Precise, cost & eco-efficient timber quality assessment for the sawmill of the future – EffiQuAss

The usage of timber in building construction will increase in the future so Europe can meet the objectives of CO2 emissions. Structural design of buildings requires the knowledge of the material that constitutes the structure. On the opposite of concrete and steel, timber mechanical properties are highly variable because of tree growth conditions and the natural heterogeneity of wood. This is the reason why quality assessment is crucial for timber. The main technology currently used to assess timber quality is dynamic vibratory excitation. However, quality assessment of timber material with this technique is unfavourable to heterogeneous species, which prevail in Southern Europe. Indeed, it fails to grade correctly about 30 % of French Douglas-fir, which is very frustrating for sawmill owners, not to mention it implies a waste of forest resources and increases the cost of timber material. Thus, it is of first importance to improve strength grading precision of the heterogeneous species, which prevail in France, for the development of eco-efficient sawmilling and timber construction building.
In recent scientific progress, laser dot scanning has shown interesting results to obtain a better quality assessment of heterogeneous species, and it also shows a high potential, which this project aims to develop. Laser dot scanning consists in projecting a row of laser dots at the four sides of a board which is conveyed longitudinally in a scanner. Laser light scatters at wood surface better in fibre direction than across, resulting in an ellipse shape which can be analyzed to measure the fibres orientation at wood surface, which is actually the fibres direction in a 3D space projected at the considered board surface and relative to board longitudinal direction. The current scientific progress uses basic interpolation methods to compute a fibres angle inside the board from these surface measurements. Then, mechanical modelling is performed based on the fact that wood mechanical properties changes according to fibres angle.
One novelty which has appeared very recently in the industry and still needs to be developed is the possibility to trace boards all along the sawmill process. A recent ANR project, TreeTrace (ANR-17-CE10-0016-03), aimed to participate to the development of log traceability from the forest to the sawmill. It has appeared that with some more research, it could also be possible to trace the boards in the following sawing steps, thus recognizing from which log and from where inside the log each board come from. By mixing the image processing knowledge of the Loria and the sawmill process knowledge of the LaBoMaP, this challenge should be met in the present project. This work will be based on TreeTrace dataset, each of the 346 board ends and each of the 32 log end images being available, with boards position inside the log known by manual labour.
The knowledge of board position inside the log is at the start of the main idea developed in this research project. Indeed, reconstituting the log from the sawn boards will allow to obtain a 3D map of fibres orientations measured at board surfaces which can be used to finally obtain the full 3d representation inside the log. This will be done in the present project by using a hydrodynamic-like model recently developed in the literature.
The results of this project should improve at mid-terme the efficiency of timber quality assessment, which is of first importance for the development of eco-efficient sawmilling and optimal usage of forest resources. They also have a very high scientific interest, modelling fibres angle inside a log being a total novelty.