Adaptive control to actively damp bistabilities in highly interrupted turning processes using a hardware-in-the-loop simulator

Abstract

Interruptions in turning make the process forces non-smooth and nonlinear. Smooth nonlinear cutting forces result in the process of being stable for small perturbations and unstable for larger ones. Re-entry after interruptions acts as perturbations making the process exhibit bistabilities. Stability for such processes is characterized by Hopf bifurcations resulting in lobes and period-doubling bifurcations resulting in narrow unstable lenses. Interrupted turning remains an important technological problem, and since experimentation to investigate and mitigate instabilities are difficult, this paper instead emulates these phenomena on a controlled hardware-in-the-loop simulator. Emulated cutting on the simulator confirms that bistabilities persist with lobes and lenses. Cutting in bistable regimes should be avoided due to conditional stability. Hence, we demonstrate the use of active damping to stabilize cutting with interruptions/perturbations. To stabilize cutting with small/large perturbations, we successfully implement an adaptive gain tuning scheme that adapts the gain to the level of interruption/perturbation. To facilitate real-time detection of instabilities and their control, we characterize the efficacy of the updating scheme for its dependence on the time required to update the gain and for its dependence on the levels of gain increments. We observe that higher gain increments with shorter updating times result in the process being stabilized quicker. Such results are instructive for active damping of real processes exhibiting conditional instabilities prone to perturbations.

Hardware-in-the-loop simulator for emulation and active control of chatter

“…HiL simulators have been shown to be useful to investigate chatter [3][4][5][6][7][8][9][10][11][12][13], and to actively control it [3][4][5][9][10][11][12][13]. Ganguli et al [3,4] and Mancisidor et al [5] used a cantilever beam approximating a flexible workpiece and actuators for cutting force emulation.…”

“…Though the setups used in [6] , [7] , [8] are real machines with an actual spindle unit, they are costly. Mancisidor et al [9] , [10] and Sahu et al [11] , [12] , [13] have used a simpler flexure that approximates a linear flexible single degree of freedom workpiece. The flexure used in [9] , [10] , [11] , [12] , [13] provided sufficient space to mount two actuators, one of which imitated cutting forces and the other served as an active damper The simple design facilitated emulations to study the influence of different nonlinear force characteristics on chatter without having to worry about potential nonlinearities in the structure influencing stability.…”

“…Mancisidor et al [9] , [10] and Sahu et al [11] , [12] , [13] have used a simpler flexure that approximates a linear flexible single degree of freedom workpiece. The flexure used in [9] , [10] , [11] , [12] , [13] provided sufficient space to mount two actuators, one of which imitated cutting forces and the other served as an active damper The simple design facilitated emulations to study the influence of different nonlinear force characteristics on chatter without having to worry about potential nonlinearities in the structure influencing stability. The simple design also facilitated investigations for gain tuning for active vibration control in the presence of cutting process nonlinearities resulting in bistabilities [11] , [12] , [13] .…”

“…Since the HiL simulator under consideration has already been validated for its capability to emulate chatter in the presence of process nonlinearities [11] , [12] , and for testing active control strategies [11] , [12] , [13] , it may be further leveraged to: investigate chatter with flexures that exhibit nonlinear behavior, investigate new models for stability prediction, investigate new strategies for (non)model-based adaptive control, investigate new advanced robust control strategies, and to generate data sets to train the AI-ML-based data-driven models that could help detect and avoid chatter during real cutting in Industry settings. …”

Design and evaluation of an MRF damper for semi-active vibration control of the machining processes

“…The primary and secondary shakers provide regenerative simulated cutting force and active vibration damping force in this approach, respectively. The simulated signal is generated through the NI LabVIEW software [4]. However, the milling machine's complex nature limits the hardware-in-loop approach's use.…”