Rogelio L. Hecker

Machining Process Monitoring and Control: The State-Of-The-Art

Research in automating the process level of machining operations has been conducted, in both academia and industry, over the past few decades. This work is motivated by a strong belief that research in this area will provide increased productivity, improved part quality, reduced costs, and relaxed machine design constraints. The basis for this belief is twofold. First, machining process automation can be applied to both large batch production environments and small batch jobs. Second, process automation can autonomously tune machine parameters (feed, speed, depth of cut, etc.) on-line and off-line to substantially increase the machine tools performance in terms of part tolerances and surface finish, operation cycle time, etc. Process automation holds the promise of bridging the gap between product design and process planning, while reaching beyond the capability of a human operator. The success of manufacturing process automation hinges primarily on the effectiveness of the process monitoring and control systems. This paper discusses the evolution of machining process monitoring and control technologies and conducts an in-depth review of the state-of-the-art of these technologies over the past decade. The research in each area is highlighted with experimental and simulation examples. Open architecture software platforms that provide the means to implement process monitoring and control systems are also reviewed. The impact, industrial realization, and future trends of machining process monitoring and control technologies are also discussed.

Predictive modeling of surface roughness in grinding

Abstract The surface roughness is a variable often used to describe the quality of ground surfaces as well as to evaluate the competitiveness of the overall grinding system. This paper presents the prediction of the arithmetic mean surface roughness based on a probabilistic undeformed chip thickness model. The model expresses the ground finish as a function of the wheel microstructure, the process kinematic conditions, and the material properties. The analysis includes a geometrical analysis of the grooves left on the surface by ideal conic grains. The material properties and the wheel microstructure are considered in the surface roughness prediction through the chip thickness model. A simple expression that relates the surface roughness with the chip thickness was found, which was verified using experimental data from cylindrical grinding.

Analysis of Wheel Topography and Grit Force for Grinding Process Modeling

Abstract The topography of grinding wheels and the forces per active grain in grinding provide the basic understanding of grinding edges and workpiece interaction. This understanding is important to the modeling, planning, and optimization of the grinding process as a whole. This paper presents a 3-D analysis of the grinding wheel topography to evaluate static parameters of the wheel such as the effective grain diameter and the static grain density as function of the radial distance from the wheel surface. To model the force at each grain, the dynamic grain density is deduced from the static grain density while considering the kinematic effects such as the shadows generated by active grains and the dynamic effects such as the local grain deflection. These effects were evaluated analytically using the grain depth of engagement and the normal force developed per grain. To calculate these effects, a probabilistic model that estimates the undeformed chip thickness was used. This model was calibrated and validated based on experimental data of total normal and tangential forces in surface grinding.

Machining Process Monitoring and Control: The State–of–the–Art

Automation at the process level for machining operations and machine tools has been a focus of research attention in both academia and industry alike for several decades. Research in this area has carried strong expectations in the context of increased productivity, improved part quality, reduced costs, and relaxed part design constraints. The basis for these expectations is two-fold. First, machining process automation, if exercised strategically and advantageously, can perform consistently for large batch production or flexibly for small batch jobs. Secondly, process automation can be set up to autonomously tune the machine parameters (feed, speed, depth of cut, etc.) in pursuit of desirable performance (tolerance, finish, cycle time, etc.), thereby bridging the gap between product design and process planning while reaching beyond the human operators’ capability. The success of manufacturing process automation hinges primarily on the effectiveness of process monitoring and control systems. This paper reviews the evolution and the state of the art of machining process monitoring and control technologies. Key issues to be presented include sensor techniques, control techniques, hardware availability, and implementation examples. Also to be reviewed are the benefits of the systems and the reasons for their delayed realization in many of today’s industrial application domains.Copyright

Editorial: A brief overview of artificial intelligence applications in machining

A brief overview of the main applications of AI in machining is carried out. This does not claim to be and exhaustive review but a simple outline of current state-of-the art and future trend in this branch. In accordance with this, only review papers or very representative and recent works are cited.

Deterministic robust control design for an iron core linear drive including second-order electrical dynamics

This work proposes a deterministic robust controller to improve tracking performance for a linear motor, taking into account the electrical dynamics imposed by a commercial current controller. The design is split in two parts by means of the backstepping technique, in which the first part corresponds to a typical deterministic robust controller, neglecting the electrical dynamics. In the second part, a second-order electrical dynamics is considered using a particular state transformation. There, the proposed control law is composed of a term to compensate the known part of the model and a robust control term to impose a bound on the effect of uncertainties on tracking error. Stability and boundedness results for the complete controller are given. To this effect, a general result on boundedness and stability of nonlinear systems with conditionally bounded state variables is derived first. Finally, experimental results for the complete controller show an improvement on tracking error of up to 31.7% when compared with the results from the typical controller that neglects the electrical dynamics.

Single-State Friction Model for Control Purposes

Friction is a complex phenomenon which has negative effects on the precision of positioning systems. Therefore, friction models for use in friction compensation have to show a good performance while remaining simple enough for its use in control algorithms. In this work, a single-state friction model for control purposes is proposed, aiming to simplify the implementation of such control algorithms, while keeping a good performance in friction compensation. This model has been tested on an experimental setup based on a linear motor, and its performance is compared to that of the Generalized Maxwell-slip (GMS) model, showing better performance on the tracking error for low accelerations.

Mechatronic Sizing of Ball-Screw Feed Drives

This work proposes a procedure to sizing a ball-screw feed drive considering the closed-loop performance. This is achieved by additional steps to the traditional mechanical sizing, where a high frequency model is used to tune the controller in order to achieve adequate stability margins. After that, the tracking error is evaluated for different screw candidates, under nominal and non-nominal operational conditions for a high speed machine tool. Finally, the optimum screw assemble is selected based on minimum torque requirement. As a main conclusion, the screw lead is identified as a high influencing factor in the system dynamics and then on the closed-loop performance, showing the need of an integrated design.

Cylindrical Grinding Cycle Design Based on Final Part Quality Constraints

A grinding process model to predict the forces and power based on a probabilistic analysis of the chip thickness is presented. The model includes tool properties, material properties, kinematic conditions, and dynamic effects. A surface roughness model based on the predicted chip thickness is formulated and validated with empirical data. A grinding cycle is proposed by combining power control during the stock removal stage and feed control during the finishing stage. The grinding power model and the surface roughness model were used to simulate the process and to find the optimal settings for the cycle. The desired grinding power and the workpiece rotational velocity for stock removal were selected to not exceed the machine capabilities and to prevent workpiece burn. The feed and the workpiece tangential velocity of the finishing stage were found based on the final part requirements such as surface roughness and the theoretical out-of-roundness. This grinding cycle was implemented and tested in an open architecture machine.Copyright

Power feedback control in cylindrical plunge grinding with an inner repetitive position control loop

This paper describes the design of a power controller for cylindrical plunge grinding, which controls the power consumed by the grinding wheel by regulating the plunge feed. A mathematical model relating the power to the plunge feed has also been developed and quantified herein with machining data. Based on this model, a proportional plus integral (PI) controller has been designed and tuned to fulfill time response specifications. An inner position loop was also designed and implemented inside the closed power loop to guarantee a stable power response. The commanded position is generated by sending small steps for the servo to follow. In this way, the stiction can be treated as a perturbation with constant period, and its effect can be mitigated using repetitive controller - a subclass of learning control. Both controllers were implemented and tested on an open architecture cylindrical-grinding machine. The repetitive controller was shown to be effective by repeated learning to eliminate the periodic friction effect. The grinding controller show a robust response to changes in the width of cut, by adjusting the feed to maintain constant grinding power.