Non-sample based parameters design for system performance reliability improvement

손영갑,김재정,신승중 대한기계학회 2009 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.23 No.10

Design methods for quality generally help to improve quality over time, but do not consider change of system performance over time, resulting from degradation in components. As design methods for quality over time (performance reliability), which minimizes effects of unavoidable component degradations as well as component variations on system performance change, system model-based sampling methods using Monte-Carlo simulations have been used. But, there are main concerns related to computational efficiency and optimization in applying the sampling methods. To overcome the concerns, we propose a non-sample method for quality over time. Based on the proposed method, the process of allocating design parameters, which could minimize the noise effects with the consequence that both quality and performance reliability are optimized, is discussed. Reliability metrics such as mean time to failure and standard deviation of time to failure are optimized simultaneously for reliability improvement. Desirability functions for the metrics are introduced to perform the simultaneous optimization. The proposed method is applied to an electrical system design and compared to a sampling based design method.

비복원 추출된 이차 형식을 가지는 가속 열화 데이터를 이용한 저장 신뢰도 추정 방법

손영갑 대한기계학회 2019 大韓機械學會論文集A Vol.43 No.12

This paper proposes a storage reliability estimation method using degradation data in quadratic form that were obtained via destructive, accelerated degradation tests. The degradation data of interest tend to increase and then decrease over time, even though most of degradation data used in storage reliability estimations show simple, monotonically increasing or decreasing trends over time. This trend is modeled as a quadratic function; how to apply accelerated life models to the function, and how to estimate storage reliability under certain storage conditions using the function are addressed in the paper. To verify the proposed method, stochastic processes were generated to simulate the non-repetitively sampled, accelerated degradation data in quadratic form. The proposed method, and a method in open literature were applied to estimate storage reliability using the degradation data. Comparison of reliability estimation results shows the proposed method is more accurate. 본 논문은 파괴적인 가속 열화 시험을 수행하여 확보한 이차 형식을 가지는 열화 데이터를 이용하여 저장 조건에서의 신뢰도를 추정하는 방법을 제시한다. 신뢰도 추정에 사용되어왔던 대부분의 열화 데이터는 시간에 따라 증가하거나 감소하는 단순한 경향을 나타내지만 본 연구의 대상인 열화 데이터는 시간이 증가함에 따라 증가한 후에 감소하는 경향을 나타내는 특성을 나타낸다. 이러한 열화 데이터 경향을 이차 함수를 이용하여 모델링하고 가속 수명 모델을 적용하는 방법과 저장 조건에서 저장 신뢰도를 추정하는 방법을 제시한다. 제안한 방법을 검증하기 위하여 통계적인 프로세스를 발생 시켜 이차 형식을 나타내는 비복원 추출된 가속 열화 데이터를 생성하였다. 생성한 가속 열화 데이터에 본 연구에서 제안한 방법과 문헌에 공개된 방법을 적용하여 저장 신뢰도를 추정하였으며 제안한 방법이 기존 방법보다 정확성이 매우 높은 결과를 제공한다.

가속수명시험을 이용한 블루투스 모듈의 수명 예측

손영갑,장석원,김재중,Son, Young-Kap,Chang, Seog-Weon,Kim, Jae-Jung 한국마이크로전자및패키징학회 2008 마이크로전자 및 패키징학회지 Vol.15 No.2

본 논문은 블루투스 모듈의 구조 및 솔더접합부에 대한 신뢰성 평가에 국한된 기존의 방법을 확장하여 정량적인 신뢰성 평가를 수행한 결과를 나타낸다. Field 환경조건에서의 정량적인 신뢰성 평가를 위해, 온도 싸이클링(temperature/thermal cycling)에서의 온도차를 가속스트레스로 선정하여 가속수명시험을 실시하였다. 가속수명시험을 통해 구한 고장시간 데이터들을 이용하여 수명분포 매개변수들을 추정하고 코핀-만슨(Coffin-Manson) 모델을 이용하였다. 가속수명시험을 수행한 결과, 블루투스 모듈의 고장모드는 Open, 고장메커니즘은 크랙(Crack)과 박리(Delamination)였다. Field에서 수거한 고장품들의 고장모드와 고장메커니즘이 가속수명시험을 통해 재현되었다. 또한 본 논문에서 블루투스 모듈이 Field에서 받을 수 있는 온도 싸이클링에서의 다양한 온도차에 대한 정량적인 수명을 예측할 수 있는 방법이 제시되었다. 제안된 방법을 이용하여 온도차 $70^{\circ}C$를 받고 있는 블루투스 모듈의 $B_{10}$ 수명은 약 4년으로 추정되었다. This paper shows quantitative reliability evaluations of a Bluetooth module through extending previous qualitative methods limited to structure reliability tests and solder joint reliability tests for Bluetooth modules. Accelerated Life Testing (ALT) of the modules using temperature difference in temperature cycling as an accelerated stress was conducted for quantitative reliability evaluation under field environment conditions. Lifetime distribution parameters were estimated using the failure times obtained through the ALT, and then Coffin-Manson model was implemented. Results of the ALT showed that the failure mode of the modules was open and the failure mechanisms are both crack and delamination. The ALT reproduced the failure mode and mechanisms of failed Bluetooth modules collected from the field. Further, a quantitative reliability evaluation method with respect to various temperature differences in temperature cycling was proposed in this paper. $B_{10}$ lifetime of the module for the temperature difference $70^{\circ}C$ using the proposed method would be estimated as about 4 years.

Storage reliability estimation of one-shot systems using accelerated destructive degradation data

손영갑,Taesoo Kwon 대한기계학회 2016 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.30 No.10

A gamma process model is widely used to estimate reliability of general systems using non-destructive degradation data. However, the daunting challenge is the application of the model to one-shot systems given that recursive measurements are unavailable and degradation trends over time are nonlinear. This study shows a storage reliability estimation method of one-shot systems using accelerated destructive degradation data when information of either chemical reaction or degradation trend is unknown. An n-th kinetic model, which is a Physics-based model, was used to evaluate the degradation phenomenon of one-shot systems using destructive degradation data. The accelerated degradation model was constructed to estimate storage reliability for normal storage temperature. The proposed method was applied to estimate storage reliability of IR flare using the stabilizer contents that were destructively measured over time at three temperature levels. The real application results show applicability of the proposed method.

Reliability prediction of engineering systems with competing failure modes due to component degradation

손영갑 대한기계학회 2011 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.25 No.7

Reliability of an engineering system depends on two reliability metrics: the mechanical reliability, considering component failures, that a functional system topology is maintained and the performance reliability of adequate system performance in each functional configuration. Component degradation explains not only the component aging processes leading to failure in function, but also system performance change over time. Multiple competing failure modes for systems with degrading components in terms of system functionality and system performance are considered in this paper with the assumption that system functionality is not independent of system performance. To reduce errors in system reliability prediction, this paper tries to extend system performance reliability prediction methods in open literature through combining system mechanical reliability from component reliabilities and system performance reliability. The extended reliability prediction method provides a useful way to compare designs as well as to determine effective maintenance policy for efficient reliability growth. Application of the method to an electro-mechanical system, as an illustrative example, is explained in detail,and the prediction results are discussed. Both mechanical reliability and performance reliability are compared to total system reliability in terms of reliability prediction errors.

다중비교를 이용한 샘플수와 샘플링 시점수의 원샷 시스템 신뢰도 추정방법 정확성에 대한 영향 분석

손영갑,Son, Young-Kap 한국군사과학기술학회 2012 한국군사과학기술학회지 Vol.15 No.4

This paper provides simulation-based results of effect analysis of sample size and sampling periods on accuracy of reliability estimation methods using multiple comparisons with analysis of variance. Sum of squared errors in estimated reliability measures were evaluated through applying seven estimation methods for one-shot systems to simulated quantal-response data. Analysis of variance was implemented to investigate change in these errors according to variations of sample size and sampling periods for each estimation method, and then the effect analysis on accuracy in reliability estimation was performed using multiple comparisons based on sample size and sampling periods. An efficient way to allocate both sample size and sampling periods for reliability estimation tests of one-shot systems is proposed in this paper from the effect analysis results.

VCR Drum 조립체의 고정나사 체결 Torque 결정에 관한 연구 ( A Study on the determination of the initial torque in tightening set screw of the drum assembly in VCR )

손영갑,김재중,백승준,장석원,유동수 대한기계학회 2000 대한기계학회 춘추학술대회 Vol.- No.-

로봇의 수명 추정을 위한 축 모터의 RMS 토크 데이터를 이용한 가우시안 과정 회귀 응용

손영갑 대한기계학회 2023 大韓機械學會論文集A Vol.47 No.12

This paper presents the results of the lifetime prediction of a six-axis robot using Gaussian process regression applied to the root mean square (RMS) torque data of each axis motor across its operating cycles. Three types of functions of the mean values were formulated and compared to construct an optimal Gaussian process regression model encapsulating the RMS torque values over these operating cycles. Contrary to monotonic increment, the RMS torque values throughout the operating cycles demonstrated a quadratic trend characterized by incremental escalation after a preceding decrement. The viability of employing Gaussian process regression for predicting the lifespan of robots was evaluated. This evaluation encompassed a comparison between the anticipated remaining useful life derived through Gaussian process regression and the prognosis rendered by the particle filter — an established technique within the realm of lifespan prediction.

로봇 시스템의 위치 반복정밀도에 대한 신뢰도 추정

손영갑,김봉석 대한기계학회 2023 大韓機械學會論文集A Vol.47 No.2

To evaluate the system position repeatability of a robot, repeatability measures at various positions have not been integrated, but evaluated individually and compared at each position in open literature. Repeatability measures at various positions must be defined as one performance metric to compare performance among different robots. Thus, this research proposes a reliability estimation method to assess the system position repeatability of robots with repeatabilities at different positions. The proposed method estimates distributions of repeatabilities at various positions and then integrates the repeatabilities into an extreme-value. The reliability of system position repeatability was modeled using the extreme-value, and the Monte-Carlo simulation method was used to estimate reliability. The proposed method was applied to a robot with five different positions, and reliability estimation results as well as the applicability were demonstrated.

신뢰성과 연구 동향

손영갑,Son, Y.K. 한국전산구조공학회 2010 전산구조공학 Vol.23 No.2