Physical and mechanical properties of heat-treated bamboo using near infrared reflectance spectroscopy
MO Junqian;ZHANG Wenbo;College of Material Science and Technology,Beijing Forestry University;
It is of great importance for the bamboo processing industry to evaluate the quality of heat-treated bamboo using the low-cost,rapid and non-destructive method. We investigated the potential of evaluating physical and mechanical properties(material color,density and mechanical properties) of the heat-treated bamboo by the near infrared reflectance spectroscopy(NIRs). Total 90 samples were averagely sorted into 3 groups by different heat treatment temperatures(150 ℃,180 ℃ and 210 ℃) with a treatment time of 2 h,and another control group had 30 samples without treatment. The NIR information was collected before and after the heat treatment. The chemical compositions of bamboo were analyzed using the wet chemical analysis method. The differences of the NIR spectra of the bamboo samples were compared through their second derivative spectra. Then,the NIRs combined with the principal component analysis(PCA) and partial least square(PLS) regression were used to sort and model the samples,respectively.The spectra analysis results showed that the absorption bands at 7 004 cm-1 and 6 452 cm-1 reflected thermal decomposition of cellulose,and the absorption band at 5 801 cm-1,4 289 cm-1 and at 6 896 cm-1,5 959 cm-1 reflected pyrolysis of hemicellulose and lignin,respectively. The results revealed that there was a well correlation between chemical components and the NIRs. The PCA results showed that the heat-treated bamboo originated from different heat temperatures could be distinguished obviously in the score plots. The prediction models for color(L*,a*,b*) of heattreated bamboo were very satisfactory,in which R2 and relative percent deviation(RPD) were over 0.93 and 3.90,respectively. The prediction models of heat-treated bamboo density also well met the requirements,in which R2 and RPD were over 0.83 and 2.42 for the density of air-dried samples,as well as 0.85 and 2.59 for the density of oven-dried samples,respectively. As for mechanical property prediction models,the R2 and RPDof modulus of rupture model were 0.82 and 2.34,while for the modulus of elastic(MOE) model,there was relative lowprediction performance,in which,R2 and RPDwere only 0.55 and 1.49 in this study. All models(except the MOE model) showed a good predictive results for the heat-treated bamboo. The NIRs method has great potential not only to discriminate the heat-trea-ted bamboo derived from different heat treatment temperatures,but also to predict physical and mechanical properties of heat-treated bamboo.
【CateGory Index】： S781.9
【CateGory Index】： S781.9