Tool and process monitoring-state of art and future prospects

Quest for process automation driven by growing costs of human labour and quality demands makes monitoring in manufacturing systems inevitable. Although numerous tool and process condition monitoring systems are now available in the market and many
of 23
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  tool monitoring, sensor, signal processing, acoustic emission, cutting force   Krzysztof JEMIELNIAK* Jan KOSMOL**   TOOL AND PROCESS MONITORING - STATE OF ART AND FUTURE PROSPECTS Quest for process automation driven by growing costs of human labour and quality demands makes monitoring in manufacturing systems inevitable. Although numerous tool and process condition monitoring systems are now available in the market and many have been installed in industry, users generally still consider them unreliable, often not worth money they cost. The bulk of the paper is centred on reasons of that defeat and measures undertaken nowadays to improve TCM/PCM systems reliability. First, the major tasks and general structure of the tool and process condition monitoring systems are presented. Then all basic elements of the monitoring systems: sensors, signal processing, feature extraction, and strategies were reviewed in terms of hitherto drawbacks and ongoing research works. The paper does not pretend to give a complete review of existing systems. Only examples illustrating discussed problems are quoted here. 1. INTRODUCTION Rising labour cost makes production automation an important priority in the major industrial countries. One of the most important factors limiting the progress in introduction of unattended machine tool operation is tool condition monitoring (TCM) and process condition monitoring (PCM). The main focus areas of TCM/PCM are: • tool condition monitoring: • tool wear monitor (detecting end of tool life) • catastrophic tool failure (CTF) detection • chip-breaking detection • chatter detection • others (eg detection of BUE formation, burr formation, collision),  Numerous different phenomena can be employed for monitoring and a variety of sensor types are available on the market. In Fig. 1 numbers of recent research publications on sensing and sensor systems [ 5 ] are shown. The numbers express research activity level in the field of tool and process monitoring in machining. It shows that the bulk of the activity is in tool wear monitoring, and tool fracture detection. Quantities most often * Warsaw University of Technology, ul. Narbutta 86, 02-524 Warsaw, Poland ** Silesian Technical University, ul. Konarskiego 18A, 44-100 Gliwice, Poland  employed in TCM/PCM monitoring is acoustic emission and cutting force components or measured variables derived from these components (extension of machine elements,  bending/displacement of the tools, torque, drive power motor current, etc.). Vibration and noise are also the focus of research work and industrial applications. Other phenomena are used occasionally, mainly in laboratories. Fig. 1. Categorisation of sensor research and development [5] Tool and process monitoring has been subject of intensive research work for several years and numerous commercial monitoring systems are available on the market.  Nevertheless it is wide recognised nowadays, that the systems do not meet shop floor requirements having too many drawbacks. In the users’ opinion [6, 12], system manufacturers give very optimistic promises (recommendations) which are not fulfilled in  practice, so the systems are often switched off after one year. Since the price of additional sensor systems is high, and its reliability is still inadequate, machine manufacturers and users are reluctant to pay the cost. Main deficits in the field of tool and process monitoring can be summarised as follows: • There is no single sensor or sensor system capable of covering all or even the majority of the possible applications. • Some of the sensors used in research work are not designed for the tough machine tool environment. This is particularly true for cutting force dynamometers and AE sensors, which have often been developed for non-destructive materials testing. • Detecting end of tool life is especially difficult, and is generally possible only after teach-in. • Most of the research results have been obtained off-line. Monitoring systems working in real time are still uncommon. There has still been no real-time realisation of some process and tool monitoring methods. • There is no effective combination of monitoring units with CNC controls, lack of interface specification, standardisation.   Fig. 2. Structure of TCM/PCM system. Before discussing the efforts undertaken to overcome these deficiencies, let us look closer at the structure of the tool and process monitoring system (Fig. 2). The cutting  process can be characterised by a variety of  physical quantities. Appropriate sensors transform the selected ones into corresponding electrical quantities (signals), which can be electronically processed and transmitted. The signal processing can be more or less complex, consisting of eg filtering (LPF, HPF, BPF), A/C conversion, FFT, RMS, wave form conditioning, standard deviation, mean value, skew, kurtosis, crossing rate, regression analysis and many others. As a result, signal representation (feature vector), sensitive to the  parameters of interest in the process, is extracted from the signal. Based on this representation and a suitable strategy, a decision of monitored process state is generated. The strategy itself is developed  basing on knowledge and experience contained in the model of the process . Deficiencies of existing monitoring systems and recent trends of their development aiming at meeting users' demands, are connected with all cells of this structure. That is how they will be presented here. 2. SENSORS 2.1. FORCE RELATED QUANTITIES Cutting force components or variables derived from these components are quantities most commonly employed in industrial TCM/PCM systems. Therefore a lot of sensors are available on the market. System producers and users have to face fundamental choice being compromise between two contradictory demands. On the one hand high accuracy both static and dynamic is required. It means that sensor should be as close to the machining point as  possible. On the other hand, sensor should be easy to retrofit without major changes in the machine tool construction and with no reduction in the static and dynamic stiffness of the machine tool. In early stages of TCM/PCM development producers of machine tools were eager to ensure their customers that machines they offered were equipped with monitoring systems. Therefore power or motor current measurement was quite popular those days  [17÷20]. The ring sensor measuring changes in the electrical current supplied to feed or spindle drive motors or servos (Fig. 3) is still available on the market. A single conductor from the power cable supplying the feed motor is fed through the current sensing ring. The current carried by the conductor passing through the sensor ring is transformed into a voltage signal proportional to the feed force [26]. It can be easily installed into almost all types of machine tools without any special engineering design and provide a comparatively low-cost yet effective (according to the producer) tool monitoring system in new or existing machines. Fig. 3 Current sensor [26]. It should be stressed however, that because of the long distance between the machining point and the sensor, a signal obtained from any motor current sensor is time lagged and has low sensitivity. Moreover, if the current supplied to a spindle drive motor is monitored in the case of turning and milling, the system sensitivity is even lower due to weak main cutting force dependence on tool wear (see below, point 4.2). Systems based on such sensors can hardly be recognised as successful [20].  Fig. 4. Piezoelectric strain transducer [29] Other example of the sensor that is easy to retrofit, requiring no special design work for adaptation and not reducing stiffness of the machine tool is the piezoelectric strain transducer shown in Fig. 4 [29]. This sensor detects the cutting force via the extension of force-carrying machine elements. Despite easy retrofitting, the suitable fitting position for the sensor can only be determined by time-consuming experimentation. The sensor was used in some commercially available systems [16, 30].   Both sensors described above possess a low level of sensitivity and are suitable only for major catastrophic tool failure identification during rough machining. Fig. 5. Feed Force Sensor: a) internal view, b) sensors fitted in the feed screws of an NCH turret lathe [24]. Fig. 6. Signals from feed force sensors [11]. Feed force sensor, suited for integration with a bearing pocket supporting a rotating shaft or spindle (Fig. 5, [24]), can be an example of a compromising design. The sensor consists of two concentric rings (Fig. 5a). The  profile of the inner ring has a special form, providing two force sensing zones, on which strain gauges are mounted. Despite much  better correlation between the actual feed force and the signal in this case, still significant inaccuracies should be taken into account [11]. Fig. 6a presents atypical feed force signal (Ff ) during air travel with programmed feed rate at the beginning of operation. Characteristic sinusoidal shape of the signal generates the feed screw bearing itself, without any relation to the real cutting force, which is equal to zero. Such sinusoidal changes in the signal can be also observed during longitudinal cutting with constant cutting  parameters when real cutting force is approximately constant (Fig. 6b). Moreover the average value of the signal slowly increases due to cumulation of the stresses in the kinematic chain, especially on the slideways. The best cutting force measurements can be achieved with force transducers placed close to the cutting edge, directly in the path of the transmitted force [2, 31, 35].
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!