26 May 1997 Computers unravel knotty problem

Through work at Lincoln University the carpenter's kit of the future may well include a computer for predicting the stresses and strains timber can cope with before it is hammered into the studs, beams and other supporting structures of your home.

Every builder knows that timber contains natural and artificial defects of all sorts – knots, drying cracks, splits and machined notches and holes – all of which can dramatically affect its load-bearing ability.

While various types of wood fracture have been studied fairly extensively there is no single, standardised approach to integrating wood fracture knowledge into wood designs procedures. Staff of the Department of Natural Resources Engineering at Lincoln University have, however, made a start.

Most knot-related cracks develop in the initial drying or ‘seasoning’ stage of timber production and result in a huge loss to the industry.

To investigate this drying deformation, Lincoln scientists have used computers to draw up ‘models’ which could be used in the future to estimate the drying deformation of knotty boards.

Applying the techniques of Digital Image Correlation to wood fracture studies, the Lincoln University team of Dr Sandhya Samarasinghe (Department of Natural Resources Engineering), Dr Don Kulasiri, (Centre for Computing and Biometrics) and Kelvin Nicole (Department of Natural Resources Engineering), have investigated whether DIC offers a useful technique for estimating the parameters governing crack propagation and ultimately the strength of cracked timber.

Using the combination of a video camera sending timber specimen images to a computer for digitisation and on-sending to another computer for analysis to determine displacement and strain fields in the vicinity of the crack tips, and testing the timber with cracks to failure point, the researchers have been able to construct plotted images depicting the magnitude and extent of stress concentrations near the tip of the cracks.

Testing various crack geometrics and orientations in order to predict the propagation of cracks parallel, perpendicular and at an angle to the grain, the researchers have shown that regardless of crack length, crack propagation is initiated at similar critical levels of stress concentration and that longer cracks reach this critical level more quickly than short ones.

They have also shown that crack propagation always follows a failure path characterised by a radial-longitudinal plane indicating that this is the weakest plane for load resistance.

These findings allow the Lincoln scientists to look ahead confidently to the common use of techniques for predicting the strength of timber containing defects.

Ian Collins, Journalist, Lincoln University, Canterbury, New Zealand