Numerical and experimental characterization of dust emission profiles for hand held wood working machines

There is a clear need of characterizing and evaluating wood dust emission profiles from wood working machines in order to predict potential exposure peaks and to ensure the safety of workers. Indeed, wood dust is classified as a well known carcinogen. Moreover, in the occupational population of the 25 member states of the European Union, around 3.6 million workers were exposed to inhalable wood dust. In order to limit the exposure of workers to excessive dust, the European directive 98/24/CE has fixed the Occupational Exposure Level (OEL) to 5mg.m-3. This directive has been transcribed into French law in 2003 declaring that the OEL would be lowered down to 1mg.m-3. Among the common wood dust emission sources, the most emissive are by far hand-held electrical wood working tools, such as sanders, routers and circular saw. In the majority of cases, the dust collecting systems are not dimensioned properly; therefore, the wood dust is not well collected. Accordingly, we are currently developing a method that should allow characterizing the emission profiles from wood working machines, so as to classify the machines with respect to their emission potential.
The evaluation procedure of wood dust emission is based on the use of a mathematical algorithm called "inversion algorithm" which allows reconstructing the emission rate by means of measurements taken at different working environment points. This procedure includes two distinct steps. The first phase consists in determining the parameters of the inversion model by using a known source of dust and corresponding concentration measurements. In the second phase, the unknown source is reconstructed by inversing the model, with corresponding concentration measurement.
After presenting the source-estimation method, this paper discusses two important approaches related to its development. The first one deals with the automatic determination of optimal positions for the sensors, based on the calculation of a term called the "optimization criterion", which allows classifying the sets of 'n' among 'N' sensors (n < N). When the best set of sensor positions is selected, an experimental approach is initiated, aiming at validating the source-estimation method, and its robustness.