Adaptive Dynamic Programming-Based Control of Chatter in Turning: Experimental Validation on a Laboratory Emulator

Chatter vibrations pose a critical challenge in internal turning operations, leading to reduced machining accuracy, tool life, and productivity. Traditional passive and model-based active control approaches often struggle with parameter uncertainties and variations in cutting conditions. This study presents an experimental validation of an Adaptive Dynamic Programming (ADP)-based control approach for suppressing chatter in internal turning using a model-free learning framework. A laboratory emulator, consisting of a cantilever beam with piezoelectric sensors and actuators, replicates the dynamics of a slender cutting tool. Chatter is simulated using an electromagnetic actuator with displacement feedback to mimic regenerative effects. The ADP algorithm iteratively refines control laws based on real-time input-output data, eliminating the need for an explicit system model. Experimental results demonstrate significant improvements in chatter stability limits, effectively increasing modal damping. Compared to traditional control techniques, the proposed approach enhances machining stability and expands chatter-free operating conditions, providing a scalable, adaptive alternative for industrial applications.