University of Natural Resources and Life Sciences, Vienna (BOKU) - Research portal

Navigation/Search

• Researchers.
• Institutes.
• Projects.
• Publications.
• Partners.
• Sponsors.

Gewählte Doctoral Thesis:

Martin Riegler (2014): ANPASSUNG DES HERSTELLPROZESSES VON HOLZWERKSTOFFPLATTEN AUF BASIS STATISTISCHER MODELLE ZUR STEIGERUNG DER RESSOURCENEFFIZIENZ.
Doctoral Thesis - Institut für Holztechnologie und Nachwachsende Rohstoffe, BOKU-Universität für Bodenkultur, pp 183. UB BOKU obvsg

Abstract:

The efficient usage of resources is becoming more important for the industrial production of wood-based panels, due to the increasing shortage of resources and rising ecological awareness. In this respect, multivariate statistical process modelling can be used to address this situation. The three significant aspects identified in the course of this thesis were “data generation”, “data management” and “process adaptation”. For the point "data generation", different methods for detecting resin were evaluated, whereas the fluorescence microscopy delivered the best results due to the high contrast between wood and resin, the high resolution obtained, the relatively low costs and simple sample preparation. Furthermore, the compression ratio of low density hardwood species, the morphology of wood fibres in fibreboards and the fracture energy of fibreboards were investigated. Determined information from raw materials and process parameters had to be aligned in time to be fused in databases ("data management"). In particular, the dynamic calculation of time lags aligned data more precisely to their respective board properties than the static calculation. This was achieved by simulating the material flow within the manufacturing process, which considers short term variations of process conditions as well as retention times within silos. In the course of the final step of multivariate statistical "process adaptation", models with high predictabilities (compared to literature) could be gained. The margin of deviation of the internal bond strength of fibreboards could be minimised by altering adaptable predictor variables. This enables lower safety margins and stable processes. Besides the challenges of implementing the presented approach of real-time process adaptation in an industrial environment, it has the potential to use raw materials more efficiently and to optimise energy intense processes.

Betreuer: Müller Ulrich

1. Berater: Gindl-Altmutter Wolfgang

2. Berater: Spangl Bernhard