Thesis for the Degree of Licentiate of Technology


Year 1999
Jukka Mäkinen

Coat weight modelling and control of a new coating station

In paper coating the trends are towards new coated paper grades and faster coating machines. New coating methods are continuously being developed, one promising new coating technique being non-contact coating. In this thesis a comprehensive study on modelling and controlling a non-contact coating process is presented. The study has three objectives: 1) development of a model and simulator for the non-contact coating process, 2) evaluation of the optimal nozzle layout for the non-contact coating station by means of the simulator and the Powell optimisation algorithm, and 3) development of the control algorithms for the machine direction and cross direction.

The model for the non-contact coating nozzle was developed using the correlations found from the literature and the results of the non-contact nozzle studies carried out at the Valmet Järvenpää Research Center. The simulator for the non-contact coating process was developed utilising the model developed for the nozzle. The non-contact coating process was studied in the trials made with the pilot coater, and the data obtained from the trial were used for the model validation.

The simulator with the Powell optimisation algorithm was utilised in the layout optimisation of the non-contact coating nozzles for the coating station. The nozzle layout was optimised with respect to the profile 2s value by changing the nozzle rotation angle, the nozzle distance and the nozzle location. The optimisation was made separately when the station had one row, and when the station had several nozzle rows. In the one-row case, the overlap ratio of the fans was the most important factor affecting the profile. In the multi-row case, additional rows had to be located to compensate for the peaks and valleys in the profile produced by the other rows.

Development of the coat weight control algorithms for both the machine direction and the cross direction was carried out using the simulator. Five algorithms were analysed for the MD control and six algorithms for the CD control. A multiple control strategy for both MD and CD controls was developed as a result of the simulation studies. Multiple control strategies of three different SISO control algorithms were utilised in both cases.

This info last modified 20 Jun 2018 - contact webmaster