View
6
Download
0
Category
Preview:
Citation preview
1
Determination of the Optimal Parameters
in Resistance Projection Welding for a
Wheel Starter of Motorcycle Engines
Dr.Chuckaphun Aramphongphun Mr.Suriya Salikul
Department of Industrial Engineering,
Faculty of Engineering, Kasetsart University THAILAND
2 Operation of the start system of the motorcycle engine
• A case study factory is an automotive and motorcycle part manufacturer. Forming and welding of sheet metals are the main manufacturing processes.
• The factory would like to reduce the manufacturing cost and time by developing alternative welding process and also maintain the same quality of the welded wheel starter.
1. Introduction
Motor Start System
Wheel Starter
(Pulley Starter)
3
1. Introduction (cont.)
• A wheel starter (or pulley starter) is a main component in the motor start system of the motorcycle engine.
• It serves as a driving gear to start the engine of motorcycle.
A wheel starter of the motorcycle engine
4
1. Introduction (cont.)
A Flow Process Chart
BOSS GEAR
RESISTANCE
PROJECTION
WELDING
SURFACE HARDENING
MACHINING
CIRCLIP &
BEARING
STORAGE
5
The wheel starter installation in the engine
A location of the wheel starter in the motor start system
1. Introduction (cont.)
6
• Plasma Arc Welding (PAW) is originally used to join the boss
and gear to make the wheel starter.
The welded part after Plasma Arc Welding
Plasma Arc Welding (Original Welding Process)
1. Introduction (cont.)
7
Comparison of various welding processes
1. Introduction (cont.)
Electron Beam Welding (EBW)
Plasma Arc Welding (PAW)
Resistance Projection Welding (RPW)
Welding cycle time = 30 sec Welding cycle time = 47 sec Welding cycle time = 9 sec
8
Comparison between PAW and RPW
1. Introduction (cont.)
Features Types of Welding Process
PAW RPW
Welding Depth 2 1
Torsion Torque 2 2
Welding Height 1 2
Welding Bead 1 2
Machine Cost 1 2
Electric Cost 2 1
Welding Wire Cost 0 2
Shield Gas Cost 0 2
Cycle Time 1 2
2 = Very good, 1 = Good, 0 = Poor
9
• By comparing various welding processes in terms of reduced cost, time, and quality of the part, Resistance Projection Welding (RPW) is the most suitable welding process used to form the wheel starter.
Resistance Projection Welding used in the experiment
Comparison among welding processes
1. Introduction (cont.)
10
An example shows the macro section of a weld
nugget after the welding time has ended.
1->2 Lowering of the top electrode
2->3 Application of the adjusted electrode force Set-up time tpre, sequence
3->4 Switching-on of the adjusted welding current for the period of the welding time tw.
Formation of the weld nugget in the joining zone of both workpieces.
4->5 Maintaining the electrode force for the period of the set post-weld holding time th.
5->6 Switching-off the force generating system and lifting the electrodes off the workpiece.
Cycle of Resistance Welding
1. Introduction (cont.)
11
• To study and analyze the factors that significantly
affect the quality of the wheel starter using
Resistance Projection Welding
• To determine the optimal parameters in Resistance
Projection Welding and set it as the production
standard
• To study the microstructure and hardness in Heat
Affected Zone after Resistance Projection Welding
2. Research Objectives
12
3. Design of Experiments
A Flow Chart of the Design of Experiments
Start
Full Factorial
Design with
Center Points
RSM Design
(CCD)
2nd Experiments
1st Experiments
Statistical
Analysis Confirm Run
Summary
Statistical
Analysis
13
RPW
Resistance welding machine used in the experiments
3. Design of Experiments (cont.)
14
Note: 1 Cycle = 50 Hz = 1/50 sec
Levels of each factor used in the experiments
3. Design of Experiments (cont.)
Factors Low (-1)
Middle (0)
High (+1)
Current (kA) 44.5 45.0 45.5
Time (Cycle) 19 20 21
Pressure (MPa) 0.35 0.40 0.45
Margin (mm) 0.500 0.5275 0.555
15
- Full Factorial design: 24 x 2 = 32 runs
- Center points = 4 runs
- Total runs = 36 runs (32 + 4)
Number of the Full Factorial experiments
3. Design of Experiment (cont.)
16
4. Results and Analysis Factorial Fit: Torque versus Current, Time, Pressure, Margin Estimated Effects and Coefficients for Torque (coded units)
Term Effect Coef SE Coef T P
Constant 1524.44 7.589 200.87 0.000
Current 16.62 8.31 7.589 1.10 0.287
Time -18.87 -9.44 7.589 -1.24 0.229
Pressure 2.75 1.37 7.589 0.18 0.858
Margin 24.37 12.19 7.589 1.61 0.125
Current*Time 15.63 7.81 7.589 1.03 0.316
Current*Pressure 1.00 0.50 7.589 0.07 0.948
Current*Margin -14.38 -7.19 7.589 -0.95 0.355
Time*Pressure 12.50 6.25 7.589 0.82 0.420
Time*Margin 10.38 5.19 7.589 0.68 0.503
Pressure*Margin -13.00 -6.50 7.589 -0.86 0.402
Current*Time*Pressure -7.50 -3.75 7.589 -0.49 0.627
Current*Time*Margin 32.38 16.19 7.589 2.13 0.046
Current*Pressure*Margin 11.25 5.63 7.589 0.74 0.468
Time*Pressure*Margin 3.00 1.50 7.589 0.20 0.845
Current*Time*Pressure*Margin -12.00 -6.00 7.589 -0.79 0.439
Ct Pt -52.19 22.768 -2.29 0.033
S = 42.9312 PRESS = 135202
R-Sq = 51.85% R-Sq(pred) = 0.00% R-Sq(adj) = 11.30%
17
80400-40-80
99
90
50
10
1
Residual
Pe
rce
nt
160015501500
50
25
0
-25
-50
Fitted Value
Re
sid
ua
l
6040200-20-40-60
10.0
7.5
5.0
2.5
0.0
Residual
Fre
qu
en
cy
35302520151051
50
25
0
-25
-50
Observation Order
Re
sid
ua
l
Normal Probability Plot Versus Fits
Histogram Versus Order
Residual Plots for TorqueNormality
Constant Variance
Independence Normal curve
4. Results and Analysis (cont.)
18
Factorial Fit: Torque versus Current, Time, Margin Estimated Effects and Coefficients for Torque (coded units)
Term Effect Coef SE Coef T P
Constant 1524.44 6.835 223.05 0.000
Current 16.63 8.31 6.835 1.22 0.233
Time -18.87 -9.44 6.835 -1.38 0.178
Margin 24.38 12.19 6.835 1.78 0.085
Current*Time*Margin 32.37 16.19 6.835 2.37 0.025
Ct Pt -52.19 20.504 -2.55 0.016
S = 38.6623 PRESS = 63807.7
R-Sq = 38.34% R-Sq(pred) = 12.26% R-Sq(adj) = 28.06%
4. Results and Analysis (cont.)
19
100500-50-100
99
90
50
10
1
Residual
Pe
rce
nt
15601540152015001480
80
40
0
-40
-80
Fitted Value
Re
sid
ua
l
806040200-20-40-60
8
6
4
2
0
Residual
Fre
qu
en
cy
35302520151051
80
40
0
-40
-80
Observation Order
Re
sid
ua
l
Normal Probability Plot Versus Fits
Histogram Versus Order
Residual Plots for TorqueNormality
Constant Variance
Independence Normal curve
4. Results and Analysis (cont.)
20
0.555
0.5
21
1945.544.5
Margin
Time
Current
1472.25
1557.50
1518.001563.75
1507.25
1504.75
1550.751503.00
1490.50
Centerpoint
Factorial Point
Cube Plot (data means) for Torque
Level of the factors that provides the highest torque is as follows: Current = 44.5 kA, Time = 19 cycle, Margin = 0.555 mm
4. Results and Analysis (cont.)
21
Response Surface Regression: Torque versus Block, Current, Time, Margin The analysis was done using coded units.
Estimated Regression Coefficients for Torque
Term Coef SE Coef T P
Constant 1536.40 10.845 141.673 0.000
Block -60.36 7.071 -8.536 0.000
Current 2.05 7.081 0.290 0.775
Time -4.60 7.081 -0.650 0.522
Margin 11.50 7.081 1.624 0.118
Current*Current 38.03 14.142 2.689 0.013
Time*Time -3.97 14.142 -0.281 0.781
Margin*Margin 12.53 14.142 0.886 0.385
Current*Time 11.88 7.917 1.500 0.147
Current*Margin -3.12 7.917 -0.395 0.697
Time*Margin 4.31 7.917 0.545 0.591
S = 31.6677 PRESS = 56396.5
R-Sq = 80.28% R-Sq(pred) = 51.79% R-Sq(adj) = 71.71%
4. Results and Analysis (cont.)
22
The optimal condition that provides the maximum torque based on the RSM model
4. Results and Analysis (cont.)
23
The Surface Plot of the RSM Model
4. Results and Analysis (cont.)
1550
1575
1600
2144.5
2045.0
1945.5
ue
TimeCurrent
Margin 0.555
Hold Values
Surface Plot of Torque vs Time, Current
Torque
24
The Contour Plot of the RSM Model
Current
Tim
e
45.5045.2545.0044.7544.50
21.0
20.5
20.0
19.5
19.0
Margin 0.555
Hold Values
>
–
–
–
–
< 1560
1560 1570
1570 1580
1580 1590
1590 1600
1600
Torque
Contour Plot of Torque vs Time, Current
4. Results and Analysis (cont.)
25
Confirmation Runs at the Optimal Parameter
4. Results and Analysis (cont.)
Factors Optimal Level
Current (kA) 44.5
Time (Cycle) 19
Margin (mm) 0.555
26
Confirmation Runs at the Optimal Parameter (cont.)
4. Results and Analysis (cont.)
Tests Measured Torque (N.m)
1 1607.62
2 1610.14
3 1607.59
4 1605.10
5 1602.60
6 1607.13
7 1612.12
8 1604.42
9 1610.18
10 1609.91
27
Hypothesis Testing of the Confirmation Runs
4. Results and Analysis (cont.)
28
One-Sample T: Confirm Run Test of mu = 1600 vs > 1600
95% Lower
Variable N Mean StDev SE Mean Bound T P
Confirm Run 10 1607.68 2.99 0.94 1605.95 8.13 0.000
P-value < (0.05), thus reject H0
According to results of the confirmation runs, it was found that
the average torque was greater than 1,600 N-m. A 95% confidence
interval was greater than 1,605.95 N-m. Therefore, the result of the experiments agreed with the values obtained from the RSM analysis.
4. Results and Analysis (cont.)
29
5. A Study of Microstructure in Heat Affected Zone
Microstructure in Heat Affected Zone (HAZ) after
Resistance Projection Welding was studied and viewed by a
microscope. This study compared (i) the microstructure in HAZ and
(ii) hardness in Plasma Arc Welding (PAW) and Resistance
Projection Welding (RPW).
30
A microstructure in HAZ of the boss at a distance of 0.2 mm from the welded joint with magnification of 2,000x
(a) Plasma Arc Welding and (b) Resistance Projection Welding
(a) (b)
5. A Study of Microstructure in Heat Affected Zone (cont.)
31
(a) (b)
A microstructure in HAZ of the boss at a distance of 0.4 mm from the welded joint with magnification of 2,000x
(a) Plasma Arc Welding and (b) Resistance Projection Welding
5. A Study of Microstructure in Heat Affected Zone (cont.)
32
(a) (b)
A microstructure in HAZ of the gear at a distance of 0.2 mm from the welded joint with magnification of 2,000x
(a) Plasma Arc Welding and (b) Resistance Projection Welding
5. A Study of Microstructure in Heat Affected Zone (cont.)
33
(a) (b)
A microstructure in HAZ of the gear at a distance of 0.4 mm from the welded joint with magnification of 2,000x
(a) Plasma Arc Welding and (b) Resistance Projection Welding
5. A Study of Microstructure in Heat Affected Zone (cont.)
34
Hardness in Heat Affected Zone after Plasma Arc Welding
Boss
(S
35C
)
Gear
(S
PH
C)
HAZ
6. A Study of Hardness in Heat Affected Zone
35
Gear (SPHC)
Boss (S35C)
HAZ
6. A Study of Hardness in Heat Affected Zone (cont.)
Hardness in Heat Affected Zone after Resistance Projection Welding
36
• The Resistance Projection Welding (RPW) has been
recently developed to join the wheel starter. • Process parameters (factors) in RPW that significantly
affected the torsion torque of the welded wheel starter
were: (i) Current, (ii) Time, and (iii) Margin.
• The pressure used in the experiments was not significant.
• The optimal setting of the significant parameters was as
follows:
Current = 44.5 kA
Time = 19 cycle
Margin = 0.555 mm
7. Summary
37
• According to results of the confirmation runs, it was found
that the average torque was greater than 1,600 N-m. (Target
specification is greater than 1,600 N-m.)
• A 95% confidence interval was greater than 1,605.95 N-m.
Therefore, the result of the experiments agreed well with the
values obtained from the RSM model.
• Moreover, RPW also significantly reduced (i) the welding
cycle time from 47 to 9 sec (-81%) and (ii) the welding cost.
7. Summary (cont.)
38
• The welded joint has experienced high temperature
and resulted in the residual stress. Microstructure in
Heat Affected Zone was therefore different from that
in the unaffected base metal zone.
• Microstructure changes resulted from Plasma Arc
Welding were more severe than those resulted from
Resistance Projection Welding and led to the non-
uniform microstructure. In addition, too high
hardness could result in brittle structure.
7. Summary (cont.)
39
Thank you for your attention.
QUESTIONS?
Recommended