Mill visits to manufacturers and users of edge-glued panels were conducted in order to characterize the quality of edge-glued joints in appearance products. During these visits, panels with gluelines of good and poor quality were collected for further analysis in Forintek’s laboratory. Microscopic measurements served to determine that the maximum acceptable width of glue joints in edge-glued panels is 0.05 mm. The main causes of troublesome gluelines resulting from ripping operations are splintering at the juncture of the edge and flat surface, excessive edge roughness, and uneven straightness of the saw kerf, although the right angle of the saw is also a critical parameter. The percentage of mill-rejected panels as a result of these problems ranges from 0.5% to 3%.
A series of edge roughness measurements taken from a sample of strips from participating mills set the stage for the development of representative roughness values for the edges of strips used in the industrial production of edge-glued panels. Edge roughness measurements taken from strips ripped in the laboratory showed the impact of various factors on edge roughness values: saw blades, feed speed and chip load. Measurements taken from edges ripped with worn saw blades indicated that edge roughness cannot be used to determine saw blade wear values because average roughness values obtained with such blades were found to be similar to those of strips ripped with well-sharpened saws.
Following the laboratory assembly of panels using strips exhibiting a wide range of edge roughness, measurements revealed that edge roughness contributes to increased glueline width, a greater proportion of gluelines wider than 0.05 mm (too apparent) and a reduction in glueline shear strength. Glueing parameters (type of glue, clamp pressure, ambient temperature, etc.) were constant throughout the production of laboratory panels.
Finally, the results of this study suggest that edge roughness values of 9 µm for Ra and 80 µm for Rt allow large-volume manufacturing of panels with good quality gluelines and that an increase in edge roughness will result in more apparent gluelines.
A study was conducted with the primary objective of gathering information for the development of a protocol for evaluating the surface quality of cross-laminated timber (CLT) products. The secondary objectives were to examine the effect of moisture content (MC) reduction on the development of surface checks and gaps, and find ways of minimizing the checking problems in CLT panels. The wood materials used for the CLT samples were rough-sawn Select grade Hem-Fir boards 25 x 152 mm (1 x 6 inches). Polyurethane was the adhesive used. The development of checks and gaps were evaluated after drying at two temperature levels at ambient relative humidity (RH).
The checks and gaps, as a result of drying to 6% to 10% MC from an initial MC of 13%, occurred randomly depending upon the characteristics of the wood and the manner in which the outer laminas were laid up in the panel. Suggestions are made for minimizing checking and gap problems in CLT panels. The checks and gaps close when the panels are exposed to higher humidity.
Guidelines were proposed for the development of a protocol for classifying CLT panels into appearance grades in terms of the severity of checks and gaps. The grades can be based on the estimated dimensions of the checks and gaps, their frequency, and the number of laminas in which they appear.
This project evaluates the potential for non-structural panels in the furniture (including cabinetry) and interior finish industries in China. It entailed two stages:
1. Review of existing information on non-structural panel markets and industry in China;
2. Survey of wood-based panel manufacturers and non-structural panel specifiers in China.
137 furniture and 132 interior finish manufacturers in eastern and southern China were surveyed by phone, mail/fax, and in personal interviews. Personal interviews were carried out with 11 panel mills in the eastern region.
The literature review is based on the Preliminary Competitor Analysis for Wood Products in China (Wahl and Gaston, 2003) that was carried out for Forestry Innovation Investment. Information specific to furniture, other non-structural panel markets and more recent publications have been added to this literature review.
A research project was carried out in collaboration with researchers from both University of British Columbia and University of Toronto to develop and test a range of hollow core composite sandwich panels based on lignocellulosic materials that can extend the current applications of wood composite products such as high density particleboard and fibreboard (hardboard and MDF). With proper engineering design and unique light weight structural features, wood fibre resources will be more effectively used and the performance of each component can be maximized in these types of novel composite panels. The outcome of this project is the development of Canadian-made light weight panels containing various low density cores, including honeycomb, low density wood wool composites and cup-shaped thin fibreboard, and high density surface panels, including plywood, hardboard and high density fibreboard (HDF) for the applications in ready to assemble (RTA) modular furniture, home and commercial cabinetry and door panels.
The work completed at Forintek included:
Development of low density wood wool panels (LCD) as the core material for the sandwich panels.
Development of cup-shaped high density fibreboard (CHDF) as the core material
Evaluation of edgewise and flat compression strength and creep behaviour of honeycomb sandwich panels fabricated by UBC.
Development of book shelf panels using four different core materials.
Performance evaluation of the book shelves developed.
The results of the experimental work suggest that:
Low density composite core materials can be made by the technology developed at Forintek laboratory using low density poplar wood wool and high viscosity phenol and formaldehyde resin with steam injection hot pressing technology. However, the strength of the panels was relatively low comparing to conventional low density particleboard, OSB or fibreboard.
The experimental work carried out on the cup-shaped high density fibreboard (CHDF) show the potential for developing various light weight core materials using current MDF process technology. The internal bond strength (IB) and water absorption (WA) of the cup-shaped panels were strongly correlated with panel density. IB increased and WA reduced when increasing the panel density. The flexibility of the technology could optimize the properties and performance of CHDF through manipulating the fibre refining process, profile design, resin system and hot pressing strategy. It shows that CHDF is a good alternative material to Kraft paper honeycombs for the manufacture of sandwich panels for higher strength and performance applications.
Test results from sandwich panels made of cup-shaped fibreboard core and HDF surface show that the nominal density of the cup-shaped core was one of the most important process parameters to adjust for the improvement of the sandwich panel properties. The flat compressive modulus, flat tensile strength and short-beam strength increased when increasing the nominal density of the core panels. Furthermore, the overall density of the sandwich panels were only fractionally increased by increasing the nominal density of the core panels due to the cup-shaped shape of the core panels. It suggests that higher nominal core density should be used when higher mechanical strength of the panels is required.
To a lesser extent, fibre type in the core panels also affects the sandwich panel properties. Longer wood fibres are recommended for use in the manufacture of the core panels.
The results of the experiment also show that increasing the thickness of the surface HDF panels increased the bending strength of the sandwich panels substantially. However, the overall density also increased.
Comparing shear properties of the four different sandwich panels developed by Forintek, we can identify that the ultimate shear strengths were different for different core materials. The sandwich panel made from polycarbonate core had the highest shear strength (0.744 MPa) followed by the panel made with CHDF (0.497 MPa). The sandwich panel made from low density wood wool core had much lower shear strength (0.012 MPa) which is lower than the paper honeycomb sandwich panels previously made by UBC with the same surface and core thickness (0.024 MPa).
The sandwich panels made with high density cup-shaped fibreboard had significantly higher core shear modulus (92.0 MPa) than any other sandwich panel studied in this project.
Laminated Veneer Lumber (LVL) and plywood are the two major veneer-based wood composite products. During LVL/plywood manufacturing, the hot pressing process is crucial not only to the quality and productivity, but also to the performance of panel products. Up to now, the numerical simulation of the hot-pressing process of LVL/plywood products is not available.
To help understand the hot-pressing process of veneer-based wood composites, the main objective of this study was to develop a computer simulation model to predict heat and mass transfer and panel densification of veneer-based composites during hot-pressing. On the basis of defining wood-glue mix layers through the panel thickness, a prototype finite-element based LVL/plywood hot-pressing model, VPress®, was developed to simulate, for the first time, the changes of temperature, moisture and vertical density profile (VDP) of each veneer ply and glueline throughout the pressing cycle. This model is capable of showing several important characteristics of the hot-pressing process of veneer-based composites such as effect of glue spread level, veneer moisture, density, platen pressure and temperature as well as pressing cycles on heat and mass transfer and panel compression. Experiments were conducted using several different variables to validate the model. The predicted temperature profiles of the veneer plies and gluelines (especially at the innermost glueline) by the model agree well with the experimental measurements. Hence, the model can be used to evaluate the sensitivity of the main variables that affect hot-pressing time (productivity), panel compression (material recovery) and vertical density profile (panel stiffness). Once customized in industry, the new model will allow operators to optimize the production balance between productivity, panel densification and panel quality or stiffness. This hot-pressing model is the first step in facilitating the optimization of the pressing process and enhanced product quality.
The highlights of a co-operative research program developed by the U.S. Forest Products Laboratory (FPL) and Forintek Canada Corp. to provide detailed creep-rupture and some creep information for composite panel products are summarized here. Support for this program has been provided by the American Plywood Association, The Waferboard Association (now the Structural Board Association), as well as the U.S. and Canadian Forest Services. Commercially produced plywood, oriented strandboard (OSB), and waferboard were tested to identify three mills that produced panels with high, low and median flexural creep performance. These three plywood, three OSB, and three waferboard products were then extensively tested to provide information on their duration of load and creep performance.
The work presented in this report is a continuation of the FPInnovations' research project on determining the performance of the CLT as a structural system under lateral loads. As currently there are no standardized methods for determining the resistance of CLT shearwalls under lateral loads, the design approaches are left at discretion of the designers. The most common approach that is currently used in Europe and North America assumes that the resistance of CLT walls is a simple summary of the shear resistance of all connectors at the bottom of the wall. In this report some new analytical models for predicting of the design (factored) resistance of CLT walls under lateral loads were developed based on connection properties. These new models were than evaluated for their consistency along with the models that are currently used in North America and in Europe.
In total five different design models (approaches) were used in the study, the two existing models and three newly developed ones. All models were used to predict the factored lateral load resistances of various CLT wall configurations tested in 2010 at FPInnovations. The analyzed walls had different aspect ratios and segmentation, different vertical load levels, different connection layouts and different fasteners in the connections (ring nails, spiral nails and screws). The design values obtained using the various analytical models were compared with the maximum forces and yielding forces obtained from the experimental tests. Ratios between the ultimate loads obtained from experimental tests and design values obtained by the five analytical design models were used as a measure for the consistency of the models. Newly developed models that account for sliding-uplift interaction in the brackets (models D3-D5) showed higher level of consistency compared to existing ones. The analytical model D4 that accounts for sliding-uplift interaction according to a circular domain, is probably the best candidate for future development of design procedures for determining resistance of CLT walls under lateral loads. In case of coupled CLT walls, contribution of vertical load to the wall lateral resistance was found to be two times lower than in case of single wall element with the same geometry and vertical load. Special attention in the coupled walls design should be given to step joints between the adjacent wall panels. Over-design of the step joint can result in completely different wall behavior in terms of mechanical properties (strength, ductility, deformation capacity, etc.) that those predicted.
It should be noted that conclusions made in this report are made based on the comparison to the tested configurations only. Additional experimental data or results from numerical parametric analyses are needed to cover additional variations in wall parameters such as wall geometry and aspect ratio, layout of connectors (hold-downs, brackets), type and number of fasteners used in the connectors, and the amount of vertical load. The findings in this report, however, give a solid base for the development of seismic design procedure for CLT structures. Such procedure should also include capacity based design principles, which take into account statistical distributions of connections resistances.
This report summarizes the experimental works that was carried out for a one-year research project developed as the continuation of previous research projects on the subject of light weight hollow core sandwich panels. The experiment focused on the investigation of creep behavior of light weight hollow core panel under long term static loading and high humidity conditions and its correlation with short term properties. Five types of surface panels were used, namely, 3.2 mm thick high density fibreboard with birch veneer on both sides, two thicknesses of M2 grade particleboard (6.3 mm and 9.5 mm) and two thicknesses of medium density fibreboard (6.3 mm and 9.5 mm). All panels were fabricated to the same final sandwich thickness of 45 mm using cell size of 12.7 mm Kraft paper honeycomb.
The results of the experiment show that the strongest facing material used to make the sandwich panels was the 3.2 mm hardboard with wood veneer lamination on both sides running along the long axis of the panel and test specimen, followed by the 6.3 mm MDF and the 9.5 mm MDF. The experiment demonstrated that exposing the panels to high humidity could cause strength loss of up to half of the original strength. However, the result of the experiment also suggested that it would be difficult to accurately predict the long term creep behavior of the sandwich panels using their corresponding short term flexural properties as the correlation between creep deformation and flexural properties was rather weak under the testing procedure and condition used.