A method and system for evaluating the failure of a diesel engine body-cylinder head composite sealing structure multi-source physical data probability

Abstract

This invention discloses a probabilistic failure assessment method and system for multi-source physical data of a diesel engine block-cylinder head composite sealing structure, belonging to the technical field of diesel engine sealing performance assessment. The method includes: acquiring physical data of the sealing band for oil holes, water holes, and air holes according to medium type and service conditions; constructing a sealing physical data tensor containing a medium-hole location dimension, a service condition dimension, and a sealing physical criterion dimension; expanding this tensor into a sealing performance matrix where each column retains a physical identifier; aggregating expert language evaluations into comprehensive cloud features; generating probabilistic weights based on Monte Carlo sampling of these comprehensive cloud features; calculating the proximity distribution and failure risk index of each sealing scheme based on the probabilistic weights and the sealing performance matrix, and determining the scheme ranking; and outputting weak hole locations, weak service conditions, and dominant failure factors based on the contribution of matrix units. This invention can maintain the physical correspondence of the sealing physical data, achieving probabilistic ranking of sealing schemes and accurate identification of weak locations.

Description

Technical Field

[0001] This invention belongs to the technical field of diesel engine sealing performance evaluation, and in particular relates to a multi-source physical data probabilistic failure evaluation method and system for diesel engine block-cylinder head composite sealing structure. Background Technology

[0002] The engine block-cylinder head mating surface of high-power-density diesel engines typically bears the responsibility of sealing multiple media, including fuel gas, cooling water, and lubricating oil. The cylinder head gasket, sealing rings, and adjacent structures at this mating surface operate under conditions of high combustion pressure, high thermal load, and strong vibration coupling, making them prone to complex sealing failures such as gas leakage, water leakage, or oil leakage. For failure assessment of diesel engine sealing structures, existing methods can disclose hierarchical index systems, finite element simulation data, normalization processing, and comprehensive sealing performance index ranking, thus solving some of the problems related to the quantification of multi-factor weights. For example, by establishing a meshed model of the cylinder block, cylinder head, cylinder head bolts, and cylinder head gasket, the finite element method is used to calculate the sealing contact pressure, thereby evaluating the sealing performance. In the sealing performance evaluation, evaluation indicators such as deformation, contact stress, contact reaction force, and contact stress distribution can be proposed. Furthermore, existing research has treated diesel engine assembly quality evaluation as a multi-attribute decision problem, using the Top-Approximation Ideal Solution Ranking Method (TOPSIS) for optimal solution selection. Other studies have determined the comprehensive weights of each parameter by coupling the Analytic Hierarchy Process (AHP) and the entropy weight method, and then combined this with the TOPSIS method to establish a calculation model for diesel engine health or sealing performance. These methods provide a certain technical foundation for the comprehensive evaluation of diesel engine sealing performance.

[0003] However, existing seal failure assessment methods typically focus on the weighting of indicators and the comprehensive index itself, failing to fully express the spatial and operational correspondences between seal physical data under multiple media, multiple orifices, and multiple operating conditions. They also struggle to directly identify weak orifices and operating conditions that increase the risk of failure. Specifically, existing assessment methods primarily analyze the sealing performance of a single medium or a single orifice, lacking a systematic data organization approach for multiple media and multiple orifices, such as oil holes, water holes, and air holes, under various operating conditions including static sealing, burst pressure, and burst pressure-temperature coupling. Furthermore, the evaluation indicators in existing methods are mostly abstract comprehensive indices. When the assessment results indicate insufficient sealing performance, engineers struggle to directly pinpoint which specific orifice, operating condition, or physical criterion is the dominant factor leading to increased failure risk from the comprehensive score. This makes it difficult to provide clear guidance for optimizing specific parameters such as bolt preload adjustment, preload sequence optimization, fluororubber hardness selection, and cylinder head gasket thickness design. Therefore, there is an urgent need for an engineering evaluation method that can maintain the physical correspondence of multi-source physical data of the diesel engine block-cylinder head composite sealing structure, and can realize scheme ranking, failure risk classification and weak hole identification. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a multi-source physical data probabilistic failure assessment method and system for a diesel engine block-cylinder head composite sealing structure, thereby resolving the issues present in the prior art.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a multi-source physical data probabilistic failure assessment method for a diesel engine block-cylinder head composite sealing structure, comprising: For multiple sealing holes in the diesel engine block-cylinder head mating surface, physical data of the sealing strip are obtained according to the medium type and service conditions. A sealing physical data tensor is constructed based on the physical data of the sealing strip. The dimensions of the sealing physical data tensor include at least the medium-orifice dimension, the operating condition dimension, and the sealing physical criterion dimension. The sealing physical data tensor is expanded into a sealing performance matrix. Each column of the sealing performance matrix retains the physical identifier of the corresponding medium type, orifice number, operating condition, and sealing physical criterion. Obtain expert evaluations of the importance of the medium-pore dimension, operating condition dimension, and sealing physical criterion dimension; convert these importance evaluations into a cloud model and aggregate cloud droplets to obtain comprehensive cloud features. Based on the comprehensive cloud features, Monte Carlo sampling is performed to generate probabilistic weights corresponding to each column of the sealing performance matrix; The proximity distribution and failure risk index of each sealing scheme are calculated based on the probabilistic weights and the sealing performance matrix, and the sealing schemes are ranked according to the proximity distribution. Based on the contribution of each matrix element in the sealing performance matrix to the failure risk index, the weak hole location, weak working condition, and dominant failure factor are output.

[0006] Preferably, the sealing holes include at least oil holes, water holes, and air holes, and the service conditions include at least static sealing conditions, burst pressure conditions, and burst pressure-temperature coupling conditions.

[0007] Preferably, the sealing physical criteria include at least contact stress margin, contact stress uniformity, and effective contact width.

[0008] Preferably, when acquiring the physical data of the sealing strip, a contact pressure sequence is extracted from the sealing strip at each sealing hole, and the width of the continuous contact area within the sealing strip that meets the contact pressure threshold is obtained; the contact pressure threshold is determined according to the medium pressure corresponding to the medium type.

[0009] Preferably, the contact stress margin is determined based on the difference or ratio between the average contact pressure and the corresponding medium pressure; the contact stress uniformity is determined based on the coefficient of variation, range coefficient, or low-pressure node ratio of the contact pressure sequence; and the effective contact width is determined based on the ratio of the width of the continuous contact area to the preset sealing width requirement.

[0010] Preferably, the process of converting the importance language evaluation into a cloud model includes: the evaluation language set of the importance language evaluation includes very low, low, medium, high and very high, and each evaluation language corresponds to a standard cloud model characterized by expectation, entropy and hyperentropy; when converting the importance language evaluation into a cloud model and performing cloud droplet aggregation, for each expert's language evaluation under the same evaluation object, the corresponding standard cloud model is called and cloud droplets are generated through a forward cloud generator; after merging all cloud droplets generated by all experts for the same evaluation object, the comprehensive expectation, comprehensive entropy and comprehensive hyperentropy are calculated through a reverse cloud algorithm, and the comprehensive expectation, comprehensive entropy and comprehensive hyperentropy are used as the comprehensive cloud features.

[0011] Preferably, the process of generating probabilistic weights corresponding to each column of the sealing performance matrix includes: performing Monte Carlo sampling and normalizing the medium-pore position weight, operating condition weight, and sealing physical criterion weight respectively; multiplying the sampled medium-pore position weight, operating condition weight, and sealing physical criterion weight to obtain column weight samples for the corresponding columns in the sealing performance matrix; and repeating the sampling to form a set of column weight samples.

[0012] Preferably, the process of calculating the closeness distribution of each sealing scheme based on the probabilistic weights and the sealing performance matrix includes: applying forward standardization to the positive indices in the sealing physical criteria and applying reverse standardization to the negative indices; forming a weighted standardized sealing performance matrix using sampling column weights under each Monte Carlo sampling, and determining the positive ideal sealing state and the negative ideal sealing state based on the matrix, and calculating the closeness of each sealing scheme; repeating the calculation to obtain the closeness distribution of each sealing scheme.

[0013] Preferably, the process of outputting weak hole locations, weak operating conditions, and dominant failure factors based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index includes: calculating the product of the deviation of each matrix unit from the ideal sealing state and the corresponding probabilistic weight to obtain the unit risk contribution; summarizing the unit risk contributions according to hole location, medium type, service condition, and sealing physical criterion respectively, and outputting the hole location, operating condition, and sealing physical criterion with the largest summed contribution value as the weak hole location, weak operating condition, and dominant failure factor.

[0014] In a second aspect, the present invention also provides a multi-source physical data probabilistic failure assessment system for a diesel engine block-cylinder head composite sealing structure, used to implement the method described in the first aspect, comprising: The physical data acquisition module is used to acquire physical data of the sealing strip for multiple sealing holes in the diesel engine block-cylinder head mating surface, according to the medium type and service conditions. The sealing unit construction module is used to construct a sealing physical data tensor based on the sealing strip physical data. The dimensions of the sealing physical data tensor include at least the medium-hole position dimension, the operating condition dimension, and the sealing physical criterion dimension. The sealing physical data tensor is then expanded into a sealing performance matrix. Each column of the sealing performance matrix retains the physical identifier of the corresponding medium type, hole position number, service condition, and sealing physical criterion. The expert cloud aggregation module is used to obtain the importance language evaluations of experts on the medium-pore position dimension, operating condition dimension and sealing physical criterion dimension, convert the importance language evaluations into cloud models and perform cloud droplet aggregation to obtain comprehensive cloud features; The probability weight generation module is used to generate probabilistic weights corresponding to each column of the sealing performance matrix by Monte Carlo sampling based on the comprehensive cloud features. The matrix sorting module is used to calculate the proximity distribution and failure risk index of each sealing scheme according to the probabilistic weights and the sealing performance matrix, and to determine the sorting of the sealing schemes according to the proximity distribution. The weak hole output module is used to output the weak hole location, weak working condition, and failure-dominant factor based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index.

[0015] Compared with the prior art, the present invention has the following advantages and technical effects: This invention acquires physical data of the sealing band by targeting multiple sealing holes in the diesel engine block-cylinder head mating surface according to the medium type and service conditions. It then constructs a three-dimensional sealing physical data tensor with medium-hole position dimension, service condition dimension, and sealing physical criterion dimension. When expanded into a sealing performance matrix, each column retains the physical identifiers of medium type, hole number, service condition, and sealing physical criterion. This ensures that the sealing physical data of oil holes, water holes, and air holes under multiple service conditions maintain a complete spatial and service condition correspondence in the same evaluation process. This solves the problem of loss of physical meaning of multi-source data in sealing evaluation and provides a data organization foundation for the accurate identification of weak points in the future.

[0016] This invention transforms expert evaluations of the importance of the medium-pore dimension, operating condition dimension, and sealing physical criterion dimension into a cloud model and aggregates cloud droplets to obtain comprehensive cloud features. This approach incorporates the ambiguity, randomness, and degree of group disagreement in expert judgments into the evaluation system, avoiding information loss caused by simply compressing expert experience into deterministic values, thereby improving the objectivity and credibility of weight assignment.

[0017] This invention generates probabilistic weights for each column of the sealing performance matrix by Monte Carlo sampling based on comprehensive cloud features. This allows the uncertainty of the importance of the medium-orifice location, the severity of the operating conditions, and the importance of the sealing physical criteria to be transmitted to the subsequent ranking process in the form of probabilistic weights, thereby quantifying the uncertainty of the evaluation results and improving the reliability of the ranking of sealing schemes.

[0018] This invention calculates the proximity distribution and failure risk index of each sealing scheme based on probabilistic weights and sealing performance matrix. It performs TOPSIS ranking in each sampling and obtains the probability distribution of proximity rather than a single deterministic score. This allows the evaluation results to reflect the impact of weight uncertainty on the scheme ranking, thereby improving the comprehensiveness and robustness of failure risk assessment.

[0019] This invention outputs weak hole locations, weak operating conditions, and dominant failure factors based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index. It can decompose the comprehensive evaluation results layer by layer to specific combinations of medium-hole location-operating condition-physical criteria, giving the evaluation conclusions clear engineering orientation. This allows it to directly serve specific parameter optimization operations such as bolt preload adjustment, preload sequence optimization, fluororubber hardness selection, and cylinder head gasket thickness design. Attached Figure Description

[0020] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a multi-source physical data probabilistic failure assessment method for a diesel engine block-cylinder head composite sealing structure according to an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating the construction of the sealing physical data tensor and sealing performance matrix according to an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the aggregation of expert language evaluation standard cloud and comprehensive cloud in an embodiment of the present invention; Figure 4 This is a schematic diagram of the probabilistic weighting and sealing performance matrix sorting mechanism in an embodiment of the present invention; Figure 5 This is a schematic diagram showing the weight comparison between the criterion layer and the indicator layer in an embodiment of the present invention; Figure 6 This is a schematic diagram of the system module structure and risk output closed loop of an embodiment of the present invention. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0023] Example 1 like Figure 1 As shown, this embodiment provides a multi-source physical data probabilistic failure assessment method for a diesel engine block-cylinder head composite sealing structure, including: S1. For multiple sealing holes in the diesel engine block-cylinder head mating surface, obtain physical data of the sealing strip according to the medium type and service conditions; Furthermore, the sealing holes include at least oil holes, water holes, and air holes, and the service conditions include at least static sealing conditions, burst pressure conditions, and burst pressure-temperature coupling conditions.

[0024] Furthermore, when acquiring the physical data of the sealing strip, a contact pressure sequence is extracted from the sealing strip at each sealing hole, and the width of the continuous contact area within the sealing strip that meets the contact pressure threshold is obtained; the contact pressure threshold is determined according to the medium pressure corresponding to the medium type.

[0025] Specifically, for the composite sealing structure of the diesel engine block-cylinder head to be evaluated, the set of sealing holes H is determined, which includes at least oil holes, water holes and air holes; the set of media L is determined, which includes at least lubricating oil, cooling water and fuel gas; and the set of operating conditions G is determined, which includes at least static sealing conditions, explosion pressure conditions and explosion pressure-temperature coupling conditions.

[0026] For each sealing hole position h and operating condition g, extract the contact pressure sequence P={ at the sealing strip node. , ,..., }, and according to the medium pressure corresponding to medium m Determine the contact pressure threshold. Statistically analyze the node contact pressure sequence to obtain the average contact pressure, standard deviation of contact pressure, proportion of low-pressure nodes, and width of the continuous contact area that meets the contact pressure threshold.

[0027] S2. Construct a sealing physical data tensor based on the sealing strip physical data. The dimensions of the sealing physical data tensor include at least the medium-hole location dimension, the operating condition dimension, and the sealing physical criterion dimension. Expand the sealing physical data tensor into a sealing performance matrix. Each column of the sealing performance matrix retains the physical identifier of the corresponding medium type, hole location number, service condition, and sealing physical criterion. Specifically, such as Figure 2As shown, the sealing scheme, medium type, orifice number, service conditions, and sealing physical criteria are each used as tensor dimensions to construct the sealing physical data tensor. .in, Indicates the sealing scheme number. Indicates the media type. Indicates the hole position number. Indicates service condition. Indicates the physical criteria for sealing; Indicates the first A sealing solution in the medium 、 Hole position 、 Operating conditions and sealing physical principles The sealing physical criteria values ​​are as follows. The sealing physical criteria include at least contact stress margin, contact stress uniformity, and effective contact width.

[0028] Furthermore, the contact stress margin is determined based on the difference or ratio between the average contact pressure and the corresponding medium pressure; the contact stress uniformity is determined based on the coefficient of variation, range coefficient, or low-pressure node ratio of the contact pressure sequence; and the effective contact width is determined based on the ratio of the width of the continuous contact area to the preset sealing width requirement.

[0029] Specifically, the contact stress margin can be expressed as the difference or ratio between the average contact pressure and the medium pressure; the contact stress uniformity can be expressed as the reverse value of the contact pressure variation coefficient, the reverse value of the range coefficient, or the reverse value of the proportion of low-pressure nodes; and the effective contact width can be expressed as the ratio of the width of the continuous contact area to the preset sealing width requirement. Thus, for each hole location, medium, and operating condition, three physical criteria values—contact stress margin, contact stress uniformity, and effective contact width—are obtained.

[0030] In one implementation, the sealing strip node contact pressure sequence P={ , ,..., The statistics and the proportion of local low-pressure nodes can be constructed using the following formula: (1) Average contact pressure: (1) (2) Standard deviation of contact pressure and percentage of low-pressure nodes: (2) (3) (3) Three sealing physical criterion values: (4) (5) (6) In the formula, Indicates the first Contact pressure at each sealing node, ; This indicates the number of contact nodes on the sealing strip at that sealing hole location that are included in the statistics; This indicates the average contact pressure of the sealing hole under the corresponding medium and operating conditions; It represents the standard deviation of contact pressure, used to characterize the dispersion of the contact pressure distribution of the sealing strip; This indicates the pressure of the corresponding medium, which may include lubricating oil, cooling water, or natural gas. This indicates the low-pressure judgment threshold, which can be determined based on the corresponding medium pressure or a preset safety factor. This represents an indicator function, which takes the value 1 when the condition inside the parentheses is true, and takes the value 0 otherwise. This indicates the proportion of low-pressure nodes, used to characterize the degree of local weakness in the sealing strip where the contact pressure is below a threshold; Indicates the contact stress margin; Indicates the uniformity of contact stress; Indicates the effective contact width index; This indicates the width of the continuous effective contact area that meets the contact pressure threshold requirement; This indicates the preset sealing width requirement.

[0031] Specifically, for each sealing scheme Seal the physical data tensor The tensor elements in the matrix are expanded into matrix rows according to the medium type, orifice number, service conditions, and sealing physical criteria to form a sealing performance matrix. .matrix Each column corresponds to a set of columns. The uniquely defined sealing physical evaluation unit retains the corresponding medium type, hole number, service condition, and sealing physical criterion identifier, so that subsequent sorting results can be traced back to the specific weak hole and weak condition.

[0032] S3. Obtain the importance language evaluation of the medium-pore dimension, operating condition dimension and sealing physical criterion dimension from experts, convert the importance language evaluation into a cloud model and perform cloud droplet aggregation to obtain comprehensive cloud features; Furthermore, the process of converting the importance language evaluation into a cloud model includes: the evaluation language set of the importance language evaluation includes very low, low, medium, high and very high, and each evaluation language corresponds to a standard cloud model characterized by expectation, entropy and hyperentropy; when converting the importance language evaluation into a cloud model and performing cloud droplet aggregation, for each expert's language evaluation under the same evaluation object, the corresponding standard cloud model is called and cloud droplets are generated through a forward cloud generator; after merging all cloud droplets generated by all experts for the same evaluation object, the comprehensive expectation, comprehensive entropy and comprehensive hyperentropy are calculated through a reverse cloud algorithm, and the comprehensive expectation, comprehensive entropy and comprehensive hyperentropy are used as the comprehensive cloud features.

[0033] Specifically, experts in diesel engine seal design, finite element analysis, and testing were invited to conduct linguistic evaluations of the medium-pore weakness, operating condition severity, and the importance of seal physical criteria. The linguistic evaluation set can be set to {very low, low, medium, high, very high}, with each linguistic term corresponding to a standard cloud model. =( , , ).in Indicates the evaluation center, Indicates evaluation ambiguity, This indicates the randomness of the evaluation. The process of aggregating the standard cloud and criteria layer corresponding to expert language evaluation is as follows: Figure 3 As shown.

[0034] When language evaluation is converted into a standard cloud model, the membership function and digital characteristics of the normal cloud can be determined by the following formula: (1) Cloud model membership mapping: (7) (2) Membership function of normal cloud: (8) (3) Standard cloud digital characteristics: (9) (10) (11) In the formula, This indicates the domain of discourse corresponding to the expert's linguistic evaluation; Domain of discourse One of the evaluation values; Indicates the evaluation value Membership degree to a certain language evaluation level; The expectation of the cloud model is used to characterize the central value of the language evaluation rating. The entropy of the cloud model is used to characterize the fuzziness and acceptable range of language evaluation; The hyperentropy of the cloud model is used to characterize the uncertainty of entropy and the random fluctuations in expert evaluation. and These represent the lower and upper limits of the numerical range corresponding to a certain language evaluation level, respectively. This represents the interval adjustment coefficient used to determine the entropy value, which can be preset according to the width of the language evaluation level interval or the evaluation scale. This represents the central expectation value of the standard cloud model.

[0035] =( , , ) indicates the first The expert commented on the first The cloud model is obtained by converting the language evaluation of each evaluation object; =( , , The ) represents the result obtained after aggregating multiple expert cloud drops. A comprehensive cloud model for each evaluation object.

[0036] For each expert's language evaluation, cloud droplets are generated using a forward cloud generator. The cloud droplets generated by all experts for the same evaluation object are then merged, and a reverse cloud algorithm is used to obtain the comprehensive expectation, comprehensive entropy, and comprehensive hyperentropy, which serve as the comprehensive cloud features for that evaluation object. Through this process, expert experience is not compressed into a single, definitive value, but rather participates in subsequent calculations as comprehensive cloud features with fuzziness and randomness.

[0037] S4. Based on the comprehensive cloud features, perform Monte Carlo sampling to generate probabilistic weights corresponding to each column of the sealing performance matrix; Furthermore, the process of generating probabilistic weights corresponding to each column of the sealing performance matrix includes: performing Monte Carlo sampling and normalizing the medium-pore position weight, operating condition weight, and sealing physical criterion weight respectively; multiplying the sampled medium-pore position weight, operating condition weight, and sealing physical criterion weight to obtain column weight samples for the corresponding columns in the sealing performance matrix; and repeating the sampling to form a set of column weight samples.

[0038] Specifically, such as Figure 4 As shown, Monte Carlo sampling is performed based on comprehensive cloud features. The media-pore location weight, operating condition weight, and sealing physical criterion weight are sampled and normalized separately. Then, the three types of weights are multiplied to obtain the column weight samples corresponding to each column of the sealing performance matrix. Repeated sampling is performed. This yields a sample set of column weights. The mean of the column weights is used to characterize average importance, while the standard deviation or confidence interval is used to characterize weight uncertainty.

[0039] When performing Monte Carlo sampling on the comprehensive cloud features, the normalized weight and mean weight of index j in the t-th sampling can be expressed as: (1) Normalized weights under the t-th sampling: (12) (2) Weighted mean: (13) (3) Coupling of three types of weights: medium-hole location, operating condition, and criteria: (14) In the formula, This indicates the Monte Carlo sampling number. ; This indicates the total number of Monte Carlo samplings; Indicates the first The first sample generated by the integrated cloud model in the second sampling The importance of each indicator is randomly assigned. Indicates the first The second sampling Normalized weights of each indicator; Indicates the first Each indicator in Average weight under sub-sampling; This indicates the number of indicators participating in the same level of normalization calculation; , , These represent the weight indices in the medium-pore location dimension, the operating condition dimension, and the sealing physical criterion dimension, respectively. Indicates the first The column weights of the sealing performance matrix are obtained by coupling the medium-orifice weight, the operating condition weight, and the sealing physical criterion weight under the second sampling.

[0040] S5. Calculate the proximity distribution and failure risk index of each sealing scheme according to the probabilistic weights and the sealing performance matrix, and determine the ranking of the sealing schemes according to the proximity distribution. Furthermore, the process of calculating the closeness distribution of each sealing scheme based on the probabilistic weights and the sealing performance matrix includes: applying forward standardization to the positive indices in the sealing physical criteria and applying backward standardization to the negative indices; forming a weighted standardized sealing performance matrix using sampling column weights under each Monte Carlo sampling, and determining the positive ideal sealing state and negative ideal sealing state based on the matrix, and calculating the closeness of each sealing scheme; repeating the calculation to obtain the closeness distribution of each sealing scheme.

[0041] Specifically, the sealing performance matrix is ​​standardized, with contact stress margin and effective contact width used as positive indicators, and contact pressure variation coefficient, low-pressure node ratio, or local risk index used as negative indicators. Under each Monte Carlo sampling, the standardized matrix is ​​weighted using column-weighted samples to determine the positive and negative ideal sealing states, and TOPSIS is used to calculate the closeness of each sealing scheme. After repeated calculations, the closeness distribution of each sealing scheme is obtained.

[0042] In probabilistic ranking, the sealing performance matrix, the normalized matrix, and the weighted normalized matrix can be represented as: (1) Matrix definition: (15) (2) Standardization of positive indicators: (16) (3) Standardization of reverse indicators: (17) (4) Weighted Standardization Matrix: (18) (5) Positive and negative ideal sealing conditions: (19) (6) Distance from positive and negative ideal states: (20) (twenty one) (7) Proximity: (twenty two) (8) Probability distribution of proximity: (twenty three) In the formula, Represents the sealing performance matrix; This represents the standardized sealing performance matrix; This represents the weighted normalized sealing performance matrix; Indicates the first The sealing scheme is in the first The original physical criterion values ​​for each sealing evaluation index; This represents the standardized index value; This represents the weighted standardized index value; ,in Indicates the number of sealing schemes involved in the sorting; ,in This indicates the number of evaluation indicators in the sealing performance matrix; Indicates the first The probabilistic column weights corresponding to each indicator; and They represent the first The maximum and minimum values ​​of each indicator across all sealing schemes; Indicates the first The ideal sealing state value of each indicator; Indicates the first The negative ideal sealing state value of each indicator; Indicates the first The distance between a sealing solution and the ideal sealing state; Indicates the first The distance between a sealing solution and a negative ideal sealing state; Indicates the first The closeness of the sealing scheme, The larger the value, the closer the sealing solution is to the ideal sealing state; Indicates the first The Monte Carlo sampling The closeness of each sealing solution.

[0043] When the maximum and minimum values ​​of a certain evaluation index are equal across all sealing schemes, the standardized value of that index is set as a preset constant, or a preset minimum positive number is added to the denominator. To avoid the denominator being zero.

[0044] S6. Output the weak hole location, weak working condition and failure-dominant factors based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index.

[0045] Furthermore, the process of outputting weak hole locations, weak operating conditions, and dominant failure factors based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index includes: calculating the product of the deviation of each matrix unit from the ideal sealing state and the corresponding probabilistic weight to obtain the unit risk contribution; summarizing the unit risk contributions according to hole location, medium type, service condition, and sealing physical criterion respectively, and outputting the hole location, operating condition, and sealing physical criterion with the largest summed contribution value as the weak hole location, weak operating condition, and dominant failure factor.

[0046] Specifically, the failure risk index is determined based on the mean, lower confidence limit, or quantile of the proximity distribution. The deviation of each matrix element from the ideal sealing state is further calculated as the product of the corresponding probabilistic weight to obtain the element risk contribution. The element risk contributions are then summarized according to orifice location, medium, operating condition, and sealing physical criteria. The orifice location, operating condition, and criterion with the highest contribution values ​​are output as the weak orifice location, weak operating condition, and dominant failure factor.

[0047] When outputting failure risk, the proximity distribution can be transformed into a failure risk index, and the risk can be further decomposed into matrix units: (1) Failure Risk Index: (twenty four) (2) Risk contribution of matrix unit: (25) (3) Summarize risk contributions by medium, orifice location, operating condition, and criteria: (26) In the formula, Indicates the first Failure risk index of a sealing scheme; Indicates the first The closeness distribution of the sealing schemes at the confidence level is as follows: quantiles below; Indicates the preset quantile level or confidence level; Indicates the first The sealing scheme is in the first Risk contribution on each matrix unit; Indicates the first Probabilistic weights corresponding to each matrix unit or evaluation index; Indicates the first The ideal sealing state value of each indicator; Indicates the first The sealing scheme is in the first Weighted standardized values ​​for each indicator; Indicates according to media type Hole position Operating conditions and sealing physical principles Summary of risk contribution; Indicates media type Hole position Operating conditions and sealing physical principles The corresponding set of matrix units; This indicates the number of the sealing hole position, which includes oil holes, water holes, and air holes; Indicates service status; This indicates the physical criteria for sealing, including contact stress margin, contact stress uniformity, and effective contact width. By comparing different... The size of the hole can determine the weak hole location, weak working condition, and the dominant failure factor.

[0048] In one specific implementation, the evaluation object is a composite sealing structure of a high-power-density diesel engine block-cylinder head. The sealing orifices include two oil holes, six water holes, and one air hole. The operating conditions include static sealing, explosion pressure, and explosion pressure-temperature coupling conditions. The contact pressure and effective contact width of the sealing strip nodes are obtained through finite element simulation or pressure-sensitive film testing, and a multi-media, multi-orifice, multi-condition sealing physical data tensor is constructed.

[0049] By evaluating cloud-based aggregation and Monte Carlo sampling using expert language, probabilistic weights corresponding to each column of the sealing performance matrix are obtained. After substituting these probabilistic weights into the sealing performance matrix and sorting them, the optimal overall sealing performance scheme and schemes with higher failure risk can be identified. Simultaneously, based on the unit risk contribution, weak points in water holes, air holes, or oil holes can be further identified, determining whether they are mainly affected by coupled operating conditions, insufficient contact stress, uneven stress distribution, or insufficient effective contact width. The correction relationship between the weights of the criterion layer and the index layer from deterministic AHP to probabilistic evaluation is as follows: Figure 5 As shown.

[0050] The beneficial effects of this embodiment: This embodiment shifts the focus of protection from general algorithm combinations to the construction of dedicated data and failure output procedures for diesel engine composite seals. Each column in the sealing performance matrix corresponds to a specific medium, orifice location, operating condition, and sealing physical criteria, thus preserving the physical meaning of the seal and preventing evaluation indicators from being abstracted into general indicators without engineering guidance.

[0051] This embodiment uses expert language cloud aggregation and Monte Carlo sampling to incorporate the importance of medium-pore location, operating condition severity, and sealing physical criteria into the sealing performance matrix in the form of probabilistic weights, which can quantify the ambiguity, randomness, and degree of disagreement of expert judgment.

[0052] This embodiment not only outputs a ranking of sealing schemes, but also locates weak holes, weak working conditions, and failure-dominant factors based on the contributions of matrix units, so that the evaluation results can directly serve the optimization of parameters such as bolt preload, preload sequence, fluororubber hardness, and cylinder head gasket thickness.

[0053] Example 2 like Figure 6 As shown, this embodiment also provides a multi-source physical data probabilistic failure assessment system for a diesel engine block-cylinder head composite sealing structure, used to implement the method described in claim 1, including: The physical data acquisition module is used to acquire physical data of the sealing strip for multiple sealing holes in the diesel engine block-cylinder head mating surface, according to the medium type and service conditions. The sealing unit construction module is used to construct a sealing physical data tensor based on the sealing strip physical data. The dimensions of the sealing physical data tensor include at least the medium-hole position dimension, the operating condition dimension, and the sealing physical criterion dimension. The sealing physical data tensor is then expanded into a sealing performance matrix. Each column of the sealing performance matrix retains the physical identifier of the corresponding medium type, hole position number, service condition, and sealing physical criterion. The expert cloud aggregation module is used to obtain the importance language evaluations of experts on the medium-pore position dimension, operating condition dimension and sealing physical criterion dimension, convert the importance language evaluations into cloud models and perform cloud droplet aggregation to obtain comprehensive cloud features; The probability weight generation module is used to generate probabilistic weights corresponding to each column of the sealing performance matrix by Monte Carlo sampling based on the comprehensive cloud features. The matrix sorting module is used to calculate the proximity distribution and failure risk index of each sealing scheme according to the probabilistic weights and the sealing performance matrix, and to determine the sorting of the sealing schemes according to the proximity distribution. The weak hole output module is used to output the weak hole location, weak working condition, and failure-dominant factor based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index.

[0054] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A probabilistic failure assessment method based on multi-source physical data for a diesel engine block-cylinder head composite sealing structure, characterized in that, Includes the following steps: For multiple sealing holes in the diesel engine block-cylinder head mating surface, physical data of the sealing strip are obtained according to the medium type and service conditions. A sealing physical data tensor is constructed based on the physical data of the sealing strip. The dimensions of the sealing physical data tensor include at least the medium-orifice dimension, the operating condition dimension, and the sealing physical criterion dimension. The sealing physical data tensor is expanded into a sealing performance matrix. Each column of the sealing performance matrix retains the physical identifier of the corresponding medium type, orifice number, operating condition, and sealing physical criterion. Obtain expert evaluations of the importance of the medium-pore dimension, operating condition dimension, and sealing physical criterion dimension; convert these importance evaluations into a cloud model and aggregate cloud droplets to obtain comprehensive cloud features. Based on the comprehensive cloud features, Monte Carlo sampling is performed to generate probabilistic weights corresponding to each column of the sealing performance matrix; The proximity distribution and failure risk index of each sealing scheme are calculated based on the probabilistic weights and the sealing performance matrix, and the sealing schemes are ranked according to the proximity distribution. Based on the contribution of each matrix element in the sealing performance matrix to the failure risk index, the weak hole location, weak working condition, and dominant failure factor are output.

2. The method according to claim 1, characterized in that, The sealing holes include at least oil holes, water holes, and air holes, and the service conditions include at least static sealing conditions, burst pressure conditions, and burst pressure-temperature coupling conditions.

3. The method according to claim 1, characterized in that, The sealing physical criteria include at least contact stress margin, contact stress uniformity, and effective contact width.

4. The method according to claim 3, characterized in that, When acquiring the physical data of the sealing strip, the contact pressure sequence is extracted from the sealing strip at each sealing hole, and the width of the continuous contact area within the sealing strip that meets the contact pressure threshold is obtained; the contact pressure threshold is determined according to the medium pressure corresponding to the medium type.

5. The method according to claim 4, characterized in that, The contact stress margin is determined based on the difference or ratio between the average contact pressure and the corresponding medium pressure; the contact stress uniformity is determined based on the coefficient of variation, range coefficient, or proportion of low-pressure nodes in the contact pressure sequence; and the effective contact width is determined based on the ratio of the width of the continuous contact area to the preset sealing width requirement.

6. The method according to claim 1, characterized in that, The process of converting the importance language evaluation into a cloud model includes: the evaluation language set of the importance language evaluation includes very low, low, medium, high and very high, and each evaluation language corresponds to a standard cloud model characterized by expectation, entropy and hyperentropy; when converting the importance language evaluation into a cloud model and performing cloud droplet aggregation, for each expert's language evaluation under the same evaluation object, the corresponding standard cloud model is called and cloud droplets are generated through a forward cloud generator; after merging all cloud droplets generated by all experts for the same evaluation object, the comprehensive expectation, comprehensive entropy and comprehensive hyperentropy are calculated through a reverse cloud algorithm, and the comprehensive expectation, comprehensive entropy and comprehensive hyperentropy are used as the comprehensive cloud features.

7. The method according to claim 1, characterized in that, The process of generating probabilistic weights corresponding to each column of the sealing performance matrix includes: performing Monte Carlo sampling and normalizing the medium-pore weight, operating condition weight, and sealing physical criterion weight respectively; multiplying the sampled medium-pore weight, operating condition weight, and sealing physical criterion weight to obtain column weight samples for the corresponding columns in the sealing performance matrix; and repeating the sampling to form a set of column weight samples.

8. The method according to claim 1, characterized in that, The process of calculating the closeness distribution of each sealing scheme based on the probabilistic weights and the sealing performance matrix includes: applying forward standardization to the positive indices in the sealing physical criteria and applying backward standardization to the negative indices; forming a weighted standardized sealing performance matrix using sampling column weights under each Monte Carlo sampling, and determining the positive ideal sealing state and negative ideal sealing state based on the matrix, and calculating the closeness of each sealing scheme; repeating the calculation to obtain the closeness distribution of each sealing scheme.

9. The method according to claim 8, characterized in that, The process of outputting weak holes, weak operating conditions, and dominant failure factors based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index includes: calculating the product of the deviation of each matrix unit from the ideal sealing state and the corresponding probabilistic weight to obtain the unit risk contribution; summarizing the unit risk contributions according to hole location, medium type, service condition, and sealing physical criterion respectively; and outputting the hole location, operating condition, and sealing physical criterion with the largest summed contribution value as the weak holes, weak operating conditions, and dominant failure factors.

10. A multi-source physical data probabilistic failure assessment system for a diesel engine block-cylinder head composite sealing structure, characterized in that, To implement the method according to any one of claims 1-9, comprising: The physical data acquisition module is used to acquire physical data of the sealing strip for multiple sealing holes in the diesel engine block-cylinder head mating surface, according to the medium type and service conditions. The sealing unit construction module is used to construct a sealing physical data tensor based on the sealing strip physical data. The dimensions of the sealing physical data tensor include at least the medium-hole position dimension, the operating condition dimension, and the sealing physical criterion dimension. The sealing physical data tensor is then expanded into a sealing performance matrix. Each column of the sealing performance matrix retains the physical identifier of the corresponding medium type, hole position number, service condition, and sealing physical criterion. The expert cloud aggregation module is used to obtain the importance language evaluations of experts on the medium-pore position dimension, operating condition dimension and sealing physical criterion dimension, convert the importance language evaluations into cloud models and perform cloud droplet aggregation to obtain comprehensive cloud features; The probability weight generation module is used to generate probabilistic weights corresponding to each column of the sealing performance matrix by Monte Carlo sampling based on the comprehensive cloud features. The matrix sorting module is used to calculate the proximity distribution and failure risk index of each sealing scheme according to the probabilistic weights and the sealing performance matrix, and to determine the sorting of the sealing schemes according to the proximity distribution. The weak hole output module is used to output the weak hole location, weak working condition, and failure-dominant factor based on the contribution of each matrix unit in the sealing performance matrix to the failure risk index.

Images