jian chen, zhili feng (pi), c. david warren · reflection reduction/minimization as-welded surface...
TRANSCRIPT
Managed by UT-Battellefor the Department of Energy
Nondestructive Inspection of RSW of
AHHS by Infrared Technology
Jian Chen, Zhili Feng (PI), C. David Warren
Oak Ridge National Laboratory
April, 2014
2 Managed by UT-Battellefor the Department of Energy
Industry has been relying on pry-check and teardown for resistance spot weld (RSW) quality inspection and control– Labor intensive and expensive (rework and scraps)
– Less effective for advanced high-strength steels, aluminum and other lightweight materials, impeding the widespread use for light-weighting
AMD 605: In-Line Resistance Spot Welding (RSW) Control and Evaluation System Assessment for Light Weight Materials
Background
In 2007-2008, USCAR conducted an assessment of the industry needs and state-of-the-art of NDE technologies (AMD605)– Industry needs: real-time inspection and postmortem (post-welding)
inspection
– Lack of viable NDE may discourage the use of lightweight materials: high strength steel and aluminum
– “No economically reliable method exists today to inspect resistance spot welds at production rates, and the only technical needs identified in the project for further development consideration were in infrared thermography technology”
3 Managed by UT-Battellefor the Department of Energy
Objective
• Develop a fully automated online non-destructive evaluation (NDE) technology for RSW quality monitoring based on infrared (IR) thermography that can be adopted reliably and cost-effectively in high-volume auto production environment for weld quality assessment An expert system including hardware and software
Capable for both post-weld and real-time on-line weld quality inspection
Weld quality database covering wide range of weld configurations (materials, thickness, coatings) common in auto-body structures
4 Managed by UT-Battellefor the Department of Energy
Past Attempts on IR NDE Postmortem NDE with mostly
limited to laboratory trials– Flash lamp heating source: pulsed
heating in milli-seconds
– Positive correlation with weld quality
– However, minimal surface temperature change (low signal-to-noise ratio) Highly sensitive to surface condition
and environment interference Requiring painting of weld surface –
impractical in auto production line
No successful attempts as real-time NDE
5 Managed by UT-Battellefor the Department of Energy
Major Problem of Conventional IR NDE
012345678
100 200 300 400
Des
tru
ctiv
ely
mea
sure
d
dia
met
er (
mm
)
Peak IR intensity at weld centerPeak IR intensity at weld center
Due to the unknown surface condition (i.e. emissivity), peak IR intensity vs. targeted weld size is very noisy.
),( totWfT
Measured IR intensity
Surface emissivity (unknown)
Temperature
6 Managed by UT-Battellefor the Department of Energy
Our New Approach
Special IR image/signal analysis algorithms have been developed – Comparing IR intensity over the entire image pixels of
a single frame and throughout multiple image frames.
– Relying on dynamic change of the IR intensity, instead of the absolute IR intensity
– Insensitive to surface emissivity.
7 Managed by UT-Battellefor the Department of Energy
Conventional IR thermography Non-contact, non-intrusive, whole field image
– Sensitive to low and non-uniform material surface emissivity
– Requiring painting of the weld surface (impractical in auto production line)
– Low signal/noise ratio
– Mostly limited to lab trials for post-weld inspection
– No successful attempts in real-time application
Our new IR-based technology Capable for both real-time and post-weld
inspection
Insensitive to surface condition
No surface treatment is needed
Easy for implementation and automation
Fast and repeatable
Advantages against Conventional IR NDE
8 Managed by UT-Battellefor the Department of Energy
Principles of new IR NDE Technique
Utilize the heat generated during welding
Post-weldReal-time
Aux. localized heating (~50 °C)
Utilize auxiliary heating source
9 Managed by UT-Battellefor the Department of Energy
Major Tasks/Efforts
Weld coupons and auto body weld structures with controlled weld quality/defects attributes– Destructive and ultrasonic NDE examinations
for weld characterizations
Post-weld IR inspection– Increasing the IR signal-to-noise ratio
Better heating techniques Digital image processing Reflection reduction/minimization As-welded surface condition
– Heating and heat transfer modeling to guide inspection and data analysis
Real-time IR Monitoring– Using the heat flow during welding
Development prototype system Beta testing in production environment Tech transfer and commercialization
10 Managed by UT-Battellefor the Department of Energy
Weld Quality Metrics
Ranked by industry advisory committee in the order of importance (high to low)– Weld with no or minimal fusion
– Cold or stuck weld
– Weld nugget size
– Weld expulsion and indentation
– Weld cracks
– Weld porosity
Most critical
Excessive indentation
Stuck weld (insufficient fusion)
Less critical
Cracks
Porosity
11 Managed by UT-Battellefor the Department of Energy
Material Matrix Tested Steel Grades, Coating,
Thickness
• DP590 galvanized 1.2mm• DP590 galvanized 1.2mm
• DP590 galvanized 1.8mm• DP590 galvanized 1.8mm
• DP600 HDG 1.0mm• DP600 HDG 1.0mm
• DP600 bare 1.0mm• DP600 bare 1.0mm
• DP600 HDG 1.0mm• DP600 HDG 2.0mm
• DP600 bare 2.0mm• DP600 HDG 2.0mm
• DP980 cold rolled 1.2mm• DP980 cold rolled 1.2mm
• DP980 cold rolled 1.2mm• DP980 cold rolled 2.0mm
• DP980 cold rolled 2.0mm• DP980 cold rolled 2.0mm
Steel Grades, Coating,Thickness
• Boron bare 1.0mm• Boron bare 2.0mm• Boron bare 1.0mm
• Boron aluminized 1.0mm• Boron aluminized 2.0mm• Boron aluminized 1.0mm
• Boron bare 1.0mm• Boron aluminized 2.0mm• Boron bare 1.0mm
• Boron aluminized 1.0mm• Boron bare 2.0mm• Boron aluminized 1.0mm
Steel Grades, Coating,Thickness
• Boron bare 1.0mm• Boron bare 1.0mm
• Boron aluminized 1.0mm• Boron aluminized 1.0mm
• Boron bare 1.0mm• Boron aluminized 1.0mm
• Boron bare 1.0mm• Boron bare 2.0mm
• Boron aluminized 1.0mm• Boron aluminized 2.0mm
• Boron aluminized 1.0mm• Boron bare 2.0mm
• Boron bare 2.0mm• Boron bare 2.0mm
• Boron aluminized 2.0mm• Boron aluminized 2.0mm
• Boron bare 2.0mm• Boron aluminized 2.0mm
Welds made with different combinations of materials, thickness, stack-up configurations and coatings.
Each combination includes spot welds with varying attributes (i.e., nugget size, indentation, etc.)
12 Managed by UT-Battellefor the Department of Energy
Controlled Welding Investigated
New caps
Worn caps
Electrodes misalignment=~1mm
Welding conditions– Varying welding current
– Varying welding hold time
– New vs. worn electrodes
– Perfectly aligned vs. misaligned electrodes
– Perfectly stacked steel plates vs. separated plates
1mm-thick washers were inserted to introduce gaps
Good weld Gap between plates
13 Managed by UT-Battellefor the Department of Energy
Weld Characterization Methods
Sectioning welds– Nugget size and shape
– Porosity and expulsion
– Surface indentation
Dye penetrants– Surface cracking
Surface micro-profiling– Surface indentation– Surface cracking
Ultrasonic C-Scan (underwater)– Nugget and weld shape– Porosity
14 Managed by UT-Battellefor the Department of Energy
Hardware Requirement
WELDING MACHINE
Post-weldReal-timeI/O
Triggering signal
I/O
Induction heater
IR camera I/O device Computer
IR camera Induction heater I/O device Computer
15 Managed by UT-Battellefor the Department of Energy
Weld Quality Analysis Software
Post-weld
Real-time
Measurable weld attributes– Nugget size and shape– Cold/stick weld defect– Weld thickness/indentation
Measurable weld attributes– Nugget size– Cold/stick weld defect– Expulsion
16 Managed by UT-Battellefor the Department of Energy
Real-time NDE System Operation Demonstration (Movie clip)
17 Managed by UT-Battellefor the Department of Energy
Post-weld NDE System Operation Demonstration (Movie clip)
18 Managed by UT-Battellefor the Department of Energy
Example 1 (3T flat coupons)
0123456789
10
0 5 10
Me
asu
red
dia
met
er (
mm
)
Post-weld predicted diameter (mm)
3.13.23.33.43.53.63.73.83.9
44.1
3.1 3.6 4.1
Me
as
ure
d t
hic
kn
es
s (
mm
)
Post-weld predicted thickness (mm)
• Boron bare 1.0mm• Boron aluminized 2.0mm• Boron bare 1.0mm
• Boron bare 1.0mm• Boron aluminized 2.0mm• Boron bare 1.0mm
0123456789
10
0 5 10
Mea
sure
d d
iam
eter
(m
m)
Real-time predicted diameter (mm)
Post-weld predicted shape
19 Managed by UT-Battellefor the Department of Energy
Example 2 (2T flat coupons)
2.42.52.62.72.82.9
33.13.23.33.4
2.4 2.9 3.4
Me
as
ure
d t
hic
kn
es
s (
mm
)
Post-weld predicted thickness (mm)
2
3
4
5
6
7
8
2 4 6 8
Mea
sure
d d
iam
eter
(m
m)
Post-weld predicted diameter (mm)
2
3
4
5
6
7
8
2 4 6 8
Me
as
ure
d d
iam
ete
r (m
m)
Real-time predicted diameter (mm)
• DP980 cold rolled 1.2mm• DP980 cold rolled 2.0mm• DP980 cold rolled 1.2mm• DP980 cold rolled 2.0mm
Post-weld predicted shape
* Note: In this case, some of the welds are not symmetric due to the misalignment of electrodes during welding.
20 Managed by UT-Battellefor the Department of Energy
no weld
Destructive measurement
Post-weld signature
Example 3 (auto body components)
21 Managed by UT-Battellefor the Department of Energy
Example 4 (Detection of stick welds)
• DP600 HDG 1.2 mm• DP600 HDG 1.2 mm• DP600 HDG 1.2 mm• DP600 HDG 1.2 mm
Normal-size welds (generated by normal welding conditions)
Stick weld (generated by decreasing welding current)
Real-time signature
Post-weld signature
Threshold value related to nugget size
22 Managed by UT-Battellefor the Department of Energy
Real-time signature
Post-weld signature
Example 5 (Detection of abnormal welds due to separation of steel plates)
Normal welding condition
Abnormal welding condition (simulated by inserting washers)
• DP600 bare 2.0mm• DP600 HDG 2.0mm• DP600 bare 2.0mm• DP600 HDG 2.0mm
0.770.64 0.72 0.67
0.31
23 Managed by UT-Battellefor the Department of Energy
More Examples
Results from a combination of
– Materials
– Plate thickness
– Stack up configurations (2T/3T)
– Surface coatings
IR
IR
Post-weld predicted shape
0123456789
10
0 5 10
Me
asu
red
dia
met
er (
mm
)
Real-time predicted diameter (mm)
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0.5 1.5 2.5 3.5 4.5
Me
asu
red
th
ickn
ess
(m
m)
Post-weld predicted thickness (mm)
0123456789
10
0 5 10
Me
asu
red
dia
met
er (
mm
)
Post-weld predicted diameter (mm)
24 Managed by UT-Battellefor the Department of Energy
Influence of Coil/Weld Alignment (Post-weld)
Weld center
offset
Aux. heater center
• Experiments were conducted on the same weld.
• Inspection at each offset location was repeated twice.
Predictions were fairly repeatable.1mm coil/weld alignment offset resulted in about 0.2mm
error for predictions of both nugget size and weld thickness
0
2
4
6
8
10
0 2 4 6
Pre
dic
ted
we
ld
dia
met
er (
mm
)
Offest from center (mm)
Trial 1Trial 2
0
0.5
1
1.5
2
2.5
3
3.5
0 2 4 6
Pre
dic
ted
th
ick
ne
ss
(m
m)
Offest from center (mm)
Trial 1Trial 2
25 Managed by UT-Battellefor the Department of Energy
Summary
Successfully developed an IR-based spot weld NDE inspection prototype system capable for both real-time and post-weld on-line applications.
Reliable detection of weld size, cold weld, and surface indents with sufficient accuracy for various combination of materials, thickness, stack-up configuration, surface coating conditions and welding conditions.
Application Measurable weld attributes Inspection time
Real-time• Nugget size and weld shape• Cold/stick weld defects• Expulsion
1.5~3s
Post-weld• Nugget size• Cold/stick weld defects• Weld thickness/indentation
~3s