NDT.net • Apr 2006 • Vol. 11 No.4

Evaluation of density distribution in wood-based panels using X-ray scanning

Xiaodong Wang, Alexander Salenikovich
Ph.D candidate and Assistant Professor, Department of Wood and Forest Sciences,
Laval University, Quebec, G1K 7P4, Canada;
xiao-dong.wang.1@ulaval.ca and Alexander.Salenikovich@sbf.ulaval.ca; Fax: 1-418-656-5262

Mohammad Mohammad, Lin J. Hu
Research Scientist and Group Leader; Research Scientist,
Forintek Canada Corp., 319 rue Franquet, Ste-Foy, Quebec, G1P 4R4, Canada;
mohammad.mohammad@qc.forintek.ca and Lin.hu@qc.forintek.ca; Fax: 1-418-659-2922

14th International Symposium on Nondestructive Testing of Wood
May 2005, University of Applied Sciences, Germany, Eberswalde.
Published by Shaker Verlag (ISBN 3-8322-3949-9).


Abstract

The density distribution within a wood-based panel is one of the most important panel quality attributes which largely determines panel end uses. Limited work has been done on the non-destructive evaluation of localized density and its distribution in a panel. This exploratory study was intended to examine the feasibility of using X-ray technology for this purpose.

A total of 20 panels of medium density fiberboard (MDF), 16-mm thick, particleboard (PB), 16 mm thick, and oriented strand board (OSB), 11 mm, 15 mm, and 18 mm thick, with 4 replications each, were scanned using a commercial X ray system VSX9811 to obtain in-plane (horizontal) density distribution of the full size panels. Then, 50-mm square samples representing all density ranges were cut from the panels to measure through-thickness (vertical) density profiles using a commercial QMS Density Profiler QDP-01X. To evaluate the accuracy of the density profiles determined by the two X-ray systems, the density of the 50-mm specimens was determined using mass/volume measurements. Three hundred 25-mm OSB square samples were tested to examine the effect of resolution on the accuracy of the X-ray measurements.

Results showed excellent correlations between both density profiles and the mass/volume measurements on 50-mm samples with the coefficients of determination as high as 0.8, except for the horizontal density profile of the MDF panels (R2 = 0.42), which had very low density variation. The correlations with the 25-mm OSB samples were significantly weaker due to low resolution of the VSX9811 X-ray system.

It can be concluded that X-ray scanning technique has great potential for non-destructive evaluation of density and density distributions for a broad range of wood composite panels. Further research efforts are needed to improve the accuracy of the current X-ray scanning techniques and systems by using proper scanning and image resolutions, scanning sensitivity, and by improving the calibration procedure used for the various wood-composites.

1 Introduction

The wood-based panel industry has grown quickly and is predicted to continue its growth with increased efficiency in the utilization of row material. As a result, the wood panel producers are exploring new markets for their products. However, such new markets require improved performance of the wood panels in terms of design and manufacturing processes to suit their end-use applications. An example of such new markets is the upholstered furniture industry, where wood panels could be used efficiently mainly due to the development of the CNC technology. Of particular interest is the control of the mean density and density distributions within the panel (both in-plane and across panel thickness) as they have a significant effect on the strength properties of these panels. In-plane (horizontal) density variation exists due to random particle or strand deposition in mat forming [1]. This variation could strongly affect the fastener holding capacity of the panel which is a major concern in certain end-use products, such as the upholstered furniture [2]. So far, panel producers have been focusing on improving the properties of the panels by increasing the mean density while maintaining the cost-effectiveness of production. Less success has been made in reducing the in-plane density variation, especially for oriented strand board (OSB).

Although the horizontal density variation has such a significant impact on panel performance, characterizing it is still difficult in many ways. Current quality control procedures used in the industry rely on destructive evaluation techniques that are inefficient, expensive, and labor-intensive. There is an imperative need for reliable and fast means of evaluating each panel quality without affecting the structural integrity of the material. The X-ray scanning technology has a good potential for providing reliable and efficient process of continuous evaluation of panels in production. However, application of the technique to measure the density of full-size and small-size panel samples is relatively new and its accuracy needs to be further explored even though commercial X-ray devices are available on the market.

This objective of this study was to evaluate the accuracy of two commercial X-ray systems in measuring the panel densities and density distributions.

2 Experimental Procedure

2.1 Material Preparation

For this study, a total of twenty full-size (1.22 by 2.44 m) panels were received from industrial partner companies with four replications of each of the following products: 1) medium density fiberboard (MDF) panels, 16-mm thick, grade 150; 2) particleboard (PB) panels, 16-mm thick, grades M2 and MS (two of each); 3) OSB panels, 11-mm thick; 4) OSB panels, 15-mm thick; and 5) OSB panels, 18-mm thick. All OSB panels were aspen, grade O2.

2.2 Measurement of Horizontal Density Profiles Using VSX9811 Scanner

To measure the horizontal density distributions, all panels were scanned using the X ray density scanning system VSX9811 at Alberta Research Council. Figure 1 shows the system and a panel being scanned. This system generates analytic X-ray beam of 60 KeV and measures the attenuations of the X-ray transmitted through the panels. In each scan of the 1.22 by 2.44 m panel, the system divides the panel area into 224x450 pixels and measures the X-ray attenuation at each pixel. The X-ray attenuation at each pixel is calibrated to the average panel density at the pixel area with the scan resolution of approximately 5 mm. By adding the densities of all the pixel areas, a complete image of the horizontal density distribution in the plane of the panel is produced, while the vertical density variation is averaged.

The density distribution of the panel is further processed to produce a colored contour image with a final resolution of 12.5 mm. The different colors represent the density variation in the plane of the full-size panel. Each color represents a range of densities of 50 kg/m3. Figures 2 shows typical horizontal density distribution images with a 25-mm square mesh for OSB, MDF, and PB panels, respectively.


Figure 1. VSX 9811 X-ray density scanning system at Alberta Research Council.

2.3 Measurement of Vertical Density Profiles Using QMS Scanner

After the horizontal density profiles were obtained, small 50-mm square specimens were cut from the panels to determine the vertical density profiles across panel thickness. At least twenty-four 50-mm specimens were cut from each panel covering the entire range of horizontal density distribution within the panel. The vertical density profiles were measured using an X-ray QMS Density Profiler QDP-01X (Figure 3), at a scan speed of 0.6 mm/sec. The principle of this machine is similar to the VSX9811 X-ray scanner, except that it is designed for small-size specimens. The scanning resolution is approximately 0.06 mm across-thickness. During scanning, the X-ray beam, parallel to the plane of the panel, passes across the thickness of the specimen. This technique averages the in-plane density of the panel. Figure 4 shows examples of vertical density profiles of the scanned panel specimens.

Figure 2.
Typical images of horizontal density variation in panels: a) OSB; b) MDF; and c) PB.


Figure 3. Schematic a cross-section of QDP-01X density profiler.


Figure 4. Vertical density profiles of specimens scanned by X-ray QMS Density Profiler.

2.4 Control Gravimetric Measurements

After scanning the full-size panels and the small size panel samples using the two commercial X-ray scanning systems, gravimetric measurements were performed on the small 50-mm and 25-mm square specimens to determine the average densities. The gravimetric densities were used as a benchmark to evaluate the accuracy of the two X-ray systems. The density of each sample was determined by measuring mass and volume in accordance with ASTM D1037 [3]. The three dimensions of each specimen were measured with calipers to the accuracy of 0.1 mm, and the mass was measured on a balance to the accuracy of 0.001g. This procedure averages the density variations in-plane and across the thickness of the specimen.

3 Results

3.1 Horizontal Density Profile

The average density of each 50 mm square specimen taken from the horizontal (in-plane) density profile in this area as measured by VSX-9811 X-ray system was compared with the gravimetric data. Results of linear regression analysis are shown in Figure 5. The analysis revealed strong correlations between X-ray horizontal and gravimetric densities for the OSB and PB panels. The coefficients of determination (R2) were between 0.81 and 0.88 for OSB panels and 0.96 for PB panels. This indicates that X-ray scanning system is fairly reliable and accurate for these products. Much weaker correlation was found for MDF panels, with R2 = 0.42, due to the narrow range of the density variation. To achieve higher correlation, the scanning resolution and sensitivity of the X-ray system should be improved.

To examine the effect of the resolution on the relationships, three hundred small 25-mm square specimens were cut from the OSB panels and their gravimetric densities were measured. Figure 6 shows the relationships between gravimetric and X-ray in-plane densities for these samples. The R2 values significantly decreased when the specimen size was reduced from 50 mm to 25 mm. This indicates that the current scanning and image resolution employed by the X-ray scanning system was too coarse to accurately measure the density variation in the small area for those wood composites that exhibit high variations in their in-plane density profiles.


Figure 5. In-plane density vs. gravimetric density (50-mm samples).


Figure 6. In-plane density vs. gravimetric density of OSB panels (25-mm samples).

3.2 Vertical Density Profile

Figure 7 shows strong relationships of vertical density profile to the gravimetric measurements for all panels. The coefficients of determination were from 0.79 to 0.82 for OSB, 0.99 for PB, and 0.92 for MDF panels. This indicates that the QMS X-ray scanning system is accurate for all types and thicknesses of products tested. Compared to the VSX9811 X-ray system, it produced similar or better results for OSB and PB panels, and certainly, much better correlations for MDF panels. The better performance of the QMS system on MDF panels demonstrated a positive correlation between accuracy and scanning resolution. The QMS Density Profiler employs higher scanning resolution (0.06 mm) than the VSX9811 system (5 mm). However, this system can only scan small size samples as opposed to the VSX9811 system which scans the full size panels.

4 Conclusion and Recommendation

It can be concluded from this limited study that X-ray scanning technique has great potential for non-destructive evaluation of density and density distributions for a broad range of wood composite panels. More research efforts are needed to improve the accuracy of current X-ray scanning systems by using proper scanning and image resolutions, scanning sensitivity, and by improving the calibration procedure for various wood composite panels. Increasing the scanning resolution to achieve reliable measurements of density while still maintaining the cost-effectiveness and high-speed of the X-ray scanning process is challenging for the wood composite panel industry.


Figure 7. Through-thickness density vs. gravimetric density (50-mm samples).

5 Acknowledgements

The authors would like to thank the Canadian Forest Service, National Resources Canada, and the National Science and Engineering Research Council (NSERC) of Canada for their financial support. Thanks are also extended to Alberta Research Council (ARC) for their technical assistance.

6 Literature

  1. Suchsland, O. and X. Xu. 1989. A simulation of the horizontal density distribution in a strandboard. Forest Products Journal 39(5): 29-33.
  2. Chen, L., S. Chen and R. Wellwood. 2002. Horizontal density variation of oriented strandboard by radiation transmission image analysis. Proceedings of the 13th International Symposium on Nondestructive Testing of Wood, University of California, California, USA.
  3. American Society for Testing and Materials. 2003. D1037. Standard test methods for evaluating properties of wood-base fiber and particle panel materials. Annual Book of ASTM Standards. West Conshohocken, PA.

© NDT.net |Top|