Official website of the New Zealand Farm Forestry Association


Measuring Strain in Wet Eucalyptus Wood by NIR

By Fei Guo and Clemens Altaner, June 2017.

Download SWP-T026 (pdf)

Executive summary

These experiments investigated whether band shifts in the NIR caused by strain can be observed in fully water saturated eucalyptus wood samples. Air-dry samples with minor changes in moisture content helped to understand the effect of strain on NIR spectra. Fully water saturated samples without tensile stress illustrated spectral changes due to drying. Whether band shifts of green wood caused by stress were detectable was revealed by NIR spectroscopy of fully water saturated samples under tensile loads.

CONCLUSION

  1. Abundant free water in fully water saturated eucalyptus wood samples resulted in strong water signals in NIR spectra. These signals were mainly in the regions between 5000 and 5300 cm-1 as well as 6100 and 7200 cm-1. The region between 5700 and 6080 cm-1 was less affected.
  2. When spectra were normalised between 6100 and 7200 cm-1, air-drying during spectra acquisition overrode the effect of mechanical strain in fully water saturated wood.
  3. Clear band shifts caused by tensile strain were observed in the 2nd derivative spectra for both, air-dry (r2 = 0.78) and fully water saturated (r2 = 0.75) samples.
  4. The rate of band shift for air-dry eucalyptus (0.94×10-3 cm-1/με) was similar to that of previously reported air-dry radiata pine (1.04×10-3 cm-1/με).
  5. Compared to the air-dry samples the band affected by strain was found at lower wave numbers in fully water saturated samples. This could be mainly an effect of baseline distortion due to a large neighbouring water signal.
  6. The band shift caused by tensile strain in the fully water saturated sample matched that of the air-dry sample when the observed band shift was corrected for the effect of moisture loss.

As a consequence of these observations it seems possible to measure growth-strain in stems by NIR spectroscopy. However, the moisture content affects the position of the signal of interest. Therefore, it seems more realistic to measure growth-strain in standing trees or freshly felled/debarked logs rather than in debarked logs in a sawmill, which might have lost some moisture. Alternatively a spectra processing method capable of removing the influence of moisture of a measurement technique which is less affected by moisture could improve this technique.

(top)


No posts yet

Farm Forestry - Headlines