Holger Eisenschmidt

Model-based Observation and Design of Crystal Shapes via Controlled Growth-Dissolution Cycles

Abstract Control of multivariate population balance systems is a challenging task in process engineering, which is in particular due to difficulties in observing the state of the dispersed phase. In this article, we present a setup based on video microscopy and Kalman filtering to observe the evolution of a crystal population in batch crystallization processes. We demonstrate the applicability of our methods on the cyclic batch crystallization of potassium dihydrogen phosphate. It is shown that the cyclic process strategy can lead to crystal shapes which would not have been attainable via pure growth processes.

Estimation of aggregation kernels based on Laurent polynomial approximation

Abstract The dynamics of particulate processes can be described by population balance equations which are governed by the phenomena of growth, nucleation, aggregation and breakage. Estimating the kinetics of the latter phenomena is a major challenge particularly for particle aggregation because first principle models are rarely available and the kernel estimation from measured population density data constitutes an ill-conditioned problem. In this work we demonstrate the estimation of aggregation kernels from experimental data using an inverse problem approach. This approach is based on the approximation of the aggregation kernel by use of Laurent polynomials. We show that the aggregation kernel can be well estimated from in silico data and that the estimation results are robust against substantial measurement noise. The method is demonstrated for three different aggregation kernels. Good agreement between true and estimated kernels was found in all investigated cases.