0
Research Papers

Evaluating the End of Maintenance Dates for Electronic Assemblies Composed of Obsolete Parts

[+] Author and Article Information
Anthony Konoza, Peter Sandborn

CALCE Electronic Products and Systems Center,
Department of Mechanical Engineering,
University of Maryland,
College Park, MD 20742

Legacy systems are defined as fielded and operational systems for which no new system production or replacement is planned.

This is a very common situation for military and civil infrastructure systems for which there is effectively no end of support date, i.e., “ageless systems” (or end of support dates are regularly extended). For example, consider air traffic control systems, while replacement of these types of aging systems has been proposed, the date on which the actual replacement will occur is unclear and depends more on economic and political pressures than system realities. In some cases the type of analysis described in this paper is a necessary step in making a business case to replace legacy systems.

It is also possible that the same part appearing in different locations in the same system has different reliability characteristics.

Alternative stochastic models that could be considered include Markov chain models and Petri nets.

Note, at the start of the analysis the system has already been fielded for a substantial length of time and the parts currently in the systems are not “good as new.” Therefore, sampling from the time-to-failure distributions begins prior to the analysis start date.

A part is defined as a component specific to a particular card and retains its own card-specific unique properties (time-to-failure distribution and quantity). Each instance of a part on a card is treated individually (represented by its own forecasted part demand based on sampling from the part's time-to-failure distribution).

The failure data is right censored because not all the fielded parts have failed to date. Right censoring occurs in reliability testing when some of the units in the population survive the full test time period without failing.

The coefficient of variation for case 1 is 0.01—to determine statistically significant deviations from the mean, a number for comparisons must fall outside this range. When the difference between the means of cases 1 and 2 are calculated, it is determined that, although small, there is statistical significance between the two cases.

This is the same as the “just in time refresh date” assumption made in general refresh planning, [8].

Contributed by the Design for Manufacturing Committee of ASME for publication in the JOURNAL OF MECHANICAL DESIGN. Manuscript received February 9, 2013; final manuscript received November 15, 2013; published online January 10, 2014. Assoc. Editor: Irem Y. Tumer.

J. Mech. Des 136(3), 031006 (Jan 10, 2014) (14 pages) Paper No: MD-13-1076; doi: 10.1115/1.4026096 History: Received February 09, 2013; Revised November 15, 2013

Long-term support of legacy electronic systems is challenging due to mismatches between the system support life and the procurement lives of the systems’ constituent components. Legacy electronic systems that are used in safety, mission, and infrastructure critical applications that must be supported for 20+ yr are threatened with diminishing manufacturing sources and material shortages (DMSMS)-type obsolescence, and their effective system support lives may be governed by existing nonreplenishable inventories of spare parts. This paper describes the development of the end of maintenance (EOM) model, which uses a stochastic discrete-event simulation that follows the life history of the population of parts in a system using time-to-failure distributions and other forecasted demands. The model determines the support life of the system based on existing inventories of spare parts and cards, and optionally harvesting parts from existing cards to extend the support life of the system. The model includes: part inventory degradation, periodic inventory inspections, and design refresh planning for selected cards. A case study using a real legacy system comprised of 117,000 instances of 70 unique cards and 4.5 × 106 unique parts is presented. The case study was used to evaluate the support life of a system with various future failure assumptions, including with and without the use of part harvesting. The case study also includes sensitivity analyses for selected design refreshes to maximize potential system life-cycle capabilities, and optional design refresh planning required to sustain the system to a specific date.

Copyright © 2014 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

EOR/EOM part failure process flow (single life history)

Grahic Jump Location
Fig. 2

Discrete-event simulation flow for multiple events

Grahic Jump Location
Fig. 3

Part degradation in inventory (shelf-life) process flow for a single part

Grahic Jump Location
Fig. 4

Part number 3798-05 failure distribution. Both data sets are equal, one shows 10% unreliable, the other 100% unreliable (censored versus uncensored).

Grahic Jump Location
Fig. 5

Part number 5004-02 failure distribution. Both data sets are equal, one shows 10% unreliable, the other 100% unreliable (censored versus uncensored).

Grahic Jump Location
Fig. 6

System-level EOM distribution (left), EOM (top right), and EOR (bottom right) results for case 1.

Grahic Jump Location
Fig. 7

System-level EOM distribution (left), EOM (top right), and EOR (bottom right) results for case 2

Grahic Jump Location
Fig. 8

System-level EOM distribution (left), EOM (top right), and EOR (bottom right) results for case 3

Grahic Jump Location
Fig. 9

Card-level support tracking and loss

Grahic Jump Location
Fig. 10

Card-level EOM distributions (left) and EOR/EOM results for case 4 (right)

Grahic Jump Location
Fig. 11

Card-level EOM distributions (left) and EOM results for case 5 (right)

Grahic Jump Location
Fig. 12

System support life gained from individual card refresh, case 1

Grahic Jump Location
Fig. 13

System support life gained from individual card refresh, case 2.

Grahic Jump Location
Fig. 14

System support life gained from individual card refresh, case 3

Grahic Jump Location
Fig. 15

Design refresh plan (left) and completed refresh date distribution (right), case 1

Grahic Jump Location
Fig. 16

Design refresh plan, case 2

Grahic Jump Location
Fig. 17

Design refresh plan (left) and completed refresh date distributions (right), case 3

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In