A simple model-based approach for fluid dispensing analysis and control

In this paper, a simple model-based approach is presented for modeling and control of the fluid dispensing. A simple model structure is derived from the pipe flow, which has the capacity to handle both Newtonian and non-Newtonian fluids. When working for unknown dynamics, parameters of the model need to be estimated from the process data. Using a simple and effective approximation method, proper operating conditions for parameter estimation can be figured out. The feasibility of this simplified model is evaluated in comparison with computational fluid dynamics for both known and unknown dynamics. Both simulation and real experiment demonstrate simplicity and effectiveness of the proposed model and its estimation method for both the steady and unsteady fluid dispensing. Based on this simple and effective model, a realistic model-based run-by-run control can be developed effectively to achieve a robust dispensing performance. Both simulation and real experiment have shown that the disturbance from the fluid variation is minimized, and the dispensing consistency is improved greatly.