REG NASA-LLIS-0783-2000 Lessons Learned - Thermal Test Levels& Durations.pdf

1、Best Practices Entry: Best Practice Info:a71 Committee Approval Date: 2000-04-14a71 Center Point of Contact: JPLa71 Submitted by: Wilson HarkinsSubject: Thermal Test Levels they also have been demonstrated to provide an effective screen of assemblies. This resulted in the JPL standard minimum test r

2、ange of -20C to +75C (for electronic assemblies in particular).These conservative test level ranges lead to several desirable features. The conservative high temperature limit restricts the permitted temperature rise from the shearplate to the junction of electronic piece parts. Thus junction temper

3、atures during the bulk of a mission are much cooler than assemblies designed and tested at lower shearplate temperatures. The increase in theoretical reliability is on the order of a factor of 10 per 25C. (refer to “Part junction Temperature“, Reliability Preferred Practice No. PD-ED-1204)There are

4、at least two failure mechanisms for both design and workmanship that should be screened by an adequate thermal environmental test of any given assembly. The first is based on Arrhenius rate related physics where time at high temperature is the key to demonstrating reliability during testing. Electro

5、nic part life is a prime example of an Arrhenius mechanism, but so are other elements Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-of assemblies including interactions between metal traces within printed wiring boards (PWBs), certain component to

6、board joints, and even solder joints to a certain extent. The other identifiable mechanism is thermally induced mechanical stress (including fatigue) as between components and the board and especially solder joints.Arrhenius Rate Physics:Contrast the test level of 75C (shearplate) to 50C short term

7、worst case transients during flight and 25C for the bulk of the mission. Based on Arrhenius reaction rate physics described in the following figures, the 75C test provides a demonstrated reliability some 2 to 8 times that of short transients to 50C, (typical of thermal cycling tests), and some 4 to

8、94 times that of long term mission shearplate temperatures (25C). These reliability ratios are based on activation energies of 0.3 eV to 1.0 eV which cover most assembly element reaction physics.refer to D descriptionD Provided by IHSNot for ResaleNo reproduction or networking permitted without lice

9、nse from IHS-,-,-refer to D descriptionD The Mariner and Viking spacecraft performed a hot dwell test (75C) of 288 hours duration. This was reduced to 144 hours for the Voyager and Galileo spacecraft. The statistical database supporting this shorter test is unique to the JPL design rules and process

10、es; therefore, the longer hot dwell duration of 288 hours is recommended for assemblies designed to non-equivalent or less conservative practices.The following figure shows the percentage of the screening test capability for Class S parts that is used by a JPL assembly test at 75C for 144 hours. A v

11、ery conservative assumption here is that all parts in the assembly test have a 35C temperature rise and that they are at 110C for the entire test. Even given this over-conservative assumption, the JPL test uses only 0.018% of the class S parts minimum screened capability. Clearly less than 2/10000s

12、of the minimum parts capability being dedicated to the assembly protoflight test is not a concern. The parts are not over-stressed by this test.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-refer to D descriptionD Thermally Induced Mechanical Stres

13、s (Fatigue):JPL has historically done a thermal dwell test rather than a specific thermal cycle test. There are data that indicate thermal cycling uses up hardware life and therefore is degrading to the flight hardware. In practice, the JPL test approach is never really just a one-cycle dwell test.

14、The assembly test program (plus any retest) and the systems test program (frequently two phases) result in a minimum of two cycles and as many as four (or more) are possible although they are not continuous and the transients are controlled to 30C/hr to prevent thermal shock. The Voyager hardware wa

15、s tested as follows:Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-refer to D descriptionD In a recent JPL study, a fatigue life relationship of equivalent thermal cycles was determined over different temperature ranges as follows:refer to D descrip

16、tionD where: C1is the number of thermal cycles over a T1range C2is the number of thermal cycles over a T2range and Y = 2.6 for eutectic solder. As a frame of comparison for workmanship purposes, the JPL protoflight test of 1 cycle over -20/75C range can be correlated to an acceptance test of 6 cycle

17、s over a 0/50C range. In this case:C1= 1, T1= 95C, C2= TBD, T2= 50Cand the equivalent cycles of the JPL test are:C2= 1(95C/50C)2.6= 5.3 cycles.Therefore, in terms of solder joint fatigue life, the JPL protoflight test equivalency to 5.3 cycles over Provided by IHSNot for ResaleNo reproduction or net

18、working permitted without license from IHS-,-,-a 50C range says that, for workmanship acceptance purposes, the JPL protoflight test is essentially the same as the example thermal cycle acceptance test, i.e., 5.3 equals approximately 6 cycles.The following figure provides comparison of solder joint f

19、atigue life. The recommended -20/+75C single cycle dwell test uses only 0.14% of the fatigue life of a solder joint qualified to NHB 5300.4 (3A-1). The point of this comparison is that the JPL protoflight test is less strenuous to solder joints than thermal cycle testing performed by most organizati

20、ons.refer to D descriptionD Ground Test & Thermally Related Problem/Failure Statistics:These practices were applied to the Mariner spacecraft series, the two Viking 75 spacecraft, the two Voyager 77 spacecraft, and more recently Galileo. These spacecraft all completed (or exceeded) their intended mi

21、ssion successfully (the Galileo mission is still underway at the time of this edition). In fact, the Voyager spacecraft have worked for over 13 years.The total number of assembly problems/failures during these missions is small, and the number of thermally induced problems even smaller. This is show

22、n in the following table where the number of problem/failures identified during assembly level thermal testing are compared with suspected flight problems/failures for the Viking, Voyager, and Galileo programs:Provided by IHSNot for ResaleNo reproduction or networking permitted without license from

23、IHS-,-,-refer to D descriptionD Impact of Non-Practice: Demonstrated design adequacy and its implications to long term reliability are affected. For example, testing at 50C instead of 75C and for about 20 hours instead of 144 hours reduces test demonstrated reliability by a factor on the order of 50.Related Practices: N/AAdditional Info: Approval Info: a71 Approval Date: 2000-04-14a71 Approval Name: Eric Raynora71 Approval Organization: QSa71 Approval Phone Number: 202-358-4738Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-