The International Journal of Advanced Manufacturing Technology

The investigation of fracture characteristics and process optimization of low-cycle

fatigue cropping by using an AET-based multi-sensor system
--Manuscript Draft--

Manuscript Number: JAMT-D-22-02814

Full Title: The investigation of fracture characteristics and process optimization of low-cycle
fatigue cropping by using an AET-based multi-sensor system

Article Type: Original Research

Keywords: multi-sensor LCFC system; fracture characteristics of 16Mn bar; loading frequency; a
RA-based optimal control scheme

Abstract: Low-cycle fatigue cropping (LCFC) is a new method for metal bar separation, which
solves the problems such as high active load and energy waste in traditional separate
method. A key issue for LCFC is to achieve high cropping efficiency and good cross-
section quality simultaneously. To solve this problem, a novel multi-sensor LCFC
system is established to investigate the fracture characteristics of the 16Mn notched
metal bar. An acoustic emission technique (AET) parameter: the ratio of rise time to
amplitude (RA) is used as a variable for optimize control scheme. 3D microscopy is
used to evaluate cross-section quality. Results showed loading frequency has a big
influence on the cross-section quality, and the optimal loading frequency is determined.
Shear failure model and tensile failure model dominate the crack propagation stage,
and tensile failure model cause poor cross-section quality. RA reflects the failure model
change well. A RA-based control scheme is proved to improve the cross-section
quality effectively. Combine with cropping time, suitable RA value is determined.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Manuscript Click here to access/download;Manuscript;manuscript.docx

Click here to view linked References

1 1 The investigation of fracture characteristics and process optimization

2
3
4 2 of low-cycle fatigue cropping by using an AET-based multi-sensor
5
6 3 system
7
8
9 4
10
11
12 5 Yujian Ren1, 2, Boyang Liu1, Yi Zhang1, Yuanzhe Dong3, Dong Jin1,
13
14
15 6 Shengdun Zhao1，Jingzhou Gao1
16
17 7
18
19
20 8 1. School of Mechanical Engineering, Xi’an Jiaotong University,
21
22
23 9 No.28, Xianning West Road, Xi’an, Shaanxi 710049, China
24
25
26 10 2. School of Mechanical and Aerospace Engineering, Nanyang
27
28 11 Technological University, Singapore
29
30
31 12 3. School of Construction Machinery, Chang’an University, China
32
33
34 13
35
36
37 14 857280283@qq.com，liuboyoung@163.com, 986593153@qq.com
38
39 15 dongyuanzhe1989@126.com, 525053923@qq.com,
40
41
42 16 sdzhao@mail.xjtu.edu.cn,
43
44
45 17 904416827@qq.com
46
47
48 18
49
50 19 Corresponding Author: Mr. Jingzhou Gao
51
52
53 20 Corresponding Author’s mailbox: 904416827@qq.com
54
55
56 21
57
58
59
60
61
62
63
64
65
 1 22 Abstract
2
3
4 23 Low-cycle fatigue cropping (LCFC) is a new method for metal bar separation, which solves the
5
6 24 problems such as high active load and energy waste in traditional separate method. A key issue for
7
8 25 LCFC is to achieve high cropping efficiency and good cross-section quality simultaneously. To
9
10
26 solve this problem, a novel multi-sensor LCFC system is established to investigate the fracture
11
12
13
27 characteristics of the 16Mn notched metal bar. An acoustic emission technique (AET) parameter:
14
15 28 the ratio of rise time to amplitude (RA) is used as a variable for optimize control scheme. 3D
16
17 29 microscopy is used to evaluate cross-section quality. Results showed loading frequency has a big
18
19 30 influence on the cross-section quality, and the optimal loading frequency is determined. Shear
20
21 31 failure model and tensile failure model dominate the crack propagation stage, and tensile failure
22
23 32 model cause poor cross-section quality. RA reflects the failure model change well. A RA-based
24
25 33 control scheme is proved to improve the cross-section quality effectively. Combine with cropping
26
27 34 time, suitable RA value is determined.
28
29 35 Keywords: multi-sensor LCFC system, fracture characteristics of 16Mn bar, loading frequency, a
30
31 36 RA-based optimal control scheme
32
33 37
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 38 Highlights
2
3
4 39 1. A novel AET-based multi-sensor LCFC system is established.
5
6 40 2. The fracture characteristics of 16Mn notched metal bar in LCFC process are investigated.
7
8 41 3. Effects of loading frequency on cross-section are investigated.
9
10
42 4. A RA-based real-time control scheme is proposed.
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 43 1. Introduction
2
3
4 44 The separation of the metal bar is the first step in many manufacturing fields. Traditional separation
5
6 45 techniques include sawing [1], plasma arc cutting [2] , laser cutting [3], fineblanking [4], and high-
7
8 46 speed cropping [5]. There are many shortcomings such as high active load, significant energy waste,
9
10
47 and material loss in these techniques. To alleviate the aforementioned issues, the low-cycle fatigue
11
12
13
48 cropping (LCFC) approach is proposed [6, 7].
14
15 49
16
17 50 A key issue for LCFC is to achieve high cropping efficiency and good cross-section quality
18
19 51 simultaneously. Many researchers devote their energy to find optimal metal bar parameters. By
20
21 52 using Acoustic emission technique (AET) and optical microscopy (OM), Ren [8] found notch
22
23 53 eccentric ratio influenced the crack propagation time and cross-section quality. The idea notch
24
25 54 eccentric ration for 16Mn bar is 0.06. Zhang [9] found bar clamping position influenced the ratio of
26
27 55 the maximum tensile stress to the maximum shear stress which near the notch bottom. For the metal
28
29 56 bar with diameter ratio L/D=8/3, the proportion distance between notch and shearing die L1 to D is
30
31 57 0.3, and the proportion distance between notch and clamping position L2 to D is 0.25. Zhong [10]
32
33 58 investigated the relationship between clearance and the ductile damage initiation criterion, and
34
35 59 found the suitable value of clearance is 0.1 mm.
36
37 60
38
39 61 Some scholars focus on the loading state optimization. Zhao [11] found that increasing displacement
40
41
62 load while decreasing loading frequency improves cross-section quality. Zhang [12] studied the
42
43
63 influence of initial load on the cross-section quality and proposed an ultimate tensile stress (UTS)
44
45
46 64 based crack initiation model. Hua [6] designed 5 control curves of vibration frequency, and found
47
48 65 linear decrement control curve is better for producing good cross-section quality.
49
50 66
51
52 67 Previous studies improve the cropping efficiency and cross-section quality. However, the fracture
53
54 68 characteristics during LCFC process are neglected. AET is a non-destructive technique that has been
55
56 69 used extensively in the field of concrete engineering [13].and fatigue damage of metals [14, 15].
57
58 70 The released energy of plastic deformation and crack growth form AE signals [16-18]. Studies based
59
60 71 on AET usually use signal features such as counts [19], amplitude [20], ring-down counts [21-23],
61
62
63
64
65
 72 kurtosis [24], energy [25-27], AE entropy [28, 29], RA [30], AF (average frequency) [31], and
1
2 73 power spectral density [32]. Based on AET, Han [33, 34] and Li [35] divided fatigue process into
3
4
74 three stages: crack initiation stage, crack propagation stage, and fracture stage. Li [36] compared
5
6
7 75 differences between high loading frequency and low loading frequency in LCFC. Ren proposed two
8
9 76 methods to balance cropping efficiency and cross-section quality: one first is based on kurtosis [37],
10
11 77 and the other is based on AE hit energy angle [38].
12
13 78
14
15 79 Above studies prove AET is benefit to advance LCFC research. In this study, an AET-based multi-
16
17 80 sensor LCFC system is established to investigate the fracture characteristics of the 16Mn notched
18
19 81 metal bar. An optimal control scheme based AET parameter RA is proposed and evaluated.
20
21
82
22
23
24
25
26
83 2. AET-based multi-sensor LCFC system
27
28
29 84 2.1 LCFC machine
30
31
32
33 85 The layout of LCFC machine is illustrated in Figure 1. Two parts in this machine: one is eccentric
34
35 86 load part which includes servo motor 1, ballscrew, bearing pedestal 1, and three-jaw chuck 1, the
36
37 87 other is cyclic load part which includes servo motor 2, bearing pedestal 2, and three-jaw chuck 2.
38
39 88 The working principle of LCFC machine is described as follows: for eccentric load part, servo motor
40
41 89 1 rotates the ballscrew, forcing bearing pedestal 1 to move along the Y axis. Three-jaw chuck 1
42
43 90 move along Y axis simultaneously. One side of the metal bar contacts with three-jaw chuck 1
44
45 91 through a sleeve which is made of ZCuZn25Al6FeMn3. The other side of the metal bar is fixed with
46
47 92 three-jaw chuck 2. The metal bar is subjected an eccentric displacement load when servo motor 1
48
49 93 work. The eccentric displacement load depends on the ballscrew rotate numbers and ballscrew’ lead.
50
51 94 Ballscrew’s lead is 4 mm in this study. For cyclic load part, servo motor 2 drives the three-jaw chuck
52
53 95 2 rotate around the X axis, putting the metal bar under cyclic loading.
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 96
17
18 97 Fig. 1 (a) The compositions of the LCFC machine (b) The photo of LCFC machine
19
20 98 The torque of servo motor is a critical value to select servo motor model. It can be calculated as
21
22 99 follows.
23
24 100 Critical crack initiation load (CCIL) is dependent on the yield strength which can be written as
25
26 101 follows [12, 39]:
27
28   k 
*
102 (1)
29
30
31  h  D  2h    h / r  1 2.4 b h

32 103 k  1   2 f   1 1     (2)
 r  D     180  
33
34
35 104 f (x) 
3 1  1 x  3 x 2

5
x 
3
35
x  0.537 x
4 5  (3)
36 8x
2.5  2 8 16 128 
37
38 4Fci L
105   (4)

 
39 D  2h
3

40 
41 2
42
43 106 here, k is the stress concentration factor, D is the diameter of the metal bar, h is the depth of
44
45
46 107 the prefabricated notch, r is the radius of the prefabricated notch, L is the distance between
47
48 108 loading position and notch certain line and Fci is the critical crack initial load.
49
50
51 109
52
53 110 The notch geometry of the metal bar is shown in Fig.2.Critical crack propagation load (CCPL) can
54
55 111 be calculated by Eq. (5) [40].
56
112
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11 113
12 114 Fig.2 The notch geometry of the metal bar
13
14
 ( D / 2  h  a ) K IC
3

15 115 Fcp  (5)

16 4L hf (( D  2h ) / D )
17
18 116 here, a is the depth of the ring crack and always simply set as 0.2 mm [12]. The notch geometry
19
20
21 117 of the metal bar is given as follows: D is 12 mm, L is 76 mm, L1 is 26 mm,  is 90 , h is
22
23 118 1 mm, and r is 0.2 mm. The mechanical properties of AISI 304, Al6061, 1045, 16Mn, and 316L
24
25 119 bars are shown in Table. The results of CCIL Fci and CCPL Fcp are shown in Table 2.
26
27
28 120 Table 1 Properties of AISI304, Al6061, 1045, 16Mn, and 316L [7, 8, 41-45]
29
30 Materials AISI 304 Al 6061 1045 16Mn 316L
31
32 Young’s Modulus (MPa), E 1950000 710000 2100000 2060000 1900000
33 Poisson’s ratio  0.30 0.33 0.30 0.30 0.3
34
35 Yield strength (MPa) ,  y 230 55.2 355 345 273
36 UTS (MPa), u 602 124 600 620 622
37
Fracture toughness ( MPa m ), K IC 228 14.68 57.98 14.31 53.34
38
39
40 121 Table 2 Critical crack initial load and critical propagation load of AISI304, Al6061, 1045, 16Mn and 316L
41
42 Materials AISI Al 6061 1045 16Mn 316L
43 304
44
45 227.84 54.68 351.63 341.76 270.43
Critical crack initial load (N), Fci
46
47
48 6318.63 406.83 1606.81 396.58 1478.22
Critical crack propagation load (N), Fcp
49
50
51 122
52
53 123 The maximum torque of servo motor 1 can be given by follow equation:
54
55 '
LP
56 124 Tr  (6)
57 2
58
59 '
60 125 p is the applied force and the maximum value is 6318.63N , L is the lead of the ballscrew, Tr is the
61
62
63
64
65
 126 required torque, and  is the efficiency of the screw which is set as 0.9 commonly. The maximum
1
2 127 torque is 4.47 N  m .
3
4
128 The servo motor 2 is subjected to a frictional force, and the friction force can be calculated as follows:
5
6 P 
Ff 129 (7)
7
8
9
130  is the friction coefficient which is set as 0.25 commonly. The maximum frictional force is
10
11 131 1579.66 N. The maximum torque of servo motor 2 can be calculated by the following equation:
12
13 132 Tr  Ff X (8)
14
15
133 X is the eccentric displacement load which is set as 4 mm in this study. The maximum required
16
17
18 134 torque of servo motor 2 is 6.32 N  m . The model of the servo motor is 130ST-M10025LFB which
19
20 135 maximum torque is 10 N  m .
21
22
23 136 2.2 Monitoring system
24
25
26
27 137 The monitoring system include four types of sensor such as AE sensor, laser-displacement sensor,
28
29 138 vibration sensor and gritting scale. Gritting scale is used to monitor the eccentric load value.
30
31 139
32
33 140 The model of AE sensor is Nano30 (Physical Acoustic Corporation, USA) which with a peak
34
35 141 frequency of 293 kHz, a sensitivity of 62dB ref. 1V/ (m/s) and the bandwidth is 125-750 kHz. The
36
37 142 operational amplifier's voltage gain is tuned to 20 dB to magnify the AE signals while filtering out
38
39 143 the background noise [36]. Because the AE signal's frequency is less than 500 kHz, a sampling rate
40
41 144 of 1 MHz is appropriate. The maximum sampling rate of PCI-1714 is 30 MS/s, which ensuring
42
43 145 adequate acquisition speed. According to a prior study [37], the AE signal's amplitude threshold
44
45 146 value is set as 0.2 V, which is twice the operating background noise.
46
47 147
48
49
148 The model of the vibration sensor is CT1010L, which frequency range is from 0.5 kHz to 5 kHz,
50
51
52 149 voltage sensitivity is 100 mv/g, and the maximum range is 50 g. The model of laser-displacement
53
54 150 sensor is HL-G108-A-C5, the distance of the testing center is 85 mm and the detection range is
55
56 151 85  20mm . The resolution ratio of the laser-displacement sensor is 2.5μm .
57
58
59
60
61
62
63
64
65
 1 152 2.3 Control system
2
3
4 153 As shown in Fig. 3 (a), the control system include a host computer, a PLC, and two servo motors.
5
6 154 The host computer sends control commands to the PLC, and the PLC subsequently controls the
7
8 155 servo motors' movement. The bending of the metal bar is controlled by servo motor 1, while the
9
10
156 rotation of the metal bar is controlled by servo motor 2. Servo motor 1 use position-control, while
11
12
13
157 servo motor 2 use speed-control. As shown in Fig. 3(b), eccentric loading system and rotate loading
14
15 158 system operating independently.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31 159
32
33 160 Fig.3 Control system (a) control system logic diagram; (b) the electrical schematic diagram of the control system.
34
35 161
36
37
38
39 162 3. The investigation of 16Mn metal bar fracture
40
41
42 163 characteristics
43
44
45
46 164 3.1 Fracture characteristics of 16Mn metal bar
47
48
49 165 AE signal features, vibration features, and laser-displacement feature are shown in Fig.4. According
50
51 166 to the fatigue mechanism, there are three stages: crack initiation stage, crack propagation stage, and
52
53 167 fracture stage. High AE kurtosis value is the transmit point among different stages. The specific
54
55 168 fracture characteristics of different stages are described as follows.
56
57 169
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18 170
19
20 171 Fig.4 Features of the AE signal, the vibration signal, and the laser displacement signal vs. time (a) AE ring-down
21
22 172 counts vs. time; (b) AE kurtosis vs. time; (c) AE energy vs. time; (d) vibration signal amplitude vs. time; (e)
23
24 173 vibration signal kurtosis vs. time; (f) smallest displacement in each loading cycle vs. time.
25
26 174
27
28 175 The crack initiation stage lasts from 0 s to 3.8 s. As demonstrated in Fig.4 (a), no ring-down counts
29
30 176 occurred during the crack initiation stage, indicating that the AE signal event amplitude did not
31
32 177 reach the threshold value. Fig.5 (a) shows the microstructure of the location of crack initiation. Lots
33
34 178 of fatigue striations and plastic deformation zones can be observed at crack initiation stage [46] . At
35
36 179 crack initiation stage, cyclic loading cause the strain localization in the form of the persistent slip
37
38 180 bands (PSBs). Repeated stretching and extrusion form plastic deformation zones. Above changes
39
40 181 release a little energy and form AE signal with low amplitude [47, 48]. Figure 4 (f) depicts the
41
42 182 smallest displacement in each cycle, which shows the metal bar's deformation degree. The
43
44
183 displacement change at the crack initiation stage accounts for only 5% of the entire process. This
45
46
184 result shows elastic deformation and a small amount of plastic deformation occurred at crack
47
48
49 185 initiation stage.
50
51 186
52
53 187 The crack propagation stage lasts from 3.8 s to 48.4 s. The intensity of the AE signal increases and
54
55 188 exceeds the threshold value as a result of the macro crack development, which results in an increase
56
57 189 of ring-down counts [49]. The energy fluctuation of the entire process is shown in Fig. 4 (c), and as
58
59 190 the metal bar is shifted from the crack initiation stage to the crack propagation stage, the AE energy
60
61
62
63
64
65
 191 increases.
1
2 192
3
4
193 At crack propagations stage, the amplitude of the stress intensity factor (SIF) K exceeds the
5
6
7 194 threshold value of SIF K th [50]. According to Paris law, crack growth rate depend on the SIF.
8
9 195 Increased SIF lead to the crack growth rate increased. High crack growth rate lead to the vibration
10
11 196 signal amplitude of metal bar increased which is shown in Fig.4 (d). As shown in Fig.4 (e), the
12
13
197 kurtosis of vibration signal is in a low value which reflect there is a small fluctuation of the
14
15
198 amplitude of vibration signal at this stage. This result is coincides with the law that the crack growth
16
17
18 199 rate increase steadily at crack propagation stage. Fig.4 (f) shows the displacement change accounts
19
20 200 for 95 percent of the entire process.
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 201
43
44 202 Fig.5 SEM images from the fracture surface (a) the location of crack imitation; (b) the location of crack
45
46 203 propagation stage phase I; (c) the location of crack propagation stage phase II; (d) the location of fracture.
47
48 204
49
50 205 Here is another interesting phenomenon：It can be seen that the amplitude of the vibration signal
51
52 206 first rises and then falls in Fig.4 (d). There are two phases in the crack propagation stage: phase I
53
54 207 lasts from 3.8 s to 20 s, phase II lasts form 20 s to 48.4 s. The ring-down counts vs. time also reflect
55
56 208 this phenomenon. As shown in Fig.4 (a), ring-down counts remain stable at 2400 in phase I, and
57
58 209 remain stable at 3900 in phase II. As shown in Fig.5 (b) and Fig.5(c), there are lots of tearing ridges
59
60 210 at the location of crack propagation stage phase I and lots of dimples at the location of crack
61
62
63
64
65
 211 propagation stage phase II. Tearing ridges are caused by shear damage. Dimples are caused by
1
2 212 tensile damage. SEM images show that shear failure model dominates crack propagation stage phase
3
4
213 I and tensile failure model dominates crack propagation stage phase II. According to the Muhammad
5
6
7 214 M.Sherif’ study [30], RA reflect failure model change well. The average RA value per second vs.
8
9 215 time is shown in Fig.6. It can be seen at crack propagation stage phase I, RA value is high.
10
11 216 Subsequently, RA decrease rapidly due to the large tensile damage replace original shear damage.
12
13 217 Finally, RA value remain a low value because of tensile failure model dominates the crack
14
15 218 propagation stage phase II.
16
17 219
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34 220
35
221 Fig.6 The average RA value per second vs. time.
36
37
38 222 To evaluate the cross-section quality of the crack propagation stage phase I and the crack
39
40 223 propagation stage phase II, a 3D microscopy is used to observe the transition area which between
41
42 224 two phases. As shown in Fig.7 (a), the transition line between phase I and phase II is shown by
43
44 225 yellow dotted line. SEM images of point A is shown in Fig.7 (c), it can be seen there are many
45
46 226 tearing ridges in phase I area and many dimples in phase II area which are coincidence with the
47
48 227 above description. Surface topography of the transition area is shown in Fig.7 (b), it can be seen the
49
50 228 value of surface roughness at phase II area is higher than that at phase I area. A measurement line is
51
52 229 cross phase I area and phase II area, and the line roughness is shown in Fig.7 (d). The height
53
54 230 difference of phase I area and phase II area are 207.37 μm and 427.73 μm , respectively. The
55
56 231 above results indicate that the cross-section quality is lower in the crack propagation stage phase II.
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 232
27
28 233 Fig.7 The cross-section quality evaluation of crack propagation stage phase I and phase II (a) OM image of the
29
30 234 transition area; (b) surface topography of the transition area; (c) SEM image of point A; (d) line roughness of
31
32 235 measurement line.
33
34 236
35
36 237 At fracture stage, because of the amplitude of SIF K approaching the material fracture toughness
37
38
39
238 K IC , the crack growth rate is incredibly fast. As shown in Fig.4 (a) and Fig. 4 (b), ring-down counts
40
41 239 and the kurtosis grow fast during this period. Lots of AE signal events occurred at this period, and
42
43 240 the metal bar is in an unstable state. The AE energy increased at the same time, which indicate that
44
45 241 the material's fracture is releasing a significant quantity of energy. Because the metal bar is separated
46
47 242 into two sections, there is insufficient vibration signal detected at the displacement load applied
48
49 243 point, resulting in the vibration signal's amplitude dropping to zero. The microstructure of the
50
51 244 fracture region is shown in Fig.5 (d), where there are more dimples and second phase particles
52
53 245 domains, and the size of the dimples is larger than what at the propagation stage. The above behavior
54
55
246 indicates pure tensile failure model dominates fracture stage.
56
57
247
58
59
60 248
61
62
63
64
65
 1 249 3.2 Effect of loading frequency on cross-section quality
2
3
4 250 Four different loading frequency are designed in this study: 20 Hz (1200 rpm), 8.33 Hz (500 rpm),
5
6 251 6.67 Hz (400 rpm), and 5 Hz (300 rpm). The displacement load is set as 3 mm. A 3D microscopy
7
8 252 which model is VHX-5000 is used to evaluate cross-section quality. Metal bars are subjected a high-
9
10
253 frequency loading at first 10 s and then subjected different low frequencies loading until fracture.
11
12
13
254 As shown in Fig.8, surface topography of metal bars subjected different loading frequencies are
14
15 255 presented. It can be seen there are three different areas: crack initiation area, crack propagation area,
16
17 256 and fracture area. The fracture area is the most uneven area of cross-section [51]. Loading frequency
18
19 257 decreasing lead to the fracture area reduce.
20
21 258
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52 259
53 260 Fig.8 Surface topography of 16Mn metal bar cross-section under different loading frequencies (a) 20 Hz ;(b) 8.33
54
55 261 Hz ; (c) 6.67 Hz; (d) 5 Hz.
56
57 262
58
59 263 To evaluate the cross-section quality, height difference of the cross-section (HDOC) is proposed.
60
61
62
63
64
65
 264 HDOC can be calculated by given equation:
1
2 265 HDOC  mean value of 10000 maximum points'height  mean value of 10000 minimum points'height (9)
3
4
266 Lower HDOC means the cross-section quality better. HDOC of metal bars subjected different
5
6
7 267 loading frequencies are shown in Table 4. It can be seen HDOC decreases significantly as the
8
9 268 loading frequency decreases. HDOC of metal bars subjected 8.83 Hz loading is the smallest.
10
11 269 Compare to HDOC of metal bars subjected 20 Hz loading, HDOC of metal bars subjected 8.83 Hz
12
13 270 loading decreased by 31%. Above results show there is a big influence of loading frequency on the
14
15 271 cross-section quality, and the optimal loading frequency is 8.33 Hz.
16
17 272
18
19 273 Table 4 HDOC of metal bars under different loading frequencies
20
21 Loading frequency (Hz) 20 8.33 6.67 5
22
23 HDOC ( μm ) 1343.71 925.78 1005.51 1217.5
24
25
26 274
27 275
28
29
30
31 276 4. RA-based control scheme
32
33
34 277 According to the above results, the most uneven area of cross-section at crack propagation stage
35
36 278 phase II and fracture stage. Both of these two stage are dominated by tensile failure model.
37
38 279 Decreasing tensile failure model influence area will improve cross-section quality. Because RA
39
40 280 reflect failure model well, a RA-based control scheme is proposed.
41
42
281
43
44
282 As shown in Fig.9 (a), the average RA values for 10 s – 20 s, 15 s-20 s, 20 s-30 s, and 30 s-40 s are
45
46
47 283 found to be 0.0029536, 0.0026321, 0.0024941, and 0.0024293 respectively. These values are set as
48
49 284 loading frequency conversion control point, which is named T. Fig.9 (b) shows the calculate
50
51 285 procedure of the control scheme. The initial high loading frequency is set as 20 Hz, and the RA
52
53 286 value is the mean value during every 0.1 s. When the RA value is greater than T, the loading
54
55 287 frequency remains unchanged, and when the RA value is less than or equal to T, the speed is reduced
56
57 288 to low frequency loading (8.33 Hz) until the bar is separated. In order to reduce the test error, a total
58
59 289 of 20 metal bar were selected for the test.
60
61
62
63
64
65
 290
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17 291
18
19 292 Fig.9 RA based control scheme (a) control position (b) the calculate procedure of the control scheme
20 293
21
22 294 As shown in Fig.10, the fracture surface of metal bars with different T are presented. It can be seen,
23
24 295 that when T increased, the fracture area decreased. When T=0.0029536 or T=0.0026321, no fracture
25
26 296 area can be found. Fig.11 presented the average HDOC and the average cropping time of metal bars
27
28 297 with different T. When T increased from 0.0024393 to 0.0024941, the average HDOC decreased
29
30 298 from 1302.15 μm to 1234.62 μm , and the average cropping time increased from 62.2 s to 80.6 s. At
31
32 299 this situation, the average HDOC decrease 5.19%, while cropping time increase 29.58%. There is a
33
34 300 few improvement of cross-section but a big decrease in cropping efficiency. When T increased from
35
36 301 0.0024941 to 0.0026321, the average HDOC decreased from 1234.62 μm to 987.53 μm , while the
37
38 302 average cropping time increased from 80.6 s to108.6 s. At this situation, the average HDOC decrease
39
40 303 20.01 %, while the average cropping time increase 34.74 %. The improvement of cross-section
41
42 304 quality is significant. When T increased from 0.0026321 to 0.0029536, the average HDOC
43
44 305 decreased from 987.53 μm to 931.21 μm , and the average cropping time increased from 108.6 s to
45
46
306 146.4 s. At this situation, the average HDOC decrease 5.70 %, while cropping time increase 34.80 %.
47
48
49 307 There is a few improvement of cross-section but a big decrease in cropping efficiency. Considering
50
51 308 the cropping efficiency and cross-section quality simultaneously, best T is 0.0026321.
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28 309
29 310 Fig.10 Fracture surface of metal bars with different T value
30 311
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45 312
46
47 313 Fig.11 The average HDOC and the average cropping time of metal bars with different T values (a) the average
48 314 HDOC vs. T value (b) the average cropping time vs. T value
49
50 315
51
52
53 316 5. Conclusions
54
55
56 317 A novel AET-based LCFC system is built. The fracture characteristics and process optimization
57
58
318 of notched 16Mn metal bar in LCFC are studied, and the specific conclusions are as follows:
59
60
61
62
63
64
65
 319 (1) AE Ring-down counts, AE kurtosis, AE energy, vibration signal amplitude, vibration signal
1
2 320 kurtosis and smallest displacement in per loading cycle reflect fracture characteristics of 16Mn
3
4
321 metal bar in LCFC effectively. According to the results, three stages exit in 16Mn metal bar
5
6
7 322 LCFC process: crack initiation stage, crack propagation stage and fracture stage. The crack
8
9 323 initiation stage cover 5 % LCFC process, while the crack propagation stage cover almost 95 %
10
11 324 LCFC process.
12
13 325 (2) There are two phases in the crack propagation stage. The crack propagation stage phase I is
14
15 326 dominated by shear failure model, and the crack propagation stage phase II is dominated by
16
17 327 tensile failure model. The most uneven area of cross-section at the crack propagation stage and
18
19 328 fracture stage, which caused by tensile failure model.
20
21 329 (3) The loading frequency has a great impact on the cross-section quality. Reducing the loading
22
23 330 frequency could reduce the fracture area. A new parameter HDOC is proposed to evaluate the
24
25 331 cross-section quality. When loading frequency is 8.33 Hz, HDOC is smallest and the cross-
26
27 332 section quality is best.
28
29 333 (4) RA reflect failure model’s change well. A RA-based control scheme is proposed, which by
30
31 334 reducing loading frequency to 8.33 Hz when RA exceed T. Results shows the cross-section
32
33 335 quality is improved by using RA-based control scheme. The effects of control scheme with 4
34
35
336 different T values are compared by combing cropping efficiency and cross-section quality. The
36
37
38 337 best T is 0.0026321.
39
40 338
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 339 Nomenclature
2
3
4 340
5
6
7 341 Abbreviation
8
9
10
11 342
12
13 343 LCFC
14
15 344 Low-cycle fatigue cropping
16
17 345 SIF
18
19 346 Stress intensity factor
20
21 347 AET
22
23 348 Acoustic emission technique
24
25 349 UTS
26
27 350 Ultimate tensile strength
28
29 351 OM
30
31 352 Optical microscopy
32
33 353 SEM
34
35
354 Scanning Electron Microscope
36
37
38 355 SCF
39
40 356 Stress concentration factor
41
42 357 PSBs
43
44 358 Persistent slip bands
45
46 359 RA
47
48 360 The ratio of rise time to amplitude
49
50 361 HDOC
51
52 362 Height difference of the cross-section
53
54 363
55
56
57
58
59
60
61
62
63
64
65
 1 364 Variable
2
3
4 365
5
6 366 D
7
8 367 Diameter of metal bar
9
10
368 h
11
12
13
369 Depth of the prefabricated notch
14
15 370 r
16
17 371 Radius of the prefabricated notch
18
19 372 L
20
21 373 The distance between loading position and notch certain line
22
23 374 L1
24
25 375 The distance between clamping position and notch certain line
26
27 376 
28
29
30 377 Notch angle
31
32 378 E
33
34 379 Young’s Modulus
35
36 380 y
37
38
39 381 Yield stress
40
41 382 K
42
43
383 Stress concentration factor
44
45
46 384 K IC
47
48 385 Fracture toughness
49
50
51 386 Fci
52
53 387 Critical crack initial load
54
55 388 Fcp
56
57
58 389 Critical crack propagation load
59
60
61
62
63
64
65
 1 390 Tr
2
3 391 Required torque
4
5 392 P
6
7 393 Required force
8
9 '
10 394 L
11
12 395 Lead of the ballscrew
13
14 396 
15
16 397 The efficiency of the screw
17
18 398 X
19
20 399 The eccentric displacement load
21
22
23 400 Fi
24
25 401 Friction force
26
27 402 
28
29 403 The friction coefficient
30
31 404 T
32
33 405 RA value at control position
34
35 406 a
36
37
38 407 The depth of the rig crack
39
40 408 
41
42 409 normal stress
43
44 410 
45
46 411 Poisson’s ratio
47
48 412 u
49
50
51 413 Ultimate tensile strength
52 414
53
54 415
55
56
57
58
59
60
61
62
63
64
65
 416
1
2
3
4
417 References:
5
6
418 [1] D. Albrecht, H. Möhring, Potentials for the optimization of sawing processes using the example of
7
8 419 bandsawing machines, Procedia Manufacturing 21 (2018) 567-574.
9 420 [2] M.A. Zeki Cinar, Z. Qasim., Developments in Plasma Arc Cutting (PAC) of Steel Alloys: A Review,
10
11 421 Jurnal Kejuruteraan (2018).
12 422 [3] K. Pandey, S. Datta, Hot machining of difficult-to-cut materials: A review, Materials Today:
13 423 Proceedings 44 (2021) 2710-2715.
14
15 424 [4] S. Thipprakmas, M. Jin, M. Murakawa, An investigation of material flow analysis in fineblanking
16 425 process, JOURNAL OF MATERIALS PROCESSING TECHNOLOGY 192-193 (2007) 237-
17
426 242.
18
19 427 [5] Y. Dong, Y. Ren, S. Fan, Y. Wang, S. Zhao, Investigation of Notch-Induced Precise Splitting of
20 428 Different Bar Materials under High-Speed Load, Materials 13 (2020) 2461.
21
22 429 [6] C.J. Hua, S.D. Zhao, L.J. Zhang, W. Liu, Investigation of a new-type precision cropping system
23 430 with variable-frequency vibration, INTERNATIONAL JOURNAL OF MECHANICAL
24 431 SCIENCES 48 (2006) 1333-1340.
25
26 432 [7] Y. Dong, J. Li, Y. Ren, S. Fan, S. Zhao, Laser-assisted cyclic chipless splitting for hard-to-cut thick
27 433 wall tubes and fatigue fracture mechanism analysis, INTERNATIONAL JOURNAL OF
28
434 MECHANICAL SCIENCES 168 (2020) 105308.
29
30 435 [8] Y. Ren, Y. Dong, J. Li, F. Zhao, S. Zhao, L. Zhang, Y. Wang, D. Jin, The investigation of low-cycle
31 436 fatigue crack propagation of 16 Mn eccentric bar based on the acoustic emission technique,
32
33 437 Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering
34 438 Manufacture 235 (2021) 1235-1247.
35 439 [9] L.J. Zhang, S.D. Zhao, J. Lei, W. Liu, Investigation on the bar clamping position of a new type of
36
37 440 precision cropping system with variable frequency vibration, International Journal of Machine
38 441 Tools and Manufacture 47 (2007) 1125-1131.
39
442 [10] B. Zhong, S. Zhao, R. Zhao, J. Liao, Numerical and experimental investigation on the influence of
40
41 443 main motor rotational frequency in fine-cropping, Proceedings of the Institution of Mechanical
42 444 Engineers, Part C: Journal of Mechanical Engineering Science 228 (2014) 514-524.
43
44 445 [11] R. Zhao, S. Zhao, B. Zhong, Y. Tang, Experimental investigation on new low cycle fatigue precision
45 446 cropping process, Proceedings of the Institution of Mechanical Engineers, Part C: Journal of
46 447 Mechanical Engineering Science 229 (2015) 1470-1476.
47
48 448 [12] L. Zhang, X. Chen, H. Wang, S. Zhao, N. Li, D. Zhang, Research on critical loading force in
49 449 precision cropping system based on hydraulic compensation, INTERNATIONAL JOURNAL OF
50
450 MECHANICAL SCIENCES 142-143 (2018) 44-50.
51
52 451 [13] R.T. Committee, Recommendation of RILEM TC 212-ACD: acoustic emission and related NDE
53 452 techniques for crack detection and damage evaluation in concrete*, MATERIALS AND
54
55 453 STRUCTURES 43 (2010) 1183-1186.
56 454 [14] J. Liu, Y. Xu, G. Pan, A combined acoustic and dynamic model of a defective ball bearing,
57 455 JOURNAL OF SOUND AND VIBRATION 501 (2021) 116029.
58
59 456 [15] D.G. Aggelis, A.C. Mpalaskas, T.E. Matikas, Investigation of different fracture modes in cement-
60 457 based materials by acoustic emission, CEMENT AND CONCRETE RESEARCH 48 (2013) 1-8.
61
62
63
64
65
 458 [16] M. Chai, J. Zhang, Z. Zhang, Q. Duan, G. Cheng, Acoustic emission studies for characterization of
1 459 fatigue crack growth in 316LN stainless steel and welds, APPLIED ACOUSTICS 126 (2017)
2
3 460 101-113.
4 461 [17] H.L. Duanegan, D.O. Harris, C.A. Tatro, Fracture analysis by use of acoustic emission,
5
462 ENGINEERING FRACTURE MECHANICS 1 (1968).
6
7 463 [18] J. Wang, J. Li, Z. Shi, Deformation damage and acoustic emission characteristics of red sandstone
8 464 under fatigue–creep interaction, THEORETICAL AND APPLIED FRACTURE MECHANICS
9
10 465 117 (2022) 103192.
11 466 [19] D.O. Harris, H.L. Dunegan, Continuous Monitoring of Fatigue-crack Growth Acoustic-emission
12 467 Techniques., Los Angeles,CA, 1973.
13
14 468 [20] D.E.W. Stone, P.F. Dingwall, Acoustic emission parameters and their interpretation, NDT
15 469 International 10 (1977) 51-62.
16
17
470 [21] B.J. Brindley, J. Holt, I.G. Palmer, acoustic emission-3 The use of ring-down counting, Non
18 471 Destructive Testing (1973).
19 472 [22] L. Dou, K. Yang, X. Chi, Fracture behavior and acoustic emission characteristics of sandstone
20
21 473 samples with inclined precracks, International Journal of Coal Science & Technology 8 (2021)
22 474 77-87.
23 475 [23] Z. He, K. Zhao, Y. Yan, F. Ning, Y. Zhou, Y. Song, Mechanical response and acoustic emission
24
25 476 characteristics of cement paste backfill and rock combination, CONSTRUCTION AND
26 477 BUILDING MATERIALS 288 (2021) 123119.
27
28 478 [24] C. Ruiz-Cárcel, E. Hernani-Ros, Y. Cao, D. Mba, Use of Spectral Kurtosis for Improving Signal to
29 479 Noise Ratio of Acoustic Emission Signal from Defective Bearings, Journal of Failure Analysis
30 480 and Prevention 14 (2014) 363-371.
31
32 481 [25] E.N. Landis, L. Baillon, Experiments to Relate Acoustic Emission Energy to Fracture Energy of
33 482 Concrete, JOURNAL OF ENGINEERING MECHANICS 128 (2002) 698-702.
34
483 [26] N.B. Burud, J.M.C. Kishen, Response based damage assessment using acoustic emission energy for
35
36 484 plain concrete, CONSTRUCTION AND BUILDING MATERIALS 269 (2021) 121241.
37 485 [27] X. Ji, W. Zhou, H. Sun, J. Liu, L. Ma, Damage evolution behavior of bi-adhesive repaired
38
39 486 composites under bending load by acoustic emission and micro-CT, COMPOSITE
40 487 STRUCTURES 279 (2022) 114742.
41 488 [28] M. Chai, Z. Zhang, Q. Duan, Y. Song, Assessment of fatigue crack growth in 316LN stainless steel
42
43 489 based on acoustic emission entropy, INTERNATIONAL JOURNAL OF FATIGUE 109 (2018)
44 490 145-156.
45
491 [29] D. D'Angela, M. Ercolino, Acoustic emission entropy: An innovative approach for structural health
46
47 492 monitoring of fracture‐critical metallic components subjected to fatigue loading, FATIGUE &
48 493 FRACTURE OF ENGINEERING MATERIALS & STRUCTURES 44 (2021) 1041-1058.
49
50 494 [30] M.M. Sherif, E.M. Khakimova, J. Tanks, O.E. Ozbulut, Cyclic flexural behavior of hybrid
51 495 SMA/steel fiber reinforced concrete analyzed by optical and acoustic techniques, COMPOSITE
52 496 STRUCTURES 201 (2018) 248-260.
53
54 497 [31] D.G. Aggelis, Classification of cracking mode in concrete by acoustic emission parameters,
55 498 MECHANICS RESEARCH COMMUNICATIONS 38 (2011) 153-157.
56
499 [32] J.D. Butterfield, A. Krynkin, R.P. Collins, S.B.M. Beck, Experimental investigation into vibro-
57
58 500 acoustic emission signal processing techniques to quantify leak flow rate in plastic water
59 501 distribution pipes, APPLIED ACOUSTICS 119 (2017) 146-155.
60
61
62
63
64
65
 502 [33] Z. Han, H. Luo, J. Cao, H. Wang, Acoustic emission during fatigue crack propagation in a micro-
1 503 alloyed steel and welds, Materials Science and Engineering: A 528 (2011) 7751-7756.
2
3 504 [34] Z. Han, H. Luo, Y. Zhang, J. Cao, Effects of micro-structure on fatigue crack propagation and
4 505 acoustic emission behaviors in a micro-alloyed steel, Materials Science and Engineering: A 559
5
506 (2013) 534-542.
6
7 507 [35] L. Li, Z. Zhang, G. Shen, Influence of grain size on fatigue crack propagation and acoustic emission
8 508 features in commercial-purity zirconium, Materials Science and Engineering: A 636 (2015) 35-
9
10 509 42.
11 510 [36] J. Li, H. Qiu, D. Zhang, S. Zhao, Y. Zhao, Acoustic emission characteristics in eccentric rotary
12 511 cropping process of stainless steel tube, The International Journal of Advanced Manufacturing
13
14 512 Technology 92 (2017) 777-788.
15 513 [37] J.L.Y.D. Yujian Ren, The investigation of the optimization scheme of the low-cycle fatigue cropping
16
17
514 based on the acoustic emission technique, Archives of Civil and Mechanical Engineering (2022).
18 515 [38] Y. Ren, B. Liu, X. Wang, Y. Dong, J. Li, S. Zhao, Optimization scheme of the precision separation
19 516 process of the notched bar based on the acoustic emission hit energy angle, THEORETICAL
20
21 517 AND APPLIED FRACTURE MECHANICS 119 (2022) 103328.
22 518 [39] АВВерхвский, Stress determination of dangerous section of complex shape parts.
23 519 Zhang Dehui, translation., Beijing, 1964.
24
25 520 [40] Handbook of stress intensity factorsScience Press, Beijing, 1993.
26 521 [41] P. Kumar, R. Jayaraj, J. Suryawanshi, U.R. Satwik, J. McKinnell, U. Ramamurty, Fatigue strength
27
28 522 of additively manufactured 316L austenitic stainless steel, ACTA MATERIALIA 199 (2020)
29 523 225-239.
30 524 [42] R.D.S.K. S. K. Balijepalli, Young's Modulus Profile in Kolsterized AISI 316L Steel, 762 (2013)
31
32 525 183-188.
33 526 [43] Y. Shang, Y. Yuan, D. Li, Y. Li, J. Chen, Effects of scanning speed on in vitro biocompatibility of
34
527 316L stainless steel parts elaborated by selective laser melting, The International Journal of
35
36 528 Advanced Manufacturing Technology 92 (2017) 4379-4385.
37 529 [44] S. Doddamani, M. Kaleemulla, Fracture toughness investigations of Al6061-Graphite particulate
38
39 530 composite using compact specimens, Frattura ed Integrità Strutturale 11 (2017) 484-490.
40 531 [45] P.Y.S.X. Qu Jia, Comparative experimental study on dynamic and static fracture toughness of 316L
41 532 three-point bending specimens, China test (2016).
42
43 533 [46] R. Masoudi Nejad, K. Farhangdoost, M. Shariati, Microstructural analysis and fatigue fracture
44 534 behavior of rail steel, MECHANICS OF ADVANCED MATERIALS AND STRUCTURES 27
45
535 (2020) 152-164.
46
47 536 [47] M.N. Batista, M.C. Marinelli, I. Alvarez-Armas, Effect of initial microstructure on surface relief
48 537 and fatigue crack initiation in AISI 410 ferritic-martensitic steel, FATIGUE & FRACTURE OF
49
50 538 ENGINEERING MATERIALS & STRUCTURES 42 (2019) 61-68.
51 539 [48] S. Lavenstein, J.A. El-Awady, Micro-scale fatigue mechanisms in metals: Insights gained from
52 540 small-scale experiments and discrete dislocation dynamics simulations, Current Opinion in Solid
53
54 541 State and Materials Science 23 (2019) 100765.
55 542 [49] R. Jiang, F. Dai, Y. Liu, A. Li, P. Feng, Frequency Characteristics of Acoustic Emissions Induced
56
543 by Crack Propagation in Rock Tensile Fracture, ROCK MECHANICS AND ROCK
57
58 544 ENGINEERING 54 (2021) 2053-2065.
59 545 [50] L.F.F.T. Ricardo, Crack Propagation in the Threshold Stress Intensity Region a Short Review. In:
60
61
62
63
64
65
 546 (Ed.), Mechanical Fatigue of Metals2019.
1 547 [51] W. Macek, R. Branco, M. Szala, Z. Marciniak, R. Ulewicz, N. Sczygiol, P. Kardasz, Profile and
2
3 548 Areal Surface Parameters for Fatigue Fracture Characterisation, Materials 13 (2020) 3691.
4 549
5
6 550
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 551 Statements and Declarations
2
3
4
5 552 Funding
6
7
8 553 This work is supported by the Aviation Joint Fund Project of National Natural Science Foundation
9
10 554 of China (U1937203), National Natural Science Foundation of China (Grant No. 52105398) ，
11
12 555 Natural Science Foundation Research Program of Shaanxi Province of China (Grant No. 2022JQ-
13
14 556 440) ，Xi 'an Science and Technology project (2017xasjl009), and China Scholarship Council (CSC
15
16
557 NO. 202006280402)
17
18 558
19
20
21
22
559 Competing Interests
23
24
25 560 The authors have no relevant financial or non-financial interests to disclose.
26
27
28 561 Author Contributions
29
30
31 562 All authors contributed to the study conception and design. Material preparation, data collection and
32 563 analysis were performed by Yujian Ren, Boyang Liu,Yi Zhang, Shengdun Zhao and Jinzhou Gao.
33
34 564 Experiment operation is performed by Yujian Ren, Yuanzhe Dong, Dong Jin. The first draft of the
35 565 manuscript was written by Ren Yujian and all authors commented on previous versions of the
36
37
566 manuscript. All authors read and approved the final manuscript.
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65