• 한국신뢰성학회지:신뢰성응용연구
• /
• 제17권2호
• /
• pp.103-111
• /
• 2017

Purpose: Accelerated degradation tests can speed time to market and reduce the test time and costs associated with long term reliability tests to verify the required service life of a product or material. This paper proposes a service life prediction method for components or materials using an accelerated degradation tests based on the relationships between temperature and the rate of failure-causing chemical reaction. Methods: The relationship between performance degradation and the rate of a failure-causing chemical reaction is assumed and least square estimation is used to estimate model parameters from the degradation model. Results: Methods of obtaining acceleration factors and predicting service life using the degradation model are presented and a numerical example is provided. Conclusion: Service life prediction of a component or material is possible at an early stage of the degradation test by using the proposed method.

• International Journal of Reliability and Applications
• /
• 제1권2호
• /
• pp.155-165
• /
• 2000

In this paper, we propose some estimators of Kullback- Leibler Information functions using the data from three step stress accelerated life tests. This acceleration model is assumed to be a tampered random variable model. Some asymptotic properties of proposed estimators are proved. Simulations are performed for comparing the small sample properties of the proposed estimators under use condition of accelerated life test.

• 대한산업공학회지
• /
• 제20권1호
• /
• pp.39-51
• /
• 1994

Almost all accelerated life tests assume that no basic failure mechanism changes within the test stresses. But accelerated life test, considering failure mechanism changes, is needed since failure mechanism changes when accelerating beyond the used stress. This paper studies the analysis when the failure mechanism changes within the test stresses. The piecewise linear regression, which the join point of two lines is estimated, is applied In particular, two accelerated life tests, with and without a change in failure mechanism are examined.

• Journal of the Korean Data and Information Science Society
• /
• 제15권4호
• /
• pp.935-943
• /
• 2004

Most of the accelerated life tests deal with tests that use only one accelerating variable and no other explanatory variables. Frequently, however, there is a test to use more than one accelerating or other experimental variables, such as, for examples, a test of capacitors at higher than usual conditions of temperature and voltage, a test of circuit boards at higher than usual conditions of temperature, humidity and voltage. A accelerated life test is extended to M-level stress accelerated life test with k-stress variables. The optimal design for Weibull distribution is studied with k-stress variables.

• 대한안전경영과학회:학술대회논문집
• /
• 대한안전경영과학회 2004년도 추계학술대회
• /
• pp.229-235
• /
• 2004

This paper treats accelerated life tests for automotive connector. The contact resistance of connector is explained by some factors; the use time(calender time, real cycle), stresses and loads adapted in auto test. The relationships between contact resistance and some factors are compared and analyzed by regression models in various test conditions; field use-environment, manufacturer's test environment, and accelerated test condition. The consistency between of manufacturer's test and field test is examined. Finally, the future study on accelerated test for automotive connector is presented.

• 품질경영학회지
• /
• 제31권4호
• /
• pp.194-202
• /
• 2003

In this paper, we propose goodness of fit test statistics for exponentiality in accelerated life tests data based on Kullback­Leibler information functions. This acceleration model is assumed to be a tampered random variable model. The procedure is applicable when the exponential parameter using the data from accelerated life tests is or is not specified under null hypothesis. And we compare the power of the proposed test statistics with Kolmogorov­Smirnov, Cramer von Mises and Anderson­Darling statistics in the small sample.

• 한국신뢰성학회지:신뢰성응용연구
• /
• 제9권2호
• /
• pp.71-80
• /
• 2009

This paper discusses the service life prediction methods for CFRP bar for concrete reinforcement using accelerated degradation tests. The relationship between performance degradation and the rate of a failure-causing chemical reaction is assumed for the temperature accelerated degradation tests. Methods of obtaining acceleration factors and predicting service life of the CFRP bar using the degradation model are presented.

• 한국강구조학회 논문집
• /
• 제19권1호
• /
• pp.79-86
• /
• 2007

도장, 용사 및 도금 등의 강교용 방청법의 장기내구성 평가에는 대기노출실험법이 사용되고 있으나, 노화데이터를 얻기까지는 수년에서 수십 년이라는 장기간의 실험기간을 필요로 한다. 본 연구에서는 기존의 방청법을 포함하여 최근에 개발, 사용되고 있는 강교용 방청법의 장기내구성의 단기간 평가를 위한 기초단계로서 무도장 강재의 부식내구성 평가를 위한 부식환경 촉진실험법의 활용에 대해 검토하였다. 본 실험에는 일본공업규격(JIS)에 규정되어 있는 S6-cycle 부식환경 촉진실험조건을 이용하여, 30, 60, 90, 120, 150일간의 무도장 강재의 부식 중량감소량 및 두께감소량의 경시변화를 정량적으로 평가하였다. 그리고 기존의 대기노출실험과 본 실험으로부터 구한 중량감소량을 이용하여 촉진배율을 산출하고, 대기노출실험지의 지역환경구분, 해안선으로부터 거리, 아황산가스량 및 비래(飛來)염분량과의 관계를 검토하였다. 그 결과, 촉진배율을 비래염분량으로 정식화하였으며, S6-cycle 부식촉진실험결과의 대기노출환경에의 활용 가능성을 제시하였다.

• 품질경영학회지
• /
• 제23권1호
• /
• pp.41-49
• /
• 1995

This paper suggests optimal accelerated constant stress life tests in Exponential distribution. The relationship between the log-mean life and the loaded stress is assumed to be linear. Optimal plans considering mean square errors of maximum likelihood estimators of the log mean life and test costs are obtained. We consider accelerated life tests with two stress levels, and as data types, failure censoring( type II) and time censoring(type I) data are used. We propose the procedure to obtain the optimal plans for each case. Some examples are also included.

• 한국산업정보학회:학술대회논문집
• /
• 한국산업정보학회 2009년도 춘계학술대회 미래 IT융합기술 및 전략
• /
• pp.107-113
• /
• 2009

This paper deals with noninformative priors for such as Jeffres' prior, reference prior and probability matching prior for scale parameter of exponential distribution when the data are collected in multiple step stress accelerated life tests. We find the noninformative priors for this model and show that the reference prior satisfies first order matching criterion. Using artificial data, we perform Bayesian analysis for proposed priors.