Downloaded from https://www.everyspec.com AMCP 7"4-11111 3. Qualitative - to see what happens factor may be defined as ,,(t) U dr(t)/dt, when the user maltreats the device as he True acceleration is illustrated in Figure probably will 7-1(A) for 2 state vector with a single di- mension - resistance ratio of a resistor. ;•,. 4, Quantitative -- to nmake a prediction It is, of course, nice it" AWl is a constant , about the life tnder actual operating con- with respect to time as in the figure and ditions. depends in some quite tractable way on the severity level, Ther.
is little question that accelerated testing is useful for the 3 quaitttllve mew- Estimates of an acceleration factor will sures; so it is mainly the quanithative prob. depend on the statistical procedures used t6 mle to which this chapter i, addressed. arrive at them, 7-2 TRUE ACCELERATION (Rf, 3) It is important to recognize the arbitravil. e d i f t c aness of the definition especially as regards -So. rl definitions' for true acceleration the word, reasonably. In order to have true 1 appeaNin the literature, some of which are acceleration, it is only necessary that the not ierst explicit. Most ensineeoj.
associate thinp in which we are immediately intf'ested true..aceleration with behavior over time. be close enough under the two sets of'con- The one liven here is chosen for its gener- dittons, To be specific, npt all failure modes ality and applicability, Acceleration need and mechanisms need be identical down to not be true to be useful even though untrue the last electron orbital, acceleration is more difficult to analyze, even qualitatively. Generally, the physical condition of the device will be included in the system state A Acceleration is true if and only if the sys- either explicitly or Implicitly In sufficient tern.
under the accelerated conditions, panes detail to permit judgments to be made about "reasonably 2 through equivalent 3 states and its deaign and construction relative to the in the same order it did at the usual con- failure modes and mechanisms. ditions. Let g(t) be the state of the system under usual ,onditions and let G(1) be the The state of a system is not uniquely de- equivalent state of the system under acceler- fined for a physical system, it is defined only ated damagers (0 is not the state at the for a conceptual model of the system.
The accelerated conditions but is the state after being transformed reversibly down to the one very poor choice is to easat that soosisrtion is usual conditions). Then there Is true accel- t. if and only if it follows the Artheniua equation. An- eration If and only if: other poor choine is to "Sociate It with the constant raaod rat,, 'The word "reasonably" in necessary because the 1. GO() W gt(t)] needs end dealrm of the stitation may be diffetent from time to time, and as enlnemn.
If things, a aoe enough for the pur•nos at hand, thae is no need to worry about 2, r(t) is a monotonically 4 increasing the discropancitr as e(ar a theme purposes ae conoertned, 'Two states of a syslrtn are equivalent If and only Ir function of the argument one wen be reversibly transformed Into the other by ohm*s. ins tim environment. For example, a resistor at a higher tempolature might never have the same resistance it would 3, G(O) 1 g(0) at a lower temperature, solely becauNe of its temperature coefficient, This Illustrated in Fig. 7-1(B). 4. r(0) 01.
4 pot those who think the term is ambigusous, mono tonic Is and here in the strict ense., L e., staying oontanit in not allowed, The acceleration factor A is a defined n If 0 and j have a one-to-one corraepondence with tim, A~T) * r An incremental acceleration ftor s argument (the reciprocal function exists), this lea logical ,.. , Af [ t/. An incremental acceleration consequence of #1 Md #3. ""0 ;04