Kim A. Stelson

Active control of acoustic reflection, absorption, and transmission using thin panel speakers

This paper explores the development of thin panels that can be controlled electronically so as to provide surfaces with desired reflection coefficients. Such panels can be used as either perfect reflectors or absorbers. They can also be designed to be transmission blockers that block the propagation of sound. The development of the control system is based on the use of wave separation algorithms that separate incident sound from reflected sound. In order to obtain a desired reflection coefficient, the reflected sound is controlled to appropriate levels. The incident sound is used as an acoustic reference for feedforward control and has the important property of being isolated from the action of the control system speaker. In order to use a panel as a transmission blocker, the acoustic pressure behind the panel is driven to zero. The use of the incident signal as a reference again plays a key role in successfully reducing broadband transmission of sound. The panels themselves are constructed using poster board and small rare-earth actuators. Detailed experimental results are presented showing the efficacy of the algorithms in achieving real-time control of reflection or transmission. The panels are able to effectively block transmission of broadband sound. Practical applications for these panels include enclosures for noisy machinery, noise-absorbing wallpaper, the development of sound walls, and the development of noise-blocking glass windows.

Creep in injection molded starch/synthetic polymer blends

Abstract Creep in uniaxial extension was measured in starch/synthetic polymer blends. In the first experiment, the starch content was kept constant at 70% and the applied stress and temperatures were varied. In a second experiment, the starch content was varied from 0 to 70% at a constant applied stress and varying temperature. The synthetic polymers in the blend were either polybutylene succinate (PBS) or polycaprolactone (PCL), both semi-crystalline biodegradable polyesters. Maleic anhydride-functionalized polyesters (5% by weight) were used to reactively compatibilize starch and the synthetic polymer. The compliance of these blends was found to increase with increasing applied stress and temperature and decrease with increasing starch content. Addition of small amounts of starch was found to significantly inhibit creep. Time–temperature superposition was attempted for blends containing 30, 50 and 70% starch. The compliance master curve was successfully obtained for the blends containing 70 and 50% starch, but empirically estimated shift factors did not obey the WLF equation. WLF constants could be calculated based on the empirical shift factors and were in the range obtained for other polymers. The compliance data could be fitted to several empirical models. Excellent fits were obtained for compliance at different stress levels and temperatures.

Prediction of Filling Time and Vent Locations for Resin Transfer Molds

A previously reported algorithm for predicting the last point to fill in a porous mold is presented. This algorithm is modified for the prediction of the filling time and vent location in anisotropic RTM molds of various geometries with time dependent boundary conditions. The proposed method is computationally efficient and requires, at most, the solution of two linear systems of equations. For a number of test cases, the performance of the method is investigated on comparing predicted vent locations with those obtained using a complete filling simulation. It is demonstrated that the proposed method can predict fill times and vent locations with the same accuracy as a direct filling simulation but at a much reduced computational cost. Finally, it is shown how an alternative interpretation of the results from the vent location algorithm can be used to predict the location of all the potential dry spots (vent locations) in an RTM mold.

On the Plastic Deformation of a Tube During Bending

The objective of this study is to develop an analytical approach to calculate the relationship between the axial curvature of a bent tube and the resulting deformation of the cross-section. The model accounts for both geometrical and material nonlinearities. An approximate expression in trigonometric form is introduced for the displacement field, which reflects the change of wall thickness and neutral axis shift during bending. The total deformation theory is employed as a constitutive relation. The solution is found using a minimization approach and the energy principle. A better approach for springback prediction might be obtained from the deformation model, which predicts a more accurate moment of inertia change during bending.

Three-Dimensional Tube Geometry Control for Rotary Draw Tube Bending, Part 1: Bend Angle and Overall Tube Geometry Control

Traditional trial-and-error springback compensation methods have the problems of high scrap rate, low efficiency, high cost of fixtures and operator experience dependency. The method presented in this paper uses on-line measured springback data from the same batch to predict and compensate for springback. Because there are no springback data for the first bend, bend-rebend control is used to make the first bend to eliminate trial tubes. In addition to springback, relaxation and radial growth are also estimated and compensated for to make a bend more accurate. A process control method is developed to optimize the overall control strategy such that the overall tube error is minimized without increasing the required hardware accuracy. The optimal process control strategy has significantly higher accuracy than the traditional trial-and-error method. The details of statistical analysis of tube tolerance and adaptive bend correction algorithm are presented in Part 2 of the paper.

Two-Stage Actuation for Improved Accuracy of Contouring

Two-stage actuation is presented as a method for improved contouring accuracy of biaxial positioning systems. The method applies the concept of dynamic compensation to a two-stage structure where the position error of the the coarse stage serves as the input to the fine stage. The objective of dynamic compensation is to obtain an overall system with the bandwidth of the fine stage and the range of the coarse stage. Based on the widely used second order, Type 1 approximation of the actuator dynamics and proportional position control it is shown that the implementation of the fine stage increases the bandwidth of the entire system and, also, decreases phase lag at lower frequencies. Moreover, higher bandwidth results in smaller comering errors in the response of a single axis. However, the magnitude of the contour error in the neighborhood of the corner is not necessarily reduced by dynamic compensation. Finally, for constant velocity, straight line contours the advantage of two-stage actuation over the conventional single-stage systems lies in its ability to eliminate steady state contour error by eliminating steady state errors in the responses of individual axes.

A model predictive control approach for a parallel hydraulic hybrid powertrain

In this paper, a model predictive control (MPC) approach is presented for solving the energy management problem in a parallel hydraulic hybrid vehicle. The hydraulic hybrid vehicle uses a variable displacement pump/motor combination to transfer energy between the mechanical and hydraulic domains and a high pressure accumulator for energy storage. The proposed controller observes the gas and brake pedal positions to estimate the desired wheel torque. It regulates the engine throttle command, pump/motor displacement, and applied brake force to track this desired torque while optimizing the powertrain efficiency. Simulation studies were conducted to evaluate the effects of engine dwell time and accumulator capacity versus on the overall vehicle fuel economy.

Three-Dimensional Tube Geometry Control for Rotary Draw Tube Bending, Part 2: Statistical Tube Tolerance Analysis and Adaptive Bend Correction

The control of each individual bend and overall process is presented in Part 1 of the paper. In Part 2 of the paper, statistical methods are used to analyze and improve 3-D tube bending accuracy. The relationship between bending process error and tube geometry error is obtained with Monte Carlo simulation. For the same tube tolerance requirement, the required process tolerance varies in a large range based on tube geometry. Among the three bending errors: bend angle, bend plane and distance between bends, bend angle error has the largest influence on tube error. For a tube with multiple bends, the overall tube geometry error can be minimized by intentionally modifying the nominal values of the bends to be made based on the errors in the existing bends. The required modification of the bending commands is calculated with an adaptive bend correction algorithm.

Modeling, Control, and Experimental Validation of a Transient Hydrostatic Dynamometer

Due to its superior power-to-weight ratio, a hydrostatic dynamometer is an ideal candidate for transient engine or powertrain testing. It can load or motor the engine to follow any desired speed and acceleration profiles for real-world applications. Given its high bandwidth, the hydrostatic dynamometer can be further used to emulate the dynamics of hybrid powertrains and, therefore, investigate the interactions between the engine and the hybrid power source in real time. This will greatly expedite the research of various hybrid powertrain architectures and control methodologies, without actually building the complete physical system. This paper presents the design, modeling, tracking control, and experimental investigation of a transient hydrostatic dynamometer. A ninth-order physics-based dynamic model for the dynamometer is formulated and then identified and validated with experimental data. To control the dynamometer to emulate the real-world engine speed/torque profiles, two different nonlinear control strategies are investigated and implemented. First, a nonlinear model-based inversion plus PID control is designed to achieve precise tracking. Then, a state feedback control via feedback linearization is designed and implemented. Experimental results demonstrate precise tracking performance with less than 5% tracking error for both transient and steady-state operations.

Active Control of Sound Transmission Through Windows With Carbon Nanotube-Based Transparent Actuators

This paper explores the development of active sound transmission control systems for windows that can achieve a significant reduction in window noise transmission. Two major challenges need to be addressed in order to make the development of such noise blocking windows feasible. These are the need for a distributed actuation system that is optically transparent and the unavailability of a real-time reference signal that can be used by the active control system to provide advance information on the noise affecting the window. To address the first challenge, a transparent thin-film actuator (speaker) is first developed for the control system, which consists of a piezoelectric poly (vinylidene fluoride) (PVDF) thin film coated with compliant carbon nanotube-based transparent conductors on both sides. The developed thin-film speaker shows excellent acoustic response over a broadband frequency range, and has the advantages of being flexible, transparent, thin, and lightweight. To address the second challenge of providing a time-advanced reference signal from a moving noise source, a small microphone array distributed on the outside wall of the home is used. New noise source identification algorithms are employed, by which an appropriate microphone from the array can be chosen to provide a reference signal. Experimental results show that over 12 dB reduction in sound transmission is achieved globally in the case of broadband sound, which demonstrates the effectiveness of the control system in blocking sound transmission.