Dipankar Sanyal

Approaching Servoclass Tracking Performance by a Proportional Valve-Controlled System

An industry-grade proportional valve is much cheaper and rugged than a servovalve. A feedforward controller for a proportional valved system has been developed here to achieve tracking controls beyond 1 Hz that are usually attained by servovalves. For compensating the higher nonlinearities, feedforward controllers have been designed offline by proposing appropriate static models for friction and valve flow and executing the supporting experiments. These controllers have been implemented with real-time PID feedback of only the piston displacement. Excellent tracking performance has been obtained up to 2 Hz that has deteriorated with an increase in cylinder friction.

An investigation of nonlinear dynamics of a thermal pulse combustor

Nonlinear dynamics of a thermal pulse combustor was investigated using a coupled fourth order lumped combustor model. The effects of optically thin radiation from the combustor gases to the wall are also investigated. The system response changes from steady combustion to extinction through oscillatory combustion as the wall temperature is lowered. The qualitative results suggest a transition to chaos through a period-doubling route prior to extinction. The presence of chaos is also confirmed by quantitative measures like correlation dimension and Lyapunov exponent. The system dynamics is qualitatively similar in presence and absence of radiation. However, inclusion of radiative heat loss leads to extinction at higher temperatures and also increases the predicted range of wall temperatures for which limit cycle behaviour is obtained. A measure for monitoring the proximity of the system to extinction has been developed using the time series data.

GA-optimized feedforward-PID tracking control for a rugged electrohydraulic system design

Rugged electrohydraulic systems are preferred for remote and harsh applications. Despite the low bandwidth, large deadband and flow nonlinearities in proportional valves valve and highly nonlinear friction in industry-grade cylinders that comprise rugged systems, their maintenance are much easier than very sophisticated and delicate servocontrol and servocylinder systems. With the target of making the easily maintainable system to perform comparably to a servosystem, a feedforward control has been designed here for compensating the nonlinearities. A PID feedback of the piston displacement has been employed in tandem for absorbing the unmodeled effects. All the controller parameters have been optimized by a real-coded genetic algorithm. The agreement between the achieved real-time responses for step and sinusoidal demands with those achieved by modern servosystems clearly establishes the acceptability of the controller design.

Effects of flow inertia modelling and valve-plate geometry on swash-plate axial-piston pump performance

The present work aims at developing a methodology for designing a swash-plate axial-piston pump whose barrel kidneys are wider than the bridges separating the kidney ports on the plate. Besides reducing pressure ripples and avoiding cavitation for minimizing the fluid-borne noise, maximization of the pump discharge has also been considered in formulating an objective function to be minimized. A comprehensive mathematical model considering oil compressibility and inertia has been developed and implemented in MATLAB/Simulink for dynamic simulation of the pump flow. The model has been found acceptable by observing a good match between its predictions of mean discharge and delivery pressure and available experimental results. A detailed parametric study has been performed and the optimum angles for the valve-plate pre-compression and the barrel kidneys have been obtained.

Theoretical simulation of ripples for different leading-side groove volumes on manifolds in fixed-displacement axial-piston pump

For many hydraulic systems, fixed-displacement axial-piston pumps often employ swash plate. The design challenge for such a pump is to minimize the pressure and flow ripples and the consequent noises. Although the use of grooves at the leading and trailing sides of the pump manifolds has been adopted by several manufacturers as a remedy to this problem, theoretical modelling and analysis of the effects of these silencing grooves are very limited. This study presents a model that puts special emphasis on analysing the effect of volume variation of the silencing grooves. All the non-linear dynamic and algebraic equations developed during modelling have been solved in a Matlab/Simulink framework. The analysis has been carried out for a pump with leading-side manifolds, since experimental results for such a pump were available. Through a constant-speed parametric analysis at fixed load, optimal dimensions for these grooves have been indicated. The major contribution of the present work is the development of a mathematical model that attempts an explicit solution of pressure within each silencing groove. The model has been presented in the analysis as an effective design analysis tool for minimizing the pressure and flow ripples.

Design of an Adaptive Fuzzy-Bias SMC and Validation for a Rugged Electrohydraulic System

An adaptive fuzzy-bias sliding-mode controller has been designed by adding biasing to an existing controller for motion tracking by an electrohydraulic system. The system has a proportional valve with large deadband and a cylinder with jump in static friction and nonlinear but continuous dynamic friction. While the biasing controller takes care of the discontinuous features, the fuzzy controller approximates the other system nonlinearities. The sliding-mode controller compensates the approximations. By a number of real-time experiments, the appropriate choices for the number of fuzzy rule base and some free parameters have been established. Then, starting with the same set of initial parameters for tracking step and sinusoidal demands of different frequencies, the responses have demonstrated sufficient adaptability of the controller.

Effects of Tailpipe Friction on the Nonlinear Dynamics of a Thermal Pulse Combustor

The effect of tailpipe friction on the combustion dynamics inside a thermal pulse combustor has been investigated using a nonlinear model consisting of four coupled first order ordinary differential equations. The dynamics of the system is represented through time series plots, time-delay phase plots, and Poincare maps. The results indicate that as the tailpipe friction factor is lowered, the system undergoes a transition from steady combustion through oscillating combustion to an intermittent combustion with chaotic characteristics before extinction. The time series data are shown to be useful indicator for early detection of extinction. In one approach (thresholding), the occurrence of local peak pressures below a predefined threshold value is identified as an event and the number of events (event count) and largest number of successive cycles with such events (event duration) are recorded as the friction factor is lowered. In another approach, the statistical moments (kurtosis) of the data are used. Number of kurtosis peaks above a prescribed value and variance of the kurtosis values are recorded for decreasing values of friction factor. All these numbers sharply increase as the system approaches extinction.

GA-Optimized Fuzzy-Feedforward-Bias Control of Motion by a Rugged Electrohydraulic System

An electrohydraulic system with a proportional valve and industry-grade cylinder has been used to target the servoclass tracking performance. Such systems have a wide range of heavy-duty applications, where the environment could be quite dirty along with the demands becoming faster and more precise every day. High static friction in the cylinder and large deadband of the valve in the system pose control challenges that are more severe than in a system with a servovalve and a low-friction cylinder. A fuzzy-feedforward-bias controller has been developed and a genetic algorithm has been used to optimize the controller parameters. The real-time control experiments revealed excellent tracking throughout the cycle for sinusoidal displacements beyond 1.5 Hz.

A Sphericosymmetric VOF Approach for Investigating Immiscible Two-Phase Systems with One Liquid Phase

A numerical model based on the volume-of-fluid (VOF) method in sphericosymmetric geometry containing two immiscible phases is developed. The model is successfully validated using a wide class of available results. These include the inward and outward solidification problems and growth of vapor film in a saturated and metastable liquid. The model is also used to simulate the very rapid collapse phenomenon of vapor film around a hot metal in subcooled water. In each case, the results indicate that the numerical simulation successfully captures the essential physics, even at small length and time scales. It is established that the proposed numerical scheme correctly predicts the advection within the VOF framework in situations with pronounced curvature.

Real-time fuzzy-feedforward controller design by bacterial foraging optimization for an electrohydraulic system

High power-to-weight ratio rugged electrohydraulic systems are extensively used for land tilling, harvesting, construction and various industrial control operations. Flow, friction and valve deadband nonlinearities of these systems make the controller design quite challenging, especially for meeting fast and precise motion tracking requirements. For such a system, a bacterial foraging optimization, or BFO, has been employed with a simple swarming method proposed here to design a real-time controller. With a fuzzy voltage compensating the flow-related continuous nonlinearities, a feedforward voltage has been employed to cater for known external loading and discontinuous internal nonlinearities due to the valve deadband and cylinder stiction. Acceptability of the optimized controller has been demonstrated by testing the performances for square, trapezoidal and triangular demands for the piston position variation with time.