Multiscale research has become an area of increasing interest and activity in chemical engineering, which is driven by the ever-growing market needs from the specific properties of products to the social and environmental constraints of industrial-scale processes. However, existing efforts have evolved from ideas and solutions that strongly reflect their original problem domains, which present a serious barrier to the application of methods to new areas. For that, there is a crucial need for general frameworks in addressing the basic concerns beyond research in specific methods and problems.

In this paper, we propose a general framework named “target-oriented multiscale systems engineering” to address the challenges arising from introducing multiscale research into chemical engineering for systematic improvements. In this framework, three major functional components are involved: (i) target component, (ii) modeling component, and (iii) utilization component.

The target block addresses the challenge of how to define the appropriate problem domains and variables associated with the desired targets, in terms of product quality and process efficiencies. In multiscale research, since the domains and variables could span various time and length scales, more ambitious targets could be obtained as compared with the traditional mono-scale approaches. On the other hand, it should avoid generating excessive domains often with unnecessary detail or inappropriate scales, due to concerns of modeling complexity. In this regard, a graphic method is introduced to state the targets with their corresponding variables and mechanisms in a 2-D length and time scale map. The target-related information will be then sent to the modeling component, where a unified multiscale modeling structure is introduced. It should help adopt appropriate modeling methods at different scales, as well as the multiscale integration strategies for model development.

As long as the quantified targets and multiscale characterization are provided, the utilization component is employed for system evaluation and further improvements. In this regard, model simplification will be discussed for utilizing multiscale information into time-constrained applications (e.g., optimization and control).

This framework provides a general, rigorous, and systematic approach for formulating and analyzing multiscale problems across a range of scientific and engineering disciplines. An automotive paint spray process is studied to illustrate the concepts of the framework, where model-based simulation reveals various usually inconceivable opportunities for improvement of product quality, energy and material efficiencies, and environmental performance.