Mobile Edition
Print
Get the Mag
Weekly UpdatesMinitab[R] software provides multiple response optimizers to help identify the combination of factor settings that jointly optimize a set of responses (Minitab Inc. 2005). Numerical and graphical optimizations were used to optimize the data from responses. In this optimization, goals were targeted at maximum values for surface hardness, MOE, and nail withdrawal resistance. During optimization, the weight and degrees of importance of all responses were held equal. A value of weight equal to 1 was used. Weight value can be set from 0.1 to 10. A weight of 0.1 puts less emphasis on the target, thus a response value far from the target may have a high desirability. On the other hand, a weight of 10 puts more emphasis on the target; therefore a response value must be very close to the target to have a high desirability. One can vary the value of degree of importance to different responses during optimization if one response is more important than the other. In this study, the degree of importance for all responses was also set to 1.
Desirability score was used to rank the tradeoffs to get the best possible goal for all responses.
The optimum conditions were verified by densifying another pregrouped set of aspen specimens. The same testing procedures aforementioned were used.
Results and discussion
After conditioning, the average MC was 9.3 percent, ranging from 8.8 percent to 9.9 percent. The average SG of surface densified aspen was 0.55, ranging from 0.43 to 0.65. Table 2 gives a summary of mean values of the selected mechanical properties tested in this study. The mean values of hardness, MOE and nail withdrawal resistance of the control (not densified) specimens were also given in Table 2.
Effects of compression ratio, press temperature and press closing time
The effects of compression ratio, press temperature, and press closing time on surface hardness, MOE, and nail withdrawal resistance were investigated. Main effect, factors interaction, and response surface plots were generated and analyzed. To illustrate the main effect and interaction plots, results for hardness were chosen since it is one of the most important indices for flooring application. Figures 1 to 3 illustrate the effects of the 3 factors chosen on each response. ANOVA indicated that compression ratio was the most significant factor affecting hardness (p-value < 0.001), MOE (p-value = 0.021) and nail withdrawal resistance (p-value = 0.005). The significant factors were analyzed at a significance level of 5 percent. Although press temperature was thought to influence the mechanical properties of densified aspen, it has no significant effect within the range selected in this study.
Figure 1 illustrates that increasing compression ratio from 9.5 to 24.0 percent resulted in an increase in hardness. There seemed to be a slight decrease in hardness if press temperature was increased from 145 to 175 [degrees]C. However, hardness increased if press temperature was increased from 175 to 205 [degrees]C. No further investigation was done for this trend since ANOVA indicated that press temperature had no significant effect on hardness within the range selected in this study. Press closing time also had no significant effect on hardness. Compression ratio (p-value < 0.001) was the most significant factor affecting hardness.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Figure 2 shows that hardness changes at different levels of each factor; however, no significant interaction effect was observed among the factors chosen for this study. Statistical linear models were determined based on the statistical analysis. The linear models were then fitted and surface plots were generated. Surface plots (Fig. 3) indicated that there was an increase in hardness, MOE, and nail withdrawal strength from the control group to that of the compressed groups regardless of what level of compression ratio, press temperature, and press closing time was used. Hardness (Fig. 3a), MOE (Fig. 3b), and nail withdrawal resistance (Fig. 3c) increased with the increase of compression ratio at all levels of press temperature (Fig. 3a and Figure 3c) and press closing time (Fig. 3b).
Process optimization
Minitab[R] software provides response optimization for experiments. Numerical and graphical optimizations were used to optimize the data from the responses. The goals for optimization of responses (surface hardness, MOE and nail withdrawal resistance) were targeted at maximum values and should not be less than the control group. During optimization, the weight and degrees of importance of chosen responses were held equal. A value of 1 was used for both weight and degree of importance. A weight equal to 1 places equal emphasis on the target and the boundary.
Desirability score was used to rank the tradeoffs to get the best possible goal for all responses. It was found that compression ratio = 24 percent, temperature = 145 [degrees]C, and closing time = 7 minutes, were the optimum conditions with a desirability value of 0.67. Figure 4 illustrates the results of process optimization. At these optimum conditions, the predicted response values for hardness, MOE, and nail withdrawal resistance were 4.8 KN, 11,630 MPa, and 0.58 KN, respectively. Individual desirability score for the predicted values of hardness, MOE, and nail withdrawal resistance were 0.65, 0.65, and 0.71, respectively. The closer the desirability score to 1, the closer the predicted response is to the set target for that particular response.
Improvement of selected mechanical properties
Results showed that there was an increase in selected mechanical properties of surface densified aspen. Figure 5 shows a comparison between the mean values of the control group, mean values of densified specimens (after verification) and predicted values (values from Minitab[R]). There was an increase of 140, 23 and 132 percent for hardness, MOE, and nail withdrawal resistance, respectively. The mean hardness of surface-densified aspen for verification was 6.0 KN which was almost equal to that of uncompressed red maple (mean = 6.1 KN) tested in this study. This indicates that the surface-densification technique proposed in this study could help the wood products manufacturers to use low- density and underutilized wood species for value-added applications such as flooring.
It would be reasonable to predict that mechanical properties of surface-densified aspen can be further increased by increasing the compression ratio. However, compression ratio should be carefully controlled below the maximum value in order to avoid any permanent cellular damage in wood. Preliminary tests showed, for example, that small visual cracks appeared in the small air-dried clear aspen specimens pressed at a localized compression ratio of 50 percent at the room temperature.
[FIGURE 5 OMITTED]
Future work
A thorough investigation of the changes in the vertical density profiles of the surface-densified wood is on-going using an x-ray scanner. It is also anticipated that mechanical densification will affect physical properties, e.g., color change, thickness swelling, water absorption, and linear expansion, which could be the focus of future work. Likewise, the effects of high-temperature heat treatment and other technologies to stabilize the dimensional property of surface densified wood will be examined.
Conclusions
The influence of compression ratio, press temperature and press closing time on the selected mechanical properties of surface densified aspen were evaluated in this study. The main findings can be summarized as follows:
1. Compression ratio was the most significant factor influencing surface hardness, MOE, and nail withdrawal resistance.
2. The optimum pressing conditions were found to be compression ratio = 24 percent, temperature = 145 [degrees]C, and closing time = 7 minutes.
3. There was an increase of 140, 23, and 132 percent in hardness, MOE, and nail withdrawal resistance, respectively, of surface-densified aspen after mechanical densification.
Literature cited
American Soc. for Testing and Materials (ASTM). 2004. Standard Test Methods for Small Clear Specimen of Timber. ASTM D 143-94. Vol. 04.10. ASTM Inter., West Conshohocken, Pennsylvania.
Ellis, S. and P. Steiner. 2002. The behavior of five wood species in compression. IAWA J. 23(2):201-211.
Gong, M., M. Nakatani, Y. Yang, and M.T. Afzal. 2006. Maximum compression ratios of softwoods produced in eastern Canada. In: Proc. of 9th World Conf. on Timber Engineering, Portland, Oregon. pp. 330.
Inoue, M., M. Norimoto, Y. Otsuka, and T. Yamada. 1990. Surface compression of coniferous wood lumber. 1: A new technique to compress the surface layer. Mokuzai Gakkaishi 36(11):969-975.
Ito, Y., M. Tanahashi, M. Shigematsu, Y. Shinoda, and C. Ohta. 1998. Compressive molding of wood by high-pressure steam-treatment. Part 1. Development of compressively molded squares from thinnings. Holzforschung 52:211-216.
Minitab Inc. 2005. Minitab[R] Statistical Software (Release 14). State College, Pennsylvania, Pennsylvania 16801-3008.
Norimoto, M. 1993. Large compressive deformation in wood. Mokuzai Gakkaishi 39(8):867-874. (in Japanese).
Uhmeier, A., T. Morooka, and M. Norimoto. 1998. Influence of thermal softening and degradation on the radial compression behavior of wet spruce. Holzforschung 52:71-81.
Wan, H. 2004. Wood Hardening technologies. National Res. Project No. 3678, Forintek Canada Corporation, Vancouver, B.C., Canada.
Wang, J. and P.A. Cooper. 2005. Vertical density profiles in thermally compressed balsam fir wood. Forest Prod. J. 55(5):65-68.
Clevan Lamason *
Meng Gong *
The authors are, respectively, Research Scientist and Research Scientist/Adjunct Professor, Wood Science and Technology Centre, Univ. of New Brunswick, Fredericton, New Brunswick, Canada E3C 2G6 (clamason@unb.cai; mgong@unb.ca). The authors would like to acknowledge Natural Resources Canada for its support and funding under its Value to Wood Program. This paper was received for publication in February 2007. Article No. 10320.
FEED