You do not have permission to edit this page, for the following reason:

You are not allowed to execute the action you have requested.


You can view and copy the source of this page.

x
1
<!-- metadata commented in wiki content
2
3
4
<big>'''A new calculation method of bearing reliability of tyre unloader based on heterogeneous dimensional interference model'''</big>
5
6
==Jingxiu Ling<sup>1,2, 3</sup>, Rongchang Zhang<sup>1,2</sup>, Jiacheng Shao<sup>1,2</sup>and Hao Zhang<sup>4</sup>==
7
8
<sup>1</sup>School of Mechanical and Automotive Engineering, Fujian University of Technology, Fuzhou 350118, China, <sup>2</sup>Fujian Key Laboratory of Intelligent Machining Technology and Equipment (Fujian University of Technology),<sup> 3</sup>School of Materials Science and Engineering, Fujian Institute of Technology, Fuzhou 350118, China,<sup>4</sup>CSCEC Strait Construction Development Co., Ltd., Fuzhou 350000, China.
9
10
Correspondence to: Jingxiu Ling  [mailto:ljxyxj@163.com ljxyxj@163.com]
11
12
<span id='OLE_LINK11'></span><span id='OLE_LINK12'></span>
13
-->
14
==Abstract==
15
16
Bearing is an important rotating support part of tyre unloader, and its fatigue reliability is an important part of the whole system reliability. Because of the huge alternating stress, the support bearing is required to have high fatigue life and reliability. In this paper, combined with stress-strength interference model and statistical theory, the life distribution of bearing steel material is predicted by using group test data; Based on the multi-rigid body dynamics and finite element numerical simulation platform, the reliability of the bearing of tyre unloader under different operating years was predicted by using the different dimensional interference model. The results show that the maximum resultant force of the bearing at the bottom rocker arm of the tyre unloader can reach 150kN, and the maximum transverse and longitudinal forces can reach 108kN and 78kN. When bearing the weight of the whole tyre and turning, the inertia force is the largest, the maximum stress value is 1316.2MPa, which occurs in the bearing inner ring and ball contact part. After the statistics, the stress amplitude distribution of the bearing conforms to Weibull distribution, and the life of the bearing follows lognormal distribution. After 10<sup>5</sup> tyre unloading, the fatigue reliability of the bearing is lower than 0.82, which is consistent with the actual working condition. Therefore, this model can be used to calculate the fatigue reliability of bearings conveniently and quickly, and provide certain theoretical support for the safety and fatigue reliability prediction of bearings.
17
18
'''Keywords''': Numerical simulation, fatigue life, reliability, different dimensional interference model
19
<!--==Article highlights==
20
21
<span id='OLE_LINK1'></span>
22
23
● Reliability prediction model for tyre unloader bearings based on heterogeneous interference theory.
24
25
● A set of fatigue reliability calculation method for the bearing of a tyre unloader is put forward.
26
27
● Probabilistic model for calculating bearing life using mathematical statistical methods.
28
29
● Fatigue life data of bearings obtained by using group method.
30
31
<span id='OLE_LINK2'></span>●Prediction of fatigue reliability of bearings based on the equivalent force method.
32
33
-->
34
35
==1. Introduction==
36
37
With the continuous improvement of China's machinery production process, the life and reliability of bearings has been greatly improved, but the bearings running on some large machinery, because of the huge alternating stress and complex working environment, making the life and reliability of bearings rapidly reduced. At the same time, with the rapid development of heavy machinery in China, the tonnage of tyres used is also rising. The object of this paper is a new type of giant tyre unloading machine, which unloads tyres weighing up to 6t. In the production process, the tyres need to be unloaded from a fixed position and flipped by the unloading machine after the completion of the previous process. During the unloading and flipping process, due to the huge weight of the tyre and the inertia force generated during the flipping, the tyre will collide and rub against the clamping mechanism of the tyre unloader, causing the system to vibrate and at the same time causing the bearings in key parts to be subjected to complex and variable random loads. Because of the randomness of the external load, the bearing material itself performance, size and other variability, the life distribution of bearings belongs to a probability distribution and with the growth of the use of years, bearing failure rate is on the rise. The reliability of the bearing life under different years needs to be studied to prevent bearing failure and subsequent safety accidents. The fatigue reliability of this new type of tyre unloader bearing is not systematically studied. In this paper, the fatigue reliability calculation method of the new tyre unloader bearing is proposed based on the theory of dynamic finite element and different dimensional interference model.
38
39
Current research on tyre unloaders is limited to the control methods and modes of operation. For example, Alessio et al. [1] introduced a robot that can assist tyre operators in the workshop to change tyres, which can be controlled in a variety of ways, such as automatic recognition of the user's gesture commands and remote operation through a control interface. In addition, Ján et al. [2] carried out a 3D modelling and dynamics analysis of a robot that unloads tyres, thus obtaining important mechanical parameters of the robot.
40
41
Existing bearing reliability predictions are different from the subject of this paper, and most of the analysis methods use isotropic interference models, which is the traditional stress-strength interference model. However, the strength distribution of general materials is difficult to predict and difficult to obtain strength distribution data through specific tests, and it needs to be combined with the material life distribution to obtain the strength distribution of materials through complex mathematical calculations, which is not conducive to the use of engineers. Domestic scholars Xie  and Wang  [3] proposed a heterogeneous interference model, which is no longer limited to the traditional reliability model in which the two variables must be of the same magnitude, but can be used to calculate the reliability through the stress distribution and the life distribution of the material. At the same time, the model no longer relies on tedious mathematical calculations but obtains the fatigue life distribution of the material through tests, which makes the fatigue reliability analysis much less difficult and simplifies the analysis steps. Liu et al. [4] provided a new analysis method for fatigue reliability of pipe structures by combining fatigue reliability of pipes under load with structural dynamic analysis. Zhou et al. [5] provided a theoretical basis for the life assessment and reliability analysis of rolling bearing systems through the fatigue reliability calculation formula of bearing systems under continuous load spectrum. Jin et al. [6] proposed an artificial intelligence method to analyze the fatigue reliability of aviation bearings. Qi and Liao [7] designed a system to evaluate the reliability of rolling bearings of traction motors based on MATLAB App Designer, which provided a new reference for the development and improvement of reliability evaluation systems.
42
43
Reuben et al. [8] improved this reliability assessment based on the Weibull diagram equation by estimating the variation in minimum bearing life and establishing confidence intervals using Monte Carlo simulations. Pape et al. [9] improved the calculation of bearing fatigue life by introducing residual stresses to the sub-surface region of the bearing. Wang [10] calculated the reliability life of ship unloader bearings by converting the tensile force of the ship unloader lifting mechanism into an average equivalent dynamic load on the bearings and then using the Miner criterion. Xia Xintao and Ye Liang et al. [11] proposed a new concept of rolling bearing performance retention reliability. Zhang et al. [12] established an analytical model for the contact fatigue reliability of main bearings based on the complex loading conditions of TBM main bearings and combined with fatigue cumulative damage and residual strength theory, based on a dynamic degradation model under complex conditions. Cheng et al. [13] developed a modified five-degree-of-freedom quasi-dynamic model considering multi-body interactions and used the modified fatigue life model proposed by Jones to assess the effect of angular displacement on bearing reliability. Herp et al. [14] proposed a bearing condition prediction method using temperature residuals and Bayesian probability statistics for wind turbine bearings, and this method can predict possible bearing failures within an average of 33 days. Tong et al. [15] proposed an improved model for calculating the operating torque of angular contact ball bearings (ACBB) by comparing the effect of different loading conditions on the fatigue reliability of angular misalignment on tapered roller bearings. König et al. [16] developed a model of bearing life which can determine the merits of hybrid bearings and allows for an analytical assessment of bearing life. Guillermo et al. [17] further extended a previously developed model for calculating bearing life based on high cycle fatigue by incorporating a new surface damage integral based on a creep mechanism into the model, providing a new approach to fatigue life prediction for bearings. Zhang et al. [18] used the maximum likelihood estimation method to calculate the actual value of the bearing, and used the SPSS curve and cumulative grey prediction model methods to train on some of the real data as a way of predicting the life of the bearing, and found that all three methods have some practical value in engineering, but the cumulative grey prediction was more effective. Lorenz et al. [19] developed a continuous damage mechanics (CDM) finite element (FE) model in order to investigate the effect of surface roughness on the rolling contact fatigue life of poor quality contact bodies and demonstrated the feasibility of the model. Pandey et al. [20] proposed a framework based on continuum damage mechanics and the finite element method to simulate the low circumferential fatigue crack expansion process, and developed a strain-based damage model to consider the effect of different strain ratios on fatigue damage. Cano et al. [21] developed a new intrinsic model for long-term prediction of creep deformation, damage and rupture by combining the Wilshire equation with continuum damage mechanics (CDM).
44
45
In summary, in terms of bearing reliability analysis and calculation, numerical analysis, theoretical calculation and experimental evaluation are mainly used at home and abroad to study the performance and life of various types of bearings, but there are fewer studies combining giant tyre unloaders and using heterogeneous interference models to analyse their bearing reliability. Therefore, this paper combines the existing information at home and abroad, mainly to fill the gap in the field of research on the fatigue reliability of bearings of giant tyre unloaders; Prediction of the life distribution of bearing steel materials using grouping method test data combined with statistical theory; Combining numerical simulation platforms such as dynamics and finite elements, the heterogeneous interference model is used to predict the reliability of bearings at different service lives. It also differs from traditional stress-strength interference models in that it directly uses the material life distribution for reliability calculations and is suitable for generalised applications in engineering.
46
47
==2. Reliability analysis process ==
48
49
<span id='_Ref96587959'></span>The fatigue reliability of the bearings of the giant tyre unloading machine is analysed and calculated by the numerical analysis platforms such as dynamics and finite element, using the theory related to fatigue reliability. The specific process is shown in [[#img-1|Figure 1]].
50
51
<div id='img-1'></div>
52
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;width:auto;" 
53
|-style="background:white;"
54
|style="text-align: center;padding:10px;"| [[File:Ling_et_al_2023a_3475_600px-Draft_Ling_717181314-image1-c.png]]
55
|-
56
| style="background:#efefef;text-align:left;padding:10px;font-size: 85%;"| '''Figure 1'''. Flow chart for calculating the fatigue reliability of a tyre unloader bearing
57
|}
58
59
60
==3. Maximum stress history of bearing==
61
62
===3.1 Structure and modelling of the tyre unloader===
63
64
The physical drawing of the tyre unloading machine is shown in [[#img-2|Figure 2]]. Because of the complex structure of the tyre unloader proper, the model (with tyre 2) is shown in [[#img-3|Figure 3]] by simplifying the 3D modeling of the threads and other parts of the tyre unloader blank that will not have an impact on the final result.
65
66
<div id='img-2'></div>
67
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;width:auto;" 
68
|-style="background:white;"
69
|style="text-align: center;padding:10px;"| [[File:Draft_Ling_717181314-image2.png|centre]]
70
|-
71
| style="background:#efefef;text-align:left;padding:10px;font-size: 85%;"| '''Figure 2'''. Physical view of the tyre unloader
72
|}
73
74
75
<div id='img-3'></div>
76
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;width:auto;" 
77
|-style="background:white;"
78
|style="text-align: center;padding:10px;font-size: 75%;"| [[Image:Draft_Ling_717181314-image3.png]]
79
80
1. Clamping plate; 2. Tyres; 3. Guide wheels; 4. Fixed ring; 5. Dynamic ring; 6. Fixed ring rotary; 7. Rocker arm
81
|-
82
| style="background:#efefef;text-align:left;padding:10px;font-size: 85%;"| '''Figure 3'''. 3D model of the tyre unloader
83
|}
84
85
86
The structure of the tyre unloader mainly consists of the Plywood 1, the Guide wheel 3, the Fixed ring 4, the Dynamic ring 5, the Rocker arm 7, etc. 16 pairs of rocker arms are mounted symmetrically on each side of the fixed ring, with the upper end connected to the moving ring by a linkage and the lower end connected to the cleat. The principle of operation is that the dynamic ring rotates at a certain angle with respect to the fixed ring, thus controlling the clamping plate on the rocker arm to clamp or unclamp the tyre. The fixed ring is equipped with 8 guide wheels, which support the fixed ring and guide it through a certain angle of rotation, while the rotation of the fixed ring and the turning of the machine are controlled by an electric motor.
87
88
===3.2 Numerical simulation of the operating conditions of the tyre unloader===
89
90
The above 3D model of the tyre unloader (including tyre) is imported into the dynamics analysis software ADAMS, and the material parameters of the model are set first. The material parameters of the whole machine and tyres are shown in [[#tab-1|Tables 1]] and [[#tab-2|2]].
91
92
<div class="center" style="font-size: 75%;">'''Table 1'''. Material parameters for tyre unloaders</div>
93
94
<div id='tab-1'></div>
95
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;font-size:85%;width:auto;" 
96
|-style="text-align:center"
97
!  Materials !! Elastic modulus (MPa) !! Poisson ratio !! Density (kg/m<sup>3</sup>)
98
|-style="text-align:center"
99
|  Rubber
100
|  7.8
101
|  0.29
102
|  1200
103
|}
104
105
106
<div class="center" style="font-size: 75%;">'''Table 2'''.  Tyre material parameters></div>
107
108
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;font-size:85%;width:auto;" 
109
|-style="text-align:center"
110
!  Materials !! Elastic modulus (MPa) !! Poisson ratio !! Density (kg/m<sup>3</sup>)
111
|-style="text-align:center"
112
|  Carbon structural steel
113
|  207
114
|  0.29
115
|  7801
116
|}
117
118
119
Constraint and actuation of each mechanism according to the actual motion of the tyre unloading machine. The whole mechanism of tyre unloading machine is mainly rotating sub and the connecting rod articulation, plywood and rocker arm and other rotating connections are rotating sub.The guidewheel shaft is fixed to the fixed ring using a fixed pair, and a rotating connection pair is applied between the guidewheel and the guidewheel shaft.The fixed ring and the moving ring are provided with a planar pair to prevent relative slippage caused by the rotation of the moving ring. To prevent over-constraints caused by the use of too many rotating subsets, resulting in instability of the solved system,the conflict constraints are replaced by primary subsets such as point overlap subsets or cylindrical amplitude, and coplanar subsets replace some of these planar subsets to ensure the stability and accuracy of the solved system.The virtual prototype of the tyre unloader is shown in [[#img-4|Figure 4]].
120
121
<div id='img-4'></div>
122
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;width:auto;" 
123
|-style="background:white;"
124
|style="text-align: center;padding:10px;"| [[Image:Draft_Ling_717181314-image4.png|441x441px]]
125
|-
126
| style="background:#efefef;text-align:left;padding:10px;font-size: 85%;"| '''Figure 4'''. Virtual prototype of the tyre unloader 
127
|}
128
129
130
After building a virtual prototype of the tyre unloader system, the simulation is based on the actual working conditions of the tyre unloader. The simulation time is set as a cycle of 2.1s. After the simulation, the load spectrum of <math display="inline">X  </math>, <math display="inline"> Y </math> and <math display="inline"> Z </math> directions of the bearing at the bottom rocker arm is extracted, and the extraction position is illustrated in [[#img-5|Figure 5]]. The loads in the <math display="inline">X  </math> and Y directions are transverse and longitudinal loads, and the loads in the <math display="inline"> Z </math> direction are those perpendicular to the bearing outward. The extraction results are shown in [[#img-6|Figure 6]].
131
132
<div id='img-5'></div>
133
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;width:auto;" 
134
|-style="background:white;"
135
|style="text-align: center;padding:10px;"| [[Image:Draft_Ling_717181314-image5.png]]
136
|-
137
| style="background:#efefef;text-align:left;padding:10px;font-size: 85%;"| '''Figure 5'''. Load extraction at rocker arm bearing 
138
|}
139
140
141
<div id='img-6'></div>
142
{| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;width:auto;" 
143
|-style="background:white;"
144
|style="text-align: center;padding:10px;"| [[Image:Draft_Ling_717181314-image6.png]]
145
|-
146
| style="background:#efefef;text-align:left;padding:10px;font-size: 85%;"| '''Figure 6'''. Three-way force load spectrum at rocker arm bearing
147
|}
148
149
150
According to the three-way force load spectrum, the average load in <math display="inline">X  </math> direction and <math display="inline"> Y </math> direction of the bearing tends to 0 within 0.8s, and there is a certain impact load, which is caused by the vibration generated when the splint contacts the tyre. After the splint clamps the tyre, the fixing pair will temporarily fail. As the weight of the tyre itself is transferred to the bearing, the load values of the <math display="inline">X  </math> and <math display="inline"> Y </math> directions of the bearing at the bottom rocker arm will rise rapidly. After about 0.2s, the components of the <math display="inline">X  </math> and <math display="inline"> Y </math> directions increase and remain at about 108kN and 78kN respectively.
151
152
==3.3 Validation of simulation results==
153
154
Because the simulation analysis model is simplified to some extent, the simulation results have some errors.Therefore, in order to verify the reasonableness and realism of the simulation results and to provide a reasonable dynamic load spectrum for the subsequent analysis and prediction of the bearing reliability. Through the proportional coefficient of the real-time output torque of the clamping motor on the control panel of the tyre unloader (as shown in Fig. 7), the output torque coefficient of the motor is converted into the actual moving ring thrust of the tyre unloader and compared with the simulation value of ADAMS, and the reasons for the error between the actual value and the simulation result are analyzed.
155
156
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
157
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image7-c.png]] </span></div>
158
159
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
160
<span style="text-align: center; font-size: 75%;">Fig. 7 Tyre unloader control panel</span></div>
161
162
The comparison data in the test is the thrust of the tyre unloading maneuver ring, which is the current torque value of the birth ring in the picture. The output proportional coefficient needs to be converted into the real thrust force on the moving ring, namely the lead screw thrust. The relation between output torque of reducer and motor parameters is shown in Eq.(1) .
163
164
{| class="formulaSCP" style="width: 100%; text-align: center;" 
165
|-
166
| 
167
{| style="text-align: center; margin:auto;" 
168
|-
169
| <span style="text-align: center; font-size: 75%;"> <math display="inline">T_a=9550\frac{P\cdot r\cdot \eta }{n}</math> </span>
170
|}
171
| style="width: 5px;text-align: right;white-space: nowrap;" | (1)
172
|}
173
174
175
Where P represents the motor power, which is 0.75kW;
176
177
''N'' represents the motor speed, which is 1500r/min;
178
179
''η ''represents the transmission efficiency, which is 98%;
180
181
''R'' represents the speed ratio of motor and reducer, which is 20:1;
182
183
The maximum output torque ''T<sub>a</sub>'' of reducer is 94N m.
184
185
The conversion formula between output torque of reducer and thrust Fa of the lead screw is shown in Eq. (2). The maximum thrust force Fa of the screw is calculated from the maximum output torque to be 56 N. The motor torque output scaling factor is converted into a value for the change in thrust of the kinematic ring, the torque factor and the actual kinematic ring thrust values are shown in Table 3.<span id='cite-_Ref96589010'></span>[[#_Ref96589010|]]
186
187
{| class="formulaSCP" style="width: 100%; text-align: center;" 
188
|-
189
| 
190
{| style="text-align: center; margin:auto;" 
191
|-
192
| <math display="inline">T_a=\frac{F_a\times t}{2\times 3.14\times {\eta }_1}</math>
193
|}
194
| style="width: 5px;text-align: right;white-space: nowrap;" | (2)
195
|}
196
197
198
Where t represents the lead of the lead screw, which is 10mm;
199
200
''η1'' represents the conversion efficiency, which is 95%.
201
202
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
203
<span style="text-align: center; font-size: 75%;">Table 3 Proportional coefficient of output torque and actual thrust value of moving ring (part)</span></div>
204
205
{| style="width: 47%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
206
|-
207
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">Proportional coefficient of output torque (‰)</span>
208
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">Actual moving ring</span>
209
210
<span style="text-align: center; font-size: 75%;">thrust (N)</span>
211
|-
212
|  style="border-top: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">232</span>
213
|  style="border-top: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">12.99 </span>
214
|-
215
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">239</span>
216
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">13.38 </span>
217
|-
218
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">244</span>
219
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">13.66 </span>
220
|-
221
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">250</span>
222
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">14.00 </span>
223
|-
224
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">258</span>
225
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">14.45 </span>
226
|-
227
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">262</span>
228
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">14.67 </span>
229
|-
230
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">269</span>
231
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">15.06 </span>
232
|-
233
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">276</span>
234
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">15.46 </span>
235
|-
236
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">283</span>
237
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">15.85 </span>
238
|-
239
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">289</span>
240
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">16.18 </span>
241
|-
242
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">295</span>
243
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">16.52 </span>
244
|-
245
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">304</span>
246
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">17.02 </span>
247
|-
248
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">302</span>
249
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">16.91 </span>
250
|-
251
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">309</span>
252
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">17.30 </span>
253
|-
254
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">317</span>
255
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">17.75 </span>
256
|-
257
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">321</span>
258
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">17.98 </span>
259
|-
260
|  style="border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">327</span>
261
|  style="border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">18.31 </span>
262
|}
263
264
265
<span id='OLE_LINK13'></span>
266
267
The drive torque was simulated by ADAMS with the drive position at the center of the tyre unloader. Based on the actual radius of the tyre unloader being approximately 1 meter, it can be assumed that the dynamic ring thrust value is numerically equal to the drive torque value. The actual calculated dynamic ring thrust variation was compared with the simulated value of the tyre unloader motor drive ring and the comparison results are shown in Fig<span id='cite-_Ref26386'></span>. 8.
268
269
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
270
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image10.png]] </span></div>
271
272
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
273
<span style="text-align: center; font-size: 75%;">Fig. 8 Comparison of simulated and actual values</span></div>
274
275
Due to the difference between the simulated and actual driving methods of the dynamic loops, the simulated and actual values are subject to certain errors. In ADAMS, a rotational drive is applied directly to the kinematic ring, and the simulation results from Fig. 8 show that the torque required to move the kinematic ring from rest to motion is large (0 to 0.1s), after which the drive torque required decreases (0.1 to 7.9s) due to the inertia of the kinematic ring itself. To maintain the movement, the torque will gradually increase, during which the collision between the moving ring and the guide wheel will also cause a sudden increase in torque, producing a peak (7.9 to 14s). In the actual working of the tyre unloader, the drive unit consists of three main parts: motor, reducer and screw. The specific mode of operation is by connecting the reducer to the motor in order to reduce the rotational speed and thus increase the output torque, which is converted into thrust on the screw by means of a screw to drive the moving ring for rotation. This process allows for customised settings of motor speed, output torque magnitude, etc. The torque magnitude is negligibly influenced by the interaction between the mechanisms during operation of the device. So the screw thrust tends to rise gradually and with little fluctuation. But the simulated force values are of the same order of magnitude as the actual force values, and the overall trend is gradually increasing, which proves that the simulation results have a certain degree of reasonableness.
276
277
==3.4 Bearing transient dynamics analysis==
278
279
<span id='OLE_LINK5'></span>From the results of the kinetic analysis, it can be seen that the bearing at the lowermost rocker is subjected to the greatest component of gravity of the tyre blank and is also furthest from the centre of rotation, where the centrifugal inertia force is greatest. The bearing selected here for analysis is therefore a deep groove ball bearing, type 61918. The bearing model is shown in Fig. 9. According to the actual bearing size parameters, the 3D modelling software is used to establish a 3D model of the bearing at the rocker arm. In order to simplify the calculation steps and save calculation resources, non-important parts such as the cage are omitted, the bearing dimensional parameters are shown in Table 4.
280
281
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
282
 [[Image:Draft_Ling_717181314-image11.png]] </div>
283
284
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
285
<span style="text-align: center; font-size: 75%;">Fig. 9 3D Model of bearing 61918</span></div>
286
287
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
288
<span style="text-align: center; font-size: 75%;">Table 4 Bearing 61918 dimensional parameters</span></div>
289
290
{| style="width: 47%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
291
|-
292
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Bearing type</span>
293
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Outer diameter(mm)</span>
294
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Inner diameter(mm)</span>
295
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Thickness(mm)</span>
296
|-
297
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">61918</span>
298
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">125</span>
299
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">90</span>
300
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">18</span>
301
|}
302
303
304
The bearing model was imported into Ansys finite element analysis software, the inner and outer rings of the bearing were constrained, and the bearing material parameters were set, as shown in Table 5. The extracted X, Y and Z forces are loaded onto the corresponding parts of the bearing for transient dynamics analysis, with the loads and constraints applied as shown in Fig. 10. The results of the transient dynamics analysis (equivalent stress clouds) are shown in Fig. 11.
305
306
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
307
<span style="text-align: center; font-size: 75%;">Table 5 Bearing material parameters</span></div>
308
309
{| style="width: 100%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
310
|-
311
|  colspan='2'  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span id='_Hlk103109913'></span><span style="text-align: center; font-size: 75%;">Materials</span>
312
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Elastic Modulus</span>
313
314
<span style="text-align: center; font-size: 75%;">(MPa)</span>
315
|  colspan='2'  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Poisson</span>
316
317
<span style="text-align: center; font-size: 75%;"> ratio</span>
318
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Yield Strength(MPa)</span>
319
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Tensile strength(MPa)</span>
320
|-
321
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">GCr15</span>
322
|  colspan='3'  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">210000</span>
323
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.29</span>
324
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1458</span>
325
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1617</span>
326
|}
327
328
329
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
330
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image12.png|709x709px]] </span></div>
331
332
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
333
<span style="text-align: center; font-size: 75%;">Fig. 10 Bearing load and constraint settings</span></div>
334
335
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
336
 [[Image:Draft_Ling_717181314-image13.png|552x552px]] </div>
337
338
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
339
<span style="text-align: center; font-size: 75%;">Fig. 11 Cloud diagram of maximum equivalent stress of bearing</span></div>
340
341
The bearing of the lowermost rocker arm of the tyre unloader is subjected to the greatest inertial force when carrying the weight of the entyre tyre and turning it over. From the results of the transient dynamics analysis, it can be seen that the maximum stress value that the bearing is subjected to is 1316.2Mpa, so the part of the bearing where the stress is high and easy to produce fatigue is the inner ring and ball contact.
342
343
==3.5 Stress history in hazardous areas==
344
In order to obtain the distribution characteristics of the stresses in the hazardous parts of the bearings, the ANSYS APDL was used to extract the nodal stress time histories corresponding to the fatigue parts of the bearings according to the results of the transient dynamics. The curve for the change in stress history is shown in Fig. 12.<span id='cite-_Ref27859'></span><div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
345
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image14.png|704x704px]] </span></div>
346
347
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
348
<span style="text-align: center; font-size: 75%;">Fig. 12 Stress time history in fatigue bearing areas</span></div>
349
350
=4. Bearing reliability calculation based on heterogeneous interference model=
351
352
==4.1 Stress history rain flow count statistics==
353
354
From the basic theory of fatigue, it can be seen that the load average, amplitude and the number of cycles are the main factors that make the component produce fatigue damage, so the bearing fatigue part of the stress time course need to be cycle counting process. In this paper, the two-parameter rainfall counting method is used to count the amplitude and mean values of the loads and to obtain important relationships between the load amplitude, mean value and the corresponding frequency.
355
356
Rain flow counting of the stress history was carried out to convert the random variable amplitude stress into a series of load cycles, the amplitude and mean distribution of the load cycles were counted and the mean and amplitude were converted into a two-dimensional histogram and the statistical results are shown in Fig. 13- Fig. 15.
357
358
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
359
 [[Image:Draft_Ling_717181314-image15-c.png]] </div>
360
361
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
362
<span style="text-align: center; font-size: 75%;">Fig. 13 Stress history rain flow counting</span></div>
363
364
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
365
 [[Image:Draft_Ling_717181314-image16.png]] </div>
366
367
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
368
<span style="text-align: center; font-size: 75%;">Fig. 14 Stress amplitude frequency statistics</span></div>
369
370
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
371
 [[Image:Draft_Ling_717181314-image17.png]] </div>
372
373
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
374
<span style="text-align: center; font-size: 75%;">Fig. 15 Stress mean frequency statistics</span></div>
375
376
Using Goodman's theory to apply an average stress correction to the stress amplitude, the stress state is converted to a load cycle with a stress ratio of -1, according to an equal life, the corrected load cycles were used as equivalent stresses for the next step of the fatigue reliability analysis. The Goodman formula is shown in Eq. (3) and the corrected results are shown in Fig. 16.
377
378
{| class="formulaSCP" style="width: 100%; text-align: center;" 
379
|-
380
| 
381
{| style="text-align: center; margin:auto;" 
382
|-
383
| <math display="inline">\frac{{\sigma }_a}{{\sigma }_{-1}}+\frac{{\sigma }_m}{{\sigma }_b}=</math><math>1</math>
384
|}
385
| style="width: 5px;text-align: right;white-space: nowrap;" | (3)
386
|}
387
388
389
Where ''σ<sub>a</sub>'' denotes the actual working stress amplitude (MPa).
390
391
''σ<sub>m</sub>'' indicates the average stress in actual working conditions (Mpa).
392
393
''σ<sub>b</sub>'' indicates the tensile ultimate strength of the material (MPa).
394
395
''σ<sub>-1</sub>'' denotes the stress magnitude (MPa) for a stress ratio of -1.
396
397
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
398
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image19.png]] </span></div>
399
400
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
401
<span style="text-align: center; font-size: 75%;">Fig. 16 Modified equivalent load frequency statistics</span></div>
402
403
==4.2 Load amplitude probability distribution fitting and test==
404
405
Due to the influence of factors such as short sampling time and large amount of data in compiling load spectrum, the load history obtained cannot fully reflect the actual load history of the bearing in the whole life stage, therefore, in order to improve the reliability of the results, the probability distribution function of random loads should be calculated, and then the load distribution of bearings in the whole life interval should be predicted.
406
407
According to the rain-flow counting statistics of stress time history in fatigue parts of bearings, the probability statistics of stress amplitude are carried out and fitted by Weibull distribution. The fitting effect and test results are shown in Fig. 17. It can be seen that the distribution of stress amplitudes generally conforms to the Weibull distribution.
408
409
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
410
 [[Image:Draft_Ling_717181314-image20.png]] </div>
411
412
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
413
<span style="text-align: center; font-size: 75%;">Fig. 17 Fitting magnitude of the Weibull distribution</span></div>
414
415
In order to verify whether the stress amplitude conforms to Weibull distribution, Fig. 18 is obtained by the graphical method. It can be concluded from the figure that the closer the data is to a straight line, the better the data fitting effect is. Therefore, it can be considered that the probability density function of stress amplitude conforms to Weibull distribution.
416
417
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
418
 [[Image:Draft_Ling_717181314-image21.png]] </div>
419
420
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
421
<span style="text-align: center; font-size: 75%;">Fig. 18 Weibull diagramming test</span></div>
422
423
Using the great likelihood method to solve for the Weibull correlation parameters, the relevant solution parameters are shown in Table 6.
424
425
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
426
<span style="text-align: center; font-size: 75%;">Table 6 Equivalent load distribution parameters</span></div>
427
428
{| style="width: 81%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
429
|-
430
|  rowspan='1' style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Scale parameterse parameters</span>
431
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Shape parameters</span>
432
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Confidence level</span>
433
|-
434
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">61.93</span>
435
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.729</span>
436
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.05</span>
437
|}
438
439
440
<span style="text-align: center; font-size: 75%;">The equivalent load probability density function is:</span>
441
442
{| class="formulaSCP" style="width: 100%; text-align: center;" 
443
|-
444
| 
445
{| style="text-align: center; margin:auto;" 
446
|-
447
| <span id='_Hlk106116567'></span><div id="_Hlk110176295" style="text-align: right; direction: ltr; margin-left: 1em;">
448
<span style="text-align: center; font-size: 75%;"> <math display="inline">f\left(x\right)=\frac{\beta x^{\beta -1}}{{\theta }^{\beta }}exp\left[-\right. </math><math>\left. {\left(\frac{x}{\theta }\right)}^{\beta }\right]</math> </span>
449
|}
450
| style="width: 5px;text-align: right;white-space: nowrap;" | (4)
451
|}
452
453
454
Where β denotes the shape parameter, θ denotes the scale parameter. Substituting the two into Eq. (4) gives:
455
456
{| class="formulaSCP" style="width: 100%; text-align: center;" 
457
|-
458
| 
459
{| style="text-align: center; margin:auto;" 
460
|-
461
| <math display="inline">f\left(S\right)=\frac{0.72S^{-0.28}}{19.51}exp\left[-\right. </math><math>\left. {\left(\frac{S}{61.93}\right)}^{0.72}\right]</math>
462
|}
463
| style="width: 5px;text-align: right;white-space: nowrap;" | (5)
464
|}
465
466
==4.3 Material life distribution of bearing steel under arbitrary stress==
467
468
A set of fatigue test data for bearing steel was obtained from the literature[22] and combined with Basquin's equation to derive a continuous probability life model for bearing steel using the great likelihood method. The test material was a high-carbon chromium bearing steel with the chemical composition shown in Table 7, which was machined to give the dimensions in Fig. 19. Then using the grouping method, a four-linked cantilever beam type rotating bending fatigue tester was used to carry out the test, starting with a load of 1700 MPa stress. By reducing the weights so that the stresses fell at intervals of 50 to 100 MPa, nine sets of fatigue test data were eventually obtained, as shown in Table 8.
469
470
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
471
 [[Image:Draft_Ling_717181314-image24.jpeg]] </div>
472
473
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
474
<span style="text-align: center; font-size: 75%;">Fig. 19 Test steel dimensions</span></div>
475
476
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
477
<span style="text-align: center; font-size: 75%;">Table 7 Chemical composition of high-carbon chromium bearing steel</span></div>
478
479
{| style="width: 59%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
480
|-
481
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Element</span>
482
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">C</span>
483
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Si</span>
484
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Mn</span>
485
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Cr</span>
486
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Cu</span>
487
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Ni</span>
488
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Mo</span>
489
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">P</span>
490
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">S</span>
491
|-
492
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Content %</span>
493
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1.01</span>
494
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.23</span>
495
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.36</span>
496
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1.45</span>
497
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.06</span>
498
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.04</span>
499
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.02</span>
500
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.01</span>
501
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.007</span>
502
|}
503
504
505
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
506
<span style="text-align: center; font-size: 75%;">Table 8 Fatigue test date</span></div>
507
508
{| style="width: 58%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
509
|-
510
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Stress(MPa)</span>
511
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Life</span>
512
513
<span style="text-align: center; font-size: 75%;">(cycle)</span>
514
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Stress</span>
515
516
<span style="text-align: center; font-size: 75%;">(MPa)</span>
517
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Life</span>
518
519
<span style="text-align: center; font-size: 75%;">(cycle)</span>
520
|-
521
| rowspan="3" |<span style="text-align: center; font-size: 75%;">1700</span>
522
|<span style="text-align: center; font-size: 75%;">4480</span>
523
| rowspan="2" |<span style="text-align: center; font-size: 75%;">1300</span>
524
|<span style="text-align: center; font-size: 75%;">2480000</span>
525
|-
526
|<span style="text-align: center; font-size: 75%;">6100</span>
527
|<span style="text-align: center; font-size: 75%;">3870000</span>
528
|-
529
|<span style="text-align: center; font-size: 75%;">6840</span>
530
| rowspan="10" |<span style="text-align: center; font-size: 75%;">1250</span>
531
|<span style="text-align: center; font-size: 75%;">40560</span>
532
|-
533
| rowspan="3" |<span style="text-align: center; font-size: 75%;">1600</span>
534
|<span style="text-align: center; font-size: 75%;">5200</span>
535
|<span style="text-align: center; font-size: 75%;">52430</span>
536
|-
537
|<span style="text-align: center; font-size: 75%;">7790</span>
538
|<span style="text-align: center; font-size: 75%;">65990</span>
539
|-
540
|<span style="text-align: center; font-size: 75%;">10000</span>
541
|<span style="text-align: center; font-size: 75%;">128800</span>
542
|-
543
| rowspan="5" |<span style="text-align: center; font-size: 75%;">1500</span>
544
|<span style="text-align: center; font-size: 75%;">8990</span>
545
|<span style="text-align: center; font-size: 75%;">2486740</span>
546
|-
547
|<span style="text-align: center; font-size: 75%;">14400</span>
548
|<span style="text-align: center; font-size: 75%;">4983540</span>
549
|-
550
|<span style="text-align: center; font-size: 75%;">34700</span>
551
|<span style="text-align: center; font-size: 75%;">8950470</span>
552
|-
553
|<span style="text-align: center; font-size: 75%;">247000</span>
554
|<span style="text-align: center; font-size: 75%;">10325090</span>
555
|-
556
|<span style="text-align: center; font-size: 75%;">609000</span>
557
|<span style="text-align: center; font-size: 75%;">11730550</span>
558
|-
559
| rowspan="7" |<span style="text-align: center; font-size: 75%;">1400</span>
560
|<span style="text-align: center; font-size: 75%;">13460</span>
561
|<span style="text-align: center; font-size: 75%;">26058040</span>
562
|-
563
|<span style="text-align: center; font-size: 75%;">15900</span>
564
| rowspan="8" |<span style="text-align: center; font-size: 75%;">1200</span>
565
|<span style="text-align: center; font-size: 75%;">125280</span>
566
|-
567
|<span style="text-align: center; font-size: 75%;">17220</span>
568
|<span style="text-align: center; font-size: 75%;">269929</span>
569
|-
570
|<span style="text-align: center; font-size: 75%;">40800</span>
571
|<span style="text-align: center; font-size: 75%;">672410</span>
572
|-
573
|<span style="text-align: center; font-size: 75%;">839140</span>
574
|<span style="text-align: center; font-size: 75%;">8771540</span>
575
|-
576
|<span style="text-align: center; font-size: 75%;">1540000</span>
577
|<span style="text-align: center; font-size: 75%;">19548550</span>
578
|-
579
|<span style="text-align: center; font-size: 75%;">4970760</span>
580
|<span style="text-align: center; font-size: 75%;">41502050</span>
581
|-
582
| rowspan="4" |<span style="text-align: center; font-size: 75%;">1300</span>
583
|<span style="text-align: center; font-size: 75%;">15360</span>
584
|<span style="text-align: center; font-size: 75%;">43386180</span>
585
|-
586
|<span style="text-align: center; font-size: 75%;">22670</span>
587
|<span style="text-align: center; font-size: 75%;">61683430</span>
588
|-
589
|<span style="text-align: center; font-size: 75%;">39830</span>
590
| rowspan="2" |<span style="text-align: center; font-size: 75%;">1150</span>
591
|<span style="text-align: center; font-size: 75%;">23276820</span>
592
|-
593
|<span style="text-align: center; font-size: 75%;">2050180</span>
594
|<span style="text-align: center; font-size: 75%;">47700000</span>
595
|}
596
597
598
Due to some differences in the actual size, shape and working conditions of the tyre unloader bearings used in this study and the shape and loading method used for the test specimens, therefore, the S-N relationship for the test material needs to be corrected to the S-N relationship for the actual zero component, and the correction formula is as follows:
599
600
{| class="formulaSCP" style="width: 100%; text-align: center;" 
601
|-
602
| 
603
{| style="text-align: center; margin:auto;" 
604
|-
605
| <math display="inline">S_a=\frac{{\sigma }_a}{K_f}\epsilon \beta C_L</math>
606
|}
607
| style="width: 5px;text-align: right;white-space: nowrap;" | (6)
608
|}
609
610
611
Where,'' '' <math display="inline">{\sigma }_{a}</math>'' ''corresponds to the stress tested in the material;
612
613
<math display="inline">{S}_{a}</math> corresponds to zero component stress;
614
615
<math display="inline">\epsilon</math>  is the dimensional coefficient of the material, the value of 0.856 was taken by consulting the 《Mechanical Design Manual》.
616
617
''β'' is the material surface quality factor, here taken as 1.
618
619
<math display="inline">{C}_{L}</math> is the loading method for the workpiece and the steel tension is taken as 0.85.
620
621
The fatigue notch coefficient K_f is related to the stress concentration coefficient, the bearing stress concentration is generally related to the roughness, here it can be considered that the bearing is consistent with the roughness of the test piece, take the value of 1, the correction results are shown in Table 9.
622
623
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
624
<span style="text-align: center; font-size: 75%;">Table 9 Fatigue test data after stress correction</span></div>
625
626
{| style="width: 48%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
627
|-
628
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span id='_Hlk106895464'></span><span style="text-align: center; font-size: 75%;">Test stress(MPa)</span>
629
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Corrected stress(MPa)</span>
630
|-
631
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1700</span>
632
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1030</span>
633
|-
634
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1600</span>
635
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">970</span>
636
|-
637
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1500</span>
638
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">909</span>
639
|-
640
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1400</span>
641
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">849</span>
642
|-
643
|  style="border-top: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1300</span>
644
|  style="border-top: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">788</span>
645
|-
646
|  style="text-align: center;"|<span style="text-align: center; font-size: 75%;">1250</span>
647
|  style="text-align: center;"|<span style="text-align: center; font-size: 75%;">758</span>
648
|-
649
|  style="border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1200</span>
650
|  style="border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">728</span>
651
|-
652
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">1150</span>
653
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">697</span>
654
|}
655
656
==4.4 Maximum likelihood method for determining the P-S-N curve of bearing steel==
657
658
The general material fatigue life follows a log-normal distribution. Assuming that the bearing steel life follows a log-normal distribution, the Anderson-Darling hypothesis test was performed on the fatigue test data at different stresses in Table 8. As the A-D test can take different confidence intervals for testing, the smaller the confidence interval, the more obvious the test effect. So 95% confidence intervals were taken for hypothesis testing of each group of fatigue data, and the test results are shown in Fig. 20- Fig. 21(partial) and Table 10(all).
659
660
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
661
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image26-c.png]] </span></div>
662
663
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
664
<span style="text-align: center; font-size: 75%;">Fig. 20 A-D test results for material life at a stress of 1030 MPa</span></div>
665
666
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
667
 [[Image:Draft_Ling_717181314-image27-c.png]] </div>
668
669
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
670
<span style="text-align: center; font-size: 75%;">Fig. 21  A-D test results for material life at a stress of 909 MPa</span></div>
671
672
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
673
<span style="text-align: center; font-size: 75%;">Table 10 Results of the A-D test for life under different stresses</span></div>
674
675
{| style="width: 49%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
676
|-
677
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">Proof stress(MPa)</span>
678
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">AD</span>
679
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">P value</span>
680
|-
681
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span id='_Hlk106895544'>1030</span>
682
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.264</span>
683
|  style="border-top: 1pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.371</span>
684
|-
685
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">970</span>
686
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.210</span>
687
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.544</span>
688
|-
689
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">909</span>
690
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.279</span>
691
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.484</span>
692
|-
693
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">849</span>
694
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.562</span>
695
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.091</span>
696
|-
697
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">788</span>
698
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.603</span>
699
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.062</span>
700
|-
701
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">758</span>
702
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.735</span>
703
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.037</span>
704
|-
705
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">728</span>
706
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.707</span>
707
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.042</span>
708
|-
709
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">697</span>
710
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;"|<span style="text-align: center; font-size: 75%;">0.250</span>
711
|  style="border-top: 2pt solid black;border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.277</span>
712
|}
713
714
715
The test shows that the p-value is greater than the significant level of 0.05, which is in line with the hypothesis test, and therefore the bearing steel material life can be considered to be in line with the log-normal distribution.
716
717
The S-N curve of the bearing is estimated using Basquin's equation combined with the method of great likelihood, Basquin's equation is
718
719
{| class="formulaSCP" style="width: 100%; text-align: center;" 
720
|-
721
| 
722
{| style="text-align: center; margin:auto;" 
723
|-
724
| <math display="inline">S^mN=C</math>
725
|}
726
| style="width: 5px;text-align: right;white-space: nowrap;" | (7)
727
|}
728
729
730
Where ''m'' is the exponent, ''N'' is the fatigue life and ''C'' is a constant.
731
732
Taking the logarithm of both sides gives:
733
734
{| class="formulaSCP" style="width: 100%; text-align: center;" 
735
|-
736
| 
737
{| style="text-align: center; margin:auto;" 
738
|-
739
| <math display="inline">lgN_p=lgC_p-mlgS</math>
740
|}
741
| style="width: 5px;text-align: right;white-space: nowrap;" | (8)
742
|}
743
744
745
Where the subscript ''p'' indicates the survival rate.
746
747
When the survival rate is 50%, the median log life is equal to the mean of the log life:
748
749
{| class="formulaSCP" style="width: 100%; text-align: center;" 
750
|-
751
| 
752
{| style="text-align: center; margin:auto;" 
753
|-
754
| <math display="inline">\mu (S)=lgN_{50}=lgC-m_{50}lgS</math>
755
|}
756
| style="width: 5px;text-align: right;white-space: nowrap;" | (9)
757
|}
758
759
760
Where <math display="inline">\mu (S)</math> is the mean log life of the sample at different stress.
761
762
When S=788MPa, the mean value of the log-life sample with stress of 788MPa is substituted as the parent mean value in Eq. (8) and (9) to obtain.
763
764
{| class="formulaSCP" style="width: 100%; text-align: center;" 
765
|-
766
| 
767
{| style="text-align: center; margin:auto;" 
768
|-
769
| <math display="inline">lgC=2.90m_{50}+5.4061</math>
770
|}
771
| style="width: 5px;text-align: right;white-space: nowrap;" | (10)
772
|}
773
774
{| class="formulaSCP" style="width: 100%; text-align: center;" 
775
|-
776
| 
777
{| style="text-align: center; margin:auto;" 
778
|-
779
| <math display="inline">\mu (S)=5.4060\mbox{-}m_{50}lgS\mbox{+}2.90m_{50}</math>
780
|}
781
| style="width: 5px;text-align: right;white-space: nowrap;" | (11)
782
|}
783
784
785
The log life of bearing steel follows a normal distribution and the mean log life is related to the standard deviation of the log life at different survival rates as follows:
786
787
{| class="formulaSCP" style="width: 100%; text-align: center;" 
788
|-
789
| 
790
{| style="text-align: center; margin:auto;" 
791
|-
792
| <math display="inline">\mu (S)-lgN_p={\upsilon }_p\sigma (S)</math>
793
|}
794
| style="width: 5px;text-align: right;white-space: nowrap;" | (12)
795
|}
796
797
798
Where <math display="inline">\sigma (S)</math> is the standard deviation of the log life of the member at different stress. <math display="inline">{\upsilon }_{P}</math> is the standard normal deviation corresponding to the probability of damage.
799
800
Substituting Eq, (8) and (9) into Eq. 12 yields:
801
802
{| class="formulaSCP" style="width: 100%; text-align: center;" 
803
|-
804
| 
805
{| style="text-align: center; margin:auto;" 
806
|-
807
| <math display="inline">\sigma (S)={\left(lgC-lgC_p\right)}^{\frac{1}{{\upsilon }_p}}-</math><math>\frac{m_{50}}{{\upsilon }_p}lgS+\frac{m_p}{{\upsilon }_p}lgS</math>
808
|}
809
| style="width: 5px;text-align: right;white-space: nowrap;" | (13)
810
|}
811
812
813
The sample log life standard deviation at a stress of 788Mpa is substituted into Eq. (13) as the parent log life standard deviation:
814
815
{| class="formulaSCP" style="width: 100%; text-align: center;" 
816
|-
817
| 
818
{| style="text-align: center; margin:auto;" 
819
|-
820
| <span style="text-align: center; font-size: 75%;"> <math display="inline">\sigma (788)=\sqrt{\frac{1}{\mbox{n-1}}\sum_{i=1}^n{\left(lgN_i-\mu (788)\right)}^2}\mbox{=}{\left(lgC-lgC_p\right)}^{\frac{1}{{\upsilon }_p}}-</math><math>\frac{m_{50}}{{\upsilon }_p}lg788+\frac{m_p}{{\upsilon }_p}lg788</math> </span>
821
|}
822
| style="width: 5px;text-align: right;white-space: nowrap;" | (14)
823
|}
824
825
826
Also known as equation:
827
828
{| class="formulaSCP" style="width: 100%; text-align: center;" 
829
|-
830
| 
831
{| style="text-align: center; margin:auto;" 
832
|-
833
| <span style="text-align: center; font-size: 75%;"> <math display="inline">\sigma (S)=\sigma (788)\mbox{+}\frac{m_{50}}{{\upsilon }_p}lg788\mbox{-}\frac{m_p}{{\upsilon }_p}lg788-</math><math>\frac{m_{50}}{{\upsilon }_p}lgS+\frac{m_p}{{\upsilon }_p}lgS</math> </span>
834
|}
835
| style="width: 5px;text-align: right;white-space: nowrap;" | (15)
836
|}
837
838
839
When the survival rate P = 84.1%,υ_p=1, which is collated to give:
840
841
{| class="formulaSCP" style="width: 100%; text-align: center;" 
842
|-
843
| 
844
{| style="text-align: center; margin:auto;" 
845
|-
846
| <math display="inline">\sigma (S)=1.1344+m_{50}lgS+m_{84.1}lgS</math>
847
|}
848
| style="width: 5px;text-align: right;white-space: nowrap;" | (16)
849
|}
850
851
852
When N follows a log-normal distribution, its probability density function is
853
854
{| class="formulaSCP" style="width: 100%; text-align: center;" 
855
|-
856
| 
857
{| style="text-align: center; margin:auto;" 
858
|-
859
| <math display="inline">f(N)=\frac{1}{\sigma (S)\sqrt{2\pi }N}exp\left\{-\frac{{\left(lnN-\mu (S)\right)}^2}{2{\sigma }^2(S)}\right\}</math>
860
|}
861
| style="width: 5px;text-align: right;white-space: nowrap;" | (17)
862
|}
863
864
865
Substituting the different stresses and corresponding lifetimes into Eq. (17) and multiplying them together, the likelihood function is obtained as equation
866
867
{| class="formulaSCP" style="width: 100%; text-align: center;" 
868
|-
869
| 
870
{| style="text-align: center; margin:auto;" 
871
|-
872
| <span style="text-align: center; font-size: 75%;"> <math display="inline">L=\prod_{n=1}^n\frac{1}{\sigma (S_i)\sqrt{2\pi }}exp\left\{-\right. </math><math>\left. \frac{{\left(lnN{-}_{}^i\mu (S_i)\right)}^2}{2{\sigma }^2(S_i)}\right\}</math> </span>
873
|}
874
| style="width: 5px;text-align: right;white-space: nowrap;" | (18)
875
|}
876
877
878
Take the logarithm on both sides of the above equation, and get it
879
880
{| class="formulaSCP" style="width: 100%; text-align: center;" 
881
|-
882
| 
883
{| style="text-align: center; margin:auto;" 
884
|-
885
| <span style="text-align: center; font-size: 75%;"> <math display="inline">lnL=-\sum_{i=1}^n\left\{In\sqrt{2\pi }+ln\sigma (S_i)+\right. </math><math>\left. \frac{{\left(lnN_i-\mu (S_i)\right)}^2}{2{\sigma }^2(S_i)}\right\}</math> </span>
886
|}
887
| style="width: 5px;text-align: right;white-space: nowrap;" | (19)
888
|}
889
890
891
Convert Eq. (18) into Eq. (19) and find the minimum value of equation (19) to find the maximum value of the likelihood function equation
892
893
{| class="formulaSCP" style="width: 100%; text-align: center;" 
894
|-
895
| 
896
{| style="text-align: center; margin:auto;" 
897
|-
898
| <span style="text-align: center; font-size: 75%;"> <math display="inline">F(m_{50},m_{84.1})=\sum_{i=1}^n\left\{In\sqrt{2\pi }+\right. </math><math>\left. ln\sigma (S_i)+\frac{{\left(lnN_i-\mu (S_i)\right)}^2}{2{\sigma }^2(S_i)}\right\}</math> </span>
899
|}
900
| style="width: 5px;text-align: right;white-space: nowrap;" | (20)
901
|}
902
903
904
Find the minimum value of F to obtain the maximum likelihood estimates of the parameters m50,m84.1, m50=15.04,   m84.1=5.32.
905
906
The values of parameters m50 and m84.1 are substituted into Eq. (11) and Eq. (16) respectively to obtain the life distribution parameters under any stress, and the log life mean and standard deviation parameter equations are as follows:
907
908
{| class="formulaSCP" style="width: 100%; text-align: center;" 
909
|-
910
| 
911
{| style="text-align: center; margin:auto;" 
912
|-
913
| <span style="text-align: center; font-size: 75%;"> <math display="inline">\sigma (S)=29.29-9.72lgS</math> </span>
914
|}
915
| style="width: 5px;text-align: right;white-space: nowrap;" | (21)
916
|}
917
918
{| class="formulaSCP" style="width: 100%; text-align: center;" 
919
|-
920
| 
921
{| style="text-align: center; margin:auto;" 
922
|-
923
| <span style="text-align: center; font-size: 75%;"> <math display="inline">\mu (S)=48.97-15.04lgS</math> </span>
924
|}
925
| style="width: 5px;text-align: right;white-space: nowrap;" | (22)
926
|}
927
928
929
The parameter-stress equation derived by the great likelihood method was plotted against the sample log-life mean and standard deviation in the same graph for comparison. The sample log-life mean and log-life standard deviation are shown in Table 11 and the results of the comparison are shown in Fig. 22 - Fig. 23. Within a certain interval, the fit is good, which proves that the parameter equations derived using the great likelihood method are more reasonable.
930
931
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
932
<span style="text-align: center; font-size: 75%;">Table 11 The mean and standard deviation of the log-life of the sample</span></div>
933
934
{| style="width: 47%;margin: 1em auto 0.1em auto;border-collapse: collapse;" 
935
|-
936
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">Stress (MPa )</span>
937
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">Mean log -life</span>
938
|  style="border-top: 2pt solid black;border-bottom: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">Standard deviation of log-life</span>
939
|-
940
|  style="border-top: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">1030</span>
941
|  style="border-top: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">3.7572</span>
942
|  style="border-top: 1pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.0951</span>
943
|-
944
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">970</span>
945
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">3.8692</span>
946
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.1433</span>
947
|-
948
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">990</span>
949
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">4.766</span>
950
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.7922</span>
951
|-
952
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">849</span>
953
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">5.1407</span>
954
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">1.0905</span>
955
|-
956
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">788</span>
957
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">5.406</span>
958
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">1.1344</span>
959
|-
960
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">758</span>
961
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">6.0801</span>
962
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">1.1266</span>
963
|-
964
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">728</span>
965
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">6.7981</span>
966
|  style="text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">1.0539</span>
967
|-
968
|  style="border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">697</span>
969
|  style="border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">7.5227</span>
970
|  style="border-bottom: 2pt solid black;text-align: center;vertical-align: top;"|<span style="text-align: center; font-size: 75%;">0.2203</span>
971
|}
972
973
974
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
975
 [[Image:Draft_Ling_717181314-image44.png]] </div>
976
977
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
978
<span style="text-align: center; font-size: 75%;">Fig. 22  Median S-N curve versus sample Log-life mean</span></div>
979
980
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
981
<span style="text-align: center; font-size: 75%;"> [[Image:Draft_Ling_717181314-image45.png]] </span></div>
982
983
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
984
<span style="text-align: center; font-size: 75%;">Fig. 23 Standard Deviation Curve Versus Sample Log-Life Standard Deviation</span></div>
985
986
=5. Bearing fatigue reliability=
987
988
According to the heterogeneous interference model, for different reliability of fatigue life is calculated, the model as shown in the equation, the above calculated stress distribution Eq. (4) and life distribution Eq. (17) substituted into the following Eq. (23), the fatigue life curve of the bearing can be obtained as shown in Fig. 24.
989
990
{| class="formulaSCP" style="width: 100%; text-align: center;" 
991
|-
992
| 
993
{| style="text-align: center; margin:auto;" 
994
|-
995
| <math display="inline">R=\int_0^{+\infty }h(S)\int_N^{+\infty }f(n,S)dndS</math>
996
|}
997
| style="width: 5px;text-align: right;white-space: nowrap;" | (23)
998
|}
999
1000
1001
Where R is the fatigue reliability, h(S) is the probability density function of the equivalent load distribution and f(n,S) is the probability density function of the component life under different stresses.
1002
1003
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1004
<big> [[Image:Draft_Ling_717181314-image47.png]] </big></div>
1005
1006
<div class="center" style="width: auto; margin-left: auto; margin-right: auto;">
1007
<span style="text-align: center; font-size: 75%;">Fig. 24 Bearing fatigue reliability variation curve</span></div>
1008
1009
As can be seen from Figure 24, the fatigue reliability decreases continuously with increasing fatigue life, which is consistent with reality. The downward trend is faster until the number of cycles is 10<sup>5</sup>, after which it slows down. When the number of fatigue cycles reaches 10<sup>5</sup>, the fatigue reliability of the bearing drops to below 0.82.
1010
1011
=6. Conclusions=
1012
1013
In this paper, the reliability of the bearings of a giant tyre unloader is predicted using a multi-body dynamics, finite and other numerical simulation platform, combined with a heterogeneous interference model, for a huge alternating load during operation:
1014
1015
(1) The joint simulation between ADAMS and ANSYS shows that the combined force on the bearing at the lowermost rocker arm of the tyre unloader can reach a maximum of 150kN, with a maximum of 108kN and 78kN in the transverse and longitudinal directions. The bearing of the lowermost rocker arm of the tyre unloader is subjected to the greatest inertial force when bearing the weight of the whole tyre and flipping, and the maximum stress value is 1316.2Mpa.
1016
1017
(2) By probability fitting and testing, the equivalent stress distribution in the fatigue part of the bearing conforms to the Weibull distribution, with the Weibull distribution scale parameter and shape parameter being 61.93 and 0.72 respectively at a confidence interval of 95%.
1018
1019
(3) Under the modified S-N curve, the life of the bearing steel material under any stress conforms to a log-
1020
1021
normal distribution, and the Basquin equation exponents are 15.04 and 5.32 for a survival rate of 50% and 84.1% respectively, as obtained by the maximum likelihood method.
1022
1023
(4) Based on the heterogeneous interference fatigue reliability model and according to the reliability calculation results, the reliability of the bearing shows a decreasing trend with the increase of the number of times it is used, which is in line with the actual situation, when the service life of the bearing reaches 10<sup>5</sup> times, the reliability drops to below 0.82.
1024
1025
The kinetic simulation, finite element analysis, fatigue reliability prediction and other technical methods used in this paper are feasible for the fatigue reliability prediction of tyre unloading machine bearings.
1026
1027
'''Funding '''This work was supported by the Natural Science Foundation of Fujian Province (Grant No.2020J01871), China Postdoctoral Science Foundation (Granted No.2020M671956).
1028
1029
==Declarations==
1030
1031
'''Conflict of Interests''' Conflict of interests The authors declare that they have no known competing financial interests or personal relationships.
1032
1033
==References==
1034
1035
[1] Levratti A , Riggio G , Fantuzzi C, et al. tyreBOT: A collaborative robot for the tyre workshop [J]. Robotics and Computer-Integrated Manufacturing,2019, 57(JUN.) :129-137. [https://doi.org/10.1016/j.rcim.2018.11.001 https://doi.org/10.1016/j.rcim.2018.11.001]
1036
1037
[2] Vavro J , Vavro J , Kováiková, Petra, et al. Kinematic and Dynamic Analysis of the Manipulator for Removal of Rough Tyres [J]. Procedia Engineering, 2016, 136 :120-124. [https://doi.org/10.1016/j.proeng.2016.01.184 https://doi.org/10.1016/j.proeng.2016.01.184]
1038
1039
[3] Xie L.Y, Wang Z. Heterogeneous interference model for fatigue reliability with random constant amplitude cyclic loading [J]. Journal of Mechanical Engineering,2008(01) :1-6(in Chinese).
1040
1041
[4] Liu K, Wang M, Yang G L. Fatigue reliability analysis of underground pipeline structures under multiple operating conditions [J]. Safety and Environmental Engineering,2019,26(04) :106-110+126(in Chinese). [https://doi.org/10.13578/j.cnki.issn.1671-1556.2019.04.017 https://doi.org/10.13578/j.cnki.issn.1671-1556.2019.04.017]
1042
1043
[5] Li R X, Zhou J Y, Sun K Z,et al. Fatigue reliability analysis of rolling bearing systems under random loading [J]. Hydromechatronics Engineering,2012,40(01) :157-160(in Chinese).
1044
1045
[6] Jin Y, Liu S J, Zhang J G. Fatigue reliability of high-speed rolling bearings under artificial neural network based on ge-etic algorithm optimization [J]. Journal of Aerospace Power,2018,(11) :2748-2755(in Chinese). [https://doi.org/10.13224/j.cnki.jasp.2018.11.021 https://doi.org/10.13224/j.cnki.jasp.2018.11.021]
1046
1047
[7] Qi M Y, Liao A H. Design of a MATLAB App Designer based rolling bearing reliability assessment system for traction motors [J]. Electronics Technology,2022,(03) :79-86(in Chinese). [https://doi.org/10.16180/j.cnki.issn1007-7820.2022.03.012 https://doi.org/10.16180/j.cnki.issn1007-7820.2022.03.012]
1048
1049
[8] Lim Chi Keong Reuben, Michael Corsar, David Mba. Bearing replacement extension without failure data[J]. Int. J. of Risk Assessment and Management, 2015, 18(1) : 38-51.  [https://doi.org/ https://doi.org/]10.1504/IJRAM.2015.068139
1050
1051
[9] F. Pape, O. Maiss, B. Denkena, G. Poll, et al. Computational approach to improve bearings by residual stresses based on their required bearing fatigue life[J]. International Journal of Computational Methods and Experimental Measurements, 2017(4), : 656-666. [https://doi.org/10.2495/CMEM-V6-N4-656-666 https://doi.org/10.2495/CMEM-V6-N4-656-666]
1052
1053
[10] Wang H B. Ship unloader rolling bearing life prediction and maintenance management platform development [D]. Jilin University,2011(in Chinese).
1054
1055
[11] Xia X T, Ye L et al. Prediction of rolling bearing performance retention reliability [J]. Bearings,2016,(06) :28-34(in Chinese). [https://doi.org/10.19533/j.issn1000-3762.2016.06.008 https://doi.org/10.19533/j.issn1000-3762.2016.06.008]
1056
1057
[12] Zhang X, Zhang Y Q, Sun Y, et al. Reliability Model of TBM Main Bearing based on Nonlinear Strength Degradation Theory[J]. International Journal of Performability Engineering,2018,14(12) :3054-3065. [https://doi.org/10.23940/ijpe.18.12.p15.30543065 https://doi.org/10.23940/ijpe.18.12.p15.30543065]
1058
1059
[13] Cheng W W, Xiang M H, Lyu Bugao et al. Influence of angular misalignment on the tribological performance of high-speed micro ball bearings considering full multibody interactions[J]. Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology,2021,235(6) :1168-1189. [https://doi.org/10.1177/1350650120948292 https://doi.org/10.1177/1350650120948292]
1060
1061
[14] Herp J, Ramezani M H , Bach-Andersen M, et al. Bayesian state prediction of wind turbine bearing failure[J]. Renewable Energy,2018,116 :164-172. [https://doi.org/10.1016/j.renene.2017.02.069 https://doi.org/10.1016/j.renene.2017.02.069]
1062
1063
[15] Tong V C,Hong S W. Improved formulation for running torque in angular contact ball bearings[J]. International Journal of Precision Engineering and Manufacturing,2018,19(1) :47-56. [https://doi.org/10.1007/s12541-018-0006-2 https://doi.org/10.1007/s12541-018-0006-2]
1064
1065
[16] König F, Ouald C A, Jacobs G, et al. A multiscale-approach for wear prediction in journal bearing systems – from wearing-in towards steady-state wear[J]. Wear,2019,426-427. [https://doi.org/10.1016/j.wear.2019.01.036 https://doi.org/10.1016/j.wear.2019.01.036]
1066
1067
[17] Guillermo E, Gabelli A. A model for rolling bearing life with surface and subsurface survival: Surface thermal effects[J]. Wear,2020 : 203446. [https://doi.org/10.1016/j.wear.2020.203446 https://doi.org/10.1016/j.wear.2020.203446]
1068
1069
[18] Zhang Y C, Shu T, Ding R X. Bearing Life Prediction Based on SPSS and Grey Prediction Model[J]. IOP Conference Series: Earth and Environmental Science,2021,634(1) :012051. [https://doi.org/10.1088/1755-1315/634/1/012051 https://doi.org/10.1088/1755-1315/634/1/012051]
1070
1071
[19] Lorenz S J, Sadeghi F, Trivedi H K, et al. A Continuum Damage Mechanics Finite Element Model for Investigating Effects of Surface Roughness on Rolling Contact Fatigue[J]. International Journal of Fatigue,2020,143(3) :105986. [https://doi.org/ https://doi.org/]10.1016/j.ijfatigue.2020.105986
1072
1073
[20] Pandey V B , Singh I V , Mishra B K . A Strain-based Continuum Damage Model for Low Cycle Fatigue under Different Strain Ratios[J]. Engineering Fracture Mechanics, 2020, 242(4):107479. [https://doi.org/10.1016/j.engfracmech.2020.107479 https://doi.org/10.1016/j.engfracmech.2020.107479]
1074
1075
[21] Cano J A ,  Stewart C M . A continuum damage mechanics (CDM) based Wilshire model for creep deformation, damage, and rupture prediction[J]. Materials Science and Engineering A, 2021, 799(4):140231. [https://doi.org/10.1016/j.msea.2020.140231 https://doi.org/10.1016/j.msea.2020.140231]
1076
1077
[22] Li W, Lu L T. Probabilistic properties of the ultra-long life S-N relationship for high-carbon chromium bearing steels [J]. Journal of Traffic and Transportation Engineering,2006(02):17-21(in Chinese).
1078

Return to Ling et al 2023a.

Back to Top