Process Modeling Efforts Research Articles

Stability modeling for chatter avoidance in self-aware machining: an application of physics-guided machine learning

Physics-guided machine learning (PGML) offers a new approach to stability modeling during machining that leverages experimental data generated during the machining process while incorporating decades of theoretical process modeling efforts. This approach addresses specific limitations of machine learning models and physics-based models individually. Data-driven machine learning models are typically black box models that do not provide deep insight into the underlying physics and do not reflect physical constraints for the modeled system, sometimes yielding solutions that violate physical laws or operational constraints. In addition, acquiring the large amounts of manufacturing data needed for machine learning modeling can be costly. On the other hand, many physical processes are not completely understood by domain experts and have a high degree of uncertainty. Physics-based models must make simplifying assumptions that can compromise prediction accuracy. This research explores whether data generated by an uncertain physics-based milling stability model that is used to train a physics-guided machine learning stability model, and then updated with measured data, domain knowledge, and theory-based knowledge provides a useful approximation to the unknown true stability model for a specific set of factory operating conditions. Four novel strategies for updating the machine learning model with experimental data are explored. These updating strategies differ in their assumptions about and implementation of the type of physics-based knowledge included in the PGML model. Using a simulation experiment, these strategies achieve useful approximations of the underlying true stability model while reducing the number of experimental measurements required for model update.

Understanding the Origin of the Low Cure Shrinkage of Polybenzoxazine Resin by Computational Simulation

Thermoset resin-based composite materials are widely used in the aerospace industry, mainly due to their high stiffness-to-weight and strength-to-weight ratios. A major issue with the use of thermoset resins in fiber composites is the process-induced residual stresses that are formed from resin chemical shrinkage during the curing process. These residual stresses within the composite material ultimately result in reduced durability and residual deformations of the final product. Polybenzoxazine (PBZ) polymer resins have demonstrated near-zero volumetric shrinkage during the curing process. Although the low shrinkage of PBZ is promising in terms of reduced process-induced residual stresses, little is known about the physical causes. In this work, Molecular Dynamics (MD) simulations are performed with a reactive force field to predict physical properties (gelation point, evolution of network, mass density, volumetric shrinkage) and mechanical properties (Bulk modulus, Shear modulus, Young's modulus, Poisson's ratio, Yield strength) as a function of crosslinking density. The MD modeling procedure is validated herein using experimental measurements of the modeled PBZ resin. The results of this study are used to provide a physical understanding of the zero-shrinkage phenomenon of PBZ. This information is also a critical input to future process modeling efforts for PBZ composites.

The effect of spindle dynamics on tool-tip radial throw in micromachining

Abstract In this work, we present a model to predict the speed-dependent radial throw at the tool-tip during micromachining with ultra-high-speed (UHS) spindles. This speed-dependent nature of the radial throw arises from the interaction between the quasi-static radial throw (tool attachment errors, tool geometric errors, and spindle error motions) and the dynamic response of the tool-collet-spindle system. The radial throw causes the cutting edge trajectory to deviate from the ideal trajectory, critically affecting the attainable dimensional accuracy and surface quality, as well as the micromachining forces. Hence, accurate determination of radial throw at the micro-tool tip is important for both practical applications and process-modeling efforts. In the current work, the proposed model describes the radial throw of the tool-axis as a dynamic response to excitation from rotating unbalance of the spindle assembly. The model parameters are first calibrated using experimentally obtained spindle dynamics and the radial throw measurements, both at two speeds. The calibrated model is then used to predict the radial throw for any spindle speed. The presented model is used to predict radial throw at two different axial locations (microtool's shaft at 2 mm and its tool-tip at 15 mm). For both the axial locations, the spindle dynamics measured at 2 mm is used for model calibration. The average error is observed to be less than 2.4% at the tool-tip (15 mm). It is concluded that the speed-dependent spindle dynamics can be used in an analytical formulation to determine tool-tip radial throw at any speed.

Radial Throw at the Cutting Edges of Micro-Tools When Using Ultra-High-Speed Micromachining Spindles

In this paper, we present an approach to determine the trajectory of the cutting edges of a micro-tool in the presence of radial throw when using ultra-high-speed (UHS) micromachining spindles. When micro tools are rotated, inaccuracies and dynamic response of the tool-collet-spindle assembly cause the microtool axis to displace radially from the average axis of rotation. This displacement is referred to as the radial throw of the axis of rotation. Unlike runout, which defines the total radial displacement of a surface within one cycle of rotation, radial throw depends on the angle of rotation (and thus, the time). The radial throw reflected at each cutting edge causes the cutting edge trajectory to vary from the ideal trajectory, which is a circle with a diameter equal to the tool diameter. As such, the radial throw at the cutting edges critically affects the attainable dimensional accuracy and surface quality, as well as the micromachining forces. For this reason, accurate determination of effective radial throw of micro-tools is important for practical applications and process-modeling efforts. This paper devises an approach for accurate measurement of radial throw at the cutting edges of micro-tools when using UHS spindles. The approach involves (1) a mathematical framework to determine the trajectory of the cutting edges from measurements of the radial throw at two locations on the tool shank, and (2) an experimental method, based on non-contact displacement measurements, for simultaneous radial throw measurements on the tool shank at two mutually-perpendicular radial directions. The approach is experimentally validated by comparing the predicted and measured radial throw at the tool tip using a micro-tool blank rotated on an UHS spindle at 120,000 rpm. The results indicate that the deviations between the predicted and measured radial throw magnitude and orientation are less than 1% and 0.1 deg., respectively. Subsequently, the validated approach is used to determine tool-tip radial throw of a commercially-available micro-endmill rotated on the same spindle at different spindle speeds. It is concluded that the presented approach provides an effective means for accurate determination of the speed-dependent radial throw at the cutting edges of micro tools towards enhancing the process performance.

Self-Assembly Kinetics of Microscale Components: A Parametric Evaluation

At the microscale, assembly by grasp and release is challenging. Self-assembly (SA) is an alternative component assembly method to significantly reduce assembly equipment costs. However, successful application of SA requires high assembly rates and high yield (few process errors). Prior modeling efforts describe SA process performance (yield and rate) for specific system configurations, but they do not use measurable process parameters and provide limited ability to predict the impact of process changes. In this paper, an experimental SA system was designed that controls process parameters independently while measuring SA outcomes. This system is used to evaluate a simple parameterization model for SA rate. The travel direction of the parts is varied to measure the path dependence of the assembly probability in the limit of low-impact velocity. Moreover, effects from changing part geometry are evaluated and accounted for in the model. Experimental results show a strong part-geometry dependence and minimal dependence on part arrival angle. This information is a key step toward a parametric kinetic model of capillary SA and complements previous SA process modeling efforts. [2014-0009]

A Structural and Behavioral Analysis Approach for Process Model Evaluation

Manufacturing of process driven business applications can be supported by process modeling efforts in order to link the gap between business requirements and system conditions. However, deviating purposes of business process modeling inventiveness have led to considerable problems of aligning related models at distinct abstract levels and distinct outlooks. Verifying the consistency of such related models is a big challenge for process modeling theory and practice. Our contribution is a concept called behavioral profile that sum up the fundamental behavioral limits of a process model. We show that these outlines can be calculated effectively, i.e., in cubic time for sound free-choice Petri nets w.r.t. their number of places and changeovers. In addition to the above Support Vector Machines (SVM) usage is helpful to improve consistency with greater confidence to evaluate behavioral and structural consistency.

Group Consensus in Business Process Modeling

Consensus is an important measure for the success of any business process modeling effort. Although intensively studied in the general literature on group processes, consensus has hardly been considered in business process modeling and never seriously measured. The author defines consensus as the level of agreement of group members’ views on the process and introduce business process similarity as a proxy. The author validates the measure by comparing it to an existing self-reported measure of consensus. The author then uses this measure to inform and guide the process of modeling.

Efficient Consistency Measurement Based on Behavioral Profiles of Process Models

Engineering of process-driven business applications can be supported by process modeling efforts in order to bridge the gap between business requirements and system specifications. However, diverging purposes of business process modeling initiatives have led to significant problems in aligning related models at different abstract levels and different perspectives. Checking the consistency of such corresponding models is a major challenge for process modeling theory and practice. In this paper, we take the inappropriateness of existing strict notions of behavioral equivalence as a starting point. Our contribution is a concept called behavioral profile that captures the essential behavioral constraints of a process model. We show that these profiles can be computed efficiently, i.e., in cubic time for sound free-choice Petri nets w.r.t. their number of places and transitions. We use behavioral profiles for the definition of a formal notion of consistency which is less sensitive to model projections than common criteria of behavioral equivalence and allows for quantifying deviation in a metric way. The derivation of behavioral profiles and the calculation of a degree of consistency have been implemented to demonstrate the applicability of our approach. We also report the findings from checking consistency between partially overlapping models of the SAP reference model.

Reengineering of design and manufacturing processes

The development of IDEF tools for modeling and analysis of processes has been motivated by the desire to increase productivity by improving communication and structure of manufacturing systems. Constructing an IDEF model is only one component of a comprehensive process modeling effort. Representing IDEF models as process graphs, performing observational model analysis, and analysis of IDEF model structure are issues addressed in this paper. The paper includes a review of current approaches to IDEF modeling in industry, as well as techniques for analysis of IDEF models. The fundamentals of IDEF0 and IDEF3 are discussed, however, the emphasis of the paper is on model analysis and reengineering design and manufacturing processes. An algorithm for model analysis is presented.