Prognostics and Health Management of Electronics by Utilizing Environmental and Usage Loads
| dc.contributor.advisor | Pecht, Michael G | en_US |
| dc.contributor.author | Vichare, Nikhil Manohar | en_US |
| dc.contributor.department | Mechanical Engineering | en_US |
| dc.contributor.publisher | Digital Repository at the University of Maryland | en_US |
| dc.contributor.publisher | University of Maryland (College Park, Md.) | en_US |
| dc.date.accessioned | 2006-09-12T05:43:14Z | |
| dc.date.available | 2006-09-12T05:43:14Z | |
| dc.date.issued | 2006-07-06 | en_US |
| dc.description.abstract | Prognostics and health management (PHM) is a method that permits the reliability of a system to be evaluated in its actual application conditions. Thus by determining the advent of failure, procedures can be developed to mitigate, manage and maintain the system. Since, electronic systems control most systems today and their reliability is usually critical for system reliability, PHM techniques are needed for electronics. To enable prognostics, a methodology was developed to extract load-parameters required for damage assessment from irregular time-load data. As a part of the methodology an algorithm that extracts cyclic range and means, ramp-rates, dwell-times, dwell-loads and correlation between load parameters was developed. The algorithm enables significant reduction of the time-load data without compromising features that are essential for damage estimation. The load-parameters are stored in bins with a-priori calculated (optimal) bin-width. The binned data is then used with Gaussian kernel function for density estimation of the load-parameter for use in damage assessment and prognostics. The method was shown to accurately extract the desired load-parameters and enable condensed storage of load histories, thus improving resource efficiency of the sensor nodes. An approach was developed to assess the impact of uncertainties in measurement, model-input, and damage-models on prognostics. The approach utilizes sensitivity analysis to identify the dominant input variables that influence the model-output, and uses the distribution of measured load-parameters and input variables in a Monte-Carlo simulation to provide a distribution of accumulated damage. Using regression analysis of the accumulated damage distributions, the remaining life is then predicted with confidence intervals. The proposed method was demonstrated using an experimental setup for predicting interconnect failures on electronic board subjected to field conditions. A failure precursor based approach was developed for remaining life prognostics by analyzing resistance data in conjunction with usage temperature loads. Using the data from the PHM experiment, a model was developed to estimate the resistance based on measured temperature values. The difference between actual and estimated resistance value in time-domain were analyzed to predict the onset and progress of interconnect degradation. Remaining life was predicted by trending several features including mean-peaks, kurtosis, and 95% cumulative-values of the resistance-drift distributions. | en_US |
| dc.format.extent | 4368463 bytes | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.uri | http://hdl.handle.net/1903/3772 | |
| dc.language.iso | en_US | |
| dc.subject.pqcontrolled | Engineering, Mechanical | en_US |
| dc.subject.pqcontrolled | Engineering, Mechanical | en_US |
| dc.subject.pquncontrolled | Reliability | en_US |
| dc.subject.pquncontrolled | prognostics | en_US |
| dc.subject.pquncontrolled | health-monitoring | en_US |
| dc.subject.pquncontrolled | HUMS | en_US |
| dc.title | Prognostics and Health Management of Electronics by Utilizing Environmental and Usage Loads | en_US |
| dc.type | Dissertation | en_US |
Files
Original bundle
1 - 1 of 1