1、14FTM04 AGMA Technical Paper Reliable Measurements of Large Gears By M. Stein, K. Kniel, and F. Hrtig, Physikalisch-Technische Bundesanstalt (PTB) 2 14FTM04 Reliable Measurements of Large Gears M. Stein, K. Kniel, and F. Hrtig, Physikalisch-Technische Bundesanstalt (PTB) The statements and opinions
2、contained herein are those of the author and should not be construed as an official action or opinion of the American Gear Manufacturers Association. Abstract Large gears have become an indispensable part of modern technical applications. The expanding industrial sectors of power generation and tran
3、smission, like shipbuilding industry, wind turbine generators and petroleum conveying systems, have led to an increasing demand for large-scale gear boxes. Thus, the qualified measurement of large gears has become more and more important as well. Their conformance with specifications according to IS
4、O 14253-1 1 has to be proved, which is not possible without a qualified statement of the task-specific measurement uncertainty. As a consequence, the manufacturing processes cannot be controlled quantitatively and at a reasonable process capability level, especially if tolerances are small compared
5、to the achievable measurement uncertainty. This specifically applies to large gears. In recent years, large gear metrology has gained in importance among the research activities at PTB. In this report, three current projects are presented. First of all the first national comparison for large involut
6、e gears is described. It has been organized by PTB and is just about to be finished. The measurement results are included in this paper. As a second step towards traceable measurements of large gears, a special calibration laboratory is intended to be established. This is part of a joint research pr
7、oject, in which a new measurement artifact of 2 m in diameter has been developed. Lastly, information about a Joint Research Project within the European Metrology Research Programme which will start in September 2014 is provided. Copyright 2014 American Gear Manufacturers Association 1001 N. Fairfax
8、 Street, Suite 500 Alexandria, Virginia 22314 October 2014 ISBN: 978-1-61481-096-4 3 14FTM04 Reliable Measurements of Large Gears M. Stein, K. Kniel, and F. Hrtig, Physikalisch-Technische Bundesanstalt (PTB) Introduction The Physikalisch-Technische Bundesanstalt (PTB) has developed new large involut
9、e gear measurement artifacts up to 2 m in diameter and corresponding calibration techniques. These can be used to quantify the quality of large gear manufacturing processes for the first time. This fulfills the essential demands from industry which can be traced back to the tremendous growth of wind
10、 energy systems during the last few years. The gear measurement artifacts represent different internal and external gears usually used in wind mill gear boxes. Their workpiece-like design allows convenient handling on conventional gear measurement instruments or coordinate measuring machines with or
11、 without rotary tables. The first large involute gear artifact representing a cut-out of an external gear of 1 m in diameter was developed in 2010. A national comparison with nine partners from industry will be finished soon. The presented results will give an impression of the reliability of large
12、gear measurements under consideration of different types of measurement machines and different measurement strategies. Within another joint research project a complete traceability chain for large gear metrology is to be established. New measurement artifacts embodying internal as well as external l
13、arge involute gears with different helix angles are manufactured. Moreover, a calibration laboratory for large gearings will be accredited. Finally, the European Metrology Research Programme (EMRP) recently decided to support a major project focusing on traceable measurements of drivetrain component
14、s. Six other European metrology institutes and designated institutes as well as partners from industry and universities will participate. For the first time metrology traceable to the SI will be provided for the measurement of highly accurate components of renewable energy systems such as wind energ
15、y systems or tidal power generators. In this paper the above-mentioned projects together with the current results are presented. On that basis the challenges of measuring large gears, the consequences for a qualified statement of the task specific measurement uncertainty and future activities in tha
16、t field are discussed. First intercomparison for large involute gears In 2010 the worlds first large measurement artifact for involute profile and helix measurements was developed at PTB, see Figure 1. It represents a cut-out gear segment with a diameter of 1 m and embodies three different helix ang
17、les and directions. Its parameters are listed in Table 1. Table 1. Gear parameters of large measurement artifact Parameter Value Gear type external (1 segment, 3 different gear specifications) Number of teeth 38/37/36 Normal module 20 mm Pressure angle 20 Face width 400 mm Helix angle/hand 0/spur; 1
18、0/right; 20/left Outside diameter 1000 mm Root diameter 905 mm Material CTE Steel/11.5 m/K/m Weight (including dismountable counter mass) 450 kg (700 kg) Reference bands diameter (form deviation) 200 mm (1 m) 4 14FTM04 Figure 1. Large involute gear artifact including holes for temperature sensors Be
19、sides the easier handling and shipping of the measurement artifact, the segmental construction allows its calibration on the established measuring devices at PTB. Nevertheless, it is still quite a challenge to cope with all the complexities involved. The measurement artifact has been calibrated on a
20、 3D-coordinate measuring machine (CMM) at PTB. It forms the first step on the way to traceable large gear metrology. If a rotary table is to be used for measuring the measurement artifact, special attention has to be paid to weight distribution along the table. In order to guarantee an equal distrib
21、ution of load, a counter mass is provided with the measurement artifact. Figure 2 shows the use of the counter mass. In this case it is necessary to adjust the height of the measurement artifact to ensure that also the lower reference band can be reached by the stylus. A positioning device helps to
22、manage a correct centering of the measurement artifact in this situation. Figure 2. Measurement artifact mounted on a rotary table with counter balance, base frame and positioning device 5 14FTM04 In 2012 a national comparison was initialized. Nine partners from industry, including research institut
23、es, testing and calibration laboratories, manufacturers of drivetrain components as well as producers of measuring machines participated. The test started in September 2012 and was due to finish in July 2014. Every partner had around six weeks to set up the artifact and undertake the measurements in
24、cluding thermalization times. Profile and helix measurements had to be done according to the measurement parameters given in Tables 2 and 3, respectively. For each of the six flanks, profile and helix deviations were to be provided. That gives 36 values for comparison per partner (Table 4). Apart fr
25、om outliers, the measurement results from all the partners, as far as evaluated yet, are in good compliance with the calibration values from PTB respecting the measurement uncertainty of the calibration. Especially among the profile deviations, there are only a few values that do not fit into the ra
26、nge given by the calibration uncertainties. Those are different for the various measurands as shown in Table 5. Table 2. Parameters for profile measurements Helix angle, 0 10 right 20 left Stylus ball diameter 8 mm 8 mm 8 mm Number of points per reference circuit 72 72 72 Point density per mm at pro
27、file 1 1 1 z-level b/2 b/2 b/2 Start evaluation LStart(length of roll) 100 mm 124 mm 100 mm End evaluation LEnd(length of roll) 220 mm 232 mm 220 mm Table 3. Parameters for helix measurements Helix angle, 0 10 right 20 left Stylus ball diameter 8 mm 8 mm 8 mm Number of points per reference circuit 7
28、2 72 72 Point density per mm at flank 1 1 1 Measurement diameter dM950 mm 950 mm 950 mm Start evaluation LStart20 mm 20 mm 20 mm End evaluation LEnd380 mm 380 mm 380 mm Table 4. Overview of all 36 measurands 0/spur Left flank profile Slope deviation Form deviation Total deviation fH, ff, Fhelix fH,
29、ff, FRight flank profile fH, ff, Fhelix fH, ff, F10/right Left flank profile fH, ff, Fhelix fH, ff, FRight flank profile fH, ff, Fhelix fH, ff, F20/left Left flank profile fH, ff, Fhelix fH, ff, FRight flank profile fH, ff, Fhelix fH, ff, FTable 5. Calibration uncertainties Measurand fHffFfHffFU (k
30、= 2), m 4.3 1 4.3 4.3 1 4.3 6 14FTM04 Figure 3. Deviations of the measurement values for profile measurands of all nine partners compared to calibration values with indication of the expanded measurement uncertainty of PTB Figure 4. Deviations of the measurement values for helix measurands of all ni
31、ne partners compared to calibration values with indication of the expanded measurement uncertainty of PTB 7 14FTM04 A traceability chain for large gear metrology Within a joint research project a complete traceability chain for large gear metrology is to be established. This project started in Septe
32、mber 2012 and will last for three years in total. The main objective is to install an accredited calibration laboratory for large gears, which is an important action towards traceable metrology (Figure 5). A brief survey among manufacturers of large gears gave an overview of the dimensions and param
33、eters of gears typically mounted in wind power plants or used in the ship building industry, for instance (Table 6). Based on the results of that inquiry, a new involute gear measurement artifact has been developed, which embodies different internal and external gears with different helix angles. On
34、e of the main demands made on the segmented measurement artifact described in Section 2 was due to the measuring facilities available at PTB. Therefore, its maximum length in the cross-section plane was limited to 600 mm. The new measurement artifact was designed without this restriction. Using a ta
35、ctile-optical measuring system, called M3D3 (mobile multi-lateration measurement system for 3D measurement application, Figure 6), which was recently developed (at PTB), it is not necessary for the CMM to be traced back to the SI. Hence, it is possible to perform the calibration outside PTB. Figure
36、5. Pyramid of calibration Table 6. Parameters of the new involute gear measurement artifact Parameter Value Gear type external and internal Number of teeth external 108/106/102 internal 95/94/90 Normal module 18 mm Pressure angle 20 Face width 420 mm Helix angle/hand external: 0/spur; 10/right; 20/l
37、eft internal: 0/spur; 10/left; 20/right Outside diameter 1980 mm Inner diameter 1685 mm Weight (estimated) 2700 kg Reference bands diameter 1870 mm 8 14FTM04 Figure 6. 3D model of the new involute gear measurement artifact The M3D3 system uses one CMM as a mover and (at least) four LaserTracers, a s
38、pecial kind of highly accurate laser tracker that is capable of automatically tracking a moving retro-reflector and measuring distances continuously 3. As shown in Figure 7, the measurements are performed in two steps. In a first step the workpiece is measured in a tactile manner on the CMM. Then th
39、e workpiece is removed from the measurement volume to make the stylus, which is replaced by a retro-reflector, visible to all four LaserTracers. In the second measurement step the CMM replays all the probing points from the first measurement step exactly only as a mover. The positions are measured b
40、y the LaserTracers based on the multi-lateration principle. At each measurement point the local error vector is evaluated as the difference between the indicated mover position by the CMM and the position measured with the LaserTracers. This is used to correct the original measurement data from the
41、first step point by point. The calibrated measurement artifact will provide the basis for the accreditation of the first calibration laboratory for large gears. The weight of the new measurement artifact is expected to be around 2.7 t. This will cause many challenges concerning its handling, particu
42、larly regarding logistics (transport, mounting, overturning). However, also from a metrological point of view special attention is to be paid to huge masses as they are much more sensitive to temperature influences resulting in fluctuating volumes. Therefore, 12 sensors equally distributed around th
43、e ring are to give information about the temperature inside the measurement artifact before and during the measuring process. Figure 7. M3D3 measurement (two steps) 9 14FTM04 For overturning the measurement artifact, two different options are discussed. Positive experience has been gained with a met
44、hod using a crane together with a tackle. This system was applied to the measurement artifact described in Section 2 (Figure 1) and worked very well. Since the new artifact is much heavier and larger, it is under consideration to use a traverse and a cross axis to guarantee an optimal weight distrib
45、ution during the process of overturning. Traceable measurements of drivetrain components In September 2014 a Joint Research Project within the European Metrology Research Programme will start under the coordination of PTB 6. This project focuses on traceable 1D-3D measurements on highly accurate com
46、ponents of large drivetrains for wind energy systems. These are shafts up to 3 m in length and 1 m in diameter, large bearings up to 3 m, internal and external gears up to 3 m, and brakes up to 1 m (Figure 8). In addition to classical gear measurement categories, solutions for measuring and characte
47、rizing waviness and surface parameters for large gears shall be provided. Especially the benefits from using microprobes on a gear measuring instrument to extend machine versatility and detect key performance characteristics on workpieces up to 3 m in diameter will be evaluated. New measurement arti
48、facts and calibration procedures will be developed for establishing traceability and estimating measurement uncertainty for drivetrain component manufacturers. The artifacts will be suitable for use in industrial settings. Another key aspect will be to establish and quantify the main additional sour
49、ces of uncertainty that influence industrial measurement capability with particular reference to environmental effects. This will result in good practice guides for industry. In addition to the Virtual Coordinate Measurement Machine (VCMM) developed at PTB 4, a Virtual Workpiece Module will be developed. This numerical model will be used to simulate the significant error influences such as surface characteristics, material influences, weight, and temperature variations. The module will allow the determina
