基于共词分析的航空发动机系统失效案例研究

邓姝娟; 王卫泽; 张英; 张成成; 涂善东

doi:10.14135/j.cnki.1006-3080.20200824002

基于共词分析的航空发动机系统失效案例研究

doi: 10.14135/j.cnki.1006-3080.20200824002

1.
华东理工大学机械与动力工程学院，上海 200237
2.
中国航发商用航空发动机有限责任公司，上海 200241

基金项目: 上海市科委资助项目（16DZ2260604）

详细信息

作者简介:
邓姝娟（1998—），女，湖北人，硕士生，主要研究方向：失效分析与表面结构完整性。E-mail：470648008@qq.com

通讯作者:
王卫泽，E-mail：wangwz@ecust.edu.cn

中图分类号: G353.1
计量
- 文章访问数: 632
- HTML全文浏览量: 435
- PDF下载量: 10
- 被引次数: 0
出版历程
- 收稿日期: 2020-08-24
- 网络出版日期: 2020-12-16

Case Study of Aeroengine System Failure Based on Co-word Analysis

1.
School of Mechanical and Power Engineering, East China University of Science and Technology, Shanghai 200237, China
2.
AECC Commercial Aircraft Engine Co. Ltd, Shanghai 200241, China

摘要

摘要: 航空发动机是多学科交叉、多组件耦合的复杂系统，其工况复杂，一旦发生故障便会造成灾难性后果。采用共词分析法将航空发动机系统失效领域内的关键词加以分类，归纳出了主要的失效形式并实现了结果可视化。研究发现考虑自相关性后的聚类分析效果更佳，且对于不同失效机理的研究更加具体和准确。对于服役过程中的常见故障，结合聚类分析结果和多维尺度分析图可以更快速地采取检查手段进行失效分析，有效避免类似事件的再次发生。
- 航空发动机 /
- 失效分析 /
- 共词分析 /
- 自相关性
Abstract: Aero-engine is a complex system with interdisciplinary and multi-component coupling, and its working conditions are complex. It can have catastrophic consequence once the failure of aero-engine occurs. Therefore, the failure analysis of aero-engine system can provide important reference for its design, maintenance and safe operation. In this work, co-word analysis method is adopted, and three multivariate analysis methods of factor analysis, cluster analysis and multidimensional scale analysis are used to classify the keywords in the field of aero-engine system failure, and the main failure forms are summarized and the results are visualized.It is found that clustering analysis with autocorrelation is better. Generally, compared with traditional statistical analysis or theoretical analysis, the analysis results of typical failure forms are basically consistent, but co-word analysis is more specific, detailed and accurate for the study of different failure mechanisms. The results of traditional statistical analysis only stay at the level of general quantitative analysis, and can only indicate the general failure direction for practical engineering application. Multi-dimensional scale analysis realizes visual similarity visualization of failure keywords. The closer to the origin of coordinates, the more core the failure forms are, the higher the frequency of occurrence is. For these common faults in the service process, combining cluster analysis results and multidimensional scale analysis charts, it is convenient for technicians to find out the failure reasons of faulty equipment more quickly, grasp the key points of work more accurately, and determine the protection scheme more effectively, effectively avoiding the recurrence of similar events and prolonging the service life of the engine.
- aero-engine /
- failure analysis /
- co-word analysis /
- autocorrelation

HTML全文

图 1 叶盘类因子分析碎石图

Figure 1. Scree plot of factor analysis of blade and wheel disc

下载: 全尺寸图片幻灯片

图 2 叶盘失效分析的聚类树状图

Figure 2. Dendrogram of failure analysis of blade and wheel disc

下载: 全尺寸图片幻灯片

图 3 未考虑自相关性的叶盘聚类树状图

Figure 3. Dendrogram of blade and wheel disc without considering autocorrelation

下载: 全尺寸图片幻灯片

图 4 叶盘失效关键词的MDS分析图

Figure 4. MDS analysis chart of failure keywords of blade and wheel disc

下载: 全尺寸图片幻灯片

图 5 机械传动系统失效关键词的MDS分析图

Figure 5. MDS analysis chart of failure keywords of mechanical transmission system

下载: 全尺寸图片幻灯片

图 6 其他附件失效关键词的MDS分析图

Figure 6. MDS analysis chart of failure keywords of other accessory equipment

下载: 全尺寸图片幻灯片

表 1 叶盘失效相似矩阵（部分）

Table 1. Similar matrix of blade and wheel disc failure (Part)

Keyword	Cosine value
Keyword	Blade	Fracture	Creep	Crack	Fatigue	Selective oxidation
Blade	1.0000	0.9842	0.8024	0.9898	0.9902	0.8410
Fracture	0.9842	1.0000	0.7965	0.9680	0.9889	0.8146
Creep	0.8024	0.7965	1.0000	0.7948	0.8018	0.6909
Crack	0.9898	0.9680	0.7948	1.0000	0.9853	0.8227
Fatigue	0.9902	0.9889	0.8018	0.9853	1.0000	0.8146
Selective oxidation	0.8410	0.8146	0.6909	0.8227	0.8146	1.0000

下载: 导出CSV

表 2 叶盘失效相异矩阵（部分）

Table 2. Dissimilar matrix of blade and wheel disc failure (Part)

Keyword	Cosine value
Keyword	Blade	Fracture	Creep	Crack	Fatigue	Selective oxidation
Blade	0	0.0158	0.1976	0.0102	0.0098	0.1590
Fracture	0.0158	0	0.2035	0.0320	0.0111	0.1854
Creep	0.1976	0.2035	0	0.2052	0.1982	0.3091
Crack	0.0102	0.032	0.2052	0	0.0147	0.1773
Fatigue	0.0098	0.0111	0.1982	0.0147	0	0.1854
Selective oxidation	0.1590	0.1854	0.3091	0.1773	0.1854	0

下载: 导出CSV

表 3 叶盘失效关键词

Table 3. Failure keywords of blade and wheel disc

Number	Keyword	Number	Keyword	Number	Keyword
A01	Blade	A16	Cyclic alternating load	A31	Recrystallization
A02	Fracture	A17	Coating shedding	A32	Surface damage
A03	Creep	A18	High temperature	A33	Low cycle fatigue
A04	Crack	A19	Etch pit	A34	Fretting wear
A05	Fatigue	A20	Forging defect	A35	Incorrect mounting
A06	Selective oxidation	A21	Machining mark	A36	Exciting force
A07	Foreign body impact	A22	Friction	A37	High-low cycle complex fatigue
A08	Hole defect	A23	Overload	A38	Overtemperature and oversintering
A09	Stress concentration	A24	Improper manufacturing process	A39	Resonance
A10	Vibration	A25	Blister	A40	Bending vibration
A11	High cycle fatigue	A26	Excessive blade bending	A41	Oxidation defect
A12	Poor processing quality	A27	Thermal corrosion	A42	Structural degradation
A13	Material defect	A28	Thermal damage	A43	Abnormal microstructure
A14	Pit corrosion	A29	Carbide
A15	Corrosion	A30	Coating degradation

下载: 导出CSV

表 4 机械传动系统失效关键词

Table 4. Failure keywords of mechanical transmission system

Number	Keyword	Number	Keyword	Number	Keyword
B01	Shaft	B16	Cyclic load	B31	Bengding and torsional fatigue
B02	Bearing cage	B17	Friction	B32	Insufficient lubrication
B03	Bearing	B18	Pit corrosion	B33	Rotating bending fatigue
B04	Bearing ball	B19	Spalling damage	B34	Fretting wear
B05	Rivet of bearing cage	B20	Machining mark	B35	Abnormal bending load
B06	Fracture	B21	Irregular design	B36	Skidding damage
B07	Crack	B22	Foreign body impact	B37	Torsional fatigue
B08	Stress concentration	B23	Overload	B38	Resonance
B09	Fatigue	B24	Electric arc damage	B39	Inappropriate fit clearances
B10	Abrasion	B25	Incorrect mounting	B40	Abrasive contamination
B11	Improper material selection	B26	Abnormal axial load	B41	Improper heat treatment
B12	Improper manufacturing or maintenance process	B27	Forging defect	B42	Governor idle gear
B13	Material inclusion	B28	High temperature damage	B43	Transfer and bevel gear
B14	Surface contact fatigue	B29	Overheating and overheating	B44	Clutch gear
B15	High contact stress	B30	High cycle fatigue	B45	Improper case hardening

下载: 导出CSV

表 5 其他附件失效关键词

Table 5. Failure keywords of other accessory equipment

Number	Keyword	Number	Keyword	Number	Keyword
C01	Cylinder	C17	Hydraulic actuator	C33	Kink
C02	Cylinder head	C18	Cited trachea clamp	C34	Surface damage
C03	Auxiliary power supply inlet valve	C19	Engine actuating cylinder	C35	Stress corrosion
C04	Combustor liner	C20	Tappet body	C36	Shear failure
C05	Hydraulic pressure line	C21	Afterburner fuel manifold	C37	Overload
C06	Exhaust system pipeline	C22	Crack	C38	Abrasion
C07	Bleed valve clamp bolt	C23	Fracture	C39	Corrosion
C08	Wheel disc connecting pin	C24	Material defect	C40	Explosion failure
C09	Fuel line	C25	Fatigue	C41	Poor quality of the roll welding seam
C10	Combustion chamber shell	C26	Improper manufacturing or production process	C42	Unqualified welding process
C11	Combustion control system adapter	C27	High cycle fatigue	C43	Pit corrosion
C12	Air radiator	C28	Stress concentration	C44	Impurity residue in molding process
C13	Reed	C29	High temperature	C45	Foreign object damage
C14	Oil pipeline	C30	Vibration	C46	Machining mark
C15	Corrugated pipe	C31	Oxidation
C16	Gas cylinder conduit	C32	Thermal fatigue

下载: 导出CSV

表 6 叶盘类因子分析总方差解释表

Table 6. Explanatory table of total variance of factor analysis of blade and wheel disc

Component number	Variance of initial eigenvalue			Variance of extraction sums of squared loadings			Rotation sums of squared loadings
Component number	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative
1	23.554	54.776	54.776	23.554	54.776	54.776	14.331	33.328	33.328
2	5.363	12.471	67.247	5.363	12.471	67.247	6.193	14.402	47.730
3	3.285	7.639	74.886	3.285	7.639	74.886	5.519	12.835	60.564
4	1.982	4.609	79.495	1.982	4.609	79.495	3.610	8.395	68.960
5	1.588	3.693	83.188	1.588	3.693	83.188	3.411	7.932	76.892
6	1.506	3.503	86.691	1.506	3.503	86.691	3.009	6.997	83.889
7	1.278	2.972	89.663	1.278	2.972	89.663	2.338	5.437	89.326
8	0.827	1.924	91.586	0.827	1.924	91.586	0.972	2.260	91.586

下载: 导出CSV

表 7 机械传动系统因子分析总方差解释表

Table 7. Explanatory table of total variance of factor analysis of mechanical transmission system

Component number	Variance of initial eigenvalue			Variance of extraction sums of squared loadings			Rotation sums of squared loadings
Component number	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative
1	13.984	31.075	31.075	13.984	31.075	31.075	9.911	22.025	22.025
2	12.586	27.969	59.044	12.586	27.969	59.044	5.569	12.375	34.400
3	5.001	11.112	70.157	5.001	11.112	70.157	5.349	11.886	46.286
4	2.910	6.467	76.624	2.910	6.467	76.624	4.905	10.899	57.185
5	2.374	5.276	81.900	2.374	5.276	81.900	4.837	10.750	67.935
6	1.798	3.996	85.896	1.798	3.996	85.896	4.303	9.563	77.498
7	1.395	3.101	88.997	1.395	3.101	88.997	3.336	7.414	84.913
8	1.146	2.546	91.542	1.146	2.546	91.542	2.983	6.630	91.542
9	0.991	2.203	93.745	0.991	2.203	93.745	2.829	6.286	90.325
10	0.627	1.394	95.139	0.627	1.394	95.139	2.009	4.465	94.790
11	0.467	1.037	96.176	0.467	1.037	96.176	0.624	1.386	96.176

下载: 导出CSV

表 8 其他附件因子分析总方差解释表

Table 8. Explanatory table of total variance of factor analysis of other accessory equipment

Component number	Variance of initial eigenvalue			Variance of extraction sums of squared loadings			Rotation sums of squared loadings
Component number	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative
1	19.670	42.762	42.762	19.670	42.762	42.762	9.624	20.922	20.922
2	5.181	11.264	54.025	5.181	11.264	54.025	4.630	10.066	30.988
3	4.085	8.880	62.906	4.085	8.880	62.906	4.344	9.443	40.432
4	3.247	7.058	69.964	3.247	7.058	69.964	3.982	8.657	49.089
5	3.114	6.769	76.733	3.114	6.769	76.733	3.592	7.810	56.898
6	2.031	4.414	81.147	2.031	4.414	81.147	3.417	7.429	64.327
7	1.907	4.145	85.292	1.907	4.145	85.292	3.119	6.781	71.109
8	1.385	3.010	88.302	1.385	3.010	88.302	3.041	6.612	77.720
9	1.090	2.370	90.672	1.090	2.370	90.672	2.652	5.764	83.485
10	0.770	1.674	92.345	0.770	1.674	92.345	2.484	5.400	88.885

下载: 导出CSV

表 9 叶盘类因子分析总方差解释表（未考虑自相关性）

Table 9. Explanatory table of total variance of factor analysis of blade and wheel disc(without considering autocorrelation)

Component number	Variance of initial eigenvalue			Variance of extraction sums of squared loadings			Rotation sums of squared loadings
Component number	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative	Total	Percentage/%	Cumulative
1	27.294	63.475	63.475	27.294	63.475	63.475	21.932	51.004	51.004
2	5.661	13.165	76.640	5.661	13.165	76.640	5.511	12.817	63.821
3	3.013	7.008	83.648	3.013	7.008	83.648	4.530	10.534	74.355
4	1.818	4.227	87.875	1.818	4.227	87.875	3.324	7.729	82.085
5	1.261	2.932	90.807	1.261	2.932	90.807	1.735	4.035	86.120
6	0.921	2.142	92.949	0.921	2.142	92.949	1.574	3.660	89.779
7	0.731	1.700	94.649	0.731	1.700	94.649	1.545	3.593	93.373
8	0.483	1.123	95.772	0.483	1.123	95.772	0.861	2.003	95.375

下载: 导出CSV

参考文献(27)

[1]	李其汉, 王延荣, 陈懋章. 航空发动机结构强度设计问题[M]. 上海: 上海交通大学出版社, 2014.
[2]	陶春虎, 钟培道, 王仁智, 等. 航空发动机转动部件的失效与预防[M]. 北京: 国防工业出版社, 2000.
[3]	王荣桥, 胡殿印. 航空发动机结构可靠性设计理论及应用[M]. 北京: 科学出版社, 2017.
[4]	汤建民, 余丰民. 国内知识图谱研究综述与评估: 2004-2010年[J]. 情报资料工作, 2012(1): 16-21.
[5]	张勤, 徐绪松. 共词分析法与可视化技术的结合: 揭示国外知识管理研究结构[J]. 管理工程学报, 2008, 22(4): 30-35, 50. doi: 10.3969/j.issn.1004-6062.2008.04.005
[6]	STOLZER A. Tutorial B-data mining for aviation safety: Using data mining recipe “automatized data mining” from STATISTICA[J]. Handbook of Statistical Analysis & Data Mining Applications, 2009: 377-389.
[7]	王有远, 张振华, 钱伟伟, 等. 基于模糊聚类的航空发动机故障预测研究[J]. 南昌航空大学学报(自然科学版), 2018, 32(1): 23-28.
[8]	王银坤, 唐庆如. 基于灰色聚类分析的FJ44发动机故障诊断技术的研究[J]. 科技创新导报, 2012(22): 15-16. doi: 10.3969/j.issn.1674-098X.2012.22.009
[9]	辜振谱, 刘晓波, 韩子东, 等. 基于改进密度峰值聚类的航空发动机故障诊断[J]. 计算机集成制造系统, 2020, 26(5): 1211-1217.
[10]	舒畅, 程铭, 许煜, 等. 航空发动机压气机叶片外物损伤规律研究[J]. 机械工程学报, 2019, 55(13): 87-94.
[11]	曹惠玲, 曹鹏双, 康力平, 等. 发动机HPT叶片失效分析及损伤发展趋势研究[J]. 中国民航大学学报, 2017, 35(3): 13-16, 26. doi: 10.3969/j.issn.1674-5590.2017.03.004
[12]	刘少武. 材料失效数据库的建立与数据分析[D]. 上海: 华东理工大学, 2018.
[13]	宋爽. 共现分析在文本知识挖掘中的应用研究[D]. 南京: 南京理工大学, 2006.
[14]	邱均平, 董克. 作者共现网络的科学研究结构揭示能力比较研究[J]. 中国图书馆学报, 2014, 40(1): 15-24. doi: 10.3969/j.issn.1001-8867.2014.01.002
[15]	张勤, 马费成. 国外知识管理研究范式——以共词分析为方法[J]. 管理科学学报, 2007(6): 65-75. doi: 10.3321/j.issn:1007-9807.2007.06.008
[16]	隋明爽, 崔雷. 基于SPSS的共现聚类分析参数选择的实例研究[J]. 中华医学图书情报杂志, 2016, 25(1): 52-56. doi: 10.3969/j.issn.1671-3982.2016.01.011
[17]	李迎迎, 郑春厚. 基于共词分析的国内知识组织现状研究[J]. 山东图书馆学刊, 2013(4): 76-80. doi: 10.3969/j.issn.1002-5197.2013.04.020
[18]	周磊, 杨威, 张玉峰. 共现矩阵聚类分析的问题与再思考[J]. 情报杂志, 2014, 33(6): 27, 32-36.
[19]	何彤宇. 基于知识图谱的国内“朝核问题”研究热点分析[J]. 情报杂志, 2017, 36(11): 78-85. doi: 10.3969/j.issn.1002-1965.2017.11.013
[20]	王曰芬, 宋爽, 卢宁, 等. 共现分析在文本知识挖掘中的应用研究[J]. 中国图书馆学报, 2007(2): 59-64. doi: 10.3969/j.issn.1001-8867.2007.02.012
[21]	王冠超. 航空发动机常见故障以及处理措施分析[J]. 黑龙江科技信息, 2016(16): 63.
[22]	张栋善, 赵成. 试分析航空发动机故障诊断技术[J]. 中国新通信, 2019, 21(17): 139-140. doi: 10.3969/j.issn.1673-4866.2019.17.110
[23]	谢广平, 武颖娜, 杨锐. 航空发动机涡轮叶片的失效分析与检测技术[J]. 民用飞机设计与研究, 2019(3): 15-26.
[24]	傅国如, 禹泽民, 王洪伟. 航空涡喷发动机压气机转子叶片常见失效模式的特点与规律[J]. 失效分析与预防, 2006(1): 18-24. doi: 10.3969/j.issn.1673-6214.2006.01.005
[25]	刘德林, 姜涛, 何玉怀, 等. 浅论国内航空轴承的失效问题[J]. 失效分析与预防, 2015, 10(5): 324-330. doi: 10.3969/j.issn.1673-6214.2015.05.012
[26]	孔德龙, 林国昌. 航空发动机主轴轴承主要损伤模式及原因分析[J]. 航空科学技术, 2011(5): 22-24. doi: 10.3969/j.issn.1007-5453.2011.05.007
[27]	LYV J H, WANG W Z, TU S T, et al. Statistical analysis of failure cases in aerospace[C]//ICSFA 2018: 20th International Conference on Structures and Failure Analysis. Tokyo: [s.n.], 2018: 3100-3104.