A manufacturing process defect analysis method based on aircraft product reliability

Abstract

The present invention provides a manufacturing process defect analysis method based on aircraft product reliability, belonging to the field of aircraft product reliability. The method specifically includes the following steps: S1. Determining defects occurring during the aircraft product manufacturing process; S2. Obtaining a design and use phase failure rate; S3. Obtaining, based on component-level defects, the product reliability and first early failure rate affected by component-level defects; S4. Obtaining, based on system-level defects, the product reliability and second early failure rate affected only by system-level defects; S5. Obtaining the use phase reliability and use phase failure rate; and S6. Determining a manufacturing process that requires control based on the degree of defect impact. The present invention is applicable to quantitative defect data analysis during aircraft product manufacturing, characterizing manufacturing process defect characteristics and reliability indicators, and specifically proposing engineering design and process control requirements to address the reliability reduction problem caused by the aircraft product manufacturing process.

Description

Technical Field

[0001] The present invention belongs to the field of aircraft product reliability, and in particular relates to a manufacturing process defect analysis method based on aircraft product reliability. Background Art

[0002] In modeling the relationship between manufacturing defects and aircraft product reliability, using aircraft after-sales maintenance data, methods have been proposed, including establishing an aircraft product reliability model that incorporates quality deviations based on assembly errors and component failures during the manufacturing process; using state-space models to explain the degradation mechanism of aircraft product reliability during system manufacturing; and using machine learning algorithms to assess and provide early warnings for mechanical aircraft product after-sales quality losses. The relationship between manufacturing defects and aircraft product early failure rates has been widely applied in semiconductor manufacturing. By analyzing the growth characteristics and clustering effects of related defects during the aircraft manufacturing phase and studying their impact on burn-in experiments and failure rate curves, models have been developed to establish an aircraft product reliability function that accounts for manufacturing defects, as well as models that relate time-varying yield and early reliability. Common defect mathematical models include the Poisson distribution and the negative binomial distribution. Furthermore, some have suggested that incorporating expert opinion and failure data from similar aircraft products could increase the data sources for assessment and further improve the accuracy of aircraft product reliability predictions.

[0003] The above background analysis shows that current research focuses more on the qualitative relationship between batch rejection rate and batch reliability of aircraft products, while less consideration is given to the correlation between relevant parameters in the manufacturing process and the degradation of aircraft product reliability during use. This leads to the neglect of the objective fact that qualified aircraft products can still experience reliability degradation, and thus the inability to accurately assess the defects in the corresponding manufacturing process, affecting the accuracy of aircraft product manufacturing system status analysis and the targetedness of manufacturing process quality control. Therefore, in order to analyze the source of manufacturing process defects and ensure aircraft product reliability, the establishment of a quantitative model that links aircraft product manufacturing process defects with aircraft product quality and supports the assessment and control of aircraft product manufacturing process defects is the focus of subsequent research. Summary of the Invention

[0004] In order to address the deficiencies of the above-mentioned prior art, the present invention addresses the above-mentioned problems and provides a manufacturing process defect analysis method based on aircraft product reliability. Combined with field failure data of aircraft products, the present manufacturing process defects are sorted out and analyzed for relevant processes and the key characteristics processed by them; the impact of manufacturing process defects is characterized by the failure rate during the aircraft product use phase, and a degradation mechanism of the design reliability of aircraft products in the manufacturing process is constructed; based on the defect distribution per unit area, the impact of component-level defects in the manufacturing process on the early failure rate of aircraft products during the use phase is quantified; based on the Weibull distribution, the impact of system-level defects in the manufacturing process on the early failure rate of aircraft products during the use phase is quantified; based on the probability of occurrence and effect of each type of defect, the failure rate of aircraft products during the use phase affected by the integration of manufacturing process defects is calculated; based on the failure rates caused by the two types of defects, the impact of the manufacturing process defects of the process on the reliability of the aircraft product is analyzed by taking the difference in the failure rates of the aircraft products during the warranty period.

[0005] The present invention provides a manufacturing process defect analysis method based on aircraft product reliability, which includes the following steps:

[0006] S1. Identify defects that occur during the manufacturing process of aircraft products;

[0007] Determine the processing sequence, process type, and resulting defects of each manufacturing process in the manufacturing process based on the aircraft product manufacturing process;

[0008] S2. Obtain the failure rate γ0 during the design use phase;

[0009] According to the manufacturing process determined by S1, for each manufacturing process, if the aircraft product is processed by the ideal manufacturing process, its design reliability R0 is:

[0010] R0(t usage )=e ∧ (-γt usage )=exp(-γt usage ) (1)

[0011] Among them, t usage is the warranty period of the aircraft product, γ is the failure rate of a single manufacturing process during design;

[0012] According to the design reliability R0 of the manufacturing process, the design failure rate γ of the manufacturing process can be obtained 0,i :

[0013]

[0014] S3. Obtaining product reliability and first early failure rate affected by component-level defects based on component-level defects;

[0015] Assuming that the total number of component-level defects per unit area of ​​aircraft products, N, follows a Poisson distribution, the product reliability R1 affected by component-level defects is:

[0016] R1(t usage ; t) = (1-q t )R0(t usage )+q t R c (t usage ;t)(4)

[0017] Among them, R1 is the product reliability affected by component-level defects, q t is the probability of component-level defects, R0 is the design reliability, R c is the product reliability affected by component-level defect activation, t is the service life of the aircraft product, and t usage It is the warranty period of the aircraft product;

[0018] According to the product reliability affected by component-level defects, the first early failure rate γ1 is obtained as:

[0019]

[0020] S4. Obtain the product reliability and the second early failure rate affected only by the system-level defects based on the system-level defects;

[0021] Assuming that the failure distribution of system-level defects is Weibull distribution, the product reliability R2 affected only by system-level defects is:

[0022] R2(t usage ; t) = R0(t usage )·(1-q St R s )(9)

[0023] Among them, R0 is the design reliability, R s To couple the impact of component-level defects on product reliability, q St is the probability of component-level defects;

[0024] According to the product reliability R2 affected only by system-level defects, the second early failure rate γ2 is obtained as follows:

[0025]

[0026] S5. Obtaining the reliability and failure rate during the use phase;

[0027] The probability of component-level defects and system-level defects occurring during processing is q t and q St The combination of {q t ,q St}, then the usage phase reliability R3 is:

[0028] R3(t usage ; t) = [(1-q t )R0(t usage )+q t R c (t usage ;t)]·[1-q St (1-F w (t usage ,t;ω))](11).

[0029] According to the reliability R3 of the use phase, the failure rate γ3 of the use phase is obtained as follows:

[0030]

[0031] S6. Determine the manufacturing process that needs to be controlled based on the impact of the defect;

[0032] The defect impact expression is as follows:

[0033]

[0034] The defects of different manufacturing processes are ranked according to their impact. The greater the defect impact, the more control is needed for the corresponding manufacturing process.

[0035] Preferably, the step S3 of obtaining the product reliability and the first early failure rate affected by the component-level defects according to the component-level defects is specifically as follows:

[0036] S31. Determine component-level defect distribution;

[0037] In a single manufacturing process, the total number of component-level defects per unit area of ​​an aircraft product, N, follows a Poisson distribution F1(t,λ), where t is time, λ is the variance of component-level defects, and component-level defects are categorized as fatal and non-fatal. The probability of being classified as a non-fatal component-level defect is τ.

[0038] S32. Determine product reliability affected by component-level defects;

[0039] Non-fatal component-level defects in t usage The probability p of being activated as a fault acti For p acti =1-exp(-t usage ), when only considering the impact of non-fatal component-level defects on aircraft product reliability, the product reliability affected by component-level defect activation can be obtained as R c for:

[0040] R c (tusage ; t) = E(1-p acti τ ) N =exp(-λτp acti )(3)

[0041] Where λ is the Poisson distribution parameter of the total number of component-level defects, τ is the probability of a component-level defect being classified as a non-fatal component-level defect, and p acti For non-fatal component-level defects, usage The probability of being activated into a fault, E() is the mathematical expectation;

[0042] Furthermore, consider the probability of non-fatal component-level defects to be q t (0≤q t ≤1), and obtain the product reliability R1 affected by component-level defects;

[0043] S33. Determine a first early failure rate;

[0044] The first early failure rate γ1 is obtained according to the product reliability affected by component-level defects.

[0045] Preferably, the fatal component-level defects are screened and eliminated through quality inspection during the manufacturing process.

[0046] Preferably, the step S4 of obtaining the product reliability and the second early failure rate affected only by the system-level defects according to the system-level defects is specifically as follows:

[0047] S41. Determine system-level defect distribution;

[0048] Use Weibull distribution to characterize system-level defect failure distribution:

[0049] F2(t usage ;α,β)=1-exp[-(t / α) β ](6)

[0050] Where F2 is the system-level defect failure distribution, α is the scale parameter, and β is the shape parameter;

[0051] S42. Determine the distribution of system-level defect failures after correction;

[0052] There is a positive correlation coupling between system-level defects and component-level defects, and the corrected system-level defect failure distribution F is obtained. ω :

[0053] F ω (F2(t usage ),F1;ω)=F2×F1×[1+ω(1-F2)(1-F1)] (7)

[0054] Where ω is the coupling parameter, which indicates the degree of correlation between the two distributions and has a value range of (0, 1);

[0055] S43. Obtain product reliability that is only affected by system-level defects;

[0056] According to the corrected system-level defect failure distribution F ω , we can get the product reliability R affected by coupling component-level defects s for:

[0057] R s (t usage ; t) = 1-F w (t usage ,t;ω)(8);

[0058] The probability of system-level defects is q St (0≤q St ≤1), the aircraft product reliability with system-level defects exists in a competing risk model, and the product reliability R2 is only affected by system-level defects;

[0059] S44. determining a second early failure rate;

[0060] The second early failure rate γ2 is obtained based on the product reliability R2 affected only by system-level defects.

[0061] Preferably, the aircraft product is a ferrite phase shift unit of an aircraft radar array.

[0062] Preferably, the manufacturing process includes 5 steps, namely: the first process type is cutting, and the defects produced are cracks and overcutting; the second process type is grinding the outer circle, and the defects produced are taper and burns; the third process type is grinding the end face, and the defects produced are vibration marks and protrusions; the fourth process type is welding, and the defects produced are shrinkage and biting; the fifth process type is winding, and the defects produced are gap defects.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] (1) The manufacturing process defect analysis method based on aircraft product reliability of the present invention can be applied to quantitative defect data analysis in the aircraft product manufacturing process, can characterize the manufacturing process defect characteristics and reliability indicators, control the process in a targeted manner, and solve the reliability reduction problem caused by the aircraft product manufacturing process.

[0065] (2) The manufacturing process defect analysis method based on aircraft product reliability of the present invention has broad application prospects and can be used in process design, preparation of process regulations, improvement of process, optimization of process systems, etc. It is of great significance to improve the manufacturing process capability of aircraft products and the consistency and stability of aircraft product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 Schematic diagram of the process of analyzing defects in the manufacturing process based on aircraft product reliability according to the present invention;

[0067] Figure 2 Schematic diagram of the degradation mechanism of design reliability during the manufacturing process of the aircraft product of the present invention. DETAILED DESCRIPTION

[0068] To fully describe the technical content, objectives and effects of the present invention, the following will be described in detail with reference to the accompanying drawings.

[0069] The present invention provides a manufacturing process defect analysis method based on aircraft product reliability, based on the following theory: the degradation mechanism of aircraft product design reliability in the manufacturing process is as follows: Figure 2 As shown in the figure, according to the composition structure of aircraft products, the manufacturing process defects of a single manufacturing process are divided into component-level defects and system-level defects. The two types of defects have different mechanisms of action on the reliability of aircraft products and there is a certain coupling relationship between the two types of defects. Let the probability of component-level defects be q t , the probability of system-level defects is q St , the failure rate γ3 of aircraft products in the service phase can be obtained based on component-level defects and system-level defects. Figure 2 Here, R0 represents the design reliability of the aircraft product after undergoing an ideal manufacturing process, R1 represents the product reliability affected by component-level defects, considering component-level defects, R2 represents the product reliability affected only by system-level defects, considering system-level defects, and R3 represents the service-phase reliability affected by the integration of manufacturing process defects. Based on the degradation mechanism of the design reliability of aircraft products during manufacturing, we first determine the design-phase failure rate γ0 based on the design of the aircraft product after undergoing an ideal manufacturing process. We then determine the service-phase failure rate γ3 based on the component-level defects and system-level defects of each manufacturing process. The difference between the service-phase failure rate γ3 and the design failure rate γ0 can be used to determine whether the manufacturing process defects generated by the corresponding manufacturing process require further control.

[0070] The present invention provides a manufacturing process defect analysis method based on aircraft product reliability, such as Figure 1 As shown, the method includes the steps of:

[0071] S1. Identify defects that occur during the manufacturing process of aircraft products.

[0072] We analyze manufacturing process defects to identify the specific manufacturing processes that lead to aircraft product failures, including the manufacturing procedures, process types, process parameters, process equipment, and steps that directly lead to the failures, as well as the key aircraft product characteristics processed in each process link. Manufacturing process defects are undesirable characteristics introduced by achieving key aircraft product reliability characteristics and are inherent characteristics of the manufacturing process output.

[0073] By sorting out and analyzing the design requirements of aircraft products, the specific manufacturing process of aircraft products is obtained, and the processing sequence and process types involved in the manufacturing process as well as the possible defects are determined. The manufacturing processes in this application all refer to the important and key manufacturing processes in the manufacturing process of aircraft products, and the processing sequence is also sorted according to the important and key manufacturing processes, rather than only containing these manufacturing processes in actual manufacturing. In addition, this patent only focuses on defects such as inclusions, cracks, and pores on the surface and inside of aircraft products, and does not involve defect types that can be measured as dimensional deviations and directly scrapped. These defects such as inclusions, cracks, and pores are difficult to detect and discover in a timely manner during the manufacturing process. The number and location are randomly generated according to the equipment status. Their impact is reflected in the service and use stage of aircraft products, making the frequency of aircraft product quality problems and the severity of failures exceed expectations.

[0074] S2. Obtain the failure rate γ0 during the design usage phase.

[0075] According to the manufacturing process determined by S1, for a single manufacturing process, if the aircraft product is processed by the ideal manufacturing process, its design reliability R0 is:

[0076] R0(t usage )=e∧(-γt usage )=exp(-γt usage ) (1)

[0077] Among them, t usage is the warranty period of the aircraft product, and γ is the failure rate of a single manufacturing process during design. Both parameters are constants. In actual operation, sometimes the failure rate of a single manufacturing process is not involved in the design, but it can be derived based on reliability analysis, function fitting software, and empirical values.

[0078] According to the design reliability R0 of the manufacturing process, the design failure rate γ of the manufacturing process can be obtained 0,i :

[0079]

[0080] S3. Obtain product reliability and first early failure rate affected by component-level defects based on component-level defects.

[0081] The manufacturing process defects of a single manufacturing process include component-level defects and system-level defects. Component-level defects include cracks, burns, etc., while system-level defects include fit and alignment, etc. The two types of defects have different mechanisms of effect on the reliability of aircraft products, and there is a certain coupling relationship between the two types of defects.

[0082] S31. Determine component-level defect distribution.

[0083] In a single manufacturing process, the total number N of component-level defects per unit area of ​​an aircraft product follows a Poisson distribution F1(t,λ), where t is time and λ is the variance of component-level defects. The probability of a component-level defect being classified as non-fatal is τ. Fatal component-level defects can be promptly screened and eliminated by quality inspections during the manufacturing process. Therefore, when considering component-level defects in this application, non-fatal component-level defects are referred to.

[0084] S32. Determine product reliability affected by component-level defects.

[0085] Non-fatal component-level defects in t usage The probability p of being activated as a fault acti For p acti =1-exp(-t usage ), therefore, when only considering the impact of non-fatal component-level defects on aircraft product reliability, the product reliability affected by component-level defect activation can be obtained as R c for:

[0086] R c (t usage ; t) = E(1-p acti τ ) N =exp(-λτp acti )(3)

[0087] Where λ is the Poisson distribution parameter of the total number of component-level defects, τ is the probability of a component-level defect being classified as a non-fatal component-level defect, and p acti For non-fatal component-level defects, usage The probability of being activated into a fault, E() is the mathematical expectation.

[0088] Furthermore, consider the probability of non-fatal component-level defects to be q t (0≤q t ≤1), the product reliability R1 affected by component-level defects is obtained as:

[0089] R1(t usage ; t) = (1-q t )R0(t usage )+q t Rc (t usage ;t)(4)

[0090] Among them, R1 is the product reliability affected by component-level defects, q t is the probability of component-level defects, which is determined by the processing status of the equipment, R0 is the design reliability, and R c is the product reliability affected by component-level defect activation, t is the service life of the aircraft product, and t usage It is the warranty period of aircraft products.

[0091] S33. Determine a first early failure rate.

[0092] According to the product reliability affected by component-level defects, the first early failure rate γ1 is obtained as:

[0093]

[0094] S4. Obtain the product reliability and the second early failure rate that are only affected by the system-level defects based on the system-level defects.

[0095] S41. Determine system-level defect distribution.

[0096] Compared to component-level defects, system-level defects have a more direct negative impact on aircraft product performance, but system-level defects are non-fatal. The failure distribution caused by system-level defects is called the system-level defect failure distribution. In this example, a two-parameter Weibull distribution is used to characterize the system-level defect failure distribution:

[0097] F2(t usage ;α,β)=1-exp[-(t / α) β ](6)

[0098] Where F2 is the system-level defect failure distribution, α is the scale parameter, and β is the shape parameter.

[0099] S42. Determine the distribution of system-level defect failures after correction.

[0100] Since the system-level processing of aircraft products is after the component-level processing, the positive correlation coupling between system-level defects and component-level defects must also be considered. The commonly used Farlie–Gumbel–Morgenstern binary distribution is selected to correct the system-level defect failure distribution, and the corrected system-level defect failure distribution F is obtained. ω :

[0101] F ω (F2(t usage ),F1;ω)=F2×F1×[1+ω(1-F2)(1-F1)](7)

[0102] Where ω is the coupling parameter, which indicates the degree of correlation between the two distributions. Its value range is (0, 1). The larger the value of ω, the weaker the correlation. The specific value is determined by experience.

[0103] S43. Obtain product reliability that is only affected by system-level defects.

[0104] According to the corrected system-level defect failure distribution F ω , we can get the product reliability R affected by coupling component-level defects s for:

[0105] R s (t usage ; t) = 1-F w (t usage ,t;ω)(8).

[0106] Product reliability R affected by coupled component-level defects s Aircraft product reliability is only considered when the positive correlation coupling effect between system-level defects and component-level defects is considered.

[0107] Furthermore, considering the probability of system-level defects to be q St (0≤q St ≤1), q St The value is also determined by the processing status of the equipment, and the reliability of aircraft products with system-level defects usually exists in a competing risk mode. The product reliability R2 affected only by system-level defects is:

[0108] R2(t usage ; t) = R0(t usage )·(1-q St R s )(9).

[0109] S44. Determine the second early failure rate.

[0110] According to the product reliability R2 affected only by system-level defects, the second early failure rate γ2 is obtained as follows:

[0111]

[0112] S5. Obtain the reliability and failure rate during the use phase.

[0113] Affected by the state of the processing equipment during the manufacturing process, the probability of component-level defects and system-level defects occurring during processing is q t and q St The combination of {q t ,q St}, then the usage phase reliability R3 is:

[0114] R3(t usage; t) = [(1-q t )R0(t usage )+q t R c (t usage ;t)]·[1-q St (1-F w (t usage ,t;ω))](11).

[0115] According to the reliability R3 of the use phase, the failure rate γ3 of the use phase is obtained as follows:

[0116]

[0117] S6. Determine the manufacturing process that needs to be controlled based on the impact of the defect.

[0118] The impact of defects is represented by the difference between the failure rate in the use phase of the aircraft product manufacturing process and the design failure rate. The greater the impact of defects, the more control is needed for the corresponding manufacturing process.

[0119] The defect impact expression is as follows:

[0120]

[0121] According to the set threshold, if it is greater than the set threshold, the manufacturing process needs to be controlled.

[0122] Alternatively, the impact of multiple defects of aircraft products can be ranked and the manufacturing process with the largest defect impact value can be selected for control.

[0123] The following describes the implementation of the present invention using a ferrite phase shifter unit in an aircraft radar array as an example. The implementation includes the following steps:

[0124] S1. Identify defects that occur during the manufacturing process of aircraft products.

[0125] By analyzing the specific manufacturing process of ferrite phase-shifter components, potential defects in the manufacturing process were identified. Based on historical information such as field failure data, failure modes, and failure mechanisms for a ferrite phase-shifter component used in an aircraft radar array, combined with historical data such as zeroing reports and failure analyses from relevant units, typical manufacturing process defects were extracted from each process step through field failure data analysis and pre-delivery burn-in experiments. The processing sequence, process type, and potential defects of the relevant manufacturing process were determined (see Table 1).

[0126] Table 1 Possible defects of ferrite phase shifter during the manufacturing process

[0127] Processing sequence Process Type Defects that may occur during the manufacturing process 1 Cutting Cracks, overcuts 2 Grinding outer circle There are tapers and burns 3 Grinding end surface Vibration marks, bumps 4 welding Shrinkage, undercut 5 Winding Gap defects

[0128] S2. Obtain the failure rate γ0 during the design usage phase.

[0129] According to the manufacturing process determined by S1, the ferrite phase shift unit has five main manufacturing processes, using different process types. For a single manufacturing process, the design reliability R corresponding to each manufacturing process can be obtained according to formula (1) combined with function fitting software and empirical values. 0,i , R 0,i It represents the design reliability of the i-th manufacturing process, specifically:

[0130]

[0131] According to the design reliability function of the above manufacturing process, the design stage failure rate γ of the i-th manufacturing process can be obtained 0,i ;

[0132]

[0133] S3. Obtain product reliability and first early failure rate affected by component-level defects based on component-level defects.

[0134] According to the processing status of the equipment during the manufacturing process of the ferrite phase shifter unit, the probability of occurrence of defects in each manufacturing process of the ferrite phase shifter unit is obtained, as shown in Table 2:

[0135] Table 2. Probability of component-level and system-level defects in each manufacturing process

[0136] Processing sequence <![CDATA[Component-level q t > <![CDATA[System-level q St > 1 0.209 0.4 2 0.5206 0.6 3 0.424 0.3 4 0.718 0.4 5 0.557 0.5

[0137] The total number of component-level defects in each manufacturing process of the ferrite phase shift unit follows the Poisson distribution. The mathematical expectation λ of the Poisson distribution and the probability τ of being classified as a non-fatal component-level defect are determined by historical data. The product reliability R affected by the activation of component-level defects in each manufacturing process is obtained by formula (3). c , the specific results are shown in Table 3:

[0138] Table 3 Product reliability affected by component-level defect activation in each manufacturing process

[0139]

[0140] According to the probability of component-level defects q in each manufacturing process in Table 2 t , the product reliability affected by component-level defect activation in each manufacturing process in Table 3, and formula (5) and formula (4), the first early failure rate γ of the i-th manufacturing process is obtained 1,i , as follows:

[0141]

[0142] S4. Obtain the product reliability and the second early failure rate that are only affected by the system-level defects based on the system-level defects.

[0143] According to the empirical values, the two parameters of the Weibull distribution in the system-level defect failure distribution of each manufacturing process of the ferrite phase shift unit are determined to obtain the system-level defect failure distribution F of the i-th manufacturing process. 2,i , as follows:

[0144]

[0145] The mutual coupling relationship between the component-level and system-level manufacturing defects introduced by each process of the ferrite phase shift unit is determined to be moderate, and the coupling parameters are set to ω1=0.30, ω2=0.50, ω3=0.45, ω4=0.30, and ω5=0.40, respectively. The coupling parameters are determined based on empirical values.

[0146] By considering the coupling effect and the probability of occurrence of system-level defects in each process, the product reliability R of the i-th manufacturing process affected only by system-level defects under competing risks is obtained. 2,i , as follows:

[0147] R 2,1 (t usage ; t) = R 0,1 (t usage )·(0.6+0.4F 0.3,1 )(34)

[0148] R 2,2 (t usage ; t) = R 0,2 (t usage )·(0.4+0.6F 0.5,2 )(35)

[0149] R 2,3 (t usage ; t) = R 0,3 (t usage )·(0.7+0.3F 0.45,3 )(36)

[0150] R 2,4 (t usage ; t) = R 0,4 (t usage )·(0.6+0.4F 0.3,4 )(37)

[0151] R 2,5 (t usage ; t) = R 0,5 (t usage )·(0.5+0.5F0.4,5 )(38).

[0152] The product reliability R of the i-th manufacturing process is affected only by system-level defects. 2,i And formula (10), the second early failure rate γ of the i-th manufacturing process is obtained 2,i .

[0153] S5. Obtain the reliability and failure rate during the use phase.

[0154] According to the probability q of component-level defects and system-level defects in the i-th manufacturing process t and q St , for {q t ,q St} combination, the reliability R of the i-th manufacturing process can be obtained 3,i for:

[0155]

[0156] According to the reliability R of the i-th manufacturing process 3,i The failure rate γ of the i-th manufacturing process in the service phase is obtained by using formula (12): 3,i .

[0157] S6. Determine the manufacturing process that needs to be controlled based on the impact of the defect.

[0158] According to the failure rate γ of the ferrite phase shift unit in the i-th manufacturing process 3,i and the design failure rate γ 0,i The difference between the two is used to obtain the defect impact, and the manufacturing process that needs to be controlled is removed based on the defect impact.

[0159] Assume that the warranty period of the ferrite phase shift unit is 90 days, that is, t usage =90, and the defect impact of the i-th manufacturing process is obtained according to formula (13). The results are shown in Table 4:

[0160] Table 4 Defect impact of each manufacturing process

[0161] Processing sequence Degree of impact of defects by process type Control priority 1 0.434 2 2 0.265 5 3 0.471 1 4 0.310 3 5 0.279 4

[0162] By sorting the defect impact results, the higher the defect impact value, the higher the control priority. According to Table 4, it can be seen that the third manufacturing process is the most affected, followed by the first manufacturing process. Therefore, in the manufacturing process of the ferrite phase shifter, according to the method proposed in the present invention, the third and first manufacturing processes are given priority for control.

[0163] The embodiments described above are merely descriptions of preferred implementations of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A manufacturing process defect analysis method based on aircraft product reliability, characterized in that: The method comprises the steps of: S1. Identify defects that occur during the manufacturing process of aircraft products; Determine the processing sequence, process type, and resulting defects of each manufacturing process in the manufacturing process based on the aircraft product manufacturing process; S2. Obtain the failure rate γ0 during the design use phase; According to the manufacturing process determined by S1, for each manufacturing process, if the aircraft product is processed by the ideal manufacturing process, its design reliability R0 is: R0(t usage )=e∧(-γt usage )=exp(-γt usage ) (1) Among them, t usage is the warranty period of the aircraft product, γ is the failure rate of a single manufacturing process during design; According to the design reliability R0 of the manufacturing process, the design failure rate γ of the manufacturing process can be obtained 0,i : S3. Obtaining product reliability and first early failure rate affected by component-level defects based on component-level defects; Assuming that the total number of component-level defects per unit area of ​​aircraft products, N, follows a Poisson distribution, the product reliability R1 affected by component-level defects is: R1(t usage ;t)=(1-q t )R0(t usage )+q t R c (t usage ;t)(4) Among them, R1 is the product reliability affected by component-level defects, q t is the probability of component-level defects, R0 is the design reliability, R c is the product reliability affected by component-level defect activation, t is the service life of the aircraft product, and t usage It is the warranty period of the aircraft product; According to the product reliability affected by component-level defects, the first early failure rate γ1 is obtained as: S4. Obtain the product reliability and the second early failure rate affected only by the system-level defects based on the system-level defects; Assuming that the failure distribution of system-level defects is Weibull distribution, the product reliability R2 affected only by system-level defects is: R2(t usage ;t)=R0(t usage )·(1-q St R s )(9) Among them, R0 is the design reliability, R s To couple the impact of component-level defects on product reliability, q St is the probability of component-level defects; According to the product reliability R2 affected only by system-level defects, the second early failure rate γ2 is obtained as follows: S5. Obtaining the reliability and failure rate during the use phase; The probability of component-level defects and system-level defects occurring during processing is q t and q St The combination of {q t ,q St }, then the usage phase reliability R3 is: R3(t usage ;t)=[(1-q t )R0(t usage )+q t R c (t usage ;t)]·[1-q St (1-F w (t usage ,t;ω))](11); According to the reliability R3 of the use phase, the failure rate γ3 of the use phase is obtained as follows: S6. Determine the manufacturing process that needs to be controlled based on the impact of the defect; The defect impact expression is as follows: The defects of different manufacturing processes are ranked according to their impact. The greater the defect impact, the more control is needed for the corresponding manufacturing process.

2. The manufacturing process defect analysis method based on aircraft product reliability according to claim 1, characterized in that: The S3, obtaining the product reliability and the first early failure rate affected by the component-level defects according to the component-level defects, is specifically: S31. Determine component-level defect distribution; In a single manufacturing process, the total number of component-level defects per unit area of ​​an aircraft product, N, follows a Poisson distribution F1(t,λ), where t is time, λ is the variance of component-level defects, and component-level defects are categorized as fatal and non-fatal. The probability of being classified as a non-fatal component-level defect is τ. S32. Determine product reliability affected by component-level defects; Non-fatal component-level defects in t usage The probability p of being activated as a fault acti For p acti =1-exp(-t usage ), when only considering the impact of non-fatal component-level defects on aircraft product reliability, the product reliability affected by component-level defect activation can be obtained as R c for: R c (t usage ;t)=E(1-p acti τ ) N =exp(-λτp acti )(3) Where λ is the Poisson distribution parameter of the total number of component-level defects, τ is the probability of a component-level defect being classified as a non-fatal component-level defect, and p acti For non-fatal component-level defects, usage The probability of being activated into a fault, E() is the mathematical expectation; Furthermore, consider the probability of non-fatal component-level defects to be q t (0≤q t ≤1), and obtain the product reliability R1 affected by component-level defects; S33. Determine a first early failure rate; The first early failure rate γ1 is obtained according to the product reliability affected by component-level defects.

3. The manufacturing process defect analysis method based on aircraft product reliability according to claim 2, characterized in that: The fatal component-level defects are screened and eliminated through quality inspection during the manufacturing process.

4. The manufacturing process defect analysis method based on aircraft product reliability according to claim 1, characterized in that: S4, obtaining the product reliability and the second early failure rate affected only by the system-level defects according to the system-level defects, is specifically: S41. Determine system-level defect distribution; Use Weibull distribution to characterize system-level defect failure distribution: F2(t usage ;α,β)=1-exp[-(t / α) β ] (6) Where F2 is the system-level defect failure distribution, α is the scale parameter, and β is the shape parameter; S42. Determine the distribution of system-level defect failures after correction; There is a positive correlation coupling between system-level defects and component-level defects, and the corrected system-level defect failure distribution F is obtained. ω : F ω (F2(t usage ),F1;ω)=F2×F1×[1+ω(1-F2)(1-F1)] (7) Where ω is the coupling parameter, which indicates the degree of correlation between the two distributions and has a value range of (0, 1); S43. Obtain product reliability that is only affected by system-level defects; According to the corrected system-level defect failure distribution F ω , we can get the product reliability R affected by coupling component level defects s for: R s (t usage ;t)=1-F w (t usage ,t;ω)(8); The probability of system-level defects is q St (0≤q St ≤1), the aircraft product reliability with system-level defects exists in a competing risk model, and the product reliability R2 is only affected by system-level defects; S44. determining a second early failure rate; The second early failure rate γ2 is obtained based on the product reliability R2 affected only by system-level defects.

5. The manufacturing process defect analysis method based on aircraft product reliability according to any one of claims 1 to 4, characterized in that: The aircraft product is a ferrite phase shift unit of an aircraft radar array.

6. The manufacturing process defect analysis method based on aircraft product reliability according to claim 5, characterized in that: The manufacturing process includes five steps, namely: the first process type is cutting, and the defects produced are cracks and overcutting; the second process type is grinding the outer circle, and the defects produced are taper and burns; the third process type is grinding the end face, and the defects produced are vibration marks and protrusions; the fourth process type is welding, and the defects produced are shrinkage and biting; the fifth process type is winding, and the defects produced are gap defects.

Images