WELDING RESEARCH

SUPPLEMENT TO THE WELDINC JOURNAL, OCTOBER 1988

Sponsored by the American Welding Society and the Welding Research Council

All papers published in the Welding Journa/'s Welding Research Supplement undergo Peer Review before publication for: 1) originality of the contribution; 2) technical value to the welding community; 3) prior publication of the material being reviewed; 4) proper credit to others working in the same area; and 5) justification of the conclusions, based on the work performed. The names of the more than 170 individuals serving on the AWS Peer Review Panel are published periodically. All are experts in specific technical areas, and all are volunteers in the program.

High-Temperature Brazing of Stainless Steel with Low-Phosphorus Nickel-Based Filler Metal

Three experimental brazing filler metals are evaluated for suitability in industrial applications

BY E. LUGSCHEIDER AND T. COSACK

ABSTRACT. The demand for boron-free, metals to a wide spectrum of applica­ using a specially developed metallo­ low-melting, nickel-based filler metals led tions. In addition to applications in spe­ graphic specimen. It has a linearly increas­ to the development of three high-tem­ cialized industries, the use of high-tem­ ing joint clearance. Results of metallo­ perature brazing alloys with low amounts perature brazing with nickel-based filler graphic investigations are summarized in of phosphorus and iron. The evidence of metals is increasing in conventional the figures. The investigations have been the industrial suitability of the proposed areas. made with a computer-controlled micro­ new brazing alloys was made with stain­ By using nickel-based brazing alloys, graph analysis system, which has graph­ less steel AISI 321 using a combined test hard phases, which decrease tensile ics. In the end, the results of the metal­ method. The results of the metallograph­ strength, are formed in the center of the lographic investigations were verified by ic investigation were confirmed by tensile braze where the maximum joint clear­ tensile strength tests of a brazed butt strength tests. The tests established opti­ ance is exceeded. This can be seen in joint tensile specimen. mum operating parameters such as braz­ metallographic investigations. The maxi­ ing temperature, time and joint clear- mum joint clearances, depending on the Experimental Work brazing parameters, were determined by Brazing Alloys Introduction The three silicide-phosphide high-tem­ perature brazing filler metals were devel­ Nickel-based filler metals are desirable KEY WORDS oped by modifying ternary Ni-Cr-Si alloys because of their alloying elements and High-Temp. Brazing with phosphorus and iron. simple method of production. Extensive 321 Stainless Steel Alloy 1 was developed by the addition composition possibilities lend these filler Low Phosphorus of 0.5% phosphorus to the alloy 20.4Cr- Nickel Base 11.6Si-bal Ni, which is almost the ternary Brazing Filler Metal eutectic point of the Ni-Cr-Si system. E. LUGSCHEIDER and T. COSACK are with the Max. Joint Clearance Hence, a melting range was reached Material Science Institute, Aachen University of Wedge Specimen between 1044° and 1060°C (1911° and Technology, Federal Republic of Germany. Investigation Method 1940°F). Combined Methods Paper presented at the 17th International A WS Alloy 2 was developed by alloying 3% Braze Formation Brazing Conference, held April 15-17, 1986, in phosphorus to the ternary alloy 27.1Q- Atlanta, Ca. 5.2Si-bal Ni, producing an alloy with a

WELDINC RESEARCH SUPPLEMENT I 215-s Table 1—The Investigated Nickel-Based Filler Metals

Brazing Nominal Melting Range Filler Composition °C(°F) Metal wt-% Tsol Tjqa

Alloy I 20.3Cr-1l.5Si- 1044 1060 0.5P-bal Ni (1911) (1940)

Alloy 2 26.3Cr-5.1Si- 952 1155 3.0P-bal Ni (1746) (2111)

Alloy 3 14.8Cr-8.0Si- 996 1058 3.0P-3.0Fe- (1825) (1936) bal Ni

the joint through capillary flow during the brazing process. For metallographic preparation, the braze is sectioned. For the investigation in a computer-controlled micrograph analy­ sis system, the wedge specimen is etched and surfaced by vapor-deposition. The computer-controlled micrograph analysis system is comprised of a microscope, TV camera, display monitor, digitizer table Fig. I — Wedge specimen used to vary the joint clearance and computer. The brazing clearance and joint geometry are registered by measur­ ing the wedge specimen. All measuring melting range of 952° to 1155°C (1746° and the melting range composition is data are stored on discs in the format of to 2111°F). Alloy 2 has a high chromium shown in Table 1. tables. They are available whenever content, which can be seen in the wide With SEM investigation, Ni solid-solu­ needed. A graphic working program melting range and the high liquidus tem­ tion silicides and phosphides were dis­ takes the measuring data and draws a perature. covered in the three brazing alloys. The wedge diagram on the printer. The dia­ Brazing Alloy 3 was developed from undesirable hard phases, such as the gram gives the geometric structure of the the ternary alloy 15.8Cr-8.5Si-bal Ni by silicides and phosphides, were evident wedge specimen with all its characteris­ alloying it with 3% phosphorus and 3% only in a low-volume amount. tics. A typical characteristic with the appli­ iron. It has a solidus of 996°C (1825°F) cation of nickel-based filler metal is the and a liquidus of 1058°C (1936°F). dependence of the tensile strength on Test Method At the beginning of testing, special key the amount of hard phases in the center alloys were used. The commercial braz­ It was shown that the brazing joint of the braze joint, which in turn, depends ing alloy BNi-7 was the phosphorus sup­ clearance depends on the development on the joint clearance. With the begin­ plier, the BNi-5 alloy and a nickel-silicon of hard phases in the center of the braze. ning of the hard phase stabilization, the alloy were the silicon suppliers, and the Therefore, a special specimen was devel­ maximum brazing clearance (MBC) is rest was pure nickel and chromium pow­ oped. The wedge specimen is composed defined. Below MBC, the micrograph of der. The alloys were made molten in a of two joining halves, tacked together by the braze joint displayed only Ni-solid- vacuum induction furnace and then GTAW after inserting spacers to make solution, indicating good tensile strength mechanically pulverized. After produc­ different joint clearances. qualities. The MBC is indicated during the tion, the brazing filler metals were chem­ Figure 1 shows the structure of the measurement of the wedge specimen. ically analyzed and the melting range was wedge specimen with 50-/im and 100- This information is used by the computer, investigated by differential thermo analy­ iiim spacers and the geometry of the joint and the maximum brazing clearances of sis (DTA). Only an insignificant deviation clearance. The filler metal powder is every wedge specimen are drawn in from the nominal composition occurred, placed in the V-groove, which then fills accordance with the brazing temperature and brazing time.

The test method is a combined meth­ od, and the results of the metallographic investigations are compared with tensile strength tests. Brazed butt joint tensile specimens are prepared by using a fixed brazing temperature and holding time, but with different joint clearances. Figure 2 shows the specimen used for tensile strength tests. The specimen con­ sists of two halves with the brazing joint in the middle. The various joint clear­ ances are made by spacers, which are removed after the brazing process. With a clearance below the maximum brazing Fig. 2 — Brazed butt joint tensile specimen clearance, the tensile strength of the

216-s | OCTOBER 1988 Fig. 3 — Microstructure of a joint brazed with Alloy 1 filler metal. The Fig. 4 — Microstructure of joint brazed with Alloy 1 filler metal. The base base metal is 321 stainless steel, brazed at a cycle of 1070"C/60 min, metal is 321 stainless steel, brazed at a cycle of 1150°C/10 min, with a with a joint clearance of 1b nm joint clearance of 20 nm braze equals the tensile strength of the During the brazing process, the molten 60 min. The grain growth in the brazing base metal. With a clearance above the filler metal fills the clearance according to joint can be seen and is contrasted with maximum joint clearance, the tensile capillary force. This capillary force entails the microstructure of the base metal. strength reaches only 20 to 50% of that an interaction between the base metal The microstructure of the solid solution of the base metal. and the filler metal. At first, there are contains 41.8% nickel, 34.1% iron, 19.1% heterogeneous chemical reactions at the chromium and 5.1% silicon. The silicon Braze Formation solid-liquid interface, whereby atoms amount was below the solution limit, so accumulate at the interface of the solid the formation of a complete solid solu­ For the brazing filler metals Alloy 1, and liquid metal. This process is known as tion was possible. The phosphorus con­ Alloy 2 and Alloy 3, the wedge specimens wetting. Secondly, there is a concen­ tent was below the detection limit. and the tensile test rods were brazed at trated equilibrium behind a thin limit temperatures and times shown in Ta­ zone. The new brazing alloy reacts with ble 2. the base metal elements such as iron, Results of the Metallographic Investigations All the specimens were stainless steel chromium and nickel. After the brazing and the Tensile Strength Tests AISI 321 brazed in a vacuum furnace with process is completed, there is a chemical AISI 321 stainless steel joints brazed 3 1 a pressure less than 10~ mbar (10~ composition in the braze different from with Alloy 1 were investigated at brazing Pa). the filler metal. A micrograph indicates a temperatures between 1070° and solid solution in the base metal in which 1150°C (1958° and 2102°F) with brazing silicon and phosphorus are dissolved. The times of 10 and 60 min. The selection of solid solutions grow on concentric circles brazing temperatures was made accord­ Table 2—The Brazing Cycles for the Three from the base metal to the brazing filler ing to the results of DTA. Investigated Brazing Filler Metals metal, meeting each other in the center Figure 4 shows a nickel solid solution of the braze. micrograph without hard phases at a joint Brazing Filler Metal Alloy Alloy Alloy It becomes a braze without hard clearance of 20 pm. The grains grow into Brazing Cycle 1 2 3 phases. At wide joint clearances, the the brazing filler metal from the base diffusion distances are already too big. metal. But the difference between base 1040 °C The hard phases formed cannot com­ metal and filler metal is enormous with (1904 °F)/10 min pletely dissolve in solid solutions. Hard this specimen brazed at 1150°C and /60 min phases, such as silicides and phosphides, having a holding time of 10 min —Fig. 5. 1070 °C form in the center of the braze, reducing With an additional diffusion heat treat­ (1958 °F)/10min tensile strength. ment, base metal assimilation can be /60 min Especially in solid solutions, iron expected. amounts of up to 40% could be analyzed The braze with a joint clearance of 93 1090 ' C by SEM, although before the brazing pm displays one-third hard phase in the S (1994 F)/10 min cycle, iron was not in the filler metal. A center of the joint. It can be clearly seen /60 min base metal solid solution can also be seen that the solid solutions are growing on with SEM analysis, but the difference concentric circles from the base metal 1120 °C between the braze, which is influenced into the filler metal. The different phases (2048 °F)/10 min in the middle of the joint are nickel solid /60 min by molten filler metal and the brazing joint clearance, is very big. This phenom­ solutions, nickel-chromium silicides and 1150 °C enon is known as erosion. This material nickel silicides. The grain boundary (2102 °F)/10min change is caused by material transport growth in solid solution has its origin at /60 min mechanisms between solid and liquid the grain boundary of the base metal. metals and by solid matter diffusion. Already, the diffusion distances for a 1190 °C The joint in Fig. 3 was brazed with the complete solubility of the hard phases (2174 °F)/10min filler metal from Alloy 1, at a temperature formed with silicon were too great at /60 min of 1070°C (1958°F) and a holding time of these brazing parameters.

WELDING RESEARCH SUPPLEMENT I 217-s tensile strength characteristics. The frac­ ture area ran through the filler metal and the base metal with shearing lips. At the brazing clearance of 12 microns, the fracture area in the base metal was sepa­ rate from the filler metal. If the joint clearance value is above the MBC, the tensile strength decreases in proportion to the increase of hard phases in the center of the filler metal. The base metal tensile strength for 321 stainless steel is shown in Fig. 8. Figure 9 shows a specimen in which the tensile test was interrupted at 565 N/mm2, and an untested specimen as a comparison. The elongation reached nearly 30%, with very little shrinkage. Brazing tests were conducted with Alloy 2 filler metal on 321 stainless steel at temperatures between 1090° and Fig. 5—Microstructure of a joint brazed with Alloy I filler metal. The base metal is 32 I stainless 1190°C (1994° and 2174°F), with holding steel, brazed at a cycle of 1150°C/10 min, with a joint clearance of 93 /um times of 10 and 60 min —Fig. 10. The brazing joint with a clearance of For all the brazed specimens, diagrams Figure 7 shows the filler metal-base 44 microns shows no continuing hard were drawn by the computer system metal characteristic of the AISI 321 stain­ phases and possesses solid solution schematically showing geometry of the less steel joints brazed with Alloy 1. For grains, a condition which extends wedge specimens. the brazing time of 10 min, the MBC's throughout the whole braze. The fine The wedge specimen diagram (Fig. 6) appear almost temperature independent branchings are remarkable. Through SEM shows a MBC of 29 pm with the given between 22 and 35 pm. At the brazing investigations, they proved to be phos­ brazing parameters. The wide beam in time of 60 min, MBC's were established phide grain boundary precipitations. The the diagram illustrates the beginning and between 35 and 42 ^m. The filler metal- precipitations show a high chromium the extent of hard phases. The straight base metal characteristic data offer a content. line pairs marked by crosses and stars basis for the estimation of manufacturing With the investigation of the maximum show the variation in joint clearance. The parameters for brazing temperature, time brazing clearance, only the hard phases in base metal is not influenced by trans and and joint clearance. the center of the filler metal were consid­ intercrystalline precipitations as with Results found by metallographic inves­ ered. The phosphide grain boundary pre­ boron-containing filler metals. The tigations are comparable to the tensile cipitations of finely shaped branchings wedge specimen diagram shown is a strength tests conducted on brazed butt could not be detected. typical example for all the wedge speci­ joint specimens. All investigated MBC's for the Alloy 2 men diagrams, produced by the comput­ At joint clearances below the MBC, system are shown in Fig. 11. They range er after analyzing data available. All which is marked as a tolerance area, from 5 to 27 pm for the holding time of wedge specimen diagrams with the MBC tensile strengths near the base metal 10 min and from 30 to 42 ^m for 60 marked appear as MBC's in the filler tensile strength were reached. The min. metal-base metal characteristic dia­ microstructure of the brazing filler metal At these ranges, the necessity of the grams. was a nickel solid solution with good combined test method is especially clear.

1.0 100 Cycle: 1150 T.I210? "Fl/lO niln Alloy 1/AISI 321 Alloy 1/AISI 321 LT • 10 1111| - un 3.2 80 " oS8 lllll -

100 - 2.1 60 -_=*=—=****] t -1 1 0 - 1.6 10 c t •****-*--» ii — "i =-=—-=i ——-— — —c _ - - c Ay'" !_ , ex , 100 •

io;o 1090 1120 •c 1150 3 1958 1991 2018 •F 2102 MBC • 29 un/1. lV10~ ln 1 1 1 1 brazing temperature Fig. 7 - Alloy I tiller metal and 32 I stainless steel base metal char jr. ter- 0.8 1.6 2.1 istic diagram bra? ing clearance Fig. 6 — Wedge specimen diagram for Alloy I filler metal, brazed at a cycle of 1150'C/10 min

218-s i OCTOBER 1988 120 800 ' i i 1 KSI 700 N/mm2 U.T.S. AISI 321 90 600 • a. o 500 > 100 Q F: Brazing defect 300 i MBC •: Fracture In the o base metal cc 30 200 - D: Fracture in the < UJ brazing seam 100 3 1/321 stainless steel, 1090 C/60 min 30"o elongation, but very little reduction of area UJ a I Tensile strengths were reached at 400 o 2 cc N/mm (58 ksi) with joint clearances < Ul below the maximum joint clearance in (/) the diagram. ui The diagram presented in Fig. 12 cr shows that there is no conformity between the metallographic investiga­ ui tions and tensile strength tests. Obvious­ Q. ly, the phosphorus amount of 3.0% is too o high for this brazing alloy. With this appli­ u—Ii cation, phosphide precipitations collect > mostly on the grain boundary, a condi­ UJ tion which can only be identified by SEM a investigation. x o Figure 13 shows the irregularly distrib­ cc uted phosphide precipitations. < Ul According to its melting characteristics, Ui the lowest brazing temperature was cc expected from Alloy 3. Brazing tempera­ z tures between 1040° and 1120°C(1904° Ui 2 and 2048 F) were selected, along with Fig. 10—Microstructure of a joint braze with Alloy 2 filler metal. The base metal is 321 stainless CL brazing times of 10 and 60 min. Fine steel, brazed at a cycle of 1190°C/60 min, with a joint clearance of 44 /um o —I Ui i.O 120 > 100 800 Ul "3ln Alloy 2/AISI 321 um KSI a • 18 Hill 700 i 3.2 2 o 80 Q hit Mill N/mm U.T.S. AISI 321 tr < ui 90 600

100 ^ 1.6 60 Brozing defect HBC CL ir?j -ir 300 Fracture ln the •2X PF base metal o 20 Fracture in the _J 200 ui 30 brazing seam > Ul 100 a 1090 1120 1150 •c 1190 o 1991 2018 2102 •F 2171 12 25 75 um 100 SO cc brazing temperature 0 0.17 o.S 1.97 2.95 3.91 < Ul Fig. 11 — Alloy 2 filler metal and 321 stainless steel base metal character­ 10"Jin brazing clearance (/> istic diagram Ul Fig. 12 — Tensile strength ot 321 stainless steel butt joints brazed with cc Alloy 2 at a cycle of 1150°C/60 min

WELDING RESEARCH SUPPLEMENT | 219-s t> * • •• .- -' x\

r^-lCNK,

v- • 9.

* * ** a *

Fig. 13 — Microstructure of phosphide precipi­ Fig. 14 — Microstructure of joint brazed with Alloy 3 filler metal. The base metal is 321 stainless steel, tations brazed at a cycle of 1090°C/60 min, with a joint clearance of 40 nm

120 800 1 i i 1 1.0 KSI 100 700 - - 3 Alloy 3/AIS1 321 10 ln N/mm2 U.T.S AISI 321 un 90 Z 3.2 • 18 IIIII 600 • | 80 • 68 IIIH — " 500 - - OJ

§ IA 60 100 - - o u F: Brazing defect CT 300 - NBC '• - •: Fracture ln the 5 1.6 10 a : : base metal 200 - 0 - igX- itij D: Fracture ln the brazing seam 201^ 100 -

0 i 12 25 50 75 um 100 0L 0 1010 1070 1090 1120 •c 0 0.17 0.98 1.97 2.95 1901 1958 1991 •F 2018

brazlng temperature brazing clearance Fig. 15—Alloy 3 filler metal and 321 stainless steel base metal character­ Fig. 76 — Tensile strength of 321 stainless steel butt joints brazed with istic diagram Alloy 3 filler metal at a cycle of 1090"C/60 min branchings are in the areas of the solid base metal at joint clearances below the metal combination. solution without hard phases. MBC. Alloy 1 filler metal showed the best As seen in Fig. 14, solid solution bridges match between metallographic investiga­ are present throughout the braze. An tion and tensile strength test. With intermediate area of hard phases exists, Conclusion brazed butt joint tensile specimens, the decreasing tensile strengths. filler metal reached the base metal tensile The results of metallographic investiga­ The demand for boron-free nickel- strengths at a brazing temperature of tions show that joint clearances have to based filler metals led to the develop­ 1070°C and a brazing time of 10 min. be adjusted between 20 and 30 pm (see ment of silicide-phosphide brazing alloys, This particular filler metal might be used in Fig. 15) to produce brazements without of which the suitability for industrial appli­ a similar manner as the commercial filler hard phases. Tensile strength tests show cations was so far unknown. metal BNi-5, where an 80°C (144°F) low­ that acceptable tensile strengths can be Using a combined test method, the er brazing temperature is possible. reached at brazing temperatures above interrelationship between joint clearance Alloy 2 filler metal is not suitable for 1070°C. Indeed, all tensile strength spec­ and hard phase formation was deter­ industrial applications. At tight joint clear­ imens brazed at 1040°C have discontinu­ mined by metallographic investigations ances and with brazing temperatures of ities, which can be traced to poor flow and tensile strength tests. The practical 1190°C, satisfactory tensile strengths and wetting of the brazing alloy. Further­ possibilities can be appraised under appli­ could not be reached. This brazing alloy is more, satisfactory results were attained cation conditions. The data from the tests not an acceptable alternative to commer­ at brazing temperatures above 1090°C. show that only through the use of the cial filler metals or the other two experi­ The tensile strength diagram in Fig. 16 two combined investigation methods is it mental filler metals. shows that tensile strengths of the filler possible to come to a final conclusion Alloy 3 filler metal can be used at metal almost reach the strength of the about the suitability of a filler metal-base brazing temperatures above 1070°C.

220-s | OCTOBER 1988 High tensile strengths can be reached at W. C. 1980. Structure of low-phosphorus BNi-7. Welding journal 62(b)Ab0-s to 164-s. temperatures of 1090°C, making it a alloyed nickel-chromium-silicon brazed stain­ 4. Lugscheider, E„ and Pelster, H. 1983. satisfactory alternative for industrial use. less steel joints. Welding lournal 59(10):283-s Nickel base filler metals of low precious metal to 288-s. Minimum brazing temperature of content. Welding lournal 62(10):261-s to 2. Lugscheider, E., and lversen, K. 1977. 266-s. 1090°C is about 60°C (108°F) lower than Investigation on the capillary flow of brazing 5. tugscheider, E., and Krappitz, H. 1985. the brazing temperature of BNi-5. filler metal BNi-5. Welding Journal 56(10):319-S The influence of brazing conditions on the to 324-s. impact strength of high temperature brazed References 3. tugscheider, E„ and Partz, K.-D. 1983. joints. Paper presented at the 16th Annual High temperature brazing of stainless steel AWS International Brazing and Soldering Con­ 1. Lugscheider, E., Klohn, K., and Burchard, with nickel-base filler metal BNi-2, BNi-5 and ference, Las Vegas, Nev.

WRC Bulletin 334 June 1988

Review of Properties of Thermo-Mechanically Controlled Processed Steels—Pressure Vessel Steels for Low-Temperature Service Japanese steelmakers have developed the Thermo-Mechanical Control Process (TMCP) that includes an accelerated cooling process in the plate mill. Fabricators have utilized various highly efficient welding technologies in their fabrication. Accordingly, a great deal of joint work has been carried out to put this steel and welding technology into practical use. This report summarizes the development of TMCP steel in Japan and was prepared by their Subcommittee on Pressure Vessel Steels. Publication of this report was sponsored by the Subcommittee on Thermal and Mechanical Effects on Materials of the Pressure Vessel Research Committee of the Welding Research Council. The price of WRC Bulletin 334 is $24.00 per copy, plus $5.00 for postage and handling. Orders should be sent with payment to the Welding Research Council, 345 E. 47th St., Suite 1301, New York, NY 10017.

WRC Bulletin 329 December 1987 Accuracy of Stress Intensification Factors for Branch Connections By E. C. Rodabaugh

This report presents a detailed examination of the stress intensification factor (SIF) formulations for perpendicular branch connections that are specified in American standard codes for use in the design of industrial and nuclear Class 2 and 3 piping systems. Publication of this report was sponsored by the Subcommittee on Piping, Pumps and Valves of the Pressure Vessel Research Committee of the Welding Research Council. The price of WRC Bulletin 329 is $20.00 per copy, plus $5.00 for postage and handling. Orders should be sent with payment to the Welding Research Council, Suite 1301, 345 E. 47th St., New York, NY 10017.

Revised WRC Bulletin 297 September 1987

Local Stresses in Cylindrical Shells Due to External Loadings on Nozzles—Supplement to WRC Bulletin 107 (Revision I) By J. L. Mershon, K. Mokhtarian, G. V. Ranjan and E. C. Rodabaugh

This Revised Bulletin 297 is intended as a replacement for the current supplement to Bulletin 107 and is specifically applied to cylindrical nozzles in cylindrical vessels. It replaces WRC Bulletin 297, August 1984. The changes in the text, figures and tables to update the 1984 edition of Bulletin 297 are described in the "Foreword to Revision I." This revised Bulletin was prepared by the Subcommittee on Reinforced Openings and External Loadings of the Pressure Vessel Research Committee of the Welding Research Council. The price of Revised Bulletin 297, September 1987, is $24.00 per copy, plus $5.00 for postage and handling. Orders should be sent with payment to the Welding Research Council, Suite 1301, 345 E. 47th St., New York, NY 10017.

WELDING RESEARCH SUPPLEMENT | 221-s