Les lames de plancher d’ingénierie sont devenues de plus en plus populaires depuis leur apparition dans les années 70. En 1999, les deux tiers des recouvrements de plancher en bois installés en Europe étaient des planchers d’ingénierie. Durant la même année, le tiers des lames de plancher installées aux États-Unis étaient des lames de plancher d’ingénierie. Malgré cette importante percée sur ces marchés, la connaissance sur le comportement du produit est limitée. Les variations des conditions hygrométriques suivant les saisons en Amérique du Nord peuvent induire des déformations hygromécaniques indésirables. L’objectif de ce travail était de comparer cinq constructions de lames de plancher d’ingénierie fabriquées de composants nord-américains. Une méthodologie pour évaluer la performance fut déterminée. Dans les limites de ce travail, la meilleure construction fut la construction faite de 4 mm d’érable à sucre en couche de surface, 8 mm de bouleau blanc comme substrat et un placage de 2 mm comme couche de contrebalancement. La déformation de tirant à cœur fut plus importante dans les constructions avec un substrat en épinette noire de même que dans les constructions sans couche de contrebalancement. L’application d’un vernis a réduit les déformations d’environ 50 %.

Engineered wood flooring is gaining in popularity since it appeared in Europe in the 70’s. In 1999, two thirds of the wood flooring installed in Europe was engineered wood flooring. The same year, one-third of the wood flooring installed in the USA was engineered wood flooring. Although engineered wood flooring captured this important market share, knowledge of its behavior is still very limited. The variations of hygrometric conditions, occurring during the transition from summer to winter in North America can induce undesirable hygromechanical deformation. The objective of this study was to compare five engineered wood flooring constructions made with North American wood components. The methodology was developed to determine the performance of engineered wood flooring. In accordance with the design parameters selected in this study, the best construction was a 4 mm-thick sugar maple surface layer, an 8 mm-thick white birch core layer and a 2 mm-thick yellow birch veneer as backing layer. The cupping distortion in black spruce core construction without backing was 20 percent higher than the white birch construction. Constructions made with white birch core showed a 10 to 40 percent lower cupping distortion than those made with black spruce core. The varnish layer plays an important role in the performance of engineered wood flooring by reducing the cupping distortion by 50 percent since it limits moisture adsorption through the surface layer. There was no significant difference in the distortion measured for two subsequent conditioning cycles indicating that the phenomenon occurs in the elastic domain.

Dimensional stability is of upmost importance in the use of wood composites in general and in layered wood composites in particular. This is especially true for appearance products such as flooring, parquet, cabinet and furniture components. Residual stresses due to the manufacturing process or the non-homogeneous adsorption and desorption of water vapor by the composite in-service may induce distortion, and consequently decrease product value. In this context, the nature of the components in a composite wood product such as engineered wood flooring determines its behavior and value.

Engineered wood flooring products are widely available in Europe and the USA. Engineered wood flooring is common in Europe where 63.8 percent of the hardwood flooring sales are for engineered wood flooring (Anonymous 1998). The market share of engineered wood flooring in the USA increased by 10 percent to reach 30 percent of the hardwood flooring market between 1988 and 1995 (Anonymous 1998, Lamy 1997). In 1999, engineered wood flooring represented 39 percent of the USA market and it is one of the fastest growing segments among floor covering materials (Anonymous 2000). Canada, is a significant hardwood strip flooring manufacturer, but it only has two engineered wood flooring manufacturers. In the USA there are approximately 35 engineered hardwood flooring plants (Anonymous 2001). Significant volumes of hardwood lumber from the USA and Canada are sold to European countries for the manufacture of these products, many of which are sold back to North America. This and the fact that engineered wood flooring uses four times less clear hardwood than traditional strip flooring, thus releasing some of the pressure on the hardwood resource, leads us to conclude that there is an opportunity to build more North American capacity to produce engineered wood flooring. This situation motivates local industries to design engineered wood flooring products better suited for the North American market and resources, thus creating a sustainable competitive advantage.

The technical literature on engineered wood flooring is limited mainly because all the research and development on those products was done on a proprietary basis. Kubler and Lempelius (1972) introduced the European technology used by this industry 30 years ago. No other work was found in the literature on this topic except for description of the manufacture of 3-ply bamboo parquet (Che and Leslie 1999, Lu et al. 1999). European and American standards were developed for those products. The French standard, NF B54-011 (AFNOR 1980) defines performance criteria such as flexural strength, bond line strength and flatness. The standard also suggests a methodology to test flooring in-service for both single-board and in service flatness. A sealed box containing air at a given temperature and relative humidity (RH) is placed over a 120 cm x 120 cm assembled flooring strips to obtain a water vapor movement only through the top surface. Two conditions are tested, 20ºC and 20 percent relative humidity (RH) and 20ºC and 85 percent RH. Flatness is determined by measuring the strip’s distortion using a 120 cm steel flat bar as reference. Distortion over the whole span of this bar must be lower than 0.4 mm to be in accordance with NF B54-011. The ANSI/HPVA LF (1996) standard does not go as far as the French standard on performance assessment of engineered wood flooring, but it defines machining requirements and tolerances for laminated wood flooring which are of 0.3 mm between the strips.

According to wood mechanics (Bodig and Jayne 1993), distortion is related to the mechanical properties of the material used in the layers of the composite and to non-homogenous moisture content. The following hypothesis was made: stresses are developed at the surface layer following the transient moisture transfer occurring when relative humidity changes between winter and summer conditions. Moisture movement occurs primarily in the surface layer of the composite. The moisture content distribution across thickness is subsequently controlled by the permeability of the other layers including the adhesive layers.

To improve its position in wood parquetry and flooring, the North American industry must develop new designs of engineered wood flooring. The objective of this paper is to present the evaluation of various core components used in engineered wood flooring constructions and to build a knowledge data base for future design development taking advantage of low-cost, locally available materials.

Five types of engineered wood flooring constructions were made following the pattern shown in Figure 2.1. The components used came from Canadian industries and are listed in Table 2.1. The core was made of low cost material such as 3B Common (NHLA) (pallet-grade) hardwood heartwood, economy grade softwood, and high density fiberboard (HDF). A 4 mm-thick sugar maple surface layer was chosen for all constructions. Following the French standard NF B 63-204-1 (AFNOR 1997) and Demange (1998), a 4 mm-thick surface layer was used, which is appropriate for high traffic intensity and offers a possibility of sanding comparable to the traditional North American hardwood flooring. Therefore, since all engineered wood floorings were made with the same surface layer, the

Figure 2.1 Engineered wood flooring construction used in this study.

stress induced in the composite by moisture content changes at the surface was nearly the same for all combinations. A 2 mm-thick yellow birch veneer was used as backing layer. It was applied to white birch, black spruce and HDF core (constructions C1, C2 and C3, Table 2.1). For the first two core materials, the composites were made with and without a backing layer (constructions C1, C2 and C4, C5, Table 2.1) to assess its influence on flooring behavior. Hygrothermal conditions typical of winter and summer conditions were reproduced in the laboratory to evaluate the cupping distortion with constructions made with different core and backing layers but with a constant surface layer. All flooring constructions were bonded by hot pressing with a urea-formaldehyde (UF) resin provided by Borden Chemicals widely used in the engineered wood flooring industry. Table 2.2 shows the resin and pressing specifications.

The constructions were evaluated following two methodologies. First, an in-service test was performed on a 61 cm by 122 cm floor made with a 19 mm-thick plywood sub-floor supported by 38 mm by 100 mm (2x4) joists (Figure 2.2) on which an engineered wood flooring was installed, alternating a 61 cm long strip and two 30.5 cm long strips bonded to the floor with a urethane adhesive as used in the industry. Second, a single sample test was performed on 3 freestanding strips sealed with silicone sealant and aluminum foil on the back face and edges to simulate the behavior in-service. On all strip samples, distortion was measured at five locations. The same measurements were made on 3 varnished freestanding strips. These strips were also moisture sealed on all faces except on top. Derived from the NF B 54-011 (AFNOR 1980) standard and simulating North American summer and winter relative humidity conditions, the samples were exposed to dry conditions (winter) established at 20±1°C and 20±5 % RH and humid conditions (summer) established as 20°C and 80 % RH. All the material was conditioned at 20°C and 50 % RH before and after manufacturing. All constructions were subjected to three cycles: humid, dry, humid, for 60, 50 and 50 days of conditioning respectively. The length of the first cycle was increased due to variation of the conditions in the conditioning room.

Figure 2.2 In-service evaluation sample with location of measurement.

The distortions were measured with a dial gauge over the width of the strip (Figure 2.3). On the large-scale floor, measurements were done along a line located at one-third of the floor edge length. Measurements were done with a gage moving along a reference line. Precision of the measurement was assured by meticulous installation of the gauge and the high number of replications on each floor (17 replications). Precision and repeatability in this measure were within the hundredth of mm. Measurements included possible distortion due to the strip end-joint (Figure 2.2).

To assess the effect of the varnish layer on the mechanical properties of maple, 10 sugar maple planks, 4 mm in thickness (radial direction), 65 mm in width and 60 cm in length were cut into four 15 cm-long pieces. Two pieces were coated on the top surface with a commercial polyurethane varnish and the two others were left uncoated. Static bending tests were performed on those samples. This test involved a center point loading with roller bearing support and a cross-head speed of 5.328 mm per minute. Span used was 96 mm (24 times the sample thickness). For the varnished samples, one test was done with the varnish layer in compression (varnish layer top surface) and the other in tension (varnish layer bottom surface).

Figure 2.3 Distortion measurement method used in this study.

A numerical criterion was used to define acceptability of cupping distortion. This was determined to be 0.25 mm for both positive and negative distortions (Figure 2.3) of a single strip. This value was obtained through a product development project with the flooring industry and it is based on qualitative and perception experiments in which twelve people were asked to walk on floor with various levels of distortion (Lefebvre and Beauregard 1999). It has already gained market acceptance with the customers of the industry partners that participated in its development. This criterion is different from overall flatness such as specified in NF B 54-011 (AFNOR 1980). It is also different from the machining tongue and groove tolerance in the ANSI/HPVA LF standard (1996).

An analysis of variance (ANOVA) was performed on the variables studied. The models used for the ANOVA are presented in Tables 2.3, 2.5 and 2.7. Figure 2.4 shows a typical curve of the distortion as a function of time and the variables studied in this project: amplitude of the distortion and slope between 13 and 28 days in a conditioning cycle. The amplitude was measured over time for each conditioning cycle. The slope was the rate of distortion in the linear segment, between day 13 and day 28 in each conditioning cycle. The maximum absolute distortion was the distortion as measured directly on the flooring strips with the dial gauge.

The objective of this study was to evaluate a set of North American components for engineered wood flooring manufacturing and to increase knowledge of engineered wood flooring behavior. The distortion in engineered wood flooring is the result of a transient phenomenon, where a high moisture content gradient in the surface layer is responsible for the cupping distortion due to shrinkage of the surface layer when the underlying layer does not shrink. Observations made on different constructions after three conditioning cycles indicated that the deformation occurs in the elastic domain. This important observation suggests that Hooke’s law of elasticity can describe the mechanical behavior of this composite within the limits of the experiment.

The results of this study show that the mechanical properties of the core layer have an important effect on the performance of engineered wood flooring. For unvarnished strips in-service and freestanding strips, the flooring constructions tested were ranked according to the MOE of the core layer material. In the case of unvarnished constructions in-service strips, it appears that the type of backing layer also influenced performance. The cupping distortion in black spruce core construction without backing was 20 percent higher than the white birch construction. Constructions made with white birch core layer showed 10 to 40 percent lower cupping distortion than those made with black spruce.

Finally, an overlaying varnish layer has a major influence on the behavior of engineered wood flooring. The cupping distortion was reduced by at least 50 percent through the use of a varnish. All varnished freestanding strips met the cupping distortion performance criterion. All significant differences between unvarnished constructions became non-significant when varnish was applied. It can be concluded that it is important to take into account the finishing system in the design of engineered wood flooring.