On Optimization of Some Parameters in Ultrasonic Metal Welding
Understanding of the weld forming process leads to a way of optimizing some parameters in ultrasonic welding
BYU. I.CHANG AND J. FRISCH
ABSTRACT. The fundamental bond ing mechanisms of ultrasonic welding are discussed and two basic Nomenclature bond forming processes are sug gested. An optimum weld condition for the electric power inputs was for
3 mulated using elastic and plastic anal A Tangential displacement of the sonotrodetip vibration 10 in. ysis and a semianalytic expression 3 Aop Optimum tangential displacement of the sonotrode tip vibration, 10~ for the optimum electric power input was thus obtained. The validity of the Radius of the contact area, in. expression was checked by a series Radius of the inner boundary ot a slip annulus, in. of experiments with 2024-T351 aluminum and OFHC copper. An Diameter of a sphere, or diameter of the disk specimen, in. expression for the coefficient of fric Young's modulus, psi tion for alternating tangential motion Constant was also obtained from this analysis. Shear modulus, psi The strength characteristics of the Normal load or weld clamping force, lb welds are compared for different test Electric power input to the transducer, watts conditions and failure modes. Optimum electric power input, watts Normal stress at the contact area, psi Introduction 2 P Maximum normal stress at the contact area, psi PmM = 3/N2 IT a max Ultrasonic welding is a method of S Tangential shear force, lb joining similar or dissimilar metals by t Ultrasonic pulse time or weld time, sec applying high frequency shear vibra tion and normal pressure to the weld Xs Minimum sonotrode tip displacement required for a fully developed 3 interface. The mechanical energy slip annulus, 10 in. transmitted to the weld area pro Xp Maximum allowable displacement for sublayer plastic deformation 3 duces a sound metallurgical bond be when an optimum weld is produced, 1 0" in. tween two metals. The major advan X A small sonotrode tip displacement by which a slip annulus is pro tage of this welding technique over duced, 10 3 in. conventional fusion joining processes " Poisson's ratio, dimensionless is low heat input at the weld. Thin
ry Shear yield stress, psi T Shear stress, psi U. I. CHANG is associated with the Weld y Maximum allowable shear strain for sublayer plastic deformation at ing Development Department, Manufac turing Development Office of the Ford the weld, % Motor Company. J. FRISCH is Professor of p. Coefficient of friction when an oscillating tangential vibration is Mechanical Engineering, University of applied to the contact between an elastic sphere and an elastic flat, California, Berkeley, California. dimensionless Paper was presented at the 54th A WS Annual Meeting held in Chicago during Apri/2-6, 1973
24-s | JANUARY 1974 foils or wires can be welded to thick cepted mechanism is solid state bond
members and the unfavorable ing some investigators have sug SONOTRODE TIP changes in material properties due to gested that ultrasonic welding is heat at or around the weld are less another form of fusion welding acti significant. Also, shrinkage and dis vated by the heat generated through tortion problems are absent from this friction and plastic deformation, or at welding technique. This joining least it is a strongly heat-assisted process is utilized in spot-type weld welding process (Refs. 8-11). ing, ring welding, line welding, and The localized temperature rise at continuous-seam welding (Ref. 1). the weld in ultrasonic welding is due For spot-type welding, the var to the combined effect of elastic hys iables under control of the welding teresis, localized interfacial slip, and a. COMMERCIAL TYPE WELDING CONFIGURATION machine operator are tip radius, nor plastic deformation. Temperature < • mal load, electric power input to the measurements made with materials UPPER DISK transducer and weld time. Proper ad covering a wide range of melting justment of these variables is essen temperatures show that the maximum tial to minimize the required energy average interface temperature when and optimize weld quality for any good welds are produced ranges material combination. The vibratory from 35 to 50% of the absolute melt frequency, which may range from ing temperature of the material, sug 10,000 to 175,000 Hz, is determined gesting no melting in the weld zone by the welding machine design and is (Ref. 1). These observations strongly not believed to be critical (Ref. 1) in support the solid state bonding mech I I I II I i I 1 I I I I 1 I I I I I II I I I I ultrasonic welding. anism. b. EXPERIMENTAL TEST CONFIGURATION The basic mechanism by which ultrasonic welds are produced is be Adhesion, one of the solid state Fig. 1 — Ultrasonic welding configurations lieved to be solid-state bonding (Refs. bonding mechanisms, requires an 1 -7). Static normal pressure and oscil intimate contact of the interface. The lating shear stresses at the weld presence of surface films is detrimen interface result in localized slip at the tal to achieve atomically close contact weld interface and plastic deforma of two metal surfaces. Surface films, tion in a thin sublayer enveloping the especially oxide films, should either interface. This process breaks up con be removed or broken in such a way al. CONTACTING SURFACES DUE TO NORMAL LOAD. taminant films and produces an area that clean metals be in contact. Bond of metal-to-metal contact. Even strength then depends upon the though this joining technique is areas where metal-to-metal contact widely used in industry, the basic is achieved. In ultrasonic welding of UN-BONDED SURFACE theory for this welding process is not metals, the breaking of contaminant completely understood because of the films for intimate metal-to-metal con complexity involved in the formation tact is accomplished by the combined bl. BONDED AREA DUE TO SLIP. of welds. Good welding practice often alternating shear stresses at or relies on a trial and error method around the weld, which result from even though some variables have an the normal load and oscillating tan experimental equation for guidance. gential force. - BROKEN SURFACE FILMS The objective of this study ic to The relative tangential displace develop clearer understanding of the ment between a pair of contacting bonding mechanism and processes, bodies can cause localized interfacial cl. FULLY GROWN BONDED AREA BY SUBLAYER and to formulate an expression of var slip and sublayer plastic deformation PLASTIC DEFORMATION. iables for the optimum welding condi around the interface if no gross slid Fig. 2 — Evolution of bonded area due to tions. In order to eliminate the com ing is assumed. Here "sliding" refers ultrasonic vibration plexity involved in a multi-interface to the uniform movement or displace problem found in commercial ultra ment of one contacting surface over sonic welding, the sonotrode tip in a another while "slip" is used for local conventional welding system was, for ized tangential displacement at the contacting surface. Gross sliding can experimental purposes, replaced by a friction when an oscillating tangen occur when the relative displacement disk specimen with a spherical radius tial force is applied to the weld, plas is large enough or the frictional force as shown in Fig. 1. This disk spec tic deformation starts when the is small enough to slide. imen was ultrasonically welded to a following condition is met: flat block specimen which was tightly In ultrasonic welding, both local fixed to a massive anvil. Such a con ized slip and sublayer plastic deforma 2 /xN = k(area) Ty=k(7ra ) TV (D figuration enabled direct application tion are desirable. The interfacial slip of vibratory energy to the weld inter breaks up surface films allowing This plastic deformation causes face, and simplified the analysis of metal-to-metal contact at higher further breakup of the contaminant weld formation. asperities and subsequently a large films, and displacement of broken number of small bonded areas are particles in the vicinity of the inter formed over the entire contacting face in a random manner. Figure 2 Mechanism of Ultrasonic Welding area (Ref. 4). illustrates this process. Ultrasonic welding of metals con The plastic deformation in the sub The oscillating plastic shear can sists of interrelated, complex pro layer enveloping the interface can induce metal circulation across the cesses such as plastic deformation, occur when the relative displacement interface and produce a severely work-hardening, breaking of contam is larger than that necessary to cause work hardened weld, which some inant films, fatigue, crack formation slip and the frictional stress is higher times self anneals to a very fine grain and propagation, fracture, generation than the flow stress of sublayer mate structure. This phenomena has been of heat by friction and plastic defor rial. If the frictional stress is lower observed in 2024-T351 aluminum mation, recrystallization, and inter than the flow stress, gross sliding will welds (Ref. 1). Plastic deformation in diffusion. Although the generally ac occur. Defining p. as the coefficient of the sublayer is the primary process
WELDING RESEARCH SUPPLEMENT! 25-s Optimum Sonotrode Tip Displacement
/ - INO SLIP ASSUMED) If an elastic sphere is pressed against an elastic flat with a normal (SLIP ASSUMED) fin load N, the normal stress distribution over the contact surface p can be represented (Ref. 14) by the ordinates of a hemisphere of radius "a" con structed on the surface of contact as shown in Fig. 4. If "a" is the radius of the circular contact surface, the F 2 63 - 18.1 A - normal stress p at any distance r is UN-SLIPPED REGION J8 given by
2 (2) P = P max Hi-) }" LIP ANNULUS Mindlin (Ref. 15) studied the effects of an oscillating tangential force on 0.2 0.4 0.6 0.8 1.0 1.4 2.0 4.0 6.0 the contact surface of two elastic TIP DISPLACEMENT — AI10"3INCHESI Fig. 4 — (a) Distribution of normal stress and shear stress over contact area be spheres. The assumption that slip Fig. 3 — Electrical power vs. displacement tween an elastic sphere and an elastic flat cannot occur at any part of the con of the sonotrode tip vibration for the ultra with normal load N and tangential force S. tact surface is untenable because it sonic welding system (b) slip annulus required the shear stress rising to in finity at the boundary of contact as shown in Fig. 4. Hence, slip was expected to begin at the boundary for high bond strength. The most to plastically deform the sublayer, a and progress inward with increasing plausible bonding mechanism during condition imposed by Eq. 1. The tangential shear force S , under the the plastic deformation stage is solid normal load range varies with power assumption that shear stress cannot state bonding such as adhesion, capacity of the machine, and desir exceed the product of constant coeffi mechanical interlocking of the sur able normal loads for any specific cient of friction and normal stress faces, recrystallization, and diffusion. welding application depend on the given by Eq. 2. As shown in Fig. 4, In summary, the dominating mech thickness and hardness of the com this slip region has an annular shape anism for ultrasonic welding is solid ponents (Ref. 1). called slip annulus. The surface is plastically deformed by slip and the state bonding, and it is accomplished Energy requirements to produce a original surface topography is dam by two different processes: slip and good weld vary with, the properties plastic deformation. The bond forma and thickness of materials being aged (Ref. 16). tion is mainly attributed to the latter welded. Insufficient power usually The slip annulus grows with in stage. results in low weld strengths due to creased value of S as can be seen insufficiently grown bonded area; from the following equation by Weld Strength and however, excessive high power dete Mindlin (Ref. 15). Welding Variables riorates the weld integrity by either introducing cracks or shearing the b = a(1--S_)V3 (3) Weld strength is the breaking load already formed weld. The degree of necessary to separate the weld, and if plastic deformation at the weld is the weld is separated by a force determined by the electric power The slip annulus starts along the acting normal to the weld interface, input and the weld strength will be boundary of contact area and grows this breaking load is known as tensile optimum for a specific value of elec inwards until it covers the whole area tric power input when other variables weld strength. If separation results of contact as S varies from O to u. N; remain constant. The electric power from a shear force acting parallel to i.e., S/MN varies from 0 to 1. Mindlin input is generally related to the vibra the weld interface, one speaks of obtained another equation, valid until shear weld strength. It is known that tional displacement of the sonotrode tip as shown in Fig. 3. Once this rela the contact surfaces start to slide weld strength is deteriorated by over each other, which relates the excessive power during ultrasonic tionship is found for a particular weld ing system, the conversion of power shear force S to the tangential dis welding. This is attributed to fatigue placement X of the top sphere. cracks developed at the interface to tip displacement is possible. The sonotrode tip displacement is more (Refs. 4, 5, 12) or excessive damage 3(2 v) S ,2/3 of the weldment by overstressing meaningful than electric power input - ,,N|I-(I (Ref. 13). Many investigators found delivered to the transducer in ana 8Ga ' /*N lyzing plastic deformation at the (4) cracking occurred in ultrasonic welds weld. The optimum tip displacements (Refs. 3, 4, 9, 10). An optimum ultra are those which allow optimum weld Combining Eqs. 3 and 4, it can be sonic weld can be made when suffi strengths when other variables are seen that the slip annulus grows with cient plastic deformation occurs with constant. increasing tangential displacement X, out serious weld deterioration by as shown in Eq. 3-a. fatigue cracking or overstressing. The effect of weld time on weld The weld strength in spot-type strength is important since prolonged 8Ga welding is affected by the setting of weld times result in fatigue damage b = a (3-a) such variables as tip radius, normal (Ref. 4), and insufficient weld times {1 3( 2-v)ju,N* r load, electric power input and weld show incomplete welds. The min time. Since a larger tip radius in imum time required appears to vary Mindlin assumed that the two creases the weld area, other var with the thickness of the workpiece spheres had the same radius and the iables are adjusted to allow proper which is in contact with sonotrode same elastic properties; however, his plastic deformation of the weld. The tips. Practical range is about 0.005 results were not seriously affected by normal load, combined with the tan sec for very fine wires to about 1.2 the radii of the spheres for a given gential force, should be high enough sec for heavier materials (Ref. 1). size of contact area (Ref. 5). Substi-
26-s I JANUARY 1 974 tuting Eq. 3 into Eq. 4 gives the follow where the plastic deformation can FATIGUE CRACK OR ing expression: DAMAGED WELD BY start first by additional tangential OVERSTRESSING shear stresses. 3(2-v) „U {l-<|)2} The theoretical plastic zone at the 8Ga (5) interface has a lenticular shape with center thickness a and diameter 2a as- Therefore, the tip displacement for a shown in Fig. 6. This may be approx fully developed slip annulus Xs is imated by a disk with a uniform thick ness a-2 and a diameter of 2a. The 3(2 -v) lenticular shape of the plastic zone is TIP DISPLACEMENT Xs" 8Ga ^N (6) proved by a weld bead (Ref. 18) pro duced at the weld interface of 17-4 This equation can also be obtained by PH stainless steel as shown in Fig substituting the condition S/(/aN) = 1 6-c. Since the radius of contact area intoEq. 4. is, as given by Hertz (Ref. 19), From Eqs. 5 and 6, a = 0.88 VfNd/E) (9)
X = A,,, TIP DISPLACEMENT s 1-fb/a)2 (7) the uniform thickness of the plastic Fig. 5 — Variation of bonding process and This expression predicts the tip dis layer tp can be assumed as expected tensile weld strength placement Xs required to have a fully developed slip annulus without tpssa/2 = 0.44 ^(Nd/E) • knowledge of the value of p. (10) < The ratio b/a can be directly mea 4-2a-* l/2a sured from the damaged slip annulus for a sphere and a flat contact. If we produced by tangential vibration with further assume that this disk with i a small tip displacement. If the tip dis uniform thickness tp has uniform ' -—•—' " placement is larger than X , either s plastic deformation by shear, the sliding at the interface or plastic de a). THEORETICAL PLASTIC ZONE AT THE INTERFACE maximum allowable plastic displace formation at the sublayer can occur. ment by this layer X can be ex Sliding at the interface will occur if P pressed as p.N is lower than the tangential force needed to plastically deform the sub layer. \i p.N is higher than that, the Xp = aY/2 = 0.44 yV^NdTi) sublayer will plastically deform, and (11) excessive displacement will be ac bl. IDEALIZED PLASTIC ZONE AT THE INTERFACE commodated by sublayer plastic where 7 is the maximum allowable deformation. shear strain of the sublayer without 9x10"3in. Since both maximum slip area at deteriorating the weld integrity by = 2a+A the contact surface and proper either fatigue cracking or over- degree of plastic deformation in the stressing. • 2.5x10'3in. sublayer are desirable for the opti Now, the expression for the optimum mum weld strength, the optimum tip tip displacement A oP is. ol. WELD BEAD OBSERVED AT THE INTERFACE 1201 displacement Aop will have the MATERIAL : 17-4 STAINLESS STEEL, NORMAL LOAD : 2.5 LBS. ELECTRIC POWER INPUT : 60 WATTS, expression Aop -Xs + Xp- WELD TIME : 0.2 SEC. , ENVIRONMENT : AIR 3 3 A p -X + X a - 3.54x10' in., A - 1.77x10" in. 0 s p (8) + 0.44 y -v'fNd/E) 1 -(b/a)2 Fig. 6 — Plastic zone at the interface where Xp is the plastic displacement (12) of the sublayer. As tip displacement increases, different bonding pro Another interesting equation is the cesses predominate as shown in Fig. expression of p. obtained by substitut 5. Also shown is the theoretical tan ing Eq. 9, shear modulus gential force and the expected tensile G = E/ 2(1+v)andv=0.3intoEq,6. weld strength as a function of tip dis gage (PPG) monitored the system placement. Appreciable bond pressure in the high vacuum range. M = 0.53(E2d/N2)1/3 x (13) strength starts to form at Xs , and the s The seal for linear motion and linear- maximum occurs at Aop. The decreas rotational motion through the vacuum ing slope of tensile weld strength Experimental Equipment chamber was provided by stainless after Aop depends on the susceptibil steel bellows. ity of the material to fatigue or over- The experimental equipments, The compressive and tensile loads stressing. shown in Fig. 7, consist of the vac were supplied by a lever system in The stress distributions at the con uum apparatus, loading and measur stalled outside of the chamber. The tact area and along the center line ing devices, and ultrasonic instrumen normal load strain gage dynamom joining the centers of spheres are tation. Pumping in the low vacuum eter was a 4-arm bending beam to known for the case of two elastic region was accomplished by a minimize anvil rotation. The anvil spheres pressed together (Refs. 14, mechanical pump of 10 cubic with specimen clamp was attached to 17). According to the analysis (Ref. feet/minute capacity, or two double the ends of the 4 arms as shown in 14), the maximum shear stress is Varian VacSorb pumps, as shown in Fig. 8. Four strain gages formed a developed on the center line at a Fig. 7. High vacuum pumping was per bridge circuit to measure the com depth of about one half the radius of formed by a 75 liter/sec ion pump. A pressive normal load N on the spec the contact area. The magnitude is standard thermocouple vacuum imens during welding and the tensile about 0.31 pma)< for v = 0.3 where pmax gauge was used during rough pump force required to separate the spec is the maximum pressure developed ing and a Bayard-Alpert type nude imens after welding. The minimum at the contact area. This point is ionization gage and a partial pressure sensitivity of the dynamometer was
WELDING RESEARCH SUPPLEMENT | 27-s Fig. 7 — Experimental apparatus
designed for a range from 0.2 to 200 lb and torque dynamometer for a rangefrom 0.2 to 200 lb in. The ultrasonic system, a schematic block diagram of which is shown in Fig. 9, was used to generate high intensity ultrasonic vibrations at the specimen interface. The ultrasonic pulse length was controlled by a elec tronic decade interval timer, and the 1 50 watt J-1 7* power supply incorpo rated with an automatic frequency control was used. Power was re corded with a WAJ2* watt meter. The 20 kHz output of the power supply was then fed into an ultrasonic transducer, which utilized a lead titan- ate zirconate crystal to convert the electrical output of the power supply into longitudinal mechanical vibra tions with a 20 kHz frequency. The coupling horn attached to the trans ducer provided a convenient location Fig. 8 — Specimen test assembly for tensile weld strength test at which the vacuum feed-through could be made. For this particular study, an exponential amplifying horn was selected (Ref. 20), and the partic 0.1 lb and the maximum allowable stainless steel bellows, and a hydrau ular dimensions yielded an amplifica load was 135 lb (Ref. 19). Figure 8 lic load applying system (Ref. 18). tion ratio of approximately 3. A typical shows the installed dynamometer This mechanism enabled the anvil to transducer displacement for full and anvil assembly with specimens have linear vertical motion of 1 in. power operation was 0.0007 in., yield in position. and independently 45 deg rotational ing a displacement of approximately To investigate the shear strength motion with excellent vacuum seal 0.002 in. at the tip of an exponential characteristics of the ultrasonically ing. Both dynamometers were placed amplifying horn. welded joints, a different loading and inside the vacuum chamber close to unloading system was used. This the anvil for more accurate measure loading system is composed of an ments. Each dynamometer has four anvil, normal load and a torque dyna strain gages to form a bridge circuit. mometer, a torque box with three The normal load dynamometer was *Branson Sonic Power Co.
28-s I JANUARY 1 974 Experimental Technique WATTMETER POWER SUPPLY INTERVAL BRANSON 20 KHz BRANSON TIMER 115 V For the two materials used in this SONIC SONIC LECTRA LABS. 60 cps investigation, the specimens shown MODEL WAJ2 MODEL J17-V MODEL TM-8 in Fig. 10 were made from 1 in. bar stock of 2024-T351 aluminum alloy SANBORN RECORDER ION GAUGE PPG. and 1 in. OFHC copper plate. The surface finish was 8 micro in. RMS 3\Z CARRIER AMPL and acetone was used for degreasing AMPLIFYING HORN TYPE 3C66 before each test. DISK The disk specimen was attached to the tip of the amplifying horn and the OSCILLOSCOPE block specimen was clamped in the TYPE 565 anvil of the dynamometer block, as DUAL BEAM shown in Fig. 8. The dynamometer assembly was then raised for contact I between the disk and block spec TRANSDUCE7 R SANBORN imen. Since the disk specimen was COUPLING HORN RECORDER the central portion of a 3/4 in. diam sphere and the block specimen had a flat surface, they modeled a sphere 75 L/S LOADING VAC SORP PUMP ION PUMP MECHANISM on flat contact configuration. The timer shown in Fig. 7 controlled the ultrasonic pulse durations of 0.2 and Fig. 9 — Schematic diagram of ultrasonic welding system for tensile weld strength tests 1.0 sec used in the experiments. As a relative measure of the ultrasonic power at the tip of the amplifying 5 6 horn, the electrical power input P to vacuum from 5 x 10 to 7 x 10 torr, the ultrasonic transducer was re with an electric power input up to 140 corded with a watt meter. watts. To find the relationship between The tensile weld strengths were electrical power input and the ampli measured for experiments of various tude of the ultrasonic vibration of the electric power input levels with disk specimen, one edge of the alumi preset normal contacting loads and num disk specimen on the horn tip ultrasonic pulse durations. Usually a was ground flat and polished. Move series of tests was done at different ment of the square edge of the flat locations of one disk and one block surface was observed with an ordi specimen to minimize Variations in nary microscope with a graduated parameters such as surface rough Fig. 10 — Ultrasonic welding specimens eyepiece. The ultrasonic vibration of ness and cleanliness. The same test the specimen extended the image of procedure was used in the atmo the edge to a value of one amplitude. spheric environment as well as a vac By measuring these extended regions sonically welded. From Eqs. 14 and uum environment in the range of 2 x 7 of the specimen image, the following 12, an expression for optimum elec 10 to4x 10 ? torr" relationship was obtained for the ex tric power input is obtained for this The shear strength tests for ultra- perimental configuration of this ultra ultrasonic welding investigation. sonically welded OFHC copper were sonic welding investigation. done in the previously described 18.1 apparatus (Ref. 21). The disk spec (b/a)2 imen was again attached to the ampli P = 18.1 A (14) [r fying horn tip and the block specimen 0.44 7
WELDING RESEARCH SUPPLEMENT! 29-s Table 1 — Calculated and Observed Values of Optimum Sonotrode Tip Displacement for 2024-T351 Al Welding, in Air
Xs Xs Xp Aop A0p Ampl., cal., obs.. a/2 cal., cal.. •op obs., 2 N. t, Power, 10-3 10 3 103 io- 10 3 10 3 obs., 10"3 lb sec w in. b/a in. in. in.lal in.(b) in. W in.
30 1 15 0.94 0.568 1.38 >1.21 0.566 0.510 1.89 90 1.84 30 0.2 15 0.94 0.54 1.32 >1.21 0.566 0.510 1.83 105 1.93 20 1 15 0.94 0.316 1.05 <1.21 0.494 0.445 1.50 45 1.41 20 0.2 — — — — — — — — 60 1.58 10 1 8 0.73 0.52 1.00 <1.21 0.392 0.353 1.35 30 1.20 >0.94 10 0.2 — — — — — — — — 45 1.40 5 1 4 0.57 0.48 0.74 — 0.311 0.279 1.02 15 0.94 5 0.2 — — — — — — — — 30 1.20 2.5 1 4 0.57 1.00 0.57 — 0.247 0.222 0.79 15 0.94 2.5 0.2 — — — — — — — — 30 1.20
6 (a) Material Properties of 2024-T351 aluminum (Ref. 22): E = 10.6 x 10 Psi., elongation (in 2 in., 1/2 in. diam specimen) = 1 9% (typical). (b) y - 9% was used for calculation.
3 a). NORMAL LOAD : 2.5. LBS. j15_|- b). NORMAL LOAD : 5 LBS. 10 in., were applied. The b/a ratios -B- AIR • 0.2 SEC. WELD TIME E> AIR -0.2 SEC. WELD TIME of the slip annuli were measured 0- AIR 1.0 SEC. WELD TIME O- AIR-1.0SEC.WELD TIME from the broken welds with a micro scope and the values of Xs were cal culated using Eq. 7. Calculated values are listed in Table 1 and agree with the observed values of Xs. Some of the slip annuli are shown in Fig. 15. © Q i 9 |fl| ? i i iT" j i For comparison, fully developed slip 20 40 60 80 100 120 annuli made under different normal 3 20 40 60 80 100 120 ELECTRIC POWER INPUT P (WATTS) ELECTRIC POWER INPUT - P (WATTS) load or power input are shown in Fig. 16. C). NORMAL LOAD : 5 LBS. d). NORMAL LOAD : 10 LBS. Calculations for the plastic dis 0.2 SEC. WELD TIME D- AIR 0.2 SEC. WELD TIME 1.0 SEC. WELD TIME _g_ AIR 1.0 SEC. WELD TIME placement Xp require a value for the allowable shear strain at the weldY . In general, ductile materials should have higher Y than brittle materials. The analytical prediction of Y is diffi cult, if possible, because of high
I I I I I I I I I strain rates involved in a high fre 0 20 40 ) 20 40 60 80 100 120 ELECTRIC POWER INPUT ELECTRIC POWER INPUT P (WATTS) quency vibration, temperature rise, fatigue problems, size effect at the e). NORMAL LOAD : 20 LBS. weld, and geometric constraint im 0 Q- AIR • 0.2 SEC. WELD TIME (TENSILE) f). NORMAL LOAD : 30 LBS. posed by surrounding material. There 6 G- AIR • 1.0 SEC, WELD TIME (TENSILE) O- AIR-0.2 SEC. WELD TIME £r- AIR -1.0 SEC. WELD TIME (SHEAR) O- AIR -1.0 SEC. WELD TIME fore, this value has been obtained do •©•• VAC. 1.0 SEC. WELD TIME (TENSILE) from one set of data (N = 30 lb) and is I I 40- - then used for the rest of the calcula tions. The computed value of Y was found to be equal to approximately Q a m u 20 one-half the typical tensile elonga 5 S tion (Ref. 22). One half of the radius UJ CC £ 30- - of contact area was calculated for different normal loads and then multi
I I I l""l | I I I I I plied by a constant value of Y to obtain 20 40 60 80 100 120 20 40 60 80 100 120 Xp. The optimum tip displacement ELECTRIC POWER INPUT P (WATTS) ELECTRIC POWER INPUT • P (WATTS) Aop was obtained by adding XP to Xs Fig. 11 — Tensile weld strength or shear weld strength vs. electric power input for 2024- as shown in Table 1. The calculated T351 aluminum values of Aop's are plotted in Fig. 14 for comparison with observed data. Equation 12 was also used to calcu late Aop for OFHC copper welding. mum weld strengths, plotted as func the relationship shown in Fig. 3, and Observed optimum power inputs for tions of normal load, are shown in these optimum tip displacements, as OFHC copper with 5 lb normal load Fig. 12. Curves of optimum electric functions of normal load, are shown were 25 W for 1.0 sec weld time and power input versus normal load are in Fig. 14. 35 W for 0.2 sec weld time (Ref. 19). shown in Fig. 13 which shows that Experiments were performed to ob A calculated value of Aop from Eq. 12 the optimum normal load increases tain slip annuli for different normal gave 1.05 * 10 3 in. for the optimum with increased electric power input. loads. To produce such slip annuli, tip displacement. Again, the value of These power values have been con electric power inputs ranging from 4 Y was taken as one-half of the re verted to corresponding tip displace to 1 5 W, which correspond to tip dis ported elongation value (Ref. 23). Cal ments of the disk specimen by using placements of 0.57 x 10 to 0.94 x culations are listed in Table 2.
30-s I JANUARY 1 974 Table 2 — Calct lated and Observed Values of Optimum Sonotrode Tip Displacement for OFHC Copper Welding in Air
A0p "op Ampl., Xs cat, a/2 Xpcal., cal., "op obs., 3 3 N, Power, 10-3 IO" i0-2 10-3 IO" t, obs., 10 3 lbs W in. b/a in. in.(a) in."" in. sec W in.
5 2 0.47 0.316 0.52 0.266 0.53 1.05 0.2 35 1.10 1.0 25 0.97 20 8 0.73 1.00 0.73 0.422 0.844 1.57 1.0 60 1.58
(a) For OFHC copper (Ref. 23): E = 1 7 * 10fipsi.; elongation (in 2 in.) = 45 to 35%. (b) y = 20% was used.
EXPERIMENTAL VALUES OF OPTIMUM TIP DISPLACEMENT 120- - Q 02 SEC. WELD TIME MATERIAL : 2024-T351 ALUMINUM -O 1JJ SEC.WELD TIME 110- - MATERIAL : 20Z4-T351 ALUMINUM WELD TIME : 0.2 SEC. H 1.0 SEC. O WELD TIME : 0.2 SEC. • 100- - 1.0 SEC. O — ENVIRONMENT : AIR ENVIRONMENT I AIR 90- -
THEORETICAL VALUES OF 0.5- \ JI OPTIMUM TIP DISPLACEMENT, Aop- I I x TIP DISPLACEMENT FUR FULLY DEVELOPED SLIP ANNULUS, Xs. et-ef / I I III H 1—M- 5 6 7 8 910 20 30 40 50 NORMAL LOAD • N (LBS.) NORMAL LOAD • N (LBS.) NORMAL LOAD • N (LBS.)
Fig. 12 — Optimum tensile weld strength Fig. 13 — Optimum power as a function of Fig. 14 — Theoretical and experimental as a function of normal load for 2024- normal load in tensile weld strength tests values of the optimum sonotrode tip dis T351 aluminum in air placement for tensile weld strength (2024- T351 Al, in air)
The values of p. were calculated by A„ = 0.58N (8-b) shows the lenticular shape of the sub using Eq. 13 and are plotted for differ layer plastic zone and both sublayer ent applied normal loads in Fig. 17. Shear stresses developed at the and peripheral cracks. An interesting These values of p. were obtained by final weld area by friction forces were phenomenon observed at the OFHC substituting Xs as listed in Table 1 calculated for the optimum welds and copper weld interface is the recrystal into Eq. 13, and are found to be ex are shown in Table 3. Since the exact lization occurring at the weld. ceedingly high compared to the ordi measurement of the final weld area is Severely deformed sublayer partially nary coefficient of friction for de difficult, broken weld areas were self annealed to a fine grain structure greased 2024-T351 aluminum. Fur used for the calculation, neglecting isshown in Fig. 1 9-c, demonstrating thermore, the curve shows that p. is the work hardening effect. The cal that recrystallization is an auxiliary not a constant but a function of culation showed that the shear bonding mechanism for ultrasonically normal load N. An experimental equa stresses developed at the final weld welded OFHC copper. tion of p. as a function of the normal area by the friction force are fairly Many of the broken welds of 2024- load for 2024-T351 aluminum was constant and are about 2.2 to 2.8 T351 aluminum revealed grooves obtained from the curve shown in times higher than the shear yield around the periphery at the leading Figure 17. strength of the material. These high and trailing edges of the weld as shear stresses for the plastic de shown in Fig. 20. These grooves are i formation in the sublayer are due to attributed to the peripheral cracks / = 10(N)- (16) the geometric constraints imposed by already present after ultrasonic the surrounding materi-al and the Another expression of X for 2024- welding. The leading and trailing s high strain rates developed during T351 aluminum has been obtained by edges of the welds are most suscep ultrasonic welding. The same geo substituting Eq. 1 6 into Eq. 6. tible for peripheral cracks. The pres metric constraint can be found in the ence of peripheral cracks are shown case of indenters pressed into flat ma clearly in Fig. 19-b for a sectioned 2 1/3 XS = 18.8(N/E d) (6-a) terial (Ref. 24). Similarly, the shear specimen of ultrasonically welded stresses calculated in Table 3 may OFHC copper. Some of the broken From Eqs. 8, 6-a, and 10 well represent the stress required to welds showed that gross sliding oc plastically deform the constrained curred at the weld, probably because 2 1/3 Aop =[18.8 WELDING RESEARCH SUPPLEMENT! 31-s a). WELD TIME 1.0 SEC. WELD TIME : 1.0 SEC. NORMAL LOAD: 5 LBS. NORMAL LOAD: 2.5 LBS. ELECTRIC ELECTRIC POWER INPUT 4 WATTS POWER INPUT : 4 WATTS (X50) (X50) b). b). WELD TIME 1.0 SEC. WELD TIME : 1.0 SEC. NORMAL LOAD: 10 LBS. NORMAL LOAD : 10 LBS. ELECTRIC ELECTRIC POWER INPUT 15 WATTS POWER INPUT : 8 WATTS (X50) (X35) c). c). WELD TIME : 0.2 SEC. WELD TIME : 0.2 SEC. NORMAL LOAD : 30 LBS. NORMAL LOAD: 30 LBS. ELECTRIC ELECTRIC POWER INPUT : 30 WATTS POWER INPUT : 15 WATTS (X35) (X35) Fig. 15 — Micrographs of broken welds showing slip annuli (2024-T351 Al.inair) Fig. 16 — Micrographs of broken welds showing fully developed slip annuli(2024-T351 At, in air) 10.0- power range it increases first, then 9.0- decreases (Ref.21 (High normal loads 8.0- are detrimental to weld strength for 7.0- low power welding, because corres pondingly small amplitude of vibra 6.0- tion cannot produce a fully developed slip annulus or proper plastic defor 5.0- mation at the weld interface. The effect of weld time on tensile 4.0- weld strength was checked for two different weld durations. The weld 3.0- time of 0.2 sec gave higher tensile weld strength, than 1.0 sec weld time, but the time effect is less signif 2.0- icant for higher normal loads. The establishment of an optimum weld 1.5 time is important to increase the production rate of reliable products. 1.5 I I 4I 5I 61 7I 8II 9 I1 0 II 20 30 II40 I50 A vacuum environment of 2 * 10 7 NORMAL LOAD- N (LBS.) to 4 x 10 7torr does not provide a sig nificant improvement in tensile weld Fig. 17 — Coefficient of friction when an oscillating tangential vibration is applied (2024- strength for 2024-T351 aluminum T351 A I, in air) with 5 lb normal load. However, higher tensile weld strengths were obtained in the same vacuum environ ment for 20 lb normal loads, especial ly when higher power inputs were Table 3 — Computed Shear Stresses by Friction Force Developed at Optimum Welds" used, as shown in Fig. 11-e. Also, the optimum power input moved from 45 W in air to 90 W in vacuum. In all vac Broken weld area, Average broken 4 uum tests the broken oxide particles, N, JLl(b) 10 in? weld area, weld area lb t = 0.2 sec t = 1.0 sec 10-"in.2 103 psi commonly found around the weld formed in air, were absent. 30 3.3 15.2 15.0 15.1 65.5 The changes in appearance of the 20 3.3 10.9 10.7 10.8 61.1 broken weld are generally related to 10 5.1 8.4 6.2 7.3 69.9 5 5.9 4.2 3.2 2:7 79.7 the electric power input and normal 25 7.3 3.1 2.3 2.7 67.5 load used. As power increased from 2 to 140 W, the following order of (a) Material: 2024-T351 aluminum; environment: air. changes was common. At low (b) See fig. 17. powers, a slip annulus was produced. 32-s I JANUARY 1 974 Then a fully developed slip annulus _ 3.0- was observed, which grew to an ellip tical weld area having shallow grooves around the periphery at the b C 2.0. leading and trailing edges of the a. weld. These grooves grew deeper and WELDING RESEARCH SUPPLEMENT! 33-s EFFECTIVE WELD AREA stresses of the optimum welds are FOR BOTH TENSILE also constant. AND SHEAR 9. The values of coefficient of fric STRENGTH tion p.for 2024-T351 aluminum when an oscillating tangential vibration is applied at the contact between an elastic sphere and an elastic flat of a). AT LOW POWERS. the same material are higher than that of sliding friction and are a func Fig. 20 — Broken welds of 2024-T351 tion of normal load. EFFECTIVE WELD AREA aluminum after tensile test in air showing FOR TENSILE STRENGTH 10. A vacuum environment of 2 x peripheral grooves. Normal load, 30 lb: 10 7 to 4 x 10 7 torr does not signif weld time 0.2 sec: (left) 60 W; (right) 75 W. icantly improve the weld strength of (X20, reduced 41%) 2024-T351 aluminum. • PERIPHERAL CRACK 11. Due to the presence of SUBLAYER CRACK EFFECTIVE sublayer cracks and peripheral cracks WELD AREA at high power levels, ultrasonic welds FOR SHEAR STRENGTH are more reliable for shear loads than b). AT HIGH POWERS. tensile loads. Fig. 22 — Effective weld area for tensile strength and shear strength of the weld References 1. "Ultrasonic Welding," Chapter 49, Welding Handbook, 5th Edition. American Welding Society, Section III, "Welding, Cutting and Related Processes." 35-i A N=2.5 LBS -O N=20 LBS. 2. Daniels, H. P. C, "Ultrasonic Weld • N=5 LBS., -H N=30 LBS. ing," Ultrasonics, October-December ^ 1965, pp. 190-196. • N=10 LBS. 3. Mindlin, R. D., Mason, W. P., Osmer, 30- T. J., Deresiewicz, H., "Effects of an Oscil lating Tangential Force on the Contact Surfaces of Elastic Spheres," Proceedings 25- of the First National Congress on Applied Mechanics, 203, 1952. o o 4. Tylecote, R. F„ The Solid Phase o Welding of Metals. New York, St. Martin's 20- Press, 1968. 5. Neppiras, E. A., "Ultrasonic Welding of Metals," Ultrasonics, July-September tr i- 1965, pp. 128-135. 15- 6. Koenigsberger, F., Adair, J. R., C3 2 Welding Technology, Third Edition, St. Martin's Press, New York 1966. < 7. Jones, J. B., Maropis, N., et al, "Phe- LU cc 10- nomenological Consideration in Ultra oo sonic Welding," We/ding Journal, Vol. 40, July 1961, Res. Suppl., pp 289-S-305-S 8. Ainbinder, S. B., "Certain Problems 5- of Ultrasonic Welding," Svarochnoe Proiz- vodstvo (Welding Production), December 1959, Trans, by B.W.R.A. 9. Weare, N. E., Antonevich, J. N., Monroe, R. E., "Fundamental Studies of I I I I I 15 30 45 60 75 90 105 120 Ultrasonic Welding," Welding Journal, ELECTRIC POWER INPUT - P (WATTS) Vol. 30, August 1960, Res. Suppl., pp. 331-s-341-s. Fig. 21 Tensile breaking stresses of 2024-T351 aluminum welds in air. Weld time, 0.2 10. Lewis, W. J., Antonevich, N. J., et sec al, "Fundamental Studies on the Mechan ism of Ultrasonic Welding" Battelle Memorial Institute, December 1960, welding is basically solid state bond grown slip annulus and that for a Wright Air Development Division. ing such as adhesion, mechanical proper sublayer plastic deformation. 11. Cl'shanskii, N. A., "On the Joining interlocking of the surfaces, recrystal 5. For an optimum welding condi of Metals by Ultrasonic Welding," Moscow Institute of Energetics, Avtoma- lization, and possibly diffusion. tion, a higher normal load requires a ticheskaya Svarka. Volume 12, 1961, 2. Bonding is accomplished by two higher electric power input or a larger Trans, by Brutcher. tip displacement. different processes, interfacial slip 12. Weare, N. E., et al, "Research and and sublayer plastic deformation. 6. Excessively high electric power Development of Procedures for Joining Both processes contribute to metal- input or large vibration amplitude Similar and Dissimilar Heat-Resisting to-metal contact. deteriorates weld integrity by intro Alloys by Ultrasonic Welding," February 3. Optimum welds are produced ducing sublayer and peripheral 1959, WADC Tech. Rep. 58-479. Wright when the maximum slip area at the cracks. Air Development Center. interface and proper plastic deforma 7. The vibration amplitude or dis 13. Bell, R. D., "Ultrasonic Seam tion in the sublayer are achieved. The placement at the sonotrode tip is a Welding of Copper Sheet," Tech. Report, No. PRN-59, Princeton, November 4, 1 962 relative displacement between two function of the electric power input 14. Timoshenko, S., Theory of workpieces by ultrasonic vibration is for a particular ultrasonic welding Elasticity, First Edition 9th Impression, important to achieve these conditions. system. McGraw-Hill, 1934. 4. The optimum sonotrode tip 8. The shear stresses developed by 15. Mindlin, R. D., "Compliance of displacement is the linear addition of friction force at the optimum welds Elastic Bodies in Contact," J. of Applied the displacement necessary for a fully are constant. The tensile breaking Mech., Volume 15, 1949, pp. 259. 34-s I JANUARY 1974 16. Bowden, F. D., Tabor, D., The Fric tion and Lubrication of Solids, Oxford Univ. Press, 1950. 17. Shigley, J. E., Mechanical Engineer 100- ing Design, McGraw-Hill Book Co., 1963. 18. Frisch, J., Chang, U., "Ultrasonic z NORMAL LOAD : 20 LBS. Welding of Metals in Vacuum," Final WELD TIME : 1.0 SEC. Report No. MD-69-2, August 1969. 80 © A AIR - SHEAR 19. Timoshenko, S., Strength of Mate rials, Part II, 2nd Edition, August 1941, pp. O AIR - TENSILE 350-356. © VAC. SHEAR 20. Pfaelzer, P. F, Frisch, J., "Ultra 60 sonic Welding of Metals in Vacuum," Final Report No. MD-67-3, December 1967, University of California. 4 40- - 21. Frisch, J., Chang, U., "Optimal Strength of Ultrasonically Bonded Metals in Air and Vacuum," Finai Report No. MD- z 70-2, September 1970. rr 20- - 22. Aluminum Standards and Data, The Aluminum Association, 2nd Edition, cc December 1969. < 23. Materials Engineering — "Mate rials Selector Issue," Mid-Oct. 1966, Vol 10 20 30 40 50 60 70 80 90 100 110 120 130 140 ume 64, No. 5. Reinhold Publishing Co. 24. Rabinowicz, E., Friction and Wear ELECTRIC POWER INPUT - P (WATTS) of Materials, John Wiley and Sons, Inc., N.Y. 1965. Fig. 23 — Shear strength and tensile strength of OFHC copper welds /If You Need to Know Filler Metals, You Need FILLER METAL COMPARISON CHARTS Before publication of the first edition of these Comparison Charts, direct comparison of two proprietary products, or classification of a filler metal by brand name alone, could only be done by examining volumes of data supplied by the 50 to 100 manufacturers and vendors. All this time and money can now be saved. Every significant manufacturer and vendor of filler metals were invited to participate by classifying his brand name desig nations according to AWS specifications for this listing. The result is this single volume of filler metal data that is just not available from any other source, regardless of price. $6.00. 21 FILLER METAL SPECIFICATIONS We bound the complete set of Filler Metal Specifications in a hard cover, to fit right in with the other books in your reference library. The binder was specially designed for us. Each of the 21 Specifications has been perforated and is contained within the binder. Each page can be read easily as the spread-open book lays flat. The binder allows room for expansion. If you buy the 21 Specs separately, the price is $44.50. If you buy all 21 Specifications together with the binder, you pay the same price and you have a bookshelf volume for the cost of the Specs alone. Order these useful and easy-to-use books from the American Welding Society, 2501 NW 7th St., Miami. FL 33125. Don't forget your Membership Discount: 25% for A&B members; 15% for C&D members. WELDING RESEARCH S U P P L E M E N T 35-s STRUCTURAL Incorporates all of the welding requirements for the construction WELDING of buildings, bridges, and CODE tubular structures. Published in September, 1972, the Structural Welding Code, AWS D1.1-72, com bines into a single document, completely updates, and replaces the Code for Welding in Building Construction, AWS D1.0-69, and Specifications for Welded Highway and Railway Bridges, AWS D2.0-69. Also, for the first time anywhere, requirements are pre sented for the design and fabrication of welded tubular structures. These are the major changes affecting the building and bridge requJements which have been incorporated into the Code: (1) the addition of requirements for visual inspec tion for and repair of defects in cut edges of plates as received from the mill, (2) revision of weld quality and inspection requirements to remove ambiguity in previous editions relative to visual and nondestructive examinations, (3) increased tolerances on warp and tilt of girder flanges, and (4) inclusion of revisions issued in April of 1970*, including those to permit use of gas metal-arc (GMAW) and flux cored arc welding (FCAW) with prequalified procedures. Fatigue stresses for use in bridge design have been extended to include all steels used under the bridge portion of the Code. To save time in the use of the Code, there is a complete index, an appendix con taining selected definitions from Terms and Definitions, AWS A3.0-69, plus other welding terms used in the Code, and an appendix for conversions to the metric (SI) sys tem. The Code is three-hole punched to permit insertion in binders if desired and to provide for the inclusion of revisions as issued. Its 8V2 in. X 11 in. size is easier to read and use than the previous 6 in. X 9 in. editions of the Building Code and Bridge Specifi cations. CONTENTS Section 1 — General Provisions Appendix C — Impact Strength Requirements Section 2 — Design of Welded —Electroslag Connections and Electrogas Welding Section 3 — Workmanship Appendix D Sample Ultrasonic Section 4 — Technique Test Report Form Section 5 — Qualification Appendix E — Sample of Welding Section 6 — Inspection Procedure Form for Prequalified Joints Section 7 — Strengthening and Repairing of Existing Appendix F — An Example of Weld Structures Quality Requirements — Bridges Section 8 — Design of New Buildings Appendix G —- Flatness of Girder Section 9 — Design of New Bridges Webs—Buildings Section 10 — Design of New Tubular Appendix H - Flatness of Girder Structures Webs — Bridges Appendix A — Plug and Slot Welds Appendix I - Terms and Definitions Appendix B — Effective Throat Thickness Appendix J Metric Equivalents The price** of the Structural Welding Code is as follows: sustaining member — $12.00; member — $12.00; associate member — $13.60; student member — $13.60; bookstores, public libraries, and schools — $12.80; and non-member (of AWS) — $16.00. Send your orders for copies to the American Welding Society, 2501 N.W. 7th Street, Miami, FL 33125. 'April 1970 issue of Welding Journal, pp. 263-272. "Prices shown include 4th class postal delivery within the United States. For other than 4th class or to foreign countries, postage will be charged accordingly. Add 4% sales tax for orders to be deliv ered within the State of Florida. A handling charge will be added if payment does not accompany order. 36-s I JANUARY 1 974