RU2783770C1 - Method for conducting multifactor equivalent-cyclic tests - Google Patents

Abstract

FIELD: testing.

SUBSTANCE: invention relates to the field of tests for reliability of confirmation of reliability indicators of elements and devices of radio-electronic equipment. A method for conducting multifactor equivalent-cyclic tests is proposed, according to which simulation of multifactor equivalent-cyclic tests is performed, including determination of a stage of a life cycle by experience of analogue product operation, determination of main types of failures and their reasons inherent to each stage of the life cycle, based on a resource consumption model, test duration is determined for each stage of the life cycle under action of factors causing failure or a group of failures, as well as a number of cycles of action and duration of multifactor equivalent-cyclic tests are calculated, then, rejection tests are conducted with subsequent localization of a failed element provided that intensity of failures corresponds to Pearson’s consent criterion with

, after which multifactor equivalent-cyclic tests are conducted, based on which measures are developed for increasing the reliability of onboard radio-electronic equipment (OREE), depending on detected defect reasons, and their efficiency of reliability increasing is assessed by the ratio of operating time for malfunction after the measure is conducted to operating time for malfunction before the measure is conducted.

EFFECT: reduction in duration and increase in the efficiency of equivalent-cyclic tests in terms of detection of failures due to degradation processes of elements and device of radio-electronic equipment and products of electronic equipment for aviation purpose due to a synergetic effect of action of factors intensifying such failures.

1 cl, 1 dwg

Description

The invention relates to the field of reliability testing to confirm the reliability of elements and devices of radio electronic equipment.

Modern trends in the development of radio electronic equipment and electronic products are aimed at increasing the density of mounting of radio elements on modules, which requires the improvement of control methods that allow determining their reliability level in the shortest possible time, such as mean time between failures and assigned resource. In order to control the reliability indicators of elements and devices of radio-electronic equipment and electronic products, equivalent-cyclic tests are carried out, which have a long duration. For example, a 20,000 hour fail-safe rating typically requires at least 1,000 hours of equivalent cycle testing conducted to industry standards based on consistent simulation of external factors.

The most efficient equivalent-cyclic

tests are multifactorial tests, in which the largest number of influencing factors is simultaneously simulated. The duration of the ECI depends on the acceleration coefficient - a dimensionless parameter that indicates the ratio of the reliability parameter (mean time between failures) to the duration of the tests.

A known method of multivariate testing, described in patent No. 2027246 (MPK H01N 49/00 dated 04/01/1992), in which tests are carried out according to special plans - a complete factorial experiment (PFE), orthogonal and rotatable central compositional plans (OCCP and RCCP, respectively) and etc. In the course of such tests, the levels of influencing factors are set at experimental facilities in accordance with the plans (PFE, OCSF, RCSF) and the values of the parameters of the object under study are measured. The disadvantage of this method is the need for a sufficiently large number of tests to build regression models, which increases the duration and cost of testing, as well as the presence of output parameters for the product under study that uniquely determine the change in the state of operability, which for electronic equipment and electronic products for aviation purposes in the form of on-board electronic equipment is difficult to implement and requires expensive diagnostic tools.

Models are known based on the principle of independence of resource consumption: the reliability of the test object, provided that the physical pattern of failures is preserved, depends on the value of the share of the resource developed by it in the past, and does not depend on how this resource was developed (Sedyakin's principle):

The resource developed during the testing process must correspond to the resource developed during operation. In this case, the acceleration factor will be defined as [1]:

The most common approach in practical application was found by the multivariate test acceleration coefficient model, where it is assumed that all influencing factors affect all failure mechanisms equally, then we get [2]:

This model has found wide application in various types of accelerated tests, assuming that each applied stress increases the base failure rate, and the next applied stress increases the total failure rate increased by the previous factor, and so on. However, the main disadvantage of such a model is the overestimation of the impact of applied influencing factors, the lack of difference in the development of failure mechanisms, in which some impacts do not have an accelerating effect.

In the European standard IEC 62506 on accelerated test methods, a model of the impact of the i-th factor accelerating one of the j-failure mechanisms with an acceleration factor K _yij is given, while the generalized multifactor test acceleration factor [2]:

The disadvantage of this model is the absence of weight coefficients of the dominant failure mechanisms in the form of a relative

share of the probability of failure q _t under the influence of the i-th factor.

This disadvantage is eliminated by the model described in GOST R 57394-2017. Integrated microcircuits and semiconductor devices, however, it does not average the total impact of factors and does not take into account the stage of the product life cycle [3]:

где d_i - число отказов изделий, вследствие развития j-го механизма отказа; S - общее число отказов; λ_Иi - интенсивность отказов, вследствие развития j-го механизма отказа; аWherein,

where d _i is the number of product failures due to the development of the j-th failure mechanism; S is the total number of failures; λ _Иi - failure rate due to the development of the j-th failure mechanism; a

The problem to which this invention is directed is to reduce the duration and increase the efficiency of equivalent-cyclic tests to detect failures of radio-electronic equipment and electronic products by modeling optimal conditions for intensifying failures.

The technical result consists in reducing the duration and increasing the efficiency of equivalent-cyclic tests in terms of identifying failures due to degradation processes of elements and devices of radio-electronic equipment and electronic products for aviation purposes due to the synergistic effect of the impact of factors intensifying such failures.

The specified technical result is achieved due to the fact that in the course of modeling control multifactor tests for fail-safe operation, the product life cycle is divided into several stages with their own failure probability and the factors that have the greatest influence on the development of failure mechanisms at this stage, a fatigue accumulation model is determined that intensifies the main types failures, acceleration coefficients are calculated for each stage of the life cycle when exposed to factors causing a failure or a group of failures, the number of exposure cycles and the duration of multifactorial equivalent-cyclic tests are calculated, then preliminary studies are carried out, including screening tests and physical and technical analysis of failures, then multifactorial equivalent-cyclic tests, on the basis of which measures are determined and their effectiveness in improving reliability is evaluated.

The method of conducting multifactorial equivalent-cyclic tests on the example of on-board radio-electronic equipment begins with the definition of the stages of the product life cycle. To do this, according to the experience of operating products - analogues, there are stages that indicate the relationship between the processes of changing the state of the product. For avionics, this will be transportation - as a stage of simulating aircraft pre-flight conditions, functioning on board - as a stage of simulating aircraft flight conditions and persistence - as a stage of simulating post-flight conditions. Further, the main types of failures and their causes inherent in one or another stage of the life cycle are determined. According to the experience of operating analogue products, as well as according to the analysis carried out in accordance with the manual on methods for assessing the safety of systems and on-board equipment of civil aviation aircraft P4761, failures caused by thermal and mechanical energy factors total from 80% to 95%. After determining the main causes of failures, by selecting the most suitable failure physics model for these failures or based on parametric models, a damage accumulation model is determined that intensifies these failures. For avionics, the failure physics model is the resource consumption model, where the resource is represented as an integral of the entropy change: at λ = const dS/dt(t), the Sedyakin principle (1) is fulfilled. The resource depletion time can be found from the ratio S(t)=S*, where S* is the critical value of the accumulated entropy.

где d_i - число отказов изделий, вследствие развития j-го механизма отказа; S - общее число отказов; λ_Иi - интенсивность отказов, вследствие развития j-го механизма отказа; a

q_i = 1; P_k - вероятность отказа на М-стадии жизненного цикла.Where

where d _i is the number of product failures due to the development of the j-th failure mechanism; S is the total number of failures; λ _Иi - failure rate due to the development of the j-th failure mechanism; a

q _i = 1; P _k is the probability of failure at the M-stage of the life cycle.

Probability of failure throughout the life cycle

Determination of the MFECI acceleration coefficient is carried out according to the following formula:

, где доли отказов при воздействии механической и тепловой энергии обозначены как q_шсв и q_цит соответственно, а вероятности отказа на стадиях «Транспортирование», «Функционирование на борту ВС» и «Сохраняемость» обозначены как Р_Трансп, Р_Функц, Р_Сохр соответственно. Расчет количества циклов и их продолжительности производим для каждой стадии отдельно. Для БРЭО количество циклов на стадии «Функционирование на борту» рассчитывается по формуле:

, где N_вп - количество взлетов/посадок ВС за год, а продолжительность одного цикла по формуле:

где Т_МФЭЦИ - планируемая продолжительность МФЭЦИ. На стадии «Сохраняемость» количество циклов изменения температуры определяется по формуле:

где N_xi - среднемесячное количество колебаний суточной температуры в течении года, а продолжительность одного цикла воздействий изменения температуры рассчитывается по формуле: t_ц=τ_о+τ_н, где τ_н, τ_о - время достижения нагрева и охлаждения установившейся температуры в наиболее массивном элементе БРЭО соответственно.

, where the proportions of failures under the influence of mechanical and thermal energy are denoted as q _wsv and q _cyt , respectively, and the failure probabilities at the stages "Transportation", "Functioning on board the aircraft" and "Storability" are denoted as Р Transp , Р _{Functs ,} Р _Save , respectively. We calculate the number of cycles and their duration for each stage separately. For avionics, the number of cycles in the "Operating on board" stage is calculated by the formula:

, where N _vp - the number of takeoffs / landings of aircraft per year, and the duration of one cycle according to the formula:

where T _MFECI is the planned duration of the MFECI. At the "Storability" stage, the number of temperature change cycles is determined by the formula:

where _{N xi is the} average _monthly number _of daily temperature fluctuations throughout the year, and the duration of one cycle of temperature change effects is calculated by the formula: avionics element, respectively.

To confirm that the block design has the necessary strength and vibration resistance, vibration strength analysis is carried out with the determination of resonant frequencies using software packages of CAE systems (ANSYS, ABAQUS, NASTRAN, MARC, Asonika, etc.), as well as full-scale tests in accordance with the regulatory and technical documentation.

Next, you need to make sure that during the multivariate tests with the specified acceleration factors, the parameter of the failure flows will be constant. To do this, preliminary studies are carried out, the essence of which is to fix and then classify failures, as well as to determine their impact on the rate of change in the reliability parameter.

To determine the levels of influencing factors of multifactorial equivalent-cyclic tests for reliability, rejection tests are carried out according to the procedures and on the equipment HALT (High accelerated life (limited) test) consisting of:

- exposure to high and low temperatures, consisting of two stages, during which the temperature changes stepwise from the initial level to the limits at which the product cannot perform its functions;

- exposure to temperature cycling at a rate of 10°C/min to 60°C/min in the range of 10-20% below and above the operating limits;

- exposure to multiaxial broadband random vibration (hereinafter referred to as SHV) with a uniform spectrum without limiting the loading range;

- combined effects of temperature cycling (hereinafter referred to as CIT) and SHV.

In each procedure, the following levels of external influencing factors for the product are sequentially determined:

- The limit value of the low temperature at which the working capacity is maintained;

- The limit value of elevated temperature at which performance is maintained;

The limiting level of SHV at which operability is maintained;

- The limit value of the low temperature at which the performance is not restored (destruction);

- The limit value of elevated temperature at which performance is not restored (destruction);

- The limiting level of SHV, at which the performance is not restored (destruction).

All failures are fixed with a certain operating time (T ₁ ,T ₂ , ..., T _K ) and analyzed by the methods described in manual 4761 (FMEA, fault tree analysis (FTA), logic diagram analysis (DD), Markov analysis (MA) and etc.). For each failure, a decision is made, for example, with an assessment of the need for corrective actions. In case of critical defects, the product is removed from testing to eliminate comments and implement corrective actions, or change the requirements of scientific and technical documentation.

Physical and technical analysis of failures of the test object is carried out in the following order:

1) Performance monitoring in accordance with the scientific and technical documentation (NTD) of the product based on:

- checking the electrical circuit;

- checks of electrical resistance and electrical strength of insulation;

- checks at extreme values of supply voltage.

2) Localization of the failed module of the test object (structurally replaceable element).

3) Monitoring the operability of the module of the test object in order to localize the failed element (interconnection, connector, printed circuit board, ECB) and identify the structural and technological element that caused the failure.

If an intermittent failure is detected, the rejection tests are repeated according to the HALT procedures, setting the limiting exposure modes. In the absence of an unambiguous conclusion about the type and cause of the failure of the test object due to the intermittent nature of the failure, all elements of the chain that are potentially responsible for the occurrence of the recorded discrepancy are replaced.

4) If a failure of a structural and technological element (printed circuit boards, interconnections, connectors, etc.) is determined, the cause of its occurrence is determined.

After screening tests, the identified failures are classified according to the nature of the change in the reliability parameter over time. Due to the fact that the determining factor is the rate of change of the reliability parameter (instantaneous for sudden and constant for gradual ones), the establishment of which, for a number of objective reasons, is associated with long-term experiments, the following condition is adopted for classifying failures, where T ₀ is the guaranteed operating time, T _i - time of occurrence of the i-th failure:

- for sudden failures: T _i ≤0.1⋅T ₀ ;

- for gradual failures: T _i >0.1⋅T ₀ .

In FIG. 1 shows a curve of failure rate over time.

The lifetime of any product is usually divided into three stages. The first stage is the running-in period, characterized by a large number of failures caused by failures in the operation of defective elements. Since the defective elements fail one after another, the failure rate decreases rapidly and reaches a plateau by the end of the first stage. The second stage is characterized by the constancy of failures and has the lowest level of failure rate. At the end of the second stage, when wear begins to affect the product, the failure rate begins to increase rapidly and the third stage begins, characterized by an increase in the failure rate.

To determine the boundaries of these stages, the hypothesis of constancy of failures is tested. The hypothesis is tested in accordance with Pearson's goodness-of-fit criterion for the number of recorded sudden failures k during HALT procedures, while the number of sudden failures should be from 3 to 40. Pearson's goodness-of-fit criterion takes the following form:

where Т ₁ , Т ₂ , …, T _K - failure moments of the test object (or group of objects);

T _Σ =T _K - total time taken into account if a plan is used to limit the number of failures;

T _Σ \u003dT _K +T, where T is the operating time after the last considered failure, if a plan with a time limit is used;

d=k-1 - when limiting the number of failures,

d=k - plan with a time limit.

The calculated value of χ ² is compared with the theoretical values of P-quantiles [4]. two-sided confidence interval, which at a significance level of 10% for the number of degrees of freedom ν=2(k-1) [4].

лежит в пределах двустороннего интервала, то принимается гипотеза, что интенсивность отказов постоянна во времени, (λ=const»).If the calculated value

lies within the two-sided interval, then the hypothesis is accepted that the failure rate is constant in time, (λ=const”).

то интенсивность отказов вероятно возрастет и необходимо повторное выполнение отбраковочных испытаний.If the calculated value

then the failure rate is likely to increase and the rejection tests need to be repeated.

, то интенсивность отказов вероятно уменьшается и необходимо дальнейшее проведение отбраковочных испытаний с учетом предыдущих результатов, для определения момента времени, когда интенсивность отказов становится постоянной.If the calculated value

, then the failure rate is likely to decrease and further screening tests are necessary, taking into account previous results, to determine the point in time when the failure rate becomes constant.

After carrying out screening tests and confirming the hypothesis of constancy of failures, planning of multifactorial equivalent-cyclic tests is carried out with the calculated value of the acceleration coefficient according to model (1) by the sequential Wald method, which provides the average expected duration of tests with a limited number of tested samples.

According to the guidelines of GosNIIAS, the duration of the MFECI, taking into account the code of the test plan, selected in accordance with OST 1 01204-2012 to confirm a given level of reliability indicator, will be calculated using the following expressions:

Mismatch line:

In the absence of failures:

Match line:

продолжительность непрерывной работы БРЭО, заданной в техническом задании; T_HP - средняя наработка на отказ и повреждение, (при контрольных испытаниях по настоящей методике назначенный ресурс T_TP); N_МФЭЦИ - число циклов испытаний, b(b_yc) - нормированный коэффициент, n - число испытываемых образцов.where

the duration of continuous operation of the avionics specified in the terms of reference; T _HP - mean time between failures and damage, (during control tests according to this method, the assigned resource T _TP ); N _MFECI - number of test cycles, b(b _yc ) - normalized coefficient, n - number of tested samples.

For a visual display of the results obtained, according to the given values of T _α , T _β , α, β, a graphical representation of the operational characteristics of the MFECI is carried out.

Next, multifactorial equivalent-cyclic control tests for reliability are carried out for three stages of the life cycle "Transportation", "Functioning on board the aircraft" and "Storability" in the HALT chamber with simultaneous exposure to temperature cycling, broadband random vibration and electrical voltage in the previously defined exposure modes at the stage of preliminary research.

After confirming or not confirming the reliability indicator (mean time between failures), a decision is made to develop measures to improve the reliability of the product, depending on the identified causes of defects (for example, during the testing of the method, a group of failures was identified associated with a violation of the solder connection of microcircuits in BGA packages , to eliminate which special technological processes for their reballing were developed) and subsequent evaluation of their effectiveness in accordance with OCT B1 00094 using the efficiency factor (K _e ), which is determined by the ratio of the time to failure after the event (T _nnm ) to the time to failure before the event (T _dnm ):

K _E \u003d T _ppm / T _dpm

Thus, the specified method of conducting multifactorial equivalent-cyclic tests based on modeling the impact of many factors makes it possible to reduce the duration and increase the efficiency of testing elements and devices of electronic equipment for reliability.

Sources of information:

1. Sedyakin N.M. About one physical principle of the theory of reliability. Proceedings of the Academy of Sciences of the USSR. Technical cybernetics. Number 3. 1966

2. IEC 62506. Methods for product accelerated testing. Edition 1.0. - 2013. - 188 p.

3. GOST P 57394-2017. Integrated circuits and semiconductor devices. Test methods for reliability.

4. RD107.460000.010-89. Radioelectronic means. Methods of technological training. Date of introduction - 07/01/1990. GNTU MRP. - 1990. - 57 p.

Claims (5)

A method for conducting multi-factor equivalent-cyclic tests of aviation electronic equipment for reliability, according to which multi-factor equivalent-cyclic tests are simulated, including determining the stage of the life cycle based on the experience of operating analogue products, determining the main types of failures and their causes inherent in each stage of the life cycle, then, based on the resource consumption model, the duration of tests is determined for each stage of the life cycle under the influence of factors causing a failure or a group of failures, and the number of exposure cycles and the duration of multifactorial equivalent-cyclic tests are calculated, then rejection tests are carried out, followed by localization of the failed element, provided compliance of the failure rate with Pearson's goodness-of-fit criterion, expressed in terms of

, where Т ₁ , Т ₂ , …, T _K - failure moments of the test object (or group of objects); T _Σ =T _K - total time taken into account if a plan is used to limit the number of failures; d=k-1 - when limiting the number of failures, d=k - plan with a time limit; at

, after which multifactorial equivalent-cyclic tests are carried out, on the basis of which measures are developed to improve the reliability of avionics (on-board radio-electronic equipment of aircraft) depending on the identified causes of defects and evaluate their effectiveness in improving reliability by the ratio of the time between failures after the event to the time on failure before the event.

Images