Lu Paper 201752 December 2017Layout.

qxp_Layout 1 11/7/17 3:40 PM Page 439

WELDING RESEARCH
SUPPLEMENT TO THE WELDING JOURNAL, DECEMBER 2017
Sponsored by the American Welding Society and the Welding Research Council

Repairing Ultrasonic Welded

Carbon Fiber­Reinforced Nylon 66 Composite
Application of a second ultrasonic vibration not only restored
joint strength but also decreased the scatter in weld strength

BY L. LU, Q. ZHI, Y.-H. GAO, Z.-X. LIU, AND P.-C. WANG

in lightweight vehicle structures (Refs.

ABSTRACT 18–23). Despite these advantages, the
use of ultrasonic welding in vehicle
In this study, a method for repairing partially welded carbon fiber­reinforced nylon 66
structural applications has been limit-
composite joints with ultrasonic welding was developed. The method included position­
ing the ultrasonic horn and anvil proximate the weld, and applying an ultrasonic vibration ed, partly because of concerns regard-
at substantially the same location on the existing welds fabricated by introducing a sepa­ ing weld quality inspection and weld
ration of 0.7 mm (0.03 in.) between 4­mm­ (0.16­in.­) thick carbon fiber­reinforced nylon repair.
66 workpieces. Welding tests were performed and the welded joints were quasi­static It has been shown that alignment
tested and analyzed. It was found that the separation between the workpieces and leveling of a welding fixture, a
influenced the contact between them, and consequently resulted in variations in weld welding horn, and the workpieces dur-
area and weld strength. The welds were repaired by applying the second ultrasonic vibra­ ing ultrasonic welding are critical for
tion with the proper energy intensity under the same horn pressure. The application of a formation of an ultrasonic weld (Ref.
second ultrasonic vibration not only restored the joint strength but also decreased the 24). When the warped workpiece is
scatter in strength of the welds. Microstructure analyses of the repaired welds revealed
positioned near a second workpiece
the main mechanism for repairing the ultrasonic welds was the increase in weld area
caused by melting of the materials at the periphery of the existing weld between the for joining, a separation is present be-
workpieces. The application of the repair method would reduce the potential need to tween the workpieces. The space be-
scrap the particular component assembly. tween the workpieces can lead to im-
proper weld formation. Therefore,
there is an urgent need to develop a
KEYWORDS method to inspect the weld quality
and repair the partially joined (or
• Carbon Fiber­Reinforced Nylon 66 Composite • Ultrasonic Weld • Flawed Joint flawed) welds in ultrasonic welding of
• Repair • Repair Mechanism
thermoplastic composites.
Weld quality can be confirmed off-
Introduction ing the composite parts (Refs. 6–11), line from the production operation by
there is a need to understand the join- periodically removing the workpieces
Ever-increasing fuel economy, envi- ing of carbon fiber thermoplastic com- from the operation and physically ex-
ronmental, and mechanical perform- posite for automotive applications. amining and testing a number of weld
ance requirements in the automotive Various joining techniques such as nuggets. Nevertheless, it would be
industry make the high strength-to- adhesive bonding, mechanical fasten- preferable to have an on-line process
weight ratio associated with carbon ing, and fusion welding are available that could be performed during the
fiber/nylon composite attractive (Refs. for joining thermoplastics (Refs. 12, welding operation to assess welding
1–4). They offer greater strength and 13). Among all of these methods, ul- conditions and weld nugget quality
better structural crashworthiness trasonic welding is particularly prom- and size. Many studies (Refs. 25–28)
when compared with other lightweight ising (Refs. 14–17). Ultrasonic welded have reported that the weld quality
materials (e.g., aluminum and magne- thermoplastic composite is known to could be reflected by various signals
sium) (Ref. 5). While a great deal of ef- have good stiffness and strength (i.e., ultrasonic wave and process sig-
fort has been focused on developing (Refs. 14, 15), providing a potentially natures, power, meltdown, etc.).
the carbon fiber composites and mold- wide range of applications, especially Among these techniques, an in-situ

DECEMBER 2017 / WELDING JOURNAL 439-s

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 440

WELDING RESEARCH

monitoring method through power

and displacement data was particular- A
ly promising (Ref. 26). Based on this
concept, an on-line method of moni-
toring the target horn displacement
and welding time of stable phase of
melt film has been developed for weld
quality inspection (Ref. 29). The pur-
pose of the present study was to devel-
op a method for repairing the flawed
joints. We first conducted tests to
study the correlation between the
weld area and strength of the welded
4-mm- (0.16-in.-) thick carbon fiber-
reinforced nylon 66 composite. Then,
the partially welded joints were fabri-
cated by introducing a separation of
0.7 mm (0.03 in.) between two work-
pieces. A second ultrasonic vibration
was applied to repair the welds and the
repair mechanism was examined. Fi-
nally, the criterion used for the devel-
opment for repairing the ultrasonic B
welded carbon fiber thermoplastic
composite and significance of the re-
pair method for vehicle structural as-
sembly were discussed.

Experimental Procedure
Materials
Commercial nylon 66 and carbon
fiber (24K, T300 type, Toray Carbon
Magic Co. Ltd.) with a length of 2 mm
(0.08 in.) were selected for this study.
Per the manufacturer’s data sheet, the
material properties of these materials
are given in Table 1.

Injection Molding Carbon Fiber­

Reinforced Nylon 66 Composite
Carbon fibers were first cleaned
with a concentrated solution of nitric
acid and then surface pretreated using
8% diglycidyl ether of bisphenol solu-
tion in acetone. Both nylon 66 and
pretreated carbon fibers were dried at
80°C (176°F) under vacuum condition
for 3 h before being used to fabricate
Fig. 1 — A — Schematic; B — photo of the specific equipment and setup for ultrasonic
welding of injection­molded carbon fiber­reinforced nylon 66 composite.
the carbon fiber-reinforced nylon 66 hopper while carbon fibers were added
composite. to the second hopper. Nylon 66 was
A twin-screw extruder with two fully melted before the carbon fibers
separate inlets was used to mold car- were added to minimize the carbon
bon fiber-reinforced nylon 66 compos- fiber from fracturing during com-
ite. Nylon 66 was added to the first pounding. Carbon fiber-reinforced ny-

Table 1 — Material Properties of Nylon 66 and Carbon Fiber

Specific Gravity Tensile Strength Elastic Modulus

(g/cm3) (lb/in.3) (MPa) (lb/in.2) (GPa) (lb/in.2)

Nylon 66 1.14 4.1 × 10–2 80 1.16 × 104 3 4.35 × 105

Carbon Fiber 1.76 6.4 × 10–2 3454 5 × 105 231 3.4 × 107

440-s WELDING JOURNAL / DECEMBER 2017, VOL. 96

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 441

WELDING RESEARCH

lon 66 composite with 30% mass car-

bon fiber was fabricated. The process-
ing temperature was 270~280°C
(518~536°F) and the screw speed was
180 rpm. After fully mixing nylon 66
with carbon fibers in a twin-screw ex-
truder, carbon fiber-reinforced nylon
66 composite was processed into the
pellets with a length of 2 mm (i.e.,
0.08 in.). The pellet was then fed into
the injection extruder to mold into the
coupons with the dimensions of 132
mm × 38 mm × 4.0 mm (i.e., 5.2 in. ×
1.5 in. × 0.16 in.). The mechanical
Fig. 2 — Schematic of a single lap­shear specimen. properties of the injection molded car-
bon fiber-reinforced nylon 66 compos-
ite were measured and the results are
A shown in Table 2.

Ultrasonic Welding Process

The ultrasonic welding process was
performed using a KZH-2026 multi-
function UW machine (Ref. 30) with a
nominal power of 2.6 kW and a nominal
frequency of 20 kHz. The output ampli-
tude of the transducer used in this
study was 25 m (9.8 × 10–4 in.), which
B was designed to be fixed and cannot be
adjusted during the welding process.
The gain ratios of the booster and the
horn were 1:1.2 and 1:1.5, respectively.
The amplitude of the assembled stack
was obtained by using a displacement
sensor that was integrated in the weld-
ing machine.
The welding setup used in this study
is schematically shown in Fig. 1. The
C piezoelectric converter converted the
electrical signal into mechanical vibra-
tions. To transfer the ultrasonic waves
to the workpieces, the transducer was
connected to the horn, which was
placed perpendicularly in contact with
the workpieces to be welded. The sup-
port frame of the transducer-booster-
horn system was attached to a pneu-
matic piston that provided vertical
Fig. 3A — Schematic for the mimic of 0.7 mm (0.03 in.) separation at the faying inter­ movement along with the static force
face; B — a discrepant weld; C — repairing the ultrasonic welded carbon fiber­reinforced (i.e., horn pressure) applied through the
nylon 66 composite. horn to the workpieces. The machine
was also equipped with a data acquisi-

Table 2 — Mechanical Properties of Injection­Molded 4­mm­ (i.e., 0.16­in.­) Thick Carbon Fiber­Reinforced Nylon 66 Composite Coupons
with 30% Mass Carbon Fiber

Tensile Strength Elastic Modulus

(MPa) (lb/in.2) (MPa) (lb/in.2)

99.2 ± 3 (1.4 ± 0.4) × 104 8936 ± 421 (1.3 ± 0.6) × 106

DECEMBER 2017 / WELDING JOURNAL 441-s

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 442
WELDING RESEARCH
Fig. 4 — Schematic of sample preparation for examining the
microstructure of the ultrasonic welded 4­mm­ (0.16­in.­)
thick injection molded carbon fiber­reinforced nylon 66 com­
posite as fabricated.
tion system, which combined with the
pressure sensor, horn-displacement
sensor, and timer, were integrated in
the controller of the UW machine. The
horn pressure, weld energy, and dis-
placement of the horn were recorded
online in a personal computer as a
function of time by the data acquisi-
tion system. The final horn displace-
ment, weld energy, welding time, horn
pressure, hold time, and delay time
were also displayed in the control panel
during the ultrasonic welding process.
To avoid the coupons’ motion during
welding, the coupons were fixed with a
fixture as shown in Fig. 1.
The machine includes three welding
modes, energy, time, and collapse mode.
The value of the weld energy for the en-
ergy mode, the welding time for the
time mode, and the horn displacement
for the horn-displacement mode are
preset prior to welding to control the
welding process. The workpieces were
then welded using nominal power of the
machine. When the weld energy, weld
time, or horn-displacement reached the
preset values for the selected weld
mode, the ultrasonic wave oscillation
was stopped. Therefore, the weld quali-
ty was controlled by the preset values in
each selected welding mode.
Sample Fabrication
The single lap-shear specimen con-
442-s WELDING JOURNAL / DECEMBER 2017, VOL. 96
Fig. 5 — Effects of welding time and horn pressure on the strength
of ultrasonic welded 4­mm­ (0.16­in.­) thick carbon fiber­reinforced
nylon 66 composite with 30% mass carbon fiber.
figuration shown
in Fig. 2 was se-
lected in this
study. There was
no energy direc-
tor introduced
on the work-
pieces. All speci-
mens were weld-
ed in energy
mode using a
KZH-2026 multi-
function ultra-
sonic welding Fig. 6 — Correlation between the joint strength and weld area of ul­
machine (Ref. trasonic welded 4­mm­ (0.16­in.­) thick lapped carbon fiber­reinforced
30) with a nomi- nylon 66 composite with 30% mass carbon fibers made with a welding
nal power of 2.6 time of 2.1 s and a horn pressure of 0.15 MPa (21.76 lb/in.2).
kW and a fre-
quency of 20 kHz. The values of the ul- the faying surface prior to welding —
trasonic energy, horn pressure, and hold Fig. 3A. Ultrasonic welds, shown in
time were preset prior to welding and Fig. 3B, were fabricated with a weld
then ultrasonic welding was performed time of 2.1 s, a horn pressure of 0.15
using nominal power. When the weld MPa (21.76 lb/in.2) and a hold time of
time or weld energy approached the 3.0 s. To repair the undersized or
preset values, the ultrasonic wave oscil- flawed welds, a second vibration with
lation was stopped and the coupons various energy intensities were ap-
were held for 3.0 s to solidify the molten plied to the existing flawed welds us-
material. All specimens were welded us- ing the same horn pressure [0.15 MPa
ing 7075 aluminum horn with a diame- (21.76 lb/in.2)], referring to Fig. 3C.
ter of 18 mm (0.71 in.).
Quasi­Static Test
Repairing Ultrasonic Welds
Quasi-static tests were performed by
To mimic the tolerance resulting loading each specimen to failure in an
from part distortion, a separation of MTS 810 tensile tester according to
0.7 mm (0.03 in.) was introduced at ASTM D1002-2001. To minimize the
Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 443

WELDING RESEARCH

Fig. 7 — Effect of separation between the workpieces on the

strength of the ultrasonic welded 4­mm­ (0.16­in.­) thick lapped
carbon fiber­reinforced nylon 66 composite with 30% mass carbon Fig. 8 — Effect of repair energy on the strength of the ultrasonic
fiber made at a welding time of 2.1 s and a horn pressure of 0.15 welded 4­mm­ (0.16­in.­) thick carbon fiber­reinforced nylon 66
MPa (21.76 lb/in.2). composite with 30% mass fiber.

A B C D

Fig. 9 — Failure modes of repaired ultrasonic welded joints: A — Schematic of a loaded joint; B — interfacial failure mode for a repaired
partially bonded weld; C — failure through the upper workpiece; D — failure through the lower workpiece for the nominal weld.

bending stresses inherent in the testing
of single-lap weld specimens, filler
plates shown in Fig. 2 were attached to
both ends of the specimen using mask-
ing tape to accommodate the sample
offset. Load vs. displacement results
were obtained, as the specimens were
loaded at a stroke rate of 2.0 mm/min
(0.08 in./min). The strength of the joint
was evaluated by the ratio of the peak
load to the overlap area of the weld [950
mm2 (1.47 in.2)]. Three replicates were
performed, and the average joint
strengths were reported.
Examinations of Weld
Microstructure
Because carbon fiber-reinforced ny-
lon 66 composite has poor transmis-
sion, it is difficult to examine the weld
microstructure using polarization mi-
croscopy (Ref. 31). To assess the char-
acteristics of the weld microstructure
of the ultrasonic welded joints, the
specimens were prepared following the
illustrations shown in Fig. 4. As
shown, the joints were notched from
the central position of the weld. Then,

DECEMBER 2017 / WELDING JOURNAL 443-s

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 444

WELDING RESEARCH

A B

C D

Fig. 10 — Appearances of the welded 4­mm­ (0.16­in.­) thick

E carbon fiber­reinforced nylon 66 composite with a separation of
0.7 mm (0.03 in.) between the workpieces before and after re­
pair with the repair energy of the following: A — 2 kJ; B — 3 kJ;
C — 4 kJ; D — 5 kJ; E — 6 kJ.

Results and the joint strength increased with in-

the prenotched specimens were im-
mersed in liquid nitrogen for 10 min
to embrittle the nylon 66. Finally, the
embrittled specimens were broken off
from the notched site. To characterize
the weld microstructure, the broken
specimens were sputter-coated with
gold for 50 s and then examined with a
scanning electron microscope
(JSM6700F).
Discussion
Ultrasonic Weld­
ing Carbon Fiber
Composite
To develop the
welding process
window, 4-mm-
(0.16-in.-) thick carbon fiber-rein-
forced nylon 66 composite with 30%
mass fiber without energy director
were welded using various combina-
tions of welding time and horn pres-
sure. The welded joints were quasi-
static tested. Figure 5 shows the ef-
fects of welding time and horn pres-
sure on the joint strength. As shown,
creasing the welding time and horn
pressure. The results indicated that in-
creasing the welding time and horn
pressure resulted in an increase in
welding energy. However, excessive
welding time and horn pressure did
not always increase the joint strength.
When the welding time exceeded 2.1 s
and the horn pressure was above 0.15
MPa (21.76 lb/in.2), the strength of
the joints decreased slightly. Referring
to Fig. 5, although significant scatter
in joint strength was observed, there
appeared an optimum welding combi-
nation (i.e., a weld time of 2.1 s and a
horn pressure of 0.15 MPa (21.76
lb/in.2) to obtain the optimal joint
strength.
To understand the cause of signifi-
cant scatter in joint strength, the frac-

444-s WELDING JOURNAL / DECEMBER 2017, VOL. 96

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 445

WELDING RESEARCH

A B

E
Fig. 11 — A — Schematic of weld growth; B — weld area of a
discrepant weld, effect of repair energy on the weld area of the
ultrasonic weld repaired with an energy of the following: C — 3
kJ; D — 4 kJ; E — 5 kJ; F — 6 kJ.

ture surfaces of the ultrasonic welded workpieces. This

joints were visually examined and the
results revealed the variation in weld
area at the faying surface was likely
the culprit. To validate this observa-
tion further, the weld areas for all ul-
trasonic welded joints made with a
welding time of 2.1 s and a horn pres-
sure of 0.15 MPa (21.76 lb/in.2) were
measured. Figure 6 shows the correla-
tion between the joint strength and
weld area. Referring to Fig. 6, the vari-
ation in weld area in the range of 180
to 400 mm2 (0.28 to 0.62 in.2) resulted
in a fluctuation of the joint strength
from 5.05 to 6 MPa (732.4 to 870.2
lb/in.2). By examining the tested speci-
mens, it was noted that the scatter in
the weld area likely resulted from the
variations in the material properties
and inconsistent surface contacts be-
tween the workpieces during the ultra-
sonic welding process.
In an attempt to check the dimen-
sional quality of the workpieces, the
dimensions and surface quality of
the injection molded carbon fiber-
reinforced nylon 66 composite
coupons were examined. The results
revealed that beside the part distor-
tion, there were some variations in
thickness and surface roughness of the
distortion could af-
fect the surface
contacts between
the horn/anvil-to-
workpiece, and F
consequently pro-
duced the separa-
tion between the
workpieces. This
could affect the
heat generation
and lead to the
variation in weld
area. To validate
this speculation, expected the scatter in the weld quali-
the workpieces with the separations in ty was inevitable. Therefore, there is
the range of 0.5 to 2.0 mm (0.02 to an urgent need to develop a method to
0.08 in.) at the faying surface, refer- inspect the weld quality and repair the
ring to Fig. 3A, were welded, and the welds in ultrasonic welding of thermo-
results of quasi-static tests are shown plastic composites.
in Fig. 7. As shown, the separation be-
tween the workpieces had a significant Repairing Ultrasonic Welded
influence on the strength of the weld-
ed joints. Examining the fractured Carbon Fiber­Reinforced Nylon
specimens revealed the separation be- 66 Composite
tween the workpieces reduced the
weld area, and thus resulted in lower A method for inspection of weld
joint strength degradation. Since the quality using horn displacement and
variations in the material property duration at the stable phase of melt
and part dimension for the injection- film during the ultrasonic welding
molded composites are common, it is process was developed in a previous

DECEMBER 2017 / WELDING JOURNAL 445-s

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 446

WELDING RESEARCH

Fig. 13 — A — Schematic of an undersized weld; B — cross section of

Fig. 12 — Effect of repair energy on the weld area in ultrasonic the discrepant joint and regions examined by SEM for the repaired ul­
welded 4­mm­ (0.16­in.­) thick carbon fiber­reinforced nylon trasonic welded 4­mm­ (0.16­in.­) thick carbon fiber­reinforced nylon
66 composite with 30% mass fiber. 66 composite.

A B

D
C

Fig. 14 — Effect of repair energy on the microstructures at region 1 of the repaired ultrasonic welded carbon fiber­reinforced nylon 66
composite with 0.7 mm (0.03 in.) separation between the workpieces: A — 3 kJ; B — 4 kJ; C — 5 kJ; D — 6 kJ.

study (Ref. 29). Once the discrepant
(or flawed) weld was identified, the de-
gree of the undersized welds could be
estimated based on the measurements
on the production line. Then, repair-
ing the discrepant or flawed welds
could be made on or off the produc-
tion line.
In this study, a second ultrasonic vi-
bration was applied to repair the dis-
crepant welds that had a separation of
0.7 mm (0.03 in.) between the work-
pieces. The ultrasonic horn and anvil
were positioned to proximate the
weld, and an ultrasonic vibration was
applied at substantially the same loca-
tion on the existing weld using the
same weld horn and horn pressure.
Repaired ultrasonic welds were then
quasi-static tested, and the results are
shown in Fig. 8. For the purpose of
comparison, the strengths of the nom-
inal welds are also included in Fig. 8.
As shown, little improvement in joint
strength was observed when the repair
energy was less than 4.0 kJ. As the re-
pair energy exceeded 5.0 kJ, it not
only increased the joint strength but
also reduced the scatter in joint
strength. These results suggest that by
applying the proper weld energy, it is
feasible to repair the discrepant ultra-
sonic welded joints with the second ul-
trasonic pulse.
Referring to Fig. 8, the increase in
joint strength with a repair energy of 5
kJ implied that the second ultrasonic vi-
bration likely resulted in an increase in
weld area by melting of the workpieces

446-s WELDING JOURNAL / DECEMBER 2017, VOL. 96

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 447

WELDING RESEARCH

A B

Fig. 15 — Effect of repair energy on the porous zone at region 2 of

the repaired ultrasonic welded carbon fiber­reinforced nylon 66
composite with 0.7­mm (0.03­in.) separation between workpieces:
A — 3 kJ; B — 4 kJ; C — 5 kJ; D — 6 kJ.

at the faying surface (Refs. 32, 33). To
validate this assessment, we examined
the failure modes of the tested speci-
mens, and the results are shown in Fig.
9. Referring to Fig. 9, the results clearly
revealed that the failure modes of the
ultrasonic welds depended upon the
weld area. While the partially bonded
joints, having the weld area < 360 mm2
(0.56 in.2), failed interfacially (shown in
Fig. 9B), the fully bonded joints frac-
tured at the workpieces, referring to Fig.
9C and D. These results suggested the
weld area of the ultrasonic welded joints
affected the stress states of the joints.
Referring to Figs. 8 and 9, it clearly
demonstrated the restoration in
strength primarily resulted from an in-
crease in weld area of the welds caused
by the second ultrasonic pulse.
Repair Mechanism
These results show the applied sec-
ond vibration with the proper weld en-
ergy repaired the welds. In order to de-
termine the correlation between the
energy required and degree of the un-
dersized weld, the tests were per-
formed to understand the repair
mechanism.
Figures 10 and 11 present the ef-
fect of repair energy on the weld ap-
pearances and growth of the repaired
welds, respectively. Referring to Fig.
10, while the weld indentation on the
upper workpiece deepened with an in-
crease in repair energy, the flash be-
tween the workpieces changed little
when the repair energy was less than 5
kJ. As the repair energy was increased
to 5 kJ, the flash between the work-
pieces increased significantly. These
results inferred that the applied sec-
ond vibration melted the materials be-
tween the workpieces.
Referring to Fig. 11B, the melting
materials occurred at a few spots and
along the edge of the overlap at the
faying surface for a partially bonded
weld. This irregular weld clearly indi-
cated the workpieces were not in inti-
mate contact during ultrasonic weld-
ing. As the second vibration was ap-
plied to a partially bonded weld, the
melting occurred at the periphery of
the undersized weld and grew with an
increase in weld energy, and the weld
geometry changed from an irregular
shape to a more or less round shape,
referring to Fig. 11C–F. Figure 12

DECEMBER 2017 / WELDING JOURNAL 447-s

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 448

WELDING RESEARCH

C
Fig. 16 — Effect of repair energy on the thickness of upper porous
zone at region 2.

shows the effect of repair energy on the decrease in the D

the weld area of the repaired joints. As porous zone of the
shown, the joints repaired with an en- upper workpiece.
ergy of 5 kJ had a comparable weld The porous zone in
area to that of the fully bonded weld, the melt was
and these results explained the find- thinned out for the E
ings shown in Fig. 8 where the welds weld repaired with
repaired with an energy of 5 kJ had a energy less than 5
comparable strength to that of the ful- kJ. As the repair en-
ly bonded welds. Combining the re- ergy exceeded 5 kJ,
sults in Figs. 8, 11, and 12, it can be the melt at region 2 Fig. 17 — Effect of repair energy on the weld indentation of the
concluded the repair mechanism was was remelted and repaired ultrasonic welded carbon fiber­reinforced nylon 66
the melting of the carbon fiber com- thickened. From the composite: A — Schematic of a weld indentation; B — 3 kJ; C —
posite at the periphery of the existing above analyses, it 4 kJ; D — 5 kJ; E — 6 kJ.
discrepant weld. The increase in the can be concluded
weld area caused by merging the new that the increase in
melt with the existing discrepant weld the weld area result-
was mainly the reason for restoring ing from the second
the weld quality. ultrasonic vibration
Figure 13 shows the cross section was indeed the re-
of a partially bonded weld and an off- pair mechanism of
center undersized weld with a size of the ultrasonic
14 mm (0.55 in.) was observed. To welds.
progressively examine the change in
microstructures of the welds, two re- Effect of Repair
gions were selected. Figure 14A–D
present the microstructures at region
Energy on
1 of the joints repaired with 3, 4, 5, Cosmetic
and 6 kJ, respectively. As shown, while Quality of
the weld at region 1 grew in both di- Repaired Joints
ameter and depth, it contained signifi-
cant pores with a repair energy of 6 kJ. One concern was
Different results were observed that while the appli- Fig. 18 — Effect of repair energy on the weld indentation depth
(Fig. 15) for the microstructures at re- of the repaired ultrasonic welded discrepant carbon fiber­
cation of the second
reinforced nylon 66 composite.
gion 2. The thickness of the melt at re- ultrasonic pulse re-
gion 2 decreased and then increased paired partially
with an increase in repair energy. Ex- bonded welds, it may degrade the cos- pieces of the repaired joints were sec-
amination of the results shown in Fig. metic quality of the workpieces. To as- tioned and the weld indentations were
16 revealed the initial decrease in melt sess the effect of ultrasonic repair on measured. Figures 17 and 18 present
thickness at region 2 was caused by the cosmetic quality, the upper work- the effect of repair energy on the weld

448-s WELDING JOURNAL / DECEMBER 2017, VOL. 96

Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 449
indentation. As shown, the weld in-
dentation virtually increased linearly
with the repair energy when the repair
energy exceeded 3 kJ. These results
suggested melting at the faying sur-
face occurred. While the application of
second pulse increased the weld area,
it also affected the surface quality.
Therefore, the key to restoring the
weld quality is to select the proper
weld energy that would increase the
weld area at the faying surface and
cause minimal degradation on the sur-
face quality.
Summary
The results presented in this study
illustrate the repair of partially bonded
welds in ultrasonic welded carbon
fiber composite with 30% mass. The
horn, part tolerance, and clamping
conditions reported here are not in-
dicative of conditions in vehicle as-
sembly plants. Furthermore, a nonde-
structive method is needed to inspect
the weld quality. Nevertheless, the
presented results indicated the appli-
cation of a second ultrasonic vibration
could repair the undersized or partially
bonded weld and reduce the scatter in
joint strength. Therefore, the process
developed here can be considered as a
potential method for postweld repair
in ultrasonic welding of thermoplastic
composites. The benefit of the method
proposed is that the joint strength for
a weld is improved, and thus reduced
the potential need to scrap the partic-
ular component assembly.
Conclusions
The present study provides a feasi-
ble method for repairing the ultrasonic
welded 4-mm- (0.16-in.-) thick carbon
fiber-reinforced nylon 66 composite
with 30% mass fibers. Extensive tests
conducted on repairing the ultrasonic
welded composite coupons concluded
the following:
1. A method of repairing the ultra-
sonic welded 4-mm- (0.16-in.-) thick lap
welds has been developed. The method
includes positioning the ultrasonic horn
and anvil proximate the poor weld area,
and applying a second ultrasonic vibra-
tion with an energy of 5 kJ at the same
location on the welds fabricated by an
introduction of a separation of 0.7 mm
(0.03 in.) between the workpieces. The
application of the second vibration not
only improved the strength but also de-
creased the scatter in strength of the re-
paired joints.
2. The main mechanism of repair-
ing the ultrasonic welds was the in-
crease in weld area caused by the melt-
ing of the materials at the periphery of
the existing undersized weld at the
faying surface.
3. The presence of a separation be-
tween the workpieces influenced the
contact of the workpieces during ultra-
sonic welding, and consequently re-
sulted in variations in weld area and
weld strength.
Acknowledgments
The authors gratefully acknowledge
the financial and technical support
provided by GM Global Research and
Development Center.
References
1. Mapleston, P. 2002. Nylons drive to
expand role in automotive applications.
Modern Plastics International 79(4): 41, 42.
2. Granowicz, P., Molteni, G. L., and
Kobayashi, T. 2011. New polymer “shield”
technology protects high-performance ny-
lon and PPA polymers to replace more met-
al for weight and cost savings under the
hood. SAE International Journal of Materials
and Manufacturing 4(1): 430–439.
3. Senthilvelan, S., and Gnanamoorthy,
R. 2006. Selective carbon fiber-reinforced
nylon 6, 6 spur gears: Development and
performance. Applied Composite Materials
13(1): 43–56.
4. Grande, J. A. 2005. New high-perfor-
mance nylons for automotive, electronics,
packaging. Plastics Technology 21(5): 53, 54.
5. Liu, Y. J., Ding, L., Li, Q. F., and
Wang, D. 2012. Finite element analysis and
shape optimization of aluminum alloy au-
tomobile energy-absorbing components.
Applied Mechanics and Materials 249–250:
954–957.
6. Feng, N., Wang, X., and Wu, D. 2013.
Surface modification of recycled carbon
fiber and its reinforcement effect on nylon
6 composites: Mechanical properties, mor-
phology and crystallization behaviors. Cur-
rent Applied Physics 13(9): 2038–2050.
7. Roh, S. C., Kim, J., and Kim, C. K.
2013. Characteristics of nylon 6, 6/nylon
6, 6 grafted multi-walled carbon nanotube
composites fabricated by reactive extru-
DECEMBER 2017 / WELDING JOURNAL 449-s
WELDING RESEARCH
sion. Carbon 60(14): 317–325.
8. Baji, A., Mai, Y.-W., Wong, S.-C., Ab-
tahia, M., and Du, X. 2010. Mechanical be-
havior of self-assembled carbon nanotube
reinforced nylon 6, 6 fibers. Composites Sci-
ence and Technology 70(9): 1401–1409.
9. Konishi, Y., and Cakmak, M. 2005.
Structural hierarchy developed in injection
molding of nylon 6/clay/carbon black
nanocomposites. Polymer 46(13):
4811–4826.
10. Senthilvelan, S., and Gnanamoor-
thy, R. 2008. Influence of reinforcement
on composite gear metrology. Mechanism
and Machine Theory 43(9): 1198–1209.
11. Bigg, D. M. 1985. Effect of com-
pounding on the properties of short fiber
reinforced injection moldable thermoplas-
tic composites. Polymer Composites 6(1):
20–28.
12. Watson, M. N., Rivett, R. M., and
Johnson, J. I. 1986. Plastics — An indus-
trial and literature survey of joining tech-
niques. Research Report No. 301/1986.
The Welding Institute, Cambridge,
England.
13. Grewell, D., and Benatar, A. 2007.
Welding of plastics: Fundamental and new
developments. International Polymer Pro-
cessing 22(1): 43–60.
14. Benatar, A., Eswaran, R. V., and Na-
yar, S. K. 1989. Ultrasonic welding of ther-
moplastics in the near-field. Polymer Engi-
neering and Science 29(23): 1689–1698.
15. Benatar, A., and Cheng, Z. 1989. Ul-
trasonic welding of thermoplastics in the
far-field. Polymer Engineering and Science
29(23): 1699–1704.
16. Stokes, V. K. 1988. Vibration weld-
ing of thermoplastics, Part I: Phenomenol-
ogy of the welding process. Polymer Engi-
neering and Science 28(11): 718–727.
17. Stokes, V. K. 1988. Vibration weld-
ing of thermoplastics, Part II: Analysis of
the welding process. Polymer Engineering
and Science 28(11): 728–739.
18. Amanat, N., James, N. L., and
McKenzie, D. R. 2010. Welding methods
for joining thermoplastic polymers for the
hermetic enclosure of medical devices.
Medical Engineering & Physics 32(7):
690–699.
19. Da Costa, A. P., Botelho, E. C., Cos-
ta, M. L., Narita, N. E., and Tarpani, J. R.
2012. A review of welding technologies for
thermoplastic composites in aerospace ap-
plications. Journal of Aerospace Technology
and Management 4(3): 255–265.
20. Zhou, A., and Keller, T. 2005. Join-
ing techniques for fiber reinforced polymer
composite bridge deck systems. Composite
Structures 69(3): 336–345.
21. Lee, J., and Roessler, L. 1997. Vibra-
tion welded composite intake manifolds —
Design considerations and material selection
criteria. Proceedings Society of Automotive En-
gineers International Congress, pp. 27–37.
Lu Paper 201752 December 2017Layout.qxp_Layout 1 11/7/17 3:40 PM Page 450

WELDING RESEARCH

22. Lee, C. S., Kagan, V., Knowlden, N.,
Kim, K. W., and Kim, H. D. 1998. Opti-
mization of vibration weld joint strength
for plastic air intake manifold. Proceedings
Society of Automotive Engineers Internation-
al Congress, pp. 111–115.
23. Bates, P. J., Mah, J. C., Zou, X. P.,
Wang, C. Y., and Baylis, B. 2004. Vibration
welding air intake manifolds from rein-
forced PA 66, nylon 6 and polypropylene.
Composites-Part A 35(9): 1107–1116.
24. Kumar, R. D., Rani, M. R., and Elan-
govan, S. 2014. Design and analysis of slot-
ted horn for ultrasonic plastic welding. Ap-
plied Mechanics and Materials 592–594:
859–863.
25. Liang, J. H., Sidiropoulos, V., Mac-
Donald, J., and Bates, P. J. 2001. Meltdown-
time profiles of vibration welded nylon 66
compounds. Society of Plastics Engineers 59th
Annual Technical Conference, pp. 1224–1228.
26. Villegas, I. F. 2015. In situ monitor-
ing of ultrasonic welding of thermoplastic
composites through power and displace-
ment data. Journal of Thermoplastic Com-
posite Materials 28(1): 65–85.
27. Villegas, I. F. 2014. Strength devel-
opment versus process data in ultrasonic
welding of thermoplastic composites with
flat energy directors and its application to
the definition of optimum processing pa-
rameters. Composite-Part A 65(10): 27–37.
28. Bates, P. J., and MacDonald, J. 2005.
Comparison of experimental and analytical
vibration welding meltdown-time profiles
for Nylon 66 and polypropylene. Polymer
Engineering and Science 45(6): 792–797.
29. Zhi, Q., Lu, L., Gao, Y.-H., Liu, Z.-X.,
and Wang, P.-C. 2017. Online weld quality
inspection in ultrasonic welding carbon
fiber/PA 66 composites without energy di-
rector. Submitted to Welding Journal, man-
uscript number: 201750.
30. Weihai Kaizheng Ultrasonic Tech-
nologies Co. Ltd., China.
31. Schlarb, A. K., and Ehrenstein, G.
W. 1989. The impact strength of butt weld-
ed vibration welds related to microstruc-
ture and welding history. Polymer Engineer-
ing and Science 29(23): 1677–1682.
32. Zhi, Q., Tan, X., Lu, L., Chen, L., Li, J.,
and Liu, Z. 2017. Decomposition of ultra-
sonically welded carbon fiber/polyamide 66
and its effect on weld quality. Weld in the
World doi:10.1007/s40194-017-0482-5.
33. Zhi, Q., Tan, X., and Liu, Z. 2017.
Effect of moisture on the ultrasonic weld-
ing of carbon-fiber-reinforced polyamide
66 composite. Welding Journal 96(6): 185-s
to 192-s.

LEI LU, QIAN ZHI, YUHAO GAO, and ZHONG­XIA LIU (liuzhongxia@zzu.edu.cn) are with the Key Lab of Materials Physics, School of Physics and
Engineering, Zhengzhou University, Zhengzhou, China. PEI­CHUNG WANG is with the Manufacturing Systems Research Lab, General Motors
Research and Development Center, Warren, Mich.

Authors: Submit Research Papers Online

Peer review of research papers is now managed through an online system using Editorial
Manager software. Papers can be submitted into the system directly from the Welding Jour-
nal page on the AWS website (aws.org) by clicking on “submit papers.” You can also access
the new site directly at editorialmanager.com/wj/. Follow the instructions to register or log
in. This online system streamlines the review process, and makes it easier to submit papers
and track their progress. By publishing in the Welding Journal, more than 70,000 mem-
bers will receive the results of your research.
Additionally, your full paper is posted on the American Welding Society website for FREE
access around the globe. There are no page charges, and articles are published in full color.
By far, the most people, at the least cost, will recognize your research when you publish in
the world-respected Welding Journal.

450-s WELDING JOURNAL / DECEMBER 2017, VOL. 96