Adaptive sawing : Yield of a concept in reality

Abstract: Glue board is the base for much solid wood furniture produced by the IKEA-owned company Swedwood. Glue board is glued together from lamellas which can be produced in different waysbut in all cases is produced from boards coming from a saw-mill. In a saw-mill there aredifferent techniques for producing boards from logs. This thesis is about the Adaptive sawingconcept and its implementation at the Swedwood site in Kostomuksha. The idea with the conceptis to raise the output from the raw material, the yield. This is partly achieved by edging away islittle as possible from the boards in the saw-mill. The boards in the saw-mill are sawn in a waynot unlike through-sawing. When brought to the glue-board factory they are scanned in order tooptimise how to rip out as much lamellas as possible. The ripped lamellas are then cross-cut intodifferent sizes which later are glued to glue board. The objective of this thesis is to examine theyield from log to glue board for the log classes 135-148 mm and 110-120 mm, to try to improveit and to identify problem factors. In order to achieve that it was necessary to be sure of the volume figures at all stages. Theincoming volume was given by the scanner at the log sorting; effort was therefore put down tomake sure the figures could be trusted. For the smaller log class the logs were instead measuredby hand and the volume was calculated. The logs were then sawn in the saw-mill. This was donewith a low production speed and some problems were noted compared with when sawing with astandard technique. The absence of edging was the main reason for these problems. The saw-millline was deemed inappropriate for sawing the smallest log class but could handle the 135 -148mm class and the yield for that class was in line with what could be expected. In the lamella production line some minor problems were noted and the boards from the fourdifferent batches from the 135 – 148 mm log class got different yields mainly due to differencesin wood quality. The yield from log to glue board for these four batches varied from 17.6 % to21.5 % with an average at 19.3 %. The yield was lower than what was achieved for earlieradaptive sawing test batches on other places. Those tests were however done on other diameterclasses and with material that was slightly different than the one in Kostomuksha. Simulationswere done to check how much the yield could be raised if other lamella widths than just thestandard 46 mm lamella width was used. A raise to an average of 20.9 % could then be expectedwhen using the lamella widths 46, 55 and 60 mm. A similar simulation was done for the adaptive material that was produced from the 110 – 120mm log class. The total yield for that material rose from 19.4 % to 22.4 % when using morelamella widths. Lamellas from this material were tested as there was fear, that they would notmeet the standards for glue board production. This did however not prove to be the case. To use the lamella production for producing such material is however not ideal, since a very smallvolume is produced.