PREFERRED RELIABILITY PAGE 1 OF 7 PRACTICES PRACTICE NO. PT-TE-144 Practice: 1 Perfrm thermal dwell test n prtflight hardware ver the temperature range f +75 C/-2 C (applied at the thermal cntrl/munting surface r shearplate) fr 24 hurs at the cld end and 144 t 288 hurs at the ht end. Benefit: This test, cupled with rigrus design practices, prvides high cnfidence that the hardware design is nt marginal during its intended lng life high reliability missin. Prgrams Which Certified Usage: Vyager, Galile, Viking and Mariner Series Center t Cntact fr Infrmatin: Jet Prpulsin Labratry (JPL) Implementatin Methd: Establish a minimum hardware test temperature level range f -2 C/+75 C and specify that a single cycle thermal dwell test be perfrmed fr the apprpriate duratins (24 hurs cld and 144 t 288 hurs ht). Technical Ratinale: In the early 196s, JPL adpted a cnservative set f thermal design and test temperature levels t demnstrate hardware design adequacy. As a starting pint, a reasnable shrt term flight temperature excursin (+5 C t +5 C) was established fr thermal cntrl surfaces (shearplates). The +5 C lwer level is a few degrees Celsius abve the freezing pint f hydrazine, thus integrated thermal cntrl f bus electrnics and prpulsin systems is pssible. The 5 C upper limit is the apprximate level reached by a luvered bus electrnics bay after abut ne hur f full (perpendicular) slar irradiance at ne A.U. (astrnmical unit) and accmdates near earth maneuvers. The lng term desired thermal cntrl range is typically 25±5 C, but this range may be brader depending n the tradeffs f lng term reliability and thermal cntrl csts. This riginal apprach reduced the verall cmplexity f the system thermal cntrl design prcess: the wide range reduced the sensitivity t luver/radiatr size, heater size, heater size, pwer variatins, etc. A margin f JET PROPULSION LABORATORY 1 Thermal dwell testing is the standard practice at JPL fr systems and cmpnents which d nt thermally cycle during flight. Fr systems and cmpnents that d thermally cycle (generally ver a range > 2 C) in flight, the JPL practice is t cycle ver a cnservative range fr three times the number f flight cycles.
PRACTICE NO. PT-TE-144 PAGE 2 OF 7 ±25 C was then applied t the allwable flight range fr qualificatin and prtflight test levels f assemblies munted t such thermal cntrl surfaces. These levels accmmdate thermal cmprmises in the design where the shrt term extremes may be apprached during steady state peratin; they als has been demnstrated t prvide an effective screen f assemblies. This resulted in the JPL standard minimum test range f -2 C t +75 C (fr electrnic assemblies in particular). These cnservative test level ranges lead t several desirable features. The cnservative high temperature limit restricts the permitted temperature rise frm the shearplate t the junctin f electrnic pieceparts. Thus junctin temperatures during the bulk f a missin are much cler than assemblies designed and tested at lwer shearplate temperatures. The increase in theretical reliability 2 is n the rder f a factr f 1 per 25 C. There are at least tw failure mechanisms fr bth design and wrkmanship that shuld be screened by an adequate thermal envirnmental test f any given assembly. The first is based n Arrhenius rate related physics where time at high temperature is the key t demnstrating reliability during testing. Electrnic part life is a prime example f an Arrhenius mechanism, but s are ther elements f assemblies including interactins between metal traces within printed wiring bards (PWB's), certain cmpnent t bard jints, and even slder jints t a certain extent. The ther identifiable mechanism is thermally induced mechanical stress (including fatigue) as between cmpnents and the bard and especially slder jints. Arrhenius Rate Physics: Cntrast the test level f 75 C (shearplate) t 5 C shrt term wrst case transients during flight and 25 C fr the bulk f the missin. Based n Arrhenius reactin rate physics described and shwn n page 5, the 75 C test prvides a demnstrated reliability sme 2 t 8 times that f shrt transients t 5 C, (typical f thermal cycling tests), and sme 4 t 94 times that f lng term missin shearplate temperatures (25 C). These reliability ratis are based n activatin energies f.3 ev t 1. ev which cver mst assembly element reactin physics. The Mariner and Viking spacecraft perfrmed a ht dwell test (75 C) f 288 hurs duratin. This was reduced t 144 hurs fr the Vyager and Galile spacecraft. The statistical database supprting this shrter test is unique t the JPL design rules and prcesses; therefre, the lnger ht dwell duratin f 288 hurs is recmmended fr assemblies designed t nn-equivalent r less cnservative practices. On page 6 we shw the percentage f the screening test capability fr Class S parts that is used by a JPL assembly test at 75 C fr 144 hurs. A very cnservative assumptin here is that all parts in the assembly test have a 35 C temperature rise and that they are at 11 C fr the entire test. Even given this ver-cnservative assumptin, the JPL test uses nly.18% f the class S parts minimum 2See "Part Junctin Temperature", Practice N. PD-ED-124
PRACTICE NO. PT-TE-144 PAGE 3 OF 7 screened capability. Clearly less than 2/1's f the minimum parts capability being dedicated t the assembly prtflight test is nt a cncern. The parts are nt ver-stressed by this test. Thermally Induced Mechanical Stress (Fatigue): JPL has histrically dne a thermal dwell test rather than a specific thermal cycle test. There are data that indicate thermal cycling uses up hardware life and therefre is degrading t the flight hardware. In practice, the JPL test apprach is never really just a ne-cycle dwell test. The assembly test prgram (plus any retest) and the systems test prgram (frequently tw phases) result in a minimum f tw cycles and as many as fur (r mre) are pssible althugh they are nt cntinuus and the transients are cntrlled t < 3 C/hr t prevent thermal shck. The Vyager hardware was tested as fllws: Prf Test Assemblies Qualificatin Test Flight Assemblies Acceptance Test Cycles: 1 Assembly (+ Retest) 1 Assembly (+ Retest) 2 Systems 1 Systems 3 Cycles (+ Retest) 2 Cycles (+ Retest) In a recent JPL study, a fatigue life relatinship f equivalent thermal cycles was determined ver different temperature ranges as fllws: C 2 ' C 1 T 1 T 2 Y where: C 1 is the number f thermal cycles ver a T 1 range C 2is the number f thermal cycles ver a T 2 range and Y = 2.6 fr eutectic slder. As a frame f cmparisn fr wrkmanship purpses, the JPL prtflight test f 1 cycle ver -2/ 75 C range can be crrelated t an acceptance test f 6 cycles ver a /5 C range. In this case: C 1 = 1, T 1 = 95 C, C 2 = TBD, T 2 = 5 C and the equivalent cycles f the JPL test are: 2.6 C 2 = 1(95 C/5 C) = 5.3 cycles.
PRACTICE NO. PT-TE-144 PAGE 4 OF 7 Therefre, in terms f slder jint fatigue life, the JPL prtflight test equivalency t 5.3 cycles ver a 5 C range says that, fr wrkmanship acceptance purpses, the JPL prtflight test is essentially the same as the example thermal cycle acceptance test, i.e., 5.3. 6 cycles. On page 7, a cmparisn f slder jint fatigue life cmparisns has been made. The recmmended -2/+75 C single cycle dwell test uses nly.14% f the fatigue life f a slder jint qualified t NHB 53.4 (3A-1). The pint f this cmparisn is that the JPL prtflight test is less strenuus t slder jints than thermal cycle testing perfrmed by mst rganizatins. Grund Test & Thermally Related Prblem/Failure Statistics: These practices were applied t the Mariner spacecraft series, the tw Viking 75 spacecraft, the tw Vyager 77 spacecraft, and mre recently Galile. These spacecraft all cmpleted (r exceeded) their intended missin successfully (the Galile missin is still underway at the time f this editin). In fact, the Vyager spacecraft have wrked fr ver 13 years. The ttal number f assembly prblems/failures during these missins is small, and the number f thermally induced prblems even smaller. This is shwn in the fllwing table where the number f prblem/failures identified during assembly level thermal testing are cmpared with suspected flight prblems/failures fr the Viking, Vyager, and Galile prgrams: Number f Prblem/ Failures Identified during Assembly Thermal Testing Number f Knwn Thermally Induced Flight Prblem Failures VIKING 251 (2 SPACECRAFT) Nne Obvius VOYAGER 123 (2 SPACECRAFT) 1 GALILEO (1 SPACECRAFT) 5 Nne t Date Impact f Nn-practice: Demnstrated design adequacy and its implicatins t lng term reliability are affected. Fr example, testing at 5 C instead f 75 C and fr abut 2 hurs instead f 144 hurs reduces test demnstrated reliability by a factr n the rder f 5.
PRACTICE NO. PT-TE-144 PAGE 5 OF 7 ARRHENIUS REACTION RATE THEORY & E a 8 k ' e 8 1 T & 1 T where: 8 - Reactin rate at temperature T (a measure f failures/time) 8 - Reactin rate at reference temperature T E a - Activatin energy, ev T - Temperature in degrees Kelvin ( K) T - Reference temperature in degrees Kelvin ( K) -5 k - Bltzmann's cnstant (8.617 x 1 ev/ K) ACTIVATION ENERGY (ev) 6.3.7 1. TEST CONDITION 75 C SHEARPLATE 8 JUNCT. 8 25 C 13.3 42.2 5595.1 SHORT TERM FLIGHT TRANSIENT 5 C SHEARPLATE 85 C JUNCTION LONG TERM FLIGHT CONDITION 25 C SHEARPLATE 6 C JUNCTION 8 JUNCT. 8 25 C 8 JUNCT. 8 25 C 7.1 95.8 676.5 3.4 17.5 59.6 TEST CONDITION OVER SHORT TERM FLIGHT TRANSIENT 8 11 C 8 85 C 1.9 4.4 8.3 TEST CONDITION OVER 8 11 C LONG TERM 8 6 C 3.9 24. 93.9 RATIO OF ARRHENIUS FAILURE RATES FOR VARIOUS ACTIVATION ENERGIES AND PRACTICE CONDITIONS
PRACTICE NO. PT-TE-144 PAGE 6 OF 7
PRACTICE NO. PT-TE-144 PAGE 7 OF 7 SOLDER JOINT FATIGUE LIFE COMPARISON NHB 53.4 (3A-1) PACKAGING QUALIFICATION TEST JPL PROTOFLIGHT -55 C t 1 C -2 C t 75 C 2 cycles 1 cycle 24 hrs cld 144 hrs ht QUALIFICATION BASELINE EXPOSURE LIFE EFFECT * 1 cycle.14% f f NHB 53.4(3A-1) 95 C slder jint ( 95 155 2.6 1 2 '.14 '.14%