Mechanical properties of timber
Looking for Good Wood
Trees are amongst the most variable of all living organisms. When trees of the same age and from the same stand are felled and the logs are sorted into stacks according to diameter, length and visible features, eg size and distribution of branches along the stem, there will still be an enormous range in the intrinsic (invisible) properties for those logs within a particular stack. For example the stiffness of visually "identical" logs varies by a factor of three. In principle one could measure stiffness by grasping the log and physically bending it, but this approach is not practical.
The long-term profitability of the solid wood processing industry depends on having tools to identify the variations in intrinsic quality of logs. Furthermore, the three "essential requirements" - stiffness, straightness and stability - are surprisingly difficult to achieve! They are most problematic in the corewood zone, which is the 10 or so growth rings adjacent to the pith. In a 25-year-old tree 50% of the merchantable stem volume is corewood.
Much of this work is supported by the Public Good Science Fund. It is multi-disciplinary in its emphasis, involving people in other departments and other institutions.
Good wood: variations between trees
At the fundamental level, the stiffness is determined by the microfibril angle, the angle at which the cellulose molecules coil round the fibre cell walls. By analogy, just as a lazy spring can be stretched easily while fencing wire is very stiff and inextensible, so the stiffness of wood increases five-fold as the microfibril angle decreases from 400 (a lazy spring) to 100 (approximating to a taut fencing wire). Our studies of the effects of microfibril angle on wood quality are supported by the x-ray diffractometry team in the Department of Chemistry (Professor Ward Robinson and Drs Ian Cave and Jan Wikaira). This facility allows us to benchmark other research to fundamental wood characteristics.
It is not possible to measure the microfibril angle in logs the skid site. An x-ray diffractometer is a $1M+ piece of laboratory equipment, so an alternative or surrogate measure is needed. This is provided by acoustics.
At the School of Forestry we have been measuring acoustic properties of wood for some 5 years. Acoustics offers a quick, reliable, field- or mill-based system for measuring stiffness directly.
Stiffness is determined from the equation:
MOEdynamic = rV2, where r is the density and V is the velocity of sound. This equation separates the contributions of density, which largely reflect quantity (mass) of wood in the tree, from that of wood quality on which the velocity of sound depends (cell diameter and wall thickness, the relative proportions of the various cell wall layers, the microfibril angle in the S2 layer, chemical composition etc.).
We are validating such technology in an on-going study of 27-yr old trees from an unpruned stand on the Mamaku Plateau. We cut these trees to give four logs and the acoustic velocity was measured along every log. The plot of the (velocity of sound)2 for all logs from every tree accurately reflects the intrinsic stiffness of the logs. There are two striking features: there is little difference in log stiffness between the butt and upper logs, and the range of values amongst each log type is huge. This justifies the earlier statement that log stiffnesses can vary by a factor of three.
The square of the velocity of sound, V2, for all logs from eighty-two trees in an unpruned stand on the Mamaku Plateau.
This is an on-going project.
Good wood: variations within trees
Knowing the grade out-turn from the knotty corewood of fast-grown radiata pine is vital for sawmills to operate more profitably. Currently the millers do not know which logs in a batch of logs can be sawn profitably into structural lumber and which should be sawn for dunnage. Work by doctoral student Ping Xu is delineating the extent of the problem, by examining wood properties within logs. In turn this will permit better use of acoustics in log sorting (discussed above). By machine stress grading the boards and "reassembling the logs" she is able to display the stiffness gradients within a log (from the pith to the cambium). In the first instance the variations in stiffness are being characterised. The diagram below emphasises the very poor stiffness of the wood surrounding the pith (in position P1).
The most interesting new observation is that the bottom of the butt log has far poorer properties to the rest of the log: yet sawmillers traditionally consider it to be the best part of the tree.
This data is of the average properties of all logs. Consideration of the properties of individual trees is critical as it will allow us to develop more detailed strategies for segregating logs using acoustics. Consider the proposition that 90% of all the problems in manufacturing good product can be attributed to 10% of the logs being milled. The unidentified poorest logs will be processed at a loss as identification of poor quality is perceived only subsequently. Boards will be dried, only to warp unacceptably; lumber will be dressed with a rough, woolly or chipped surface; structural lumber will be graded only to find that it fails to meet specifications. This low quality tail in a population is a feature of all processing whether machining screws, mixing concrete, or casting metal. The reason why it is so important to the timber industry is that trees are amongst the most variable of all living organisms. The range of strength in a batch of steel is insignificant compared to the range of property values found in lumber. This is why many see the future of the industry in engineered wood panels where wood may be chipped or fiberised before being blended and glued together again. The average property values may not change much but the range - the variability - is greatly reduced.
Stiffness variations within logs. P1 are the 90x35 mm boards which include the pith, P2 boards are adjacent to the P1 boards being slightly removed from the pith, etc. Only the three locations closest to the pith (P1-P3) are shown here.
Good wood in seedlings and clonal plantlets
The University of Canterbury has a multiclient programme involving forest companies in New Zealand and overseas. Our interest is in identifying material which is intrinsically stiffer than average. Dr Hakan Lindstrom, a post-doctoral research fellow from Sweden, is resident at the School for two or three years and is the principal investigator. In addition Brian Butterfield (Plant & Microbial Sciences), Ryogo Nakada (Forest Tree Breeding Centre of the Japanese Forest Agency at Ibaraki) and John Walker are involved in this programme. Hakan has been developing methods to measure the stiffness of whole bolts of clearwood cut from between branch nodes.
Measuring wood properties of young trees arising from a tree breeding trial.
Currently, tree breeders are reluctant to select trees until their final properties are well established, often sampling trees when aged between 5 and 15 years. This programme focuses on trees that are less than 5 years old.
Studies of rootwood
This is a new initiative with Plant and Microbial Sciences (Drs Brian Butterfield and Sandra Jackson) and follows some pioneering work by Dr Junji Matsumura of Kyushu University while on sabbatical leave at Canterbury. Doctoral student Linda Ching-Yi Hsu is exploring rootwood ultrastructure (SEM, TEM, Confocal laser microscopy etc) with regard to genetic and environmental factors. The contrasts between rootwood and stemwood can be considerable. Attention is being given to the reasons for such contrasts, to increase our understanding of the wood quality and to see how properties might be manipulated.
Studies in wood drying and wood products engineering
Both timber drying and the manufacture of wood panels are undertaken at the postgraduate level in joint programme with staff in the Department of Chemical & Process Engineering (CAPE). Research facilities are located in both departments. Recently the University has establishment a panel laboratory and this area is considered a priority research field. Participating staff include Kelvin Chapman and John Walker (Forestry) and Pat Jordan, Roger Keey, Laurence Wetherley and Chris Williamson (CAPE) together with four postgraduates and research fellows.