Metrological performance of igps in measurement assisted assembly processes

Size: px

Start display at page:

Download "Metrological performance of igps in measurement assisted assembly processes"

Felicity Williamson
6 years ago
Views:

1 Metrological performance of igps in measurement assisted assembly processes Gustavo Heiden 1, Maurício de Campos Porath 1 1 Universidade Federal de Santa Catarina Campus Joinville mauricio.porath@ufsc.br Abstract: This paper presents a performance evaluation of igps in assembly processes that include alignment and positioning requirements. An experimental device was designed to simulate an assembly process with positioning and alignment in three degrees of freedom. The assembly of the device was performed with the assistance of the igps and the result evaluated with a coordinate measuring machine. The results confirm the applicability of the igps in assembly processes with positioning tolerances larger than 1 mm. Keywords: assembly, positioning, alignment, igps, industrial geodesy, large volume metrology Resumo: Este trabalho apresenta a avaliação do desempenho do sistema igps em processos de montagem, utilizando um dispositivo que simula uma tarefa de montagem com operação de posicionamento e alinhamento em três graus de liberdade. A montagem do dispositivo foi realizada com apoio do sistema igps e o resultado avaliado utilizando-se uma máquina de medir por coordenadas. Os resultados em laboratório confirmam a aplicabilidade do sistema no apoio ao processo montagem de estruturas com requisito de posicionamento que não sejam menores que 1 mm. Palavras-chave: montagem, posicionamento, alinhamento, igps, geodésia industrial, metrologia de grande escala 1. INTRODUCTION The assembly of large structures is one of the key challenges in sectors such as the shipbuilding industry, the aerospace industry and the power generation equipment industry. In shipbuilding, for instance, one of the most important stages is named erection in which the previously constructed blocks are welded together to form the hull. These blocks, weighting hundreds of tons, are commonly aligned and positioned by 8 o Congresso Brasileiro de Metrologia, Bento Gonçalves/RS, 2015 special vehicles, cranes or hydraulic jacks. Theodolites, laser levels, plumb-bobs and measuring tapes are used as measurement systems. The assembly is performed in an iterative way, including alignment, positioning and measuring processes. A paradigm shift in large-scale assembly is expected through the development of measurement assisted assembly processes using advanced 3D measurement systems. One of the 1

most promising measurement systems is the Indoor Global Positioning System (igps) that allows six degrees of freedom dynamic tracking in large-scale assembly processes. Mosqueira et al.

2 most promising measurement systems is the Indoor Global Positioning System (igps) that allows six degrees of freedom dynamic tracking in large-scale assembly processes. Mosqueira et al. evaluated this system as feedback for the robotic alignment of fuselages, using laser radar as reference measurement equipment [2]. Muelener et al. [3] and Schmitt et al. [4] assessed the overall performance of igps in 3D position measurement. They also identified the main sources of uncertainty of measurement processes using this system. However, we could not locate research results on the evaluation of the metrological performance of igps for positioning and alignment tasks using an accurate calibrated measurement system. This paper presents the performance evaluation of the igps in assembly processes that include alignment and positioning steps in one degree of freedom for translation and two degrees of freedom for rotation. The experiment we performed consists in assembling a special device with assistance of igps and verification of the results with a coordinate measuring machine (CMM). LED ring that emits a strobe signal at the beginning of every second rotation of the rotor. Photosensitive sensors in the receivers detect these light pulses and convert them to an electrical signal. This signal is amplified, digitized and processed in the PCE units that send the position data to a measurement workstation. The speed of the transmitter s rotors lies between 40 and 50 Hz, but each one operates with a slightly different frequency, enabling signal isolation [5]. Azimuth and elevation angles of a transmitterreceiver vector are determined by evaluating the detection times of the light beam signals [5,6]. A transmitter with +30 and -30 vertically inclined fan-shaped beams is depicted in figure 1. The detection times (t 1, t 2, t 3 ) of the three signals is illustrated in the graphic in the lower part of the same figure. 2. MATERIALS AND METHODS 2.1. Indoor Global Positioning System (igps) The optical measuring system igps, also known as rotary-laser automatic theodolites (R-LAT) [3], operates according to the triangulation principle. On triangulation-based systems the position of a target point is determined based on azimuth and elevation angle measurements from at least two stationary measurement systems with known position and orientation. The igps consists of three basic components: transmitters, which act as measurement stations, receivers and position calculation engines (PCE). The transmitters are equipped with a rotary head that sweeps two fan-shaped laser beams and a Figure 1. Azimuth and elevation angle determination of a transmitter-receiver vector Azimuth is calculated from the strobe time t 1 and the average of t 2 and t 3 and elevation is derived from the difference between t 2 and t 3 [6]. 2

3 The receivers are integrated in measuring tools. The standard measuring tools are the scalebar, which is used in the system-setup, the minivector bar, used as static reference points (monuments) as well as in dynamic tracking tasks and the handheld vector bar probe (HHVB) which is the device used to capture single points on the measuring object. Each one of these tools is equipped with two receivers, enabling the determination location and orientation in five degrees of freedom. A six degrees of freedom tracking frame is created by combining at least two mini vector bars that are fastened to the object to be dynamically tracked. We used this kind of frame to track the alignment and positioning movements of the assembly process. The typical static length measurement uncertainty of the igps under good environmental conditions is 0.5 mm [5]. Dephental et al. found that the highest accuracy of an igps system with four transmitters can be obtained, if these are installed on the corners of a rectangle and if the measurements are taken next to the centre of this rectangle [5]. The configuration we used in our experiments is depicted in figure 2. We performed the assembly on a workbench of 1 m height located approximately at the centre of the rectangle formed by four transmitters. Figure 1. Transmitter configuration The locations of the transmitters are listed in table 1. Coordinate z refers to the height of the transmitter in relation to the ground. The origin of coordinates x and y is located at transmitter T 4. Table 1. Transmitter location at working volume x [m] y [m] z [m] T1 0,03 9,88 1,84 T2 7,83 9,88 1,83 T3 7,92 0,00 1,86 T4 0,00 0,00 1, Experimental device We developed a simple experimental device to simulate an assembly process with two degrees of freedom of rotation and one degree of freedom of translation. This device consists of two parts that can be aligned and positioned with aid of igps and then rigidly fastened to a base plate. The size of the device was defined to fit the measurement volume of the available coordinate measuring machine. The L-shaped parts are made from aluminum profiles [7] and are fastened on a plate by three sets of screws, bolts and nuts (figure 3). This configuration allows three degree of freedom: 3

translation in z and rotation around x (Rx) and y (Ry). Figure 2. Assembly of the L -parts on the base plate Two integrated detector units (IDU) with mini vector bars are placed over each L.

HHVB or CMM probe) featuring a spherical tip of appropriate size by centering the sphere in the cone. We called one of the L part A and the other part B.

Experimental device The coordinate system was defined as follows: xy plane defined by A 1, A 2 and A 3 ; z axis defined by A 1 and A 2 and the origin of the system defined by A 1 (figure 5). Figure 4.

4 translation in z and rotation around x (Rx) and y (Ry). Figure 2. Assembly of the L -parts on the base plate Two integrated detector units (IDU) with mini vector bars are placed over each L. Three nests are also fixed over each part. A nest is a metallic cylindrical piece with an inner cone. The location of a nest can be determined with a probe (e.g. HHVB or CMM probe) featuring a spherical tip of appropriate size by centering the sphere in the cone. We called one of the L part A and the other part B. All nests of each part are numbered from 1 to 3 related to its L part (figure 4). Figure 3. Experimental device The coordinate system was defined as follows: xy plane defined by A 1, A 2 and A 3 ; z axis defined by A 1 and A 2 and the origin of the system defined by A 1 (figure 5). Figure 4. Coordinate system. 2.3 Experiment procedure For each L we created a frame to track the assembly process. We called them frame A and frame B. Frame A was kept stationary and fastened to the base plate. Frame B was moved until its nests where brought to the xy plane, i.e., until z coordinates of all nest where as close to zero as possible (we admitted a tolerance of ± 0.3 mm). After the alignment process, we also fastened part B to the base plate and evaluated 4

5 the result by measuring the position of all nests 10 times with the HHVB probe. After the assembly and measuring processes, we took the device to a metrology laboratory where the positions of the nests where measured using a CMM (figure 6). A third measurement cycle with igps was performed after the return of the device to our laboratory to discard misalignments during transport. Figure 5. CMM measurement 3. RESULTS AND DISCUSSION The results of the three measurements cycles are listed in table 2. Table 2. Results of the three measuring cycles IGPS results before CMM measurement [mm] A1 0,00 0,00-0,04 A2 489,75 0,00 0,00 A3 624,74 664,13-0,03 B1 433,60 818,60 0,23 B2-54,86 818,83 0,06 B3-189,19 153,86 0,19 CMM measurement results [mm] A1 0,00 0,00 0,00 A2 489,79 0,00 0,00 A3 624,83 664,17 0,00 B1 433,62 818,56-0,01 B2-54,73 818,80 0,41 B3-189,06 153,88 0,49 IGPS results after CMM measurement [mm] A1 0,01 0,00-0,01 A2 489,79-0,02-0,01 A3 624,53 664,21-0,02 B1 433,28 818,53-0,10 B2-55,02 818,72 0,27 B3-189,11 153,84 0,37 Table 3 lists the differences between igps and CMM results. Table 3. Measuring error of the igps Difference between igps (before MMC measurements) and MMC [mm] A1 0,000 0,004-0,038 A2-0,045-0,002 0,000 A3-0,095-0,041-0,026 B1-0,017 0,038 0,248 B2-0,130 0,029-0,353 B3-0,128-0,016-0,295 Difference between igps (after MMC measurements) and MMC [mm] A1 0,007 0,005-0,010 A2 0,001-0,024-0,014 A3-0,303 0,042-0,019 B1-0,338-0,037-0,090 B2-0,284-0,075-0,142 B3-0,046-0,044-0,111 The repeatability (two standard deviations) of the igps measurement results and the expanded measurement uncertainty are presented in table 4. 5

6 Table 4. Repeatability of the igps measurement results and expanded measurement uncertainty of the CMM measurements [2] Mosqueira G, Apetz J, Santos K M, Villani E, Suterio R and Trabasso L G 2012 Analysis of the indoor GPS system as feedback for the robotic alignment of fuselages using laser radar measurements as comparison 2012 Robotics and Computer- igps repeatability (2s) before CMM measurement [mm] A1 0,008 0,034 0,043 A2 0,029 0,049 0,046 A3 0,161 0,134 0,045 B1 0,089 0,037 0,047 B2 0,073 0,074 0,053 B3 0,060 0,057 0,065 Expanded uncertainty of CMM results [mm] A1 0,002 0,002 0,002 A2 0,002 0,002 0,002 A3 0,003 0,003 0,002 B1 0,002 0,004 0,002 B2 0,002 0,004 0,002 B3 0,002 0,002 0,002 igps repeatability (2s) after CMM measurement [mm] A1 0,025 0,094 0,029 A2 0,115 0,089 0,025 A3 0,051 0,044 0,023 B1 0,078 0,078 0,078 B2 0,084 0,062 0,028 B3 0,021 0,089 0,015 We conclude from the z coordinates of the MMC results that there is a maximum misalignment of 0.5 mm between the nests of the two L. Although the device is much smaller than an industrial structure, we consider this result representative for the performance of the system in an assembly process. Due to the measurement principle of igps, size-proportional error sources are expected to be less significant than effects such as triangulation condition and transmitter visibility. The repeatability of the igps results ranges from 0.01 mm to 0.16 mm, which shows that stable measurement values are strongly dependent on the measurement conditions at a specific point. Although the length measuring performance evaluation is not in the main scope of this paper, it is possible to verify from table 3 that all the measured errors were smaller than 0.5mm. This result confirms the value specified by the manufacturer. 4. CONCLUSIONS In this paper we evaluated the performance of the igps system in assembly processes that include alignment and positioning requirements evaluated using a self-designed device. The alignment quality of the assembly was evaluated with a coordinate measuring machine. The presented results show that it is possible to introduce igps as support to assembly processes. However, it is important to note that due to metrological limitations of igps, the positioning error tolerance shouldn t be less than one millimetre. A next project task is set to evaluate igps system robustness in shipbuilding conditions. The practicability of outdoor measurements and the sensitivity to optical and electromagnetic interferences will be studied. Simultaneously, we re aiming to integrate igps in a self-developed Stewart Platform for assembly purpose. 5. REFERENCES [1] Maropoulos P G, Muelaner J E, Summers M D and Martin O C 2014 A new paradigm in large scale assembly research priorities in measurement assisted assembly Int. J. Adv. Manuf. Technol DOI: /s

7 Integrated Manufacturing, 28 DOI: /j.rcim [3] Muelaner J E, Wang Z, Martin O, Jamshidi J and Maropoulos P G 2010 Estimation of uncertainty in three-dimensional coordinate measurement by comparison with calibrated points Meas. Sci. Technol. 21 DOI: / /21/2/ [4] Schmitt R, Nisch S, Schönberg A, Demeester F and Renders S 2010 Performance evaluation of igps for industrial applications International Conference on Indoor Positioning and Indoor Navigation (IPIN), September, Zürich, Switzerland DOI: /IPIN [5] Depenthal C and Schwendemann J 2009 IGPS A New System for Static and Kinematic Measurements 9 th Conference on Optical 3D Measurement Techniques, Vienna v [6] Gao W, Kim S W, Bosse H, Haitjema H, Chen Y L, Lu X D, Knapp W, Weckenmann A, Estler W T and Kunzmann H 2015 Measurement technologies for precision positioning, CIRP Annals - Manufacturing Technology DOI: /j.cirp [7] FAMAK 2011 Catálogo de produtos linha 2011 FKLEAN [Internet]. [cited 2015 Aug 26]. Available from: ACKNOWLEDGMENTS The authors thank Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the financial support. 7

Horizontal scale calibration of theodolites and total station using a gauge index table

Journal of Physics: Conference Series PAPER OPEN ACCESS Horizontal scale calibration of theodolites and total station using a gauge index table To cite this article: L H B Vieira et al 015 J. Phys.: Conf.