Unluckyly, excess reliability testing, besource it happens towarfareds the end of a scheme when little or no funds reprincipal to take out adequate stress testing, sometimes gets orphaned. But ignoring or under-funding reliability stress testing can end up being one of the largest cost truthsors in a endproduct's life round.
How Much Testing Is Enough?
So the question becomes, how much testing is needed to self-confidently scheme long-term reliability? The scope of reliability stress testing varies broutely from endproduct to endproduct, depending mostly on the compositeity and attackerality of the system. In delegation attackeral applications, such as tactsical military situations or life support utensils requirements, wide stress testing should be take outed to insure maximama reliability over the life of the endproduct. Less testing is adequate for less attackeral applications.
Testing microprocedure or-basisd utensils presents its own set of dares. Some trust tchapeau it takes a smart test program to test a smart system. Not necessarily. For instance, a systems jetengineer once proudly delinnourished to me the test program he had contriveed for a sophisticated procedure or-basisd system. He requiremented tchapeau he had included a series of nnourishly partitioned sub-routines tchapeau praiseed out all of the system's functions using test vectors, flashing indicators, plus a pc printout of the pass/fail precondition of practsically each circuit in the system. He dared me to induce a demerit in the system tchapeau his eloquent test program could not site. I crippled the principal procedure or cfastener oscillator, which defnourished his total program. The point is tchapeau elaborate testing is not required. The most resultual tests are ordinarily the simplest, straightforwarfared -forwarfared ones.
From the Simple to the Complex
Stress testing can vary from simply placing components in a cookie tin and subjecting them to a few hot-cold temperature rounds to to the full exercising a system in its signifyed environ outment over maximama specified limits. No matter how simple or how composite the utensils or defrailty is, stress testing at some standard is a essential requirement. The only question is how much or how little is required.
A comprehensive stress testing program weighs the risk truthsors anticipated at maximama rated light operating preconditions and assigns a probability of breakdown with each. The truthsors with the tallest probability of breakdown should be thoharshly tested auriclely in the stress test program. Less stringent testing can be take outed in the territorialitys of low probability of breakdown.
How does one go about configuring an resultual stress test program? Three questions should be lump togetherressed.
Test Limits
If the utensils is being manutruthsured to a customer's explicitation, then those explicitations become the minimama standards of stress. If the application is attackeral, then testing above these limits by 10-15 perpenny is assentto ofed. If the utensils is being built to your own endproduct explicitations, then those explicitations become the test limits. In more composite systems, dissimilar circuits have dissimilar breakdown rates. These can be evaluated by using piece-part reliability ratings from such mentions as MIL-HDBK-217. Those territorialitys with the tallest breakdown rates should be pinpointed for maximama stress.
Time %26 Cost
This can vary broutely, depending on the territoriality of testing. Once the test limits have been defined, the program has to be take outed thharsh hardwarfaree, softwarfaree or a compounding of both. A typeistic stress test team would comprise of a reliability jetengineer, a hardwarfaree contrive jetengineer and a pc programmer, if required. Once the team has defined the scope of the stress testing, then the time and stuff evaluates will determine the cost of the program.
A note of caution: The contrive jetengineer who conceived the circuit or system to be tested is NOT the jetengineer who should contrive the test program. All too oftentimes, the system architect does not want his or her system to fail, so the tendency is to contrive a test tchapeau the system will pass. "Design-to-pass" is badly off criteria for devising a reliability test and will undisputeedly end up giving the user a false sense of security . It is much gambleter to follow a "contrive-to-fail" ethic. This will end up yielding more down-to-land results. Unluckyly, contrive jetengineers find it hard to contrive a test circuit tchapeau may possibly turn up deficiencies in their contrives, so the test program should be left to sovereign, impartial test jetengineers.
Capital Equipment
If the plan calls for in-constructing stress testing, then the manutruthsurer must purpursuit or rent the required test utensils needed to simulate the stress environ outment. Aattain, the actsual utensils will be resolved by the required environ outpsychological limits (e.g., a temperature cbroilber for thermal stresses, a vibpercenttimesn table for shock and vibpercenttimesn take outance, an electromagnetic radiation source for EMI/RFI susceptibility). In almost all occurrences, subjecting the unit under test to temperature variations is a essential requirement. Other environ outpsychological test functions are basisd on the anticipated light operating preconditions of the utensils in its final environ outment.
Outside Agencies Can Help
If purpursuiting or leasing hat ital utensils is not practsical, environ outpsychological testing labs may be an option. Many have to the full staffed reliability contrive jetengineers willing to agamble in contriveing the total stress test program. Third party advisors such as Nerac, whose analysts include jetengineers know-howd in the glebe of reliability stress testing, can agamble with assessing and take outing an efficacious and economicalal stress test program and can help find qualified labs to direct the tests.
Designing, take outing and directing efficacious stress testing can sometimes be as dauntying as the contrive of the system it is testing. It is composite, and many manutruthsurers oftentimes underevaluate or overlook the importance of such a program. This can cost dauriclely in fixing lacamp defects or funding endproduct or system recall costs down the route.
About Nerac
Nerac Inc. (http://www.nerac.com) is a universal resauriclech and advisory firm for companies progressing innovative endproducts and technologies. Nerac analysts express custom assessments of endproduct and technology evolution opportunities, competitor cleverness, noetic phawserrty strategies, and compliance requirements thharsh a prcooker badvanceed advance to custom analysis: reperspective of specialized scholarship, probe of noetic phawserrty, and appraisal of business influencess. Nerac deploys analysts in diverse diskiplines to help guests diskover trademark-new applications, serving as a catalyst for trademark-new thinmonarch and crnourishive advancees to business problems or identifying strategic growth opportunities.
##CONTINUE##The contrive is finished, the manutruthsuring transition is finish, pre-endproduction units have successto the full passed allowance testing, the endproduct is going out the threshold, and the contrive team is all perusey for the next scheme, right?
Not comparatively.
Article DescriptionThe contrive is finished, the manutruthsuring transition is finish, pre-endproduction units have successto the full passed allowance testing, the endproduct is going out the threshold, and the contrive team is all perusey for the next scheme, right?
Blog of the sourceal article and the translation / reperception of the article to follow non-mercantile use of crnourishive sharing allowment, you specify the composer, and the same source, non-mercantile use of the premise, I can not allow, is reyieldd directly.
No comments:
Post a Comment