It is shown in this paper that knowledge of the spatially measured temperatures inside the refining zone in TMP refiners is essential in describing non-linear dynamics of high consistency refining. By expanding the pulp and handsheet properties using piece-wise linear functions into the time domain, an auto-regressive model can be applied to verify that the tempera-ture and the consistency profiles, in combination with the motor load and production rate, are key input candidates when modeling changes in different pulp properties. The model accuracy is analyzed using process information captured at different time and operating conditions. It is also shown that it is more complex to estimate and validate the tensile index than the mean fiber length and Somerville shives content, especially close to refiner operating limits where a shift in the process gain may occur. This type of switched dynamics in tensile index estimation at a specific consistency is related to non-linear behaviors where the fiber pad distribution most likely undergoes a local collapse.
Funding Details: Norske Skog