Downloaded from https://www.everyspec.com detailed specificatlofl of the systeni state humidity, and corrosive elements, There IsA will vary with out needs and desires iind with a large body of material In the mechanical the required tractability of the resulting avd metsallurgical fields dealingj with those equations. The state of its'stont ordinarily gross failure modes. Since the behavior of will have several dimensions (components); electronic components is organized different- so it can be classed as a vector, For example, ly, there is no organize0l body of literature consider a resistor.
If we are concerned only which deals with gross faralw modes that cut about Ito resistiltce and nothing else then the across all components, A number of inform&- state of the system will be given by the re- tion sources on accelerated testing of electran- uiatance of the device (or something equlva- ic components are available, many of which lent thereto such as a ratio of the resistance ame listed In Ref. 3; some of thenm contain'con- ' to an InItia resistance). On tie o ther hand, ceptual errors sid ought to be read critically. i we may be concerned about the resistance, Ref.
3 Itself is dated (as are Its references) the temperature coefficient of resistance, the because it was issued In 1968. The failure voltage coefficient of resistance, and the modes of semiconductors, or at least their chemical composition of the resistive mate- relative Importance, have changed drastically rial. 7Then there will be several dimensions since then. for the systemn state, and two states will not 44 be the same anleso all corresponding dlmen- 74 CONSTANT SE VERITY-LI V9L dosian. r pair-wiae the same.
METHOD a Lst one be concaeda~ that assocuiatiingh raiing ye faceeae asystem state only wihi systemmoeThsIthtrdinatyefaclrtd rather than with the system Itself Is too test In which the severity level remains con- sloppy, An analogy can be mnade tu therio. stant throughout the liffe of the items on dynamics. There can be many thermody- test. It Is customary to rnm tests at several namic models or a systein depending on what severity levols and to plot a ourve of some is of concern. The entropy is rnot defined for psrometef such as failure rato vs severity-. the system itself but only with regard to a level.
A sample of several itemns usually Is particular thermodynamic model of that sys- put on test, and thja test stopped when some tern. fraction of the ou~lgnal sample list failed or a specified teat thime has eliepsd. For re- In addition to verifying that true acceler- liability prediction purpose., the early ation exists, a great deal of effort musi be fractlor that falls is most imijortint because expended In determining the acceleration only thj sthort-lived Items are g~Mng I.c. affect flanction. It Is virtually always pros~aned the reliability seriously, that the acceleration function is a constant. V~-5 STEP-STRESS AND PRlOGRESSIVE.
743 FAILURE MODES AND MECHANISMS SThU$S METHODS Some gross failure modes which can be The woro "strews" Is used In the tense of isocelerated for me.hanlcal parts are failgue, ieverity level. In this raethra, tho severity corrosion, critep-flapture, stress corrosion, and opplied to a saiiupla oit'tcmi Is Increased 1,r various combination. of 'hom, In ciectron- stops or lticretm~nts until some criterion It Ics, one does not ordlairrily speeitV the gruss m~et for termiretins the test. All steps dto failure niA'.es for acceleration, but, rather, not have to be tho same size, even thokigh V,: specifies the "stresses" which aru being ifl- this is a coinniosa nrattice, ( ciensed.
Some cif these are toniperaturi,~4 supply voltage, powevr dissipation. Yvibration, Tito term step-stress uised In th~e litereture A.ja