CN106021783A - Multiple impact stress screening condition optimizing method and device - Google Patents

Abstract

The invention discloses a multiple impact stress screening condition optimizing method and device. The multiple-impact stress screening condition optimizing method comprises the steps that product reliability modeling is performed to obtain a working life distribution function of a zero-damage product and a working life distribution function which fails in the using process only due to impact stresses; multiple impact stress screening process modeling is performed according to the working life distribution function of the zero-damage product and the working life distribution function which fails in the using process only due to the impact stresses to obtain a normalized service life distribution function of an un-failed product after multiple impact stress screening; the normalized service life distribution function is optimized according to the expected life of the product to obtain optimized multiple impact stress screening conditions, wherein the screening conditions include screening times and screening stresses. The multiple impact stress screening condition optimizing method can conduct multiple impact stress screenings and can play a certain guiding effect on screening condition optimization of the multiple impact stress screenings.

Description

The optimization method of a kind of repeat impact stress screening condition and device

Technical field

The present invention relates to reliability engineering technique field, particularly to the optimization side of a kind of repeat impact stress screening condition Method and device.

Background technology

During product development, environmental stress screening test is that a kind of conventional rejecting product introduces in the fabrication process Potential defect, the effective means of damage.Such as, for aerospace electron product, before with satellite launch, often need to carry out device The environmental stress screenings at different levels such as level, plate level, unit level.Currently for choosing of environmental stress screening experimental condition, root often Determine according to engineering experience or all kinds of specification, seldom determine according to actual development, the applicable cases optimization of specific products.This ring The mode of choosing of border stress screening experimental condition there may be problems with: first, probably due to too high experimental condition causes There is defect, damage in the most intact product；Second, probably due to too much screening cause unnecessary testing expenses and time Between.

The research optimized currently, with respect to environmental stress screening Experiment Modeling and experimental condition is little, main in prior art For single-impact stress, carry out screening test condition to the research of the impact of life of product characteristic after test.Due to existing skill Art only accounts for the process of the test of single-impact stress screening, and is likely that there are to carry out in actual product screening process and repeatedly rushes Hit the situation of stress screening, therefore, it is necessary to propose a kind of screening conditions prioritization scheme based on repeat impact stress screening.

Summary of the invention

In view of the above problems, the invention provides optimization method and the device of a kind of repeat impact stress screening condition, with The problem solving prior art inapplicable repeat impact stress screening.

For reaching above-mentioned purpose, the technical scheme is that and be achieved in that:

On the one hand, the invention provides the optimization method of a kind of repeat impact stress screening experimental condition, described method bag Include:

Reliability modeling to product, obtains the working life distribution function of not damaged product, and in use Only cause the working life distribution function of the product lost efficacy because of impact stress；

Working life distribution function according to described not damaged product and described the most only lead because of impact stress Cause the working life distribution function of the product lost efficacy, to repeat impact stress screening process model building, obtain repeat impact stress sieve Do not lost efficacy after choosing the normalizing life distribution function of product；

Life expectancy according to product optimizes described normalizing life distribution function, the repeat impact stress sieve after being optimized Condition, described screening conditions is selected to include impact screening number of times and screening stress.

Preferably, the working life distribution function of described not damaged product is specially

Described the most only because of impact stress cause lost efficacy product working life distribution function be specially

Wherein, T_NFor the life-span of not damaged product, T_EFor the most only causing the product longevity of inefficacy because of impact stress Life, r (x) is the substrate crash rate of not damaged product, the accumulative number of shocks that N (t) is suffered when being working time t, W_iFor work During the lesion size that causes of i & lt impact stress, W_MFor the original damage size of product, D_wFailure threshold for product.

Preferably, the described working life distribution function according to described not damaged product and the most only because of impact Stress causes the working life distribution function of the product lost efficacy, and includes repeat impact stress screening process model building:

Product is carried out repeat impact stress screening, and the product by not losing efficacy after repeat impact stress screening is carried out Classification；Wherein, there is not damage after including initial not damaged and screening in the product by not losing efficacy after repeat impact stress screening Product, initial not damaged but there is the product of damage after screening and be initially present damage but by the product of screening；

Calculate every series products all by the product of repeat impact stress screening in shared ratio, and according to described The working life distribution function of not damaged product and the work of the described product the most only causing inefficacy because of impact stress Make life distribution function, calculate every series products life distribution function in screening process；

According to the ratio shared by every series products and the life distribution function in screening process thereof, set up repeat impact stress Screening process model.

Preferably, there is not the ratio shared by product of damage and in screening process after described initial not damaged and screening In life distribution function be followed successively by respectively:

,

$P(T_{ESS}>t|Z=1)=\exp (-{\∫}_0^{t}r(x\left)dx\right);$

Described initial not damaged but screening after there is the ratio shared by product of damage and the life-span in screening process thereof Distribution function is followed successively by:

,

$P(T_{ESS}>t|Z=\left(2,n\right))=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}P(\underset{i=0}{\overset{N\left(t\right)}{\mathrm{\&Sigma;}}}{W}_i+\underset{i=1}{\overset{N-n}{\mathrm{\&Sigma;}}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described being initially present is damaged but by the ratio shared by product of screening and the distribution of the life-span in screening process thereof Function is followed successively by:

,

$P(T_{ESS}>t|Z=3)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}P(\underset{i=0}{\overset{N\left(t\right)}{\mathrm{\&Sigma;}}}{W}_i+\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described repeat impact stress screening process model is

Wherein, π is the ratio of initial not damaged product, and s is the value of screening stress, and ρ (s) is initial not damaged but screening The ratio of the rear product that there is damage, N is the impact screening number of times of repeat impact stress screening, and n is initial undamaged product Produce damage when n-th impact stress is screened, M is initial not damaged but produces damage when n-th impact stress is screened Product number, W_siThe lesion size caused for i & lt impact stress in screening process, W'_MFor initial undamaged product because of punching Hit the lesion size first that stress screening causes, T_ESSFor the service life by the product after repeat impact stress screening.

Preferably, the described life expectancy according to product optimizes described normalizing life distribution function, many after being optimized Secondary Shocks stress screening condition includes:

According to formulaDo not lost efficacy after optimizing described repeat impact stress screening product The normalizing life distribution function of product；Or,

According to formulaDo not lose after optimizing described repeat impact stress screening The normalizing life distribution function of effect product；

Wherein, T_gLife expectancy for product.

On the other hand, the invention provides the optimization device of a kind of repeat impact stress screening experimental condition, described device Including:

Reliability modeling unit, for the Reliability modeling to product, obtains the working life distribution letter of not damaged product Number, and the most only cause the working life distribution function of the product lost efficacy because of impact stress；

Repeat impact stress screening process model building unit, for the working life distribution function according to described not damaged product The most only the working life distribution function of the product lost efficacy is caused, to repeat impact stress because of impact stress with described Screening process models, the normalizing life distribution function of the product that do not lost efficacy after obtaining repeat impact stress screening；

Screening conditions optimize unit, optimize described normalizing life distribution function for the life expectancy according to product, obtain Repeat impact stress screening condition after optimization, described screening conditions include impact screening number of times and screening stress.

Preferably, the working life distribution function of described not damaged product is specially

Described the most only because of impact stress cause lost efficacy product working life distribution function be specially

Wherein, T_NFor the life-span of not damaged product, T_EFor the most only causing the product longevity of inefficacy because of impact stress Life, r (x) is the substrate crash rate of not damaged product, the accumulative number of shocks that N (t) is suffered when being working time t, W_iFor work During the lesion size that causes of i & lt impact stress, W_MFor the original damage size of product, D_wFailure threshold for product.

Preferably, described repeat impact stress screening process model building unit includes:

Sort module, for product being carried out repeat impact stress screening, and to by after repeat impact stress screening not The product lost efficacy is classified；Wherein, the product by not losing efficacy after repeat impact stress screening includes initial not damaged and sieve Do not exist after choosing the product of damage, initial not damaged but there is the product of damage after screening and be initially present damage but by screening Product；

Computing module, for calculate every series products all by the product of repeat impact stress screening in shared ratio Example, and according to the working life distribution function of described not damaged product with described the most only cause because of impact stress The working life distribution function of the product lost efficacy, calculates every series products life distribution function in screening process；

Set up module, for according to the ratio shared by every series products and the life distribution function in screening process thereof, build Vertical repeat impact stress screening process model.

Preferably, there is not the ratio shared by product of damage and in screening process after described initial not damaged and screening In life distribution function be followed successively by respectively:

,

$P(T_{ESS}>t|Z=1)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right);$

Described initial not damaged but screening after there is the ratio shared by product of damage and the life-span in screening process thereof Distribution function is followed successively by:

,

$P(T_{ESS}>t|Z=\left(2,n\right))=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}P(\underset{i=0}{\overset{N\left(t\right)}{\mathrm{\&Sigma;}}}{W}_i+\underset{i=1}{\overset{N-n}{\mathrm{\&Sigma;}}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described being initially present is damaged but by the ratio shared by product of screening and the distribution of the life-span in screening process thereof Function is followed successively by:

,

$P(T_{ESS}>t|Z=3)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}P(\underset{i=0}{\overset{N\left(t\right)}{\mathrm{\&Sigma;}}}{W}_i+\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described repeat impact stress screening process model is

Wherein, π is the ratio of initial not damaged product, and s is the value of screening stress, and ρ (s) is initial not damaged but screening The ratio of the rear product that there is damage, N is the impact screening number of times of repeat impact stress screening, and n is initial undamaged product Produce damage when n-th impact stress is screened, M is initial not damaged but produces damage when n-th impact stress is screened Product number, W_siThe lesion size caused for i & lt impact stress in screening process, W'_MFor initial undamaged product because of punching Hit the lesion size first that stress screening causes, T_ESSFor the service life by the product after repeat impact stress screening.

Preferably, described screening conditions optimize unit, specifically for according to formula Do not lost efficacy after optimizing described repeat impact stress screening the normalizing life distribution function of product；Or, according to formulaDo not lost efficacy after optimizing described repeat impact stress screening normalizing life-span of product Distribution function；Wherein, T_gLife expectancy for product.

The embodiment of the present invention provides the benefit that: the present invention, by setting up the reliability model of product, respectively obtains lossless Injured labor product and the most only cause, because of impact stress, the working life distribution function of product that lost efficacy, and according to obtained The working life distribution function of two types product set up repeat impact stress screening process model, thus obtain repeat impact Do not lost efficacy after stress screening the distribution function in service life of product；Therefore, the screening technique of the present invention can be suitable for enter product The situation of row repeat impact stress screening；The present invention does not loses always according to after the life expectancy optimization repeat impact stress screening of product The normalizing life distribution function of effect product such that it is able to the repeat impact stress screening condition after being optimized, to said circumstances Certain directive function is played in the screening conditions optimization realizing repeat impact stress screening.

Accompanying drawing explanation

The optimization method flow chart of the repeat impact stress screening condition that Fig. 1 provides for embodiment one；

Fig. 2 necessarily screens in numbers range for what embodiment two provided, the product task longevity under the conditions of different impact stress The reliability curves figure at life end；

The optimization apparatus structure block diagram of the repeat impact stress screening condition that Fig. 3 provides for embodiment three.

Detailed description of the invention

For making the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention Formula is described in further detail.

Embodiment one:

Owing to prior art only accounts for the process of the test of single-impact stress screening, and in actual product screening process very There may be the situation carrying out repeat impact stress screening.For carrying out the situation of repeat impact stress screening, not only need to consider Impact stress size, on the impact of product reliability after screening, also needs to consider that test number (TN) is to the shadow of product reliability after screening Ring.In view of above-mentioned consideration, for the situation of repeat impact stress screening, the open a kind of repeat impact stress screening condition of the present invention Optimization method.

The optimization method flow chart of the repeat impact stress screening condition that Fig. 1 provides for the present embodiment, as it is shown in figure 1, should Method includes:

S110, the Reliability modeling to product, obtain the working life distribution function of not damaged product, and using Journey only causes the working life distribution function of the product lost efficacy because of impact stress.

In manufacture course of products, due to off-gauge material or other factors, may produce some exist latent defect, The product of damage (such as: micro-crack).These defective products, in use easily lose because of the damage of accumulation Effect, relative to there is not the product of latent defect, damage, just has a kind of extra failure mode because what accumulated damage lost efficacy Pattern, the product reliability model that therefore the present embodiment is set up distinguishes not damaged product and because impact stress causes the product of inefficacy Product, and obtain the working life distribution function of this two series products.

In the work process of product, it is frequently subjected to meet the impact of nonhomogeneous Poisson distribution under work condition environment, it is assumed that produce Accumulation number of shocks suffered in product working time t is N (t), and nonhomogeneous Poisson distribution intensity function is λ (t), there will be no latent It is denoted as T in defect, the product of damage, namely the life-span of lossless product_N, corresponding substrate crash rate is r (t).Further, this enforcement Official holiday is located in work condition environment, and impact stress can make the damaged products size that originally there is latent defect expand further until reaching To failure threshold D_w, and the product that originally there is not latent defect is insensitive to impact stress, i.e. will not be because of under work condition environment Impact stress cause the generation that damages accordingly, finally quit work because of other failure modes.

Based on foregoing description, working life distribution function the most following formula (1) institute of the not damaged product in the present embodiment Show:

$P(T_N>t)=\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)---\left(1\right)$

The substrate crash rate that r (x) is not damaged product in formula (1)；It should be noted that the base described in the present embodiment End crash rate is the crash rate not considering accumulated damage failure mode.

During product work time t, shown in the lesion size such as formula (2) of the product that originally there is latent defect:

$W\left(t\right)=\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+W_M---\left(2\right)$

W in formula (2)_iThe lesion size caused for i & lt impact stress in work process, W_MInitial damage for product Hinder size, usual W_MMuch smaller than W_iSize, thus formula (2) can be approximately:

$W\left(t\right)\&ap;\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i---\left(3\right)$

Can obtain according to formula (2) or (3), the most only cause the work of the product lost efficacy because of impact stress Life distribution function, concrete as shown in formula (4):

$\begin{array}{c}P(T_E>t)=P\left(W\right(t)<D_w)\\ =\underset{j=0}{\overset{\mathrm{\&infin;}}{\mathrm{\&Sigma;}}}P(\underset{i=0}{\overset{N\left(t\right)}{\mathrm{\&Sigma;}}}{W}_i<D_w)\&CenterDot;P\left(N\right(t)=j)\end{array}---\left(4\right)$

For ease of understand that the present embodiment formula (4) defines the most only because impact stress causes the product that lost efficacy The implication of working life distribution function, the present embodiment is illustrated by following specific implementation.

In this specific implementation, it is assumed that failure threshold D_wMeet the exponential that parameter is θ, W_iFor independent same distribution Stochastic variable time, above-mentioned formula (4) can be derived as further:

$P(T_E>t)=\exp \{-{\&Integral;}_0^t(1-M_W\left(-\mathrm{\&theta;}\right))\mathrm{\&lambda;}\left(x\right)dx\},t\&GreaterEqual;0---\left(5\right)$

W in formula (5)_W() is variable W_iMoment generating function.

For a collection of product newly developed, it is assumed that the ratio of no defective product is π, the ratio of the most defective product is 1-π, is denoted as T by the life-span of product, then the working life distribution function of product is:

$P(T>t)=\mathrm{\&pi;}\&CenterDot;\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)+(1-\mathrm{\&pi;})\&CenterDot;\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;P(T_E>t),t\&GreaterEqual;0---\left(6\right)$

Formula (6) defines the product randomly selected from a collection of product produced working life in use and divides Cloth function.

S120, according to the working life distribution function of not damaged product and the most only causes losing because of impact stress The working life distribution function of product of effect, to repeat impact stress screening process model building, obtains after repeat impact stress screening not The normalizing life distribution function of inefficacy product.

In this step " according to the working life distribution function of not damaged product with the most only because of impact stress Cause the working life distribution function of the product lost efficacy to repeat impact stress screening process model building " particularly as follows:

Product is carried out repeat impact stress screening, and the product by not losing efficacy after repeat impact stress screening is carried out Classification；Wherein, there is not damage after including initial not damaged and screening in the product by not losing efficacy after repeat impact stress screening Product, initial not damaged but there is the product of damage after screening and be initially present damage but by the product of screening；

Calculate every series products all by the product of repeat impact stress screening in shared ratio, and according to lossless The working life distribution function of injured labor product and the most only cause the working life of the product lost efficacy to be divided because of impact stress Cloth function, calculates every series products life distribution function in screening process；

According to the ratio shared by every series products and the life distribution function in screening process thereof, set up repeat impact stress Screening process model.

It should be noted that owing to when product is carried out repeat impact stress screening, above-mentioned three types can be obtained Do not lost efficacy the life distribution function of product, in order to obtain impact screening time optimal in repeat impact screening process in subsequent treatment Number and screening stress, the life distribution function of the product that do not lost efficacy of these three type is weighted summation and processes by the present embodiment, The life distribution function of product it is thus possible to do not lost efficacy after representing screening by the expression formula of a life distribution function；The i.e. present invention In normalizing life distribution function be diagrammatically only by with the life-span distribution of all kinds of products that do not lost efficacy after the statement screening of expression formula, not It is that the life distribution function to all kinds of products that do not lost efficacy is normalized.

The method for building up of the repeat impact stress screening process model for being more fully understood that in step S120, the present embodiment leads to Cross following implementation to illustrate.

In the impact stress screening test of this implementation, a magnitude is impact (the generally punching in work condition environment of s Hit the several times of stress intensity) put on each product participating in screening.

For original defective product, it is W that i & lt impact screening test is introduced into an extra size_siDamage, Its size increases and random increase with s.Additionally, the most intact product of a part creates scarce in impact stress screening test Fall into, the ratio of this portioned product is denoted as ρ (s), have 0≤ρ (s)≤1, and ρ (s) increases and random increase with s equally.Wherein, After impact stress screening test, the sample of inefficacy will be rejected, and the sample not lost efficacy then comes into operation.

This implementation is after terminating impact stress screening test, and the product not lost efficacy can be divided three classes: (1) initial nothing The product of damage is there is not after damage and screening；(2) initial not damaged but there is the product of damage after screening；(3) it is initially present Damage but by the product of screening.

The most then having, the ratio shared by product that there is not damage after the initial not damaged of first kind product and screening is:

For Equations of The Second Kind product, also can segment further: i.e. initial not damaged but produce when n-th impact screening The product of defect；Given M such product, then produce defect and eventually through n times impact screening when n-th impact screening The ratio of product be:

$M\&lsqb;{(1-\mathrm{\&rho;}(s\left)\right)}^{n-1}\mathrm{\&rho;}\left(s\right)\&rsqb;\mathrm{\&pi;}P(D_w>\underset{i=1}{\overset{N-n}{\&Sigma;}}{W}_{Si}+{W_M}^{\&prime;})---\left(8\right)$

Wherein, W_siThe lesion size caused for i & lt impact stress in screening process, W'_MFor initial undamaged product Because impact stress screens the lesion size first caused, usual W'_MRelative to W_SiThe least.

According to foregoing description, the initial not damaged of Equations of The Second Kind product but there is the ratio shared by product of damage after screening For:

3rd series products is initially present damage but by the ratio shared by the product of screening is:

Working life distribution function P (T according to the not damaged product obtained in step S110_N＞ t) and in the process of use In only cause, because of impact stress, the working life distribution function P (T of product that lost efficacy_E＞ t) above-mentioned three series products can be obtained exist Life distribution function in screening process.

The life-span distribution in screening process of the product of damage is there is not after the initial not damaged of first kind product and screening Function is:

$P(T_{ESS}>t|Z=1)=\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)---\left(11\right)$

The initial not damaged of Equations of The Second Kind product but n-th impact screening time produce defect product in screening process Life distribution function is:

$P(T_{ESS}>t|Z=\left(2,n\right))=\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N-n}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j)---\left(12\right)$

3rd series products is initially present damage but by the product of screening life distribution function in screening process is:

$P(T_{ESS}>t|Z=3)=\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j)---\left(13\right)$

According to the ratio shared by above-mentioned three series products and the life distribution function in screening process thereof, can obtain repeatedly Impact stress screening process model is:

Wherein, in formula (14)Special Not, when failure threshold D_wMeet the exponential that parameter is θ, W_iDuring for independent identically distributed stochastic variable, have:

T in above-mentioned formula_ESSFor the service life by the product after repeat impact stress screening, correspond to respectively One series products, Equations of The Second Kind product produce product and the 3rd series products of damage when accepting the screening of n-th impact stress.

S130, optimizes described normalizing life distribution function, the repeat impact after being optimized according to the life expectancy of product Stress screening condition, described screening conditions include impact screening number of times and screening stress.

For the product developed, the reliability of the life expectancy of product is an important expectation index, generally Wish the reliability of the life expectancy of improving product.

Step S130 is mainly by the normalizing longevity of the product that do not lost efficacy after following two schemes optimization repeat impact stress screening Life distribution function:

The optimization of screening conditions is carried out, to obtain the screening conditions of optimum according to following formula:

${(N,s)}^{*}=\underset{N,s}{\mathrm{argmax}}\&lsqb;P(T_{ESS}>T_g)\&rsqb;---\left(16\right)$

Or, carry out the optimization of screening conditions according to following formula, to obtain the screening conditions of optimum:

${(N,s)}^{*}=\underset{N,s}{\mathrm{argmax}}\&lsqb;{\&Integral;}_0^{T_g}P(T_{ESS}>x)dx\&rsqb;---\left(17\right)$

Wherein, T_gLife expectancy for product.

It should be noted that formula (16) and formula (17) typically can be investigated in impact screening by the method for simulation calculation Times N and the different reliability of product or the expectsted of working life in the scope possible for value s of screening stress, and finally choose The figure of merit.

The method of the present embodiment is particularly suited for the product of failure threshold index of coincidence distribution, and the present embodiment is produced by foundation The reliability model of product, respectively obtains not damaged product and the most only because impact stress causes the work of the product lost efficacy Make life distribution function, and set up repeat impact stress sieve according to the working life distribution function of obtained two types product Select process model, thus the distribution function in service life of the product that do not lost efficacy after obtaining repeat impact stress screening；Therefore, the present invention Screening technique can be suitable for the situation that product is carried out repeat impact stress screening；The present invention is always according to the life expectancy of product Optimize the normalizing life distribution function of the product that do not lost efficacy after repeat impact stress screening such that it is able to the repeatedly punching after being optimized Hitting stress screening condition, certain guidance work is played in the screening conditions optimization that said circumstances realizes repeat impact stress screening With.

Embodiment two

The optimization method of the repeat impact stress screening condition in embodiment one is suitable for and product is carried out repeat impact stress The screening scene of screening, is particularly suited for the product of failure threshold index of coincidence distribution.As accepted with the solder joint of micro-crack During external impact, crackle will gradually extend, the fracture failure when crack size reaches a threshold value, thus embodiment one In the optimization method of repeat impact stress screening condition can be used for the repeat impact stress screening of solder joint.

When to repeat impact stress screening process model building, it is assumed that substrate crash rate r (t)=0.1 of not damaged product, non- Homogeneous Poisson distribution intensity function λ (t)=1, failure threshold D_wParameter θ=1, task time τ=4.0, initial lossless injured labor Ratio π=0.7 of product, initial not damaged but there is ratio ρ (the s)=1-exp [-0.015s] of product of damage after screening, i-th Lesion size W that Secondary Shocks stress causes_iMeet the exponential that parameter is 3, then have W_iMoment generating functionAnd In order to portray damage W_siWith the relation of screening stress s, define W_siMeeting parameter isExponential, i.e.

It should be noted that the present embodiment chooses above-mentioned characterisitic parameter according to estimates according to product attribute and history cumulative number Numerical value.

According to above-mentioned parameter, Fig. 2 gives in certain screening numbers range, the product task under the conditions of different impact stress The reliability curves figure of life-span Mo.According to the result in Fig. 2, optimum option screening times N=6 and screening stress s=6.

Embodiment three:

The present embodiment uses the expectation design identical with embodiment one, it is provided that a kind of repeat impact stress screening condition Optimize device.

The optimization apparatus structure block diagram of the repeat impact stress screening experimental condition that Fig. 3 provides for the present embodiment, such as Fig. 3 institute Showing, this device includes:

Reliability modeling unit 31, for the Reliability modeling to product, obtains the working life distribution of not damaged product Function, and the most only cause the working life distribution function of the product lost efficacy because of impact stress.

Wherein, the working life distribution function of not damaged product is specially The most only the working life distribution function of the product lost efficacy is caused to be specially because of impact stress

T in above-mentioned formula_NFor the life-span of not damaged product, T_EFor the most only causing losing efficacy because of impact stress Life of product, r (x) is the substrate crash rate of not damaged product, the accumulative number of shocks that N (t) is suffered when being working time t, W_iThe lesion size caused for i & lt impact stress in work process, W_MFor the original damage size of product, D_wMistake for product Effect threshold value.

Repeat impact stress screening process model building unit 32, is distributed letter for the working life according to described not damaged product Number and the life distribution function of the described work product the most only causing inefficacy because of impact stress, should to repeat impact Power screening process models, the normalizing life distribution function of the product that do not lost efficacy after obtaining repeat impact stress screening.

In the implementation of the present embodiment, repeat impact stress screening process model building unit 32 includes:

Sort module, for product being carried out repeat impact stress screening, and to by after repeat impact stress screening not The product lost efficacy is classified；Wherein, the product by not losing efficacy after repeat impact stress screening includes initial not damaged and sieve Do not exist after choosing the product of damage, initial not damaged but there is the product of damage after screening and be initially present damage but by screening Product；

Computing module, for calculate every series products all by the product of repeat impact stress screening in shared ratio Example, and according to the working life distribution function of described not damaged product with described the most only cause because of impact stress The working life distribution function of the product lost efficacy, calculates every series products life distribution function in screening process；

Set up module, for according to the ratio shared by every series products and the life distribution function in screening process thereof, build Vertical repeat impact stress screening process model.

Wherein, there is not the ratio shared by product of damage and the longevity in screening process thereof after initial not damaged and screening Life distribution function is followed successively by respectively:

,

$P(T_{ESS}>t|Z=1)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right);$

Initial not damaged but there is the ratio shared by product of damage and the life-span distribution in screening process thereof after screening Function is followed successively by:

,

$P(T_{ESS}>t|Z=\left(2,n\right))=\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N-n}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j);$

It is initially present damage but by the ratio shared by product of screening and the life distribution function in screening process thereof It is followed successively by:

,

$P(T_{ESS}>t|Z=3)=\exp (-{\&Integral;}_0^{t}r\left(x\right)dx)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described repeat impact stress screening process model is

π in above-mentioned formula is the ratio of initial not damaged product, and s is the value of screening stress, and ρ (s) is the most lossless Hindering but there is the ratio of the product of damage after screening, N is the impact screening number of times of repeat impact stress screening, and n is the most lossless The product of wound produces damage when n-th impact stress is screened, and M is initial not damaged but produces when n-th impact stress is screened The product number of raw damage, W_siThe lesion size caused for i & lt impact stress in screening process, W'_MFor the most undamaged Product screens the lesion size first caused, T because of impact stress_ESSFor the use by the product after repeat impact stress screening Life-span.

Screening conditions optimize unit 33, optimize described normalizing life distribution function for the life expectancy according to product, Repeat impact stress screening condition after optimization, described screening conditions include impact screening number of times and screening stress.

In the implementation of the present embodiment, screening conditions optimize unit 33, specifically for according to formulaOptimize the normalizing life distribution function of the product that do not lost efficacy after repeat impact stress screening； Or, according to formulaDo not lost efficacy after optimizing repeat impact stress screening product Normalizing life distribution function；Wherein, T_gLife expectancy for product.

The specific works mode of each unit module of apparatus of the present invention embodiment may refer to the embodiment of the method for the present invention, Do not repeat them here.

In sum, the invention provides optimization method and the device of a kind of repeat impact stress screening condition, by building The reliability model of vertical product, respectively obtains not damaged product and the most only because impact stress causes the product of inefficacy Working life distribution function, and according to the working life distribution function of obtained two types product set up repeat impact should Power screening process model, thus the distribution function in service life of the product that do not lost efficacy after obtaining repeat impact stress screening；Therefore, originally The screening technique of invention can be suitable for the situation that product carries out repeat impact stress screening；The present invention is always according to the expectation of product Life-span optimizes the normalizing life distribution function of the product that do not lost efficacy after repeat impact stress screening such that it is able to many after being optimized Secondary Shocks stress screening condition, certain guidance is played in the screening conditions optimization that said circumstances realizes repeat impact stress screening Effect.

The foregoing is only presently preferred embodiments of the present invention, be not intended to limit protection scope of the present invention.All Any modification, equivalent substitution and improvement etc. made within the spirit and principles in the present invention, are all contained in protection scope of the present invention In.

Claims (10)

1. the optimization method of a repeat impact stress screening condition, it is characterised in that described method includes:

Reliability modeling to product, obtains the working life distribution function of not damaged product, and the most only because of Impact stress causes the working life distribution function of the product lost efficacy；

Working life distribution function according to described not damaged product and described the most only cause losing because of impact stress The working life distribution function of the product of effect, to repeat impact stress screening process model building, after obtaining repeat impact stress screening Do not lost efficacy the normalizing life distribution function of product；

Life expectancy according to product optimizes described normalizing life distribution function, the repeat impact stress screening bar after being optimized Part, described screening conditions include impact screening number of times and screening stress.

Method the most according to claim 1, it is characterised in that

The working life distribution function of described not damaged product is specially

Described the most only because of impact stress cause lost efficacy product working life distribution function be specially

Wherein, T_NFor the life-span of not damaged product, T_EFor the most only causing the life of product of inefficacy, r because of impact stress X () is the substrate crash rate of not damaged product, the accumulative number of shocks that N (t) is suffered when being working time t, W_iFor work process The lesion size that middle i & lt impact stress causes, W_MFor the original damage size of product, D_wFailure threshold for product.

Method the most according to claim 2, it is characterised in that the described working life according to described not damaged product is distributed Function and the most only cause the working life distribution function of product that lost efficacy, to repeat impact stress because of impact stress Screening process modeling includes:

Product is carried out repeat impact stress screening, and the product by not losing efficacy after repeat impact stress screening is carried out point Class；Wherein, the product by not losing efficacy after repeat impact stress screening includes: there is not damage after initial not damaged and screening Product, initial not damaged but there is the product of damage after screening and be initially present damage but by the product of screening；

Calculate every series products all by the product of repeat impact stress screening in shared ratio, and according to described lossless The life distribution function of injured labor product and described the most only cause the life-span distribution letter of the product lost efficacy because of impact stress Number, calculates every series products life distribution function in screening process；

According to the ratio shared by every series products and the life distribution function in screening process thereof, set up repeat impact stress screening Process model.

Method the most according to claim 3, it is characterised in that

There is not the ratio shared by product of damage after described initial not damaged and screening and the life-span in screening process divides Cloth function is followed successively by respectively:

$P(T_{ESS}>t|Z=1)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right);$

Described initial not damaged but screening after exist damage the ratio shared by product and in screening process life-span distribution Function is followed successively by:

,

$P(T_{ESS}>t|Z=\left(2,n\right))=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N-n}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described being initially present is damaged but by the ratio shared by product of screening and the life distribution function in screening process thereof It is followed successively by:

$P(T_{ESS}>t|Z=3)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described repeat impact stress screening process model is

Wherein, π is the ratio of initial not damaged product, and s is the value of screening stress, and ρ (s) is initial not damaged but deposits after screening In the ratio of the product of damage, N is the impact screening number of times of repeat impact stress screening, and n is that initial undamaged product is n-th Producing damage during Secondary Shocks stress screening, M is initial not damaged but produces the product of damage when n-th impact stress is screened Number, W_siThe lesion size caused for i & lt impact stress in screening process, W'_MFor initial undamaged product because of impact stress The lesion size first that screening causes, T_ESSFor the service life by the product after repeat impact stress screening.

Method the most according to claim 4, it is characterised in that the described life expectancy according to product optimizes the described normalizing longevity Life distribution function, the repeat impact stress screening condition after being optimized includes:

According to formulaDo not lost efficacy after optimizing described repeat impact stress screening product Normalizing life distribution function；Or,

According to formulaDo not lost efficacy after optimizing described repeat impact stress screening product The normalizing life distribution function of product；

Wherein, T_gLife expectancy for product.

6. the optimization device of a repeat impact stress screening experimental condition, it is characterised in that described device includes:

Reliability modeling unit, for the Reliability modeling to product, obtains the working life distribution function of not damaged product, with And the most only cause the working life distribution function of the product lost efficacy because of impact stress；

Repeat impact stress screening process model building unit, for the working life distribution function according to described not damaged product and institute State the most only because impact stress causes the working life distribution function of the product lost efficacy, to repeat impact stress screening Process model building, the normalizing life distribution function of the product that do not lost efficacy after obtaining repeat impact stress screening；

Screening conditions optimize unit, optimize described normalizing life distribution function for the life expectancy according to product, are optimized After repeat impact stress screening condition, described screening conditions include impact screening number of times and screening stress.

Device the most according to claim 6, it is characterised in that the working life distribution function of described not damaged product is concrete For

Described the most only because of impact stress cause lost efficacy product working life distribution function be specially

Wherein, T_NFor the life-span of not damaged product, T_EFor the most only causing the life of product of inefficacy, r because of impact stress X () is the substrate crash rate of not damaged product, the accumulative number of shocks that N (t) is suffered when being working time t, W_iFor work process The lesion size that middle i & lt impact stress causes, W_MFor the original damage size of product, D_wFailure threshold for product.

Device the most according to claim 7, it is characterised in that described repeat impact stress screening process model building unit bag Include:

Sort module, for carrying out repeat impact stress screening, and to by not losing efficacy after repeat impact stress screening to product Product classify；Wherein, after the product by not losing efficacy after repeat impact stress screening includes initial not damaged and screening Do not exist the product of damage, initial not damaged but there is the product of damage after screening and be initially present damage but by the product of screening Product；

Computing module, for calculate every series products all by the product of repeat impact stress screening in shared ratio, with And according to the working life distribution function of described not damaged product with described the most only cause losing efficacy because of impact stress The working life distribution function of product, calculate every series products life distribution function in screening process；

Setting up module, for according to the ratio shared by every series products and the life distribution function in screening process thereof, foundation is many Secondary Shocks stress screening process model.

Device the most according to claim 8, it is characterised in that there is not the product of damage after described initial not damaged and screening Ratio shared by product and the life distribution function in screening process thereof are followed successively by respectively:

,

$P(T_{ESS}>t|Z=1)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right);$

Described initial not damaged but screening after exist damage the ratio shared by product and in screening process life-span distribution Function is followed successively by:

,

$P(T_{ESS}>t|Z=\left(2,n\right))=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N-n}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described being initially present is damaged but by the ratio shared by product of screening and the life distribution function in screening process thereof It is followed successively by:

,

$P(T_{ESS}>t|Z=3)=\exp (-{\&Integral;}_0^{t}r(x\left)dx\right)\&CenterDot;\underset{j=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}P(\underset{i=0}{\overset{N\left(t\right)}{\&Sigma;}}{W}_i+\underset{i=1}{\overset{N}{\&Sigma;}}{W}_{si}<D_w)P\left(N\right(t)=j);$

Described repeat impact stress screening process model is

Wherein, π is the ratio of initial not damaged product, and s is the value of screening stress, and ρ (s) is initial not damaged but deposits after screening In the ratio of the product of damage, N is the impact screening number of times of repeat impact stress screening, and n is that initial undamaged product is n-th Producing damage during Secondary Shocks stress screening, M is initial not damaged but produces the product of damage when n-th impact stress is screened Number, W_siThe lesion size caused for i & lt impact stress in screening process, W'_MFor initial undamaged product because of impact stress The lesion size first that screening causes, T_ESSFor the service life by the product after repeat impact stress screening.

Device the most according to claim 9, it is characterised in that described screening conditions optimize unit, specifically for according to public affairs FormulaDo not lost efficacy after optimizing described repeat impact stress screening normalizing life-span of product divides Cloth function；Or, according to formulaAfter optimizing described repeat impact stress screening Do not lost efficacy the normalizing life distribution function of product；Wherein, T_gLife expectancy for product.