Combinatorics of Contact-Application to Fixturing
P. M. Ferreira,* R. Marin
University of Illinois; University Scholars Award; Caterpillar, Inc.

Contact plays an important part in a number of manufacturing planning problems: automated fixturing, robotic grasping, assembly planning, NC code generation. In this project we study problems in feasibility, synthesis, and analysis of contact. For example, the kinds of questions we seek answers to are: Given n bodies with m contact relations defined on them, is it possible to produce a spatial configuration of the bodies to simultaneously satisfy all stated contact constraints? If such a configuration exists, what is the dimension of the solution space (or, does the specification produce a structure or a mechanism)?

Cutting Fluid Recycling through the Ultrafiltration Process
R. E. DeVor,* S. G. Kapoor,* S. Skerlos, N. Rajagopalan
Partnership Illinois; NSF Machine-Tool Agile Manufacturing Research Institute; IRMCO Advanced Lubricant Technologies; Illinois Waste Management Research Center

The scope of this research is to develop ultrafiltration technology which will alleviate or eliminate liabilities (environmental, health, performance, financial) by cleaning and reusing the cutting fluid. This ultrafiltration research aims to drastically reduce pollution emissions to the environment from cutting fluid waste while eliminating risk factors for occupational airway disease and allergic contact dermatitis as well as improving process performance while greatly reducing disposal and acquisition costs. Ceramic membranes are being studied to filter synthetic and semi-synthetic cutting fluids. Experimentation to date has focused on the flux characteristics of uncontaminated cutting fluids as a function of additives including lubricants, defoamers, and biocides.

Cutting Insert Selection
P. Ferreria,* S. Roy
Caterpillar, Inc.

This project develops a software system to aid in the selection of cutting tool inserts for machining operations. Using experimental performance data on the insert, the selection process is formulated as a nonlinear optimization problem.

Decision Support Modules for Machine Tools
P. M. Ferreira,* S. Roy, S. Jain, V. Subramanian
Caterpillar, Inc.

Modern NC machine tools have different configurations and capabilities and produce vastly different results when executing the same NC program to produce the same workpiece geometries. This causes problems, both in process planning and in capability analysis. In this project, we are developing an integrated environment for storing and updating machine capability and performance data.

Development and Calibration of a Parallel-Link High-Speed X-Y Table
P. M. Ferreira,* Z. Abadin, M. Lee
University of Illinois; Ford Machine-Tool Research

In this project we attempt to design and fabricate a novel X-Y translation system. In contrast to conventional 'stacked axes' configurations, this design is such that both (x and y) actuators are simultaneously grounded (so the table is the only moving member in the mechanism), and there is no asymmetry in the inertial loads carried by the actuators. This allows for uniform behavior across the workspace and low inertial loads, making this a design particularly useful in high-speed, high-accuracy applications.

Dry Machining with Applications to Drilling and Tapping
S. G. Kapoor,* R. E. DeVor,* S. Kalidas
Hayes-Lemmerz International; NSF I/UCRC Center for Machine Tool Systems Research

Recently, a considerable amount of attention has been directed to the problem of dry machining, in large part because of the environmental and health concerns associated with the use of cutting fluids. In drilling and tapping, dry machining is complicated by the problems associated with chip clogging and chip removal. This project is investigating reduced-cutting fluid and dry drilling. Attention is being focused on the use of thin-film coatings on drills to improve lubrication and chip removal characteristics of the drilling process. Current modeling work is focusing on the prediction of temperatures. Temperature measurements during drilling experiments are driving the model building and validation processes.

Enhancement of Hole Quality in Drilling
R. E. DeVor,* S. G. Kapoor,* K. Ehmann* (Northwestern Univ.), J. Ni* (Univ. of Michigan), K. Gupta
National Science Foundation

Research has been directed at the modeling and prediction of the drilling process such as torque and thrust (U of I). This work must now be extended to link these drilling process parameters to process responses that describe hole quality, including cylindricity, roundness, location errors, and oversize errors. This multi-university project is directed toward the development of models that will predict hole quality as a function of drill geometry and drilling process conditions. Drilling process dynamic characteristics which give rise to initial drill skidding and wandering are being included in the models as well as models for cutting forces and drill temperature effects.

Five-Axis Machining of 3-D Free-Form Surfaces
S. G. Kapoor,* R. E. DeVor,* R. Zhu
NSF I/UCRC Center for Machine Tool Systems Research

With recent advances in CNC machining centers, 3-D free-form surfaces such as dies, molds, turbines, and other aerospace components are being machined using five-axis machining centers. In order to assure part quality and higher productivity, the knowledge of cutting force and vibration system during machining is crucial. The purpose of this project is to develop a mechanistic model for predicting cutting forces and tool deflection in machining of free-form surfaces. A generalized chip load calculation procedure is being developed which will take into account the cutter geometry like ball end mill as well as tool path interference with the workpiece geometry.

Logical Control of Large-Scale Discrete Event Systems with Application to Flexible Automation
P. M. Ferreira,* C. Yuan
University of Illinois; University Scholars Award; University of Illinois Manufacturing Research Center

Structural control refers to the shaping of the structure of the state space of a discrete-event system. The state space is a directed graph and one of the most important structural properties required of this graph is that the component containing the initial state be strongly connected. This guarantees that the system is free of deadlocks under normal operation. In this project we devise control policies (which are essentially cuts on this directed graph) that are polynomially computable and guarantee strong conductivity while ensuring that the size of the strongly connected component is large. Special system structures under which these cuts are 'optimal' are also explored.

Machining Advisor
R. E. DeVor,* S. G. Kapoor,* J. Lu
NSF I/UCRC Center for Machine Tool Systems Research

Given inputs for part material, tool material, tool geometry, and cutting conditions, models for machining operations such as milling and drilling predict cutting forces, tool and workpiece deletions, and associated part quality characteristics. In a process planning environment, however, it is desired to determine the necessary inputs for a given set of output requirements. This inverse problem is being addressed in this project. A machining advisory system is being developed wherein the process model is used to identify the values of a set of decision variables which optimize a set of process objectives such as cycle time, tool life, and part dimensional accuracy.

Mechanistic Model for the Prediction of Cutting Forces under Tool Wear
S. G. Kapoor,* R. E. DeVor,* X. Liu
NSF I/UCRC Center for Machine Tool Systems Research

Mechanistic models have been developed to predict cutting forces in machining operations. These models consider only the case for which the tool is sharp. In this project, a contact stress field model is being used together with the mechanistic model for sharp-tool to predict forces due to flank wear. This model incorporates both the elastic and plastic contact characteristics on the flank of the tool. Only one wear test in conjunction with a series of sharp tool tests is required to calibrate the worn tool force prediction model. Good results have been obtained for orthogonal and oblique machining and more recently for turning.

Modeling and Analysis of Turning Processes with Contoured and Interrupted Surfaces
R. E. DeVor,* S. G. Kapoor,* R. Reddy
NSF I/UCRC Center for Machine Tool Systems Research

Turning simulation models considered in the past were suitable for uniform cross-section along the length of cut/axis of the part. However, while machining an axial contour geometry, these models fail to consider the variation of cutting parameters along the length of cut. The purpose of this project is to develop a process model for prediction of cutting forces and surface errors in contour turning and to study the effects of workpiece contour geometry and varying workpiece dynamics on process performance. The effect of cutting parameters on burr formation in turning interrupted surfaces is also being studied.

Modular Machining Fixture Test-Bed Development
R. E. DeVor,* S. G. Kapoor,* B. Fang
NSF Machine-Tool Agile Manufacturing Research Institute

Over the last four years, research on fixture design and fixture analysis has been conducted within the Machine-Tool Agile Manufacturing Research Institute. With the progress in theoretical understanding and model development, there is an increasing need to conduct experimental studies for model validation. The objectives of this project are to develop an experimental fixture test-bed that will take into account the needs of our industrial university partners by supporting a wide spectrum of experiments to study both static and dynamic characteristics of workpiece-fixture systems and to validate results from theoretical model development on fixture design and performance analysis.

Next-Generation Intelligent Monitoring System
R. E. DeVor,* S. G. Kapoor,* J. A. Stori,* L. Yang, R. Zhu, Q. Mezentsev, J. Lim
National Institute of Standards and Technology, Advanced Technology Program; Montronix, Inc.

Current machine tool monitoring systems are limit-based and do not function well when process variations such as stock size and hardness variations arise. The result is false alarms that seem to indicate other problems that are really not present. The goal of this project is to develop a machining process model-based on-line intelligent monitoring system to detect, isolate, and identify process variations and faults in machining processes. Using an open architecture control enabling environment improved fault detection and diagnostics will lead to adaptive control strategies for process optimization.

Noncircular Turning Process for Camshaft Machining
T. C. Tsao,* R. E. DeVor,* S. G. Kapoor,* A. G. Alleyne,* A. Babinski, Z. Sun, B. S. Kim, X. C. Tan, R. Gajjela
National Institute of Standards and Technology, Advanced Technology Program; Saginaw Machine Systems, Inc.

The purpose of this project is to develop a breakthrough technology that will revolutionize the manufacturing of camshafts. The technology is based on the use of variable-depth-of-cut machining in a single-point turning environment to produce noncircular shapes using a combination of rapid actuation of the tool slide and high-speed, real-time, digital signal processing and precision motion-control schemes. This technology enables the generation of a wide variety of cam profiles in software, creating an agile manufacturing process that will meet evolving trends and competitive needs for U.S. camshaft manufacturing in the years to come.

On-Line Tool Wear and Cycle Time Optimization
J. A. Stori,* J. Lim
National Institute of Standards and Technology; Montronix, Inc.

During the launch phase of a production facility, balancing the inevitable trade-off between cycle-time, tool-wear, part-quality, scrap-rate, and failure-rate is of primary concern to the process engineer. In the domain of machining operations, such decisions are typically ad hoc. Data that may be available from a process monitoring system is rarely utilized during the part programming stage. Algorithms are being developed that adaptively update process parameters based on real-time sensor data and key parameters specified by the process engineer to improve part count and optimize cycle times and tool wear.

Parametric Cost Modeling
M. L. Philpott,* M. Marini, W. Wu, D. Wood
John Deere Harvester

A methodology for Parametric Cost Modeling is being developed which seeks to interactively provide designers with manufacturing cost information as they create a design using a 3D CAD solid modeler. A user interface has been created for Parametric Technology's Pro/ENGINEER? solid modeler (ProE). The interface is being linked to a number of critical cost drivers via parametric cost models derived directly from John Deere's manufacturing cost database. Feature recognition algorithms are employed to extract key geometry features from the part model. A cost window is automatically created each time the designer adds geometric features to the evolving design.

Precision Machining-Sensors and Control
M. L. Philpott,* T. C. Tsao,* R. E. DeVor,* S. G. Kapoor,* D. Kim
NSF/ARPA Machine-Tool Agile Manufacturing Research Institute, DMI 93-20944; Montronix, Inc.

The objective of this research is to develop a common hardware and software platform for implementing sensor-based precision machining control. Focus is on the processes of grinding and single-point turning, as they offer the most potential for improvement in light of present usage in industry. The project addresses basic research in sensors and control to improve the performance of existing machine tools as well as to provide the basis for improvements in design for the next generation of machine tools.

Prediction of Residual Stresses in Machining
S. G. Kapoor,* R. E. DeVor,* Y. Liu
NSF I/UCRC Center for Machine Tool Systems Research

This project is developing models for the prediction of machining-induced residual stresses. Experimental studies have been conducted to understand the nature of the stress profiles induced by orthogonal and three-dimensional processes viz., turning and the end milling processes. X-ray defraction techniques are used to measure the residual stresses. The predictive model includes both mechanical and thermal effects and links machining processing conditions to the deformation fields at and below the machined surface. The results are being used to develop a model to predict part warpage due to the machining-induced residual stresses. Recent work is focusing on model enhancement to accommodate pre-stressed conditions.

Process-Conscious Tool Path Generation
J. A. Stori,* H. Wang, P. Huang
University of Illinois

Tool-path generation for machining operations has traditionally been approached from a purely geometric perspective. When the cutting mechanics and process dynamics are considered, existing tool path strategies are found to be significantly lacking. Excessive plunging and slotting, sharp velocity discontinuities, and changing cut geometry limit production rates, reduce part quality, and increase tool wear. New algorithms are developed to reduce variations in cutter engagement and chip geometry resulting in a stable, predictable, and controllable process. Particular emphasis is placed on accommodating the dynamic limitations of modern high-speed machining centers.

Productivity and Compatibility of Cutting Fluids with Microfiltration Membranes
R. E. DeVor,* S. G. Kapoor,* S. J. Skerlos
NSF I/UCRC Center for Machine Tool Systems Research

While the ability of membrane filtration to remove cutting fluid contaminants is well proven, the capacity of the technology to accomplish this without disrupting cutting fluid chemistry is not known. This research aims to identify the chemical characteristics of cutting fluid ingredients that determine the compatibility and productivity of a given cutting fluid/membrane combination. The specific research objectives are to assess the compatibility of both non oil-containing (synthetic) and oil-containing (semisynthetic and soluble oil) cutting fluids with aluminum oxide microfiltration membranes and to investigate different microfiltration membrane materials and geometries for compatibility with synthetics, semi-synthetics, and soluble oils.

Productivity and Quality Improvement in Deep-Hole Drilling
R. E. DeVor,* S. G. Kapoor,* B. Ozdoganlar, J. Mellinger
Delphi Automotive Systems; NSF I/UCRC Center for Machine Tool Systems Research

The goal of this project is to improve the productivity and quality of the deep-hole drilling process by understanding the major contributors to chip clogging and poor chip evacuation. Extensive experimental studies will support model development and model validation leading to the determination of the process and drill geometry variables that can facilitate the evacuation of chips in the drilling process. The project results should lead to reduced tool breakage and therefore less downtime as well as shorter machining cycles and better hole quality.

SMARTCUTS-Simultaneous Machining with Real-Time Control of Tooling Systems
P. M. Ferreira,* R. E. DeVor,* S. G. Kapoor,* M. Johnson, Z. Abadin
NSF Machine-Tool Agile Manufacturing Research Institute; NSF I/UCRC Center for Machine Tool Systems Research

In this project we have designed and built a machine tool based on a novel parallel-link mechanism called the tetrahedral tripod. To do so, we have developed a number of analysis and synthesis tools to aid in the design process of such machine tools. For example, we have developed analytical procedures for producing bounds on the stiffness of parallel-link machine tools across the workspace. Currently, we are developing a second high-speed, high-stiffness, 3-axis, parallel-link machine to serve as a universal tooling system (i.e., either for workholding or for cutting tool positioning).

Scalable Flexible Manufacturing
P. M. Ferreira,* S. Jain, V. Subramanian
Rockwell-Collins

In this project, the problem of developing scalable flexibility in printed circuit board production is addressed. The goal of the project is to develop a system capable of taking a printed circuit board design and generating its process plans and insertion programs for each process on its route. Furthermore, for a given set of boards being manufactured, the overall system control logic is simultaneously generated.

Simulation-based Machining Parameter Optimization
J. A. Stori*
University of Illinois

Mechanistic process simulations have proven valuable for the prediction of machining state variables over a wide range of operating parameters. Such simulation tools, however, are seldom an integral part of machining parameter optimization modules. To increase the predictive accuracy of parameter selection decisions, simulation feedback is used to tune approximate analytic models during the optimization process. This iterative procedure makes efficient use of the computationally expensive process simulation. Gradient-based descent methods are combined with parameter-space decomposition techniques to explore the complex topology of the operating space.

Spindle-based On-Line Force Monitoring and Spindle Speed Variation for Machine-Tool Chatter Prevention
S. G. Kapoor,* R. E. DeVor,* B. Ozdoganlar, O. Mezentsev
Kistler Instruments; Ingersoll Milling Machine Co.; NSF I/UCRC Center for Machine Tool Systems Research

The spindle speed variation (SSV) method results in an augmentation of the machining stability and improvement in the surface quality. The University of Illinois and the Ingersoll Milling Machine Company have been collaborating on this project to develop a chatter-avoidance technology based on the SSV method. A special spindle/motor/drive system has been designed and constructed and a test plan is being developed to validate the theory developed for automotive industry powertrain applications. Kistler Instruments, a second industrial partner on the project, has developed a new spindle-based force ring monitoring technology that is being employed in the testbed to facilitate on-line adaptive control research for vibration reduction.

Strategic Quality Deployment and Value Benchmarking
H. E. Cook,* E. Cowan, C. Bush, M. Pozar, M. Neidlinger, A. Wu, M. Lee, T. Mosely
C. J. Wicall Gauthier Professorship; University of Illinois; Ford Motor Co.; Caterpillar, Inc.; General Electric Co.

The objectives of this research are to (1) develop the relationships between the metrics of quality, cost, lead time, and innovation and the traditional bottom line functions of return on investment and market share; (2) to generate a fundamental and encompassing definition of quality that includes considerations of value and cost and apply it to the entire product realization process; and (3) to explore the role of organization structure and corporate culture on manufacturing effectiveness. Initial results, based upon a market model that incorporates value as well as the more traditional elements of cost and price, show that a single universal metric governs manufacturing effectiveness. The new quality function being developed yields the traditional Taguchi formalism as a limiting case.