Cirp Annals - Manufacturing Technology: Sciverse Sciencedirect
Cirp Annals - Manufacturing Technology: Sciverse Sciencedirect
Cirp Annals - Manufacturing Technology: Sciverse Sciencedirect
A R T I C L E I N F O A B S T R A C T
Keywords: This paper presents a review of the latest research activities and gives an overview of the state of the art in
Machine tool
understanding changes in machine tool performance due to changes in thermal conditions (thermal
Thermal error
errors of machine tools). The topics are focused on metal cutting machine tools, especially on turning and
Compensation
milling machines as well as machining centres. The topics of the paper thermal issues in machine tools
include measurement of temperatures and displacements, especially displacements at the tool centre
point, computations of thermal errors of machine tools, and reduction of thermal errors. Computing the
thermal errors of machine tools include both, temperature distribution and displacements. Shortly
addressed is also to avoid thermal errors with temperature control, the inuence of uids and a short link
to energy efciency of machine tools. The paper presents the summary of research work in the past and
current. Research challenges in order to achieve a thermal stable machine tool are discussed. The paper
apprehend itself as an update and not a substitution of two published keynote papers of Bryan et al. [28]
in 1990 and Weck et al. [199] in 1995.
2012 CIRP.
options in analysing the thermal behaviour of machine tools. In reduce waste of energy. Section 5 provides an overview of
Section 2, the advancements in measurement of thermal errors and temperature control, uids, and energy efciency.
temperatures are presented. This paper presents an overview of the latest research activities
Numerical methods are currently used to compare different in the eld of thermal errors in machine tools. The paper is
machine tool designs in early stages of development or to simulate organised as follows: The advances in measurement of thermal
effects of temperatures on machine tools to detect the sources of errors and temperatures are given in Section 2. Methods and
thermally induced tool centre point (TCP) displacements. Because research work used and developed for computing thermal errors in
of the costs of computation time for transient simulations, machine tools are presented in Section 3. In Section 4, the research
engineers often just use steady state results for their simulations. activities in reduction of thermal errors are summarised. A
However, it is observations of transient behaviour that show the summary of the activities of researchers in the area of temperature
TCP displacement can change its direction during operation control, the inuence of uids and energy efciency are
[16,49,126] if two different time constants are involved or if the recapitulated in Section 5. The paper is concluded by summarising
temperature eld spreads out from a heat source and becomes the activities and future trends.
homogeneous after having been governed by steep gradients.
Furthermore, steady state results do not yield the time dependant 2. Advances in measurement of thermal errors and
behaviour leading up to the steady state situation, this period can temperatures
represent a number of hours.
Advances in mechatronics [162] lead to higher precision of 2.1. Measurement of thermal deformations
machine tools. However, to achieve higher precision, the predict-
ability of thermal stability of machine tools becomes more critical, Today, many solutions exist to measure the displacement
especially to avoid costly design modications in later stages of (position and orientation) of machine tool components. Not all of
machine development based on experimental studies. Improve- these systems can be used for the measurement of thermal errors
ments in computational techniques lead to better estimation of as these measurements are specically challenging because of the
temperature distribution and thermally induced displacements at following requirements:
the TCP [97]. Today the processing power of state of the art
personal computers is high enough to handle such computing measurement of all relevant geometrical error parameters;
requirements. Even the computationally intensive simulations of which includes the relevant working volume;
transient effects of temperature on a complete machine tool with with sufciently low measurement uncertainty; and
the nite element method (FEM) can be done in a reasonable and within a short time, so that the effect of changing
calculation time. An overview of modelling and computing thermal temperatures on the geometrical error parameters can be
errors is given in Section 3. monitored by the measurement system.
The technical denition for all arrangements reducing thermal
errors at the TCP of machine tools is thermal reduction, which is Choosing the correct measurement system depends on the
considered in Section 4. Early thermal reductions can be found in error sources. Environmental inuences like machine shop
the gridiron pendulum of Harrison developed in 1726. In this clock temperature generally lead to slow changes of machine tool
pendulum the thermal inuence on a clock is minimised by temperatures, but affect the whole volumetric performance.
combining brass and steel, two materials with different coef- Internal inuences like heat generated by bearings and guideways,
cients of expansion (Fig. 1). lead to a local deformation of the machine tool structure and
In machine tools the design for thermal stability is used e.g., for therefore lead to displacements, which change the volumetric
mountings of linear scales. Other arrangements, such as heating performance partially [4,123125,133135,139,168,173,182,196].
and cooling devices to stabilise temperature distribution, are used Displacements caused by internal heat sources are less predictable
for reducing the thermal error [93,181]. On the other hand, and can change faster than those caused by the environment.
mechatronics are frequently used for error compensation. Thermal In the last two decades the International Organization for
errors are computed with various numerical algorithms and a Standardization (ISO) published several standards: ISO 230-3 [88],
movement to compensate for the thermally induced errors is temperature distortion of machine tools, ISO 10791-10 [85],
generated by a controlled actuator. temperature distortion of machining centres, and ISO 13041-8
Controlling temperature is still a critical requirement for high [86], temperature distortion of turning machines. These standards
accuracy manufacturing. Different media are used to stabilise the provide methods for a systematic analysis of the thermal
temperature distribution of machine tools as well as the behaviour of machine tools with main spindles. This analysis
environmental temperature in the shop oor. The material includes the ETVE (environmental temperature variation error),
properties of the chosen uid primarily inuence the design and the thermal distortions caused by rotating (main) spindles and
energy efciency of the cooling system. In the discussion of energy thermal distortions caused by moving linear axes. Measurement of
efciency of machine tools, it has been identied that temperature the thermal distortion between the tool side and workpiece side is
rise in machine tools is always provoked by components with common for the standards. Usually a test mandrel is clamped in the
energy loss and further power expenditure is required to reduce spindle and a measurement setup with ve displacement
errors induced by this loss. More efcient components also help measurement devices is xed onto the table of machining centres
(Fig. 2) or in the case of a turning centre the tool holder. In [8,23] it
is shown that all structures have a thermal resonance frequency. If
a machine tool is loaded with ambient temperature variations, any
frequency can be predicted with the step response method.
For measuring the thermal distortion caused by moving a linear
axis, the standards advise to detect the error at both ends of the
axis travelling distance as shown in Fig. 5 [85,88]. If in case of such
a measurement the thermal distortion causes a tilt motion of the
axis, the angular displacements at both ends of the travelling
distance are different. If just one position is measured, the angular
displacement can also be interpreted as a thermal location error
like a deviation in the squareness of two axes. With laser
interferometers the displacements in axial directions along the
Fig. 1. Simplied model of a gridiron pendulum, 1: steel, 2: brass [44]. travelling distance can be measured at various positions
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 773
Fig. 2. Measurement setup for thermal distortions caused by a rotating spindle and
environmental temperature variation error (ETVE) for a machine tool with a vertical
spindle [88].
Fig. 4. Telescopic double ball bar measurement setup for thermal distortions of a
ve-axis machine tool [41,50].
[51,88,178]. Also, comparator systems are used to detect the
thermal distortion in up to two directions in various positions
along the machine tool axis [141]. distortion is minimised by balancing out the magnitude and
To measure environmental inuence on machine tools, a direction of thermal expansion by different materials, the chosen
temperature chamber was developed. The chamber is able to geometry, and by taking into account different heat loading of the
control the time-dependent air and foundation temperatures adapter parts.
[159]. The thermal inuence of a foundation block was investi- Using touch trigger probes clamped in the machine tools
gated in [160]. spindle has some advantages. With one probe the thermally
Measuring the thermal distortion of machine tools with rotary induced TCP displacements in up to three directions can be
axes is not yet included in the standards. With the R-test device detected and the thermal distortion is detected at the centre line of
[201] and two additional length measurement probes, the thermal the machine tool spindle.
distortion caused by rotating main spindles are detected in up to Touch trigger probes are often used to measure the displace-
ve directions in one measurement position of a ve-axis machine ments between machining operations. In such cases the tool is
tool. With a master ball clamped in the spindle and the R-test changed for a touch trigger probe that detects the actual thermally
sensor nest at the machine tools table, it is possible to measure the induced TCP displacements. Usually at the table a measurement
thermal distortion in three directions for various axis positions of setup with several detecting points is clamped. The measurement
ve-axis machines [143]. The thermal distortion of a rotational setup is either chosen according to the material to be machined,
table can be detected using a master ball clamped at the table. The the structure material of the machine tool, or material with low
temperature rise of the table causes a dimensional growth of the thermal expansion. Traditionally invar, ceramic glass, or carbon-
table. Such measurements are advisable in order to understand bre-reinforced plastic (CFRP) is used for this purpose [192]. CFRP
workpiece errors of ve-axis machine tools [48]. has very low thermal expansion in the direction of the bre [180].
If touch probes are used to detect the thermally induced TCP Deformation sensors with strain gauges and a rod of invar are
displacements when having high relative velocity, e.g., detecting used for measuring the thermal deformations of the machine tool
thermal distortion caused by a rotating spindle, the machine tool frame (Fig. 46). The deformation of the machine tool frame is used
has to be stopped during measuring. Consequently, contactless, to determine the TCP displacements [150].
inductive and capacitive probes (Fig. 3) are most commonly used to The measurement procedure described in the standards is used
detect thermally induced TCP displacements with rotating for machine tools under no-load or nishing conditions. For
spindles [40,183]. To detect them using a telescoping ball bar, a simulating the workpiece inuence sometimes the machine tool
special adapter was designed [41]. The adapter was mounted near table is loaded with a mass. In a hydraulic brake emulates the
the milling head of a running spindle for measuring the TCP reaction torque from the cutting process and thus loads the spindle
displacement when the inuence of the spindle needs to be as shown in Fig. 6 [17]. The magnitude of the load depends on the
detected. Therefore, the adapters thermal distortion itself can oil pressure and is controlled by a choke which is regulated
cause wrong measurement results. In the horizontal direction a manually. A loaded double ball bar (LDBB) [2] is applied to evaluate
rotational symmetric design was chosen to minimise the adapters the static behaviour of machine tools under loaded conditions.
inuence (Fig. 4). In the vertical direction, the adapters thermal Machine errors caused by temperature change in time and
space can be experimentally determined by temperature invariant
reference objects or independent length measurements. Methods
and procedures are described in [4,162,203]. New measurement
techniques based on multilateration enable the online correction
of task-specic machine movements. For instance the M3D3
system developed at PTB [203] uses a set of high-accuracy tracking
Fig. 5. Measurement setup for thermal distortions caused by a moving linear axis of
Fig. 3. Measurement setup for detecting axial expansion of an air spindle with a machining centre using a touch-trigger probe, 1: machine table, 2: machine
capacitive-type measurement probe and a master ball [154]. spindle, 3: touch-trigger probe, 4: probing points (total of 12) [88].
774 J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791
2.1.2. Measuring local displacement vectors Fig. 7. Left: three axis milling machine; right: infrared temperature measurement of
Similar to the volumetric calibration, there are several the Z-axis with (a) surfaces prepared with adhesive tape and (b) unprepared
measurement systems that can be used for the measurement of metallic surfaces of spindle housing [142].
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 775
Fig. 8. Reection of heat on a Rugotest reference part for turning with Ra 0.4 mm to
50 mm at room temperature of 20.5 8C, from IWF.
stability lobes, the stability lobe generated from the measurements research became much more effective when the FEM started to be
obtained on a cold spindle did not predict the cutting stability applied and developed.
properly when the machining experiment was conducted using a The FEM has enabled in-depth analysis of the thermal behaviour
warmed up spindle. On the other hand, the stability lobe obtained of machine tools under the inuence of heat sources present inside
just before the machining (with a warm spindle), taking into the machine tool structure and in its surroundings. Moreover, thanks
account the frequency shift, properly predicted the machining to FEM one can examine the effect of the individual structural
stability. components, both the ones incorporating heat sources and the ones
subjected to the inuence of external heat sources, e.g., varying
3. Computing thermal errors of machine tools ambient temperature. The FEM is also used to determine the
inuence of heat transfer coefcient (lm coefcient) due to free and
Temperature inuences the geometry of workpieces, measur- forced convection [69,127,128,164]. The accuracy of the geometrical
ing equipment, and machine tools. Deviations from the reference modelling of the machine tool structure has increased signicantly,
temperature of 20 8C [32,84,130], temporal and spatial tempera- if thermal displacements are caused directly by different tempera-
ture variations [82], as well as the material coefcient of thermal tures, by strains, or even by power losses.
expansion (CTE) have to be known for thermal error compensation. A machine tool error in a numerical controlled (NC) axis results
A deviation of the temperature from the reference temperature from the mutual displacements of the individual components
causes, in case of an isotropic CTE, a linear length change in space. depending on the operating conditions. The error changes with the
In case of an anisotropic CTE, the length change varies in space. heat generation and heat transmission conditions. Therefore the
Temporal temperature variations cause varying length changes in computing of thermal errors must be based on the very precise
time. Spatial temperature variations cause deformations depend- modelling of all the major thermal phenomena taking place in the
ing on position. machine tool as it operates [92,94,95,103]. In order to accurately
In high-precision length measurements it is common practice represent the behaviour of the individual machine tool compo-
to numerically correct linear length changes due to constant nents, a model should be ne-tuned on the basis of precise
temperature deviations for both workpieces and machine scales. measurements of the temperature and displacements at specied
The challenge nowadays lies in the determination and correction of points of the machine tool. The precise identication of tempera-
non-linear length changes. In Fig. 13, for instance, the inuence of a tures and displacements in a machine tool prototype is vital for the
constant temperature gradient of a machine tool is illustrated. It creation of an accurate model and for its evaluation, especially
bends the machine bed which nally contributes to straightness, when thermal errors are to be compensated on the basis of the
rotational, and squareness errors of the guideways. The tempera- model and numerically simulated displacements [100,104,107].
ture gradient can be measured by means of temperature sensors. Today, the ambition of every designer of highly efcient
This kind of error can be described in a thermal kinematic model machine tools, particularly the ones to be used for precision
and nally corrected. machining, is to be able to accurately predict thermal errors
Fig. 14 shows geometric machine deviations caused by a local through numerical simulations. Accurately predicted errors are the
heat source. basis for their effective and easy compensation.
The most rational, although difcult and laborious, way of
3.1. Criteria and ways of determining thermal errors modelling is the integrated modelling of entire machine tool
structures, which takes into account the thermal interactions
Tracing back the history of research on the identication and between the individual assemblies and the machining processes.
reduction of machine tool thermal errors, one can notice that the An integrated computing model enables one to effectively improve
the thermal performance of the whole machine tool, i.e., to
minimise thermal errors, and to precisely predict thermal errors
for error compensation purposes [109,110].
In many cases, however, machine tool designers need a quick
assessment of the possibility of improving the main (e.g., spindle
and feed) assemblies. The modelling of assemblies isolated from
the whole machine tool is much less time-consuming. Such
modelling of thermal errors is often justied and in many cases
precedes integrated modelling.
Fig. 20. Algorithm for calculating power losses in torque motors and bearings for
Fig. 18. Experimental and computed temperatures of two spindle bearings [105]. modelling of thermal behaviour of tilting rotary table with direct drives [5].
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 779
Fig. 23. Machined surface with thermal error compensation (thermal distortion in
Z), left: without smoothing, right: with smoothing [121].
rotational speed, the work cycle, the load resulting from the torque
Fig. 21. Thermal FEM model of a bearing and its surroundings [76].
of elastohydrodynamic friction in the lubricating lm of the
turning parts, and the heat transfer conditions. In order to reduce
deformations determine the thermal error in the rotation axis of power losses, airoil lubrication is used on high-speed screws with
the spindle. a large pitch.
Other error models aimed at compensating the thermal axial Relations for losses due to load and friction in the lubricating
displacement of machine tool spindles are models based on data medium can be found in ball screw manufacturer catalogues. But
from measuring temperatures alone, displacements alone or both the modelling of heat transmission, both natural and forced (the
temperatures and thermal displacements. This group 4 also cooling of the nut and the internal cooling of the screw), is difcult.
includes error models exploiting: articial neural networks, linear Similarly as in the case of the fast change of the headstock position
and non-linear regression, dynamic models, transfer function, relative to the machine tools bed or stand, an effective method of
adaptation models, and other [31,33,34,119,193,197,198,207 modelling heat transmission and temperature and strain distribu-
209]. tions in thermally non-stationary states is sought for the fast travel
Among the models in group 4, the model developed by Kim of the screw relative to the nut.
et al. [121] deserves to be presented. For the compensation of Z- A simplied approach to the modelling of the thermal
axis errors in a machining centre with a maximum spindle speed of behaviour of the ball screw transmission was presented in [59].
25,000 rpm, it distinguishes two errors: an axial offset error, which An attempt to experimentally and computationally determine
is assumed to stem from the behaviour of the test bar/spindle joint, temperatures in the nut area and the temperature at one point of
and a thermal error which is dened as the sum of the the screw for an intermittently working, pre-tensioned screw was
temperature-dependent deformations and deections of the made in [122].
headstock and column components. Each thermal mode is When determining thermal errors arising in ball screws, it is
correlated with the temperature of the corresponding component very important to accurately identify the distribution of tempera-
through a thermal mode gain. Mathematical models for Y- and Z- ture along the screw and on this basis determine its axial thermal
axis thermal distortions are expressed as elongation. In the research undertaken by Heisel et al. [70], an
infrared camera was used to identify temperatures. An example of
dy GTy T y (1) an experimentally determined temperature distribution for 4000
for six-temperature measuring points, where cycles, modelled and measured positioning errors are shown in
Fig. 24.
T
Gy GEy ; GBy ; GEy By Similar investigations were carried out by Horejs et al. [77]. A
h h c ; Gc (2)
simple thermalmechanical model of a ball screw with bearings at
T
T3 T4 T5 T6
Ty ; T 3 T 4; ; T5 T6 (3)
2 2
Fig. 24. Ball screw thermal behaviour for a travel distance of 100 mm: top:
temperature distribution measured on the screw (at the beginning: blue, after 4000
cycles: red), bottom: modelled positioning errors compared with experimental data
Fig. 22. Comparisons of spindle axial shift with and without compensation [120]. [122].
780 J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791
both ends (Fig. 25) was used to perform FEM. The numerical model
covers the friction torque of the bearings, all heat transfer
conditions, the nut friction torque, and the external load. The
model was veried by comparing the measured (by resistance
thermometers) temperatures at points located along the screw
with the ones calculated using the numerical model. The Fig. 27. Thermal error variation under dry and wet machining conditions [170].
discrepancy was found to amount to 7%. The positioning error
along the screw with one xed bearing, calculated from a
temperature prole obtained by means of infrared thermography, the difference amounted to 10 mm. This is unsatisfactory in the
differed from the measured one by 10% in the loaded part of the case of machine tools for precision machining. But one should bear
screw and by 8% in the free part. in mind that the machining process introduces many disturbances
The thermo-mechanical model is a substantial step towards the which until now have not been taken into account in thermal error
development of an accurate integrated model for calculating compensation.
power losses, temperatures, and thermal displacements of a ball Articial neural networks (ANN) have been used to model
screw with pretensioned bearings on both sides in its natural thermal errors in positioning for axes with preloaded ball screws
operating conditions. The positioning errors under thermal load [171,172]. A new generation articial neural network based on the
will be smaller with xed bearings at both ends, as schematically wavelet theory deserves attention [113]. Wavelet neural networks
shown in Fig. 26. supported by the evolutionary particle swarm optimisation (PSO)
In [205] thermal error modelling by FEM is limited to the technique, dramatically increase convergence and assure much
heating and thermal elongation of the ball screw alone, neglecting smaller screw nut temperature and positioning prediction errors
the effects originating from the machine tool structural bodies. than conventional ANNs.
Only the inuence of the bearings and the trapezoidal distribution In the case of high power losses in the nut and high rates of feed,
of thermal load in the screwnut joint were taken into account. a substantial reduction in thermal errors can be achieved by
Good agreement with measurements was obtained, but the cooling the nut and the whole screw from the inside. In [204] one
analysis was limited to the table system isolated from the machine can nd an analysis of the effect of internal cooling of the screw in
tool. axes X and Z of a lathe slide on the thermal errors in these axes.
In [58] attention was drawn to the fact that it is necessary to Fig. 28 shows calculated losses in the ball screwnut unit for X- and
take into account changes in the tension of the ball screw that Z-axes when moving with and without a load, as a function of
accompany the changes in its temperature. A relationship for travel rate. Also the effect of cooling in the two axes on the thermal
screw stiffness was presented. It was also shown that the thermal error of the Z-axis is presented. It was demonstrated that using
errors of the screw can be reduced by modifying its mounting even small amounts of cooling oil reduces the thermal error
stiffness and reducing the signicant inuence of the machine tool
body in which the screw is mounted.
In [169,170] an attempt was made to develop a model for
predicting the thermal errors of a three-axis machining centre due
to heat generation in its linear NC axes as a function of varying
operating conditions. The model was based on experimental tests
which indicated that the rise in the temperature of the ball screw
nut during operation has the strongest effect on the thermal errors
in the NC axes. The main factors that determine the magnitude of
thermal errors in the NC axes were: the machine tool operating
conditions, the power losses in the ball screw nut, and the rate of
travel. Mutual interaction between NC axes and the table was
observed. It was found that in the cold state the table load and the
load generated by the machining forces have a signicant effect on
the thermal errors. It was shown that the thermal error rapidly
increases at the moment when wet cutting with a coolant becomes
dry cutting (Fig. 27). The hybrid Bayesian network for the
classication of tests and the powerful regression tool support
vector machine model (SVM) for the efcient mapping of
temperature data with a positioning error were used to predict
the thermal error of positioning. A comparison of the predicted
positioning errors in axis X with the measured ones showed that
Fig. 28. Top: computed power losses in ball screw-nut unit for X- and Z-axes of a
lathe, bottom: cooling effect (cooling of nut and ball screw) for thermal error in Z-
Fig. 26. Impact of bearings conguration on positioning errors [77]. axis [204].
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 781
Fig. 29. (a) Thermal equivalent circuit diagram of the ball screw system, R: thermal
resistance, TA: ambient temperature, T: rolling element temperature, (b)
computation results of TCP-displacements without nut cooling and with two
different types of nut cooling: nut cooling 1 and nut cooling 2 [148].
twofold and that cooling in the two axes mutually affects the
thermal error in each of the axes. This provides an argument for the
integrated computing and analysis of thermal errors for similar
designs, particularly considering applications of high loading
generated through high machining forces.
In [147,148] a thermal equivalent circuit model of a ball screw is
presented. It has been shown that cooling the nut can reduce the
thermal errors (Fig. 29). Furthermore, the inuence of the coolant
temperature variation is considered in the simulation. Nut cooling
was chosen in this study since the nut was the non-rotating Fig. 30. Measurement results: (a) arbitrary feed rate cycle of the Y-axis, (b)
element making it more practical to feed through with cooling temperature variations of Y-axis linear motor, (c) temperature variations and
uid. thermal error (LM: linear motor) [120].
The calculated heat uxes were applied as heat sources with the
simulation set for 90 min. The resulting maximum displacement
appeared at the top of the Z-carriage, whereas the maximum
temperature was at the XZ-plate. The thermal displacements at the
TCP were also generated to investigate the effect of the thermal
error on the TCP. Maximum displacements of the TCP of Dx = 5 mm,
Dy = 10 mm, and Dz = 6 mm arise when taking the linear motors
cooling system into account.
With a knowledge-based description of the boundary condi-
tions, simple types of load were simulated using FEM [163]. A
procedure for computing the heat transfer coefcient at a machine
tool surface depending on air temperature, temperature distribu-
tion of the machine tool, and orientation of the surface was
developed [164]. The adaptation of the heat transfer coefcient
allows a more accurate modelling of convective heat transfer. To
model the inuence of surface roughness and pressure [89] in the
heat transfer in joints, special FEM elements are developed in
[161]. In [116] a formula to compute the thermal conductivity of
joints is given.
analytical model;
Fig. 35. Schematic illustration of hydrostatically supported precision spindle and
numerical model, such as an FEM model;
experimental setup (numbers indicate temperature sensors) [153].
transfer function model; and
neural network.
Fig. 39. Measured proles of face turned surfaces machined with and without
compensation [154].
Fig. 37. Relative displacements measured and estimated for randomly changing 4.2.3. Estimation of thermal deformation based on temperature
spindle speed and ambient temperature [154].
distribution information with use of a neural network model
In [155] a schematic illustration of the experimental setup to
by the growth of the spindle in its axial direction was compensated estimate the relative displacement between the spindle and the
to improve the machining accuracy [154]. Fig. 3 shows an workpiece table of a vertical machining centre is given. Thermo-
illustration of the experimental setup. couples are attached to various parts of the machine tool.
In this case both the heat generated due to the spindle rotation Information on the temperature distribution of the machine is
and the temperature variation of the ambient air are considered to input to the neural network to estimate the thermal deformation.
estimate the growth of the spindle. The transfer functions between The relative displacement between the rotating tool and the
the rotational speed of the spindle and the thermal deformation workpiece caused by the heat generated due to the spindle rotation
and that between the ambient temperature variation and the was estimated by employing neural network models.
thermal deformation were separately identied by experiments. A It was shown experimentally that the relative displacement
simple method was developed to estimate and compensate the was successfully estimated by the neural network. It was also
thermal deformation by utilising the estimated transfer functions shown that the points of temperature measurement can be
and the convolution integral. reduced without sacricing the estimation accuracy by utilising
Fig. 37 shows an example of experimental results showing the the time-series data of temperature. Fig. 40 shows examples of
relative displacements between the spindle face and the workpiece experimental results obtained while the spindle speed was
measured and estimated for randomly changing spindle speed and randomly varied, which compare the measured and estimated
ambient air temperature. The thermal displacement was estimated relative displacements. The relative displacements in both Y and Z
during actual ultraprecision face turning of aluminium and directions were estimated within 3 mm deviation.
compensated by adjusting the depth of cut accordingly. Fig. 38
shows an example of the experimental data. The machining error 4.2.4. Estimation of thermal deformation based on strain distribution
due to the above-mentioned heat sources was successfully reduced information with use of a neural network model
to less than 15% as shown in Fig. 39. It is expected that the estimation accuracy will be improved if
Brecher et al. [14,15,18,19] use PT1PT3 time delay elements as one can measure the strain distribution of the machine structure.
transfer functions. Depending on the actual speed and power of the Conventional strain gauges cannot be utilised for this purpose as
main spindle, the compensation algorithm calculates the displace- their thermal properties are so chosen that only the mechanical
ment of the TCP. The main advantage is that only internal data from strain is detected and the thermal effect is cancelled when they are
the control system are used as input. By combining the indirect attached to the surface of structure.
with the direct compensation with few measuring points during Fig. 41 shows a schematic illustration of the arrangement of
compensation the modelling effort can be reduced without a loss of conventional strain gauges to measure the total strain of the object,
accuracy. or the sum of the thermal strain and the mechanical strain. Two
For a combination of different inputs, temperature sensors dummy gauges are attached to the block which is not exposed to
placed on the machine tool structure, and rotational spindle speed, the mechanical stress but placed under the same thermal condition
the transfer functions are presented in [78,79,140]. The model uses
Fig. 38. Thermal displacement estimated during machining and measured error of Fig. 40. Comparison of tool work table relative displacements measured and
machining with compensation [154]. estimated while spindle speed is randomly varied [155].
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 785
Fig. 43. Examples of face turned surface proles without (top) and with (bottom)
compensation [157].
Table 1
Film coefcients of cooling uids used in machine tool applications [47].
a thermal shower;
the linear bearing hydrostatic oil;
the work spindle bearing oil;
the work spindle motor; and
Fig. 48. Cross section of a Coanda-effect tubing to control the temperature of a process uid.
machine tool spindle [46].
References
[1] Abele E, et al, (2010) Machine Tool Spindle Units. CIRP Annals Manufacturing
Technology 59/2:781802.
[2] Archenti, A., 2011, A Computational Framework for Control of Machining
systems Capability From Formulation to Implementation, PhD Thesis, KTH
Royal Institute of Technology, Stockholm, Sweden, ISBN 978-91-7501-162-2.
[4] Bayer-Helms F, et al, (1985) Langenstabilitat bei Raumtemperatur von Proben
der Glaskeramik Zerodur. Metrologia 21(2):4957.
[5] Blazejewski A, et al, (2010) Modelling Thermal Deformation of Tilting Rotary
Table with Direct Drive System. Journal of Machine Engineering 10(4):1832.
ISSN 1895-7595.
[6] Bonse, R., 1998, Thermisches Last-Verformungsverhalten von Werkzeug-
maschinen, Diss. RWTH Aachen. ISBN 3-8265-6102-3.
[7] Bonse R, Weck M (1994) Indirekte Kompensation Thermo-elastischer Verlager-
ungen bei Einwirkung mehrerer Warmequellen, VDW 8493.
[8] Bosch J, Bryan J (1995) Chapter 8, Temperature Fundamentals. Coordinate
Measuring Machines and Systems, Marcel Dekker Inc..
[9] Bossmanns B, et al, (1999) A Thermal Model for High Speed Motorized
Spindles. International Journal of Machine Tools and Manufacture 93:
13451366.
Fig. 52. Box precision optics grinding machine temperature control [151].
[10] Brecher C, et al, (2006) Messtechnische Untersuchung des thermoelas-
tischen Verlagerungsverhaltens von Werkzeugmaschinen, 11. Dresdner
WZM-Fachseminar.
[11] Brecher C, et al, (2009) Interaction of Manufacturing Process and Machine
6. Conclusion Tool. Annals of the CIRP 58(2):588607.
[12] Brecher C, et al, (2009) Modelling of Thermal Behaviour of a Milling Machine
In the past 20 years since Bryan et al. [28] published the keynote Due to Spindle Load. 12th CIRP Conference on Modelling of Machining Opera-
tions, vol. 2, 673678. ISBN 978-84-608-0866-4.
paper on thermal effects, research has continued as documented [13] Brecher C, et al, (2009) Optimierung des thermischen Verhaltens von Fras-
by many publications in this area. The investigations in the past 20 maschinen. Zeitschrift fur wirtschaftlichen Fabrikbetrieb 104:437444.
years are driven by developments in measurement devices, like [14] Brecher C, et al, (2010) Compensation of Thermo-depend Machine Tool
Deformations Due to Spindle Load Based on Reduced Experimental Procedure
infrared cameras, and measurement strategies, by increased
and Modelling Effort Synthesis Between Direct and Indirect Compensation.
computational power, and by the increased use of numerical Eighth International Conference on High Speed Machining, 978-2-9534170-2-9.
controlled machine tools. pp. S.89S.95.
In the future, precise workpieces of different materials with [15] Brecher C, et al, (2010) Compensation of Thermo-Dependent Machine Tool
Deformations Due to Spindle Load Based on Reduced Modeling Effort. 14th
different coefcients of thermal expansion have to be manufac- International Conference on Mechatronics Technology (ICMT 2010), Icho-kaikan,
tured on machine tools. One of the main error sources in S.295S.302.
production is a changing ambient temperature as well as ambient [16] Brecher C, et al, (2010) Thermisch bedingtes Verformungsverhalten von
Frasmaschinen. VDI-Z 152(9):7881.
temperatures other than 20 8C. In order to compensate the errors [17] Brecher, C., et al, (2009) Stressing Unit for Modelling of Thermal Behaviour of
caused by the ambient temperature, the thermal behaviour of the a Milling Machine, 12th CIRP Conference on Modelling of Machining Opera-
machine tool itself, the coefcients of thermal expansion of the tions, vol. 2, 727730. ISBN 978-84-608-0866-4.
[18] Brecher C, Hirsch P (2004) Compensation of Thermo-elastic Machine Tool
machine tool components and the workpieces, and the thermal Deformation Based on Control internal Data. Annals of the CIRP 53(1):299304.
conductivity of the materials have to be known well. More precise [19] Brecher C, Hirsch P (2006) Kompensation thermoelastischer Verlagerungen,
measurement devices and measurement strategies should be 11. Dresdner Werkzeugmaschinen-Fachseminar.
[20] Brecher C, Wissmann A (2006) Messtechnische Untersuchung des thermo-
developed to reduce the uncertainties of temperature and
elastischen Verlagerungsverhaltens von Werkzeugmaschinen, 11. Dresdner
displacement measurement of machine tools and workpieces. Werkzeugmaschinen-Fachseminar.
In order to identify and prioritise critical parameters inuen- [21] Brecher C, Witt S (2004) Static, Dynamic and Thermal Behaviour of Machine
cing the machining results, the simulation-based sensitivity Tools with Regard to HPC. High-Performance Cutting 227242.
[22] Bryan J (1969) International Status of Thermal Error Research. Annals of the
analyses as described in [43] should be improved. These sensitivity CIRP 16(2):203215.
analyses should be focused on material parameters, model [23] Bryan J, et al, (1965) Thermal Effects in Dimensional Metrology, ASME 65-
description, model simplications, and the description of bound- PROD-13.
[24] Bryan J, et al, (1967) Expansion of a Cutting Tool During Chip Removal. Annals
ary conditions. of the CIRP 16(1):4952.
Compared to the research effort on measurement, simulation, [25] Bryan J, et al, (1978) A Practical Solution to the Thermal Stability Problem in
and compensation strategies related to thermal errors, the Machine Tools, SME, MR-72-138.
[26] Bryan, J., et al., 1982, An Order of Magnitude Improvement in Thermal
inuence of the machining process and the inuence of the Stability with the Use of Liquid Shower on a General Purpose Measuring
coolant have not been studied with the same intensity. These are Machine, SME Precision Machining Workshop, UCRL-87591.
certainly elds where future work should focus to reduce thermal [27] Bryan J, et al, (1986) Shower and High-Pressure Oil Temperature Control.
Annals of the CIRP 35(1):359364.
errors of machined parts.
[28] Bryan J, et al, (1990) International Status of Thermal Error Research. Annals of
There are many indications that thermal effects on machine the CIRP 39(2):645656.
tools have become more important. However, the discussions on [29] Buchman K, Jungnickel G (1978) Die Warmeubertragung an Be-und Verarbei-
tungmaschinen PWR, vol. 20. Wroclaw, ITMiA. p. 279.
energy efciency may cause a shift in the industry. In the near
[30] Carlisle, K., 1996, Design, Build and Development of an Ultra Precision
future not just thermally stable machine tools will be successful at Machining Facility, MSC Thesis, Craneld.
the market. It is conceivable that the market desires thermally [31] Chang CW, et al, (2005) Dynamic Model Based ion Genetic Algorithms of
stable and energy efcient machine tools [112]. Prediction for the Thermal Deformation of Machine Tools. Materials Science
Forum 505507:163168.
[32] Chapman M (2003) Limitations of Laser Diagonal Measurements. Precision
Acknowledgements Engineering 27(4):401406.
[33] Chen JS (1996) Neural Network-based Modelling and Error Compensation of
Thermally-induced Spindle Errors. International Journal of Advanced Manu-
The authors would like to thank all CIRP colleagues and experts facturing Technology 12:303308.
who have given input to this keynote paper. A special thanks to the [34] Chen JS, et al, (1995) Quick Testing and Modelling of Thermally-induced
following colleagues who have contributed with detailed informa- Errors on CNC Machine Tools. International Journal of Machine Tools and
Manufacture 35(7):10631074.
tion: Ess M., ETH Zurich, Haber D., WZL Aachen, Heisel U., IFW [35] Chen JS, et al, (2003) Characterization and Models for Thermal Growth of a
Stuttgart, Horejs O., RCMT Prague, Ihlenfeldt S., Fraunhofer IWU Motorized High Speed Spindle. International Journal of Machine Tools and
Chemnitz, Jatzkowski P., WZL Aachen, Konig J., TU Berlin, Manufacture 43:11631170.
[36] Chen TY, et al, (1999) Optimum Design of Headstocks of Precision Lathes.
Lundholm T., KTH Stockholm, Mitsuishi M., University of Tokyo, International Journal of Machine Tools and Manufacture 39:19611977.
Stehle T., IFW Stuttgart, Weikert S., ETH Zurich, Wissmann A., WZL [37] Chou C (1990) Liquid Temperature Control for Precision Tools. Annals of the
Aachen CIRP 39(1):535543.
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 789
[38] Ciddor PE (1996) Refractive Index of Air New Equations for the Visible and [80] Hornych J, et al, (2009) Issues in Identication of Thermal Transfer Functions
Near Infrared. Applied Optics LP 35(9):15661572. of Machine Tools. Proceedings of 9th EUSPEN 451454.
[39] DeBra DB, et al, (1986) Shower and High Pressure Oil Temperature Control. [81] Hornych J, et al, (2009) Thermomechanical Transfer Functions and Control
CIRP Annals 35(1):359363. of a Machine Tool Cooling System. Modern Machinery (MM) Science Journal
[40] Dehaes, J., 1998, Studie over de Thermische Beinvloeding van de Bewer- 9697.
kingsnauwkeurigheit van een Freesmachine, Diss. K.U. Leuven. [82] Hou D, et al, (2004) A Novel FEA Model for the Integral Analysis of a
[41] Delbressine FML, et al, (2006) Modelling Thermo-mechanical Behaviour of Machine Tools and Machining Processes. Key Engineering Materials 257
Multi Axis Machine Tools. Precision Engineering 30(1):4753. 258:4550.
[42] Denkena B, et al, (2007) Kompensation thermischer Verlagerungen. Werk- [83] Huang Y, et al, (1999) Optimisation of Fixture Design with Consideration of
statttechnik Online 97(1112):913917. Thermal De-formation in Face Milling. Transactions of NAMSRI/SME 27:
[43] Denkena B, et al, (2009) Sensitivitatsanalyse fur ein Simulationsmodell. 322340.
Werkstatttechnik Online 99(5):294299. [84] ISO 1. (2002) Geometrische Produktspezikation (GPS) Referenztemperatur fur
[44] Dobrinski P, et al, (2006) Physik fur Ingenieure, Teubner Verlag978-3-8351- die Geometrische Produktspezikation und-prufung, Verein Schweizer Maschi-
0020-6. nen-Industrielle Zurich.
[45] Donmez MA, et al, (1986) A General Methodology for Machine Tool Accuracy [85] ISO 10791-10, 2007, Test Conditions for Machining Centres Part 10:
Enhancement by Error Compensation. Precision Engineering 8(4):187196. Evaluation of Thermal Distortion, Genf, Schwitzerland.
[46] Donmez MA, et al, (2007) A Novel Cooling System to Reduce Thermally- [86] ISO 13041-8, 2004, Test Conditions for Numerically Controlled Turning
Induced Errors of Machine Tools. Annals of the CIRP 56(1):21524. Machines and Turning Centres Part 8: Evaluation of Thermal Distortions,
[47] Dornfeld D, et al, (2009) Precision Manufacturing, Springer, New York, ISBN: Genf, Schwitzerland.
978-1441940841. [88] ISO 230-3, 2007, Test Code for Machine Tools Part 3: Determination of
[48] Ess M, et al, (2011) Thermal Displacements of Rotary Axes. Annual Meeting Thermal Effects, Genf, Schwitzerland.
MTTRF. [89] Itho S, et al, (1992) Behavior of Interface Pressure Distribution in a Single
[49] Eversheim W, et al, (2006) 100 Jahre Produktionstechnik: Werkzeugmaschi- Bolt-Flange Assembly Subjekt to Heat Flux. Journal of Engineering for Industry
nenlabor WZL der RWTH Aachen von 1906 bis 2006, 978-3540333159. 114:231236.
[50] Florussen GHJ, et al, (2003) Assessing Thermally Induced Errors of Machine [90] Ito Y (2010) Thermal Deformation in Machine Tools, Mcgraw Hill Book Co.978-
Tools by 3D Length Measurements. Annals of the CIRP 52(1):333336. 0071635172.
[51] Franke M, et al, (2010) Measuring Large 3D Structures Using a Portable 4-arm [91] Jedrzejewski J (1985) Kompensation thermischer Verlagerungen einer Dreh-
Laser Interferometer. Advances in Metrology 3542. maschine. Werkstatt und Betrieb 118:8587.
[52] Fraser S, et al, (1998) Modelling, Identication and Control of Thermal [92] Jedrzejewski J (1988) Effect of the Thermal Contact Resistance on Thermal
Deformation of Machine Tool Structures, Part 1: Concept of Generalized Behaviour of the Spindle Radial Bearings. International Journal of Machine
Modelling. Journal of Manufacturing Science and Engineering 120:623631. Tools and Manufacture 28(4):409416.
[53] Fraser S, et al, (1998) Modelling, Identication and Control of Thermal [93] Jedrzejewski J, et al, (1973) Zur Erwarmung von Drehmaschinen-Spindelkasten,
Deformation of Machine Tool Structures, Part 2: Generalized Transfer Func- vol. 4. WZM International. pp. 4750.
tions. Journal of Manufacturing Science and Engineering 120:632639. [94] Jedrzejewski J, et al, (1977) Warmeubergangsverhaltnisse an Spindelkasten
[54] Fraser S, et al, (1999) Modelling, Identication and Control of Thermal von Drehmaschinen. Industrieanzeiger 99(74):14361439.
Deformation of Machine Tool Structures, Part 3: Real-Time Estimation of [95] Jedrzejewski J, et al, (1988) Description of the Forced Convection along the
Heat Sources. Journal of Manufacturing Science and Engineering 121:501508. Walls of Machine-tool Structures. Annals of the CIRP 37(1):397400.
[55] Fraser S, et al, (1999) Modelling, Identication and Control of Thermal [96] Jedrzejewski J, et al, (1989) Beurteilung der Berechnungsmethode fur die
Deformation of Machine Tool Structures, Part 4: A Multi-Variable Closed- Bestimmung der Energieverluste in Walzlagern. Schmierungstechnik 20:
Loop Control System. Journal of Manufacturing Science and Engineering 243244.
121:509516. [97] Jedrzejewski J, et al, (1990) Numerical Optimization of Thermal Behaviour of
[56] Fraser S, et al, (1999) Modelling, Identication and Control of Thermal Machine Tools. Annals of the CIRP 39(1):379382.
Deformation of Machine Tool Structures, Part 5: Experimental Verication. [98] Jedrzejewski J, et al, (1991) Numerical Modelling of the Thermal Behaviour of
Journal of Manufacturing Science and Engineering 121:517523. Spindle Bearing Assemblies for Precise Machine tools. Proceedings of the CIRP
[57] Gibson AO, et al, (1996) A linear Observer Based Approach for Estimation Conference on PE&MS, 7-5618-0280-3. pp. 525536.
Spindle Bearing Loads. Proceedings of the ASME Dynamic Systems and Control [99] Jedrzejewski J, et al, (1992) A new Approach to Modelling Thermal Beha-
Division 251258. viour of a Machine Tool under Service Conditions. Annals of the CIRP
[58] Gim T, et al, (2001) Ball Screw as Thermal Error Compensator. Proceedings 41(1):455458.
form ASPE Annual Meeting. [100] Jedrzejewski J, et al, (1992) Thermal Displacements Compensation of Man-
[59] Gleich S (2007) Approach for Simulating Ball Bearing Screws in Thermal ufacturing Cells Using a Universal Correcting Temperature Function. Pro-
Finite Element Simulation. Journal of Machine Engineering 7(1):101107. ceedings of CSME Forum 675680.
[60] Gontarz A, et al, (2010) Energy Consumption Measurement with a Multi- [101] Jedrzejewski J, et al, (1994) Directions in Improving Thermal Behaviour of
channel Measurement System on a Machine Tool. Intech 16. Spindle Bearing Assemblies in FMS Moduls. Manufacturing Systems
[61] Grossmann G, et al, (1998) Thermisches Verhalten veranderlicher Strukturen. 23(4):317322.
Konstruktion 50(6):2731. [102] Jedrzejewski J, et al, (1996) Articial Intelligence Tools in Diagnostics of
[62] Grossmann K, Jungnickel G (1999) Genauigkeitssteigerung an Werkzeug- Machine Tool Drives. Annals of the CIRP 45(1):411414.
maschinen. Zeitschrift fur wirtschaftlichen Fabrikbetrieb 94(6):320323. [103] Jedrzejewski J, et al, (1996) Thermisches Verhalten von Werkzeugmaschi-
[63] Grossmann K, Jungnickel G (2003) Instationares thermoelastisches Verhalten nen-Gestellen. Industrie Anzeiger 99(65):12431245.
von Vorschubachsen mit bewegtem Walzkontakt, 3-86005-381-7. [104] Jedrzejewski J, et al, (1998) An Approach to Integrating Intelligent Diagnos-
[64] Grossmann K, Jungnickel G (2006) Prozessgerechte Bewertung des thermischen tics and Supervision of Machine Tools. Journal of Intelligent Manufacturing
Verhalten von Werkzeugmaschinen, 3-86005-547-X. 9:295302.
[65] Grossmann K, Jungnickel G (2008) Thermisch Modellierung von Prozessein- [105] Jedrzejewski J, et al, (2004) Hybrid Model of High Speed Machining Centre
ussen an spannenden Werkzeugmaschinen, 978-3-86780-089-1. Headstock. CIRP Annals Manufacturing Technology 53(1):285288.
[67] Hahn, M.H., et al., 2006, Evaluation of an Inexpensive Method to Stabilize the [106] Jedrzejewski J, et al, (2005) High-speed Precise Machine Tools Spindle Units
Temperature of Machine Tool Components, National Institute of Standards Improving. Journal of Materials Processing Technology 162163:615621.
and Technology Internal Report, NISTIR 7292. [107] Jedrzejewski J, et al, (2005) Numerical Analyses and Compensation of HSC
[69] Heisel U, et al, (2003) Warmeubertragungsbedingungen an Werkzeug- Machine Tools Thermal Displacements.pp. 268275 Proceedings Lamdamap,
maschinenwanden. Dima 4:2427. vol. 71861941188.
[70] Heisel U, et al, (2006) Thermography-Based Investigation into Thermally [108] Jedrzejewski J, et al, (2007) Precise Modelling of HSC Machine Tool Thermal
Induced Positioning Errors of Feed Drives By Example of a Ball Screw. Annals Behaviour. Journal of Achievements in Materials and Manufacturing Engineering
of the CIRP 55(1):423426. 24(1):245252.
[71] Heisel U, et al, (2010) Modelling of Interaction Processes in Cutting. Proceed- [109] Jedrzejewski J, et al, (2008) Operational Behaviour of High Speed Spindle
ings of 2nd International Conference on Process Machine Interactions, 978-0- Unit. Modern Machinery (MM) Science Journal 10:4043. ISSN 1803-1269 or
9866331-0-2. 1085-0476.
[72] Henneke KD (2007) Thermisch geeicht-praziser gefertigt. Technica 7:3839. [110] Jedrzejewski J, et al, (2008) Precise Model of HSC Machining Centre for
[73] Herbst U (2000) Kompensation thermoelastischer Verlagerungen an Werk- Aerospace Parts Milling. Journal of Mechanical Engineering 8(3):2941.
zeugmaschinen. Schleiftechnisches Kolloquium 135145. [111] Jedrzejewski J, Kwasny W (2010) Modelling of Angular Contact Ball Bearings
[74] Herbst, U., 2002, Analyse und Kompensation thermoelastischer Verlagerun- and Axial Displacements for High-speed Spindles. CIRP Annals Manufactur-
gen, Diss. RWTH Aachen. ing Technology 59(1):377382.
[75] Hoffmann E (1988) Konvektiver Warmeubergang an Werkzeugmaschinen- [112] Jedrzejewski J, Kwasny W (2011) Study on Reducing Energy Consumption in
Bauteilen. Industrie-Anzeiger 9:3031. Manufacturing Systems. Journal of Machine Engineering 11(3):720.
[76] Holkup T, et al, (2010) Thermo-mechanical Model of Spindles. CIRP Annals [113] Jin C, et al, (2010) Wavelet Neural Network Based on NARMAL2 Model for
Manufacturing Technology 59(1):265268. Prediction of Thermal Characteristics in a Feed System. Chinese Journal of
[77] Horejs O, et al, (2007) Determination of Positioning Error of Feed Axes Due to Mechanical Engineering 23.
Thermal Expansion by Infrared Thermography, ATEM07, JSME-MMD, Septem- [114] Jorgensen BR, et al, (1996) Dynamic of Machine Tool Spindle/Bearing System
ber, 1214. Under Thermal Growth. Proceedings of the ASME Dynamic Systems and Control
[78] Horejs O, et al, (2009) A Compensation Technique of Machine Tool Thermal Division 333340.
Errors Built on Thermal Transfer Functions. Proceedings of 5th International [115] Jungnickel G (2000) Simulation des thermischen Verhaltens von Werkzeug-
Conference on Leading Edge Manufacturing in 21st Cent, Osaka, Japan, 197202. maschinen, Lehre Forschung Praxis.
[79] Horejs O, et al, (2010) Compensation of Machine Tool Thermal Errors Based [116] Jungnickel G (2000) Thermische Simulation von Werkzeugmaschinen, Lehre
on Transfer Functions. Modern Machinery (MM) Science Journal 162165. Forschung Praxis.
790 J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791
[117] Jungnickel G (2006) Modellgestutzte Kompensation von thermisch beding- [151] Morantz P, et al, (2010) Precision grinding for rapid fabrication of segments
ten Verlagerungen in Echtzeitfahigkeit. AG Struktur und Prozessanalyse for extremely large telescopes using the Craneld BoX. Proceedings SPIE
147149. 7739:773905.
[118] Jungnickel G (2006) Prozessgerechte Bewertung des thermischen Verhaltens [152] Mori M, et al, (2009) Design Optimization and Development of CNC Lathe
von Werkzeugmaschinen. AG Struktur und Prozessanalyse 138140. Headstock to Minimize Thermal Deformation. Annals of the CIRP 58(1):
[119] Kang Y, et al, (2007) Modication of a Neural Network Utilizing Hybrid Filters 331334.
for the Compensation of Thermal Deformation in Machine Tools. Interna- [153] Moriwaki T (1988) Thermal Deformation and Its On-Line Compensation of
tional Journal of Machine Tools and Manufacture 47:376387. Hydrostatically Supported Precision Spindles. Annals of the CIRP 37(1):
[120] Kim JJ, et al, (2004) Thermal Behaviour of a Machine Tool Equipped with 393396.
Linear Motors. International Journal of Machine Tools and Manufacture [154] Moriwaki T (1998) Analysis of Thermal Deformation of an Ultraprecision Air
44:749758. Spindle System. Annals of the CIRP 47(1):315319.
[121] Kim KD, et al, (2004) Real Time Compensatory Control of Thermal Errors for [155] Moriwaki T, et al, (1995) Estimation of Thermal Deformation of Machine Tool
High Speed Machine Tools. Proceedings of Instn. Mech. Enrs, 218, Part B, by Applying Neural Network. Transactions of JSME C 61(584):16911696.
Engineering Manufacture, 913924. [156] Moriwaki T, et al, (1997) Prediction of Thermal Deformation of Machine Tool
[122] Kim SK, et al, (1997) Real-time Estimation of Temperature Distribution in a with Strain Sensor. Manufacturing Science and Technology ASME MED 6/
Ball-screw System. International Journal of Machine Tools and Manufacture 1:137142.
37:451464. [157] Moriwaki T, et al, (1998) Development of an Intelligent Turning Machine
[123] Kim SM, et al, (2001) Prediction of Thermo-elastic Behaviour in a Spindle- Equipped with Open-Architecture CNC Controller to Compensate Thermal
bearing Surroundings. International Journal of Machine Tools and Manufacture Deformation of Machine Tool. Proc. CIRP Sponsored International Seminar on
41:809831. Improving Machine Tool Performance, San Sebastian, 317325.
[124] Kim SM (2002) Effect of Bearing Surroundings on the High-Speed Spindle- [158] Niemeier W (2008) Ausgleichungsrechnung, 978-3-11-019055-7. 426 ff.
Bearing Compliance. Advanced Manufacturing Technologies 19:551557. [159] Neugebauer R, et al, (1998) Thermozelle zur Untersuchung des thermischen
[125] Kim SM, et al, (2005) Spindle Housing Design Parameter Optimization Verhaltens von Maschinen. Maschinenmarkt MM das Industriemagazin
Considering Thermo-Elastic Behaviour. International Journal of Advanced 104(40):3437.
Manufacturing Technology 25:10611070. [160] Neugebauer R, et al, (2001) Optimization of Thermal Behavior of machine
[126] Ko TJ, et al, (2003) Particular Behaviour of Spindle Thermal Deformation by Tools with Separated Machine Bed Components. Production Engineering
Thermal Bending. International Journal of Machine Tools and Manufacture VIII(1):137140.
43:1723. [161] Neugebauer R, et al, (2002) A modelling approach to optimize the thermal
[127] Kohut P, et al, (2010) The Inuence of Convective Boundary Condition on behavior of machine tool components. Production Annals of the German
Thermal-Deformation State of Machine Tool (in Czech). Vysoke ucen tech- Academy of Society for Production Engineering 9(1):131134.
nicke v Brne 2127. [162] Neugebauer R, et al, (2007) Mechatronic Systems for Machine Tools. Annals of
[128] Konvica J, et al, (2004) Simulation, Experimental Investigation and Control of the CIRP 56(2):656686.
Thermal Behavior in Modular Tool Systems, Nonlinear Dynamics of Production [163] Neugebauer R, et al, (2009) Improving the Precision by Thermal Simulation.
Systems. pp. 265285. ATZproduktion 0304(2):49.
[129] Koscsak, G., 2007, Ermittlung des instationaren thermischen Verhaltens von [164] Neugebauer R, et al, (2010) An extended Procedure for Convective Boundary
Vorschubachsen mit Kugelgewindetrieb mit Hilfe der Verarbeitung thermo- Conditions on Transient Thermal Simulations of Machine Tools. Production
graphischer Messdaten, Diss. Universitat Stuttgart. ISBN 978-3-00-0236372. Engineering Research and Development 6:641646.
[130] Kruth JP, et al, (2001) Compensation of Static and Transient Thermal Errors on [165] Oliver MA, Webster R (1990) Kriging: A Method of Interpolation for Geo-
CMMs. Annals of the CIRP 50(1):377380. graphical Information Systems. International Journal of Geographical Informa-
[131] Lee JH, et al, (2001) Development of Thermal Error Model with Minimum tion Science 4(3):313332.
Number of Variables Using Fuzzy Logic Strategy. KSME International Journal [166] Pierse M (2007) A Simple Method for Thermal Error Correction of a Grinding
15(11):14821489. Machine. Proceedings Lamdamap, vol. 81861941188.
[132] Lee K-J, et al, (2007) Repeatability Analysis on the Tool Point Dynamics for [167] Popov G (1988) Einuss der Konvektion auf das thermische Verhalten von
Investigation on Uncertainty in Milling Stability. ASME Conference Proceed- Werkzeugmaschinen. Industrie-Anzeiger 23:3839.
ings IMECE2007, vol. 3, 477485. ISBN 0-7918-4297-5 or 978-0-7918-4297-3. [168] Postlethwaite SR, et al, (1999) Machine Tool Thermal Error Reduction An
[133] Li H, et al, (2004) Analysis of Bearing Conguration Effects on High Speed Appraisal. Proceedings Institution of Mechanical Engineers 213(B):19.
Spindles Using an Integrated Dynamic Thermo-mechanical Spindle Model. [169] Ramesh R, et al, (2000) Error Compensation in Machine Tools A Review Part
International Journal of Machine Tools and Manufacture 44:347364. II: Thermal Errors. International Journal of Machine Tools and Manufacture
[134] Li H, et al, (2004) Integrated Dynamic Thermo-mechanical Modelling of High 40:12571284.
Speed Spindles Part1: Model Development. Journal of Manufacturing Science [170] Ramesh R, et al, (2002) Support Vector Machine Model for Classication of
and Engineering 126:148158. thermal Error in Machine Tools. International Journal of Advanced Manufac-
[135] Li H, et al, (2004) Integrated Dynamic Thermo-mechanical Modelling of High turing Technology 20:114120.
Speed Spindles Part 2: Solution Procedure and Validation. Journal of Manu- [171] Ramesh R, et al, (2003) Thermal Error Measurement and Modelling in
facturing Science and Engineering 126:159168. Machine Tools Part I: Inuence of Varying Operating Condition. International
[136] Li S, et al, (1997) A study of Pre-compensation for Thermal Errors of NC ma- Journal of Machine Tools and Manufacture 43:405419.
chine Tools. International Journal of Machine Tools and Manufacture [172] Ramesh R, et al, (2003) Thermal Error Measurement and Modelling in
37:17151719. Machine Tools Part II: Hybrid Bayesian Network-support Vector Machine
[137] Luttrel D (2007) Tutorial on Fundamentals of Thermal Effects: Precision Model. International Journal of Machine Tools and Manufacture 43:405419.
Design Principles, Measurement und Control of Temperatures. ASPE Annual [173] Salje` E, et al, (1988) Comparison of Machine Tool Elements Made of Polymer
Meeting. Concrete and Cast Iron. Annals of the CIRP 37(1):381384.
[138] Maisch M (1993) Software korrigiert geometrische und thermische Fehler. [174] Schwenke H, et al, (2005) Error Mapping of CMMs and Machine Tools by a
Werkstatt und Betrieb 126(11):691694. Single Tracking Interferometer. CIRP Annals 54(1):475478.
[139] Mannan MA, et al, (1992) Investigation into Temperature Dependence of [175] Schwenke H, et al, (2008) Geometric Error Measurement and Compensation
Motor Current Measurements Applied to Monitoring and Adaptive Control. of machines An Update. Annals of the CIRP 57(2):660675.
Annals of the CIRP 41(1):451454. [176] Schwenke H, et al, (2009) On-the-y Calibration of Linear and Rotary Axes of
[140] Mares M, et al, (2010) Application of Mechatronic Approach to Modelling, Machine Tools and CMMs Using a Tracking Interferometer. Annals of the CIRP
Identication and Control of Machine Tool Thermal Errors. Proceedings of 58/1:477480.
29th IASTED International. Conf. Mod.. ISBN 978-0-88986-833-5. [177] Shore, P., 1995, Ultra Precision Machining Facility, PhD Thesis, Craneld.
[141] Mayr, J., 2010, Beurteilung und Kompensation des Temperaturgangs von [178] Song D, et al, (2000) Real-time Measurement of Spindle Thermal Deformatio
Werkzeugmaschinen, Diss. ETH Zurich. Using Interferometers. Optical Engineering 39(8):21142118.
[142] Mayr J, et al, (2007) Comparing the Thermo-mechanical Behaviour of ma- [179] Song HB, et al, (2000) Flow and Heat Transfer Characteristics of a Two-
chine Tool Frame Designs Using a FDM FEM Simulation Approach. Proceed- dimensional Oblique Wall Attaching Offset Jet. International Journal of Heat
ings ASPE Annual Meeting 1720. and Mass Transfer 43:23952404.
[143] Mayr J, et al, (2007) Nutzen thermische Messungen auf Werkzeugmaschi- [180] Spur G, et al, (1988) Thermal Behaviour Optimization of Machine Tools.
nen? MB-Revue 110117. Annals of the CIRP 37(1):400405.
[144] Mayr J, et al, (2008) Simulation and Prediction of the Thermally Induced [181] Spur G, et al, (1989) Ausgleich von Axial- und Winkelverlagerungen an
Deformations of Machine Tools Caused by Moving Linear Axis Using the Drehmaschinen durch Beheizen. Industrie-Anzeiger 89(2):3233.
FDEM Simulation Approach. Proceedings ASPE Annual Meeting. [182] Spur G, et al, (1992) Konstruktionskatalog zur Optimierung des thermischen
[145] Mayr J, et al, (2009) Calculating thermal location and component errors on Verhaltens von Werkzeugmaschinen, Teil1: Aufbau des Katalog, Spindelbau-
machine tools. Proceedings ASPE Annual Meeting. ISBN 978-1-887706-55-1. gruppe. Zeitschrift fur wirtschaftlichen Fabrikbetrieb 87(12):690693.
[146] Mayr J, et al, (2009) Compensation of Thermal Effects on Machine Tools [183] Srinivasa N, et al, (1996) Spindle Thermal Drift Measurement Using the Laser
using a FDEM Simulation Approach. Proceedings Lamdamap, vol. 9. ISBN Ball Bar. Precision Engineering 18:118128.
1861941188. [184] Teeuwsen JWMC, et al, (1989) A General Method for Error Description of CMMs
[147] Mayr J, et al, (2010) Comparing Different Cooling Concepts for Ball Screw Using Polynomial Fitting Procedures. Annals of the CIRP 38(1):505510.
Systems. Proceedings ASPE Annual Meeting. ISBN 978-1-887706-55-1. [185] Turek P, et al, (2010) Methods of Machine Tool Error Compensation. Journal of
[148] Mayr J, et al, (2011) Thermal behaviour improvement of linear axis. Proceed- Machine Engineering 10(4):526.
ings of 11th euspen International Conference, V1, 291294. [186] Uhlmann E, et al, (2010) Simulation of the Thermal Behavior of a Machine
[149] McClure E, Watts R (1968) Thermal Expansion of the Workpiece During Tool Equipped with Linear Motors. Deutsch-Polnisches Seminar.
Turning. ASME Winter Meeting, 68-WA/PROD-24. [187] Uhlmann E, et al, (2007) Adaptronic Compensation of Thermal Strain at
[150] Mitsuishi M, et al, (2001) Development of an Intelligent High-Speed Machin- Machine Tool Spindles using CRP-Bandages. Proceedings of 2nd Manufacturing
ing Centre. Annals of the CIRP 50(1):275280. Engineering Society International Conference.
J. Mayr et al. / CIRP Annals - Manufacturing Technology 61 (2012) 771791 791
[188] Uhlmann E, et al, (2007) Milling Machine Evolution in Area of Conict [201] Weikert S, Knapp W (2004) R-Test, A New Device for Accuracy Measurements
between Efciency. Accuracy and Social Ecology, 12th International Seminar on Five Axis Machine Tools. Annals of the CIRP 53(1):429432.
on High Techn. UNIMEP. [202] Wendt K, et al, (2003) Inspection of Large CMMs by Sequential Multilatera-
[189] Uhlmann E, et al, (2008) Compensation of Thermal Deformations at Machine tion Using a Single Laser Tracker, Laser Metrology and Machine Performance
Tools using Adaptronic CRP-Structures. Proceedings of 41st CIRP Conference on VI. 6th International Conference on Laser Metrology, Machine Tool, CMM and
Manufacturing Systems: Manufacturing Systems and Techniques for the New Robot Performance, 121130.
Front, 183186. [203] Wendt K, et al, (2010) Mobile Multi-lateration Measuring System for High
[190] Uhlmann E, et al, (2010) Application of CFRP Structures for Compensation Accurate and Traceable 3D Measurements of Large Objects. Proceedings of the
of Thermal Strains at Machine Tools. CIRP Paris January Meet STC M 10th ISMQC Paper No 25 14.
Pap Sess. [204] Winiarski Z, et al, (2008) Decreasing of Thermal Errors in a Lathe by Forced
[191] Uhlmann, E., et al., 2010, Kompensation thermischer Verlagerungen an Cooling of Ball Screws and Headstock. Journal of Machine Engineering
Werkzeugmaschinen durch Einsatz von CFK-Strukturen, Fortschr.-Ber. 8(4):122130.
VDI, Reihe 2 Fertigungstechnik, Nr. 675: Hybride Techn. in der Prod., [205] Wu CH, et al, (2003) Thermal Analysis for the Feed Drive System of a CNC
VDI-Verlag, 6577. Machine Centre. International Journal of Machine Tools and Manufacture
[192] Van den Bergh, C., 2001, Reducing Thermal Errors of CMM Located on the 43:15211528.
Shop-Floor, Diss. K.U. Leuven. [206] www.precitech.com, 10.03.2011, www.makino.com, 10.03.2011.
[193] Veldhuis SC, Elbestawi MA (1995) A Strategy of Compensation of Errors in [207] Yang H (2005) Dynamic Neural Network Modelling for Nonlinear, Nonsta-
Five-Axis Machining. Annals of the CIRP 44(1):373377. tionary Machine Tool Thermally Induced Error. International Journal of
[194] Walter R (2006) Mit direkter Kuhlung zu mehr Genauigkeit. Werkstatt und Machine Tools and Manufacture 45:455465.
Betrieb 6:129130. [208] Yang H, et al, (2005) Adaptive Model Estimation of Machine-tool Thermal
[195] Weck M, Brecher C (2006) Werkzeugmaschinen Automatisierung von Maschi- Errors Based on Recursive Dynamic Modelling Strategy. International Journal
nen und Anlagen, 978-3540225072. of Machine Tools and Manufacture 45:111.
[196] Weck M, Brecher C (2006) Werkzeugmaschinen Messtechnische Untersuchung [209] Yang H, Ni J (2005) Adaptive Model Estimation of Machine-tool Thermal
und Beurteilung, 978-3540225058. Errors Based on Recursive Dynamic Modelling Strategy. Machine Tools &
[197] Weck M, et al, (1998) Compensation of Thermal Errors in Machine Tools with Manufacture 45:111.
a Minimum Number of Temperature Probes Based on Neural Networks. [210] Yang SH, et al, (2004) Measurement of Spindle Thermal Errors in Machine
Proceedings of the ASME DSC 64:423430. Tool Using Hemispherical Ball Bar Test. International Journal of Machine Tools
[198] Weck M, et al, (2002) Kompensation thermoelastischer Strukturverformun- and Manufacture 44:333340.
gen. Werkstatttechnik Online 92:327332. [211] Yun WS, et al, (1999) Thermal Error Analysis for a CNC Lathe Feed Drive System.
[199] Weck M, et al, (1995) Reduction and Compensation of Thermal Errors in International Journal of Machine Tools and Manufacture 39:10881101.
Machine Tools. Annals of the CIRP 44(2):589598. [212] Zhao Y, et al, (2010) Optimization and Temperature Mapping of An Ultra-
[200] Weidlich D, Nestmann S (2001) Kompaktfuhrungen an Mineralgussgestellen. High Thermal Stability Environmental Enclosure. Precision Engineering
Werkstatt und Betrieb 78:120123. 34:164170.