Issue link: https://iconnect007.uberflip.com/i/1011746
32 PCB007 MAGAZINE I AUGUST 2018 To get a feel for the factors influencing Mil Handbook 217 MTBF prediction, the following information is necessary: • Thirteen various environments where defense electronics will be deployed • Whether the construction obeys IPC Class 3 or lower IPC class construction • The number of board layers from one to 18 (an 18-layer board was complicated in 1995) • A quality factor for whether the assembly is to be machine soldered or hand soldered So, why has there not been an update of Handbook MIL 217 since 1995? The book "Re- liability Growth, Enhancing Defense System Reliability" contains an explanation [1] . The lack of an update is something of a "hot pota- to" within the DoD, for the following reasons: 1. The general methodology of MIL 217 is to analyze field failures to compile the impor- tant reliability factors and derive a formula for their influence. However, there is not really a good mechanism for collecting field failure data across the various programs within the DoD. Failure data is hard to accumulate. 2. There is little precise information about the operating conditions of deployed defense electronics. Sometimes a weapon system has different environments— storage temperature, humidity, operating temperature when active, etc. Consider the operating conditions of a fighter aircraft: storage at various land or sea locations, operating at elevated desert or frigid high- altitude temperatures, standby for days but then extreme stress conditions for a one-hour mission, and so on. 3. The formulas are statistically based— whereas there is frequently an understand - ing of the influence of operating conditions. The latter was traditionally called physics of failure (PoF), but has recently been giv - en the more positive name "reliability phys- ics analysis" and has come to be a highly developed branch of reliability engineering. Within the DoD, the original critic of MIL Handbook 217 was the U.S. Army. That dates to 1995, when Assistant Secretary of the Army Gilbert Decker noted in "Policy on Incorporat- ing at Performance-Based Approach to Reli- ability in Requests for Proposals" that Army systems reliability should include: • Quantified reliability requirements and allowable uncertainties • Failure definitions and thresholds • Life-cycle conditions That is, reliability of Army systems should not be derived just from mathematical formu- las but should try to understand the conditions that influence failure. MIL 217 is still in use today, mainly for spare parts purchases at the start of, and during the useful life of weapons systems. The formulas, when use data is gathered, will generate fail- ure numbers for spare parts purchases. The purchase of new weapons systems can be de- fined with required spare parts and the whole deployment going forward—with a best esti- mate of how to keep the weapons system oper- ational for 20, 30, 40, or even 63 years. Is MIL 217 dead? The pioneering work on MIL 217 was done at the Rome Air Defense Center near Utica, New York. As recently as 2010, there has been a Defense Standards Pro- gram Office effort to generate MIL 217 Hand- book Revision G. However, Naval Support Ac- tivity–Crane, Indiana has been designated now as the "preparing activity" for any revisions that would be directed by the DoD. While there is still some hope to incorporate some physics reliability analysis into MIL 217, the ef- fort is minimal. So, how is the defense community to deal with reliability in the future? Recent advances in computer-based modeling now allow a rea- sonable prediction of assembled systems life. This modeling has moved beyond prediction of solder joint life and now includes other stress- es such as vibration and various mechanical shocks, properties of laminates, final board finishes, and plating properties of the copper board conductors. Great strides are being tak-