Mechatronic Systems for Machine Tools

R. Neugebauer1 (1), B. Denkena2 (2), K. Wegener3 (2)1Fraunhofer-Institute for Machine Tools and Forming Technology IWU, Chemnitz, Germany

2Institute of Production Engineering and Machine Tools (IFW), Hannover University, Garbsen, Germany 3Institute of Machine Tools and Manufacturing (IWF), ETH, Zurich, Switzerland

Abstract This paper reviews current developments in mechatronic systems for metal cutting and forming machine tools. The integration of mechatronic modules to the machine tool and their interaction with manufacturing processes are presented. Sample mechatronic components for precision positioning and compensation of static, dynamic and thermal errors are presented as examples. The effect of modular integration of mecha-tronic system on the reconfigurability and reliability of the machine tools is discussed along with intervention strategies during machine tool operation. The performance and functionality aspects are discussed through active and passive intervention methods. A special emphasis was placed on active and passive damping of vibrations through piezo, magnetic and electro-hydraulic actuators. The modular integration of mechatronic components to the machine tool structure, electronic unit and CNC software system is presented. The pa-per concludes with the current research challenges required to expand the application of mechatronics in machine tools and manufacturing systems.

Keywords Machine tool, Mechatronic, Adaptronic

1 INTRODUCTION The technological transformation of the information soci-ety is associated with considerable changes in the de-mands on machine tools, requiring new solutions for the inherent conflict in design between precision and produc-tivity [1][2][3]. Future performance requirements of ma-chine tools will be characterised by trends in microsys-tems technology and nanotechnology. For mechanical manufacturing, this results in constantly growing de-mands for precision in machine tools (Fig. 1). The scien-tific and technical challenge does not lie entirely in the capability to produce mechanical structures with ever greater precision, but also mass produced workpieces at reasonable production costs. Globalisation is putting mechanical production in high-wage countries under immense economic pressure [2]. One fundamental initiative to find a solution in the area of process development is the shortening of process chains by technology substitution. This must be accomplished by increasing the precision of highly productive procedures, since increasing the productivity of high-precision proce-

dures is mostly limited by the underlying physical princi-ples of those procedures. Consequently, it can be deduced that the contradiction posed by the demands for productive precision in large working areas and flexible, reliable manufacturing re-quires a universal system architecture for machine tools with the typical properties of mechatronic systems – con-version capability (reconfigurability) and self-optimisation (immunity to disturbance) [5]. The concrete demands identified from surveys of cus-tomers and manufacturers in this respect [2] are: • An increase in availability as a result of prescient

maintenance; status and process monitoring; intelli-gent maintenance; diagnostic functions.

• Highly economic manufacturing (life-cycle costs) as a result of function-oriented design; improved perform-ance due to dynamics, flexibility as a result of recon-figurability.

These factors are of necessity associated with: • A high level of standardisation and modularisation:

interfaces, technology modules, independent func-tional assemblies.

• Efficient control systems. In recent decades, mechatronics has become established as a developmental method for such systems and today constitutes the key discipline in production technology [6]. The term mechatronics was coined in 1969 by the com-pany Yaskawa Electric Cooperation to signify the exten-sion of the function of mechanical components by means of electronics. As a result of the immense speed of devel-opments in information processing, the aspect of integrat-ing “intelligence” in technical systems in mechanical engi-neering is now increasingly at the forefront. Mechatronic systems are essentially characterised by the function-oriented expansion of a mechanical system by the spatial and/or functional integration of sensors and actuators and the use of a control system to guarantee functionality [7]. The benefits of mechatronics go beyond a mere additive effect, as is emphasised by the definition given in [8]:

Figure 1: Machining capability over time [4].

Annals of the CIRP Vol. 56/2/2007 -657- doi:10.1016/j.cirp.2007.10.007

-658-

“[Mechatronics needs]...a synergetic cross-fertilization between the different engineering disciplines involved: mechanical engineering, control engineering, microelec-tronics and computer science. This is exactly what mechatronics is aiming at; it is a concurrent-engineering view on machine design.” The term “adaptronics” was coined in the 1980s in the USA. Under the umbrella of this term, developments in materials science and engineering were begun which were inspired by bionic construction principles, as in the structure of a human muscle, aimed at producing efficient lightweight structures for air and space travel by means of direct integration of transducer materials acting as sen-sors and actuators in the construction material itself (Fig. 2). Transducer materials which are used include piezo ceramics, shape memory alloys, magnetostrictive materi-als or electromagnetically activated fluids and polymers. By connecting composite components acting as sensors and actuators, using appropriately adapted electronic controls, the structural dynamics properties of the basic mechanical structure can directly be influenced. One ma-jor application area is vibration control in lightweight struc-tures, extending as far as ASAC - Active Structural Acoustic Control [9]. Today, the initiatives to find solutions featuring distributed integration of these materials in con-struction materials to form an active composite material by means of complex mechanical and information tech-nology are only just beginning and are currently the sub-ject of intensive research [10]. The use of actuators made from transducer materials in highly integrated mechatronic components, such as pre-cision positioning systems, is technically more advanced and already employed in industrial applications. If such components use integrated sensors for autonomous im-provement of higher-level mechanical or mechatronic structures, they are defined in the vocabulary of mechani-cal engineering as adaptronic [11]. In future, mechatronics will be fundamentally enriched by optical technologies as a new basic system type. Even today, they are established directly in machine tools via the laser beam, as an alternative machine drive and tool for mechanical processing.

Optical waveguides permit interference-free communica-tion between sensors, information processes and actua-tors with high bandwidths. Sensors based on optical prin-ciples, including industrial image processing, achieve the very highest information density. At the same time, the speed of data processing by optoelectronics will further increase in the future. The following can therefore be identified as driving the further development of mechatronics technology and hence, as a consequence, the development of machine tools:1. Information technology based on Moore’s law [12]. 2. Adaptronics (and the associated manufacturing proc-

esses of microsystem technology) for lightweight con-struction and for increasing the level of integration of the components.

3. Optics as the universal new basic system type. The basic approach of mechatronics is inherently an-chored in the development of machine tools. The elec-tronically controlled movement behaviour of a mechanical structure was first implemented in machine tools in 1952 with the first NC control system made at MIT. Since the 1970s, mechatronics has fundamentally changed the functionality and efficiency of machine tools as a result of the increasing integration of NC movement controls and the automation of processes. In order to be able to make optimum use of the modern technical options offered by mechatronics, modified processes [7] and tools [13] are required for a unified design. These are described in de-tail in Chapter 2. In particular, a change in the design paradigm is needed to prevent the inevitable increase in system complexity due to the increased functionality from mechatronics resulting a priori in a decrease in reliability. Because of the differing scale and bandwidth of machine errors (Fig. 3), it makes sense for the further development of the machine tool mechatronic system to adhere to a hierarchical conceptual framework.

0

100

freq

uenc

y / [

Hz]

transfer errors

300

thermal behaviourgravitation

Calibration

Model basedwith inherent

Actuators

Mechatronic/Adaptronic

Components

Compensation Methods • natural oscillationsof the mechanics

• elastic deformationsby process- andinertial forces

drive errors

natural oscillations of the machine foundation

Figure 3: Bandwidth of machine tool errors and options for intervention.

In doing so, it is possible to distinguish the following sys-tem levels: 1 – The machine tool mechatronic system. Component sensors and the lowest possible number of additional sensors capture an image of process and ma-chine that approximates to reality. With the support of models, the intrinsic drives of the machine tool are used by the control system to correct processing errors. The primary application for which this is suitable today in-volves quasi-static error sources, such as thermally de-termined displacements or calibration of machines. How-ever, the comprehensive data from the sensors also pro-vide initiatives for increasing the reliability and availability of the machine tool. The increasing importance of solu-tions of this kind is a consequence of the high speed of development of CNC technology. They are strongly influ-enced by design and system integration and are detailed in examples in Chapters 2 and 4. 2 – The mechatronic components of the machine tool. These include in particular the main and feed drives. Past research work has concentrated on this system level and has already been analysed and published in detail. These components and their integration in terms of information technology define the capability of the actuator and the bandwidth of the solution initiatives in system level 1. Ma-jor progress can therefore be expected as a result of op-timum system integration (Chapter 4). 3 – The integration of additional mechatronic solutions in machine tool components.

Mechanical Systems Mechatronic Systems Adaptronic Systems

System

Answer

PassiveStructureActuatorsSensors

Sensors Actuators Control

System

AnswerSystem

AnswerPassive

StructurePassive

Structure

Control

Figure 2: Characteristics of mechatronic systems.

-659-

The aim of intervention in this system level is to improve machine tool malfunctions by means of autonomous components which either act directly as ancillary drives to compensate for dimensional errors near to the source of the interference, or to improve the behaviour of the con-trolled process of the basic structure of a machine tool component for the higher-level control system. The ad-vantage of this kind of system intervention lies in a design which is optimally adapted to the local problem and the consequent removal of restrictions with regard to the bandwidth of malfunctions to be compensated. As Plug & Play modules, such solutions are interesting for the im-plementation of reconfigurability. These are the focus of Chapter 3. In relation to the overall function of the machine tool, the above system levels, which are not independent, have both a specific suitability for use and, at the present time, different levels of advancement in application. The opti-mum effect can therefore only be achieved with thorough system integration concepts (Chapter 4), which take into account both the mechanical involvement and the signal and energy flows, but above all the architecture of the data processing. Examples of optimisation criteria con-taining fundamental parameters, including economic pa-rameters, are the Plug & Play capability, the energy effi-ciency and the reliability of the overall solution.

2 SPECIAL DESIGN CHARACTERISTICS 2.1 Engineering Process Mechatronic systems are complex systems whose solu-tion field is interdisciplinary. They show a greater func-tional density and closer of function than traditional me-chanical and electrical systems. In their design, therefore, the functional aspect is clearly at the forefront compared with the purely mechanical/geometrical aspect. The con-sequence of this is that master procedures tailored to the design of mechatronic systems have been setup for the development process and their application leads to a principally usable product. Nevertheless it has been shown in design research that a generally applicable stringent process cannot exist and that the flow diagrams have more of the character of a guideline (VDI 2206 [7], Kallenberg [14], Gausemeier [15], Albers [16]). The underlying process for designing mechatronic systems and the associated system support is shown using VDI 2206. In principle, the mechatronic system consists (as shown in Fig. 4) of a basic system, of mechanical, electrical or any other physical kind, which controls flow of material, energy or even information. It is connected via sensors and actuators to elements for processing information. With the help of sensors state variables of the system can be sampled; the sensors in this case can also be observers, i.e. sensors represented within the software. From the measured values obtained, data processing determines the effects required influenc-ing the status parameters in the desired manner and ac-tuators are employed to exert an effect on the basic sys-tem.A complex mechatronic system can be subdivided into mechatronic modules as shown in Fig. 4. These modules are themselves interconnected via the information proc-essing system. The modularisation facilitates the defini-tion of interfaces and the formation of a product structure, in which the relationships between the modules are less prominent than those within the module. In addition, the

modules are not just arranged on one level but ordered in a hierarchical structure in the sense that several modules are interconnected by means of their functional structure. Examples of hierarchical organisation are monitoring and fault diagnosis functions, the generation of parameters for a series of mechatronic modules and the learning capabil-ity of systems. The integration of mechatronic modules has two aspects to consider, spatial integration and functional integration, which are achieved by means of streams of material, en-ergy and information. The product structuring specified by the integration must above all be capable of reflecting the entire product generation process so that, for example, housing-free integrations are possible, the assembly structure is shown, testable sub-systems ready for instal-lation are created, and installation of replacements is fa-cilitated in the event that repairs are required. Product development and production development must therefore be geared closely to one another. The spatial integration of components which belong together in terms of their function enables, by avoiding housings, the functional density to be increased and savings to be made in terms of weight and assembly space, and where applicable the avoidance of plugs and connections enables an increase in reliability, while at the same time guaranteeing the compatibility of the components in terms of the frequently lethal environmental conditions in machine tool construc-tion.The mechatronic initiative results in a dramatically enlarged range of approach and conversely in a dramatic increase in system complexity. Controlling the latter is a pure bottom up – initiative, i.e. the integration of sepa-rately developed and optimised assemblies is no longer viable. Instead, a prior top-down design is needed, i.e. the design of a gross structure from which, by means of divi-sion into functional elements and an increase in the level of detailing, individual elements can be more accurately specified. A transition to bottom-up design is required as a subsequent iterative step towards system optimisation. Mastering the complexity will require, in addition to a change in approach, more intensive data processing sup-port. For shortening development times, virtual prototyp-ing tools are gaining more and more importance. These use simulation to represent system properties in a model capable of being used for experimentation to gain knowl-edge which is transferable to reality. Simulation moves into a new dimension, since several disciplines are to be covered in a single simulation model.

ACTUATORS SENSORS

forces,torques

disturbances deviations

nominalsignals

MECHANICAL SYSTEM

CONTROL

actualvalues

Sensors Actuators

Control

System

AnswerPassive

Structure

Figure 4: Mechatronic system; mechatronic basic mod-ules.

-660-

The V- model envisages a macrocycle, which based on the requirements of the product development provides for the system design which is subsequently transferred to the now parallelised domain design (Fig. 5). This is fol-lowed by integration, resulting in a (virtual) prototype and the verification and validation of the system design. This cycle can be repeated several times in order to achieve a greater level of maturity through iteration. The design cy-cle is flanked by modelling and simulation. For individual parts of the task, a microcycle is adopted from Systems Engineering [17]. In addition, special process building blocks are defined, which can be used as standard proc-esses for individual tasks (Fig. 6).The model has two basic defects: • The special function of field observation of products or

systems, via service, is not explicitly taken into ac-count; it can undoubtedly be added easily and adapted to the individual case. Log book functions are a requirement for all components to record and store footprints, life cycle history etc.

• There is no basis for the parallelisation of the domain-specific design, i.e. the exchange of information dur-ing the domain-specific design is not defined. It is as-sumed that the entire domain-specific design pro-ceeds in concentric layers from basic to detailed.

More recent publications address this (cf. Bathelt [18], Gausemeier [19]), by defining an interface medium which contains all the information of the interfaces between the

domain-specific designs and institutionalises them from an organisational point of view, for example, in the form of synchronisation gates with workflows for request and re-lease. While in the case of Bathelt it is, for example, the expanded function structure according to Fig. 6 from which the mechanical product structure, the I/O list and the sequential function chart for the SPS programming, as well as the mechanical function structure and a sen-sor/actor list are derived, further development is moving in the direction of a uniform development environment for all domains, which at the present time is still unrealistic due to the diversity of the data processing systems avail-able. 2.2 Modelling and Analysis Fig. 7 illustrates the notion of an all-inclusive mechatronic database [20], which completely defines the product at the end of the design process and manages to do so while avoiding redundant data input during the develop-ment process. Such a database must be capable of being visualised and evaluated for the design process accord-ing to a diverse criteria (behaviour forecasts, circuit dia-gram, geometry, control software, control structure, etc.), shown in Fig. 7 as perspectives on the data model or as a facade of it. Such perspectives can, exactly like their as-sociated input tools, already be based on evaluations of the database as, for example, the circuit diagram is a graphical display based on the equipment and connection databanks which are part of the database. A structure for system data must be developed so that interrelationships between data can be mapped and a general integration of design and modelling tools achieved, almost like an “Es-peranto for design and modelling”. Initiatives in this area include for example UML, Modelica and State graph [16]. Similarly, it should be possible for any kind of simulation to be supplied by this database, without the need for addi-tional data and data going beyond the necessary design specifications of the product. Altintas et al [21] show the benefits of simulation technol-ogy and the use of virtual prototypes. In addition to this, because it is easy to vary the data model, a larger num-ber of development steps can be undertaken in a shorter time, thereby producing a more mature, more progressive product. When building such a model, however, working steps are always required which must be performed out-side the product definition, thus reducing the achievable advantage. Furthermore, the shaping of the model at the present time remains dependent on the questions to be answered, i.e. on which forecasts relating to the future product are used. The aim of model-based system design is the compre-hensive forecasting and optimisation of behaviour, includ-

verification / validation

requirements product

systemdesign

mechanical engineeringDomain-specific design

syst

emin

tegr

atio

n

electrical engineeringinformation technology

degree of maturity degree of maturity

modelling and model analysis

Figure 5: V-model as a process model for the design of mechatronic systems [7].

Figure 6: Standard work flow for system design after Bathelt [18].

thermalbehavior

Mechatronic Modelgeometric CAD-model

controlsoftware

dynamicbehavior

electric circuit planmechatronic data base

controllerstructure

Figure 7: Mechatronic model as an all-inclusive data base.

-661-

ing all the various domains, and this requires the afore-mentioned comprehensive database. The benefit of model-based design also lies in the virtual start-up, i.e. the testing of the control software with the relevant con-troller even before the real system is in existence (cf. Dierssen [22]). The more aspects of the product, which are mapped in the model, the better this is. The concept of a model-based design system is repre-sented by the MECOMAT system [13] in accordance with Fig. 8, which is an integrated software system for the syn-thesis, analysis and optimisation of a class of machine tools, in which, in addition to the structural elements in accordance with Fig. 9, drives and control elements are stored as parameterised elements and the designer syn-thesises his design from these. One method of achieving the interdisciplinary definition of machinery function and deducing the PLC code is de-scribed by Poernbacher and Zäh [23]. In this method, the function groups of the machine are modelled per UML, its behaviour as status machine and time-dependent behav-iour are modelled via simple laws of motion. The interac-tion between the elements is established in a sequence diagram, from which the PLC software can be derived. Even the basic design in the system design phase must

be examined in terms of its functionality, because design faults can no more be eradicated in the detailed phases, with the result that the development process has to be started all over again. There is therefore a need for mod-elling and analysis in the early phases of development as long as the database remains sparse, and for these to be expanded steadily as more detailed knowledge of the system is acquired [24]. Fig. 10 after [25] shows the gross modelling required for thermal analysis, which is per-formed using the thermobalance method.

Since mechatronic aids are used for the purpose of influ-encing the precision of machine tools, a calibration is re-quired which must be decided, as well as evaluated and optimised during the design phase. Model-based design for mechatronic systems therefore has the task of deciding a calibration strategy, and of es-timating the qualities of a system improved in this way in order to be able to ensure as a result that the overall sys-tem is capable of optimisation. As it is only with difficulty that reliable calibration can be undertaken in a straightforward manner by including axis-related control points, because of the uncertainties due to operating conditions and errors in positioning, a calibra-tion procedure is required which according to Weikert [26], and Schwenke [27] is based on an error model of the machine (Fig. 11). Its parameters are determined by means of the calibration according to a pre-defined strat-egy and associated system of measurement, and with its help it is possible to compensate for the kinematics errors of the machine. For a procedure such as this, forecasts are required regarding the uncertainty of calibration and the uncertainty of parameters as a consequence of the calibration strategy, in order to be able to carry out model-based error budgeting as well as a virtual acceptance test.

fixedtable

fixedheadstock

cross slidemovingheadstock

movingtable

base withguide

Figure 9: Combination of parameterised variants in the MECOMAT system [13].

Figure 8: MECOMAT model-based design tool [13].

frontbearings

rearbearingscooling

Figure 10: Gross modelling using the thermobalance method [25].

(1)

(1)

(2)

(3)(4)

(A) (B)

Figure 11: Calibration by means of laser interferometer to identify the parameters of a machine model after [27].

-662-

The universal generation and use of the uniform database ultimately facilitates model-based commissioning, in which the real control interacts as programmed hardware in the loop simulation model of the mechatronic remain-der of the system, the behaviour of the system being made visible by means of an appropriate visualisation method. Such approaches have been proposed for in-stance by Dierssen [22] and by Poernbacher and Zäh [23] as well as Pritschow and Roeck [28]. In addition to the visualisation of the machine, in [23] real components can be connected via the real PROFIBUS. The design of mechatronic systems is iterative because the challenge is to seek a system that has specific prop-erties. However, all simulation-supported methods initially require a system, the properties of which can then be predicted. For them, the design task is an inverse prob-lem. To solve this, Seo et al. [29] propose a unified ap-proach to the prior topological optimisation with the aid of a Bond graph-based representation of the system and an evolution algorithm, before beginning to implement the design. 2.4 The model as mechatronic component The aim of using a machine model for controlling machin-ery is to improve the behaviour of the machine by virtue of the fact that the behaviour of the machine is described and can be predicted in the control system itself. This means that actions can be initiated in such a way that un-desirable side effects are suppressed. Model-based initia-tives are available for • Compensation of kinematic errors (calibration model). • Processing the signals of additional status sensors. • Dynamic compensation. • Thermal compensation. • Failure prediction by detecting any deviation from

normal behaviour. • Avoidance of critical states (chatter) by model-based

prediction of areas of instability and • Rule-based decision models for reaching stable zones

in the vicinity (Altintas [33]). The aim of dynamic compensation is to improve the be-haviour of the machine, as a result of the increase in in-formation about the mechanical system which is available in the control system releasing the drive commands, so that the control system can take account of the physical behaviour of the entire system. Undesirable movements of the system are eliminated with the aid of axis regula-tion. A large proportion of these undesirable movements, above all vibrations of the system, is however determinis-tic and hence predictable, and could be taken care of up-stream of the controller by means of set point generation. The closed loop controller would then only have to sup-press the discrepancies between the model and reality and disturbances which could not have been anticipated, e.g. from the process. The published systems can be dis-tinguished with regard to the complexity of the model and hence the physical effects which are taken account of in the set point generation and at the point of engagement, i.e. at the point between the NC code and electrical signal generation at which the evaluation by the physical model comes into play. Other differences relate to the evaluation of additional sensors, the signals from which are processed with the aid of the machine model in such a way that they can connect actuators. In addition to the direct effect of the machine axes, systems are known which involve com-mand actuators, either as redundant and highly dynamic axes or as additional short correction axes. The integra-

tion of acceleration sensors near the tool centre point (TCP) is discussed in [34]. From the acceleration signals an observer for the TCP position is generated from a dy-namic model of the machine. The system can be used to master flexible structures and enhances the vibrational deflections by 70 percent in the given system with one axis and two masses. A Kalman filter is introduced to take into account that the model does not come arbitrarily close to the real system. Pritschow et al. [35] use a Ferraris sensor to improve the velocity signal instead of the measuring systems. These are used to observe disturbance, by which means it is possible to halve the sensitivity to disturbance. Zirn et al. [36] propose a controller for the positioning of a single rotational axis, where the cross-talk is taken into consid-eration via an observer of the mass angle of rotation formed from the mechanical model of the rotational axis. Symens et al. [37] opt for an H -controller for a 2-axis pick-and-place system, which adjusts to the compliance of the system according to the position of the holder. In this case however the mechanical model is not exploited directly, which generates non-linear changes in the con-trol parameters into nominator and denominator of the transfer function, thereby causing poor numeric condition-ing; instead the positions of poles and zero settings are changed directly in a linear relationship with the length of the radial arm. Altintas and Erkorkmaz [38] and Hadorn [39] propose a system using a model to achieve optimum-velocity con-trolled trajectory, in which a parallel-kinematics machine, because of its non-linearity, can achieve significant time savings compared with conventional velocity set point generation. Denkena and Holz [40] and Jaeger et al. [41] (cf. Fig. 12) demonstrate the use of the dynamic machine model for force/momentum set point generation and in addition expand the concept of the set point generation to adapt the assignment of parameters to the dynamic con-ditions of the controlled system.

Response to temperature changes is increasingly being seen as a problem for the manufacturer of machinery, and for this reason the thermal characteristics must be taken into consideration and optimised in the design phase. In the design phase, the design rules for optimum temperature sensitivity must be exploited and, where nec-essary, temperature control systems must be provided, the efficacy of which must be tested using a thermal model of the machine. Published concepts are basically distinguished in respect of the actuators: • Temperature sensitivity compensation via the ma-

chine axes (cf. [42]).

Figure 12: Dynamic machine model for force/momentum set point generation [41].

-663-

• Temperature sensitivity compensation via additional axes, such as for example heating and cooling ele-ments (cf. [43]).

They are different in relation to sensors: • Measurement of temperature (cf. [42][44]). • Measurement of thermal expansion (cf. [43][45]). • Direct mechanical probing of TCP (normal practice for

lathes) [46]. • Measurement of power of drives and assemblies. The thermal machine model build up in the control system follows the established data from the sensors and actua-tors and is distinguished in terms of structure and predic-tive quality. The basic variants are: • Simplified heat conduction equations and thermo-

elastic equations with a numerical solution. • Phenomenological equations, mostly physically moti-

vated [42]. • Neural networks [47][48][49]. • Stored look up tables [50]. Solutions for temperature compensation are summarised in a 1995 keynote paper by Weck et al. [51]. The direct displacement measurement of the TCP has the disadvantage that it requires extra time and that it records the deformation status of the machine only at one point from which the deformations of the structure must be ex-trapolated. 2.4 Allocation of Functionality Section 2.1 dealt with modularisation and partitioning as tools of mechatronic design and to reduce complexity. Another design issue relates to spatial and functional in-tegration. The aim is: 1. To create modules that function autonomously and

from the outside have manageable, simple interfaces, and as few of them as possible. This results in the creation of adaptronic solutions, i.e. modules in which sensors and controls, with intelligence between the two, are integrated, including spatial integration.

2. To operate with smallest distance to the causal loca-tion, i.e. to measure and to control (cf. Bansevicius and Giniotis [53]), where the final element in each in-teract chain is subject to the corrective influence of its own dedicated control device).

3. To exploit existing control devices and sensors as far as possible, in order to avoid introducing additional failure scenarios into the system by incorporating ad-ditional elements, which gives preference to corrective action via the main axes.

4. Mechatronic systems, depending on information from the process, often require the necessary values to be derived indirectly in order to protect the sensor sys-tems, and thus require complex signal processing and evaluation.

Modularisation of machine tools results in autonomously operating units. Defined interfaces make it possible to re-place them quickly, and in addition enable speedy instal-lation of these units. Strictly modular design thus permits, firstly, reconfigurable machines and machine systems, and secondly facilitates auto-commissioning and auto-surveillance. The need for transformable systems capable of adaptation to changes in the range of tasks requires that the structural and functional limits must be compliant to each other or must be identical (cf. Heisel et al. [54]). Fig. 13 shows a possible communications structure for reconfigurable machines, where the modules have their

own internal bus system, connected for instance to a special machine bus system. Reconfiguration by the user will only succeed if the control system itself recognises the configuration and adapts itself to it. For this, an open control system is required, which recognises new mod-ules at start-up and derives measures for adjusting. The control system uses a mechatronic model expanded and adapted by the new modules from which necessary con-trol sequences can be deduced. The parameters are set with the aid of the module parameters stored in the mechatronic modules. These parameters are entered in the configuration list and are interpreted by the control system, which then adapts to them, during start-up. Interfaces and structures for reconfigurable systems are also proposed in [56][57], while in Chen et al. [58] a modularisation strategy is proposed according to features of a family of parts to be manufactured. The necessity of ascertaining the mechanical properties of possible con-figurations in advance in the case of reconfigurable sys-tems is outlined in [59][60], for which in [59] sub-structuring via non-linear receptance coupling is used. Thus the circle is closed via [61]. Reconfigurable systems must, in accordance with Fig. 14, construct together with the configuration matrix a mechanical and thermal model controlled by the matrix, which subsequently serves in the control of the overall system to improve its dynamic and thermal behaviour. While approaches considered to date have taken as their starting point a hierarchical connection of the individual mechatronic modules, Rzevski [62] emphasises that with the increasing complexity of the systems, the self-

MACHINE TOOL

drive A drive B sensor A

field bus

machine bus

bus interface

e.g. filed bus

businterf.

controlmodul

sens.sens.

moduln

businterf.

controlmodul

sens.

modul1

Figure 13: Communications structure of a reconfigurable machine after [54].

HardwarePhysischer Zusammenbau

SimulationVirtueller Zusammenbau

WZM-Modul-Bibliothek - VirtuellWZM-Modul-Bibliothek - Physisch

Bahninterpreter

RegelungParameterdeterminierung

Software

hardware(physical assembling)

simulation(virtual assembling)

machine tool modules library(physical / virtual)

controlparameter determination

path interpreter

software

Figure 14: Structure of a reconfigurable machine system where the model required for dynamic and thermal com-pensation is created in the control system [61].

-664-

occupation rate increases to such an extent that the per-formance of the communication system is no longer suffi-cient.

3 MECHATRONIC SOLUTIONS In this chapter the current limitations on the functionality of machine tool components and their influence on the overall behaviour of the system is shown. Based on this, both general and specific mechatronic initiatives and solu-tions for the relevant problems faced are indicated. Inte-gration aspects are then discussed in greater detail in Chapter 4. 3.1 Solutions for Cutting Machine Tools 3.1.1 Feed Drives Functionality Characteristics Feed drives are the central elements of machine tools. The functionality of feed drives is in addition heavily de-pendent on the properties of other machine components, in particular the guide systems and the frames. This ne-cessitates a closely meshed assessment of individual systems. Direct feeds are increasingly being used in highly dynamic machines in particular, but in addition in-creasing drive bandwidths and fast traverse speeds have been recorded in electromechanical feed drives in recent years. However, the rising dynamic values lead to a more pronounced increase in the thermal load in the feed sys-tem, to which e.g. the cooling systems of the screw drives have to be adjusted. Model-based compensation of ther-mally determined deformations within the sphere of feed mechanics is addressed in publications including [46], where the problems of signal transmission in moving parts is the focus of particular attention. As an alternative to fixed temperature sensors, the use of thermography, which works without motion, is proposed in [46]. Coulomb friction has an essentially limiting influence on positioning accuracy in the feed system. The effects of friction can be minimised by using low-friction or friction-free drive and guide systems. Examples include hydro-static [63], magnetostatic [64] and aerostatic [65] screw drive spindle nuts and bearings. The disadvantage of these solutions is that they are usually technically more complex and more expensive in economic terms than roller bearing elements, and passive systems in particular have limited stiffness. In order to achieve a high dynamic feed, it is necessary to have both greatly increased gain of control – which is ex-pressed by means including an increased speed gain fac-tor Kv in the position control system – and a high level of jerk, which achieves high acceleration and high speeds by the shortest possible route. A high jerk is necessary for good path accuracy, and high gains are also required for good positioning accuracy and/or stability [66]. The con-trol bandwidth is normally limited, both in the case of elec-tromechanical drives and in direct drives, by the lowest natural frequency of the mechanical system [67]. In elec-tromechanical drives, the ratio of characteristic angular frequencies of the mechanical vibration system (mass-stiffness-damping of the mechanical transfer elements and the slides) to the drive is of decisive importance. High stiffness and damping of the mechanical structure is an essential prerequisite for a high bandwidth of the feed axes.With direct drives, a high bandwidth and high jerk as a basic principle can be achieved due to the absence of mechanical transfer elements. This leads however to in-creased dynamic excitation of the machine structure.

Mechatronic Solutions Static and quasi-static behaviour: In the case of large machines with heavy axes in particular, positioning accu-racy can be limited by friction in the guide system, even though the resolution of the measuring system is ade-quate. A mechatronic approach which can increase local positioning accuracy is the superposition of inaccurate NC axes with lighter, more accurate NC axes. In order to be able to achieve the required stiffness in the case of low mass axes also, they must normally be much shorter. At the same time, other drive and guide principles are of-ten used, e.g. piezo actuator systems and solid flexures. A decisive factor for absolute (absolute cutting position) accuracy is the configuration of the measuring systems. If the redundant axes have separate length measuring sys-tems, then with measuring systems connected in series it is only possible to achieve a relative (relative to the NC basic axis) improvement in precision. Here the limited control behaviour of the basic axis must be taken into ac-count and switched off, e.g. by clamping. Another ap-proach is to integrate the redundant measuring system in such a way that errors in the basic axis are recorded and can be compensated by the second axis. By this means it is possible to achieve an improvement in precision. Approaches or solutions to increase positioning accuracy using redundant axes are described in applications in-cluding a large lathe [68] (see Fig. 15), a milling table [69] and for the Z-axis of a five-axis ultra-precision milling ma-chine [70].

If extremely high standards are required in terms of sur-face finish and dimensional accuracy of workpieces (nm-range, e.g. grinding and diamond turning), conventional drive structures are limited [71]. An alternative approach to the achievement of extreme positioning accuracies is the combination of guide and drive functionality [71][72] [73], e.g. in steppers. Piezo actuators are frequently used as actuators. In order to minimise friction, the actuators are integrated in the mechanical system via solid flexures. In [73], ferro fluids are used for a drive and guide system. Dynamics of feed axes: In the following, mechatronic systems are proposed which further improve the func-tionality of the optimised feed system in terms of stiffness and damping. Based on the complex associations which limit the bandwidth of feed axes, mechatronic approaches are not limited solely to the feed axis. Initiatives and solu-tions, relating for instance to frame components or guide systems, are however dealt with in the relevant chapters. The problems of damping between components which are moved in relation to one another are discussed in de-tail in [74][75]. An adaptive electro-rheological film (ERF) damper is presented. A high level of damping, as required

direction of motion

tool holdersensor output

positioning screw for tolerance compensation

energy input frame

VDI-interfacepreload

solid flexure

(membrane)

slide

piezo actuator

Figure 15: Micro axis for large lathes [68].

-665-

for example during processing, can be set by means of a high-viscosity damping fluid. However, since high viscos-ity damping hinders rapid traversing movements, the damping can be lowered further during the traverse. Ac-tive fluids for semi-active damping are presented in [76] and other sources. A semi-active approach based on a controllable eddy-current damper for precision machinery is described in [77]. Solutions for increasing the axial dynamic stiffness of screw drives are presented in [78][79]. Autonomically act-ing compensating units are described, which are config-ured between spindle nut and slide. This allows compen-sation for vibrations in the drive train for which the band-width of the feed drive is inadequate. The solution pro-posed by Neugebauer [78] uses piezo actuators inte-grated in the compensation unit via solid flexures (Fig. 16). Piezo fibre sensors serve as vibration sensors.

An approach, which in principle can achieve maximum bandwidths on the tool and workpiece respectively, and which is similar to the solutions already described for in-creasing positioning accuracy, is to use superposed NC axes. It is also based on the combination of drives and kinematics with markedly differing characteristics [80][81]. Superimposing of large axes with low dynamics and short and high-dynamic axes (working in the same direction) results in a hierarchical division of the overall dynamics on several functional levels [82][83] (cf. Fig. 17). The overall dynamics (actual values) are obtained from the individual dynamics working in the same direction; see Fig. 18 (top). The highest possible overall dynamics can usually only be achieved locally. For example, roughing operations are carried out globally by moving the slow axes. Finish-ing operations take place locally by moving the fast axes in conjunction with positioning movements of the slow axes.Many solutions, in particular those in most Fast Tool Ser-vos (FTS), do not superpose motions, but use the main

axis for positioning only. Based on working principles, FTSs can be categorised mainly according to three types [84]: Piezo FTS [81][85][86][87][88][89][90][91][92] [93][94], Lorentz FTS [90][95][96][97][98][99], and normal-stress FTS [84][100][101]. With regard to moving strokes, Lu [84] defines short stroke as less than 100 μm, inter-mediate as between 100 μm and 1 mm, and long stroke as above 1 mm. A modular concept with redundant axes for five-axis mill-ing on large machine tools for tool- and die making is pre-sented in [82]. The machines normally designed for the demands of heavy machining need to be limited in terms of their dynamic parameters in order to achieve the re-quired precision, in particular during finishing, because of the long axes lengths and the low natural frequencies. Because of the superposition of motions, a compromise value for the agile axes in terms of their dynamic parame-ters and required length must be calculated from the characteristics of the slow axes (in particular the path until the maximum speed has been reached). If in addition the simultaneous superposition of several axes is considered, then the use of parallel structures with electromechanical actuators emerges as a rational solution for the agile unit. The agile hexapod unit in Fig. 18 has 5(6) degrees of freedom (DOF) on the TCP. In order to avoid causing the machine to vibrate during highly dynamic use of linear direct drives, a compensation of inertial force was devised in [103] in the form of a com-pensation slide, travelling on the same guide path as the processing slide but opposite to the direction of feed. Re-sidual parasitic vibrations are compensated by an active tool holder [106]. Another approach is impulse decoupling [66][104][105], in which the half of the motor which is normally attached firmly to the frame is also supported so that it can move in the direction of feed. Mechatronic solutions, which adjust the preloading de-pending on the loads, have been described in [78][107][108].Increasing thermal loads are a further aspect of electro-mechanical drives operated at high speeds. One solution to extend the temperature-dependent operating range of

ACTUATORS SENSORS

forces,torques

disturbances deviations

nominalsignals

MECHANICALSYSTEM

CONTROL

forces,torques

disturbances deviations

MECHANICALSYSTEM

CONTROL

SENSORS ACTUATORS

+

actualvalues

Figure 17: Principle of superposed axes.

piezo fibre sensorfor load detection

multilayer stack actuatorsin differential setup

solid flexure

driveby the spindle nut

power take-offto the carriage

capacitive measuring systemfor controlling the actuator stroke

Figure 16: Axial vibration compensation unit [78].

time [s]

path

[m]

AO original motionSA agile motionSI inertial motion

X, A

Z, CY, B

Figure 18: Motion characteristics of superposed axes (top) and agile unit of a modular machine tool [82][102].

-666-

high-speed electromechanical feed drives by means of pre-loaded spindles with bearings on both sides is de-scribed in [109]. To prevent thermally determined defor-mations of the lead screw resulting in inadmissible loads on the bearings, the axial fixation of a spindle bearing is equipped with an additional sensor-actuator unit. 3.1.2 Guide Systems Functionality Characteristics In conjunction with drive systems, guideways constitute the interfaces between frame components that move rela-tive to one another. Extreme demands are made on them regarding freedom from constraint (low friction), uniform movement and guidance precision, stiffness and damping characteristics. Damping in guideways plays an essential role in the system damping of the machine. An increase in system damping in roller guide systems can be achieved by the integration of additional damping shoes. The concept of a hydrostatic guide system in the compact guide system format in accordance with DIN 645 was presented at EMO’05 by [110], essentially reducing the costs of system integration. Use of larger carrying pockets enabled Aesop [111] and Hardinge [112] to achieve greater loading capability and stability in com-parison with previous development initiatives according to DIN 635. Because of the mainly mechanical control of the guideway gaps, (e.g. by means of membrane throttles or ring groove sheaths), hydrostatic guide systems are very resistant to process and disturbing forces. The high cost of production and operation is particularly disadvanta-geous, as is the fact that it is impossible to intervene in the control system behaviour during operations [113]. Coulomb friction in guide systems is particularly disad-vantageous when very high positioning accuracies or very high axis speeds are required. Minimal friction is the ad-vantage of magnetic and aerostatic guide systems [75]. They represent the best approach where there are ex-treme demands in relation to freedom from constraint and axes dynamics. However, passive systems possess very low stiffness and damping and are inherently unstable. Dynamic instabilities can be minimised by passive optimi-sation in aerostatic bearings. Mechatronic Solutions Active hydrostatic guide systems: A mechatronic ap-proach to hydrostatic guide systems is presented in [113]. The mechanical control system of the guidance gap is extended by a mechatronic control system. The hydro-static pressure pocket unit builds on the functional princi-ple of a servo valve. The fundamental distinction here is the nature of the deflection of the valve piston or control piston. The preliminary phase, consisting of a torque mo-tor/impact plate, is replaced by the sensor/actuator unit and the control arm, see Fig. 19. In order to record the vertical position, the dynamic pressure measurement principle is used.

Electro-/magnetorheological fluids constitute another ap-proach with regard to active level control of hydrostatic guide systems [73][74][114][115][116]. They change properties under the influence of electrical/magnetic fields, in particular their viscosity. Forces are generated via the fluids as a function of the effect of the fields. In addition to level control, drive functions with up to six de-grees of freedom can also be realised. Active magnetic guide systems: The great potential of linear direct drives in relation to achievable motion dy-namics will only be fully exploited when the friction of tra-ditional guideways does not limit the accuracy and speed of positioning. By integrating non-contact magnetic guide systems in feed axes it is possible to cancel out the feed velocity and acceleration limited by traditional sliding or roller guide systems [117][118][119][120]. However, in order to achieve a high static stiffness they demand highly dynamic control systems. The associated costs do however encompass the possibility of accommodating influences during operations, e.g. by correction of spatial positions. The Z-axis of a milling machine with active magnetic guides [120] as shown in Fig. 20 uses a con-figuration of eight electromagnets arranged in pairs, each at an angle of 45°, see also Chapter 4.2.

Active aerostatic guide systems: There are essentially two methods of realising active aerostatic bearings (AAB) and guide systems: the air-film-thickness feedback method (Fig. 21 a, b) and the bearing pressure feedback method (Fig. 21 c). In the first method, changes in the air film thickness are detected and electric signals are sent to actuators for controlling the flow resistance, thus keeping the gap height constant [121][122][123]. This approach is very accurate; however it is complex and expensive. In the second approach, the change in the pressure on the surface of the bearing controls the opening of the active restrictor. A disadvantage of this approach is that the film thickness on the bearing surface is not measured directly. However this method needs no electrical supply, and hence it is possible to incorporate the active restrictor

Fpiezo - sensor/actuator-unit

control-piston

dynamic pressure for distance measurement

ppumpQpump

h

machine base

pressure pocket

measure-panel

pre-throttle

control gear

Figure 19: Adaptronic pressure pocket for hydrostatic guide systems [113].

objective (bearing platen)

amp.(PI)

control3 piezo

act.

amp.measured

hdesired

h

capacitivesensor

airsupply

plat

en h

eigh

th

+ -

air filmpressurecontrol

sensor

u

x objectivemass

(c)

air film

controller

sensor

u

x

air pad

objectivemass

(b)

objectivemass

air padair film

controller

sensor

u

x

(a)

(d)

Figure 21: Types of AAB, in accordance with [124] and [123], and actively compensated aerostatic bearing [125].

Z-SLIDE

motor spindleprimary part of linear direct drive

electromagnetic guiding slide

Figure 20: Active Z-slide frame with magnetic bearings (left), view of an electromagnet (right) [120].

-667-

within the aerostatic bearing [123]. In [124], the use of (a) for linear guide systems and (b) for rotary guide systems is proposed. Figure (d) shows a constructional solution for a rotary table with active aerostatic bearings in accor-dance with principle (b). 3.1.3 Frame Components Functionality Characteristics Frame components are crucial for static and thermal ma-chine stiffness because of their frequently high elastic and thermal effective lengths. (Quasi-) static deformations can be compensated by calibration or model-based solutions (by the existence of corresponding NC axes). Frame components are frequently the cause of the lowest and most dominant natural vibrations. In many cases they significantly restrict the achievable feed bandwidth and hence the performance capability of machine tools. In ad-dition, because of their inertia, they have a decisive de-termining influence on the performance required from the feed drives. Therefore the primary design target for frame components is to achieve high stiffness and damping with low mass. In addition to optimisation of the topology, there are interesting initiatives in the area of lightweight construction materials for moving parts [127].Passive measures to increase the damping of frame com-ponents include, in addition to targeted choice of materi-als, the optimisation of joint damping or passive dynamic add-on systems such as absorbers or auxiliary mass dampers. However, the add-on systems have the disad-vantage of being tuned to a more or less fixed frequency range, which is detrimental in view of the frequently posi-tion-dependent machine characteristics. Mechatronic Solutions Static and quasi-static behaviour: Comprehensive compensation for thermally determined deformations by axes readjustment is only possible in 5-axis machines. In the case of lathes in particular, tilting normally cannot be compensated completely by NC axes. These problems have been addressed in [128][129][130]. Integration of additional actuators and sensors do permit active com-pensation for tilting of the spindle head. In the compensa-tion of deformations [128][129], an additional drive is inte-grated in the spindle head in order to compensate for thermally determined tilting around the Y-axis, see Fig. 22. Because of the high time constant of thermally deter-mined deformations, and in order to achieve a high static stiffness, an electromechnical drive is used. As a function

of the measured temperature, a trapezoidal thread spin-dle is driven via a stepper motor and gear unit. Because of the differing spindle leads, the actuator is splayed, changing its length. This results in a compensating tilting movement around the Y-axis. The thermally determined tilt error can be compensated by more than 90%. A unit with three degrees of freedom for compensation of thermally determined tilt error of the spindle head around the X and Z axis, as well as shifts in the Y axis, is pre-sented in [130]. Here, the spindle head is supported on bearings at three points and driven by three piezo actua-tors. Thermally determined diameter errors on the work-piece can also be lowered by more than 90%. For milling machines, an actively supported table is pre-sented in [131], which is controlled via four piezo actua-tors. By this means it is possible to compensate tilting de-viations around the X-axis and the Y-axis and for shifts in the Z-axis. Approaches involving the strategic use of what is in fact the undesirable thermoelastic behaviour of frame compo-nents as actuators to compensate for thermally deter-mined deformations is presented for example in [43][126][132]. The strategic generation of thermally de-termined deformations by means of thermal actuators (heating/cooling) can equalise undesirable thermally de-termined deformations. In contrast to widely used passive tempering, e.g. by means of coolant, there is no compen-sation for the cause of the disturbance. The initiatives can therefore be distinguished from the electromechanical and piezo-electrical actuators described above, in that thermal energy is applied instead of mechanical or elec-trical energy. This approach has the advantage that it is not usually necessary to introduce additional joints in the frame components. The relatively high inertness is how-ever a disadvantage. Dynamic behaviour and process stability: In active vibration suppression in particular, the issue of placement of additional sensors and actuators and the control strategies used is of crucial importance [133]. In relation to this, investigations have been carried out on portal ma-chine models, by Ehmann and Nordmann [134] amongst others. It is noticeable that there are hardly any initiatives for structural integration of actuators in frame compo-nents. Instead, ancillary systems with absolute effects (e.g. mass dampers) are preferred for dynamic problems or the actuators are placed within the flow of force (e.g. between the guide and the frame) [134]). An expansion of the limited fields of application of passive dynamic ancillary systems can be achieved by means of semi-active and active initiatives. In semi-active add-on systems, stiffness and/or damping are adapted to the cur-rent behaviour of the machine. An adaptive mass damper with a magnetostrictive actuator is described in [135] and an adaptive friction damper with pneumatic actuators for controlling friction force is described in [136]. Semi-active joint damping [137] achieved by controlling friction force by means of piezo actuators is also a known technique. In the case of major fluctuations in vibration behaviour, active absorbers (active mass dampers) provide an alter-native to passive and semi-active damping systems. They introduce into the structure the reaction force of an accel-erated inertial mass which is 180° out of phase of the vi-bration. Electromagnetic actuators are known from Nord-mann [138] and industrial solutions [140], while an elec-tro-hydraulic actuator is presented in [139]. Solutions for structurally integrated actuator systems are known from parallel kinematics. Because of the often clearly defined strut loads and the simple strut structures,

electric motor

gear (i=3.5)

lead screw with two different lead length

mechanical actuator (solid flexure)

electromechanical compensation unit

solid flexure (left side)

temperature sensor on the upper front of head stock

X

Z

Y

solid flexures (right side)

head stock of the lathe

Figure 22: Electromechanical unit for compensation of thermally determined tilting of spindle heads [128][129].

-668-

these provide, in contrast to the often complex frame to-pologies of serial kinematics, favourable conditions for the integration of additional compensation actuator systems. For struts with torsional load, a structurally integrated unit with piezo actuators is described in [141], see Fig 23. Solutions to compensate for axial strut deformations by means of piezo actuators are described in [142][143]. In finishing procedures such as polishing, even minimal vibration amplitudes are undesirable. For the active dy-namic increase in stiffness of a serial polishing machine, a brace with integral piezo actuators is integrated in [79] (cf. Fig. 24). The vibrations are measured via absolute acceleration sensors. The brace is mounted between the tool holder and the guide of the spindle head, hence par-allel to the flow of force.

An active connection to a grinding wheel spindle on the spindle head with the same aim of influencing machine compliance was presented in [144]. However, the piezo actuators are in this case configured within the flow of force.Components for active isolation of vibration also result in improvements in terms of the restricted field of application of passive systems, with hydraulic [145] and pneumatic actuators [146] being used amongst others. 3.1.4 Main Drives Functionality Characteristics In order to realise the very high speeds which are possi-ble today and the required precision, direct drives which are stiffer and without backlash are increasingly being used because of the availability of extensively adjustable motors. The high revolutions and accompanying high output requirements place extremely high demands on the static, dynamic and thermal stiffness and the useful life of bearings and spindles. Pre-stressed bearings with ceramic rolling elements guarantee a high static stiffness. Because of the high centrifugal forces and thermal loads, however, the pre-loading changes as a function of the revolution number, with the result that it is always neces-sary to find a compromise between adequate bearing stiffness and maximum revolution number. This problem

does not as a general principle occur with non-contact magneto bearings and aerostatic bearings. However, due to their passivity, they only have very low stability, al-though they stabilise themselves in the super-critical revolution range. Except in hydrostatic and hydrodynamic bearings, main spindles are only weakly damped. Since main spindles form a coupled vibration system with the various tools [147], the dynamic behaviour of the main spindle-tool-system is changeable, which limits the effect of purely passive damping systems. Online solutions therefore ex-ist for spindle speed variation or selection (variation of metal removal parameters), in particular to avoid chatter vibrations [149][148], based on the dynamics of stability lobe pockets or sensor integrated chatter detection [150]. Another popular on-line method for chatter avoidance is the continuous periodic spindle speed modulation tech-nique [148]. Tuning of the natural frequencies of the cut-ting tool and the spindle and tool holder setup is proposed in [151].Mechatronic Solutions Static and quasi-static behaviour: The active control of the bearing gap in particular in magnetostatic and aero-static bearings does however also improve static stiffness actively. Active non-contact bearings are used today in industry. The problems of pre-loading as a function of the speed in high-speed spindles with roller bearings can be signifi-cantly improved as a result of autonomous adjustment of the pre-loading by means of integrated actuators [105][128][129][152]. The mechanical conversion is es-sentially based on the same principle as that described in Chapter 3.1.1 for screw drives. Dynamic behaviour and process stability: A funda-mental distinction can be drawn with regard to the active damping of main spindles between two variants of the active bearing. Active bearing supportThere are currently three types of technically viable ac-tuators for active bearing support available, acting on the basis of different physical principles, i.e. electromagnetic actuators, piezo actuators and hydraulic actuators (piston actuators or active sliding bearings) [153]. Active magnetic bearings can be used for active damping or chatter suppression by means of appropriate control algorithms [110][154][155][156]. They are based on con-trol of the air gap. In addition, an option to expand the technology also exists. It is also possible to a certain ex-tent to use, for example, oscillating movement of the spindle rotor for lapping drilled holes for non-circular processing of drilled holes, for circular milling to compen-sate for the thermal drift of the machine or the drift of the milling cutter [157]. The disadvantage is the higher engi-neering cost. Piezo actuators are used primarily because of their high dynamic performance. Technical solutions with the aim of actively improving the chatter tendency of spindle ma-chine tool systems with roller bearings over a wide revolu-tion range, consist of integrating radially operating actua-tors, evenly distributed across the range, which exert their influence on the spindle shaft via the roller bearing [153] [158]. A solution with two piezo actuators was presented in [160][161]. A similar solution with 4 electrorestrictive piezo actuators is described in [159]. Because of their high dynamic performance, piezo actua-tors are often used as actuators in the air-film-thickness feedback method in main spindles with aerostatic bear-

fixed

twis

ted

piezoactuators

claws for supporting the actuators

twisting-elasticsolid flexure

shortening extension

Figure 23: Unit for compensation of torsional vibrations in parallel kinematics bracing [141].

module B for increasing static and dynamic stiffness

module A for increasing dynamic stiffness

Figure 24: Modules for increasing the dynamic and static stiffness [79].

-669-

ings. Motor spindles with active aerostatic bearings are commercially available from sources including [122]. Active sliding bearings with piezo actuators are presented in [162][163]. The work in [163] deals with the develop-ment of an adaptive hydrodynamic bearing made of a mobile housing mounted on piezoelectric actuators. The device is placed near one of the two bearings supporting a slender shaft.Hydraulic actuators can generate higher strokes than piezo actuators [164]. With the active milling spindle bear-ing shown in Fig. 25 it is possible to achieve strokes of ±0.5 mm [165][166]. The main view of Fig. 25 shows how the active ring is incorporated in the housing and how three separate pressure chambers are created.

Active ancillary bearingsThe active ancillary bearings are based on the principles of active bearing support. In [167], a milling spindle was examined in which the critical number of revolution per minute can be influenced by an additional adaptive mag-netic bearing B between the main bearing (A) and secon-dary bearing (B). This circumvents the attainment of the first critical rotational speed, resulting in a fundamental expansion of the stable working area. A similar solution for spindles with roller bearings in which the additional adaptive magnetic bearing is placed in front of the main bearing is described in [32]. Automatic balancing systems: For spindles with active magnetic bearings, there exists the possibility of compen-sating for residual imbalances from the main spindle-tool-system during processing (in-process balancing). How-ever, automatic mechanical balancing systems also exist [168][169][170]. As a rule they displace individual masses [171] or the centre of gravity of eccentric rings. This latter variant is used by [169]. The level of balancing, which can be controlled during operations, is realised by means of the control synchronised with the revolutions of two axi-ally very closely positioned imbalanced disks on the spin-dle shaft. The transfer of force takes place via permanent magnets. 3.1.5 Tools and Clamping Devices Functionality Characteristics Tools and tool holders have a decisive influence on the achievable processing precision and metal cutting capac-ity. Firstly, because of their slender form, tools have very low static bending strength and torsional stiffness. Sec-ondly, in the flow of force between tool and workpiece, other compliances inevitably arise as a result of additional joints. In [174] it was established that the tool - tool holder - main spindle - bearing system can contribute over 50% to the total compliance at the cutting edge. A much higher proportion can be expected in the case of particularly long tools. In addition, the low stiffness of tools also has a lim-iting effect on process stability (regenerative chatter) and

hence on the achievable metal cutting volumes in a given time. Secondly, however, the additional joints also have a fundamental effect on the overall damping of the tool-spindle system. A number of activities have been undertaken to increase the static and dynamic stiffness of tools, in particular those with a large overhang. Passive add-on systems for vibration suppression are usually easy to optimise, since the properties of tools are less susceptible to fluctuations in parameters. A greater need for intervention is apparent in the static behaviour of thin tools. There is a real need for an active correction system for very low frequencies but with reli-able and affordable sensing devices [172]. Mechatronic Solutions Static and quasi-static behaviour: The model-based correction of the milling deflection is described in [175] (mathematical process force model) and [176] (overview), [177] and elsewhere. In machines which cannot execute all the correctional movements due to the absence of NC axes, the integration of additional actuators is advanta-geous. A solution of this kind is presented in [178] (see also Chapter 4.2). Piezo actuators facilitate dynamic posi-tioning of the tool-main spindle system in the micrometer range by a tilting motion around the X and Y axes and by movement in the Z axis (Fig. 26). Movements of around ±130 μm can be generated.

All these principles are however scarcely suited to com-pensation for the deflection of thin boring bars. In this area, mechatronic initiatives with additional actuators have great potential. In the solutions already known [179][180][181][181][182], the tool cutting edges can be adjusted to compensate for deflections of the boring bar. The active boring tool shown in the Fig. 27 consists of a rotating boring shaft and a vertical measuring sleeve [182].

An active chuck has been presented in [183], see Fig. 28. The mechatronic chuck equalises eccentricities in the clamped workpiece, arising for example as a result of er-rors in assembling the chuck itself, or as a result of errors

oil

active ring

hydrostaticallybalancessealing oil

piezovalveactive ring

positionsensor

bearings inO-arrangement

CHIRONmachine tool

hydrostaticbearing

preload spindle bearing

Figure 25: Active spindle bearing system with hydraulic actuators and servo valves with piezo pilot controls [166].

inductive energy transmissioninformation transmission with slip rings

spindle

piezo actuator

cutting edge

static sleeve withdistance sensors

rotary encoder

eddy current distance sensors

cutting edge

solid flexures and strain gauge sensors

Figure 27: Active boring bar to compensate for the deflec-tion of the cutting edges boring processes [182].

withoutcompensation

withcompensation

nominal positionpart

cutter

main spindlesolid flexure(membrane)

piezo actuatorprestressing clamping

ring

supportingring

Figure 26: Compensation of deflection of the milling head by a mechatronic unit [178].

-670-

in clamping the workpiece. Communications with the electronic elements mounted in the chuck take place via Bluetooth.Tools for polishing optical surfaces such as wafers must realise an even distribution of the polishing pressure. An industrial application [185][186] has presented a seg-mented active chuck for this purpose. Dynamic behaviour and process stability: As illus-trated in [172], the use of clamping systems with good damping properties is an essential approach for increas-ing process stability. Fluids are often integrated in order to increase damping. These systems are capable of fur-ther optimisation if the damping force can be adapted to changing process conditions. A semi-active tuneable-stiffness boring bar containing an electrorheological fluid (ERF) is proposed in [187], to suppress chatter in boring. The ERF undergoes a phase change when subjected to an external electrical field and serves as a complex spring behaving non-linearly in the structure. As a result, the global stiffness and energy dissipation properties of the boring bar, under an electric field, change drastically at different amplitudes of vibration. It is found that the ampli-tude of chatter can be prevented from increasing by using the nonlinear vibration characteristic of the ERF. A similar semi-active damping concept, but with a magneto-rheological fluid (MRF) in the tool holder in a boring bar for single lip drilling, is described in [188]. The efficiency of active damping with feedback control on tools or tool holders has been demonstrated in various projects [189][190][191][192][193]. Because of positive developments in terms of price, another picture will now no doubt emerge, confirming the commercial availability of active turning tool holders. In [192], various turning tool holders with embedded piezo actuators are offered, see Fig. 29. According to the manufacturer, these should re-sult, in conjunction with an external controller, in a 15% production improvement, 75% product surface improve-ment, 10% reduction in costs for tools and work holders and a 90% reduction in noise. A procedure for active damping of vibrations in wheel-shaped cutting tools is described in [194]. The forces re-

quired are introduced into the structure via piezo actua-tors, where the force transfer to the rotating cutting wheel in phase opposition to the wheel is accomplished via ball bearings. A similar paper on this topic is presented by Pape and Carlson [195]. The reduction in vibrations in circular saw blades, is investigated in [196] via the princi-ple of multi-modal vibration absorbers by means of inte-grated piezoceramic components in conjunction with pas-sive or semi-active electrical networks. The aim of most of the systems described above is to im-prove the dynamic behaviour of tools. In addition, there are also initiatives to integrate mechatronics in the tool or in the tool holder for the purpose of improving the behav-iour of the machine. Since tools and clamping systems represent the ends of the force flow of the machine in the direction of the process, they are especially suitable loca-tions in which to integrate sensors and actuators to de-termine and compensate for working space errors. Vibration suppression for turning operations using integral force feedback strategy is implemented in a smart tool holder in [197]. It is equipped with a standard VDI-3425 interface to fit into the turret of a CNC-lathe. Similar solutions have been presented in [30] (piezo ac-tuators) and [198] (magnetostrictive actuator). In [30], an optical sensor records the relative displacement between workpiece and turning tool holder. In [106], an active tool is described for lathes. Use of a highly dynamic infeed of the tool cutting edge, using piezo actuators, reduces the disturbing effects on the cutting depth caused by the me-chanical vibrations (measurement of machining force) and at the same time records and compensates for the radial deviation of the spindle via 3 capacitive sensors. The solution presented in [199] decouples the glass cut-ting wheel from portal vibrations. Like the tool, the workpiece can also be isolated from machine vibrations. The effect on the workpiece of an ad-ditional support (active steady rest) with piezo actuators and capacitive position sensors is presented in [31] in re-lation to lathes. In [200], a system is presented in which several process and state variables, including waviness on the workpiece/tool in process, are detected. Vibration assisted machining: Significant reductions of cutting and friction forces can be realised by introducing an additional motion into the cutting zone. Positive results (reduction of cutting and friction forces as well as tem-perature in the cutting zone, tool wear and surface rough-ness) were recorded for a broad frequency range 20 Hz – 40 kHz [172]. The superposition of vibrations on the process has been investigated in many machining procedures. Grinding and lapping in particular have scope for significant industrial application. In order to generate the necessary vibrations, the band-width for the existing machine drives (NC axes) would only be sufficient for very low frequencies (up to approx. 100 Hz, depending on the size of the machine). In addi-tion, this excitation variant is dubious from an energy-related viewpoint because of the large masses to be moved. The consequence of this is that an additional ac-tuator system is required to generate vibrations. For the aforementioned reasons, this is integrated as closely as possible to the process. For optimum transfer of the excitation energy to the tool cutting edges, a corresponding natural frequency of the system capable of vibration, usually the tool is excited, in particular in the higher frequency range. In addition to longitudinal natural frequencies to generate axial move-ments of the cutting edge [201] (approx. 20 kHz). Bending

Measurement system for the gear contour

turning tool

part

frame components, connected to the chuck

piezo actuators and strain gauge sensors

inductive energy transmission

flange

elasto-kinematicelement 2D

(solid flexure)

frame components, connected to flange

Figure 28: Precision positioning system for lathes [183].

cutting speed direction

part boring bar

embedded piezo

actuator

control

excitation introduced by the process

Figure 29: Smart cutting system for reduction of harmful tool holder vibrations [192].

-671-

natural frequencies [202] (approx. 20 kHz) are excited in order to generate elliptical vibrations of the cutting edges. 3.2 Solutions for Forming Machine Tools 3.2.1 Main Drives Functionality Characteristics An optimisation of the kinematic behaviour of path-bound presses can be achieved, e.g. by means of unequal translation of the drives. The disadvantage of this is, however, that it is not possible to respond to changing requirements in terms of the process technology. In the tryout field, in particular, and in smaller batch sizes, con-trolled hydraulic main drives are now used to emulate the kinematic behaviour of path-bound presses [203]. They are characterised by variable press forces and change-able ram strokes and forming speeds. In addition to classic path-bound mechanical or force-bound drives, research activities are focusing on alterna-tive drive principles such as direct linear drives for micro-forming [204], innovative machine structures [205][206] and the magnetic forming of sheet metals [207][208]. These aspects will not however be included in the follow-ing analysis. Mechatronic Solutions Path-bound presses have a high energetic efficiency and high stiffness. Through the use of a controllable second drive superposed on the main movement, it is possible to suspend the path-bound property of the ram motion of the flywheel for a limited time, which makes it possible to im-prove adaptation to process requirements [209]. By this means, for example, the impact speed can be reduced while the number of strokes remains unchanged, resulting in an improvement in the dynamic behaviour at the be-ginning of the forming process. Alternatively, it permits the kinematics of the press drive to be changed in such a way that the process cycle is shortened at the same (lim-ited) impact speed. The principle of a solution with a servo motor and planetary gearing is shown in the Fig. 30.

3.2.2 Frame Components Functionality Characteristics Forming machines and their frames need above all to be able to cope with large process forces. These introduce large deflections of the frame, which in a large car body panel press may range up to several mm. While the lengthening of the uprights, the deflection of the drive sys-tem and crown are easily compensated during setup by the ram adjustment, the deflection of the table introduces faults in the workpiece and needs to be compensated in the dies.

Mechatronic Solutions It is of significant interest that try-out presses can create the forming behaviour of concrete production machines. For this various investigations were carried out in which additional hydraulic cylinders, needed for the emulation of the forming behaviour, were integrated in the machine bed (see Fig. 31). Systems were mainly used for this to reduce the stiffness by eliminating stiffness structure ele-ments such as walls, ribs, etc. In connection with deflec-tion- or force measurement systems, this system offers an alternative to real mechatronic systems.

For press brakes, which are directed towards manufactur-ing of net shape sheet parts, compensation of frame de-flections is necessary and has been developed into a very sophisticated and highly integrated mechatronic system (see Chapter 4). The deflection results in a typical boat shape of the part [212] (Fig. 32). Besides the varying an-gle over the length, the bending line becomes curved, depending on the deflection partition of the lower and of the upper beam. Therefore deflection compensation of only one of the two beams necessarily leaves the two faults uncorrected.

The compensation of the deflection of the lower beam is done by crowning, pre-deflection of the lower beam with either wedge systems or short stroke hydraulic cylinders, which lift the tool carrying frame against some supporting frame, see Fig. 32. Today crowning is done automatically. The pressures in the two cylinders during closed loop controlled synchronous movement of both sides of the beam are measured. From those measurements, the de-flections of the die and the lower and upper beams are calculated using a mechanical model of the lower and upper beam, and the required crowning is deduced from this so that the two beams become parallel in the vicinity of the operating die (Fig. 32). Crowning shall be synchro-nous with the loading to avoid marks and internal stresses. This still does not compensate for the variation of the bending angle, which also needs a pre-deflection of the upper beam, which could be achieved in a manner similar to the crowning of the lower beam. In free bend-

main drive

additional drive

epicyclic gear train with flywheel

epicyclic gear with flywheel

Figure 30: Second drive with planetary gearing [210].

lower die

die carriercompensation cylinders

upper die

press bed

Figure 31: Hydraulic cylinders in the press bed for com-pliance adjustment [211].

short stroke cylinders

upper traverseand punch

lower traverse and die

90° > 90°90°

90°90°

90°

Compensation of Bending Compensation of Tilting

Figure 32: Dynamic crowning of lower beam, after [212][213].

-672-

ing, the bending angle is determined by the position of the upper punch relative to the lower die. Therefore the de-flection of the frame as well as thermal expansion needs to be controlled. As the final bending angle is largely in-fluenced by spring-back, the functionality of press deflec-tion compensation on the machine side, and the spring-back compensation on the materials side are integrated. Two main systems can be distinguished by the meas-urement inputs: direct measurement of the bending angle or indirect measurement of the punch position and the bending force. For the direct measurement the closed loop control may adjust the bending angle iteratively. To speed up produc-tion, in [216] a process model is introduced for bending, which enables the machine to find the correct bending angle to within an absolute uncertainty of 1° within two steps.The sensors range from sensors that detect the bending angle with different kinds of probes from the punch side [212][214][215] as well as from the die side (LVD, Gas-parini) [216] or optically [217]. In the die area, the main disadvantage is that the sensors need to be exchanged, or adjusted for every new part, besides which it is in an unsafe area due to material han-dling. This decision is perhaps the main aspect in the al-location of mechatronic functionality in forming machines. The indirect measurement, introduced by Stelson [218] and Reissner and Meier [219] avoids extra sensors in the die area, but measures at a distance from the process and thus results in disturbed signals. 3.2.3 Dies Functionality Characteristics The material flow required for the production of sheet metal components is decisively influenced in conventional metal forming by the tuning of the force regime of punch and blank holder and by the design of the drawing gap between punch and matrix. Optimum in-feed of the sheet metal material into the matrix is a prerequisite for the pro-duction of good components and the avoidance of cracks or folds. Using special machine tool technology meas-ures, it is possible to expand the limits of forming technol-ogy, but above all to ensure the viability of forming proc-esses, especially when processing high-strength and ul-tra-high-strength sheet metal materials. Mechatronic Solutions Intelligent drawing technology – Multipoint drawing technology: In addition to further development of con-ventional drawing technology, entailing the challenge of mastering system parameters, innovative solutions are being used to control the complex processes involved in controlling the flow of materials, especially with regard to complex component geometries [220][221]. There will be additional pressure associated with this in the context of materials which are difficult to form. Multi-point drawing devices in which there is a direct transfer of force from the cushion cylinder to the blank holder can be used to control the flow of materials during deep drawing, see Fig. 33. Where there is corresponding flexibility of the blank holder, it is possible to regulate the pressure locally and to optimise the in-feed behaviour of the sheet metal. The consequence is a significant in-crease of the process window, leading to more robust manufacturing procedures. Responses to changes in the process parameters are only possible provided there is integration of a process control system. The drawing in of the sheet metal into the flange area is used as a process control variable, while

the local hold-down force facilitated by multi-point tech-nology is used as the manipulated variable. Various sens-ing elements and strategies (including laser triangulation measurement) are used to measure the in-feeding of the sheet metal (material flow), see Fig. 34. The basis for control is a master curve (sheet metal in-feed curve), which is recorded by the laser-sensor during production of a drawn component, regarded as optimal in terms of shape and dimensional accuracy. This master curve of the sheet metal in-feed serves as the target value in the control circuit. The data recording of the master curve takes place after optimisation of factors such as the blank, the lubrication regime and the hold-down force. The effects of using process control on a real component are a significant reduction in the rate of defective compo-nents, stabilisation of the process and improvement of the workpiece quality.

Tool-integrated solutions: One initiative designed to adapt the hold-down force to the process at critical points, despite pressing without multi-point cushions, is the inte-gration of actuators in the deep-drawing tool. For this purpose, the blank holder (or the tool) is designed to have local flexural elasticity. Piezo actuators were used for the tool shown in the Fig. 35. Intelligent drawing technology – integrated parallel holding: When using transfer tools, the eccentric location of the drawing stage leads to tilting of the cushion plate, as a result of which undefined clamping bolt forces are introduced into the hold-down device with consequent wear and tear of the guide- and clamping bolt plate. The drawing cushion presented in [224] has a special cushion control system with integral parallel controls especially for

ram

cylinderblock

distribution-safety block

storage unittable boxcushion cylinderpinsblank holder

power unit

Figure 33: Multi-point deep drawing cushion [222].

Drawing depth

Rad

ial l

engt

heni

ng Δ

L

failure-free part

ΔLWrinkles

ΔLFailure-free

ΔLCracks

Monitoring of the material flow

elasticblank

holder

stick punch

sensor

die

Fbh5

Fbh2Fbh1Fbh6

process computer with fuzzy-controller

Controlling(partial blank holder forces Fbhi)

blank

sensor position

sticks

Punch

Fbh6

Fbh8 Fbh1

Fbh3

die

opticalsensor

press table

cylinders

bottom of drawing cushion

blank

© IFUM

wrinkles

crack

Figure 34: Measurement of flow of materials [221].

-673-

the above application in order to be able to dispense with external equipment (push rods, balance pins), see Fig. 36. The 7-point cushion can compensate for a tilting movement introduced by the process due to the me-chanical construction and the architecture of the cushion control system.

Parallel holding of the ram: Eccentric loads cause the ram to tilt, especially in the case of less stiff hydraulic presses but also in the case of mechanical presses. For this reason, passive and active parallel holding systems are used. Parallel ram holding can also be used in tryout machines for emulating the compliance behaviour of the production presses. This requires pre-defined, stroke-dependent skewing of the ram. There are various solutions with regard to the design of the parallel ram holding system, distinguished by their construction and precision parameters: • Parallel ram holding by the main press cylinder. • Parallel ram holding by means of four separate cylin-

ders which introduce a force between the table and the ram [222].

• Parallel ram holding by means of four separate cylin-ders between the head and the ram [225].

The most accurate results are obtained, depending on stiffness, by the variant with four separate cylinders be-tween the table and the ram. Vibration superposition in the forming process: The process window in sheet metal forming is limited by the use of new, higher strength materials and the growing demands in relation to quality. One option for extending the forming limits of the materials by technological means is the superposition of vibrating process forces during forming. Strategic intervention in the tribological system can achieve both improved material flow by reducing fric-tion and exert a positive influence on the structure of the material [226]. In addition, reduced forming forces are re-quired when realising the vibration-superposed forming process. The quantity of lubricant can be significantly re-duced.

A direct effect on the drawing punch is only applicable in dual-function hydraulic machines. In single-action presses, the option thus exists to excite the drawing punch via an additional plate. When the punch vibrates, the additional plate and punch are both subjected to an oscillating lifting motion (sinusoidal movement) which is realised independently of the external force (drawing force).In the case of direct excitation via existing drives, the achievable frequencies are limited by inertia. For the tools of a 500 tonne press, punch vibrations of 20 Hz and blank holder vibrations of 50 Hz are achieved. Superposition of ultrasonic waves on massive forming processes also facilitates a reduction in friction and an improvement in surface quality [226]. In [227] it is shown that the complex effect of ultrasonics on friction in forming processes can be theoretically described. From various investigations [228][229], it can be con-cluded that when the direction of vibration is in the direc-tion of the friction force, the friction is primarily affected by the ultrasonic vibrations. If the drawing direction is per-pendicular to the direction of vibration and the matrix is positioned in the maximum tension direction, then it is primarily the element of drawing force required for the forming that is affected by the ultrasonic vibration. Fig. 37 shows ultrasonic-superposed tube drawing [230]. The mandrel is excited to vibrate axially.

4. SYSTEM INTEGRATION This chapter examines the system integration of mecha-tronic components in an existing machine tool. The sys-tem boundary between the mechatronic component and the machine itself is often smooth and not always well-defined. In Fig. 38, the three domains of mechatronics (mechan-ics, electronics, information technology), are shown in each circle. The physical integration describes the elec-trical and the mechanical domain and the functional integration is shown with the information integration do-main.Regarding Fig. 38 from the bottom up, the lowest layer represents the design phase. In this layer the integration of the mechatronic component and the machine will be illustrated, divided by the domains. This phase was con-sidered in Chapter 2. Above the design layer the model layer illustrates the interaction of the component in the information technical domain. The model layer discusses the integration of the component in its functional integra-tion in the information technical domain. The real level constitutes the physical integration of the component in the machine. Finally the real and the design level repre-sent an illustration of physical integration of one mecha-tronic component during the developing (design level) and during the application phase (real level). The inner ring on each individual level represents the component and the outer ring symbolises the machine. The whole circle therefore represents the system. Interactions be-tween the domains are indicated by arrows. For example

controlled (bending-elastic) ranges

4 piezo actuators

Figure 35: Deep-drawing tool with local flexural elasticity and integrated piezo actuators [223].

ram

Kippmoment

eccentric load

die carrier

parallel-holding cylinders

low pin force

equal cushion forces

ram

Kippmoment

without tilting compensation with tilting compensation

high pin force

low cushion forces

high cushion forces

equal pin force

tilting moment

tilting moment

Figure 36: Principle of integrated parallel holding [224].

ultrasonic-generator

die

gripper transformatorpiezoelectric transducer

λ/2-sonotrode

FZ

mandrel vibration

Source: after Buckley, Freemanmachine bedmoveable slide

Figure 37: Ultrasonic-superposed tube drawing [230].

-674-

there is an interaction between the electrical domain to the mechanical domain in the form of an actuator system. Conversely, mechanical actions are transformed into electric current via sensors. Electrical signals are trans-formed via digital analogue converters into digital signal flows for the control system in the information domain. The system integration in Fig. 38 is shown by the red in-tegration arrows between the inner ring of the component and the outer ring of the machine. In principle, integration forms can be realised in a physical way on the real or de-sign level, or in the functional integration on the level of models. When controlling the component or the ma-chines, it is possible by modelling the complete system to achieve an increase in quality control of the individual systems. 4.1 Options for System Integration There are various options for integrating components in an overall system. The following are of particular impor-tance:• Mechanical energy exchange via coupling. • Energy exchange via electronic coupling. • Information exchange via regulating and control

components. There are basically two options for mechanical integra-tion of components. Firstly, the design of a modular com-ponent which is capable of flexible integration in various machines. Secondly, the integrated component is adapted specifi-cally for the individual machine and usually cannot be transferred to other machines. From the perspective of electrical integration, the energy supply to the component is of great interest. Here, stan-dard plugs and voltages are necessary to connect sen-

sors, actuators or microcontrollers. The energy supply to mechatronic components is expensive in modular inte-grated systems for the reasons already outlined. Consequently, non-contact solutions are often used in this field. These can be inductive, optical or even decen-tralised energy generating systems [177][195][231][232]. The information technology integration relates essen-tially to data transfer and data processing. Today data transfer primarily involves digital data. Be-cause of the various demands and market situations, a large number of possible bus structures and general data transmission channels became accepted. Fig. 39 shows communication channels in production via field buses.

In highly integrated mechatronic components, the aspects of data exchange and energy transfer were kept in mind during the construction process and therefore were taken into account in terms of the configuration. Mechanical components were adapted to accommodate demands such as the passage of cables. In contrast, the connec-tion of modular systems, integrated as modifications to existing systems, is far more complex. In many cases data transfer and energy transfer via cables is only possi-ble under very limited conditions; therefore non-contact types of transfer are getting ever more widespread. The most important solutions are set out with the relevant characteristic data in Table 1. Rational data processing and reliable data evaluation of-fer great potential for future innovation. One of the chal-lenges is the intelligent evaluation of data from a number of mechatronic components and the rejection of irrelevant data. Thanks to a growing number of modular compo-nents, there is an increase in interfaces for machine tool control, undeniably leading to an increase in the volume of data which has to be processed by the control system. The control systems in common use are: • Stored program control systems. • Numerical control systems. By increasingly computing power there is an option to perform more complex tasks, as well as to consider more complex selection procedures for making decisions to ac-complish tasks. Complex calculations can be performed with an NC con-trol system such as any kind of Digital Signal Processing Units (DSP).In the field of control, model based control concepts be-come more and more interesting, not only for robotics but also for machine tools. Using independent controllers for each axis like the single actuator control for serial industry robots [253][254], has been a well known and established procedure for several decades. Such concepts reach ac-

Figure 38: System integration of components in a ma-chine system.

Figure 39: Communications in production [234].

-675-

ceptable positioning accuracies for pick and place opera-tions as well for slow processing tasks. On the other hand the progression of the dynamic performance of today’s available actuators requires more sophisticated control concepts to maintain the accuracies achieved to date, for operations with improved dynamic performance. Thus, model based control concepts consider the individual characteristics of the drives and actuators by modelling the systems’ behaviour as well as modelling the dynamics of the complete system. Such controllers have been well known for several years, although only the increasing cal-culating capacities of modern computers allow their im-plementation today [233]. The development of “intelligent” machine tool compo-nents is being planned in the German collaborative re-search centre (SFB) 653. The vision is a “feeling” compo-nent which possesses different network sensory capabili-ties as well as an inherent information storage device and independently initiated communications. As a result it is capable of autonomously recognising an extremely wide variety of process status and their effects on the proper-ties of the workpiece, as well as the behaviour of the tool, and if necessary of initiating appropriate correcting meas-ures [231][234]. In the field of identification of objects and positions, artifi-cial intelligence has been the source of many methods which have been investigated and applied. Initially, pa-rameters are extracted from pictures, followed by a classi-fication of the pictures, i.e. an interpretation of the content of the pictures. This can be achieved with the aid of re-sources such as neural networks. All three measures of effectiveness of simulation systems, i.e. operating time, storage requirements and applicability, can be fundamen-tally improved by the use of AI (artificial intelligence) tech-niques [235]. The foundations, common features and applications of AI methods are being researched in SFB 531 “Design and management of complex technical processes and sys-tems with methods of computational intelligence”. In addi-tion, AI methods and their significance for production technology are described by Pritschow and Kissling [236]. By means of AI and more complex data processing sys-tems, the workpiece could be integrated in the way mechatronic components are considered. By means of sensors on the workpiece, data from the production proc-ess are fed into the control system of the machine. In ad-

dition, data from quality control for example can be stored in the workpiece. In this way, a workpiece represents a mechatronic component integrated within the machine for a limited period of time. On the IT level, efforts are being made to accomplish data transfer during processing. On the electrical level, verification is required regarding the way in which the workpiece is supplied with energy by the machine. This can be achieved using a wireless method. In future, the complete life cycle of the product can be monitored, for example using Radio Frequency Identifica-tion (RFID) technology. The systems become more com-plex it beams more important to investigate system-failures and their consequences. 4.2 Design for Reliability of Mechatronic Systems A frequent task and application of mechatronic concepts is the improvement of functionality of originally mono-disciplinary, mostly mechanical systems. This is accom-plished via additional sensors, actuators, in order to achieve the transition from one domain to the other, i.e. in each case using additional functional elements. Assuming that each functional element has its own failure probabil-ity, the reliability of a system would suffer as a result without any additional measures. On the other hand, mechatronic components often serve to secure the func-tionality of the mono-disciplinary basic system, and there-fore also serve to widen the status parameter range in which the system delivers good results. The reliability of a system is a quality characteristic and requires the same care as the realisation of functionality itself [237]. Con-trary to general opinion, that a hierarchically structured system is as poor as the weakest level [238], the situation is much worse. For serial systems in which the function is accomplished through interaction of all the components, the overall system is always more unreliable than the worst component. Hence it is the system complexity that is addressed di-rectly in terms of the uncertainty of the system’s capability to meet the required functional standards [239]. In order to improve reliability, therefore, the measures specified in [239] for reducing system complexity are an effective pat-tern for action.In line with the importance ascribed to it, dependability must be secured during the design phase (Design for De-pendability) [237]. Even in very early phases of the design process, it is necessary to work towards the dependability of the overall system. In [237][238], procedures are de-scribed by means of which it is possible to make depend-ability-oriented decisions at the beginning of the system design process, despite data which is still very uncertain. In [237][240] it is shown how design for dependability is incorporated in the work plan in accordance with VDI 2206 and how the epistemic uncertainty, as a measure for the reliability of the data, converges with the specified reliability as the data becomes more concrete. The basic causes of undesirable events are drawn up with the aid of the fault-tree analysis. The link between reliability characteristics and a fault tree is given in [241]. In [242] it is shown how a reliability analysis is built up using status models. Systems for which the safety aspect is critical must be fail-safe, i.e. they must be able to suffer certain distur-bances without losing their safety function. In addition to this, the failure orientation of components is critical [244] (fail safe, fail silent). The fail silent principle must be real-ised for mechatronic systems in particular, which serve to improve characteristics, i.e. the defective component is identified by means of an error message without disturb-ing continued operations. Paired with redundancy, there-

Table 1: Characteristics of wireless connections.

1900-2170 MHz

FDD mode (transmission in diff. frequency ranges): 384 kBit/s for Downlink;TDD mode (transmission at diff. times: 10 ms, trans-mission time): 1920 kBit/s

Antennaca. 20 WHandy0,125-0,25W

Preferablyfrom50 m to >20km

UMTS

5 GHz range23 Mbit/s – 155 Mbit/s0,2 – 1 W30 m – 5 kmHIPER-LAN

2,4 GHz range1 Mbit/s10 mWibree(new dev. 2007)

2,402–2,480GHz

Up to 1.2: 723,2 kbit/s2.0: 2,1 Mbit/s3.0: 480 Mbit/s

100 mW2.5 mW1 mW

Class 1: 100 mClass 2: 10 mClass 3: 1 m(in the open)

BluetoothIEEE802.15.1

Infrared850-900 nm

IrDA 1.0: 9,6 – 115 kBits/sIrDA 1.1: up to 16 MBit/s

40-500 mW/Sr

1 mIrDA

a: 5,15-5,75 GHz, 19 ch.;b: 2,4-2,4835GHz, 13 ch. in Europe;h: 2,4-2,4385 GHz, 13 ch. in Europe

IEEE 802.11 2 MbpsIEEE 802.11a: 54 Mbpsb: 11 Mbpsg: 54 Mbpsh: 54 Mbpsn: 600 Mbps

30 mW(802.11a)Max. 1000

mW(restricted)

100-300 m in the open, 80 m

in buildings

WLAN(IEEE802.11-Family)

2.46 GHzFallback 868 MHz

20-250 kbit/s1…10 mWApprox. 100 mZigBee

FrequencyBandwidthTransmitter power

RangeName

-676-

fore, a viable safety concept emerges. In [245] it is pointed out how redundancy and modularisation of the system can significantly increase reliability. Basic safety-related standards include IEC 61508, IEC 62061, ISO 13849 and also in terms of availability in general VDI 4004, IEC 61070. 4.3 Variations / Concepts of Systems integration This section contains an analysis of system components considered from the point of view of integration in the ma-chine. The introduction of chapter 4 describes, via Fig. 38, the integration of one component in detail. Figure 40 classifies different components by their types of integra-tion, such as modular and integrated characteristics with respect to different aspects, for example physical inte-gration and functional integration. In physical integra-tion, defined or even standardised mechanical, electrical and information technology interfaces support the Plug & Play principle. This initiative is particularly interesting from the point of view of the increasing reconfigurability of ma-chine tools. A further reduction in the complexity of integration can be achieved by means of independent energy supply and decentralised computing units. The functional integration takes account of the task of the mechatronic component as well as the degree of information exchange with the higher-level system. Functionally autonomous compo-nents do not need to have resources in the higher-level system in order to fulfil their task. They therefore possess a functionally independent sensor system, their own con-trol algorithms and an independent actuator system. If the dependency on resources of the higher-level system in-creases, the degree of cooperation rises. An example of co-operating components is the NC motion separation in redundant axes. However, it should also be emphasised that most solu-tions are hybrid forms of the aforementioned integration aspect.In contradiction to pure software integration are solutions with their own hardware which contain at least one ambi-ent temperature sensor and frequently a measuring sys-tem, and which undertake corrections by means of addi-tional actuators. A system like this can be constructed in modules, but is typically a highly integrated special solu-tion for the respective particular application, as outlined in section 2.3. Independent or modular solutions are mostly very local in their effect or are limited to a single axis direction. The close integration with the machine is linked to the fact that the temperature response is evoked throughout the struc-ture. The actuators using existing drives represent an ex-

tremely favourable solution in terms of cost and only compensation in axes directions not realised on the ma-chine require their own actuators. Modular autonomous components: In [196] an autonomous vibration-damped circular saw blade with integrated piezo actuators is presented (Fig. 41). The pie-zoceramic elements integrated in structural conformity are connected to passive or semi-active electrical net-works supplied via a generator on the rotating shaft. En-ergy supply, actuators, sensors and electrical networks are thus completely independent from the rest of the ma-chine tool. The standardised tool holder serves as the mechanical interface.

A vibration suppression concept for boring processes is presented by [188]. It involves a semi-active damped tool holder with magneto-rheological fluids. The concept is modular and can be integrated via standardised tool holder. Other initiatives for integration of actuators espe-cially for damping of chatter on main spindle tool systems are presented in [159][161]. The adaptronic spindle unit presented in [178] is a further example of a modular and autonomous mechatronic component. In principle, the concept already functions autonomously and independently of the machine controls. Three piezo actuators (Fig. 42) permit positioning of the tool in three spatial directions and thereby enable active compensation of milling head displacement and vibra-tions.Figure 42B shows the surface measured with a laser pro-filometer of a workpiece made with the adaptronic milling spindle. This has involved displacements of the milling head of 50 μm and 220 μm. It shows that precision ad-justments in the μm range are possible with a high de-gree of repeatability.

Figure 41: Adaptive circular saw blade [196].

physical integration

func

tiona

l int

egra

tion

highly integrated“special solution”

modularity“plug and play”

Figure 40: Aspects of system integration 1: Saw blade [196], 2: AdSpin [40][178], 3: Mechatronic Chuck [184], 4:

Model based compensation [213], 5: HDM [106], 6: Dynapod [82], 7: Axis [248], 8: High Speed Machine

[120].Figure 42: Adaptronic spindle unit for milling machines

[178].

-677-

A mechatronic chuck is an example of a mechanically modular integrated component with wireless energy and data transmission [184]. The communication between a control computer and the mechatronic chuck is imple-mented via Bluetooth. Radio transmission enables inter-ference-free transmission of actuating signals, even dur-ing processing. The most favourable options for commu-nications between the control computer of the mecha-tronic component and the machine tool control system are connections via LAN or WLAN. Modular co-operating components: In the following analysis, modular components are presented which dem-onstrate an increased degree of cooperation with the higher-level system. These basically include modular axes systems and technology modules. An axis module is presented in [246]. It possesses de-fined mechanical, electrical and information technology interfaces. However, functionally it must cooperate closely with the basic system, because as an NC axis it forms part of the basic system, see Fig. 43. By the use of decentralised periphery and integration of the inverter in the module, the electrical supply cables are reduced to a bus system and two energy supply cables. The individual axis modules are connected via the bus system using in-formation technology when the processing system is con-figured. A procedure for creating mechatronic modules is shown in [248].

Similarly co-operative in nature, is the agile hexapod module of a machine tool with redundant axes described in Chapter 3.1.1 [82]. It can be exchanged for other proc-essing platforms. This permits further adaptation to the process requirements. From an information technology point of view the modules are integrated via SERCOS Interface in the andronic 2000 control system. The parti-tioning of movement takes place at interpolator speed in such a way that for a particular point in time a corre-sponding partitioning for two moving components is pos-sible, based purely on the past axis movement [83][247] (Fig. 44). It is advantageous here to comply simultane-ously with all the restrictions on the axis involved with re-spect to range of movement, axis velocity, axis accelera-tion and axis torque [83].

The starting point is a slow drive structure )(tsS with significantly more inertial behaviour, but which guaran-tees the required range of movement, together with an agile drive structure )(tsA with a significantly restricted range of movement which guarantees the required movement dynamics. According to the principle of a mas-

ter-slave connection, the sluggish element is always try-ing to catch up with the movement of the dynamic ele-ment and increasingly takes over the total movement

)(tsO and total velocity without infringing the restrictions in terms of dynamics for this axis, see Fig. 44. By this means, the paths of the more agile components are again restricted and are independent of the total range of move-ment.Highly integrated autonomous components: Highly integrated systems are primarily characterised by de-creasing flexibility with regard to mechanical and energy-related interfaces. In order to realise their function, how-ever, they are not in principle reliant on the higher-level system from the information technology perspective. In [143] an adaptronic component for vibration control of a machine tool with scissor-like parallel kinematics has been developed. Sensors and actuators are included in the active strut and all functions of the device are com-pletely autonomous. Different local and model-based con-trol algorithms, developed with a flexible multibody sys-tem model, can be implemented on a digital signal proc-essor (DSP) platform to investigate their effect on the dy-namics of the machine tool. Highly integrated co-operating components: These basically include all fundamental functions of the main system which are characterised by minimal interface defi-nition (minimal flexibility). An important example is pro-vided by NC axes, which must of necessity communicate with the other axes in order to be able to realise depend-ent movement. Solutions for feed systems which go fur-ther than this basic function include active inertial force equalisation and jerk impulse decoupling. Firmly inte-grated in the system, they need data on the movement of the NC axis. In the high-speed lathe with active inertial force compen-sation (see also Chapter 3.1), an open control system has been used [42]. The control system is distinguished by higher dynamics and improved contour description com-pared with current control systems. The system is based on the Fire-Wire-Bus. This open bus system offers the option of synchronised transmission of data from I/O components and optical systems as well as the data re-quired for the control system. Position-related data are transmitted by the control system as cubic splines of vari-able duration, which can only be transmitted in the posi-tion control circuit of the servo in digital form at a fre-quency of 4000 Hz. At the same time, the position signals from other mechatronic components can be evaluated. Modular integration of an active tool holder is possible. In order to realise a dependent movement, however, an ex-change of information with the main system is required. A highly integrated active magnet guide system is de-scribed in [120]. Its primary function, to guide, is autono-mous in nature. However, a more detailed examination is required of the secondary functions, which demand in-creased co-operation with the basic system. The com-plete mechanical integration of magnetic guidance in the flow of force of the machine already offers extensive op-

calculation:inertial component

with limited dynamics++

+ _sO sA

sI sO

result:agile component

with limited strokeIOA sss −=

),( OIOI sssfs −=A general description of reactive partitioning of dynamism

Figure 44: Reactive partitioning of dynamics [83][247].

Figure 43: Functional integration by harmonisation of structural and functional system limits [246].

-678-

portunities for influencing the behaviour of the overall sys-tem.In the context of first explorations, as shown in [120], clear increases in accuracy have already been achieved. The positioning error is given out at 333 Hz and read in at the controller speed of 5 kHz. In the milling process, for instance, the increases in precision achieved by this structure permit a reduction in errors of circularity of up to 44%. In order to investigate the positioning accuracy, test workpieces were milled on the machine and their geome-try then was examined with a coordinate measurement machine (Fig. 45). Further options are offered by the use of integrated sensor systems for process monitoring. Be-cause of the friction-free design of the component, the motor current from the drive control system for the Z-axis can be used for diagnostic purposes. In conjunction with the force components of the remaining five degrees of freedom, in the guidance control systems, a process force analysis with high precision is possible. These data can again be used to influence the control system, for exam-ple in stopping the machine.

In addition to the components described for cutting ma-chine tools, the aspects of integration with reference to two solutions for forming machines (see also Chapter 3) are described below. Both solutions are highly inte-grated into the basic system. The first solution examined expands on the designs for compensation on press brakes [213]. In the system of Beyeler [213], all force measurements are accomplished by precision pressure measurements in

the cylinders, while Lutters et al. [217] prefer the meas-urement in the lower beam by piezo dowels, which are calibrated. The material data in the process model need to be identi-fied and ideally stored within a database of the machine control system (Fig. 46). Mentink et al. [249] favour the use of bending-dedicated parameters from bending tests rather than general material-related data. YANG et al. [250][251] use neural networks for self- learning systems and fuzzy modified models to describe the influences of different parameters on the force-stroke curves. Mentink et al. [249] also discuss the self-adaptation of the process to in-process measurements of the force – stroke curve. The second solution presented is devoted to the integra-tion of multi-point drawing technology (cf. Fig. 47). The 16 independent drawing cushion axes of a multi-point cush-ion presented in [222] can be used as passive pressure axis for a drawing cushion function, or as opposing axis for holes or punchings. The openness of NC controls (OEM-NCK interface) enables the introduction of inde-pendent cushion-type control structures which include and use existing NC functionalities.

4.4 Conclusion and Outlook of System Integration In the preceding sections, components and their integra-tion within systems have been examined. • The mechanical integration of components is at an

advanced stage in respect of optimisation and im-provements.

• Electrical integration could be promoted by means of standards which are applicable at industry level rather than at corporate level.

• The domain of information technology has special po-tential in terms of future integration.

The trends which can be observed in production, such as for example the shortening of the product life cycle, de-creases in batch sizes and an increase in the number of product variants, run counter to the life and the cost-intensive nature of capital goods [2]. As a result future production locations will need to be more flexible, both in terms of the working steps and the spatial installation, which must be accomplished at the lowest possible cost. This will require modular compo-nents which are flexible from a mechanical and electrical point of view and in terms of information technology. In order to be able to evaluate the vast amounts of data generated in a rational manner, software structures will be required which are capable of making an intelligent choice from a large volume of information and of taking

10 μm

D = 25 mm

x

y

ae

ap

1 Without Compensation2 With Compensation

1,74 μm2

Mean Circular Deviation

9,3 μm2

16,8 μm1

3,63 μm1

Deviation on Roundness

Processae= 50 μmap= 2 mmfz = 0,1 mmvf = 8 m/minvc = 1005 m/minn = 40.000 rpmMaterialAl 7449

Figure 45: Measuring the circularity of a milled test work-piece [120].

pres

sure

programmedsheet thickness

new sheetthickness

punch travel

Figure 46: In-process measurement of the sheet thick-ness as an input parameter in the process model by sam-

pling the pressure-load-free pre-stroke [213].

16 p

oint

die

cus

hion

NCUE/R

pressure

pressure- and power control

HLA HLAHLA HLAHLA HLAHLA HLA

SPS

32 x pressuresensor

16 x position measurement system (absolut)

16 x nominal(±10V)

NCU-Link to cushion main drive

16 x cylinder

Sinumerik 840 D

VISU PCWINCC

safety control

PCMCIA card

Figure 47: Structure of the control of multi-point cushions [128][222].

-679-

the appropriate action as a result. AI (artificial intelli-gence) is one of the possible initiatives in this area. Pioneering modular systems have already been devel-oped in robotics. Highly integrated products for mecha-tronic systems include for example the “Powercube” module [252]. These third-generation modular compo-nents constitute individual axis. Development of the first generation of these components was carried out in the Lissy project [253]. The axes can be combined with one another by simple manual intervention. The modules are supplied with power and data by means of a special plug connection. The software supplied with the product en-ables the geometric elements to be assembled and sub-sequently controlled in a computer. The modules are fit-ted with encoders and force measurement devices. The development of similar lightweight construction com-ponents has been undertaken for space flight, weight op-timisation being of crucial importance, and a ratio of 1 to 1 of force to own weight was achieved. An arm with 7 DOF weighs approx. 14 kg [255]. The arms and modules are designed in such a way that they detect the slightest dis-turbances and achieve adjustable compliance behaviour via high-speed control cycles. These prototypes incorpo-rate force-momentum sensors, encoders on the transmis-sion output side and their own drives with power electron-ics. A superposed joint controller is realised via a DSP with a speed of 330 μs, and the motor controller operates at a speed of 25 μs. The joint DSP communicates with the higher-level control unit via the SERCOS Bus. With suitable central controls, self-configuring production cells could, in future production locations, control a com-plete production chain. Predictions about machine tooIs in the year 2010 [2] make mention of the “digital factory”. In such a factory, in which all the components are kept in-formed of the status and the current production step of the other machines, manufacturing could be flexible and production completed swiftly. Concepts such as “Just in Time” production or production for stock, which depends on suppliers and inventory, could be automatically initi-ated and adapted. Requirements: With a shared communications standard between all sensors and machines, these could be con-trolled and evaluated by a central intelligence. Defined System Limits: System limits play a crucial role in any discussion about integration of components in me-chanical systems. Defined system limits could simplify the integration of components in machine systems in the fu-ture.Some standards in the field of mechanical integration are already existing [173]. Components are frequently de-signed with the clear aim of being mechanically modular or integrated. This is a function of the component itself and its intended purpose. For this reason, defined system limits are not necessary in all components, but could be very useful in most of them. Definition of system limits in relation to energy transfer is however often desirable. When considering uniform volt-age for connecting components it is noticeable that domi-nance in this respect is determined primarily by the mar-ket. This applies in particular to the mechanical plug con-nection to the machine or the control system. One field which has great potential for future integration is information technology. Intelligent data evaluation and data transfer are associated with this. Defined system lim-its in the form of uniform data formats, bus structures [256][257] and standardised wireless transmission links could be used to define a universal communications standard.

A recent project in this area [256] constitutes a technical leap forward in the direction of key functions suitable for routine application. The developments include compo-nents, reference architectures, convergence of technolo-gies via a technology platform and spin-offs and products. By analysing applications, it will be possible to identify the required characteristics of the reference architecture, pro-cedures and components to be developed. A further project Papas [257] was concerned, amongst other things, with the specification of draft designs for “Plug and Play” modules. A model-based control proce-dure with real-time model calculations was produced us-ing resources which included the open-source modelling language Modelica. It is usually companies who set or prescribe defined stan-dards by virtue of their market penetration. Definition of an “open” interface for developers and other suppliers of mechatronic components would open the market for new developments. This opposes efforts in the market econ-omy by companies who control the market.

5 CONCLUSION Mechatronics enables a supreme development method-ology for machine tools. The technological developments in the individual domains of the basic structure – mechan-ics/materials, transformation systems – sensor/actuator systems and data processing will characterise the future development of “intelligent” machine tools. Over the next few years, the emerging trends will be the increasing use of self-optimising, in part adaptronic components and the use of ever more efficient control systems for model-supported compensation of machine errors and process control. All functionality of machines will become elec-tronically enhanced and thus mechatronic functions. Reconfigurability as a necessary pre-requisite for the flexibility of machine tools demands mechatronic machine tool components with Plug & Play functionality. This re-quires the creation and standardisation of interfaces which are uniform in terms of their mechanical, energy and information technology aspects. These machine tool components must at the same time be strictly function-oriented in their design. They must be subjected to a comprehensive description of their functionality in terms of both hardware and software structures. Analogies with developments in computer technology or robotics will be-come more prominent. In our vision of the future, the boundaries between robot-ics and machine tools will become diffuse in mechatronic manufacturing resources. This trend will go hand in hand with a fundamental transformation in terms of kinematic structures, but above all in the architecture of control sys-tems. One of the major challenges will be to guarantee the reliability of these mechatronic systems in order to meet economic demands in terms of the availability of production systems. The challenge is to fully exploit mechatronic options and capabilities already in the earliest design phase. The de-sign tools for such mechatronic systems cannot be re-duced simply to the sum of modelling, simulation of indi-vidual domains and securing appropriate exchange of data. The task of design technologies which are today subsumed under the heading of “virtual machine tool” [21] will undergo fundamental expansion. Beyond evaluating the standard of quality met by the specification profile, we shall in the future be guiding the creation in the virtual world of configurations of hardware and software which are already essential: with the implementation of a tech-nological specification profile achieved by integrating mechatronic modules, the software for controlling the en-

-680-

tire system will be created automatically, as will the soft-ware of corresponding agents in the components. Optimi-sation tools will achieve the optimum division of functions between components, in terms of energy criteria also. At the same time, a reliability model will be produced which will take account of all the available sensors in the sys-tem. This will ensure, what is inherently retrievable in any operating status in mechatronic systems, the availability of the production system. This will be achieved by means of a certain degree of hardware and software redundancy and by means of the extensive knowledge of the current status of the system. As a result, the transformation of the paradigm “reliability despite mechatronics” into “reliability by means of mechatronics” will be complete.

6 ACKNOWLEDGEMENTS The authors would like to thank all the members of the CIRP Editorial Committee for their reviews and com-ments. We gratefully received contributions of a number of colleagues (in alphabetical order and with CIRP mem-bers denoted by *): Prof. Y. Altintas*, Prof. T. Aoyama*, Prof. B.-A. Behrens*, Prof. C. Brecher*, Prof. D.B. Debra*, Prof. K. Erkorkmaz*, Prof. J. Fleischer*, Prof. P. Groche*, Prof. K. Großmann, Prof. U. Heisel*, Prof. J. Hesselbach*, Prof. F. Jovane*, Dr. X. Lu, A. Paul, Prof. G. Pritschow*, Prof. H. Shinno*, Prof. G. Tani*, Prof. E. Uhlmann*, Prof. M. Weck*, Prof. K. Weinert* and Dr. M. Zatarain*. This paper would not have been realized with-out the dedicated effort of Dr. W.-G. Drossel, Dipl.-Ing. M. Wabner, Dipl.-Ing. A. Wedler, Dipl.-Ing. J. Immel, Dipl.-Ing. Ihlenfeldt, Dr. T. Päßler, Dipl.-Ing. H.-C. Möhring and Dipl.-Ing. C. Will.

7 REFERENCES [1] European Commission, 2004, Manufuture: A Vision

for 2020, Report of the High-Level Group. [2] Reinhart, G., Zielasko, W., Spath, D., 2002, Ab-

schlussbericht: Werkzeugmaschine 2010: Vorunter-suchung zum Thema Werkzeugmaschine 2010, Karlsruhe.

[3] Chryssolouris, G., 2005, Manufacturing Systems: Theory and Practice, Springer-Verlag, New York.

[4] Byrne, G., Dornfeld, D., Denkena, B., 2003, Ad-vancing Cutting Technology, CIRP Annals, 52/2:483-507.

[5] Neugebauer, R., Drossel, W.-G., Roscher, H.-J., Schwaar, M., 2002, Intelligent Strut, 3rd Chemnitz Parallel Kinematics Seminar, Chemnitz, 23.-25.04.2002:17-47.

[6] Jovane, F., Boër, C. R., Koren, Y., 2003, Present and Future of Flexible Automation: Towards New Paradigms, Annals of the CIRP, 52/2:543-561.

[7] VDI 2206 Design Methodology for Mechatronic Sys-tems, Beuth-Verlag GmbH, Berlin.

[8] Van Brussel, H. M. J., 1996, Mechatronics – A Powerful Concurrent Engineering Framework, IEEE/ASME Transactions on Mechatronics, 1/2:127-136.

[9] Janocha, H., 1999, Adaptronics and Smart Structu-res, Springer-Verlag Berlin.

[10] www.pt-piesa.tu-chemnitz.de. [11] www.tu-braunschweig.de/adaptronik/

ausschreibung. [12] Moore, G. E., 1965, Cramming More Components

onto integrated Circuits, Electronics 38(8):114–117.

[13] Van Brussel, H., Sas, P., Németh, I., De Fonseca, P., Van den Braembussche, P., 2001, Towards a Mechatronic Compiler, IEEE/ASME Transactions on Mechatronics 6(1):90-105.

[14] Kallenbach, E., Zöppig, V., Birli, O., Feindt, K., Ströhla, T., Saffert, E., Schmidt, J., 2001, Integrati-on mechatronischer Systeme, 4. VDI Mechatronik Tagung Innovative Produktentwicklungen, Franken-thal, VDI-Berichte 1631, Düsseldorf, 57-72.

[15] Gausemeier, J., Grasmann, M., Kespohl, H.D., 1999, Verfahren zur Integration von Gestaltungs- und Berechnungssystemen, VDI-Berichte 1487, Düsseldorf, 71-87.

[16] Albers, A., Matthiesen, S., 2002, Konstruktionsme-thodisches Grundmodell zum Zusammenhang von Gestalt und Funktion technischer Systeme - Das Elementmodell, Wirkflächenpaare & Leitstützstruk-turen, zur Analyse und Synthese technischer Sys-teme, Konstruktion 54, 7/8: 55-60.

[17] Daenzer, W .F., Huber, F., 1994, Systems Enginee-ring - Methoden und Praxis, 8. Auflage, Zürich, Ver-lag Industrielle Organisation.

[18] Bathelt, J., Jönsson, A., Bacs, C., Dierssen, S. Meier, M., 2005, Applying the New VDI Design Guideline 2206 on Mechatronic Systems Controlled by a PLC, ICED05 Int. Conf. on Eng. Design, Mel-bourne, Australia, 415-416.

[19] Möhringer, S., Gausemeier, J., 2002, An Interface Specification for Principle Solutions Supporting the Cross-Domain Design of Mechatronic Systems, Proc. of the 7th Int. Design Conf. Dubrovnik, Croatia.

[20] Weber, C., Muth, M., Bär, T., Werner, H., 1999, In-telligente Kopplung von Gestaltung und Berech-nung, VDI Berichte 1487, Düsseldorf, 47-69.

[21] Altintas, Y., Brecher, C., Weck, M., Witt, S., 2005, Virtual Machine Tool, Annals of CIRP, 54/2:651-673.

[22] Dierssen, S., 2002, Systemkopplung zur kompo-nentenorientierten Simulation digitaler Produkte, Diss. ETH Zurich.

[23] Zaeh, M.F., Poernbacher, C., 2005, A Model-Based Method to Develop PLC Software for Machine Tools, Annals of the CIRP, 54/1:371-374.

[24] Zirn, O., Weikert, S., 2006, Modellbildung und simu-lation hochdynamischer Fertigungssysteme, Sprin-ger, Berlin, Heidelberg, New York.

[25] Jedrzejewski, J., Kowal, Z., Kwasny, W., Modrzycki, W., 2005, High-speed Precise Machine Tools Spin-dle Units Improving, J. Materials Processing Tech-nology 162-163:615-621.

[26] Weikert, S., Knapp, W., 2004, R-Test, a New De-vice for Accuracy Measurements on Five Axis Ma-chine Tools, Annals of the CIRP, 53/1:429-432.

[27] Schwenke, H., Franke, M., Hannaford, J., 2005, Er-ror Mapping of CMMs and Machine Tools by a Sin-gle Tracking Interferometer, Annals of the CIRP, 54/1:475-478.

[28] Pritschow, G., Roeck, S., 2004, Hardware in the Loop Simulation of Machine Tools, Annals of the CIRP, 53/1:295-298.

[29] Seo, K., Fan, Z., Hu, J., Goodman, E.D., Rosenberg, R.C., 2003, Toward a Unified and Auto-mated Design Methodology for Multi-domain Dy-namic Systems Using Bond Graphs and Genetic Programming, Mechatronics 13:851-885.

-681-

[30] Choudhury, S.K., Goudimenko, N.N., Kudinov, V.A., 1997, Online Control of Machine Tool Vibration in Turning, International J. of Machine Tools and Manufacture, 37/6:801-811.

[31] Henmi, N., Murata, R., 1998, Vibration Control in Cutting Process by an Active Work-rest, Int. J. of the Japan Society for Precision Eng., 32/2:133-134.

[32] Abele, E., Kreis, M., Roth, M., 2006, Electromag-netic Actuator for in Process Non-Contact Identifica-tion of Spindle-Tool Frequency Response Func-tions, Conf.-Speech, CIRP 2nd Int. Conf. on High Performance Cutting, 12.-13. Jun., Vancouver, Can-ada.

[33] Altintas, Y., Weck, M., 2004, Chatter Stability of Metal Cutting and Grinding, Annals of the CIRP, 53/2:619-642.

[34] Zatarain, M., Ruiz de Argandoña, Illaramendi, A., Azpeitia, J.L., Bueno, R., 2005, New Control Tech-niques on State Space Observers for Improving the Precision and Dynamic Behaviour of Machine Tools, Annals of the CIRP, 54/1:393-396.

[35] Pritschow, G., Eppler, C., Lehner, W.-D., 2003, Fer-raris Sensor - The Key for Advanced Dynamic Drives, Annals of the CIRP, 52/1: 289-292.

[36] Zirn, O., Batzies, E., Weikert, S., Schöller, T., 2006, State Control of Servo Drives with Flexible Struc-tural Components, IEEE Industry Applications Conf. 41, Tampa, Florida, USA, 1760-1766.

[37] Symens, W., Van Brussel, H., Swevers, J., 2004, Gain-scheduling Control of Machine Tools with Varying Structural Flexibility, Annals of the CIRP, 53/1:321-324.

[38] Altintas, Y., Erkorkmaz, K., 2003, Feedrate Optimi-zation for Spline Interpolation in High Speed Ma-chine Tools, Annals of the CIRP, 52/1:297-302.

[39] Hadorn, M., 2004, A Model-Based NC Postprocess-ing System for Parallel Kinematic Machine Tools, Proc. 4th Chemnitz Parallel Kinematics Seminar, 24:291-312.

[40] Denkena, B., Holz, C., 2006, Advanced Position and Force Control Concepts for the Linear Direct Driven Hexapod PaLiDA, Proc. 5th Chemnitz Paral-lel Kinematics Seminar, 33:359-378.

[41] Jaeger, C., Hadorn, M., Wegener, K., 2007, Modell-basierte Steuerungskonzepte für Werkzeugmaschi-nen, WB, to be published.

[42] Brecher, C., Hirsch, P., 2004, Compensation of Thermo-elastic Machine Tool Deformation Based on Control Internal Data, Annals of the CIRP, 53/1:299-302.

[43] Mitsuishi, M., Warisawa, S., Hanayama, R., 2001, Development of an Intelligent High-Speed Machin-ing Center, Annals of the CIRP, 50/1:275-280.

[44] Sriyotha, P., Nakamoto, K., Sugai, M., Yamazaki, K., 2006, Development of 5-Axis Linear Motor Driven Super-Precision Machine, Annals of the CIRP, 55/1:381-384.

[45] www.hueller-hille.com. [46] Heisel, U., Koscsák, G., Stehle, T., 2006,

Thermography-Based Investigation into Thermally Induced Positioning Errors of Feed Drives By Ex-ample of a Ball Screw, Annals of the CIRP, 55/1:423-126.

[47] Yang, H, Ni, J., 2005, Dynamic Neural Network Modelling for Nonlinear, Nonstationary Machine

Tool Thermally Induced Error, Int. J. Machine Tools and Manufacture, 45:455-465.

[48] Chen, J.-S., 1996, Neural Network-based Modelling and Error Compensation of Thermally-induced Spindle Errors, Int. J. Advanced Manufacturing Technology, 12:303-308.

[49] Yang, Q.D., Van den Bergh, C., Vanherck, P., Kruth, J.-P., 1999, Study of the Thermal Deforma-tions on a Coordinate Measuring Machine and Compensation via a Neural Network Model, Int. J. Flexible Automation and Integrated Manufacturing, 7:129-147.

[50] Tseng, P.-C., Ho, J.-L., 2002, A Study of High Pre-cision CNC Lathe Thermal Errors and Compensa-tion, Advanced Manufacturing Techn., 19:850-858.

[51] Weck, M., McKeown, P., Bonse, R., Herbst, U., 1995, Reduction and Compensation of Thermal Er-rors in Machine Tools, Annals of the CIRP, 44/2:589-598.

[52] Attia, M.H., Fraser, S., 1999, A Generalized Model-ling Methodology for Optimized Real Time Com-pensation of Thermal Deformation of Machine Tools and CMM Structures, Int. J. Machine Tools and Manufacture, 39:1001-1016.

[53] Bansevicius, R., Giniotis, V., 2002, Mechatronic Means for Machine Accuracy Improvement, Mecha-tronics, 12:1133-1143.

[54] Wurst, K-H., Heisel, U., Kircher, C., 2006, (Re)konfigurierbare Werkzeugmaschinen – not-wendige Grundlage für eine flexible Produktion, wt-online, 96:257-265.

[55] Landers, R.G., Min, B.-K., Koren, Y., 2001, Recon-figurable Machine Tools, Annals of the CIRP, 50/1:269-274.

[56] Pritschow, G., Kramer, C., 2005, Open System Ar-chitecture for Drives, Annals of the CIRP, 54/1:375-378.

[57] Oldknow, K.D., Yellowley, I., 2001, Design, Imple-mentation and Validation of a System for the Dy-namic Reconfiguration of Open Architecture Ma-chine Tool Controls, Int. J. Machine Tools & Manu-facture, 41:795-808.

[58] Chen, L., Xi, F., Macwan, A., 2005, Optimal Module Selection for Preliminary Design of Reconfigurable Machine Tools, J. Manufacturing Science and Engi-neering, 127:104-115.

[59] Yigit, A.S., Ulsoy, A.G., 2002, Dynamic Stiffness Evaluation for Reconfigurable Machine Tools In-cluding Weakly Non-linear Joint Characteristics, Proc. Int. Mech. Eng., 216/B:87-101.

[60] Katz, R., Yook, J., Koren, Y., 2004, Control of a Non-Orthogonal Reconfigurable Machine Tool, J. Dynamic Systems, Measurement and Control, 126/2:397-405.

[61] Lorenzer,T., Kunz, A., Wegener, K., 2007, Rekonfi-gurierbare Maschinensysteme - Produktionsmittel mit Zukunft, WB, to be published.

[62] Rzevski, G., 2003, On Conceptual Design of Intelli-gent Mechatronic Systems, Mechatronics, 13:1029-1044.

[63] www.hyprostatic.de. [64] Shinshi, T., Shimokohbe, A., Sato, K., Izumi, S.,

Hashimoto, J., 2000, Magnetic Lead Screw Mecha-nism and its Application to Linear Positioning Sys-tem, 7th Int. Symposium on Magnetic Bearing, Switzerland, 431-436.

-682-

[65] Tachikawa, H., Fukuda, M., Shinshi, T., Sato, K., Shimokhbe, A., 1997, Precision Positioning Using Air Bearing Lead Screw, Int. Conf. on Micromecha-tronics for Information and Precision Equipment, Tokyo, 238-243.

[66] Berkemer, J., Altenburger, R., Koch, T., Lehner, W.-D., 2004, Effective Use of the Dynamic Potentials of Direct Drives, wt-online, 94/05:233-241.

[67] Pritschow, G., 1998, A Comparison of Linear and Conventional Electromechanical Drives, Annals of CIRP, 47/2:541-548.

[68] Großmann, K., Müller, J., Schween, A., 2001, Mik-ro-Achse als Zusatzaggregat für Großdrehmaschi-nen, ZWF, 96/9:470-473.

[69] Gloess, R., 2006, Aktive Piezosysteme im Maschi-nenbau, Conf.-Speech, MaVo-FASPAS-Workshop, Dresden.

[70] Takeuchi, Y., Sakaida, Y., Sawada, K., Sata, T., 2000, Development of 5-axis Control Ultraprecision Milling Machine for Micromachining Based on Non-friction Servomechanism, Annals of the CIRP, 49/1:295-298.

[71] Van Brussel, H., Reynaerts, D., Vanherck, P., Ver-steyhe, M., Devos, S., 2003, A Nanometre-precision, Ultra-stiff Piezostepper Stage for ELID-grinding, Annals of the CIRP, 52/1:317-322.

[72] Shamoto, E., Murase, H., Moriwaki, T., 2000, Ultra-precision 6-axis Table Driven by Means of a Walk-ing Drive, Annals of the CIRP, 49/1:299-302.

[73] Uhlmann, E., Bayat, N., 2006, High Precision Posi-tioning with Ferrofluids as an Active Medium, An-nals of the CIRP, 55/1:415-418.

[74] Aoyama, T., Inasaki, I., 1997, Application of Elec-trorheological Fluid Dampers to Machine Tool Ele-ments, Annals of the CIRP, 46/1:309-312.

[75] Shinno, H., Hashizume, H., Sato, H., 1999, Nano-meter Positioning of a Linear Motor-Driven Ultra-precision Aerostatic Table System with Electror-heological Fluid Dampers, Annals of the CIRP, 48/1:289-292.

[76] Uhlmann, E., Bayat, N., 2005, Intelligente magne-tofluidische Dämpfer, Futur 2:10-11.

[77] Hafla, A., Pritschow, G., 2005, Adaptronische Schwingungsdämpfung in Präzisionsmaschinen, Fortschritt-Berichte VDI (2), Fertigungstechnik, VDI-Verlag, 29-39.

[78] Neugebauer, R., Wittstock, V., Gebhardt, S., Schönecker, A., 2005, Strukturintegrierte Aktorik im Maschinenbau, Conf.-Speech, Adaptronic Congress Göttingen.

[79] Simnofske, M., Hesselbach, J., 2006, The Increase of the Dynamic and Static Stiffness of a Grinding Machine, Conf.-Spech, ASME IDETC/CIE 2006 (Int. Design Eng. Techn. Conf. & Comp. and Information in Eng. Conf.), 99740, Philadelphia, USA.

[80] Neugebauer, R., Drossel, W., Ihlenfeldt, S., 2004, Components and Equipment for Intelligent Produc-tion Systems, Proc. of MDP-8, Conf. on Mech. De-sign and Production Cairo, Egypt, Jan. 4-6, 57-60.

[81] Altintas,Y., Woronko, A., 2002, A Piezo Tool Actua-tor for Precision Turning of Hardened Shafts, An-nals of CIRP, 51/1:303-306.

[82] Neugebauer, R., Ihlenfeldt, S., Kirchner J., Adam, F., 2000, Redundante Hybridkinematik für den Großwerkzeug- und Formenbau, 2nd Chemnitz Pa-rallel Kinematics Seminar, 153-168.

[83] Neugebauer, R., Schröder, T., Krabbes, M., 2005, Optimization of a Trajectory Splitting Algorithm for Machine Tools with Hierarchical Drive Structures, 50. Int. Wiss. Koll. der TU Ilmenau, Sept. 19.-23., 153-154.

[84] Lu, X., Trumper, D.L., 2005, Ultrafast Tool Servos for Diamond Turning, Annals of the CIRP, 54/1:383-386.

[85] Kouno, E., 1984, A Fast Reponse Piezoelectric Ac-tuator for Servo Correction of Systematic Errors in Precision Machining, Annals of the CIRP, 33/1:369–372.

[86] Cuttino, J.F., 1999, Performance Optimization of a Fast Tool Servo for Single-point Diamond, IEEE/ASME Trans. on Mechatronics, 4/2:169-179.

[87] Rasmussen, J.D., 1994, Dynamic Variable Depth of Cut Machining Using Piezoelectric Actuators, Int. J. of Machine Tools and Manufacture, 34/3:379–392.

[88] Patterson, S., Magrab, E., 1985, The Design and Testing of a Fast Tool Servo for Diamond Turning, Precision Engineering, 7/3:123-128.

[89] Dow, T.A., Miller, M.H., Falter, P.J., 1991, Applica-tion of a Fast Tool Servo for Diamond Turning of Nonrotationally Symmetric Surfaces, Prec. Eng, 13/4:243-250.

[90] Weck, M., Mehlkopp, K., Terwei, T., Özmeral, H., 1995, A New Hybrid Concept for a Long Stroke Fast Tool Servo System, Proc. of ASPE, 10:93-98.

[91] Okazaki, Y., 1998, Fast Tool Servo System and its Application to Three Dimensional Fine Surface Fig-ures, Proc. of ASPE, 17:100-103.

[92] Eda, H., 1992, Ultra-precise Machine Tool Equipped with a Giant Magnetostriction Actuator - Development of New Materials and Their Applica-tion, Annals of the CIRP, 41/1:421-424.

[93] Liu, D., 1998, Surface Texture Improvement in the Turning Process via Application of a Magnetostric-tively Actuatored Tool Holder, ASME J. of DSMC, 120, 193-199.

[94] www.nanowave.co.jp [95] Douglass, S., 1983, A Machining System for Turn-

ing Non-axis-symmetric Surfaces, PhD dissertation, the University of Tennessee, Knoxville.

[96] Greene, W., Shinstock, D., 1997, Design of a Linear Voice Coil Actuator for Fast Tool Servo Applica-tions, Proc. of ASPE 16:187-190.

[97] Ludwick, S., Trumper, D.L., 1999, Design of a Ro-tary Fast Tool Servo for Ophthalmic Lens Fabrica-tion, Precision Engineering, 23/4:253-259.

[98] Montesanti, R.C., Trumper, D.L., 2003, High Band-width Short Stroke Rotary Fast Tool Servo, Proc. of ASPE, 30:115-118.

[99] Todd, M.W., Cuttino, J.F., 1997, Development of a Long Range, Traction Drive Fast Tool Servo for Diamond Turning Applications, Proc. of ASPE, 16:398-401.

[100] Lu, X., Trumper, D.L., 2003, Electromagnetically Driven Fast Tool Servo, Proc. of ASPE, 103-106.

[101] Montesanti, R.C., 2005, High Bandwidth Rotary Fast Tool Servos and a Hybrid Rotary/Linear Elec-tromagnetic Actuator, PhD dissertation, M.I.T. Cambridge.

[102] www.mikromat-wzm.de

-683-

[103] Weck M., Hennig J., Wetter O., 2003, Long Stroke Fast Tool Servo System with Dynamic Mass Com-pensation and Hydrostatic Bearings, Proc. of the Int. Topical Conf. on Precision Engineering, Aa-chen, 136-140.

[104] Stoiber, D., Knorr, M., 2000, Ruckfrei – Dank der Impulsentkopplung 1 m bei 30g Beschleunigung, WB, 133/6:31-35.

[105] Brecher, C., Ostermann, T., Weck, M., 2004, Pick&Place: Hochdynamischer Bestückungsauto-mat mit impulsentkoppelten linearen Direktantrie-ben, Elektrisch-mechanische Antriebssysteme - In-novationen - Trends - Mechatronik, Fulda, Sept. 6.-7., VDI/VDE Verlag Berlin, 317-328.

[106] Hilbing, R., 2004, Genauigkeitssteigerung von Prä-zisionsdrehmaschinen durch aktive Kompensation dynamischer Sörungen, Diss. TH Aachen.

[107] Weule, H., Golz, H.-U., 1991, Preload-control in Ball Screws - a New Approach for Machine Tool Build-ing?, Annals of the CIRP, 40/1:383-386.

[108] Chen, J.S., Dwang, I.C., 2000, A Ballscrew Drive Mechanism with Piezo-Electric Nut for Preload and Motion Control, Int. J. of Machine Tools and Manu-facture, 40:512-526.

[109] Michos, G., 2005, Mechatronische Ansätze zur Op-timierung von Vorschubachsen, Diss. University of Erlangen-Nürnberg.

[110] www.ina.de. [111] US Patent 5,971,614 – Aesop-1999. [112] Patentanmeldung US2004/0042689 Hardinge. [113] Fleischer, J., Knödel, A., Munzinger, C., 2006, Ac-

tive Hydrostatic Guiding System with Integrated Adaptronic Sensor/Actuator-Unit, Conf.-Speech, 5. Int. Fluid Power Conf. - Customer Benefits of Fluid Power, Aachen, Germany.

[114] Holmes, M.L., Trumper, D.L., 1996, Magnetic/Fluid Bearing Stage for Atomic-Scale Motion Control, Precision Engineering, 18/1:38-49.

[115] Gleichner, A., 1999, Aktive hydrostatische Führun-gen mit elektrorheologischen Flüssigkeiten, Diss. TU Braunschweig.

[116] Hesselbach, J., Abel-Keilhack, C., 2002, Active Hy-drostatic Bearing with Magnetorheological Fluid, 8th Int. Conf. on New Actuators, Bremen, Germany, 343-346.

[117] Williams, M.E., Trumper, D.L., 1993, Magnetic Bearing Stage for Photolithography, Annals of the CIRP, 42/1:607-610.

[118] Kim, W.J., Trumper, D.L, Lang, J.H., 1998, Model-ing and Vector Control of a Planar Magnetic Levita-tor, IEEE Transactions on Industry Applications, 34/6:1254-1262.

[119] Wahner, U., 2002, Lineare Magnetführung für di-rektangetriebene Vorschubachsen, Diss. RWTH Aachen.

[120] Denkena, B., Kallage, F., Ruskowski, M., Popp, K., 2004, Machine Tool with Active Magnetic Guides, Annals of the CIRP, 53/1:333-336.

[121] Sato, K., Horikawa, O., Shimokohbe, A., 1996, Spindle Motion Control by Active Air Rotary Bear-ing, The Japan - China Bilateral Symposium on Ad-vance Manufacturing Engineering, Hayama, Japan, 44-49.

[122] www.leuvenairbearings.com.

[123] Al-Bender, F., Van Brussel, H., 1994, Active Dy-namic Stiffness Control of Aerostatic Bearings, Proc. 3rd Int. Conf. Ultraprecision in Manufacturing Engineering, Franz Rhiem Duisburg, 143-146.

[124] Shimokohbe, A., Aoyama, H., 1991, An Active Air Bearing: A Controlled-type Bearing with Ultra-precision, Infinite Static Stiffness, High Daming Ca-pability and New Functions, Nanotechn., 2/1:64-71.

[125] www.mech.kuleuven.ac.be. [126] Uhlmannn, E., Neumann, C., 2005, Adaptronik in

Werkzeugmaschinen – Kompensation thermisch bedingter Verlagerungen, Futur, 2:8-9.

[127] Neugebauer, R., Hipke, T., 2006, Machine Tools with Metal Foams, Advanced Engineering Materials, 8/9:858–863.

[128] www.iwu.fraunhofer.de. [129] www.niles-simmons.com. [130] Mussa, S., 2002, Aktive Korrektur von Verlage-

rungsfehlern in Werkzeugmaschinen, Diss. Univer-sity of Karlsruhe.

[131] Jungnickel, G., Wunderlich, B., 1988, Korrektursys-tem für thermisch bedingte Neigungen an Werk-zeugmaschinen, wt Werkstattechnik, 88/3:92-95.

[132] www.epucret.de. [133] Preumont, A., 2002, Vibration Control of Active

Structures, an Introduction, Kluwer Academic Pub-lishers, Dordrecht, Netherlands.

[134] Ehmann, C., Nordmann, R., 2003, Comparison of Control Strategies for Active Vibration Control of Flexible Structures, Conf.-Speech, 6th Conf. on Ac-tive Noise and Vibration Control Methods, Cracow.

[135] May, C., Pagliarulo, P., Janocha, H., 2006, Optimi-sation of a Magnetostrictive Auxiliary Mass Damper, Actuator 2006, Proc. 10th Int. Conf. on New Actua-tors, 344-348.

[136] Weck, M., Schulz, A., 2003, Adaptiver Reibungs-dämpfer, wt online, 93:7-8.

[137] Gaul,L., Albrecht,H., Wirnitzer,J., 2000, Damping of Structural Vibrations Using Adaptive Joint Connec-tions and Neural Control, Smart Structures: Theory and Applications, CISM Courses and Lectures No. 429, Udine, June 19-23, 86-97.

[138] Ehmann, C., Nordmann, R., 2002, Low Cost Actua-tor for Active Damping of Large Machines, Conf.-Speech, Mechatronics Conf., Berkeley, California.

[139] Weck, M., Schulz, A., 2004, Active Auxiliary Mass Damper, Production Engineering XI/1:39-43.

[140] www.micromega-dynamics.com. [141] Neugebauer, R., Wittstock, V., Illgen, A., Kranz, B.,

Naumann, G., 2006, The 3rd Generation of an Ac-tuator-Sensor-Unit for Tripod-Structures, Proc. 5th

Chemnitz Parallel Kinematics Seminar, Chemnitz, Apr. 25./26., 325-340.

[142] Wauer, J., Rudolf, C., Fleischer, J., Munzinger, C., 2005, An Approach for Compensation of Geometric Faults in Machine Tools, Proc. of the IDETC/CIE 2005, 5th Int. Conf. Multibody Systems, Nonlinear Dynamics and Control, 75-81.

[143] Heisel, U., Braun, S., 2007, An Adaptronic Ap-proach to Active Vibration Cancellation in Machine Tools with Truss-Like Parallel Kinematic Structure, Annals of the WGP (submitted for publication).

-684-

[144] Michels, F., 1999, Stabilisierung des Schleifprozes-ses mit aktiven Systemen, Diss. RWTH Aachen.

[145] Abicht, C., Ulbrich, H., Riebe, S., 2002, Active Vi-bration Isolation of a Stewart-Platform using High Response Hydraulic Actuators, Proc. of the 6th Int. Conf. on Motion and Vibration Control (MOVIC 2002), Saitama, Japan, Aug. 2002, 30-35.

[146] www.bilz.ag. [147] Weck, M., Krell, M., 2000, Tool and spindle systems

with high damping, WGP-Annals, VII/2:79-84. [148] Smith, S., Tlusty, L., 1992, Stabilizing Chatter by

Automatic Spindle Speed Regulation, Annals of the CIRP, 41/1:433-436.

[149] Sexton, J. S., Stone, R.J., 1975, The Stability of Machining with Continuously Varying Spindle Speed, Annals of the CIRP, 24/1:321-326.

[150] Ismail, F., Soliman, E., 1997, A New Method for the Identification of Stability Lobes in Machining, Int. J. of M. Tool Design and Reserach, 37/6:763-774.

[151] Duncan, G.S., Tummond, M.F., Schmitz, T.L., 2005, An Investigation of the Dynamic Absorber Effect in High-speed Machining, Int. J. of Machine Tools and Manufacture, 45:497-597.

[152] Tönshoff, H.K., Denkena, B., Götz, T., 2002, Piezo-electric Actuator Based Preload Control Unit for Machine Tool Spindles, WGP Annals, IX/1:117-122.

[153] Ehmann, C., Alizadeh, A., Nordmann, R., 2003, Schwingungsdämpfung aktiv gelagerter Rotoren mit robuster Regelung, 6th Int. Conf. Schwingungen in rotierenden Maschinen - SIRM, Darmstadt, 1-10.

[154] www.axomat.de. [155] www.levitec.de. [156] Nordmann, R., 2005, Use of Mechatronic Compo-

nents in Rotating Machinery, Conf.-Speech, 7th

Conf. on Vibration Problems, Istanbul, Sept. 05.-09. [157] Thiele, J., Urban, F., 1998, Motorspindeln mit akti-

ven Magnetlagern erhöhen Leistungsfähigkeit von Maschinen, MM, 52/104:30-34.

[158] Li, W., Maißer, P., Enge, H., 2004, Self-learning Control Applied to Vibration Control of a Rotating Spindle by Piezopusher Bearings, Proc. Int. Mech. Eng., J. Systems and Control Engineering, 218/1:185-196.

[159] Dohner, J.L., Lauffer, J.P., Hinnerichs, T.D., Shankar, N., Regelbrugge, M., Kwan, C., Xu, R., Winterbauer, B., Bridger, K., 2004, Mitigation of Chatter Instabilities in Milling by Active Structural Control, J. of Sound and Vibration, 269:197-211.

[160] Hill, S., Stallaert, B., Symens, W., Sas, P., 2006, Force Transmission Loss as a Mechanism for the Active Control of Noise Using an Active Bearing: Preliminary Results, Proc. ISMA2006, Noise and Vibration Eng., Leuven, Sept. 18-20, 211-222.

[161] www.weissgmbh.de. [162] Santos, I.F., Scalabrin, A., Nicoletti, R., 2001, Ein

Beitrag zur Aktiven Schmierungstheorie, Schwin-gungen in rotierenden Maschinen, Vieweg-Verlag, Wiesbaden.

[163] Carmignani, C., Forte, P., Rustighi, E., 2001, Active Control of Rotor Vibrations by Means of Piezoelec-tric Actuators, Proc. of the ASME Design Engineer-ing Technical Conf., 6(A), 757-764.

[164] Althaus, J., Ulbrich, H., 1990, Active Chamber Sys-tem for Vibration Control of Rotating Machinery, Proc. of the 3rd Int. Symposium on Transport Phe-

nomena and Dynamics of Rotating Machinery, Honolulu, USA, 543-558.

[165] Breit, H., 2002, Aktive Spindellagerung, Ölhydraulik und Pneumatik 46(7):424-432.

[166] Kohmäscher, T., 2004, Active Spindle Bearing De-vice for Chatter Control within Milling Machines, Conf.-Speech, 3rd PhD Symp., Terrassa, Spain.

[167] Klaffert, T., 2006, Selbstoptimierende HSC-Motorspindel, Diss. TU Chemnitz.

[168] www.lord.com. [169] www.auswuchttechnik.de. [170] www.schmitteurope.com. [171] Wieland, F., Hipke, T., Reinke, K., 1997, Offenle-

gungsschrift DE 19711726 A1, Vorrichtung zum Unwuchtausgleich, Fraunhofer Gesellschaft.

[172] Rivin, E.I., 1999, Stiffness and Damping in Me-chanical Design, Marcel Dekker Inc., New York.

[173] Fleischer, J., Denkena, B., Winfough, B., Mori, M., 2006, Workpiece and Tool Handling in Metal Cut-ting Machines, Annals of the CIRP, 55/2:817-841.

[174] Levina, Z. M., Zverew, I. A., 1986, DEA of Static and Dynamic Characteristics of Spindle Units, Stanki I instrument, 8:6-9 (in Russian).

[175] Tönshoff, H. K. et al., 1999, Compensating Milling Tool Deflection for Five Axis Machine Tools, Int. Conf. Quality Manuf., Stellenbosch, South Africa, 91-95.

[176] Pritschow, G., Korajda, B., Fanitza, T., 2005, Kom-pensation der Werkzeugabdrängung, Geometrische Betrachtungen und Korrekturstrategien, wt-online 95/5:337-341.

[177] Tauchen, M., Findeklee, J., 2000, Reduktion der Werkzeugabdrängung beim HSC-Schlichtfräsen, VDI-Z, III/2000, 32-35.

[178] Denkena, B., Will, J.C., Sellmeier, V., 2006, Predic-tion of Process Stability and Dynamic Forces of an Adaptronic Spindle System, Conf.-Speech, Adap-tronic Congress, Göttingen.

[179] Bushuev, V. V., 1991, Compensation of Elastic De-formations in Machine Tools, Stank I instrument, No. 3:42-46 (in Russian).

[180] Katsuki, A., Onikura, H., Sajima, T., Akashi, T., Ma-tsuo, T., 1992, Development of a Deep Hole Boring Tool Guided by Laser, Annals of the CIRP, 41/1:83-87.

[181] Koren, Y., Pasek, Z. J., Szuba, P., 1999, Design of a Precision, Agile Line Boring Station, Annals of the CIRP, 48/1:313-316.

[182] Brecher, C., Schauerte, G., Lange, S., 2005, A-daptronisches Bohrwerkzeug zur Feinbearbeitung von Zylinderhülsen, INN Innovative Technik, 10(32):14-15.

[183] Denkena, B., Götz, T., 2005, Aktive rotierende Spannvorrichtung mit Bluetooth-Anbindung, wt-online, 95/05: 118-121.

[184] Denkena, B., Immel, J., Götz, T., 2005, Drehspann-futter mit integrierter, mechatronischer Feinpositio-niervorrichtung, Conf.-Speech, Int. Forum Me-chatronik, 15.-16.6., Augsburg.

[185] Weber, C. T., Weiser, J., 2003, Formadaptiver Chuck, Conf.-Speech, Proc. 11th CMP-user mee-ting, Munich.

[186] www.igam-mbh.de.

-685-

[187] Wang, M., Fei, R., 1999, Chatter Suppression Based on Nonlinear Vibration Characteristic of Elec-trorheological Fluids, Int. J. of Machine Tools and Manufacture, 39:1925-1934.

[188] Weinert, K., Kersting, M., 2006, Conceptual Devel-opment of an Adaptronic Tool Holder for Single Lip Drills, Production Engineering, 13/1:9-12.

[189] Cowley, A., Boyle, A., 1970, Active Dampers for Machine Tools, Annals of the CIRP, 18/1:213-222.

[190] Tewani, S.G., Rouch, K.E., Walcott, B.L., 1995, A Study of Cutting Process Stability of a Boring Bar with Active Dynamic Absorber, Int. J. of Machine Tools and Manufacture, 35/1:91-108.

[191] Browning, D.R., Golioto, I., Thompson, N.B., 1997, Active Chatter Control System for Long-overhang Boring Bars, Proc. of the SPIE, 3044:270-280.

[192] www.acticut.com. [193] Tellbüscher E., 1985, Konstruktion von Dämpfern

und deren Einsatz an Rundschleifmaschinen, VDI-Z 11/80:37-40.

[194] Jendryschik, J., 1987, Dynamisches Verhalten rotie-render scheibenförmiger Trennwerkzeuge, Diss. University of Hannover.

[195] Pape, D.A., Carlson, W.B., Fowler, S.A., 1993, De-sign of an Electroceramic Actuator for Control of Radial Saw Blade Vibration, Vibration and Control of Mechanical Systems ASME 61, 1001-1008.

[196] Homann, S., Rose, M., Breitbach, E., 2005, Hybride elektromechanische Netzwerke mit piezoelektri-schen Komponenten, Conf.-Speech, Adaptronic Congress, Göttingen.

[197] Denkena, B., Harms, A., 2004, A Tool Adaptor for Vibration Damping in CNC-Lathes, Proc. of the 37th

CIRP Int. Seminar on Manufacturing Systems, May 19-21, Budapest, Hungary, 127-131.

[198] Sturos, T.J., Sutherland, J.W., Moon, K.S., Liu, D., Kashani, A.R., 1995, Application of an Aactively Controlled Magnetostrictive Actuator to Vibration Abatement in the Turning Process, Proc. of the SME, 57/1:539-544.

[199] Glaser, S., Hesselbach, J., Last, P., Pietsch, I., 2003, Funktionsverbesserungen von Glasschnei-demaschinen durch aktive Schwingungsdämpfung, Conf.-Speech, Adaptronic Congress, Wolfsburg.

[200] Gosebruch, H., 1990, Rundschleifen im geschlos-senen Regelkreis, Diss. University of Hannover.

[201] Neugebauer, R., Stoll, A., 2004, Ultrasonic Applica-tion in Drilling, J. of Materials Processing Technol-ogy, 149:633-639.

[202] Suzuki, K., Uematsu, T., Tochinai, H., Mishiro, S., Nakagawa, T, 1993, A New Grinding Method Ce-ramics Using a Biaxially Vibrated Non-rotational Ul-trasonic Tool, Annals of the CIRP, 42/1:375-378.

[203] Neugebauer, R., Päßler, T., Kroos, J., 2000, Tryout-Presse-Anforderungsbild, strittige Gestaltungsfra-gen, Prototyp, Int. Conf. Neuere Entwicklungen in der Blechumformung, Stuttgart, May 23.-24., 107-124.

[204] Groche, P., Schultheis, V., Schneider, R., 2005, New Production Perspectives Through Direct Drive Systems, WGP Annals, XII/1:97-100.

[205] Hesselbach, J., Rathmann, S., 2005, Hexapoden für die flexible Umformung - Aufbau und Inbetrieb-nahme einer flexiblen Umformmaschine, wt-online 10:793-797.

[206] Neugebauer, R., Noack, S., 2004, The Accuracy of Hydraulically Driven Struts for Production Machines, 4th Chemnitz Parallel Kinematics Seminar, Chem-nitz, Apr. 20.-21., 24:503-516.

[207] Neugebauer, R., Löschmann, F., Putz, M., Koch, T., Laux, G., A Production-Oriented Approach in Elec-tromagnetic Forming of Metal Sheets, 2nd Int. Conf. ICHSF, Dortmund, Mar. 20.-21., 129-139.

[208] Kleiner, M., Brosius, A., 2006, Determination of Flow Curves at High Strain Rates using the Elec-tromagnetic Forming Process and an Iterative Finite Element Simulation Scheme, Annals of the CIRP, 55/1:267-270.

[209] Patent DE 198 10 406.5, 1998, Exzenterpresse mit variabler Stösselbewegung, Schuler Pressen.

[210] Jung, E., 1998, Hydromechanische Pressenhaupt-antriebe auf der Basis sechsgliedriger Lenkerhebel-getriebe, Diss. TU Chemnitz.

[211] www.mueller-weingarten.de. [212] Trumpf, 2005, Faszination Blech, proprietary print. [213] Beyeler AG, 2005, Die neue Beyeler PR - Serie mit

IPC, proprietary printing. [214] Liet, R., 2001, Darley’s ACS Angle Measuring Sys-

tem for Press Brakes, 9th Int. Conf. on Sheet Metal Leuven, 467-474.

[215] Serruys, W., 2001, Adaptive Bending, 9th Int. Conf. on Sheet Metal Leuven, 503-512.

[216] Von Finckenstein, E., Austerhoff, N., Heller, B., Su-laiman, H., 1997, In-process Control System for Flexible Air Bending of Sheet Metal, Production En-gineering, IV/2:109-114.

[217] Lutters, D., Streppel, A.H., Kals, H.J.J., 1997, Mod-elling and Control of the Air Bending Process, Sec-ond World Congress on Intelligent Manufacturing Processes & Systems, Budapest, 473-478.

[218] Stelson, K. A., 1986, A Adaptive Pressbrake Control for Strain-Hardening Materials, J. of Eng. for Indus-try, 108:127-132.

[219] Reissner, J., Meier, M., 1985, Computer Aided Three Point Bending of Multiple Bend Profiles, An-nals of the CIRP, 34/1:259-264.

[220] Klose, L., Elend, K.-E., 2000, Elastisches Niederhal-tersystem erweitert Tiefziehbereich von einfach wir-kenden Pressen, MM, 106/19:30-37.

[221] Behrens, B.-A., Vogt, O., Hübner, S., 2006, Tiefzie-hen mit hydroelastischem Niederhalter - Elastische Niederhaltersysteme bieten beim Tiefziehen ein ho-hes Kostensenkungspotential, wt-online, 3:128-130.

[222] Päßler, T., 2003, Parallel geregeltes Ziehkissen, Jahresbericht Fraunhofer IWU, Chemnitz.

[223] Neugebauer, R., Roscher, H.-J., Wolf, K., Hoff-mann, M., Scheffler, S., 2006, Control of Sheet-Metal Forming Processes with Piezoactuators in Smart Structures, Proc. of SPIE 61710E, Feb. 27.-28., San Diego, California, USA, 1-9.

[224] Hoffmann, M., Päßler, T., Urschl, G., Egerer, E., 2004, Offener Steuerungskern hält Ziehkissen im Lot, Blech InForm, 3:38-40.

[225] www.dieffenbacher.de. [226] Lehfeldt, E., 1968, Beeinflussung metallischer Rei-

bungsvorgänge durch Schall im 20 kHz-Bereich, Diss. RWTH Aachen.

-686-

[227] Siegert, K., Ulmer, L., 2001, Influencing the Friction in Metal Forming Process by Superimposing Ultra-sonic Waves, Annals of CIRP, 50/1:195–200.

[228] Panknin, W., Naujoks, B., Siegert, K., 1973, Der Einfluß longitudinaler Ultraschall-Schwingungen auf Umformprozesse unter besonderer Berücksichti-gung des Wirkungsgrades, Zeitschrift für Metallkun-de, 64/3:145-151.

[229] Malek, R., 1997, Beitrag zum Rohrziehen mit ultra-schallerregtem Dorn, Diss. University of Stuttgart.

[230] Leser, H., 2004, Ultraschallbeeinflussung von Draht- und Rohrziehprozessen, Forschung am Insti-tut für Umformtechnik der Univ. Stuttgart, Sept. 16., MAT-INFO Werkstoff-Informationsgesellschaft.

[231] Jäger, T., Hobler, M., Sulzberger, P., 2006, Ener-gieautarke Transpondersysteme für die Zustands-überwachung rotierender Komponenten von Motor-frässpindeln, Conf.-Speech, 4. GMM-Workshop E-nergieautarke Sensorik, Karlsruhe, Sept. 14.-15.

[232] Denkena, B., Brinkhaus, J., Lange, D., 2005, Possi-bilities of Sensor Transponders in Machine Tools, 10th Int. Scientific Conf. on Production Engineering CIM, Zagreb, 11-13.

[233] Denkena, B., Grendel, H., Holz, Ch., 2004, Model Based Feedforward and State Control of the Paral-lel Kinematics PaLiDA, Proc. 4th Chemnitz Parallel Kinematics Seminar, Chemnitz, April 20 - 21, 185-202.

[234] Weck, M., Peters, A., 2001, Mobile Kommunikation-Mobile Produktion, WT, H9, 565-570.

[235] Szczerbicka, H., Uthmann T., 2000, Modellierung, Simulation und Künstliche Intelligenz, Fortschritte in der Simulationstechnik, SCS Publishing House, Er-langen.

[236] Pritschow, G., Kissling, M., 1989, KI-Methoden und ihre Bedeutung fuer die Fertigungstechnik, Künstli-che Intelligenz in der Fertigungstechnik, 3-16.

[237] Limbourg, P., Germann, D., Petersen, J., Kochs, H.-D., 2005, Integration von Verlässlichkeitsbetrach-tungen in die VDI 2206: Entwicklungsmethodik für mechatronische Systeme, VDI-Berichte, 1892.

[238] Tomatis, N., Brega, R., Arras, K., Jensen, B., Moreau, B., Persson, J., Siegwart, R., 2001, A Complex Mechatronic System: from Design to Ap-plication, IEEE/ASME Int. Conf. on Advanced Intel-ligent Mechatronics Proc., Como, Italy, 278-283.

[239] Suh, N.P., 2005, Complexity in Engineering, Annals of the CIRP, 54/2:581-598.

[240] Kochs, H.-D., 2004, Key Factors of Dependability of Mechatronic Units – Mechatronic Dependability, Proc. 28th Annual Int. Computer Software and Ap-plications Conf. COMPSAC, 584-586.

[241] Guérin, F., Todoskoff, A., Barreau, M., Morel, J.-Y., Mihalache, A., Dumon, B., 2002, Reliability Analysis for Complex Industrial Real-time Systems: Applica-tion on an Antilock Brake System, IEEE SMC (7).

[242] Biegert, U., 2003, Ganzheitliche modellbasierte Si-cherheitsanalyse, atp - Automatisierungstechnische Praxis, 8/45:42-49.

[243] Limbourg, P., Germann, D., 2004, Reliability As-sessment and Optimization under Uncertainty in the Dempster-Shafer Framework, 27th ESReDA Semi-nar, Glasgow, GB, 77-89.

[244] Dilger, E., Karrelmeyer, R., Straube, B., 2004, Fault Tolerant Mechatronics, Proc. 10th IEEE Int.l On-Line Testing Symposium, 214-218.

[245] Gray, J., 1985, Why do Computers Stop and What Can be Done About It, Proc. German Association for Computing Machinery Conf. on Office Automa-tion, Erlangen, 3-12.

[246] www.isw.uni-stuttgart.de. [247] Schröder, T., 2007, Entwicklung und Evaluation von

Algorithmen zur zeitoptimierten Bewegungszerle-gung bei kinematisch redundanten Werkzeugma-schinen, Diss. TU Chemnitz.

[248] Pritschow, G., Wurst, K.-H., Seyfarth, M., Kircher, C., 2005, Self-adaptable Control Systems for RMS, Conf.-Speech, CIRP 3rd Int. Conf. on Reconfigur-able Manufacturing, May 10.-12., University of Michigan, Ann Arbor, USA

[249] Mentink, R.J., Lutters, D., Streppel, A.H., Kals, H.J.J., 2003, Determining Material Properties of Sheet Metal on a Press Brake, J. Materials Proc-essing Technology, 141:143-154.

[250] Yang, M., Kojima, N., 1997, High Accuracy V-Bending Process Control with an On-Line Database and Modified Fuzzy Models, JSME, Series C, 40:157-162

[251] Yang, M., Shima, S., Watanabe, T., 1992, Devel-opment of Control System Using Neural Network Model, Proc. Japan USA Symposium on FA, 2:1485-1491.

[252] Gattringer, H., 2006, Johannes Kepler University of Linz: The Bipedal Robot, dSPACE NEWS, dSPACE GmbH, Paderborn.

[253] Autorenkollektiv, 2001, Leichtbauarm mit innovati-ven Sensoren und sensorbasierten Steuerung-systemen für Produktion und Dienstleistung LISSY, Forschungzentrum Karlsruhe.

[254] www.de.schunk.com. [255] Hirzinger, G., Sporer, N., Schedl, M., Butterfaß, J.,

Grebenstein, M., 2004, Torque-controlled Light Weight Arms and Articulated Hands - Do We Reach Technological Limits Now?, Wessling.

[256] Schraft, R.D., Hägele, M., Prassler, E., Bischoff, R., Nebel, B., Lawitzky, G., Plöger, P., Ritter, H., 2006, DESIRE: Schlüsseltechnologien für alltagstaugliche Serviceroboter, BMBF Statustagung, Berlin.

[257] Grunewald, G., 2005, Papas Abschlusspräsentati-on, Conf.-Speech, F124, Robotik und Handha-bungstechnik in der Produktion, Dec. 8.-9., Stutt-gart.