Influence of chromium content on the dry machining performance of cathodic arc evaporated TiA1N coatings

Physical vapour deposition (PVD) titanium aluminium nitride coated cutting tools are used extensively in global manufacturing for reducing production costs and improving productivity in a number of aggressive metal-cutting operations, namely, dry and high-speed machining. In this investigation, the performance of Ti1−xAlxN and Ti1−x−yAlxCryN coatings was assessed on Co-HSS twist drills used to machine grey cast iron. The failure criterion for drills was defined as a critical sized flank wear land at the outer corners of the drills. Using this criterion, the average tool life of uncoated twist drills was increased by factors of 2.5, 3.0 and 3.0 by Ti0.59Al0.41N, Ti0.27Al0.19Cr0.54N and Ti0.21Al0.14Cr0.65N coatings, respectively. Notwithstanding the similar increase in average tool life, the Ti1−x−yAlxCryN coatings produced more consistent results than the Ti1−xAlxN coated drills with standard deviations of 67, 3 and 19 holes, respectively. This result has significant practical implications in manufacturing, since drills are not replaced on an individual basis, but rather on a preset tool change frequency. The present paper discusses the performance of Ti1−xAlxN and Ti1−x−yAlxCryN coated drills in terms of average and practical drill life and concludes with remarks on the characterisation of PVD coatings and their significance on the performance of Co-HSS twist drills when dry machining grey cast iron.

Improved support vector machine generalization using normalized input space

Data pre-processing always plays a key role in learning algorithm performance. In this research we consider data pre-processing by normalization for Support Vector Machines (SVMs). We examine the normalization affect across 112 classification problems with SVM using the rbf kernel. We observe a significant classification improvement due to normalization. Finally we suggest a rule based method to find when normalization is necessary for a specific classification problem. The best normalization method is also automatically selected by SVM itself.

Nonlinear discrimination using support vector machine

Appropriate training data always play an important role in constructing an efficient classifier to solve the data mining classification problem. Support Vector Machine (SVM) is a comparatively new approach in constructing a model/classifier for data analysis, based on Statistical Learning Theory (SLT). SVM utilizes a transformation of the basic constrained optimization problem compared to that of a quadratic programming method, which can be solved parsimoniously through standard methods. Our research focuses on SVM to classify a number of different sizes of data sets. We found SVM to perform well in the case of discrimination compared to some other existing popular classifiers.

Press shop machine analysis and trending

Historically downtime data collection and reporting systems in many automotive body panel press shops has been somewhat adhoc. The impetus for this study stems from frustration in respect of how this data is collected, assessed for trends and presented. Ideally this data should be used to identify costly repetitious faults for actioning of maintenance work and for feedback to tool design for consideration when designing new parts.

Presently this data is stored largely in the form of tacit knowledge by press shop operators; the encumbrance of transferring such information being that there is very often only limited channels to quantify it into something more tangible. Findings show that there tend to be two related obstacles to plant data recording. The first is that automation of down time data collection alone cannot determine fault causes as the majority of press shop events are initiated primarily from operator observation. The second is that excessive subjective operator input can often result in confusion and end up taking greater time in recording than remedying the actual fault.

This Paper presents the development of a system that through press mounted touchscreens encourages basic subjective operator input and relates this with basic objective data such as timekeeping. In this way all responses for a given press line become valuable and can be trended and placed in a hierarchy based on their percentage contribution to downtime or statistical importance. This then is capable of statistically alerting maintenance, line flow and/or toolbuild areas as to what issues require their most urgent attention.

Design of minimal-order stabilisers for multi-machine unstable power systems

A reduced dynamics stabiliser for multi-machine power systems is presented in this paper. The design of the stabiliser is based on the theory of linear functional observers and the solution of a simple parameter optimisation problem. The order of the stabiliser could be as low as the number of machines in the system. The design is applied to an open-loop unstable multi-machine power system.

The reality of quality systems software : working for the machine, or smart people around the machine?

This case study-based research examines the outcomes of using quality system software (QSS) in terms of the support it provides to, and the constraints it imposes on, its users. The results indicate that QSS can help facilitate knowledge access and sharing, and to a lesser extent, facilitate communication. It can be an effective mechanism for initiating and managing changes to work processes. Although some writers have opined that information technology (IT) is another way of controlling workers, this was not evidenced in this study. Technical shortcomings did hamper performance of some tasks, and for some managers, extra work was involved.

Review of applications of the Cutting Angle methods

The theory of abstract convexity provides us with the necessary tools for building accurate one-sided approximations of functions. Cutting angle methods have recently emerged as a tool for global optimization of families of abstract convex functions. Their applicability have been subsequently extended to other problems, such as scattered data interpolation. This paper reviews three different applications of cutting angle methods, namely global optimization, generation of nonuniform random variates and multivatiate interpolation.

Machine to die for: the quest for free energy

This is about the search for 'perpetual motion' and 'free energy'. Conventional science claims that it is impossible, yet generations of inventors have been mesmerised by the promise of an engine that powers itself. The world's reliance on diminishing fossil fuel resources and the associated problems of pollution serve to spur them on. It showcases a number of dedicated, sometimes eccentric, and always obsessive individuals who have devoted their lives to this quest.

Dehairing Australian alpaca fibres with a cashmere dehairing machine

Many classes of alpaca fibres contain a certain amount of coarse fibres, which are strong and stiff, and cause discomfort to the end users of the alpaca fibre products. It is therefore desirable to separate the coarse fibres from the fine alpaca fibres. This paper reports trial results on alpaca dehairing using a cashmere dehairing machine. The diameters of alpaca fleece, dehaired alpaca fibres and removed alpaca fibres were analysed, and the fibre lengths before and after dehairing have been compared. The results indicate that it is feasible to dehair alpaca fibres using a cashmere dehairing facility. The dehaired alpaca fibres are cleaner, bulkier and softer, with around 1.5 μm reduction in average fibre diameter, but the dehairing process shortens the dehaired fibre length considerably. The dehairing effectiveness of coarse fibre removal using the cashmere dehairing technology has also been discussed in this paper.

Extended cutting angle method of global optimization

Methods of Lipschitz optimization allow one to find and confirm the global minimum of multivariate Lipschitz functions using a finite number of function evaluations. This paper extends the Cutting Angle method, in which the optimization problem is solved by building a sequence of piecewise linear underestimates of the objective function. We use a more flexible set of support functions, which yields a better underestimate of a Lipschitz objective function. An efficient algorithm for enumeration of all local minima of the underestimate is presented, along with the results of numerical experiments. One dimensional Pijavski-Shubert method arises as a special case of the proposed approach.