Michal Pohanka finished his habilitation and became associate professor on 17th October. Michal Pohanka operates in the field of heat transfer. Mainly, he is specified in solution of inverse task of heat transfer in metallurgical processes.
His habilitation is called ‘Rozvoj inverzních metod vedení tepla a jejich aplikace’. This work focuses mainly on the inverse tasks of heat conduction and their use in the metallurgical industry such as applications for continuous heat treatment, roll cooling and descaling. Numerical models for design and control in the metallurgical industry require a precise description of the heat transfer on the surface of the cooled material. Comprehensive heat transfer information is not yet available for cooling of hot surfaces by water sprays or laminar jets. Therefore, the necessary boundary conditions must be obtained experimentally, which leads to ill‑posed inverse heat conduction problems. In the non-stationary direct heat conduction task used in simulations and in on-line control, temperature distribution as a function of time is calculated based on known boundary conditions, geometry, material properties, and known initial temperature distributions. In the inverse task that we are dealing with, the time dependent boundary conditions are calculated based on the measured temperature history at one or more points within the body. There are also three other types of inverse tasks: determination of material properties from measured temperature history, determination of body geometry and determination of initial temperatures. To solve these problems, two main approaches are described: sequential and whole‑domain. This works deals with both of these approaches and introduces new ones. New Sequential identification approach is described and it is shown that it is more universal than widely used Sequential Beck’s approach. Another new Sub‑domain method for linear models is also described. This new method uses benefits from both approaches: the speed from sequential approach and accuracy from whole‑domain approach. For determination of thermophysical properties of scale layer on steel new approach is described. This method is based on flash method and uses complex heat transfer and fluid flow model of measuring apparatus. The speed‑up approach using acceleration on GPU is described and results shows that the computational time of heat conduction can be significantly shorten. The usage of new approaches is presented in research projects.
Congratulations to associate professor Michal Pohanka.