Predicting the strength of sawn wood products: a comparison between x-ray scanning of logs and machine strength grading of lumber
2005 (English)In: Forest products journal, ISSN 0015-7473, Vol. 55, no 9, 55-60 p.Article in journal (Refereed) Published
For a long time, process improvements in sawmills have focused on high productivity measured in volume per hour and volume recovery. As sawmills become increasingly efficient, the importance of focusing on value recovery becomes obvious. In order to maximize value recovery, sawmills need the ability to sort logs according to properties such as strength. The aim of this study was to compare the results of predicting the strength of center boards based on x-ray scanning of logs with the results obtained by machine strength grading using a bending machine. The study was based on 131 Norway spruce (Picea abies (L.) Karst.) sawlogs that were scanned with an x-ray LogScanner and then sawn into boards. The bending stiffness of the center boards was tested using a strength-grading machine, and the bending strength was tested according to EN 408. Models for prediction of bending strength based on machine strength grading and x-ray LogScanner measurements were calibrated using partial least squares regression. The study showed that the x-ray LogScanner (r2 = 0.44) and machine strength grading (r2 = 0.43) had similar accuracy in predicting bending strength. The combination of both methods resulted in significantly higher accuracy (r2 = 0.56). The root mean square error (RMSE) was 8.5 MPa for the bending machine, 8.4 MPa for the x-ray LogScanner, and 7.4 MPa for the combination of both methods. Consequently, the combination is an interesting alternative. Future studies should include a larger number of boards and focus on the effect of log carriers and on finding the reasons behind outliers
Place, publisher, year, edition, pages
2005. Vol. 55, no 9, 55-60 p.
Research subject Wood Technology
IdentifiersURN: urn:nbn:se:ltu:diva-12849Local ID: bff24680-bb52-11db-b560-000ea68e967bOAI: oai:DiVA.org:ltu-12849DiVA: diva2:985800
Validerad; 2005; 20061017 (cira)2016-09-292016-09-29Bibliographically approved