Thermal parameters of a work-roll play an important role in the modeling of the rolling process, due to periodic thermal loading. The knowledge of thermal parameters is also vital in understanding the fatigue life of the work-roll and the thermal crown. However, estimation of the thermal parameters viz., thermal conductivity, thermal diffusivity and convective heat transfer coefficients at both, inner and outer roll periphery is tough to realize during the rolling process. Various methods employed earlier for measuring the thermal properties of work-rolls in the rolling process requires intrusion in the surface of the work-rolls, mainly to embed the thermocouples inside the rolls. These methods are easy to implement, but it is really hard to achieve truthful estimation. A possible way out is to measure the average thermal parameters of a work roll in the rolling process by utilizing the measured temperature at two specified locations on the work-roll surface. In this work, an inverse method is proposed to estimate the thermal properties and convective heat transfer coefficients of a roll in the rolling process. The inverse method makes use of a direct model of temperature determination considering plane strain problem, which is based on the integral transform method. For minimizing the error between the computed and experimentally recorded data, a quasi-Newton method is used. In lieu of shop floor experiments, a finite element method (FEM) based package ABAQUS 6.10 is used to obtain the temperature distribution in the work-roll. Further, an additive white Gaussian error is added in the FEM simulated measurements to assess the inverse method for stability towards mild measurements. The inverse estimation is successfully validated and can be used in shop floor for the online determination of thermal parameters of the work-rolls in the rolling process.
The vector optimization and modeling of a technical systems
