View
216
Download
2
Category
Preview:
Citation preview
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Professor: Dr. Marcelo Dapino
Acknowledgements:
• NSF I/UCRC Program, SVC IAB
• NSF Graduate Research Fellowship Program
• Ohio Third Frontier Wright Project
Process-Property Relationships in
Ultrasonic Additive Manufacturing of
Lightweight Structures
Adam Hehr
hehr.7@osu.edu
Department of Mechanical and Aerospace Engineering
The Ohio State University
Eighth Annual Meeting, Honda R&D, 20-21 Aug 20152
Talk Overview
• Ultrasonic Additive Manufacturing- UAM
• Bond Formation in UAM
• Welder Energy and Parameters
• Effect of Build Compliance on Effective Weld Power
• Power Compensation
• UAM Process Modeling
• Concluding Remarks
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Ultrasonic Additive Manufacturing - UAM• Recent technology that combines:
Ultrasonic metal welding
Additive manufacturing
CNC machining center
• Low temperature process:
Interface near ½ Tmelt
Measureable temp near 100 oC for
Al 6061-H18 [1]
• Enabling technology for dissimilar
metal joining and low temperature
applications
[1] M. Sriraman, M. Gonser, H. Fujii, S. Babu and M. Bloss, "Thermal transients during processing of materials by very high power ultrasonic additive
manufacturing," Journal of Materials Processing Technology, vol. 211, p. 1650– 1657, 2011. 3
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
UAM Video
4
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
NiTi Al
Smart Materials
Integration
Al
Fiber Optic Cable
PVDF in Al
Embedded Temperature Sensitive Sensors
and Electronics
Al
Carbon
Fiber
Al
Ti
Light-Weighting with Dissimilar
Materials and Metals
Alum
inumFib
er
Op
tic
s
SS tube
Cu
Al
Complex Internal
Cooling
Solid-State ActuationMulti-Material Joints and ReinforcementEfficient Cooling and Embedded Sensing
Flu
id f
low
pump
heater
cooling
thermo-
couple
Automotive Aerospace
http://jwst.nasa.gov/www.dolphin.fr
UAM Applications and Strengths
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Bond Formation in UAM
• Recent advancements in delivered power and down force have remedied interface
voids, i.e. gapless structures [2]
• Weld microstructure is composed of recrystallized small (~1 micron) equiaxed
grains within narrow region of weld interface (~ 20 micron) [3-4]
• The degree of recrystallization is foretelling of the shear strain at the interface due
to dynamic recrystallization being a function of (i) strain and (iii) temperature [5]
6
Bulk Tape
Bulk Tape
Mixing
New
Grains
New
Baseplate
Grains
Baseplate
Original Baseplate
Grain Boundary
Bulk Tape
New Tape
Grains Mixing
New Baseplate
Grains
Welded Layers
Baseplate
[2] K. Graff, M. Short and M. Norfolk, "Very High Power Ultrasonic Additive Manufacturing (VHP UAM)," in International Solid Freeform Fabrication Symposium, Austin, 2011.[3] K. Johnson, "PhD Thesis: Interlaminar Subgrain Refinement in Ultrasonic Consolidation," Loughborough University, 2008.[4] R. Dehoff and S. Babu, "Characterization of interfacial microstructures in 3003 aluminum alloy blocks fabricated by ultrasonic additive manufacturing," Acta Materialia, vol. 58, pp. 4305-4315, 2010. [5] F. Humphreys and M. Hatherly, Recrystallization and Related Annealing Phenomena, Elsevier, 2004.
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Welder Energy and UAM Parameters
• Controllable Parameters:
Vibration amplitude, δ (μm, % of max)
Normal force (N)
Sonotrode travel speed, Vt (in/min)
Baseplate temperature (°F)
• Fixed Parameters:
Vibration frequency (~20 kHz)
Sonotrode roughness (~7 or 14 micron Ra ) and material (tool steel)
Tape thickness (~0.006”)
• Typical weld power for Al 6061-H18 is near 2-3 kW
• Typical weld power for as rolled 4130 steel is near 5-7 kW
7
𝐸𝑤𝑒𝑙𝑑 = 𝑃 ∙ 𝑑𝑡 =1
𝑉𝑡 𝐹 ∙ 𝜔 ∙ 𝛿 ∙ 𝑑𝑥
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Effect of Build Compliance on Effective Weld Power
• Build compliance refers to the mechanical deformation of the part when subjected
to shear force during the UAM process – compliance increases with part height [6]
• Build compliance leads to a decrease in plastic deformation, i.e., a decrease in
effective weld power
• Power compensation achieved by increasing weld amplitude manually
8
0 5 10 15 20 2560
70
80
90
100
110
120
No. of Welded LayersPe
rce
nt A
ve
rag
e E
lectr
ic P
ow
er
(W/W
)
UncompensatedCompensated
Fs
UAM Stack
with keff
stiffness
Baseplate
New Layer
Imparted amplitude
Amplitude lost to
compliance
25% power loss
Amplitude available
to weld foil
[6] A. Hehr, P. Wolcott and M. Dapino, "Effect of Weld Power and Build Compliance on Ultrasonic Consolidation," Rapid Prototyping Journal, In Press.
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Power Compensation
9
• Does power compensation lead to stronger welds?
• Approach:
Measured weld power for compensated and uncompensated stack builds
Push-pin testing to evaluate bond strength
Focused Ion Beam (FIB) imaging used to analyze interface
microstructure
0 5 10 15 20 25 30
400
600
800
1000
1200
1400
1600
Uncompensated Power
Weld Length (cm)
We
ld P
ow
er
(W)
Layer 2
Layer 5
Layer 10Layer 15
Layer 20
0 5 10 15 20 25 30
400
600
800
1000
1200
1400
1600
Compensated Power
Weld Length (cm)
We
ld P
ow
er
(W)
Layer 2
Layer 5
Layer 10
Layer 15
Layer 20
0 5 10 15 20 25800
900
1000
1100
1200
1300
1400
1500No. Layers vs. Steady-State Avg. Power
Number of Welded Layers
Ave
rag
e P
ow
er
(W)
Uncomp.Compensated
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Power Compensation: Push-Pin Testing
• Comparative test
• Used to evaluate UAM
interfacial bond strength
• Successfully used in
recent Al 6061-H18 weld
study [7].
10
F
Base plate
Notch
Width
3-4 Layers
Thick
Base Plate
Plus 5-6
Layers Deep
20 Total
Layers
UAM Build
“Good”“Poor”
0 0.5 1 1.5 2 2.5 3 3.50
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Stroke [mm]
Fo
rce [kN
]
Energy Compensated
Uncompensated
Similar Peak Load
Enhanced Mechanical Work
Interface
Delamination
Delaminationf
Failure Thru All
the Layersf
CompensatedUncompensated
[7] P. Wolcott, A. Hehr, M.J. Dapino, “Optimized welding parameters for Al 6061 ultrasonic
additive manufactured structures,” Journal of Materials Research 29 (17).
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Power Compensation: Weld Microstructure
11
Uncompensated: Layer 15
Unmixed zone
Uncompensated: Layer 5 Compensated: Layer 5
Compensated: Layer 15
• Enhanced interface recrystallization with power compensation
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Power Compensation: Energy Balance
• Energy balance can be utilized to analyze bonding process
• Eplastic measured remotely via transducer power consumption (Eweld)
• Erecryst measured via quantity of new small grains at interface [8]
• Stronger welds achieved with larger Erecryst due to Hall-Petch
relationship [8]
• First time weld microstructure has been correlated with energy
input
12
𝐸𝑠𝑢𝑟𝑓 + 𝐸𝑝𝑙𝑎𝑠𝑡𝑖𝑐 = 𝐸𝑏𝑢𝑙𝑘 + 𝐸𝑟𝑒𝑐𝑟𝑦𝑠𝑡 + 𝐸𝑡ℎ𝑒𝑟𝑚𝑎𝑙Input Output
[8] R. Abbaschian, L. Abbaschian and R. E. Reed-Hill, Physical Metallurgy Principles, Stamford: Cengage Learning, 2010.
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
UAM Process Modeling: LTI Model
13
Welding
Assembly𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐 𝑐𝑢𝑟𝑟𝑒𝑛𝑡: 𝑖𝑠ℎ𝑒𝑎𝑟 𝑓𝑜𝑟𝑐𝑒: 𝐹𝑠
𝑉𝑜𝑙𝑡𝑎𝑔𝑒: 𝑉𝑤𝑒𝑙𝑑 𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦: 𝑗𝜔𝛿
𝑰𝒏𝒑𝒖𝒕𝒔 𝑶𝒖𝒕𝒑𝒖𝒕𝒔𝑳𝒊𝒏𝒆𝒂𝒓 𝑺𝒚𝒔𝒕𝒆𝒎
𝑗𝜔𝛿𝑉
=𝐻𝑒𝑚∗ 𝐻𝑒
∗
𝐻𝑚∗ 𝐻𝑚𝑒
∗𝑖𝐹𝑠
𝑳𝑻𝑰 𝑴𝒐𝒅𝒆𝒍
𝛿
Fs
Piezo
Compression Rod
Actuator 1
V
Actuator 2
V
Sonotrode
𝑖 𝑖
V
Ct
Mt1:Φt
𝛿Φt 𝛿𝑖
Φt 𝑉
1/KtDt
Fs
Lumped System: Welder Characterization
V
1/Mt
Ψt:1
𝛿𝑖 Ψt 𝛿
Kt 1/Dt
FsΨt𝑖
Lumped System: Welder Operation
• Electroacoustics theory [9-12]
[9] B. Richter, J. Twiefel, J. Wallaschek, Energy Harvesting Technologies: Ch4: Piezolectric Equivalent Circuit Models, Springer, 2009.[10] W. P. Mason, Electromechanical Transducers and Wave Filters, D. Van Nostrand Company, Inc., 1942[11] F. Hunt, Electroacoustics: The Analysis of Transduction and Its Historical Backaground, Cambridge: Harvard University Press, 1954.[12 C. van der Burgt, H. Pijls, Motional Positive Feedback Systems for Ultrasonic Power Generators, IEEE Transactions on Ultrasonics Engineering 10 (1)
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
UAM Process Modeling: Characterization
• Shear force: High frequency modal hammer
• Weld velocity: Laser vibrometer
• Voltage and current: Linear amplifier
• Frequency response functions: Quattro analyzer
• Boundary conditions: In UAM machine
14
Welding
Assembly
Laser Vibrometer
PowerAmplifier
Charge
Amplifier
Quattro
Analyzer
Analysis
Computer
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
UAM Process Modeling: Model Fit
• Good agreement with measured and closed form FRFs
• Fit procedure found in literature [9]
15
19.6 19.7 19.8 19.9-38
-36
-34
-32
-30
-28
-26
Frequency (kHz)
He d
B M
agnitude (
A/V
)
Exp.
Fit
19.6 19.7 19.8 19.9-70
-60
-50
-40
-30
Frequency (kHz)
Hem
dB
Magnitude (
m/s
/V)
19.6 19.7 19.8 19.9-80
-70
-60
-50
-40
-30
Frequency (kHz)
Hm
dB
Magnitude (
m/s
/N)
19.6 19.7 19.8 19.920
40
60
80
100
120
Frequency (kHz)
He P
hase A
ngle
(deg)
19.6 19.7 19.8 19.9-100
-50
0
50
100
Frequency (kHz)
Hem
Phase A
ngle
(deg)
19.6 19.7 19.8 19.9-150
-100
-50
0
50
100
Frequency (kHz)
Hm
Phase A
ngle
(deg)
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
UAM Process Modeling: In-Situ Measurements
LTI model assumption valid?
• Yes, welder dynamics are pseudo-stable during operation
• Measurements support model assumptions
16
0 0.5 1
19.7
19.75
19.8
19.85
Weld
er
Fre
quency (
kH
z)
Time (s)
0 0.5 11.7
1.75
1.8
1.85
1.9
1.95
2
Peak V
elo
city (
m/s
)
Time (s)
No Welding
Welding
Peak Weld Velocity Welder Frequency
0 0.5 10
200
400
600
800
1000
Avera
ge E
lectr
ic P
ow
er
(W)
Time (s)
Electric Power Draw
40 50 60 70 801
1.5
2
2.5
3
Pe
ak V
elo
city (
m/s
)
Amplitude Level (%)40 50 60 70 80
0
100
200
300
400
Ave
rag
e E
lectr
ic P
ow
er
(W)
40 50 60 70 80600
700
800
900
1000
1100
1200
1300
1400
Amplitude Level (%)
Pe
ak V
olta
ge
(V
)
(a)
40 45 50 55 60 65 70 75 8050
100
150
200
250
300
350
Amplitude Level (%)
Ave
rag
e E
lectr
ic P
ow
er
(W)
Power Calibration
Measured
Estimate
Adj. Estimate
Welder VoltageWelder Velocity and Power
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
UAM Process Modeling: Shear Force and Efficiency
• Shear force near 2000 N, which is similar to de Vries [13].
• Welder efficiency near 82%, which is near ultrasonic metal welding
estimates [14] and below piezoelectric transducer efficiency [15].
• Upward frequency shift from UAM build stiffening system.
17
0 0.5 119.75
19.8
19.85Excitation Frequency
Weld
er
Fre
quency (
kH
z)
Time (s)
0 0.5 11.6
1.7
1.8
1.9
2Peak Weld Velocity
Peak V
elo
city (
m/s
)
Time (s)
0 0.5 1
0
500
1000
1500
2000Peak Shear Force
Peak F
orc
e (
N)
Time (s)
0 0.5 10
500
1000
1500
2000Weld Power
Avera
ge E
lectr
ic P
ow
er
(W)
Time (s)
No Welding
Welding
𝑖
𝑖𝑟𝑒𝑓=
𝑃
𝑃𝑟𝑒𝑓
𝐹𝑠 =𝐻𝑒𝑚∗ 𝑖 − 𝑗𝜔𝛿
𝐻𝑚∗
𝑒 =𝑗𝜔𝛿 ∗ 𝐹𝑠
𝑃
∆𝑃 =1
2
V
Ψ𝑡Δ𝐹𝑠
𝑺𝒉𝒆𝒂𝒓 𝑭𝒐𝒓𝒄𝒆 𝑬𝒔𝒕𝒊𝒎𝒂𝒕𝒊𝒐𝒏
𝑬𝒇𝒇𝒊𝒄𝒊𝒆𝒏𝒄𝒚
𝑷𝒐𝒘𝒆𝒓 𝑪𝒐𝒏𝒕𝒓𝒐𝒍 𝑳𝒂𝒘
[13] E. de. Vries, Mechanics and Mechanisms of Ultrasonic Metal Welding, Ph.D. thesis, The Ohio State University, Columbus, OH, USA (2004).[14] K. Graff, AWS Handbook 9th Edition: Volume3: Ultrasonic Welding of Metals, American Welding Society, 2001.[15] R. Friel, Power Ultrasonics: Ch13: Power ultrasonics for additive manufacturing and consolidation of materials, Elsevier, 2015.
Eighth Annual Meeting, Honda R&D, 20-21 Aug 2015
Conclusion
• UAM enables fabrication of unique materials and products
• Weld power/energy correlation with bond quality
• Weld power as an in-situ process variable
• LTI model which uses shear force as a system input
• In-situ measurements of welder
18
Recommended