Effect of sensor locations on the solution of inverse Stefan problems

Abstract

The aim of this study is to investigate the effect of
sensor location on the solution of inverse Stefan problems.
A unidirectional conduction driven solidification process is
considered. The enthalpy formulation along with conjugate
gradient method is used to simulate the direct problem and
minimize the objective function. The sum of square deviation
between the measured and the calculated temperatures
at sensor location is defined as objective function. Measured
temperatures are simulated using direct solver for triangular and step shape boundary heat fluxes. Differentsensor locations in the spatial extent of the computational
domain are selected. The results show that as the sensor is
taken further from the active boundary (the boundary
which heat flux applied on it) the error in reconstructed
heat flux becomes larger and vice versa. Also the effect of
noisy input data is investigated which indicate that the solution
is stable even in high noise levels in measured data.