quently and more often than in conventional cutting (Figure 5.2-4). Wear may be critical and therefore wear sensors are of interest. They should be able to deter- mine the end of tool life reliably. There are several approaches, as discussed in Section 4.3. One of the main accuracy problems with automated machine tools is derived from the thermal stability. The temperature field in the machine structure changes according to the effect of several heat sources. The most important heat sources are very often the spindle bearings. The monitoring of the bearing tem- perature is recommended because of the high investment that an HSC machine represents and the critical power losses with high spindle frequencies. The heat- ing of fast-running main spindles can lead to an unstable state: heating increases the pre-stressing of the spindle-bearing system, which increases power losses and heating, etc. Monitoring is therefore advisable. This can be done fairly easily and reliably by thermocouple sensors. Similar measuring devices might be advisable to monitor feed drive components such as spindle-nut systems and direct linear drives to ensure a tolerable increase in temperature. 5.3 Micro-machining M. Weck, RWTH Aachen, Aachen, Germany The manufacture of micro-components using high-precision machine tools, so- called ultra-precision machines, imposes new demands on integrated sensor sys- tems. In micro-machining, extremely filigree turning, planing, or milling tools are frequently used. In addition, the machining forces are very low, typically in the range below 1 N when natural diamond tools are used. Very few sensor systems meet the requirements for micro-machining. To deter- mine process forces, piezoelectric force sensors with very high resolutions have to be used to generate a useful measurement signal. Attempts have been made to 357 Fig. 5.2-4 Tool life with high cutting speeds. Source: kindly provided by B. Denkena, University of Hannover Sensors in Manufacturing. Edited by H.K. Tönshoff, I. Inasaki Copyright © 2001 Wiley-VCH Verlag GmbH ISBNs: 3-527-29558-5 (Hardcover); 3-527-60002-7 (Electronic) supervise the cutting process via acoustic emission (AE) systems. However, only the tool-workpiece contact can be determined precisely with these methods. An in-process measurement with AE systems has not been developed successfully. One of the most needed measurement systems for micro-machining are tool measurement systems. If the tool which is clamped in is to be oriented precisely, it is vital that the relative position of the cutting edge to the machine coordinate system, the tool geometry, and the distance between the tool and the workpiece are known. The exact determination of the workpiece geometry permits any cen- tering error in the milling tools, for example, to be corrected. The use of forming tools in micro-turning operations demands that the relative position can be deter- mined precisely and the measurement of the distance between the tool and the workpiece permits defined tool feed motion. The degrees of measuring accuracy required are listed in Table 5.3-1. Non-contact optical measuring techniques are particularly well suited to this ap- plication owing to the small dimensions of the cutting edges of micro-tools and their susceptibility to cutting edge fracture. These techniques entail the use of a charge-coupled device (CCD) camera with a microscope to measure the cutting edge of the diamond tool. The position and geometry of the tool in relation to the machine coordinate system are subsequently measured using image-processing techniques. As an example, a tool measuring system integrating a micro-milling machine is presented below. The three-axis ultra-precision milling machine and the optical tool measure- ment system were designed and built at the IPT in Aachen (Figure 5.3-1). This milling machine is equipped with two air bearing spindles and linear air bearing guides. The y-guide is provided with a high-speed milling spindle, permitting ul- tra-precision milling operations with optical surface quality to be carried out on the machine. To measure the current position, a laser interferometer system is integrated into the linear slides. In combination with high-precision direct current (DC) torque motors and servo systems, a position accuracy in the submicrometer range has been achieved. The geometry of the rotating milling tool is directly measured using an optical system, mounted on the y-slide. To achieve a precise tool measurement, telecentric imaging systems, which per- mit the measuring operation to be performed largely independently of the per- 5 Developments in Manufacturing and Their Influence on Sensors358 Tab. 5.3-1 Precision requirements for tool measuring systems in micro-machining operations Position Angular position in relation to the machine axes < 0.18 Tool distance from the workpiece (z)<1lm Tool distance from the rotational axis (y) < 0.1 lm Tool distance from rotational axis (x) < 0.1 lm Geometry External tool contour (turning) < 1 lm Tool radius (milling) < 1 lm Wear/spalling Qualitative spective, were used (Figure 5.3-2). This was the only way of recording high-preci- sion geometric measurements of three-dimensional objects. The illumination unit used is a telecentric, low-wavelength transillumination (blue spectral region). This minimizes diffraction effects at the tool cutting edge, thus increasing the level of absolute measuring accuracy. Because of the low overall dimensions of the micro- machining tools, a measuring range of 0.5 ´0.5 mm at 100-fold magnification of the optic has proved suitable. If the exacting requirements in terms of evaluation accuracy are to be met, it is not sufficient to use standardized methods of image processing which draw a con- clusion as to the position of the tool on the basis of a straightforward, pixel-orient- ed evaluation of camera images. The geometric resolution is insufficiently high for this. A combined method of pixel and so-called sub-pixel evaluation has there- fore been developed. This technique permits all quantities required to be deter- mined with a resolution lower than 1 lm via high-precision edge detection and exact definition of the tool contour via straight and circular segments. Appropriate illumination is used in the camera image, which is recorded by a very precise, distortion-free lens system, to contrast the object clearly against the background. Once this object image has been recorded, an edge detection opera- tion is conducted. A number of gradient techniques, which exploit the fact that 5.3 Micro-machining 359 Fig. 5.3-1 Ultra-precision milling machine ‘UPM’ (Fraunhofer IPT) there is a considerable difference in the gray scale value at one edge, are suitable. Good results are recorded particularly by the techniques which use Sobel filters and filters with similar characteristics. The outcome of these techniques is a multi-pixel breadth edge in the image, which is thinned out by mass techniques (non-maximum suppression). The remaining curve, which is the breadth of one pixel, is converted by an edge-following algorithm into a series of coordinates re- presenting the corresponding points on the contour. The result is a list of coordi- nates which describe the contour in discrete pixel units. Straightforward evaluation algorithms are stretched to their limits in applica- tions of this nature. They produce a rough grid image of an edge which is actu- ally smooth. Moreover, it is impossible to distinguish between straight lines and circles. In order to achieve a higher degree of accuracy, the contour is approxi- mated in an additional, more extensive step involving the detachment of discrete points. Complex segmentation algorithms can be used to detect corner points. De- pending on the course of the contour between the corner points, these sections are allocated an appropriate geometric form (straight or circular segment). The specific parameters of this form are determined and filed in a table. This table contains, for example, data about the starting and end points, the radius, and the mid-points of circular segments and also the tool angles or the angle which is as- sumed by a straight line in relation to the x-axis. The openness of the data structure makes it possible to prepare it in such a way as to ensure that the data can be read and processed by CAD programs with suitable interfaces. It is a particularly important feature in this context that all in- formation about the contour of the object can be presented with extreme preci- sion, at a resolution < 1 lm. The approach adopted in the tool measuring operation is described in the fol- lowing in greater detail, on the basis of the example of an end milling cutter and 5 Developments in Manufacturing and Their Influence on Sensors360 Fig. 5.3-2 Optical tool measuring system a turning tool. In order to measure a micro-milling cutter, images are recorded continuously while the milling tool rotates at low speed. Each individual image is linked to the result image via minimum formation so as to ensure that after the milling tool has rotated several times, the resultant groove geometry becomes clearly visible (Figure 5.3-3). This is then evaluated with sub-pixel accuracy using an edge detection algorithm. The resolution achieved is < 1 lm. The side walls of the groove are approxi- mated by sections of lines. A geometric evaluation permits the resultant diameter of the groove to be determined subsequently at various distances from the chan- nel base. A centering error of the milling tool which influences the breadth of the groove can be compensated via the machine control system in a later step. Addi- tionally, any angle error which may occur when the milling tool is clamped in, and which results in the course of the machining operation in a V-shaped chan- nel edge geometry, is detected at an early stage. Should this be the case, the tool must be re-clamped and the tool must be measured again. In addition to measuring end milling cutters, the system can also be used to de- termine the tool contour of diamond spherical cutters. When a deviation from the spherical reference contour is recorded by the image processing system, the adju- stable holder can realign the cutting edge on the machine. Additionally, the opti- cal measuring system permits the exact relative position of turning tools in rela- tion to the machine coordinate system to be determined (Figure 5.3-4). This is of particular importance in the case of forming tools since any alignment error re- sults immediately in a deviation of the workpiece from the reference geometry. When the workpiece to be machined is recorded by the camera, the distance be- tween the tool and the workpiece can also be recorded using image processing techniques (Figure 5.3-4). This is of particular significance to the machine operator during the so-called ‘scraping operation’. When the optical technique of measuring the distance be- tween the tool and the workpiece is applied, a reproducible absolute precision of <1 lm is achieved. When the operation is started in the conventional mode, the 5.3 Micro-machining 361 Fig. 5.3-3 Image recording in optical tool measuring operations 1. Camera picture of the non rotating milling tool 5 Developments in Manufacturing and Their Influence on Sensors362 Fig. 5.3-4 Determining the position and geometry of an end milling cutter and the angular posi- tion of a turning tool workpiece is moved at a low feed speed until the first chip formation is visible with the naked eye. The precision and consistency of this feed operation depend heavily on the experience of the machine operator and generally fluctuate by at least 1–2 lm. This, however, is intolerable for demanding applications of the mi- cro-structure technique or for measuring coated sample parts. 5.4 Environmental Awareness F. Klocke, RWTH Aachen, Aachen, Germany Ecological issues are assuming increasing importance in many areas of the econo- my as a result of legislation and growing public awareness. Manufacturing, char- acterized by a chain of resource-intensive processes and by large quantities of waste materials and emissions, is frequently the focus of interest. Government-im- posed environmental regulations and increased cost pressure in conjunction with the need to prevent the production of waste materials or to dispose them appro- priately are forcing companies to introduce innovative, environmentally compati- ble manufacturing processes. In the manufacturing environment, the starting points for ecologically oriented improvement lie in the need to prevent the genera- tion of waste materials and pollutants in the first place, or to reduce the volumes produced and re-use them. The advantages which stand to be gained as a result of the application of more environmentally compatible technologies are clear: re- duced levels of energy consumption, waste, and disposal costs, together with high- er employee motivation and lower rates of absenteeism due to illness [1]. In the successful, practical application of process monitoring systems and com- ponents, monitoring- and sensor-related solutions are adapted to meet the specific requirements of the machining task concerned. This demands precise knowledge of the machining operation, ie, the manufacturing environment, the machining process, and any potential process malfunctions [2–4]. The requirements relating to the monitoring system may, however, differ considerably in terms of the objec- tives and the implementation of the monitoring system. A reduction in the quantity of cooling lubricant used in machining operations is a good example of the specific demands imposed on process monitoring and sensor systems by manufacturing processes which have been optimized in terms of environmental compatibility. On the one hand, the application of sensors is simplified since the requirements relating to the robustness and coolant resis- tance of the sensors are lower in this case. On the other hand, a reduction in the amount of cooling lubricant used frequently increases the degree of thermal load to which the parts are subjected. It therefore becomes more important to monitor any temperature-related change in dimensional and form accuracy or in the struc- ture of the material of the finished parts. The influence exerted on the structure is critical, particularly in machining operations conducted on hardened materials. Demands made on the monitoring system and on measurement engineering, which arise from the specific boundary conditions of environmentally compatible 5.4 Environmental Awareness 363