacceptably long. There is also a risk that the composite wood material will not be stabilised.

The process time is dependent on the process temperature. A high process tempera- ture results in a short process time. Typical process times vary between 0,5 and 24 hours but both longer and shorter process times can be used depending on the par- ticular conditions.

Another important is the water transport out from the composite wood material during the heat treatment. It is important that this water transport is prevented as much as possible. This can be achieved by covering the composite wood material to at least 80 % with a water-impermeable material. One way of covering the compos- ite wood material is shown in figure 4. A flexible composite wood specimen (1) is bent around a metal cylinder (2). Then, the flexible composite wood specimen is fixed in this position by a bent metal piece (3). By this arrangement all surfaces but two small end surfaces are covered with a water-impermeable material, thus pre- venting water transport out from the specimen. In a preferred embodiment, the com- posite wood material can be completely covered with a waterproof and heat-stable material, such as a metal foil. A third alternative is to carry out the heat treatment in a heated humidity chamber where the relative humidity can be maintained at at least 65 % even when the temperature is above 90 °C. Preferably, the composite wood material is both completely covered with a waterproof material and heat-treated in a humidity chamber.

The present invention will now be described with reference to the enclosed experi- mental part. These experiments only illustrate particular embodiments of the inven- tion and should not be construed so as to restrict the scope thereof.

Experimental part Materials and methods Sample discs All experiments below were carried out with aspen discs of the dimension 200x200x5 mm. These discs were made in accordance with the method of prepara- tion 2 starting with lamellae of the dimension 200x28x5 mm.

Methodforfortyzing tlae sanaple discs Figure 4 shows one way of how to form a composite wood specimen during stabili- sation according to the present invention. A sample disc (1) is bent around a metal cylinder (2). Then, the sample disc is fixed in this position by a bent metal piece (3) and cramps (not shown).

In some experiments, the sample disc was completely covered with aluminium foil before being bent according to the method shown in figure 4.

Stiffness measurements Figure 5 shows how the stiffness was measured. An optionally bent sample disc 10 is attached to a slab 12 using two clamps 14,16 and a rod 18. A strip 20 is arranged in such a way that it touches both top ends 22,24 of the bent composite wood specimen 10. When measuring the stiffness, one end 24 is loaded with a certain force F, which in this case was 5 N. The wood specimen 10 is then bent downwards and the distance 8 is a measurement of the stiffness. Accordingly a low 8 corre- sponds to a stiff specimen.

The enclosed tables all disclose stiffness change values A8. The stiffness change is defined as ##=#original-#stiff, where #original is the stiffness before the stabilisation treatment and 5stiff is the stiffness after the stabilisation treatment. A high A8 corre- sponds to a stiff sample. A negative A6 means that the sample disc is less stiff than before the treatment.

Measurements of net water transport The net water transport during the stabilisation process is measured as percent weight loss. A positive value means that the water content of the sample disc has increased.

Experiments 15 sample discs were heated for 2 hours at different temperatures and humidity con- ditions. Some were bent in accordance with the method shown in figure 4 during the treatment whereas others were not bent. Furthermore, some of the bent discs were also wrapped in aluminium foil. In table I below, following abbreviations are used: 55 %, 55 % relative humidity; 65 %, 65 % relative humidity; oven, the disc was treated in a dry oven; wrapped, the sample disc was bent as shown in figure 4 and completely covered by aluminium foil; bent, the sample disc was bent in an open bending device as shown in figure 4 without any aluminium foil; not bent, the sam- ple disc was just inserted into the oven/humidity chamber without being neither bent nor wrapped.

The table clearly shows that the largest stiffness changes were obtained when the sample disc was wrapped and exposed to high temperatures.

Table 1<BR> Sample disc 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Temperature 70 70 70 95 95 95 95 95 95 140 140 140 180 180 180 (°C) Humidity 55% 55% 55% 65% 65% 65% Oven Oven Oven Oven Oven Oven Oven Oven Oven Conditions Bending Wrap- Bent Not Wrap- Bent Not Wrap- Bent Not Wrap- Bent Not Wrap Bent Not ped bent ped bent ped bent ped bent ped Bent Net water (%) 0.05 0.13 -0.66 0.01 -0.10 -1.22 0.26 -2.50 -3.64 -4.77 -6.11 -6.62 -6.29 -7.62 -7.82 transport ## (mm) -0.4 -1.8 -7.9 11.8 3.0 -3.0 4.0 2.1 -1.9 10.8 9.3 3.5 11.8 2.8 2.8