Developing a computational model of a pilot-operated control valve with advanced numerical methods
Ahmed Aissa Berraies defended his PhD thesis at the Department of Mechanical Engineering on September 3rd.
Pilot-operated control valves (POCV) are a specific type of high-tech miniature valves that have potential for complex applications. However, these valves encounter a significant challenge when proportional control is needed. This is a mode with continuous adjustments to flow instead of ON or OFF flow switching. As soon as the valve reaches a certain setpoint, it begins to vibrate, creating noise and an unstable flow. Within his PhD research, Ahmed Aissa Berraies uses advanced numerical methods for fluid-structure contact interaction (FSCI) to create a computational model of the valve. The results from this research can be used to make POCV more suitable for complex applications and increase its market value.
The model that was created enables the analysis of the complex interplay between the fluid and moving parts and allows assessing the onset of the vibrations. demonstrates in his research that the vibrations are a form of flutter called self-excited oscillations. They are triggered by rapid pressure changes and a Venturi effect, whereby fluid accelerates through a narrow gap, causing pressure to drop right when the valve is supposed to seal. These findings provide important insight to mitigate flow-induced vibrations.
Redesigning and evaluating the valve
Berraies subsequently redesigned the valve computationally and leveraged available FSCI simulation tools to evaluate it. The resulting simulations depict that the new design not only eliminates the vibrations during proportional control operations but also makes the valve respond faster. This way the research demonstrates the merits of virtual prototyping in the design process and optimization for more reliable and efficient products.
Challenges with the numerical methods
Nevertheless, this research also revealed a formidable challenge with regard to the numerical methods themselves. When the POCV was modeled as nearly closed or at early opening setpoints, simulations tended to fail or converge slowly. In these nearly closed situations, the way the numerical iterative method manages the interaction between fluid and structure introduces an unexpected damping force that can disrupt the calculations. A previously uncharacterized phenomenon called the added-damping effect was identified and described during the research. Such an effect explains why standard FSCI simulation tools using the Dirichlet-Neumann scheme may struggle in such specific, yet common scenarios.
Therefore, this research provides a twofold contribution. First, it provides a clear fluid-structure-contact analysis of a problematic microfluidic valve and a numerical design tool for design and process optimization. Second, it offers a new fundamental understanding of the added-damping effect, which can help improve existing numerical methods and develop more robust simulation tools.
Title of PhD thesis: . Supervisors: Prof. Harald van Brummelen and Prof. Ferdinando Auricchio.