Microstructural Observations of Arc Welded Boron Aluminum Composites, JR Kennedy

Tags: filaments, boron, filler metal, filament, arc welding, weld, welding, composites, electron microprobe analysis, Solar Division of International Harvester Company, energy input, resistance welding, composite, California, welding power supply, fusion processes, Electron Micrograph, copper welding, titanium carbide, scanning electron microscopy, thermal energy, welding torch, scanning electron micrograph, SEM observation, specimen, Research Symposium, Reactive Materials, Technical Sessions, McCormick Place, Refractory Metals Committee, Wrought Nickel Alloys Subcommittee of the High Alloys Committee, Air Force Materials Laboratory, Explosion Welding, Welding Research Council, molten aluminum, boron filaments, energy inputs, Reactive Metals, Chicago Special, arc weld
Content: Microstructural Observations of Arc Welded Boron-Aluminum Composites Investigation indicates it may be possible to join B/AI composites by arc welding B Y J . R. K E N N E D Y
ABSTRACT. Welding studies were conducted on boron-aluminum (B/AI) composites to observe the effects of gas tungsten-arc welding on the boron reinforcing filaments and aluminum matrix. The objective of this investigation was to determine the basic potential of arc welding B/AI composites for possible structural applications. Microstructural observations after welding revealed matrix fusion without apparent boron filament damage. Analysis of weld metal regions indicated that aluminum filler metal additions intermixed with the matrix and altered its chemical composition. It is concluded that arc w e l d ing of B/AI composites may be possible by control of welding energy input. Introduction Exposure of boron filaments to molten aluminum raises the question of chemical reactivity and its effect on filament properties. Boron-aluminum interactions are time-temperature dependent and may be sluggish in the solid state, or very rapid in the presence of a superheated liquid aluminum matrix. Thermal treatments J. R. KENNEDY is with the Research Department, Grumman Aerospace Corporation, Bethpage, New York
such as diffusion welding, casting, and arc welding may induce various interfacial reactions detrimental to filament strength and composite structural efficiency. For example, it has been found that amorphous boron dissolves noticeably in molten aluminum at 1000 C.1 In other work on cast boron-aluminum composites, several minutes of exposure to molten aluminum at about 740 C caused considerable interaction leading to partial dissolution and edge scalloping of the boron-filaments.2 It was also determined2 that exposures up to three minutes at 740 C resulted in minimal observable reaction effects. In a study of the interaction between boron and aluminum,3 boron
filaments were placed in molten aluminum at 6 8 0 C. After exposure times of 1 min and 15 min, interaction layers were measured at 23/im and 5^im, respectively. In a study of B/AI solid state reactions,4 it was found that a time-temperature dependent incubation period exists, after which a loss of strength occurs. In some cases, tensile strength increased during this period, suggesting some beneficial effects of the interfacial reaction. Studies of B / A I interactions during fusion welding have been generally quite limited. It is reported that gas tungsten-arc, electron beam, and plasma welding usually result in severe weld embrittlement and filament degradation.5'6 Typical effects on the boron included fil-
.oopoooop
a 9 ^^K.^*.
axDOpoo o.
Q 3 U 3ri 0 C*
O l / r y > 5 '"y"T >5fQetti630aQc&B.aQDOoC( 3 D O O Q 0 0 3 D J OOOO GOD* e r f Fig. 1 -- Transverse section of arc weld in B/AI, made with 4043 filler wire at 6435 joules.
X30, not reduced
120-s I M A R C H 1 9 7 3
3©00 cP F/ff. 2 -- Comparison of arc welds made at 6435 joules; (a) 4043 Al filler addition; (b) no filler metal addition, X150, reduced 29%
ament cracking, break-up, misorientation, and partial or complete dissolution. In the worst cases, reductions in composite strength at the weld joint have been as high as 90 percent. On
the other hand, resistance-spot welding of boron-aluminum, w h i c h also requires matrix fusion, has shown more promise in minimizing adverse thermal effects on the boron filaments.5,3
The most significant qualitative difference between resistance welding and the other fusion processes is the relatively lower Thermal Energy input of the former. The influence of welding energy input on interfacial reactions has also been demonstrated in welding studies on titanium-tungsten and titanium-graphite composites.9 As welding energy input was increased, tungsten dissolution became greater and titanium carbide formation around the graphite filaments grew thicker. It was concluded that thermal energy delivered to the composite during welding is a significant factor in controlling the nature of filament-matrix reaction products. High welding heats can increase dissolution between components, producing extensive diffusion zones and larger quantities of additional phases. Quantitative evaluation of these effects and their contributions to composite efficiency will be necessary for practical utilization in future applications. It is clear that the specific effects of short time, high thermal energy exposure on boron-aluminum composites have not been completely characterized. This program was undertaken
Weld [in ©i © _ O ,, G ,, G ,,© ~©,-,G,-,0 GGO0GeG°GGGQG°GQG
Filament Direction
Weld " " " ^ Bead Original Boundary of Composite
B/AL Composite
^y?77?
Weld Bead
A-A
r--
Optical Micrograph
lOOx Magn
K Scanning Electron Micrograph
Fig. 3 --B/AI weld region selected for examination: (a) optical micrograph; (b) scanning electron micrograph. X100, reduced21% W E L D I N G RESEARCH S U P P L E M E N T ! 121-s
3 pi yt'^Wm yi*m%M
Fig. 4 -- Comparison of selected region on arc welded B/AI composite: (a) light micrograph (X500); (bj scanning electron micrograph (X500); (cj silicon distribution electron microprobe (X400J. Entire group reduced 35%
to obtain more insight into arc welding of B/AI composites. The present work describes an initial study to observe and to assess qualitatively the condition of boron filaments and aluminum matrix after exposure to gas tungsten-arc welding. Experimental procedure Tests were conducted on 0.025-in. (50%-6 ply) and 0.050-in. (40%-9 ply) thick B/AI composite sheets, made from unidirectionally aligned 0.004in. diam boron filaments and 6061 aluminum. The composites were fabricated by hot pressing. Arc welding was performed using a standard 300 A ac/dc welding Power supply with a manual gas tungsten-arc welding torch. An 0.040 in. diam thoriated tungsten electrode, alternating current, and argon gas shielding were employed during welding. All specimens were bead-on-sheet welds; 122-s I M A R C H 1 9 7 3
when filler metal was added during welding, 1/16-in. diam 4043 aluminum wire was employed. Immediately prior to welding, the B/AI specimens were degreased by cloth wiping with isopropyl alcohol, followed by a light surface milling in the intended fusion region with a hand draw file. The specimens were clamped in a copper welding fixture with a 1/4-in. space between the hold-down bars. A copper backing bar with a 1/4-in. wide x 1/16-in. deep groove was also used. Specimens were metallographically prepared by minimizing final polishing time, first w i t h diamond to 1/um, and then with alumina to 0.05Mm. Boron filaments were also extracted for examination by leaching the aluminum matrix with dilute hydrochloric acid. In addition to optical microscope observation of composite microstructures, Scanning Electron Microscopy
(SEM) was performed with a Cambridge Stereo-Scan instrument, and electron microprobe analysis was conducted with a Philips AMR-3 analyzer.
w 4b Fig. 5 -- B/A I interf
region
*·%& Ј '
sPfe'S... I j j · TJLx i ·
P'4
Fig. 6 -- Scanning electron micrographs of exposed boron filaments: (a) as received; (b) as welded. Both X400, reduced 5
Results and Discussion The effect of increases in welding energy input on 0.050-in. thick B/AI composite was studied. The specimens used for this study were 1 / 4 - i n . wide and 2-in. long; they were clamped and welded from one side only under a stationary arc, essentially producing a localized arc spot weld. Under these conditions, the energy input (joules) is the product of arc power and fusion time (A x V x sec). Welding With and Without Filler Metal A transverse section of a weld made with 4043 Al filler addition at 6435 joules (26 A, 16.5 V, 1 5 sec) is shown in Fig. 1. Large pores were formed in the weld metal interior, as contrasted to fine pores in the weld crown. Matrix melting has caused some shifting of the boron filaments in the fusion zone. A comparison of two welds made at 6435 joules with and without filler metal is shown in Fig. 2. These results generally typify the entire group of specimens made in this study that w e r e welded up to energy levels of 1 8,000 joules. When filler metal was not added, the upper boron filaments were subjected to intense arc heating that caused severe filament fragmentation and dissolution. In addition, the difficulty in maintaining a stable weld puddle increased, usually resulting in the aluminum's being drawn away from the puddle center to its edges, indicative of poor wetting between boron and aluminum.
On the other hand, these problems were largely eliminated when filler metal was present. Apparently, the additional Si-enriched filler metal helps to shield the top layers of boron and promotes metal flow into the matrix, as evidenced by the penetration observed in Fig. 1. Due to the limited scope of this initial work, a full spectrum of fusion effects have not been determined. This includes the effects of higher energy inputs, the welding of thicker sheet, and the effect of longer passes. However, it is known qualitatively that excessive energy inputs will cause significant filament damage and displacement regardless of filler metal additions. It should be noted that a certain percentage of filament damage, such as radial cracking and contact fragmentation, as shown in Figs. 2 and 4, is considered "normal" and also exists to varying degrees in as-received and welded B/AI composite sheet. This damage is believed to occur during primary fabrication of the composite. Microstructural Examination An analysis was made of the microstructure from an arc welded beadon-sheet specimen of 0.025-in. thick B/AI, shown schematically in Fig. 3. This specimen was welded from both sides (not simultaneously) using 4043 Al filler addition, at an energy level of about 4950 joules/in. (20 A, 16.5 V, 4 ipm). A weld region was randomly selected from this specimen for subsequent examination by various techniques, as shown in Fig. 4. Our interest was to observe particular effects on the boron and the matrix
resulting from this exposure. In Figs. 3 and 4, the SEM region shows the filaments intact within an apparently sound matrix. The distinct grain boundaries in the matrix are clearly evident in the light microscopy photograph of Fig. 4. The presence of Si in this region is confirmed by the microprobe analysis that shows a definite correlation between the grain boundaries and Sirich regions. It can also be noted that the Si distribution is relatively uniform through the specimen thickness. The average Si concentration after welding was calculated to be about 4.7%; the Si concentration in unwelded as-received composites was 0.58%. The nominal Si distribution in 4043 Al filler metal is 5% and in 6061 Al about 0.5%. The matrix appears to have been completely melted and fairly uniformly enriched with filler metal, while the filaments remained apparently undamaged. This microstructure appears to be typical of a Si-rich condition in aluminum welds and castings,10" where relative insolubility of Si in Al causes dispersion of small areas of Al-Si eutectic in the Al matrix and grain boundary enrichment. A Si layer or ring is also seen immediately adjacent to each filament (in some cases, a double Si ring is apparent). It is likely that as freezing progressed in the matrix, a solidification front rich in Si advanced to the filament periphery where the Si then segregated as along a grain boundary. After being polished and etched, the boron usually stands slightly in
W E L D I N G RESEARCH S U P P L E M E N T ] 123-s
relief above the matrix, making sharp planar focusing of both B and Al difficult with the optical microscope; this results in the characteristic shadowing around each filament. Examination of the B/AI interfacial region at higher SEM magnifications revealed a portion of the interface configuration to be in the form of a small concave fillet between the filament and matrix, as shown in Fig. 5. A somewhat rough or jagged texture was also discernible in the transition zone vicinity of various filaments. Further SEM observation of B/AI interactions was made on welded specimens in which the matrix was leached to expose boron filaments. Figure 6 is a comparison between asreceived and as-welded B/AI composite. In the welded specimen, to which 4043 Al filler was added, the variable mottled surface texture of the boron filaments is evident. The filigreed matrix between the filaments, resulting from a differential etching rate, probably represents a skeletal grain boundary network, enriched in Si as a result of welding. Subsequent electron microprobe analysis showed these areas to be Si rich relative to the as-received composite. Microprobe scans over the filaments in the welded specimen also showed significant concentrations of Al and Si. The work by Klein4 on solid state reactions in B/AI composites showed that the interaction
phase has an uneven, acicular appearance with the necdlelike protrusions into the matrix. The interaction product in that work has been tentatively identified as A I B 2 . The reaction product in the welded condition has not yet been identified, but the presence of a combination of complex intermetallic compounds is not unlikely. Conclusions It has been s h o w n that thin-sheet B/AI composites can be subjected to certain arc welding thermal conditions without severely damaging the boron filaments. It is possible to add and to intermix filler metal through the matrix to alter its chemical composition significantly. These results indicate that arc welding of B/AI may be possible by control of welding energy input Identification of the subsequent fusion reaction products, and the effects of those products on composite mechanical properties, has not yet been determined but is planned in continuing studies. In addition, knowledge of reaction growth rate kinetics and the means to control the reaction products during welding would increase the potential of B/AI for consideration as welded structures. References 1. Hansen, M., Constitution of Binary Alloys, McGraw-Hill Book Co., New York, 1958. 2. Hill, R. et al., "The Development and
Properties of Cast Boron/Aluminum Composites," Advances in Structural Composites. 12th SAMPE Symposium, Vol. 12, p. AC-18, Western Periodicals Co., N. Hollywood, California, October 1967. 3. Restall, J. et al., Metals and Materials, pp. 467-473, November 1 970. 4. Klein, M. and Metcalfe, A., Effect of Interfaces in Metal Matrix Composites on Mechanical Properties, AFML-TR-71 -1 89, Solar Division of International Harvester Company, San Diego, California, October 1971. 5. Schaefer, W. and Christian, J., Evaluation of the Structural Behavior of Filament Reinforced Metal-Matrix Composites, Convair Division of General Dynamics, San Diego, California, AFMLTR-64-36 (Vols. I, II, III), January 1 969. 6. Structural Design Guide for Advanced Composite Applications: Vol. 3 -- Manufacturing, 2nd Ed., pp. 8.4.28.4.18, Air Force Materials Laboratory, Wright-Patterson Air Force Base, Ohio, January 1 971. 7. Hersh, M,, Welding Journal, Vol. 47, No. 9, Research Suppl., pp. 404s-409s, 1968. 8. Development of Improved MetalMatrix fabrication techniques for Aircraft Structure, Report No. GDC-DBG70-002-1, Convair Division of General Dynamics, San Diego, California, July 1 970. 9. Kennedy, J. and Geschwind, G., presented at the 2nd International conference on Titanium, Boston, Massachusetts, May 1972. 10. Knight, J ,, Welding Kaiser Aluminum, Oakland, California, 1 967. 11. Brick, R. and Phillips, A., Structure and Properties of Alloys, 2nd Ed., McGraw-Hill Book Co., New York, 1949.
Technical Sessions Sponsored by WRC at Chicago Special committees of the Welding Research Council are sponsoring several technical sessions at the 54th Annual Meeting of AWS in Chicago, McCormick Place, April 2-6. Research Symposium: Weldability Testing Session 5, Tuesday morning, April 3 (Sponsored by the Wrought Nickel Alloys Subcommittee of the High Alloys Committee) Workshop on the Joining of Refractory and Reactive Metals Special Session, Tuesday, April 3, 4:30-6:00 p.m. (Sponsored by the Reactive and Refractory Metals Committee) Welding of Refractory and Reactive Materials Session 10, Wednesday morning*April 4 (Sponsored by the Reactive and Refractory Metals Committee) "Characteristics of Postweld Aging of Metastable Beta Titanium Alloy" by M. A. Greenfield and C. W. Pierce, Wright-Patterson Air Force Base "Microstructure-Property Control with Postweld heat treatment of Ti-6AI-5V-2Sn" by R. P. Simpson, Air Force Materials Laboratory, and K. C. Wu, Northrop Corp. "Explosion Welding" by H. B. Hix, E. I. du Pont Co. "Recent Developments in the Welding of M o l y b d e n u m " by A. J. Moorhead and G. M. Slaughter, Oak Ridge National Laboratory. 124-s I M A R C H 1 9 7 3

JR Kennedy

File: microstructural-observations-of-arc-welded-boron-aluminum-composites.pdf
Author: JR Kennedy
Published: Mon Feb 11 16:05:34 2013
Pages: 5
File size: 0.61 Mb


Send in the Clowns, 2 pages, 0.06 Mb

, pages, 0 Mb
Copyright © 2018 doc.uments.com