Online Inspection of Weld Quality in Ultrasonic Welding of ...€¦ · the shop floor, the tests...

Introduction Automobile exhaust emission hasdrawn wide public concern and onepossible solution to this issue is reduc-ing the weight of vehicles (Refs. 1–4).Lightweight materials, such as ad-vanced polyamide composites rein-forced by primary carbon fibers, arebeing increasingly used in automotiveand aerospace applications due to

their good mechanical properties andlow density (Refs. 1, 5, 6). To obtaincomplex components, which oftenconsist of small parts, joining thepolyamide parts becomes a primarypriority (Refs. 7–10). Among all thejoining techniques available, ultrasonicwelding is one of the most promisingmethods because it is fast, energy effi-cient, potentially suitable for massproduction, and offers sound cosmetic

quality (Refs. 11–14). The key feature of ultrasonic weld-ing is that the ultrasonic vibrationsproduce the intermolecular diffusionand the entanglement of the molecu-lar chain in the molten state on the in-terfaces of the welded polymer due tothe intermolecular friction within thethermoplastics (Refs. 11–14). To in-crease the welding speed and weldquality, the energy director is usuallyrecommended to direct and focus theultrasonic weld energy to a smallercontact area (Refs. 15, 16). However,due to the difficulty in locating the en-ergy directors in the lapped work-pieces with a robot, ultrasonic weldingof the workpieces without energy di-rectors was assessed. To facilitate theuse of the ultrasonic welded polymericcomposites for lighter, stronger, andmore cost-effective vehicle structures,manufacturing guidelines for compos-ite structures are required. The devel-opment of these guidelines requiresnot only detailed understanding of theultrasonic welding process but also thequality monitoring and controlling ofthe ultrasonic welded joints. To inspect the weld quality, variousmethods (e.g., destructive such as thechisel test) or nondestructive testing(e.g., ultrasonic technique (Ref. 15))have been applied to assess the weldquality. Although the chisel test hassome advantages, such as its relativeease of use, low cost, and usability on

WELDING RESEARCH

MARCH 2018 / WELDING JOURNAL 65-s

SUPPLEMENT TO THE WELDING JOURNAL, March 2018Sponsored by the American Welding Society

Online Inspection of Weld Quality in Ultrasonic Welding of Carbon Fiber/Polyamide 66

without Energy DirectorsA method of inspection was developed based on

target horn displacement and duration

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

ABSTRACT In this study, a method of inspecting the weld quality (i.e., static strength) in ultrasonic welding of 4-mm- (0.16-in.-) thick carbon fiber/polyamide 66 (Cf /PA66) composite with 30 wt-% fiber without energy directors was investigated.The transient horn displacements and dissipated powers were recorded to corre-late with the weld quality. It was found that the transient horn displacement ofthe joint during the ultrasonic welding process displayed four distinct stages: Cf

/PA66 expansion phase (stage I), unsteady melting phase (stage II), equilibriumphase of the materials melting and melt outflow (stage III), and the material cooling down phase (stage IV). The weld quality was closely related to the horndisplacement and duration in stage III of the ultrasonic welding process. Targethorn displacement and duration in stage III were determined by desired staticstrengths of the welded 4-mm- (0.16-in.-) thick Cf /PA66 composite. The horn dis-placements and durations of the welds in stage III were recorded and the meas-urements were compared with the target values during welding. A quality weldwas determined by judging whether the measured values of horn displacementand duration in stage III were within the tolerance range of the target values. Consequently, an online inspection method to evaluate the weld quality was developed based on target horn displacement and duration.

KEYWORDS • Ultrasonic Welding • Carbon Fiber/Polyamide 66 • Weld Quality • Online Inspection

https://doi.org/10.29391/2018.97.006

Zhi Supplement correctednew.qxp_Layout 1 2/9/18 3:17 PM Page 65

the shop floor, the tests are qualitativeand subjective. The ultrasonic methodoften requires a relatively long cycletime and couplant gel. It is often usedoffline for quality inspection of thewelded workpieces. Therefore, it is im-perative to develop an online nonde-structive inspection method to evalu-ate the quality of the ultrasonic weld-ed thermoplastic composites. An extensive literature review re-vealed that few studies have reportedon online nondestructive inspectionfor ultrasonic welds. Tolunay (Ref. 17),Stokes (Ref. 18), and Wijk (Ref. 19)previously found signals (i.e., ultrason-ic wave and process signatures) couldindirectly reflect the weld quality.Bates et al. (Ref. 20) studied the effectof imperfect mating parts on the vi-bration welding of a polyamide 66compound and found the part qualitywas correlated to the meltdown pro-file. Villegas et al. (Refs. 21, 22) dis-covered that the quality of the ultra-sonic welded lap joints of CF/PEI witha flat energy director was related tothe process signatures (e.g., horn dis-placement and dissipated power of theprocess). However, the quantitativecorrelation between the joint strengthand process signatures was not reported. The present study was undertakento develop a method for inspecting on-going welding operations as a se-quence of ultrasonic welded carbonfiber/polyamide 66 composite jointswith 30 wt-% fiber without energy di-rectors being formed. We first focused

our attention on conducting the teststo understand the effect of processvariables on the joint strength. Then,microstructures and weld formationmechanism of the ultrasonic weldedjoints were analyzed and the correla-tion between the joint strength andprocess signatures was investigated.Finally, the application of this methodfor weld quality inspection was dis-cussed.

Experimental Procedure

Materials

Commercial polyamide 66 and car-bon fiber (24 K, T300 type, Toray Car-bon Magic Co. Ltd.) with a length of 2mm (0.08 in.) were used. The fiberswere first cleaned with a concentratedsolution of nitric acid and then surfacepretreated using 8% diglycidyl ether ofbisphenol solution in acetone. Bothpolyamide 66 and pretreated carbonfibers were dried at 80C in a vacuumcondition for 3 h before being used tofabricate 30 wt-% carbon fiber/polyamide 66 composite. A twin-screw extruder with twoseparate inlets was used to mold 30wt-% carbon fiber/polyamide 66 com-posite. Polyamide 66 was added to thefirst hopper, and carbon fibers wereadded to the second hopper.Polyamide 66 was also fully melted be-fore carbon fibers were added to mini-mize the fracture of carbon fiber dur-ing compounding. The processing tem-

perature was 270°–280°C(518°~536°F), and the screw speed was180 roations per min (rpm). After fullymixing polyamide 66 with carbonfibers in a twin-screw extruder, thecarbon fiber/polyamide 66 compositewas processed into the pellets with alength of 2 mm (0.08 in.). The pelletwas then fed into the injection extrud-er to mold into coupons with dimen-sions of 132 38 4 mm (5.2 1.5 0.16 in.). All coupons were stored in anambient laboratory environment(20C (68F) and 50% R.H.) and driedin a vacuum oven at 80C (176F) for48 h before welding to completely re-move moisture in the specimen. Theinjection molded carbon fiber/polyamide 66 composite exhibited atensile strength of 99.2 ± 3 MPa ((1.4± 0.4) 104 lb/in.2) and an elasticmodulus of 8936 ± 3 MPa ((1.3 ± 0.6) 106 lb/in.2), respectively.

Ultrasonic Welding Process

The ultrasonic welding process wasperformed using a KZH-2026 multi-function UW machine (WeihaiKaizheng Ultrasonic Technologies Co.Ltd., China) with a nominal power of2.6 kW, nominal frequency of 20 kHz,and nominal amplitude of 25 m (9.8 104 in.). The machine was equippedwith a data acquisition system thatcombined a horn pressure sensor,horn-displacement sensor, and atimer, which were integrated in thecontroller of the UW machine, asshown in Fig. 1. In addition, the horn

WELDING RESEARCH

WELDING JOURNAL / MARCH 2018, VOL. 9766-s

Fig. 1 — Schematic for ultrasonic welding of injection moldedcarbon fiber/polyamide 66 composite.

Fig. 2 — Sketch of the temperature measurements duringthe ultrasonic welding of injection molded 4-mm- (0.16-in.-)thick lapped carbon fiber/polyamide 66 composite withoutenergy directors.


pressure, weld energy, and horn dis-placement were recorded online in apersonal computer (PC) as a functionof time by the data acquisition system.The final horn displacement, weld en-ergy, weld time, horn pressure, holdtime, and delay time were also dis-played in the control panel during theUW process. To avoid coupon motionduring welding, the coupons were heldin place by using a fixture. The machine had three modes: en-ergy, time, and horn displacement.The value of weld energy, weld time,and horn displacement for the threemodes, respectively, were preset tocontrol the welding process. The work-pieces were then welded using thenominal power of the machine. Whenthe weld energy, weld time, or horndisplacement reached the preset val-ues for the selected weld mode, the ul-trasonic wave oscillation was stopped.Therefore, weld quality was controlledby the preset values in each selectedwelding mode. When time mode wasselected, the values of the delay time,weld force, hold time, and ultrasonictime were preset prior to welding.When the ultrasonic triggering wasperformed, the horn was pressed ontothe workpieces for 2 s, and then it ul-trasonically vibrated until the presettime was reached. The welded work-pieces were held for 3 s to solidify themolten material. All specimens werewelded using a 7075 aluminum hornwith a diameter of 18 mm (0.71 in.).

Temperature Measurement

To analyze the weld initiation andgrowth during ultrasonic welding, thetemperature evolutions at the locationsnear the horn-workpiece interface andfaying surface were measured. Figure 2shows the experimental setup for tem-perature measurements. As shown, twosmall holes with a diameter of 1.0 mm(0.04 in.) and a depth of 12.5 mm (0.49in.) were drilled at the side of the upperworkpiece. Two holes were drilled at 0.2mm (0.008 in.) from the top and bot-tom surfaces of the upper workpiece, re-spectively. Two K-type thermocoupleswere imbedded into the two small holesand secured with epoxy compound sothat the thermocouples were secured.The temperature evolutions at thesetwo locations were recorded as a func-

tion of time by a data acquisition sys-tem during ultrasonic welding.

Weld Microstructure

To assess the characteristics of theweld microstructure of the ultrasonicwelded joints, the specimens were pre-pared using the procedures shown inFig. 3 for the tested joints. In Fig. 3, thejoints were notched from the central po-sition of the weld. Then, the prenotchedspecimens were immersed in liquid ni-trogen for 10 min. The embrittled speci-mens were broken off from the notchedsite; the broken specimens were sput-ter-coated with gold for 50 s to increasethe conductivity, and the microstruc-tures of the welds were examined with ascanning electron microscope(JSM6700F).

WELDING RESEARCH


Fig. 3 — Schematic of a sample preparation forexamining the microstructure of the ultrasonicwelded 4-mm- (0.16-in.-) thick injection moldedcarbon fiber/polyamide 66 composite.

Fig. 4 — Influences of welding time and horn pressure on — A — staticstrength; B — weld area of the ultrasonic welded 4-mm- (0.16-in.-) thick lapcarbon fiber/polyamide 66 composite with 30 wt-% carbon fiber.

A

B


Quasi-Static Test

Quasi-static tests were performedby loading each specimen to failure inan MTS 810 tensile tester according toASTM D1002-2001. To minimizebending stresses inherent in the test-ing of single-lap weld specimens, fillerplates were attached onto both ends ofthe specimen using a masking tape toaccommodate the sample offset. Loadvs. displacement results were ob-tained, as the specimens were loadedat a stroke rate of 2 mm/min (0.08in./min). The strength of the joint isdefined as the ratio of the peak load tothe overlap area of the weld (950 mm2

(1.47 in.2)). Three replicates were per-formed, and the average jointstrengths were reported.

Results and Discussion

Quality of Ultrasonic Weld

To understand the weld quality inultrasonic welding of 4-mm-(0.16-in.-)thick lap carbon fiber/polyamide 66composite with 30 wt-% fiber, exten-sive welding tests were performed. Insetting up the welding machine to pro-duce a series of uniform welds, initialvalues of suitable horn pressure, weld-ing time, and clamping conditionswere established for the welding hornand specific workpieces. The weldingcontroller was programmed in an at-tempt to maintain these values so thatthe same welds are produced duringextended manufacturing operations. Figure 4A, B shows the effect ofprocess variables on the weldstrength and weld area for an ultra-sonic welded 4-mm- (0.16-in.-) thicklap carbon fiber/polyamide 66 com-posite, respectively. As shown, signifi-cant scatter in the weld strength andweld area for the joints made underthe same welding conditions was ob-served. Careful analyses of the resultsrevealed that the scatter likely result-ed from the variations in weld area.To validate this speculation, the fail-ure modes of the tested joints madewith a welding time of 2.1 s and ahorn pressure of 0.15 MPa (21.76lb/in.2) were examined. Figure 5A, B presents an interfacialfailure mode and a mixed mode of in-terfacial and workpiece fracture, re-spectively. Examination of the failure

mode shown in Fig. 5A exhibited thatthe cracks initiated at the intersectionof the weld and opening between theworkpieces, and grew through theweld nugget. Ultimate failure was byoverload of the remaining unseparatedsections. A different failure mode, re-ferring to Fig. 5B, revealed that thecracks initiated at the pores inside theweld, and then propagated throughthe thickness of the upper workpiece. In combining Figs. 4 and 5, it canbe seen the joints that failed with aninterfacial failure mode had greaterstrength than the joints that failedwith a mixed failure mode. Figure 6A,B showed the microstructures of thejoints failed with interfacial andmixed fracture modes, respectively.As shown, significant pores werepresent at the porous zone for thejoints with a mixed failure mode, asseen in Fig. 6B. The presence of poresin Fig. 6B was likely the culprit forthe reduction in joint strength.Therefore, significant scatter in thejoint strength, as in Fig. 4, likely re-sulted from the variations in weldquality. While the exact causes ofpores are beyond the scope of thisstudy, the implication of these resultsshowed there is an urgent need tohave an online system developed tomonitor weld quality.

Weld Growth Mechanism

To develop the weld monitoringsystem, it is necessary to understandhow the weld initiates and grows dur-ing the welding of 4-mm- (0.16-in.-)thick lap carbon fiber/polyamide 66composite. From the results shown inthe “Quality of Ultrasonic Weld” sec-tion, the optimal welding parametersfor UW of Cf /PA 66 were a weldingtime of 2.1 s and a horn pressure of0.15 MPa (21.76 lb/in.2). Thus, tran-sient temperature histories near thefaying surface and horn-workpiece in-terface, as well as the transient horndisplacement for the joints made withthe optimal welding parameters, wererecorded. The results are shown in Fig.7. In addition, the transient horn dis-placement during welding was alsorecorded and shown in Fig. 7. Asshown, the weld initiation and growthcould be divided into four stages dur-ing the welding, and each stage is dis-cussed next.

Stage I

In this stage, ultrasonic vibrationbegan, and the friction between thehorn-workpiece and workpiece-work-piece resulted in an increase in tem-perature at both interfaces. Small as-

WELDING RESEARCH

WELDING JOURNAL / MARCH 2018, VOL.9768-s

Fig. 5 — Failure modes for the ultrasonic welded 4-mm- (0.16-in.-) thick lap carbonfiber/polyamide 66 composite with 30 wt-% carbon fiber and without an energy di-rector fabricated with a welding time of 2.1 s and a horn pressure of 0.15 MPa (21.76lb/in.2): A — interfacial failure; B — mixed interfacial and workpiece fracture.

A

B


perities were in contact at the fayingsurface, where high stresses were con-centrated on these asperities, and con-sequently the temperature increasedto 70C (158F) near the faying sur-face within 0.35 s. As the welding con-tinued, the temperature near thehorn-workpiece interface barely in-creased after a welding time of 0.35 sand the horn displaced slightly upwardprimarily resulting from the expansionof the materials. Figure 8A, B showsthe weld indentation and weld areawith a welding time of 0.3 s, respec-tively. As shown, little weld indenta-tion and a small weld with an irregularshape were observed in stage I.

Stage II

In stage II, friction between inter-faces and viscoelastic dissipation ofthe material resulted in an increased

temperature at the interfaces. Refer-ring to Fig. 7, the temperature nearthe faying surface virtually increasedlinear with the welding time andreached a peak temperature of 379C(714F) at a welding time of 1.7 s,which exceeded the melting point(259C (498F)) of carbon fiber/polyamide 66 composite. The meltingat the faying surface began to growand spread, and the correspondinghorn displacement increased slowly.Due to the molten material serving asa “lubricating agent” (Ref. 23), the fric-tion energy dissipated as the fayingsurface decreased, and consequentlythe melt rate of materials at the fayingsurface began to decrease. Althoughthe raising temperature rate near thefaying surface decreased, the tempera-ture continuously increased to reach astable 379C (714F). Figure 9A to D shows the weld in-

dentation and weld area for the weldingtime of 0.5, 0.9, 1.3, and 1.7 s, respec-tively, in stage II. As shown, the weld in-dentation and weld area grew withwelding time, which resulted in a melt-ed film forming at the faying surface,and the horn moved downward. Refer-ring to Fig. 7, as the temperature in-creased near the faying surface, thetemperature near the horn-workpieceincreased as well. The temperature nearthe horn-workpiece interface reachedabout 360C (680F) with a weldingtime of 1.7 s, which resulted in a weldindentation of about 0.05 mm (0.002in.) on the upper workpiece.

Stage III

In stage III, referring to Fig. 7, asthe temperature near the faying sur-face stabilized, the temperature nearthe horn-workpiece interface in-creased with time. Figure 10A, Bshows the weld indentation and weldarea with a welding time of 2.1 s, re-spectively. As shown, as the horn in-dented into the upper workpiece, amelt film formed, and some moltenmaterials were flashed out of the fay-ing surface. As a result, a sharp in-crease in horn displacement was ob-served in stage III. Under this condi-tion, the melt rate of materials was inequilibrium with the spread rate of themelt (Ref. 24). It is noted from Fig. 7 that al-though the temperature near thehorn-workpiece interface was lowerthan that of the region near the fayingsurface at the beginning of the ultra-sonic welding process, it reached about

WELDING RESEARCH


Fig. 6 — Microstructures of the ultrasonic welded 4-mm- (0.16-in.-) thick lapped carbon fiber/polyamide 66 composite with 30 wt-% carbon fiber and without energy directors fabricated with a welding time of 2.1 s and a horn pressure of 0.15 MPa (21.76lb/in.2): microstructure of — A — nominal; B — discrepant weld.

Fig. 7 — Temperature-time histories near the horn-workpiece and faying surfaces aswell as horn displacement for ultrasonic welding of 4-mm- (0.16-in.-) thick lap carbonfiber/polyamide 66 composite with 30% mass fiber and without an energy directorfabricated with a welding time of 2.1 s and a horn pressure of 0.15 MPa (21.76 lb/in.2).

A B


404C (759F) in stage III. These re-sults inferred that as the energy dissi-pated and the faying surface de-creased, more energy was dissipated atthe horn-workpiece interface, whichwas mainly because the upper work-piece softened from dissipated heatand made the ultrasonic vibration dif-ficult to transmit into the faying inter-face. The temperatures near the fayingand horn-workpiece interfaces weremuch higher than the melting point(259C (498F)) of carbonfiber/polyamide 66 composite, whichlikely degraded polyamide 66 and re-sulted in volatile gases observed at theporous zone (Ref. 25).

Stage IV

Ultrasonic vibration was stoppedand the melt began to solidify in stageIV. Referring to Fig. 7, the tempera-tures near the faying and horn-work-piece interfaces decreased with time asthe horn moved slightly downward re-sulting from the material contracting.From the above analyses, we can con-clude that the weld initiation andgrowth mainly depended on the ener-gy consumed by the materials at thefaying surface.

Correlation Between Horn Displacement and DissipatedPower

From the analyses above, it wasnoted that the energy dissipated at thefaying surface was related to the weldarea. In this section, the correlationbetween the weld area and energy dis-sipation, and how the dissipated pow-er relates to the horn displacementthat can reflect the weld growth, areexamined. It is well known that the heat rateof the faying surface during the ultra-sonic welding process is determined bythe viscoelastic dissipation and fric-tion between two workpieces, andthus by the dissipated power (Refs. 11,19, 26, 27). The lumped parametermodel for the simulation of the ultra-sonic welding of thermoplastics pro-posed by Benatar et al. demonstratedthat the instantaneous dissipatedpower (Ws) during ultrasonic weldingwas closely related to the impendenceof horn-workpieces interface, fayingsurface, and workpiece-anvil interface(Ref. 12).

P1 = 2fE” (A0/L1)2(0.5 – Z2

cos2 + 0.5Z22) (1)

P2 = 2fE”(A0/L2)2[0.5(Z22

+ Z23)+ Z2Z3cos(2 – 3)] (2)

P3 = 2fE”(A0/L3)2 Z23/2 (3)

Ws = P1 + P2 + P3 (4)

where P1, P2, and P3 are the instanta-neous dissipated power for the horn-workpiece, workpiece-workpiece, andworkpiece-anvil interfaces, respective-ly. E" is the loss modulus; f is the vibra-tion frequency; A0 is the amplitude ofvibration; L is the contact length ofthe faying surface; Z2 and Z3 are theimpedances of the faying surface andworkpiece-anvil interface, respectively;and 2 and 3 are the phase angles ofthe vibration on the faying surface andworkpiece-anvil interface, respectively.Ws is the total energy dissipated dur-ing the ultrasonic welding process.Furthermore, experiments and simu-lation done by Benatar et al. (Ref. 12)and Villegas (Refs. 21, 22) showed thatwhile the melting of asperities and en-ergy directors at the faying surfacecaused an abrupt increase in the im-pedance of the faying surface (Z2), acontinuous decrease in dissipatedpower was observed. Therefore, simi-lar observations were drawn that therewas a connection among the ultrasonicvibration phase, dissipated power, andhorn displacement. To examine if the horn displace-ment relates to the dissipated power,the dissipated power and horn dis-placement of ultrasonic welding of 4-mm- (0.16-in.-) thick carbon fiber/polyamide 66 composite made with awelding time of 2.1 s under a hornpressure of 0.15 MPa (21.76 lb/in.2)were measured. Figure 11A, B showsthe results for the joints with thestrengths of 6.1 (884.7 lb/in.2) and4.7 MPa (681.6 lb/in.2), respectively.Referring to Fig. 11, the dissipatedpower in stage I linearly increased asthe friction/viscoelastic dissipationoccurred between the horn-work-piece and faying surface. Because ofheating, the workpieces expanded,and consequently led to the hornmoving upward. As the welding con-tinued, more asperities at the fayingsurface melted in stage II, and thedissipated power reached a maximumvalue. This is followed by all asperi-ties melting at the faying surface,and a steady melt film was formed in

WELDING RESEARCH


Fig. 8 — A — Weld indentation; B — weld size at the faying surface of the ultrasonicwelded 4-mm- (0.16-in.-) thick carbon fiber/polyamide 66 composite with 30 wt-%fiber and without an energy director fabricated with a welding time of 0.3 s and ahorn pressure of 0.15 MPa (21.76 lb/in.2) (stage I).

A

B


stage III. As the melting and spread-ing rates of material at the fayingsurface were in equilibrium, less vi-bration energy was required to meltthe material at the faying surface tocompensate the spread melt (Ref.21). The horn displacement increasedand dissipated power decreased withtime until the ultrasonic vibrationwas stopped due to the formation ofthe melt at the faying surface. Theseresults suggest the horn displace-ment and dissipated power wereclosely related. Careful examinations of the resultsshown in Fig. 11 indicated the jointwith a strength of 6.1 MPa (884.7

lb/in.2) had a greater increase in horndisplacement and duration than thatof the joint with a strength of 4.7 MPa(681.6 lb/in.2) in stage III. Similar re-sults were observed for the jointsmade with other welding variables.These results imply that the horn dis-placement and duration in stage IIImay be a good indicator of the qualityin ultrasonic welding of carbonfiber/polyamide 66 composite.

Extraction Horn Displacement and Duration of Stage III inUltrasonic Welding

As shown above, the changes of

the horn displacement and durationin stage III play an important role indetermining the joint strength. Inthis section, the extraction of thechanges of horn displacement andduration in stage III during ultrason-ic welding is studied. Figure 12 showsthe horn displacement and partialdissipated power (2000 to 2060 W)of Fig. 11A. Referring to Fig. 12, thechange of the horn displacement(DT3) in stage III can be determinedfrom the duration (tT3) that is deter-mined by the change of dissipatedpower. As shown in Fig. 12, the melt-ing of asperities with various heightsat the faying surface resulted in afluctuation series of dissipated powerin stage II. A melt film formed andthe dissipated power dropped dra-matically in stage III. Therefore, thesharp drop in dissipated power wasan indicator of the starting time ofstage III. The duration in stage III(tT3) was determined from the sharpdrop in dissipated power to the endof the vibration, referring to Fig. 13.Therefore, the change of horn dis-placement (DT3) was determined asthe duration where the dissipatedpower started to decrease in stage IIIuntil the end of vibration. The com-bination of the duration (tT3) andchange of displacement (DT3) instage III were then used to correlatethe strength of ultrasonic welded car-bon fiber/polyamide 66 composite.

Correlation between Horn Displacement and Duration in Stage III and Joint Strength

Once the method of inspecting theweld quality was proposed, it was neces-sary to validate this approach. Ultrason-ic welding of carbon fiber/ polyamide 66composite with 30% mass fiber was per-formed with a welding time of 2.1 s anda horn pressure of 0.15 MPa (21.76lb/in.2). The horn displacement and dis-sipated power histories were recorded,and the changes in horn displacementand duration in stage III for each testwere measured using the aforemen-tioned method, referring to Fig. 12. Fig-ure 13 presents the correlation betweenthe joint strength and combination ofduration (tT3) and change of displace-ment (DT3) in stage III. As shown, agrey zone that defines the joints withacceptable strength (>5.5 MPa (797.7

WELDING RESEARCH


Fig. 9 — Effect of welding time on the weld indentation and weld formation at thefaying surface of the ultrasonic welded 4-mm- (0.16-in.-) thick carbon fiber/polyamide 66 composite with 30 wt-% fiber and without an energy director fabri-cated with a horn pressure of 0.15 MPa (21.76 lb/in.2): A — 0.5 s; B — 0.9 s; C — 1.3 s;D — 1.7 s (stage II).

A

C

B

D

Zhi Supplement correctednew.qxp_Layout 1 2/9/18 3:17 PM 1

lb/in.2)) was determined. These resultssuggest that if the duration and changeof horn displacement (DT3) in stage III of the ultrasonic weldedjoint fall in the range of the grey area,the joint with desired strength can be obtained. Further analyzing the results shownin Fig. 13 revealed that to obtain asound weld, the duration (tT3) andchange of horn displacement (DT3) instage III must be kept in a proper range.The welds with small horn displacement(DT3) had a thin film thickness, andconsequently had a weak strength. Onthe other hand, the welds having longduration (tT3) and a large change inhorn displacement (DT3) had a thickfusion zone with significant pores there.Both had a negative impact on the jointstrength. Therefore, the key to produc-ing a quality weld is to control the dura-tion (tT3) and horn displacement(DT3) in proper range for ultrasonicwelding of carbon fiber/polyamide 66

composite. Referring to Fig. 13, to ob-tain the sound weld, the duration andchange of horn displacement were inthe ranges of 0.25 to 0.35 s and 0.04(0.0016 in.) to 0.1 mm (0.004 in.), re-spectively, for ultrasonic welding of 4-mm- (0.16-in.-) thick lap carbon fiber/polyamide 66 composite.

Application and Prospect of HornDisplacement for Weld QualityInspection

The aforementioned test resultsdemonstrated that the correlation of theweld strength and horn displacementand duration in stage III of ultrasonicwelding could be used to nondestruc-tively assess the weld quality online.Target values for the horn displacementand duration in producing a sound weldare determined for the workpieces. Dur-ing welding operation, the horn dis-placement and duration during the weldgrowth in stage III are continually taken

and analyzed, and the results were com-pared with the target displacement andduration values. While the measuredvalues are within a tolerance range ofthe target values, welding is continuedusing the preset parameters. When themeasured horn displacement and dura-tion fall outside the target range, eitherthe fault alarm is flagged or welding pa-rameters can be adjusted to correct thedisplacement and duration and, there-after, welding is continued. Consequent-ly, the horn displacement method wouldpermit easy and efficient production ofa series of substantially sound weldswithout interruption of the welding op-eration. Checking the horn displace-ment online is done without stoppingthe welding process.

Conclusions

1. The weld initiation and growth inultrasonic welding of 4-mm- (0.16-in.-)thick lap carbon fiber/polyamide 66

WELDING RESEARCH


Fig. 10 — A — Weld indentation; B — weld at the faying sur-face of the ultrasonic welded 4-mm- (0.16-in.-) thick carbonfiber/polyamide 66 composite with 30 wt-% fiber at a weld-ing time of 2.1 s and a horn pressure of 0.15 MPa (21.76 lb/in.2)(stage III).

Fig. 11 — Horn displacement and dissipated power curvesfor the ultrasonic welded carbon fiber/polyamide 66 com-posite joints fabricated with a welding time of 2.1 s under0.15 MPa (21.76 lb/in.2) and with strengths of — A — 6.1 MPa(884.7 psi); B — 4.7 MPa (681.6 lb/in.2).

A

B

A

B


composite with 30 wt-% fiber withoutan energy director was composed offour stages, namely, the Cf /PA66 expan-sion phase (stage I), the unsteady melt-ing phase (stage II), the equilibriumphase of the materials melting and themelt outflow (stage III), and the materi-al cooling down phase (stage IV). 2. Under a given horn pressure,melt spread (i.e., weld area) was closelyrelated to the horn displacement aswell as duration in stage III of thewelding process. 3. The criterion for producing quali-ty welds was to keep the duration(tT3) and horn displacement (DT3) instage III in proper range for ultrasonicwelding of 4-mm- (0.16-in.-) thick car-bon fiber-reinforced polyamide 66composite. Moreover, the target horndisplacement and duration were deter-mined by desired static strengths ofthe welded workpieces. 4. An online inspection method toevaluate the weld quality was devel-oped. The horn displacement and dura-tion of the weld in stage III during weld-ing were recorded, and the measure-ments were compared with the targetvalues. A quality weld was determinedby judging whether the measured valuesof horn displacement and welding timewere within the tolerance range of thetarget values.

The authors gratefully acknowledge

the financial and techni-cal support provided by GM Global Re-search and Development.

1. Elalem, A., and El-Bourawi, M. S.2010. Reduction of automobile carbondioxide emissions. International Journal ofMaterial Forming 3(1): 663–666. 2. Van den Brink, R. M. M., and VanWee, B. 2001. Why has car-fleet specialfuel consumption not shown any decreasesince 1990? Quantitative analysis of Dutchpassenger car-fleet special fuel consump-tion. Transportation Research Part D 6(2):75–93. 3. Duflou, J. R., De Moor, J., Verpoest,I., and Dewulf, W. 2009. Environmentalimpact analysis of composite use in carmanufacturing. Manufacturing Technology58(1): 9–12. 4. Witik, R. A., Payet, J., Michaud, V.,Ludwig, C., and Manson, J.-A. E. 2011. As-sessing the life cycle costs and environ-mental performance of lightweight materi-als in automobile applications. Composites:Part A 42(11): 1694–1709. 5. Hodkinson, R., and Fenton, J. 2000.Lightweight construction materials andtechniques. Lightweight Electric/Hybrid Ve-hicle Design 44(3): 173–198. 6. Davies, G. 2003. Future trends in au-tomotive body materials. Materials for Au-tomobile Bodies: 547–396. 7. Ageorges, C., Ye, L., and Hou, M.2001. Advances in fusion bonding tech-niques for joining thermoplastic matrix.Composites: Part A 32(6): 839–857. 8. Fiorini, M., Campana, G., Pagano, N.,and Morelli, R. 2015. Comparison betweenjoining technologies for polymeric films.

9th CIRP Conference on Intelligent Compu-tation in Manufacturing Engineering —CIRP ICME ‘14. Procedia CIRP 33:412–417. 9. Amancio-Filho, S. T., and dos Santos,J. F. 2009. Joining of polymers and polymer-metal hybrid structures: Recentdevelopments and trends. Polymer Engi-neering and Science 49(8): 1461–1476. 10. Grewell, D., and Benatar, A. 2007.Welding of plastic: Fundamentals and newdevelopments. International Polymer Pro-cessing 22(1): 43–60. 11. Levy, A., Le Corre, S., Chevaugeon,N., and Poitou, A. 2011. A level set basedapproach for the finite element simulationof a forming process involving multi-physics coupling: Ultrasonic welding ofthermoplastic composites. European Jour-nal of Mechanics A: Solids 30(4): 501–509. 12. 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(29): 1689–1698. 13. Benatar, A., and Cheng, Z. 1989. Ul-trasonic welding of thermoplastics in thefar-field. Polymer Engineering and Science29(29): 1699–1704. 14. Schell, J. S. U., Guilleminot, J., Bi-netruy, C., and Krawczak, P. 2009. Compu-tational and experimental analysis of fu-sion bonding in thermoplastic composites:Influence of process parameters. Journal ofMaterials Processing Technology 209(11):5211–5219. 15. Srajbr, C., Tanasie, G., Dilger, K.,and Bohm, S. 2011. Active thermographyfor quality assurance of joints in automo-bile manufacturing. Welding in the World55(7–8): 90–97. 16. Khoy, Y., Chien, L.-H., Chang, B. C.,and Liu, S.-J. 2000. Effects of the shape ofthe energy director on far-field ultrasonic

WELDING RESEARCH


Fig. 12 — Method to extract the horn displacement (DT3) and weld-ing time (tT3) in stage III in ultrasonic welding of carbonfiber/polyamide 66 composite with 30 wt-% fiber under a weldingtime of 2.1 s and a horn pressure of 0.15 MPa (21.76 lb/in.2).

Fig. 13 — Correlation between the duration (tT3) horndisplacement (DT3) in stage III and strengths for the ul-trasonic welded 4-mm- (0.16-in.-) thick lap carbonfiber/polyamide 66 composite with 30 wt-% fiber.

References

Acknowledgments


welding of thermoplastics. Polymer Engi-neering and Science 40(1): 157–167. 17. Tolunay, M. N., Dawson, P. R., andWang, K. K. 1983. Heating and bondingmechanisms in ultrasonic welding of ther-moplastics. Polymer Engineering and Science2(13): 726–733. 18. Stokes, V. K. 1988. Vibration weld-ing of thermoplastics Part I: Phenomenolo-gy of the welding process. Polymer Engi-neering and Science 28(11): 718–727. 19. Van Wijk, H., Luiten, G. A., VanEengen, P. G., and Nonhof, C. J. 1996.Process optimization of ultrasonic welding.Polymer Engineering and Science 36(9):1165–1176. 20. Bates, P. J., MacDonald, J.,Sidiropoulos, V., and Liang, H. 2001. Meltdown-time profiles of vibration weld-ed polyamide 66 compounds. Society ofPlastics Engineers 59th Annual TechnicalConference, pp. 1224–1228. 21. Fernandez Villegas, I. 2015. In situmonitoring of ultrasonic welding of ther-moplastic composites through power anddisplacement data. Journal of Thermoplastic

Composite Materials 28(1): 66–85. 22. Fernandez Villegas, I. 2014.Strength development versus process datain ultrasonic welding of thermoplasticcomposites with flat energy directors andits application to the definition of opti-mum processing parameters. Composites:Part A: Applied Science and Manufacturing65(10): 27–37. 23. Bates, P. J., and MacDonald, J. 2005.Comparison of experimental and analyticalvibration welding meltdown-time profilesfor nylon 66 and polypropylene. PolymerEngineering and Science 45(6): 789–797. 24. Stokes, V. K. 1988. Vibration weld-ing of thermoplastics. Part II: Analysis ofthe welding process. Polymer Engineeringand Science 28(11): 728–739. 25. Zhi, Q., Tan, X., Lu, L., Chen, L., Li,J., and Liu, Z. 2017. Decomposition of ul-trasonically welded carbon fiber/polyamide66 and its effect on weld quality. Weld inthe World 61(5): 1017–1028. 26. 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-sto 192-s. 27. Nonhof, C. J., and Luiten, G. A.1996. Estimates for process conditionsduring the ultrasonic welding of thermo-plastics. Polymer Engineering and Science36(9): 1177–1183.

WELDING RESEARCH


QIAN ZHI, YU-HAO GAO, LEI LIU, andZHONG-XIA LIU ([email protected]) are with the Key Lab ofMaterials Physics, School of Physicsand Engineering, Zhengzhou Univer-sity, Zhengzhou, Henan, China. PEI-CHUNG WANG is with the Manufactur-ing Systems Research Lab, GeneralMotors Research and DevelopmentCenter, Warren, Mich.

Welding Journal Now Publishing Direct Object Identifier (DOI)Numbers

Dear members of the welding research community,

Note that with this issue of the Welding Journal Research Supplement, we are includ-ing Direct Object Identifier (DOI) numbers with each of the papers published inprint and online. (We have also backnumbered the papers published in the Januaryand February 2018 issues.) A direct object identifier is a unique alphanumeric stringassigned by a registration agency (we are using Crossref.org) to identify content andprovide a persistent link to its location on the Internet. Our decision to beginassigning a DOI for each paper comes directly from a request by the researchcommunity.

As part of our obligation to Crossref.org, we are asked to provide DOI numbers, whenavailable, in the references section of papers. So, if you have submitted a paper tothe Welding Journal or are planning on submitting a paper, we ask that you updateyour references to include DOI numbers whenever possible.

Thank you.

Mary Ruth Johnsen, Publisher, Welding Journal


Online Inspection of Weld Quality in Ultrasonic Welding of ...€¦ · the shop floor, the tests...

Documents

Transcript of Online Inspection of Weld Quality in Ultrasonic Welding of ...€¦ · the shop floor, the tests...