Dissertations / Theses on the topic 'Gage repeatability and reproducibility'

Thesis: M. Eng. in Manufacturing, Massachusetts Institute of Technology, Department of Mechanical Engineering, 2015.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 80-81).
The purpose of this project is to set up a statistical process control (SPC) system in a high volume machining center to reduce the scrap rate and improve the manufacturing quality. The system is demonstrated on a machining center at Waters Corporation as part of a team internship project. This thesis focuses on the gage repeatability and reproducibility study (Gage R&R study) for the implementation of the SPC system. Based on the knowledge about the machining processes and the gages available, we select the proper gages for different dimensions to conduct the Gage R&R study. Gage capabilities are analyzed and root-cause analysis for incapable gages is performed. Related reaction plans are developed and implemented in order to improve the gage capabilities. Discussion about tolerance redesign leads to the adjustment of specifications in the manufacturing area. As a result of these efforts, we find that the existing measurement system is capable for the SPC real time inspection system. As for the final result for this entire project, we demonstrated that with the SPC system, we successfully reduce the scrap rate by half and thus offer substantial cost savings as well as improved product quality.
by Shaozheng Zhang.
M. Eng. in Manufacturing

This work deals with measurement system analysis in the production of rotary tools. The aim is to carry out literature search on the topic and then do analysis Gage R&R of the measurement system in the company ANAJ Czech s.r.o. and on the basis of the obtained data to recommend appropriate corrective arrangements to improve the current measurement system.

This Master’s thesis deals with the evaluation of quality pneumatic linear drives. The theoretical part describes the problems in the field of pneumatic cylinders and statistical methods to perform the analysis. The practical part includes the execution and evaluation of the statistical analysis of the measurement process and pneumatic cylinders, also brings recommendations to optimize the measurement process. The methods used included analysis Gage R & R, Analysis control charts and data analysis using the process capability. The study was made for the purpose of improving test equipment Norgren Ltd. Used computer programs was Minitab Statistical Software and MS Excel.

Aim of the thesis is suggestion of Measurement system analysis for a company in the automotive industry. There were made an analysis of requirement of ČSN EN ISO 9001 and ČSN P ISO/TS 16949. Next there are procedures for various measurement system analysis. There are camparison between company and standart’s requirements. Then there are choosen suitable method for measurement system analysis.

Gage Repeatability & Reproducibility (GRR) Studies have become critical in process improvement projects in the manufacturing sectors. There are various methods to conduct GRR study. However, the most widely used is the Automotive Industry Action Group (AIAG) method, which was standardized after the recognition of the importance of measurement systems. In this study, AIAG method and Wheeler’s method are compared, with specific interest in the proportions of the estimates of variation. An experimental study was designed, with factors being the operators and parts. The spectrum analyzer – Quattro, was tested for its adequacy of measurement and to understand the variability in the measurement system. In this research, vibration-impact testing was performed on Stereolithography (SL) parts and the measured feature was the natural frequency. The data was analyzed following the AIAG method and that proposed by Wheeler. From the results obtained, the Repeatability and Reproducibility were over estimated by AIAG method in comparison to Wheeler’s. The Wheeler’s method gave a better understanding of the sources of variation. Due to which, the measurement system capability could be judged without bias. Also, Wheeler’s method helps in making right decisions about the measurement system. Therefore, Wheeler’s method is strongly recommended over the AIAG.
Thesis (M.S.)--Wichita State University, College of Engineering, Dept. of Industrial and Manufacturing Engineering.

Presentation given on October 27, 2016 at Data Reproducibility: Integrity and Transparency program as part of Open Access Week 2016.
We describe a study into the extent to which Computer Systems researchers share their code and data. Starting with 601 papers from ACM conferences and journals, we examine the papers whose results were backed by code to see for what fraction of these we would be able to obtain and build the code. Based on the results of this study, we propose a novel sharing specification scheme that requires researchers to specify the level of sharing that reviewers and readers can assume from a paper.

Study objectives were to determine the effectiveness, reproducibility, and repeatability of dental evaluation for estimating cattle age. Cattle (n = 400) aged 1.5 to 20 yr were evaluated by 3 technicians for estimated age via dentition. A subset (n = 383) was aged again 4 wk later by the same observers. Age estimations were most accurate in YOUNG (≤ 5 yr old) cattle with at least 95.7% rate of accuracy within 2 yr across all observers and observation events. For MIDDLE (6 to 10 yr old) and OLD (> 10 yr old) cattle, these accuracies were 81.5 and 62.1%, respectively. Reproducibility proved high, with all observers consistently assigning age estimates within 1 yr of one another for more than 9 out of 10 YOUNG animals; all observers agreed on at least every 8 out of 10 estimations for all age groups within 3 yr. Repeatability was less consistent.

The purpose of this thesis was to assess the repeatability and reproducibility of two high-speed macrotexture measuring systems. The first portion of the study collected macrotexture measurements using the two high-speed systems on the Virginia Smart Road facility and validated the reproducibility of the mean profile depth (MPD) measurements with reference CT Meter measurements. The various data sets were then compared with each other. The objective was to determine whether the two systems are collecting repeatable and reproducible data. The analysis showed that the two high-speed systems investigated have good repeatability (0.105 mm for the Ames and 0.113 mm for the SCRIM) when measuring the average MPD of the sections investigated. The two systems produce measurements that are highly-correlated (Ames R2 = 0.9591 and SCRIM R2 = 0.9157) with the reference ones obtained with the CT Meter. While the Ames systems, with the data processed using the Virginia Tech filter, measures MPD values that are very close to those of the CT Meter, with a virtually zero systematic bias. The SCRIM obtains slightly lower readings. The differences are thought to be due to the filtering of the raw pavement elevation measurements used by the SCRIM processing software to eliminate dropout and spikes in the laser measurements.
Master of Science

It is the responsibility of pavement engineers to design pavements that provide safe and smooth riding surfaces over their entire life cycle. Each year many people around the world lose their lives in vehicle crashes, which are one of the leading causes of death in the United States (US). One of the contributing factors in many of these crashes is inappropriate friction between tires and the pavement. To minimize the impact of this factor, state Departments of Transportation (DOTs) must monitor the friction of their pavement networks systematically and regularly. Several devices are used around the world for measuring friction. Locked-wheel skid trailers are the predominant technology for roadways in the U.S. However, Continues Friction Measuring Equipment (CFME) is emerging as a practical alternative, especially for network-level monitoring. This type of technology has been used for monitoring runway friction for many years and is starting to be used also for measuring roadway friction. This thesis evaluates the different operational characteristics of CFME to provide guidelines for highway agencies interested in using this technology for supporting their friction management programs. It follows a manuscript format and is composed of two papers. The first part of the thesis presents a methodology to objectively synchronize and compare CFME measurements using cross-correlation. This methodology allows for comparing the “shape” of the friction profiles, instead of only the average friction values. The methodology is used for synchronizing friction measurements and assessing the repeatability and reproducibility of the CFME using friction measurements taken on a wide range of surfaces at the Virginia Smart Road. The proposed approach provides highway agencies with a rigorous method to process CFME measurements. The second part of the thesis evaluates the impact of several operational characteristics on the CFME measurements using a field experiment. The results of the experiment confirmed that the measurements are significantly affected by (1) the direction of testing while testing on sections of road with a significant grade, (2) water film thickness, and (3) testing speed. The experiment showed that measurements taken downhill on a 6% grade were significantly higher than those taken uphill. The analysis also verified that, consistent with previous studies, the measured friction decreases with higher water depth and testing speeds. It also showed that the change of friction with speed is approximately linear over the range of speeds used in the experiment. In general, the thesis results suggest that CFME can provide repeatable and reproducible friction profiles that can be used to support friction management programs and other asset management business functions. However, care should be taken with regard to the operational conditions during testing since the measurements are affected by several factors. Further research is needed to (1) quantify the effect of these, and potentially other, operational factors; and (2) establish standard testing condition and approaches for correcting measurements taken under other conditions.
Master of Science

Within the Master´s Thesis „Repeatability and Reproducibility Improvement of Suspension Components Tests“ is resolution of a problem measurement of the radial elasticity of ball joint with regard for Repeatability and Reproducibility measurement.

This work is focused on extruded fiber diameter measurement problem. For this purpose a procedure has been proposed. This procedure makes use of numerical methods for image processing, which are described in theoretical part of work. The proposed procedure has been processed into single-purpose software and in the final part is assessed its repeatability and reproducibility.

The object of this work is design braking device with high repeatability and reproducibility that is used to analyse the measurement system. I developed this work in cooperation with TRW-DAS Dačice a.s. At the beginning, I made a search of the chassis system, testing of ball joints and steering. Furthermore, I made search of the MSA and the types of braking devices. The rest of the work is focused on the design and the thermal analysis of braking device. Then there are the actual test.

The aim of this thesis is to twofold, in order to demonstrate the variabilities that can be faced within the compressive strength of blended cements, one blended cement namely CEM IV / B (P-V) 32.5N is selected and the 28-day compressive strength is obtained by 16 different laboratories following TS EN 196-1 standard. Later, to show the variabilities that could be faced by different standards, three different cement types were selected and their compressive strengths are determined following two procedures first with TS EN 196-1, later with similar procedure described in ASTM. The strength of cement is determined by TS EN 196-1 in Turkey that is the same for all types of cements. However, American cement producers use different standards for testing the strength of Portland cement and blended cements. The main difference is the amount of water utilized in producing the cement mortar. It was observed that for Portland and Portland composite cements
there is not any significant difference in between the compressive strength results of cement mortars prepared by both methods. However, for pozzolanic cements, there is much deviance in the compressive strength results of cement mortars prepared by TS EN 196-1.

Les animaux de compagnie présentent un intérêt majeur dans les études cardiovasculaires soit comme modèles expérimentales soit comme modèles spontanées. De plus, le développement de méthodes d'exploration non invasives en médecine vétérinaire est devenu une nécessité à cause de régulations strictes concernant le bien être de l'animal. Un facteur limitant demeure l'exactitude et la fiabilité (répétabilité/reproductibilité) des techniques d'exploration pour leur utilisation répétitive chez des animaux vigiles. La mesure de la pression artérielle systémique (PA) non invasive et l'échocardiographie conventionnelle et avancée ont gagné de la place dans le domaine de recherche ; cependant, il n'existe pas de recommandations spécifiques sur la validation métrologique de méthodes quantitatives utilisées en recherche clinique ou fondamentale en médecine vétérinaire. Le présent travail a premièrement focalisé sur la validation métrologique de la mesure de la PA chez le chien et le chat en étudiant la répétabilité et la reproductibilité de la méthode Doppler (chez les deux espèces) et de la méthode oscillométrique (chez le chien) et en déterminant l'influence du niveau d'expérience de l'opérateur sur la variabilité intra- et inter-jour et le pourcentage de mesures réussies. Puis l'effet de la position de l'animal et du nombre de mesures répétées sur de variables sélectionnées d'échocardiographie bidimensionnelle (2D) et temps-mouvement (TM) a été évalué. De plus, la méthode de l'aire de la surface de l'isovélocité proximale (PISA) pour la quantification de la régurgitation mitrale (RM) ainsi que le strain (St) et strain rate (SR) dérivé du Doppler tissulaire ont été validé chez le chien par le calcul de la répétabilité et la reproductibilité des techniques. Dans un deuxième temps les techniques précédemment validées ont été appliquées chez des chiens sains, des chiens souffrant de maladie valvulaire mitrale dégénérative (MVD) spontanée et dans un modèle expérimental de myopathie centronucléaire (CNM) chez des Labradors retrievers. Toutes les études ont été réalisées à l'Unité de Cardiologie d'Alfort, à l'Ecole Nationale Vétérinaire d'Alfort. Pour les mesures de la PA, toutes les deux études ont montré que l'entrainement est important, notamment pour la mesure de la pression diastolique avec le Doppler, tandis que la répétabilité et reproductibilité sont bonnes pour un opérateur expérimenté. Concernant l'échocardiographie 2D et TM, la technique peut être réalisée chez des chiens couchés ou debout avec une répétabilité/reproductibilité similaire. La fraction régurgitée évaluée par la méthode PISA est une variable répétable et reproductible pour l'évaluation de la RM chez le chien souffrant de MVD. De façon similaire le St et SR peuvent être mesurés de manière fiable chez le chien vigile et leur utilisation combinée avec d'autres paramètres échocardiographiques offre une nouvelle approche à l'étude de la fonction systolique dans cette espèce et peut complémenter l'échocardiographie conventionnelle pour la mesure de la performance myocardique. L'intérêt de ces méthodes validées a été illustré dans des études avec des chiens souffrant de MVD de gravité variable et une différence significative des paramètres étudiés (PA et fraction régurgitée) a été retrouvée dépendant du stade de la maladie. Enfin dans le modèle de CNM, des altérations sous-cliniques ont été détectées et considérées comme significatives étant donnée les faible coefficients de variation de techniques utilisées. En conclusion, le présent travail a évalué la répétabilité et la reproductibilité de quelques techniques d'exploration cardiovasculaires et a démontré l'importance de la validation en les appliquant dans des circonstances cliniques. Cependant ces résultats ne sont valables que dans les conditions spécifiques utilisées dans ces études
Small animais present a major interest in cardiovascular studies either as experimental or as spontaneous models. Additionally, the development of non-invasive exploration methods in veterinary medicine has been impeding since regulations applying to animais are getting stricter and animal welfare has been the subject of many debates recently. A limitation factor regarding noninvasive cardiovascular exploration by different methods remains the accuracy, and reliability of the different techniques, that is to say, if the methods are repeatable and reproducible, and able to be used in repeated manner in awake animais. Although non-invasive arterial blood pressure (ABP) measurement and conventional, as weil as, advanced echocardiographic techniques have gained place in the domain of veterinary research, there are no specifie guidelines on how to metrologically validate a quantitative imaging technique in veterinary medicine for use in clinical or fundamental research. The present work focused firstly on metrological validation of non-invasive ABP measurement in dogs and cats by assessment of repeatability and reproducibility of Doppler ultrasonography (DU, in both dogs and cats) and high definition oscillometry (in dogs) and by determination of the influence of the observer's level of experience on within-day and between-day variability and the percentage of successful measurements. Then, the effect of animal position and number of repeated measurements on selected 2-dimensional (20) and M-mode echocardiographic variables in heal~hy dogs was evaluated. Moreover, the proximal isoveJocity surface area (PISA) method for quantification of mitral regurgitation (MR) and the tissue Doppler-derived strain (St) and strain rate (SR) technique in the dog were validated by calculation of repeatability and reproducibility. In a second step, the previously validated techniques were applied to healthy dogs, dogs suffering from spontaneous mitral valve disease (MVD) and to an experimental model of centronudear myopathy (CNM) in labrador retrievers. A total of 7 studies were realized. Ali studies were conducted at the Cardiology Unit of Alfort, in the National Veterinary School of Alfort. For ABP measurements, bath studies showed that training is important especially for evaluation of diastolic ABP with DU, whereas repeatability and reproducibility are good for sufficiently experienced observers. Regarding transthoracic 20 and M-mode echocardiography, itcan be performed on dogs in both standing position and lateral recumbency, and the repeatability of measurements is similar. Regurgitant fraction assessed by the PISA method is a repeatable and reproducible variable for the non-invasive evaluation of MR in awake dogs with MVD. This Doppler index is reliable, and its assessment can therefore be recommended for use, in combination with evaluation of other Doppler echocardiographic variables, in longitudinal studies of mitral valve disease in dogs. SimilarJy, myocardial St and SR can be measured non-invasively and in real time from regional myocardial velocities by TOI with good repeatability and reproducibility in the awake dog. The combined use of. St and SR indices with other tissue Doppler parameter's offers a new ultrasound approach to the systolic function in this species and may complement conventional echocardiographic measure of myocardial performance. The interest of those validation methods was firstly iIIustrated in studies involving dogs suffering from MVD of differing severity, where a relevant difference of the evaluated parameters (ABP and regurgitant fraction) was found depending on the stage of the disease

Theme of the Master’s thesis is analysis of process for production of small plates. Main parts of the thesis are analysis of the production process, process mapping, data capture, analysis of the data by Pareto analysis and find causes of nonconformities. Causes analysis is based on data analysis and process observation. It is divided to chapters by factors which are influencing the process. Optimization measures are designed with regard to their simplicity of realization and system in organization.

Esta dissertação tem como principal objetivo modelar e analisar dados de ensaios de comparação interlaboratorial realizados no Instituto Hidrográfico relativos a nutrientes em águas salinas. Neste contexto aplicam-se técnicas estatísticas usuais no tratamento de informação oriundas de procedimentos laboratoriais, tais como os testes de comparação interlaboratorial (definido como sendo um teste para avaliar a variabilidade entre laboratórios), Dixon (utilizado para detetar a presença de outliers) e Cochram (definido como sendo um teste para estudar a variabilidade interna de um laboratório). São aplicadas cartas de controlo a sistemas de medição, sendo dado um especial destaque à formalização de linhas gerais para uma melhor operacionalização das cartas já existentes e para a implementação de outras cartas de controlo, como as EWMA, procurando definir a carta mais apropriada para diferentes variáveis e contextos. Também se avalia um sistema de medição não replicável, isto é, um sistema cujas leituras não podem ser repetidas em cada amostra, pois as amostras analisadas sofrem alterações durante o ensaio ou são destruídas. Para tal realizam-se estudos de repetibilidade e reprodutibilidade, procurando também evidenciar a mais-valia da sua introdução tendo em vista os objetivos definidos pelo laboratório; Statistical quality control in laboratory tests: particular case of nutrients in saline waters ABSTRACT This work aims to model and analyze inter laboratory test data held in the Hydrographic Institute concerning nutrients in saline waters. In this context common statistic techniques are applied deriving information processing laboratory procedures, such as inter laboratory comparison tests (defined as a test to evaluate the variability among laboratories), Dixon (used to detect the presence of outliers) and Cochran (defined as a test to study the variability of an internal lab). The measurement systems control charts are applied, and a special emphasis is given on formalizing guidelines for a better implementation of existing cards and to implement other control charts such as EWMA, seeking to define the most appropriate letter for different variables and contexts. Also evaluating a non-replicable measuring system, i.e. a system whose readings cannot be repeated in each sample, because the analyzed samples changed during the assay or are destroyed. So repeatability and reproducibility studies are held also seeking to highlight the added value of its introduction in view of the objectives defined by the laboratory.

This work deals with questions about the measuring of the tantalum capacitor leakage current. This thesis answers to many questions of this sphere e.g. Study of basic parameters, analysis of the measurment LI Head, creating software for the connection between the measuring LI Head, the computer and the analog measuring card, activation of the measuring set, verification of set functionality and economic evaluation of the whole project.

Os ensaios laboratoriais para avaliação técnica ou estudo de material de irrigação envolvem a medição de diversas grandezas, bem como o monitoramento e controle das condições de ensaio. Qualquer sistema de medição e controle naturalmente apresenta instabilidades que podem afetar a qualidade dos resultados de ensaio, implicando em aumento da incerteza no processo de medição. A implementação do Controle Estatístico de Qualidade pode ser promissora para detecção de causas externas de variação, avaliação da variação máxima permitida aos dados de um processo e discriminação das principais componentes de variabilidade de um sistema de medição. Nesse sentido, considerando a hipótese de possibilidade de aprimoramento da qualidade dos resultados de ensaios, a presente pesquisa teve por objetivo utilizar técnicas de controle estatístico de qualidade e análise de sistemas de medição para avaliar o grau de adequação de bancadas de ensaio de emissores de microirrigação, aperfeiçoar os sistemas de medição e controle adotados nas bancadas e diagnosticar a porcentagem de contribuição dos fatores inerentes aos sistemas de medição e controle na variabilidade total das medidas de vazão em ensaios de uniformidade. A pesquisa foi conduzida no Laboratório de Ensaios de Material de Irrigação (LEMI/INCTEI/ ESALQ/USP). Foram elaborados gráficos de controle de Shewhart, média móvel exponencialmente ponderada (MMEP), além de índices de capacidade para o processo de medição da pressão de ensaio, monitoramento de temperatura da água e vazão, consideradas como as variáveis características da qualidade para ensaios de determinação das curvas de vazão em função da pressão de entrada e uniformidade de vazão de emissores de microaspersão e gotejamento. Além disso, realizou-se estudo de repetitividade e reprodutibilidade (Gage RR) do sistema de medição da vazão, para o qual foram utilizadas 10 repetições, sob três turnos de trabalho (manhã, tarde e noite), com 13 emissores na bancada de microaspersão e 25 emissores na bancada de gotejamento. Não houve influência do turno de trabalho nas medidas de vazão obtidas e o aprimoramento do processo de ajuste da pressão de ensaio pela implantação de um controlador proporcional integrativo-derivativo (PID) ao ensaio de microaspersores foi suficientemente detectado pela aplicação dos gráficos de controle. A pressão e temperatura da água se mantiveram estáveis durante cada ensaio, e não foram identificadas causas não aleatórias de variabilidade na rotina de ensaios. O processo de controle e medição da pressão, para as duas bancadas, foi classificado como\"excelente\" quanto ao desempenho e capacidade de se manter entre os limites de especificação e centralização em relação ao valor de referência (alvo desejado). Como esperado, a variabilidade entre emissores foi a componente de maior contribuição na variância total da vazão, correspondendo a 95,47% e 96,77% para microaspersão e gotejamento, respectivamente. A repetitividade e reprodutibilidade do sistema de medição de vazão (Gage RR) foi \"aceitável\" para as duas bancadas, com as respectivas contribuições de 4,53% e 3,23%, em relação à variância total, associados às incertezas do Gage RR.
Laboratory tests for technical evaluation or irrigation material testing involve the measurement of various greatnesses, as well as monitoring and control of test conditions. Any measurement and control system naturally presents instabilities that can affect the quality of the test results, resulting in increased measurement uncertainty. The implementation of the Statistical Quality Control may be promising for detecting causes of variation non-random, evaluation of the tolerance permitted to data of the process and breakdown of the main variability components of a measuring system. In this sense, considering the hypothesis of possibility of improving the quality of test results, the present study aimed at using statistical quality control techniques and measurement systems analysis to assess the reliability of test benches of microirrigation emitters, improve measurement and control systems adopted in the benches and to diagnose percentage of contribution of the factors inherent in the measurement and control systems in the total variability of flow measurements on uniformity tests. The research was carried in Laboratório de Ensaios de Material de Irrigação (LEMI/INCTEI/ ESALQ/USP). Shewhart control charts were developed, exponentially weighted moving average (EWMA), and capability index for the process of measuring the test pressure, temperature monitoring of water and flow, regarded as the quality variables for determination the flow curves as a function of inlet pressure and uniformity of flow for microsprinkler emitters and drip. In addition, it was made study of repeatability and reproducibility (Gage RR) of flow measurement system for which were used 10 repetitions in three work shifts (morning, afternoon and evening), with 13 emitters in the microsprinkler\'s bench and 25 emitters in the drip bench. There wasn\'t influence of the work shift in the obtained measures for flow and the improvement of the adjustment process of the test pressure for the implementation of the integrative-derivative proportional controller (PID) to microsprinklers test was sufficiently detected by control charts. The pressure and water of temperature remained stable during the tests and weren\'t identified non-random causes of variability in routine tests. The process control and measurement of pressure was classified \"excellent\" to performance and capability to remain in the range of specification and centralization in relation to the reference value (desired target), to the two benches. As expected, the variability between emitters was the greater contribution component in the total variance of flow, corresponding to 95.47% and 96.77% for microsprinkler and drip, respectively. The repeatability and reproducibility for flow measurement system (Gage RR) was \"acceptable\" for the two benches, with the contributions respective of 4.53% and 3.23% relative to the total variance, associated with uncertainties of Gage RR.

Despote the indisputable advantages of fine-pitch components, is need to calculate with a few trouble during production, especially increased requirements for accuracy of mounting and solder printing. In this work I’m concerned with problems of solder printing for these components and evaluation using SPC. For the evaluation I used 3D paste inspection based on laser scanning of the surface. The output of this work is to describe the principles of solder printing and elaborating of GR&R, SPC analysis and histograms of solder printing for some outputs. I focused in my master thesis on motive design change of problematic components and economic evaluation of the adjustments.

碩士
明志科技大學
工業管理研究所
97
In the past, most scholars emphasized the research of single quality characteristics in the analysis of Gage Repeatability and Reproducibility Study. However, in most practical measuring results, the multiple characteristics occur. If the quality characteristics of the interaction between each other, when conducting Gage R&R analyses, it might be possible to ignore the affected relationship among every quality feature and thus mistaken estimate the extent of varying measurement, which may result in the invalid face validity of the whole measuring capacity. This paper is mainly aims at measured object using the same measuring different position, and the formation of multi-variable measurement data, and for this multi-qualities data, put forward a new measurement system assessment model. Firstly, on tests for normality of the collected data from experiments will be conducted. Secondly, following that would be the application of the Principal Components Analysis (PCA) of multivariate statistics to transfer multiple characteristics into one or a few irrelevant traits of synthetic quality. Next use the Multivariate analysis of Variance to find the variables of covariance matrix and eigenvalue. Then, using relationship between the principal component and multivariate normal elliptical, to estimate multivariate extensions of gage approval criteria precision to tolerance (P/Tm) ratio, measurement variation to total variation of measurement system (%R&Rm) ratio, and measurement signal to noise (SNRm) ratio, will be employed as the pointer to indicate measuring capacity. In the method verification, use to the literature of the two examples and the actual measurement panel to explore the film thickness. In the Wetting balance case, applying multivariate extensions of gage approval criteria precision to tolerance ratio yielded P/Tm=0.2674 , measurement variation to total variation of measurement system ratio yielded %R&R=30.38%, measurement signal to noise ratio yielded SNRm=4.43 . This paper methods have greater measurement variability arising, due to the original case is using the geometric mean estimate of a measure index, which have become smaller variation. In the automotive metal door case, P/Tm=0.2274 is better than the original case for the analysis of variance, whereas %R&R=10.21 and SNRm=13.72 results showed that the original case studies better measurement capability. Finally, in the film thickness measurement case, the results found in this paper study method of three multivariate measurement indicators will be consistency; the P/Tm=0.0822, %R&Rm=27.03% and SNRm=5 have shown measurement capability can be accepted. Therefore, using the multivariate criteria the measurement system would actual appear worse than when using the single variable approach.

碩士
大葉大學
機械工程研究所碩士在職專班
94
Today, industrial parts need more accurate dimensions for manufacturing processes than those in the past. These industrial parts with high precision need more reliable and accurate measuring equipments to do the measurement tasks for quality control. Coordinate measuring machines have been predominatly used to generate measurement points for a surface. The measurement data are analyzed to yield geometric tolerance information for the surface features such as flatness. With the availability of tolerance information, it is necessary to check if the surface is within the specifield limit. In this paper, I compare the effectiveness of Hemmersley sequence sampling, Halton-Zarmba sequence sampling and the Aligned systematic sampling. The mathematical model is used for the measuring points of the workpiece. The experiment is used to find the optimal number of measuring points and an appropriate sampling method to measuring coplanarity with Gage Repeatability and Reproducibility analysis.

碩士
華梵大學
工業工程與經營資訊學系碩士班
97
A manufacturing company of quantity from measurement, if quality is close to customer's demand accords with the customer and market need. The good quality company determines to the market demand in the society. Gage Repeatability and Reproducibility (GR&R) is one of the measurements in the measurement system analysis (MSA) included in QS9000. Based on system variation to manufacture process deviation or part tolerance with percentage, which is P/T（Precision-to-Tolerance）value. If P/T value is less than 10%, the measurement system precision is acceptance. If P/T value falls between 10% to 30% , whether the measurement system precision is acceptance depends on company’s decision. If P/T value is greater than 30% , the measurement system precision can't be accepted. GR&R combine estimates method for the variations with repeatability and reproducibility. Applying DOE (Design of Experiment) to handle the gage repeatability and reproducibility estimation of a multivariate measurement data.We compared ANOVA (Analysis of Variance) method with LFM (Long Form Method) two software method and one real case study. Used the one real case for ten same sample and three measure workers and repetition three times for multivariate measure data to compare different.

碩士
國立臺灣科技大學
工業管理系
88
The developments of technology and upgrades in industry have a causative relation with whether the metrology is good or bad. To achieve the goal for quality improvements , an accurate measurement is needed. Otherwise, there will be little or even no success at all no matter how hard you work. The asking for quality assurance have become stricter and stricter in the international community. Therefore, how to assure a good measurement quality is an important task for those engaging in work related to quality. This research is to study gauge repeatability and reproducibility (GR&R) and to develop the applied structure of measurement assessment in the fixed model (fixed measurement personnel, random measurement sample). Based on the relevant theory and practical experience of GR&R. First, we complete the linear statistical model of GR&R ,and construct the GR&R assessment method, the assessment process diagram, and the assessment standards. We use ANOVA in the research and successfully gain the goal of assessing GR&R with the assistance of fast-speed computer software SAS and Excel. Finally, we carry on empirical study for measurement ability especially in the mixed model to illustrate the measurement assessing processes. We also obtain the showing of the whole measurement ability by using assessing standards. By means of the applications of empirical study, we can illustrate that the analysis process in this research does promote and maintain the measurement quality. It can also furnish the references to those who engage in quality work when doing relevant measurement work in applied measurement techniques.

碩士
國立臺灣科技大學
工業管理系
94
This study investigated the POBREP (Process-Oriented Basis Representations) method to handle the repeatability and reproducibility estimation of a multivariate measurement data. POBREP can provide the meaningful pattern for multivariate measure data that help us understanding the causality of measurement data abnormally. Furthermore, we compared POBREP method with PCA (Principal Components Analysis) method using two simulation examples and one real case study. The result shows that PCA can only offer whether the measurement system is acceptable or not, but it can not provide gauge is unable to know which reasons caused it. Fortunately, POBREP can point out the abnormality of pattern and the variation of manufacturing processes. Therefore, we can conclude that POBREP method with multivariate data is better than PCA in explaining manufacturing process and measurement variation.

碩士
國立成功大學
統計學系
102
A gauge repeatability and reproducibility (GRR) study for analyzing gauge variation needs to be conducted prior to the process capability analysis. In this research, we proposed a new precision-to-tolerance ratio (P/T) for multivariate measurement system by considering the correlation coefficients among tolerances. And the simulation results show that our revised P/T ratio outperforms the existing one in terms of robustness and accuracy. Moreover, the optimal allocation of several parameters such as the number of quality characteristics (v), sample size of parts (p), number of operators (o) and replicate measurements (r) is discussed using the confidence interval (C.I.) of the revised P/T. Finally, a standard operating procedure (S.O.P.) to perform GRR study for multivariate measurement systems is summarized based on the research results. Hopefully, it can be served as a useful reference for quality practitioners when conducting GRR study for a multivariate measurement system analysis (MSA) in industries.

碩士
國立臺北科技大學
工業工程與管理研究所
92
Which particularly the researches of the Gauge Repeatability and Reproducibility (GR&R) in Gauge Variation Study has been paid much attention to, only that most researches in the literature are restricted to the studies on Mono Characteristics. However, in most practical measuring results, the multiple characteristics occur. There exists relationship among these characteristics. As there is more than one quality feature, it is appropriate to regard these features as inter-dependent parts instead of independent parts. If they are considered independent parts, when conducting GR&R analyses, it might be possible to ignore the affected relationship among every quality feature and thus mistaken estimate the extent of varying measurement, which may result in the invalid face validity of the whole measuring capacity. The aim of this research attempts to analyze variables of measuring the multiple characteristics. Firstly, on tests for normality of the collected data from experiments will be conducted because when there are significant differences between data distribution and normal distribution, the results adopting statistical testing would be meaningless. Secondly, analyze the individual varying situation of quality features according to their quality features, Analysis of Variance (ANOVA). Then, the accurate Precision to Tolerance Ratio (P/T ratio) will be employed as the pointer to indicate measuring capacity so as to look deeper into measuring capacity of separate quality feature. Following that would be the application of the Principal Components Analysis (PCA) of multivariate statistics to transfer multiple characteristics into one or a few irrelevant traits of synthetic quality. Based on these new traits, next the Analysis of Variance is employed to discuss the reciprocal affected influences among multiple characteristics, the situation of inter-dependence among every quality trait, and to calculate the pointer such as P/T ratio and measurement variation to total variation of measurement system ratio of measuring capacity so as to judge the whole synthetic performances of capacity. At last, the methods of the Gauge Variation Study of multiple characteristics and ways of the estimation of measuring capacity will be proposed to provide both academic researches and the industries with more practical, feasible, and further research resources in the field of the Gauge Variation Study of multiple characteristics.

碩士
義守大學
電子工程學系碩士在職專班
100
Emerging high-speed bus technologies are presenting new test challenges for testing engineering device mass-production, debug and validation. These challenges include: Multi-gigabit operating speeds, tester correlation problem, and signal integrity of test fixture issues. In the past, ATE measurements were commonly accepted as correct. The pursuit of improved quality of high-speed testing has proven this to be a fallacy for high-speed tester. Because of the lack consistency in the more accuracy timing specifications of automatic test equipment (ATE), the usefulness of suppliers’ specification sheets are very limited for high-speed testing. A key aspect of mass-production device testing is the requirement for at-speed functional testing, requiring stability signal quality and high-bandwidth signal at multi-Gbps data rates. Further, these new communications-like buses require more tight of timing parameters. Conventional low-speed tests may no longer yield meaningful results or even be applicable for tester issues. In this report we present a yield prediction methodology for high-speed digital tester: starting from the analysis specifications of high-speed tester, high-speed measurements with Agilent DSO81204B Infiniium high performance oscilloscope and collection data by Laboratory Virtual Instrument Engineering Workbench (Labview) program, we analyze the timing and swing voltage results of test-pin with Gauge Repeatability and Reproducibility (GR&R) method by Microsoft Excel in order to predict reliability and performance. The proposed timing parametric yield analysis is enables analyzing the trade-offs that arise between the predicted quality of tester and the required overall test environment timing accuracy (OTETA).

碩士
南台科技大學
工業管理研究所
96
A reliable and accurate gauge system is closely linked with each people. Correct analysis of the measurement system can assure the exactness of the process capability calculation and it is related to the development of science and industry technology. Good measurement system is the key to stable process and thus makes the product conform to the quality specifications. The purpose of the research is to expand the ANOVA method for calculating the gauge repeatability and reproducibility (GR&R) since it only considers two factors, part and operator, in the models. In this research the third factor, gauge, is added to the model. Also various experimental designs are considered in the GR&R analysis to make them more applicable to industry. The research discusses crossed design and nested design for their application on the GR&R analysis. A full model is considered on the crossed design while the nested design discusses both the cases of interaction and without interaction among factors. Also each design can be fixed effects model, random effects model, or mixed effects model. By doing this, one can compare the differences of the GR&R analysis for various designs and models. The research shows that the GR&R variations are different using ANOVA method while applying it on different designs and effect models. This suggests one have to carefully select correct design and effect model according to practical situation, otherwise it may cause a totally wrong conclusion.

碩士
南台科技大學
工業管理研究所
95
Gauge study considers both repeatability and reproducibility of the measurement system. The assessment of GR&R provides objective policy-making data not only ensuring accuracy but also ensuring precision of the measurement system within the controllable range for monitoring the manufacturing process. Therefore, only qualified measurement technology can guaranty the manufacturing process and the product quality in the state of statistical control. Methods to estimate repeatability and reproducibility contain ANOVA method, Classical GR&R method, and Long Form method. In the past, the measurement system usually utilized the method of ANOVA to analyze the measurement system while there was interaction between operators and parts. Classical GR&R and Long Form methods can not be used to estimate the variation with interaction between operators and parts. Therefore, the research modified the two methods not only making them applicable in the case of interaction between operators and parts but also providing unbiased gauge variance estimators. This improved the shortcoming of gauge analysis using the concept of control chart. Meanwhile, the research compared the gauge variance estimators of different estimation methods by simulating a measurement system based on the criterions of the biasness, variance, and mean squares error of the gauge variance estimates. The research showed that the ANOVA method produces unbiased gauge variance estimator. While taking accuracy and precision into consideration simultaneously, the modified gauge variance estimator of the research is the best one. In addition, the research also discussed the parameter values of the measurement system which are sample size, number of operators, and repetition number per operator. It is suggested that one must consider increasing the number of operators first, sample size the second and repeated measurements the last.

碩士
南台科技大學
工業管理研究所
95
The development of technology and upgrade in industry has a causative relationship with metrology. To achieve the goal for quality improvements, an accurate measurement system is required. The gauge repeatability and reproducibility (GR&R) in the measurement system analysis (MSA) applies experimental design to analyze the variation due to operators and the variation due to measurement equipments. The purpose of the research is to improve the measurement evaluation system nowadays, modify the flaw of Classical GR&R and Long Form methods to develop unbiased gauge variance estimators for the case of without interaction between operators and parts. Finally we compare the accuracy and precision of three methods and the unbiased estimators method proposed in the research. Thus, we provide one gauge variance estimation method which is more accurate and precise. The modified method will have enormous impact on the current MSA handbook and the software applying Classical GR&R as a method to design it. I addition, the research uses simulation method to compare the gauge variance estimators of four different estimation methods based on the three criterion of mean ratio, sample variance, mean squares error of the estimated gauge variances. We discovered the gauge variance estimators of the modified method are close to those of ANOVA. The research also confirms the lager value of npk, the more accurate and precise the gauge variance estimates are. The research also validates that increasing the number of operators is the best policy to enhance the gauge accuracy and precision.

碩士
大葉大學
工學院碩士在職專班
100
ISO9001:2000 and TS 16949 have become the major quality system management models in presently traditional industries and Hi-tech industries. Measurement System Analysis (MSA) manual, on the other hand, is one of the core tools in ISO/ TS 16949. MSA aims to evaluate Gauge Repeatability and Reproducibility (GR&R) where control, monitoring, and the maintenance of measuring process are required in measuring systems so that the measuring capability could be ensured under statistical control. An ideal measuring system should present the statistical characteristic of zero error on any measured products. Nevertheless, such an ideal measuring system hardly exists. Managers therefore have to adopt such measuring systems with unsatisfactory statistical characteristics. When proceeding GR&R variability, personnel in quality-related industries have to follow the standards formulated in MSA manual in AIAG to determine the appropriateness of the measuring system. People are likely to ignore the computing process as it is simple; the determination of measuring systems do not merely depend on some simple indexes. Traditional MSA indexes are constructed with definitely observed values. However, measurements with observed values are not entirely error-free. For this reason, this study proposes the research and evidence on the three cases in a case company and applies the integration of Fuzzy Theory and Gauge Repeatability and Reproducibility to discuss the differences in the evaluation index %GR&R and Number of Distinct Categories (NDC). Having fuzzy numbers to substitute for definite numbers, it is found that the data of %GR&R are increased and NDC is decreased after the fuzzification. Such results verify that the fuzzified %GR&R and NDC become stricter on the determination criterion. The research outcomes could assist the case company in improving the reference data of measuring systems as well as promoting the measuring quality. Key words: Measuring system, Fuzzy Theory, Gauge Repeatability and Reproducibility

Aim: To evaluate repeatability and reproducibility of macular thickness measurements in visually normal eyes using the Topcon 3D OCT-1000. Methods: Phase 1 investigated scan repeatability, the effect of age and pupil dilation. Two groups (6 younger and 6 older participants) had one eye scanned 5 times pre and post- dilation by 1 operator. Phase 2 investigated between-operator, within and between-visit reproducibility. 10 participants had 1 un-dilated eye scanned 3 times on 2 separate visits by 2 operators. Results: Phase 1: No significant difference existed between repeat scans (p=0.75) and no significant difference was found pre- and post-dilation (p=0.54). In the younger group variation was low (95% limits ± 3.62 m) and comparable across all retinal regions. The older group demonstrated greater variation (95% limits ± 7.6 m). Phase 2: For a given retinal location, 95% confidence limits for within-operator, within-visit reproducibility was 5.16 m. This value increased to 5.56 m for the same operator over two visits and to 6.18 m for two operators over two visits. Conclusion: A high level repeatability, close to 6 m, of macular thickness measurement is possible using the 3D OCT- 1000. Measured differences in macular thickness between successive visits that exceed 6 m in pre-presbyopic individuals are therefore likely to reflect actual structural change. OCT measures are more variable in older individuals and it is advisable to take a series of scans so that outliers can be more easily identified.

A reliable and robust laboratory method is essential for the forensic analysis of deoxyribonucleic acid (DNA), particularly for low-template samples. Electropherogram peak heights are important to the identification of STR alleles, and these peak heights are prone to error. Since error can be introduced into the process during sample preparation, quantification, amplification, or analysis, validation studies are performed in an attempt to characterize the signal variation associated with the process. While current practices assess aspects of a method, such as sensitivity and reproducibility, the effects of daily laboratory alterations are often not considered. Additionally, samples used in a validation study may be prepared using serial dilutions. Therefore, understanding the extent to which error is propagated through the series and the effect it has on the results could help improve validation practices. This work aimed to assess the effect daily laboratory modifications have on the signal in a forensic electropherogram. Specifically, the variability in signal when different capillary and amplification kit lots were used was evaluated against the variability observed when a single sample was either injected or amplified multiple times. The variability was determined via the examination of peak heights, peak height ratios, stutter, and drop-out. The effect of serially diluting samples was examined via an in silico model of the dilution process, polymerase chain reaction (PCR), and capillary injection. The peak heights from simulated serially diluted samples using the concentration of a stock DNA were compared to the peak heights from simulated samples that were quantified after the dilution series was generated and prior to amplification. The different capillary lots and amplifications were found to result in greater variation compared to the multiple injections. Additionally, when the stutter percentages obtained from using multiple kit lots were compared to those obtained using the same kit lot, differences in stutter percentage deviations resulted in different stutter thresholds. Drop-out rates were also different between the samples amplified with one kit versus the same samples amplified with multiple kit lots. Therefore, at a minimum, multiple amplifications should be run on multiple capillary lots during validation. Further, if available, the use of multiple kit lots is recommended, particularly in cases where stutter thresholds or drop-out models are used during interpretation. Creating validation samples via serial dilutions was also found to increase the variation observed in peak height in the simulated samples, suggesting that samples should be quantified post-dilution. In addition to characterizing the variability of several components of DNA analysis, an alternative quantification method was investigated in order to decrease the overall variability associated with the quantification process. This work sought to develop an electrochemical biosensor using a single-stranded DNA (ssDNA) probe chemically adsorbed to a gold electrode. This would allow for the direct quantification of DNA and eliminate the need for qPCR and fluorescent-based oligonucleotide detection systems. The DNA probe was successfully adsorbed to the surface of the gold disk electrode, hybridized to a single-stranded complementary DNA sequence, and detected using square wave voltammetry. Additionally, the ability to control the amount of DNA chemisorbed to the electrode surface was investigated by varying the incubation time in the probe solution. The measured current increased as the incubation time increased from 15 minutes to one hour, after which it plateaued. The use of an electrochemical biosensor is a promising alternative to qPCR for the quantification of DNA, with one hour being the optimal incubation time in the probe solution.