Production Engineering

, Volume 7, Issue 2–3, pp 309–318 | Cite as

Simulative investigation of point diffraction interferometry with regard to machine tool calibration

  • Christian Brecher
  • Christian Krella
  • Florian Lindner
Machine Tool

Abstract

One major effect that can deteriorate the geometric accuracy of a calibrated machine tool is the elastic deformation of its structure due to heat transferred into it. Different solutions to this problem have been presented in the past—e.g. temperature-controlled structural parts and environments—but they are complex and expensive. In an on-going Collaborative Research Centre funded by the German Research Foundation scientists are working on model-based compensation and correction strategies. In this paper, a different approach is discussed using point diffraction interferometry to directly measure the three-dimensional distance between tool centre point and workpiece coordinate system. Commercially available three-dimensional measurement systems are introduced representing the state of the art before explaining the basics of point diffraction interferometry. Different parameters are investigated using simulation tools and the achievable accuracy of the approach is evaluated. Additionally, some suggestions are made concerning hardware requirements of a planned demonstrator.

Keywords

Optical measurement Interferometry Distance measurement Calibration Thermal deviation 

1 Introduction

In many fields of industrial manufacturing tight tolerances are required in functional areas of the produced parts. If these tolerances are not met, cost and labour extensive rework will be necessary or the parts will have to be declared as scrap. Compliance with the requested tolerances leads to high demands regarding the manufacturing processes and the accuracy of the employed machine tools [1]. The main sources for a reduced accuracy can be classified into kinematic errors, thermo-mechanical errors, loads, dynamic forces, motion control and control software [2].

Thermo-mechanical errors are caused by internal heat sources and environmental influences [3, 4]. Depending on material and structural properties the resulting temperature distribution causes unwanted displacements and inclinations at the tool centre point (TCP) of the machine [5]. These thermally induced structural deformations commonly dictate the achievable accuracy [2, 6] and will directly lead to dimensional and shape deviations of the finished product [6, 7]. The general relationship between heat transfer and workpiece quality is depicted in Fig. 1.
Fig. 1

Relationship between heat transfer and thermal deviation, cf. [5]

Existing solutions which try to establish a thermal equilibrium increase the energy consumption and impact both the profitability and the environment. The Collaborative Research Centre SFB/TRR 96, which was started in July 2011, will therefore focus on the thermo-energetic design of machine tools. One approach is to correct thermo-elastic deformations using small movements of the machine axes based on complex mathematical models. The other approach is to compensate for those deformations by different means (e.g. controlled temperature distribution, drive optimisation) [6].

2 Three-dimensional distance measurement techniques

A different approach to correct thermal deviation is based on the direct distance measurement between TCP and workpiece coordinate system and comparison of this distance to the value of the machine tool’s numerical control (NC). A difference between the two values is either caused by a remaining geometric error or by thermo-elastic deformation and can be corrected by small movements of the machine axes. In the following chapter, some commercially available three-dimensional measurement systems will be introduced representing the current state of the art.

2.1 iGPS (triangulation)

In the US patent US6535282B2 a three-dimensional measurement system allowing for the position determination of a receiver with regard to at least two known transmitters via triangulation is described [8]. Figure 2 shows the principle of the system and its mode of operation.
Fig. 2

Principle of iGPS and its mode of operation, cf. [9]

The transmitters emit two continuously rotating laser fans which are tilted to the vertical and to each other. The relative position of a receiver is defined as the intersection of the lines of sight towards at least two transmitters (azimuth and elevation angles) which can be calculated from simple time measurements if the geometric setup and the rotational speeds of the transmitters are known [10].

Realistically achievable accuracies fall into the range of 100–500 μm for a measurement field of about 3 × 3 m² [10, 13]. Possible error sources are multipath propagation (reflexions) of the laser fans, background noise, suboptimal sensor positioning and orientation as well as kinematic effects [11, 12].

The use of this system as a recalibration device for shop floor machining centres is questionable, since the thermally induced errors are typically only as big as or even smaller than the accuracy of the measurement device [4, 5]. In addition, the minimal distance between sensor and transmitters, required for proper operation [12], hinders the deployment in entirely enclosed machines. Furthermore, the geometric setup of the transmitters has to be known precisely and may not change between different measurements. The detection of angular displacements is only possible using a set of at least three sensors with known distances to each other.

2.2 μ-GPS (trilateration)

Within the scope of the research project ACCOMAT a measurement system based on trilateration has been developed [14, 15]. Figure 3 shows the main components of the system and its mode of operation. The system is based on white light interferometry and matches the optical path length of a reference unit (comparator) to the path lengths between a transmitter and several reflectors inside the working volume [16].
Fig. 3

Principle and mode of operation of μ-GPS, cf. [16]

Simulative investigations and one-dimensional experiments have been carried out successfully to estimate the performance of the system under real working conditions [17]. Using improved hardware the standard deviation of three-dimensional measurements was determined to be 7.5 μm. Estimates for the overall system costs were in the range of 33.000–37.000 € [16].

Due to the high system costs, a widespread deployment as a recalibration device permanently built into a machine tool is not very likely, although the performance is sufficient with regard to repeatability of the measurements. Comparable to the iGPS device mentioned before, the geometric configuration of the reflectors must be known in advance and held constant during the measurements. Angular displacements can also only be detected using a set of sensors with known distances to each other.

3 Distance measurement using point diffraction interferometry

3.1 Basic approach and physical background

The idea of a point diffraction interferometer is based on the interference of at least two coherent light waves emanating from point sources that are positioned closely to each other. Assuming a spherical shape the mean intensity of two interfering waves (indexed 1 and 2) can be expressed by Eq. (1) [18]. In this equation e is the electric field strength at the origin, r is the position vector originating from the source, c is the speed of light, ε the electric constant, φ the initial phase, Δ stands for the difference between two variables and the suffix 0 denotes vacuum condition.
$$ \bar{I} = c_{0} \varepsilon_{0} \left[ {\frac{{{\mathbf{e}}_{{\mathbf{1}}}^{{\mathbf{2}}} }}{{{\mathbf{r}}_{{\mathbf{1}}}^{{\mathbf{2}}} }} + \frac{{{\mathbf{e}}_{{\mathbf{2}}}^{{\mathbf{2}}} }}{{{\mathbf{r}}_{{\mathbf{2}}}^{{\mathbf{2}}} }} + 2\frac{{{\mathbf{e}}_{{\mathbf{1}}} {\mathbf{e}}_{{\mathbf{2}}} }}{{{\mathbf{r}}_{{\mathbf{1}}} {\mathbf{r}}_{{\mathbf{2}}} }}cos\left( {{\mathbf{k}}\Updelta {\mathbf{r}} + \Updelta \varphi } \right)} \right] $$
(1)
The equation shows that the local phase is a function of the path difference Δr between the two waves. The other variable expressions can be interpreted as offset and scale factors. The cosine in the equation causes the intensity distribution to oscillate spatially between a minimum and a maximum value resulting in the pattern of dark and bright fringes seen in Fig. 4.
Fig. 4

Two interfering spherical waves and their resulting intensity and phase distribution

Because the same intensity value can be measured at different positions, there is no bijective relationship between intensity and position for single points. This problem can be overcome by looking at two-dimensional cross-sections of the intensity pattern, e.g. captured via a charged coupled device (CCD) camera. In this case, the relationship between the waves’ origins and the position of the capture device is uniquely determined [18].

The spacing of the extrema in the intensity image (and the discontinuities in the phase image accordingly) depend on the distance between the wave sources and the camera. The lateral offset defines the curvature that can be observed in the images. The initial orientation between camera and light sources is arbitrary but has to be determined via calibration and should be held constant between consecutive measurements.

3.2 Mathematic algorithm for distance determination

The steps in Fig. 5 represent the algorithm for the determination of the distance between the light sources and a two-dimensional receiver. In principal, the position is determined via a cost function by minimising the difference between the real phase, extracted out of a series of intensity images, and a synthetic phase, calculated for a certain position inside the working volume.
Fig. 5

Steps to determine the three-dimensional distance based on a series of intensity images

Although the position information is also contained in each individual intensity image, these images are not very suitable for position determination due to physical constraints. First of all, the initial field strengths of the light waves are unknown and difficult to determine. In addition, the intensity measured at each point of the CCD array varies depending on the background noise, the transmissibility of the medium and the accuracy of the individual photo cells so that the absolute intensity measured at each pixel is not very reliable. These problems can be circumvented by transforming the intensity information into phase information. For further details on the mathematic algorithm and the associated techniques reference is made to [19, 20, 21, 22, 23].

The final step of the software algorithm consists of an iterative approach which tries to minimise the sum of the error squares between a synthetic phase image Φs, calculated for a position vector a using theoretical equations, and the phase image Φr, extracted and unwrapped from the intensity images, as seen in Eq. (2). Φ0 is the initial phase offset between the two waves. This regression problem is of a nonlinear kind due to the characteristics of the underlying mathematical equations. Several techniques are available to cope with this kind of regression problem, e.g. Newton’s method and quasi-Newton methods [24].
$$ E = \hbox{min} \sum\limits_{i = 1}^{n} {\left[ {\left( {{\varvec{\Upphi}}_{{\mathbf{r}}} - {\varvec{\Upphi}}_{{\mathbf{0}}} } \right)_{i} - \left( {{\varvec{\Upphi}}_{{\mathbf{s}}} - {\varvec{\Upphi}}_{{\mathbf{0}}} } \right)_{i} } \right]^{2} = \hbox{min} \sum\limits_{i = 1}^{n} {\left( {\left\| {f({\mathbf{a}})} \right\|_{2}^{2} } \right)} } $$
(2)

3.3 Existing diffraction interferometer setups

A first demonstrator based on point diffraction interferometry was described in 2001 by Korean scientists [21]. The system, similar to the setup shown in Fig. 6, was based on two optical fibres which were connected on one side to the same coherent light source via a fibre coupler. Due to diffraction caused by the small core diameter of the fibres, light exits as two almost perfect spherical waves. The resulting intensity distribution of the two interfering waves was captured using a CCD array. The relative phase shifts between different intensity images, which are required for phase extraction, were realised by elongation of the optical fibres with the help of piezoelectric transducers (PZT).
Fig. 6

Possible setup for a point diffraction interferometer, cf. [22]

For the one-dimensional case a measurement uncertainty of 0.2 μm over a distance of 150 mm with a repeatability of 30 nm was reported [21]. Further tests carried out with an optimised version a year later indicated that the expected volumetric accuracy should be better than 1 μm [22]. A three-dimensional verification of this assumption was successfully carried out using a coordinate measuring machine as reference for a working volume of 60 × 60 × 20 mm³ [23].

However, some limitations hinder the practical deployment of the fibre-based interferometer setup in machine tools. Our experiments have shown that the optical fibres are extremely susceptible to mechanical and thermal stress which causes uncontrollable changes in the optical path difference between the two waves. In harsh environments without air-conditioning, like shop floors, the result will be a drift of the interference pattern which can have a magnitude significantly bigger than the shift of the interference fringes induced by a controlled phase shift. In addition, the use of piezoelectric transducers for phase shifting adds a considerable degree of complexity and costs to the entire system. Finally, the opening angle of a spherical wave emitted from an optical fibre is limited due to its numerical aperture [25], which makes it difficult to cover bigger measurement volumes.

3.4 Alternative interferometer setup for the use in machine tools

In order to eliminate these problems, a new system setup has been designed which features optical lenses for spherical wave generation and a simpler phase shifting technique. Coherent light from a frequency-stabilised laser is coupled into a single optical fibre and transmitted to an emitter head. Here, the light is decoupled, split into two separate beams and guided towards focussing lenses generating spherical waves that emanate from their focal points. The phase shift is realised using a liquid crystal retarder which induces a phase shift in one of the beams depending on the applied voltage. The problem of pattern drift is also avoided as the light is separated behind the connecting fibre. This way the phase relation between the two separated beams remains constant. A patent application for the system design has been filed [26].

A schematic overview of the system setup and its possible deployment inside a machine tool is depicted in Fig. 7. The camera is connected to the spindle housing while the emitter head is positioned on the machining table. The laser and phase controller electronics can be situated outside the machine, connected to the emitter via a single optical fibre.
Fig. 7

Schematic overview of the alternative system setup and deployment inside a machine tool

Figure 8 shows exemplary results of a first trial which was executed after the system had been deployed to a test bench. In a) the two dimensional intensity distribution resulting from the interference is shown. The white circle marks an image disturbance which is probably caused by small objects on the lens. A sectional view marked by the white line is displayed in b) clearly showing the non-uniform intensity distribution caused by the effects described before. Ideal phase shifts of equal step size between consecutive images and the results of the liquid crystal retarder are depicted in c). The setup is still in the process of being established so no final measurement results are currently available.
Fig. 8

a Part of intensity image captured by CCD camera, b Sectional view of intensity profile (indicated by white line in a), c Comparison between real and ideal phase shifts

4 Simulative investigation of the system behaviour

After the principle functionality had been verified [21, 22, 23], the main task in testing the point diffraction interferometer system regarding its applicability on machine tools was to optimise the system parameters with respect to machine tool specific boundary conditions. While machine tools can be found with working volumes in almost any size, even an average value exceeds the volume tested in [23] by several multiples. Further boundary conditions result from environmental conditions within the machine tool working volume, such as air temperature and contamination or fluctuation in ambient light. Additional system constraints result from the technical equipment used to perform the diffraction measurements, like the CCD array size and resolution or the algorithm implementation and number precision.

In order to deal with the large number of influencing parameters and their wide spectrum of possible values, a simulation framework has been created enabling the user to evaluate the systematic influence of each parameter on the volumetric accuracy of the system. In the next paragraph, the simulation framework and test procedure will be presented briefly, followed by results obtained from corresponding simulations.

4.1 Simulation framework and test procedure

The simulation procedure can be divided into three consecutive steps. In a first step, a sequence of inference images is calculated for a given theoretical position and orientation. Based on the assumption of spherical wave fronts the calculated intensity images represent 2D-cuts through the volume in which these wave fronts interfere. Influencing parameters which can be used as simulation variables are listed in Fig. 9. In step two of the simulation process the mathematical algorithm for distance determination is applied for a starting position, which differs from the theoretical position, yielding a calculated position. In the last step, this calculated position is compared to the theoretical position for overall performance measurement. Furthermore, the theoretical phase image can be compared to the phase image extracted by the algorithm.
Fig. 9

Boundary conditions and data flow within the simulation framework

The configuration shown in Table 1 was used as a standard setup for the simulation framework. While intensity values for all photodiodes were used for extracting the phase image out of the inference image sequence, only a reduced number of 80 × 36 = 2,880 photodiodes and corresponding phase values spread equally over the CCD array was used for the calculation of the workspace position. This is due to the fact that the convergence of the numerical search is not much affected by the number of data points as long as it exceeds 20, while computation time increases rapidly with it [22].
Table 1

Standard simulation framework setup

Parameter

Value

Wavelength (nm)

630

Initial amplitudes

1

Distance between wave sources (mm)

10

Number of CCD array photodiodes

(1,280 × 960)

Distance between adjacent photodiodes (μm)

5

Initial phase offset (rad)

0

Phase modulation window (rad)

(0, 1.5π)

Number of phase modulations

4

Reduced number of CCD array photodiodes

(80 × 36)

Initial offset (mm)

20

Machine epsilon

2−52

4.2 Results of the simulative investigation

Using a simulation framework as described in the previous paragraph the systematic influence of different parameter has been investigated. Some exemplary results will be presented in the following chapters. If not otherwise specified, a configuration as stated in Table 1 was used for all simulations.

4.2.1 Position dependent accuracy of the algorithm

Assuming a fixed parameter setup one can expect the system accuracy to be dependent on the spacing between the wave sources and the CCD array. For wave sources located around [x, y] = [0 m, 0 m], with their common axis pointing in x-direction, the partial position dependent algorithm accuracy is displayed in Fig. 10. The CCD array centre is located in the x–y-plane with x ∈ [−0.12 m, 0.12 m] and y ∈ [0.35 m, 0.85 m], the array normal pointing in negative y-direction. The total position dependent accuracy is composed of the three partial accuracies in the spatial domain of which the component in x-direction is picked as an example in Fig. 10. It can be seen and should be noted that the accuracy does not decline neither linearly nor exponentially with increasing CCD array distance, but in a more complex manner.
Fig. 10

Position dependent error in x-direction

4.2.2 Tilt dependent accuracy of the algorithm

While for former measurements it was assumed that the orientation of the CCD array does not change, for the application within a machine tool this is not necessarily true. The angular deflection around the y- and z-axis can be detected and compensated by the algorithm, so that it does not show any significant influence on the position determination. Tilts around the x-axis cannot be detected due to the system concept. The tilting influence is shown in Fig. 11 for a maximum angular deviation twice as high as the tolerance specified in the ISO 10791 standard series [27]. It can be seen that for an increasing deviation of the CCD array orientation from its original orientation, with its normal pointing in negative y-direction, the error in position determination increases almost linearly with small tilting angles around the x-axis. As it can be expected with regard to the geometric setup, a small angular deflection affects the error’s z-component more than the y-component, while its influence on the x-component is insignificant. Furthermore the effect depends on the array’s position in the working volume.
Fig. 11

Error in position determination as a function of the CCD array tilt around the x-axis, simulated at position [x, y, z] = [0.03 m, 0.35 m, −0.01 m]

4.2.3 Influence of camera parameters

Charged coupled device arrays are commercially available with numerous different photodiode numbers, spacings and resolutions in terms of intensity. The size of the error in position determination induced by the limitation of the CCD array area is displayed in Fig. 12, assuming a constant number of photodiodes being used and ensuring a sufficient maximum spacing between them that avoids aliasing. The simulation indicates that the error converges towards a constant level with increasing CCD array area. Furthermore, the level and speed of convergence was found to depend on the distance between the wave sources and the CCD array.
Fig. 12

Error in position determination as a function of the CCD array area, simulated at position [x, y, z] = [0.03 m, 0.60 m, −0.01 m]

The influence of the CCD intensity resolution on the error in position determination is illustrated in Fig. 13. Again, only the spatial error in x-direction has been chosen exemplarily. The figure shows an approximately linear decrease in logarithmic error going along with the logarithmic increase in intensity resolution until a minimal error is reached resulting from the rest of the parameter setup.
Fig. 13

Error in position determination as a function of the CCD intensity resolution, simulated at position [x, y, z] = [0.01 m, 0.35 m, 0.00 m]

4.2.4 Impact of wavelength disturbances

Modern laser sources often come with a wavelength stability of 0.2 nm or better. For so called “standard air” this value corresponds to the maximum deviation of wavelength in air compared to vacuum when applying the Edlén equation for temperatures between 5 and 40 °C, cf. [28, 29, 30]. The effect of wavelength disturbances on the error in position determination for differences of up to 2 nm between assumed and real wavelength is displayed in Fig. 14. As it can be seen from the three plots the relationship between error and wavelength disturbance appears to be linear, yet with different gradients for the three spatial directions.
Fig. 14

Error in position determination as a function of the difference between assumed and real wavelength, simulated at position [x, y, z] = [0.03 m, 0.35 m, 0.00 m]

5 Conclusion and outlook

The results of the simulative investigation confirm the results of the experimental tests carried out with the interferometer setup described in [21, 22, 23] but expand the measurement volume being examined. They further show the principal dependency of the geometric accuracy on different parameters, for example indicating that a tilt of the CCD array around the x-axis will lead to large errors that cannot be compensated within the algorithm. However, this problem could be solved by introducing a third wave source or by moving the transmitter for a known increment which allows for two consecutive measurements with all other parameters held constant. The results indicate that a point diffraction interferometer in theory can be used as a recalibration device for machine tools if further corrections to the hardware are made.

Future work will primarily focus on the improvement of the alternative diffraction interferometer setup. Final tests have to show if the accuracy is in the range indicated by the simulations and if external disturbances can be handled sufficiently via compensation or correction. Wavefront distortion of the beam caused by the optical components (aberration) might be a problem and needs to be investigated. In addition, the phase shift quality of the liquid crystal retarder has to be evaluated. Furthermore, improvements to the simulation environment are planned.

Notes

Acknowledgments

The authors would like to thank the European Commission for funding this work as part of Project Chameleon within the 7th Framework Programme of the European Union.

References

  1. 1.
    Kinkel S, Lay G (2006) Technologietrends in der Produktion: Praxis der Anlagenmodernisierung in der deutschen Metall- und Elektroindustrie. Mitteilungen aus der Produktionsinnovationserhebung 39, Fraunhofer Institut für System—und Innovationsforschung (ISI), KarlsruheGoogle Scholar
  2. 2.
    Schwenke H, Knapp W, Haitjema H, Weckenmann A, Schmitt R, Delbressine F (2008) Geometric error measurement and compensation of machines—an update. CIRP Ann 57(2):660–675. doi:10.1016/j.cirp.2008.09.008 CrossRefGoogle Scholar
  3. 3.
    Bryan J (1990) International status of thermal error research. CIRP Ann 39(2):645–656. doi:10.1016/S0007-8506(07)63001-7 MathSciNetCrossRefGoogle Scholar
  4. 4.
    Weck M, Brecher C (2006) Werkzeugmaschinen 5—messtechnische untersuchung und beurteilung, dynamische stabilität. Springer, HeidelbergGoogle Scholar
  5. 5.
    Brecher C, Hirsch P (2004) Compensation of thermo-elastic machine tool deformation based on control internal data. CIRP Ann 53(1):299–304. doi:10.1016/S0007-8506(07)60702-1 CrossRefGoogle Scholar
  6. 6.
    Großmann K (2012) Thermo-energetische gestaltung von werkzeugmaschinen. ZWF 107(5):307–314Google Scholar
  7. 7.
    Großmann K (2012) Energieeffizient zerspanen. Maschine + Werkzeug 113(4):78–79Google Scholar
  8. 8.
    Hedges T, Takagi H, Pratt T, Sobel MJ (2001) Position measurement system and method using cone math calibration. Patent, US6535282B2Google Scholar
  9. 9.
    Wang Z, Mastrogiacomo L, Franceschini F, Maropoulos P (2011) Experimental comparison of dynamic tracking performance of iGPS and laser trackerGoogle Scholar
  10. 10.
    Müller T, Schwendemann J (2012) iGPS—ein vielseitiges Messsystem hoher Genauigkeit. avn 119(4):146–157Google Scholar
  11. 11.
    Muelaner JE, Wang Z, Martin O, Jamshidi J, Maropoulos P (2010) Verification of the indoor GPS system, by comparison with calibrated coordinates and by angular reference. J Intell Manuf (Online First: 12/06/2010)Google Scholar
  12. 12.
    Depenthal C, Schwendemann J (2009) iGPS—a new system for static and kinematic measurements. In: Grün A, Kahmen H (ed) Optical 3-D Measurement techniques IX, ViennaGoogle Scholar
  13. 13.
    Schmitt R, Nisch S, Schönberg A, Demeester F, Renders S (2010) Performance evaluation of iGPS for industrial applications. In: Mautz R, Ingensand H, Kunz M (ed) Proceedings of the 2010 international conference on indoor positioning and indoor navigation, Zurich, IEEE, New YorkGoogle Scholar
  14. 14.
    Spath D (2002) ACCOMAT—die genauigkeitsgeregelte Maschine. Abschlussbericht zum BMBF-Leitprojekt ACCOMAT (1998–2002). Final report, University of KarlsruheGoogle Scholar
  15. 15.
    Weule H, Plutowsky A, Höller F, Spieweck M, Werner J (2001) A three-degree-of-freedom measurement system for machine tools. Prod Eng Res Devel 8(2):95–98Google Scholar
  16. 16.
    Schmalzried S (2007) Dreidimensionales optisches messsystem für eine effizientere geometrische Maschinenbeurteilung. Dissertation, University of Karlsruhe, Shaker, AachenGoogle Scholar
  17. 17.
    Plutowsky A (2002) Charakterisierung eines optischen messsystems unter den bedingungen des Arbeitsraums einer werkzeugmaschine. Dissertation, University of KarlsruheGoogle Scholar
  18. 18.
    Steel WH (1983) Interferometry. University Press, CambridgeGoogle Scholar
  19. 19.
    Bruning JH (1978) Fringe Scanning Interferometers. In: Malacara D (ed) Optical shop testing. Wiley, New York, pp 409–437Google Scholar
  20. 20.
    Ghiglia DC, Pritt MD (1998) Two-dimensional phase unwrapping. Wiley, New YorkMATHGoogle Scholar
  21. 21.
    Kim S-W, Yang DY (2001) New design of precision CMM based upon volumetric phase-measuring interferometry. CIRP Ann 50(1):357–360. doi:10.1016/S0007-8506(07)62139-8 CrossRefGoogle Scholar
  22. 22.
    Rhee H-G, Kim S-W (2002) Absolute distance measurement by two-point-diffraction interferometry. Appl Opt 41:5921–5928. doi:10.1364/AO.41.005921 CrossRefGoogle Scholar
  23. 23.
    Rhee H-G, Chu J, Lee Y-W (2007) Absolute three-dimensional coordinate measurement by the two-point diffraction interferometry. Opt Express 15(8):4435–4444. doi:10.1364/OE.15.004435 CrossRefGoogle Scholar
  24. 24.
    Fletcher R (1987) Practical methods of optimization. Wiley, New YorkMATHGoogle Scholar
  25. 25.
    Young M (2000) Optics and lasers: including fibers and optical wave guides. Springer, BerlinGoogle Scholar
  26. 26.
    Brecher C, Klein W, Linder F (2011) Verfahren und System zur Korrektur der Position eines durch eine Steuerung bewegten Objektes. Pending patent, DE102011011286A1Google Scholar
  27. 27.
    DIN ISO 10791 (2001) Machine tools—test conditions for machining centres. DIN German institute for standardization (ed)Google Scholar
  28. 28.
    Edlén B (1966) The refractive index of air. Metrologia 2:71–80CrossRefGoogle Scholar
  29. 29.
    Birch KP, Downs MJ (1994) Correction of the updated Edlén equation for the refractive index of air. Metrologia 31:315–316CrossRefGoogle Scholar
  30. 30.
    Lide DR (ed) (2003) CRC handbook of chemistry and physics. A ready-reference book of chemical and physical data. Tayler & Francis, Boca Raton, FLGoogle Scholar

Copyright information

© German Academic Society for Production Engineering (WGP) 2012

Authors and Affiliations

  • Christian Brecher
    • 1
  • Christian Krella
    • 1
  • Florian Lindner
    • 1
  1. 1.Laboratory for Machine Tools and Production Engineering (WZL), Chair of Machine ToolsRWTH Aachen UniversityAachenGermany

Personalised recommendations