Edward C. De Meter

Fast support layout optimization

Abstract Machining-fixture supports are used to increase workpiece rigidity. A critical problem of machining-fixture design is where to place a fixed number of supports in order to minimize workpiece deformation during machining. This paper presents the Fast Support Layout Optimization (FSLO) model. The objective of the FSLO model is to determine support locations that will minimize the maximum displacement-to-tolerance ratio of a set of workpiece features subject to a system of machining loads. The FSLO model utilizes a Finite Element Analysis (FEA) model to characterize workpiece stiffness. Solution of the FSLO model improves an existing support layout by systematically altering the boundary conditions applied to the FEA model. The FSLO model is unique in that its solution time is both very small and insensitive to the size of the FEA model, the sizes of machined features considered, and the sizes of the candidate regions of the supports. In addition to describing the formulation of the FSLO model, this paper describes a set of experiments that were used for its verification.

An investigation into the use of spatial coordinates for the genetic algorithm based solution of the fixture layout optimization problem

The fixture layout optimization problem is highly modal. It has been shown that layout optimization methods that use the genetic algorithm (GA) are effective at finding high quality solutions. To date, such methods have used FEA node numbers to represent the locations of fixture elements. This paper presents a new GA based optimization method that uses spatial coordinates to represent the locations of fixture elements. This method integrates a number of important GA concepts such as real encoding, increased mutation, and sharing. Test results show that the new method provides higher quality solutions with substantially less numbers of generations than the old method. These results also show that the use of increased mutation and sharing significantly improves the performance of the new method.

A model to predict minimum required clamp pre-loads in light of fixture–workpiece compliance

Abstract The determination of minimum required clamp pre-loads is an important process in the design of machining fixtures. This paper presents a linear, clamp pre-load (LCPL) model that can be applied to fixture–workpiece systems whose compliance is load invariant. The model considers the static deformation of the fixture–workpiece system in response to the clamping process and the machining process. Sources of compliance throughout a fixture–workpiece system are considered. The model computes the minimum required pre-loads necessary to prevent workpiece slip at the fixture–workpiece joints throughout the machining process. This paper also describes an experimental study that was used to characterize the accuracy of the LCPL model with regard to the application of a ramping external load to a fixture–workpiece system. Over the contact conditions tested, the LCPL model was observed to overestimate the minimum required clamp pre-loads by an average of 7%. This experimental study also revealed the sensitivity of the computed pre-loads to the relative compliance of the fixture elements as well as the coefficient of friction.

An investigation of the effectiveness of fixture layout optimization methods

Abstract Deriving the optimal layout of fixture elements is critical to minimizing the impact of fixture–workpiece deformation on machined feature error. Various optimization methods for solving this problem have been reported. Unfortunately no investigation has been executed to compare their relative performance. This paper presents the methodology and results of an extensive investigation into the relative effectiveness of the main elements of these competing methods. All methods were tested over a broad range of conditions. Performance measures that were tracked included solution quality, solution repeatability, and computation time. The results of this investigation show that the best overall performance is provided by optimization methods that use both the genetic algorithm and continuous interpolation for the distribution of boundary conditions.

Part location algorithms for an intelligent fixturing system part 1: system description and algorithm development

An Intelligent Fixturing System (IFS) is currently being developed to hold a family of cylinder heads for machining operations. This system incorporates a Part Location System (PLS) to locate the workpiece reference frame of a cylinder head relative to a pallet. This two-part paper describes the development and evaluation of various part location algorithms that have been created for this application. Part 1 of this paper provides a detailed description of the IFS and the PLS and of the cylinder head location problem. It also provides a detailed description of three different algorithms that were developed for the PLS. The first algorithm, SeQuential Least Squares (SQLS), is based on the sequential, parametric regression model. The second class of algorithms, SiMultaneous Least Squares (SMLS), is based on the simultaneous, parametric regression model. The third algorithm, SQLS-SMLS (fixed cd) is a hybrid of the first two methods.

Light Activated Adhesive Gripper (LAAG) workholding technology and process

Abstract This paper describes the underlying concepts and proofof- concept testing of a new workholding technology referred to as Light Activated Adhesive Gripper (LAAG). Using this technology, a workpiece is held by a system of discrete adhesive joints. These joints are very strong, very tough, and dynamically stiff. Furthermore, light is used to cure or destroy these adhesive joints, on demand, within seconds. LAAG technology enables workpieces to be held with minimal pre-load distortion, with maximum dynamic rigidity, and with maximum accessibility to the machining process.

Restraint analysis of assembly work carriers

Abstract Work carrier design is critical to the performance and economic feasibility of automatic assembly systems. Unless parts are securely held throughout transport, insertion, and fastening operations, precision assembly is impossible to achieve. This paper presents analytical techniques for determining whether carrier locators provide sufficient kinematic restraint. It presents techniques for determining whether carrier clamps and locators have the potential to resist arbitrary external forces. Finally, it presents a model for determining the minimum clamp actuator intensities necessary to restrain parts throughout transport and assembly.

Global workpiece positioning system (GWPS) Part 1: Concept and development of a laser probe based system

Abstract One component of a modern machining system that has remained virtually unchanged over time is part location. Current methods of part location use physical datum surfaces on the workpiece. Workpieces are positioned relative to the axial reference frame of the machine tool by placing these datum surfaces in contact with locators in the fixture. Due to datum surface variation and fixturing errors, workpiece positioning is often inconsistent and variable for machining operations requiring multiple setups. In turn, this leads to datum-related geometric errors of the machined surfaces. This paper presents a potential solution to these problems in the form of the Global Workpiece Positioning System (GWPS) concept. Using the GWPS concept, artifacts attached to the workpiece are used to define a system of datum target points that, in turn, are known relative to a global workpiece reference frame fixed to the workpiece. Using a target point sensing process, the locations of subsets of artifact target points are mapped into the axial space of the machine tool with a level of uncertainty equivalent to the axial motion uncertainty of the machine tool. In this way, the location of the global workpiece reference frame with respect to the machine tool axial reference frame is defined. All machining operations for all setups are performed relative to this reference frame. This paper, which is part one of a two-part paper, describes the GWPS concept. It further describes the development, testing, and evaluation of a laser probed based GWPS. The average positioning accuracy of this system was determined to be 1.3 μm, which is within the axial motion uncertainty of the machine tool on which it was implemented. This measure is based on the location of a drilled and reamed hole. The second paper describes the implementation and testing of the GWPS with ultrasonic transducers and receivers. This system utilizes time-of-flight ranging and is similar in principle to the current satellite-based Global Positioning System (GPS).

Characterization of the quasi-static deformation of LAAG joints adhering machined steel surfaces

Light Activated Adhesive Gripper (LAAG) workholding technology is a means by which a workpiece is held by adhesive joints that can be instantaneously cured or destroyed, on demand. A LAAG joint is the adhesive bond between the gripper pin and workpiece. Due to the novelty of this concept, no knowledge exists with regard to how LAAG joints deform and fail during quasi-static loading. This paper describes an investigation that was carried out to characterize the strength, ductility, and failure modes of a LAAG joint adhering a machined, steel surface subject to axial loading and shear loading.

Part location algorithms for an Intelligent Fixturing System Part 2: algorithm testing and evaluation

Abstract An Intelligent Fixturing System (IFS) is currently being developed to hold a family of cylinder heads for machining operations. This system incorporates a Part Location System (PLS) to locate the workpiece reference frame of a cylinder head relative to a pallet. This two-part paper describes the development and evaluation of various part location algorithms that have been created for this application. Part 2 of this paper describes a series of simulation experiments that was carried out to assess the accuracy and computation speed of the SeQuential Least Squares (SQLS) algorithm, the SiMultaneous Least Squares (SMLS) algorithm, and the SQLS-SMLS (fixed cd) algorithm. The results of this study reveal that the SQLS algorithm is fast to execute, but its accuracy is very sensitive to random variations in the coordinate data. Alternatively, the SMLS algorithm is insensitive to random variations in the coordinate data but is computationally more expensive to execute. Furthermore, its accuracy is sensitive to bias errors in the chordal distance between the dowel holes. Only the SQLS-SMLS (fixed cd) is insensitive to both types of variations.