Abstract
A reaction calorimeter, CRC, was taken as an example by which to demonstrate the systematic method for a parameter optimization of the time constant correction when heat transfer was not negligible. An essential role of the time constant was also examined with a simple mathematical model. As a result, four conclusions were obtained; a) With CRC the optimum reference interval exists when the square error between the correction data and the calibration signal shows the minimum. b) The deviation does not always decrease when the square error decreases by increasing the reference points with CRC. c) It is possible to underestimate the reaction rate when a long reference interval is taken. d) Noise enhancement occurs with a short reference interval. Therefore, when correction by the time constant parameter is carried out, it is necessary to specify the reference interval and the sampling rate.
