Tool Control Research Articles

Noise Injection Techniques to Expose Subtle and Unintended Message Races

Debugging intermittently occurring bugs within MPI applications is challenging, and message races, a condition in which two or more sends race to match with a receive, are one of the common root causes. Many debugging tools have been proposed to help programmers resolve them, but their runtime interference perturbs the timing such that subtle races often cannot be reproduced with debugging tools. We present novel noise injection techniques to expose message races even under a tool's control. We first formalize this race problem in the context of non-deterministic parallel applications and use this analysis to determine an effective noise-injection strategy to uncover them. We codified these techniques in NINJA (Noise INJection Agent) that exposes these races without modification to the application. Our evaluations on synthetic cases as well as a real-world bug in Hypre-2.10.1 show that NINJA significantly helps expose races.

A Comparative Study of CNC Part Programming Addressing Energy Consumption and Productivity

This paper reports on a comparative study realized to contrast between the part code generated by two different CAM systems and one part program generated by the authors. This comparison addresses in particular tool path, but also includes number of lines, cycle time and energy consumption. The comparison is carried out through a case study: a mechanical component was designed so as to produce a number of CNC part programs generated with two different CAM systems; one more program was directly programmed by the authors in the machine tool's controller. Different sample parts were produced using the same machine tool, tooling and process parameters; and measurements of energy consumption and cycle time were gathered. The results show differences between the tool paths generated by each approach and are used by the authors to further define strategies to produce trajectories that could reduce cycle time and energy consumption.

Machine Availability Monitoring for Adaptive Holistic Scheduling: A Conceptual Framework for Mass Customization

Although the market started expressing the need for high product variety three decades ago, manufacturing systems have not kept the pace with these social and economic developments. Modern industrial shop-floors are highly affected by the ever-increasing product variety and volatile market demands introduced by the currently established mass customization paradigm. The increasing complexity and high unpredictability in manufacturing activities require immediate reactions to emerging disturbances, such us reducing bottlenecks, and avoiding brake-downs and idleness of critical equipment in order to increase productivity and therefore, the company's competitiveness. The framework proposed in this research work includes machine monitoring techniques for the near real-time identification of machine status, in order to allow a predictive maintenance engine to diminish machine tool failures. Moreover, an adaptive short-term scheduling mechanism is employed, using monitoring data for the refinement of production schedules based on the current and future conditions of the shop-floor. Data acquired from a multi- sensory pilot installation together with information directly retrieved from the machine tool's controller are fused and comprise the input to subsequent software modules. The monitoring module is described, which analyses sensory data during the machine operation, as well as the Adaptive Scheduling Engine that dispatches schedules and refines future schedules based on awareness regarding machine availability. The software architecture and hardware setups are described and comprise the roadmap for the framework development. A possible application of the framework is described in the context of a European SME manufacturer.

Development and Optimization of the Die-Sinking EDM-Technology for Machining the Nickel-based Alloy MAR-M247 for Turbine Components

In this paper, the results regarding the optimization of the die-sinking EDM-technology for the fabrication of turbine components are presented. The main objective of this work was the reduction of machining time while producing cavities with a depth of 11mm in the alloy MAR-M247, with respect to the requirements for surface roughness and sub-surface damage. The study was conducted using graphite electrodes of three different dimensions, totalizing 12 electrodes and a total electrode area of 89.50 mm2. The machine tool GENIUS 1000 THE CUBE was used in the experiments and IonoPlus IME-MH was used as dielectric fluid. The Design of Experiments Methods (DoE) were applied in the planning and execution of all experiments. Firstly, experiments were conducted to estimate the maximal discharge current that can be applied for fulfilling the quality criteria regarding both surface roughness and thickness of affected sub-surface, since discharge current is known as the main influencing process parameter affecting both factors. In the second stage, the process was optimized by varying following electrical parameters of the machine tool: discharge current (ie), discharge duration (ti), pause duration (t0), ignition voltage (V) and duty factor (τ). Finally, both process parameters as well as machine tool's control parameters were adjusted to achieve optimized results regarding the total machining time and electrode wear with respect to the surface requirements of produced parts. A total machining time reduction of more than 50% in comparison to the technology currently employed by the industry was achieved. Requirements on both arithmetical surface roughness as well as affected sub-surface thickness could be fulfilled. This robust technology was developed for producing the cavities in MAR-M247 and all applied methods, studied parameters as well as the results regarding machining time, electrode wear and the quality of the achieved work piece are presented.