SE1550718A1 - Brazing material testing - Google Patents

Abstract

Abstract A brazing method for brazing article of Stainless steel comprises the steps of:Providing a brazing paste comprising a metal powder comprising: Trace amounts of C and S; 2-2,5% Mo; 12-13% Ni; 17,5-18,5% Cr; 6-8.3% Si; 4.5-5.5% Mn; all percentages being given by weight, balance being iron (Fe). The metal powder has a particle size less than 106 um,wherein the metal powder is made into a paste by addition of binder and solvents.The paste is applied on or close to contact points between the stainless steel articlesto be brazed, whereafter the articles are placed in a furnace. The furnace is heated toa temperature lower than the temperature at which the metal powder of the brazingpaste is completely melted, whereafter the stainless steel articles are allowed to setby lowering the temperature.

Description

Brazing material testing Summary Joining tests have been conducted with 316 base materials and different blends ofpure 316L PM and boron free F312 fillers. The Boron free F312 fillers have,according to DTA-TGA (Differential Thermal Analysis - Thermo Gravimetry Analysis)measurements, an approximate solidus temperature, i.e.. melting temperature, of1250 °C. The heat treating temperatures in this study are conducted at temperaturesequal to or lower than 1250 °C. Thus, the joining process can be seen as activateddiffusion bonding or brazing with partly melted fillers, depending of the Si-content inthe fillers in this study ...... ._ Purpose The purpose of this study is to evaluate the possibility to join AlSl 316 stainless steelbase material with boron free fillers. Boron is known to combine with the Chromium instainless steel and generate Chromium borides. The formation and precipitation ofChromium borides reduce the mechanical strength as well as the corrosionproperties of the stainless steel base metal.

Fillers The tested alloy, below called G577, comprises trace amounts of C and S, 2.3% Mo,12.9% Ni, 18,4% Cr, 8.3% Si, 5,6 % Mn, balance being lron (Fe). To evaluate fillerswith different Silicon contents, different blends of alloy G577 powder were mixed withpure 316L sinter powders (PM) into a paste. Alloy G577 was diluted with 316L PMpowders in ratios 75/25 and 50/50. Also, fillers comprising 316L PM powders andbinder only were tested, showing surprisingly good results.

Moreover, the bonding properties with respect to particle size distribution were alsoevaluated. Two different sievings of Alloy G577, i.e. >45 um and <106 um, weremixed with the 316L PM of <22 um in the ratios 75/25 and 50/50 respectively. Theproperties of the different filler pastes are presented in table 1.The chemicalcompositions for the F313/316L blends in table 1 are calculated values.

Table1 filler ro rties F313-B-D-9302 204 1649135 1280 1340:F313P-91XX 15 009 G577 59 1250 1315 316LP-92XX 15 OO3B 614637 645 _ 4 F3 6L-91XX 15 O15A _ 3 6 F3 6L-91XX 15 O15B _ 4 86L-9OXX 15 O16A 7 3 6 ln the below figures, DTA/TGA measurements for the first alloy of Table 1 is shown.As can be seen, there is a melting onset of the alloy at about 1280 degrees C, and asecond peak at about 1340 degrees C. The onset temperature is the temperaturewhere the alloy starts to melt, and the 1340 degrees C peak represents roughly thetemperature where the alloy is completely melted.

Dsc amw/mg) i exo 0 n '5 I oneea: 1 gçg fåcwwm*_10 ._15 _ Peak; _20 _ 200 400 600 800 1000 12'00 1400Temperature /°C Meln 201370771607 51 user sEEvDl Instrument I NETZSCH STA 449F3 STAIMQFSAVÛKSSVM File I Lil0MDlTerrnoanaly/serlDTA dlVlOWE MârS\2013'07'15\F312 lol 1649135J1gi>dS3 Pfgjeex; sefnpie; F312, 28.5 mg se e een/To ; Dsc/TG cp s / s 're Corn/fn. lange; 020/35000 mgidentity; F3121011ß49135 Merenei ; sample M rype grfneee. ; nsc7TG / eernplewln conecllon Dsc earn/fn. lenge; 02015000 livnere/lime; 201370771515:34:42 cgneengn me ; AR RT71500°c 2013701723.ngl>7l>e3 segfnenre; 1/2 Lebgmgry; Tennoleb 'refnpceL/sene. Files; 2012071ß,10|<,A _ 1171:312012071ß,10|<,AR.ngl>7ee3 cmeibie; Dsc/TG pen A12o3 opemgf; sEEvm Renge; 30/10.0(| Creflïed Wlïh NETZSCH PIUIQUS SO/'TWGIQ Figure 1) DTA-TGA for 1649135 ln Figure 2 below, DTA-TGA measurements for two different brazing materials, theG577 (red line) and a brazing material being equal to G577 but with an additionalcontent of boron (green line) are shown. As can be seen, the melting temperature ofthe Boron containing brazing material is significantly lower than for the G577 brazingfiller - 1180 degrees C for the brazing material containing boron and 1310 degrees Cfor G577, respectively.

Figure 2) DTA-TGA for G577 (red) and a brazing material containing Boron (green) Tests of mechanical strength The mechanical strength of the braze was evaluated by tensile testing of “Innovationdisc samples (|DS)”, i.e. test samples in the form of small discs provided with apressed pattern resembling the herringbone pattern provided on heat exchangerplates in order to provide contact points between neighbouring heat exchanger plateswhile holding the plates on a distance from one another. The samples were coatedwith brazing material using a stencil, see figure 3. The stencil is identical to a stencilused for actual coating of heat exchanger plates with brazing material. The chevronangle, i.e. crossing angle, of the neighbouring lDS:es is 45°. Although the samepattern is used for the paste application and measures are taken to select sampleswith similar amounts of paste, the amount of filler varies from sample to sample. Thevariation in paste amount results in different cross sections of the braze joints andthereby differences in “rupture force”. The gap clearance between the beams of the two lates ma alsovar.D Y Y x \\ Tests concerning actual heat exchangers 4 heat exchangers were initially manufactured, where the heat transfer area,including port regions, wherein a paste comprising G577 alloy was applie to areassurrounding port openings and the heat exchange area, i.e. the area provided withherringbone pattern. Flanks were robot dispensed with F312D paste, i.e. a pastehaving low melting point by addition of boron with the standard dispensing program.All units leaked in the port area. lt was therefor concluded that G577 brazing material did not provide “fluid tight” joints. This is probably due to inherent cavities in thebrazing joint, which will be described later. The robot dispensing program wastherefor modified so as to apply paste containing brazing material comprising boronand having a melting temperature lower than the brazing temperature on flanks aswell as in the ports to be sealed.

Brazing cycles “Boron free” fillers should theoretically not gain anything from a diffusion step duringbrazing. Contrary, by minimising time at elevated temperatures, grain growth couldbe reduced and thereby increasing the base material strength. Different brazingcycles have therefore been tested. See Figs 5 to b below, wherein temperature andpressure of the brazing furnace is shown as a function of time. lt should be noted thatthe following results, if not otherwise stated, were obtained with a brazing cycleaccording to Fig. 6.

Figure 5) 1160 °C; 120 min. Pink line represents furnace pressure.

F|gure 6. 1i2i50°C; iiåöimin + TiÖÖÖÖj» 180 minfPink ||ne represents furnace pressure.

Figure 7.1250°C; 120 min. Pink line represents pressure.

Figure 8.1250°C; 90 min + 1150°C; 60 min. Pink line shows furnace pressure.

Results Time- Temperatureüfšmlwæ ltošš thickness»§_ - : _* Ü cti? was - __ _ åk šäšflcfïia - ._ ' på:maa . .fiilšer r ~ _42; x få w wgg; . g' f; ,. lg; f; .mnof ß Åaëx Qtå Åås »ffš_ -zß g» m 4521» »Q g»Sääf? få? aïeëä åk få*BT _ *så :ü*Lift w? Figure 9) Tensile strength of G577 applied on SS316 IDS having a plate thickness of0,3 mm for different brazing temperatures. The best results are achieved for abrazing temperature of 1250degrees C (brazing cycle according to Fig. 6). »fïflfixnï ïflfniåš üßtmfl cušâ ïhšckuessïölišaïjtsfiítüö - ~ ä fåE _. Q“E “få | | iïzïâüfzïin 125% + Iïiiïüiíï Eflëïüznín IÉSQQ Bfyåë-Qnfišn; ïššïfiäï + »âü snitt íílâüí ïamp Figure 10) Time-Temperature effect on mechanical strength for G577 having a grainsize of less than 106 pm. Please note that there is a trend indicating that shorter timeat the highest temperature yieids a strongerjoint. This might be due to the fact thatshorter time periods at the elevated temperatures reduces grain growth of the basematerial.

»Individual Uašufl Fšot of Fflrcefl Q 3 man cošš thïšcknessJ' I»SölišaïfišFüfüi?s Q? »alsjzrttlß “ 51:11:12 - a å s* _ Éß å, *ßå ' ä ß* s i s*w s i*å å __ & :alfa - ä* man - Ü i 'i | i | 'i | i |gšx m n, y _ '_ t _ \. *'si? Mif .- fis s* .- så »så få »så å*:tftfššaä- :Fä Qvßáfi *Fy :ašjçç å;Flikar type, partïïcie size fllirstrmuftifin and Sii-wnftent Figure 11) Comparison of tensile strength for 0,3 mm coil thickness with differentfillers, heat treated at 1250 C according to the brazing cycle of Fig. 6. “Gen||” is abrazing material identical to G577, except for the addition of 1,1% B. 316L PMcontains 0,4% Si. Cu-foil is heat treated at 1130 C. The brazing materials containing4% and 6% Si were obtained by mixing G577 with stainless steel powder in therelations 50/50 and 75/25, respectively. Please note that the small-grain (less than 45pm) filler with 6% Si gives the same strength as the “Gen||” filler containing boron,but the variance between strongest and weakest joint is significantly smaller, which isbeneficial.

»B 4 um! flošš thickness , de Gsànïïa* f fiíïrïfifš -~ åsfl" 'ÜÜÜ å; änïl: : _ ä :_å. ß* §E ÅS" Ü §:E3011431213lülšiš -lQfiÉlUÜ | | | | | i f? _ 'x *vx *ti ~2= :agge 41 gåva ä Äšåå ß üßü. få sig: f i? ifiilw* Figure 12) Tensile strength 0,4 mm coil thickness with different fillers, heat treated at1250 °C according to the brazing cycle of Fig. 6. Cu-foil is heat treated at 1130 °C.Please note that also for the 0.4 mm coil, the best result concerning variance of thestrength is obtained with small-grain braze filler containing 6% Si. fiyfi taxin cuii thzš-cknefis, Cïcšie Genlfiï1010130-*3 3:00 - est "c - å*s __ _ S? æ w2 230m - i * f*ä s sL ä*i? seas- - _1% 5000- - , | | | i _ | _ | -< íüfisng: »äíufisrïf flfiEm-ï Wšrnyf Cu--fcšifiller Figure 13) Tensile strength 0,5 mm coil thickness with different fillers, heat treated at1250 °C. Cu-foil is heat treated at 1130 °C. Most consistent results achieved withsmall-grain braze filler containing 6% Si.

Tensile strength testingln the following, mic aphs of cut-up brazing ' ints are shown. lf not otherwisestated, the samples e been heat-treated ac ding to the brazing cycle of Fig. 6.

Figure Brazing filler comprising 8% Si, grain size less than 106 pm. Heat treatedat 125 according to Fig.6.

Figure Brazing filler coprising 8% Si, <45prn grain size (top) _ <106 prn grainsize (b m).

\ F|gure 16) Brazing f|||er comprising pure 316L powder having a grain size of 22 um.Please note that the structure of the base material seems totally unaffected by thebrazing material (no entrainment) and that the joint is porous _ The joint is, however,surprisingly strong. _ Please note that thebase material seems unafiected by the brazing material an at the pores aresignificantly larger than for the joint obtained by the 22 pm pure stainless steelpowder brazing filler. g 8) Brazing joint with a brazing material comprising 6% Si (i.e. a mixture of75% G577 with a grain size of106 and 25% pure stainless steel powder havingrain size of 22 pm). Please note the brazing material has entrained well intobase material, however without eroding the same. The pore size of the brazed jointis, however, rather large. e 20) Joint obtained by small-grain (less than 45 pm) braze filler with 6% Si. 'sfacto entrainment of brazing filler material into base material, excellent erosionproperties significantly smaller pore size than with 160 pm braze filler having thesame percentage of Si.

Rupture samples Below, some rupture photographs of strength tested ' ' s are shown. Please notethat for almost all of the samples, the rupture occurs ' e base material, not the bra ' joint itself. lt can hence be concluded that most of the brazing materials(exce , maybe the bra ' materials comprising only 4% Si) are “strong enou ” - itis of no use providing a zing material that gives a strongerjoint than the m 'al itis supposed to join, since the resulting strength of the system brazing joint- basematerial will not be stronger than its weakest link, i.e. the base material.

Figure 21) G577. , coil thi tea with ou fiiier mateHeat treated at 1160 de rees C) Figure 28) Brazing material comprising 4% Si, grain size of 106 _ 0,5 mm coilthickness. As can be seen, in this test, the brazing material was weak link forthree out of fourjoints. lt should be borne in mind, however, that this test was madefor the weakest brazing material and the strongest base material. ___________ .~........... .\ Brazing material com sinpm PSD. All ruptures in brazing material.

Figure 30) Brazing material comprising 6% Si, 45 um grain size. All ruptures occurredin base material.

Ductility ofjoints. lt is a well known problem of brazed heat exchanger brazed with the prior art brazefillers comprising boron that they tend to be less ductile than copper brazed heatexchangers. This problem is, however, overcome according to the present invention.ln Fig. 31, the force vs. elongation for different brazing joints are shown. One brazingmaterial stands out from the rest, namely the prior art brazing joint comprising boron.As can be seen, the force required for elongating the brazing joint with this material issignificantly larger than for all the other brazing material, including copper, i.e. it isless ductile than copper. All brazing materials according to the invention are,however, about as ductile as the copper brazings. ............................................................ ._thickness, BT 1250 °C. \“\\\.w.\\»~ “w \\\\ lïigure 32) Force vs Elongation for “best off' tensile samples and 0,3 mm coilthickness, BT 1250 °C.

Burst pressures of prototype units Burst tests have also bee performed for actual heat exchangers. The testedexchangers comprise 18 plates. ln figure 33, some burst pressure for an identicalheat exchanger brazed with 106 um braze filler comprising 4%, 6% and 8% Si,respectively, a 45 um braze filler comprising 4% Si and a prior art braze fillercomprising boron (“Genll”) Are shown. As can be seen, the tested braze fillerscomprising 6 and 8 % Si show good results concerning burst pressure, whereas thebraze fillers comprising 4% do not. ln Fig. 34, comparative measurements are shown for brazing materials according tothe invention having 4, 6 and 8% Si, respectively, for plate thicknesses of 0.3, 0.4and 0.5 mm, respectively. A copper braze is shown for comparison. As can be seen,the copper brazing joints have comparable strength for 0.3 mm plate thickness, butthe brazing joints according to the invention are significantly stronger for thickerplates.

Bâï Bttrst garessure Fï-FB.. =Cycš=e fienïí:zo- -~ - Qs* “åært' F E' få: åEi å' _ | | | | iftïfi-ënnfy-wš-faåi -6>'~t»"=-5t fiïfi-finafy- <45nfy sannfiíler Figure 33) Burst pressures for B5T-18 units < 106my 4%§ < 106my 6%§sooo sooo7ooo- wow sss7z 55955 zä Ä ä sooo-E E å*4176 5000.4ooo-ä*| | | | | |0,3 0,4 o,s 0,3 0,4 o,sBM thidmess BM thidmess< 106n1y 7-8°/oSi Cu-f0i|sooo sooo\\\“7ooo- 664432\\\å \ §o9o,24 \\“ \\\*= sooo- 6000' ss41,7ë ä š* s.i iiwsooo- Ä4ooo-4ooo-ofs o,'4 ofs of: o,'4 ofsBM Thidmess BM thidmess Fig. 34 Joint strength plots for iron brazed brazing materials containing 4, 6 and 8%Si, respectively, for different base material thicknesses. A copper brazed joint isshown for comparison. Please note that the iron based brazing joints are significantiystronger than copper brazed joints for plate thicknesses of 0,4 and 0,5 mm.

Discussion Activated Diffusion Healing (ADH) Effect of Time-Temperature on sample strength The IDS samples heat treated at 1160 °C and 1210 °C, have lower mechanicalstrength compared to IDS heat treated at 1250 °C, figure 9. This is discussed in moredetail in RD101184. Low strength is also obtained for the pure 316L PM, heat treatedat 1250 °C, figure 9. Although diffusion bonding is obtained between the differentparticles of the 316L PM as well as between the 316L PM and the base material,figure 14, the high porosity of the joint results in low mechanical strength.

Figure 10) shows the effect of different “heat loads” on the same filler and basematerial. The three sub groups have been heat treated at 1250 °C. The sub-groupeindicating the lowest strength have been subjected to the highest “heat load", i.e. 2hours at 1250 °C + 3 hours at 1100 °C. The sub-groupe only subjected to 1,5 hoursat 1250 °C + 1 hour at 1150 °C give higher values. This may be due to stronger basematerial.

Joint strength for 0,3 mm coil thickness For the IDS-316 in 0,3 mm coil thickness, heat treated at 1250 °C, the rupture alwaysoccur in the base material in the vicinity of the joint, figures 13-15. Higher mechanicalstrength is obtained for these samples compared to samples heat treated at lowertemperatures, figure 9. Moreover, no significant difference in strength can be seenbetween fillers containing 4% Si, 6% Si or 8% Si, when subjected to the same “heatload”. No difference can be seen due to the particle size distribution of the filler either.See figure 11.

Joint strength for 0,4 mm coil thickness For the IDS-316 in 0,4 mm coil thickness, heat treated at 1250 °C, the rupture alsooccurs in the base material in the vicinity of the joint, figure 19-21. Higher mechanicalstrength is obtained for these samples compared to 0,3 mm coil thickness. Nosignificant difference can be seen between 4% and 6% Si. No significant differencecan be seen for <45 pm particle size compared to <106 pm particle size.

When the rupture occurs in the base material, the integrity of the joint is satisfactory.When the samples have been subjected to similar “heat loads”, the variation in jointstrength between the different fillers, as well as between the samples with the samefillers, is probably due to variations in the amount of applied paste. More paste resultsin largerjoints and bigger cross section areas. Figure 3 shows two samples aftercapillary dispensing, where the left sample gave 4085N and the right sample 4896N.As can be seen, the amount of applied paste is lower on the sample to the left, despite being applied at the same occasion. The fracture area can not be measuredaccurately after tensile testing, since the rupture runs through the base material, ascan be seen from figure 22 and 23.

The Genlll samples with s 6% Si have longer elongation prior to rupture compared tothe Genll samples and “boron free” samples with 7-8% Si. This indicates a higherductility for s 6% Si samples, figures 25-26.

Erosion versus Si-content The “Boron free” filler with 7,3% Si, have an approximate solidus temperature (Tsm) of1280 °C, figure 1. The TS0| for the filler with 8,3% Si is approx 1250 °C, figure 2.However, metallographic investigations of CP-1501 14-06 and CP-150223-01 stillshowed a substantial dissolution of the base metal at 1250 °C, figures 13 and14,despite heat treating temperatures being lower than measured TS0| of the fillers. Ascan be seen from figures 13 to 19, the erosion rate decreases with decreasing Si-content. Slight dissolution of the base metal can be seen for 6% Si at 1250 °C.Hardely any erosion for 4% Si.

Summary lt has been shown that it is possible to braze articles of stainless steel 316, morespecifically plate heat exchangers with a boron free, iron based brazing materialcomprising Si and Mn as melting point depressants. Moreover, it has been found thatthe article of stainless steel may be joined in temperatures lower than thetemperature at which the brazing material is completely melted. lnitial tests show, however, that brazing joints obtained at temperatures below themelting point of the brazing material, although sufficiently strong, are not fluid tight,probably due to micropores. Therefore, in the case of brazed plate heat exchangers,the circumferential skirt and the areas surrounding the port openings may be brazedwith a brazing material having a melting point lower than the brazing temperature (NB-this brazing material may comprise more silicone or even boron as melting pointdepressant), while the heat exchanging areas may be brazed with a brazing materialaccording to the invention.

Moreover, it has been found that it the particle size of the brazing material is crucialfor achieving the best results. Smaller particle sizes give stronger joints. Moreover,tests have shown that it is possible to join articles of stainless steel by applying small-grain particles of pure stainless steel between the surfaces to be joined. These jointsdo, however, not yet provide sufficient reliability and strength, but results indicate thatbetter results may be achieved by using even higher temperatures and smallergrains.

Boron free brazing material showing excellent binding properties between articles ofstainless steel when brazed at a temperature of 1250 degrees C comprises a metalcontent comprising: Trace amounts of C and S; 2.3% Mo; 12,3% Ni; 17,9% Cr; 6.4% Si; 4.8% Mn; Balance being iron (Fe), the metal content of the brazing material being in powderform and having a particle size less than 106 pm. The powdered metal content ismade into a paste by addition of 7% binder and solvents. All percentages are givenby weight.

The metal content of the brazing material according to the above may bemanufactured by mixing an alloy powder comprising: 2,3% Mo; 12,9% Ni; 18,4% Cr; 8,3% Si; 5,9% Mn, Balance being iron (Fe), with stainless steel powder in the ratio 75/25.

Concerning particle size, provisional results have shown that the properties of ablend of grain sizes of the metal powders comprised in the brazing material arecharacterised by the smallest particle size in terms of larger contact area per weight unit, significant pore size of the resulting joint, and the resulting creep characteristics.

This means that it is possible to attain the characteristics of a small grain materialeven for a blend of coarse (large size) particles and small size particles. Sincebrazing alloys having a coarse grit (i.e. large particles) are significantly lessexpensive than fine particles, this is beneficial from a cost standpoint.

Claims (21)

1. A brazing method for brazing article of stainless steel, comprising the steps of:Providing a brazing paste comprising a metal powder comprising: Trace amounts of C and S; 2-2,5% Mo; 12-13% Ni; 17,5-18,5% Cr; 6-8.3% Si; 4.5-5.5% Mn; all percentages being given by weight, balance being iron (Fe), the metal powder having a particle size less than 106 um,wherein the metal powder is made into a paste by addition of binder and solvents;applying the paste on or close to contact points between the stainless steel articles tobe brazed; placing the articles in a furnace; Heating the furnace to a temperature lower than the temperature at which the metalpowder of the brazing paste is completely melted; and Allowing the stainless steel articles to set by lowering the temperature.

2. The method according to claim 1, wherein the particle size is less than 45 um.

3. The method of claim 1 or 2, wherein the particle size is less than 20 um.

4. The method of claims 1-3, wherein the metal powder is manufactured by blendinga powder of an alloy comprising: 2-2,5% Mo; 12-13% Ni; 17,5-18,5% Cr; 6-8.3% Si; 4.5-5.5% Mn; Balance being iron (Fe), with stainless steel powder in a ratio giving the desired Siand Mn percentage.

5. The method according to claim 4, wherein the particle size of the stainless steelpowder is smaller than the particle size of the alloy.

6. The method of claim 5, wherein the particle size of the stainless steel powder isless than 22 um.

7. The method according to claims 1-6, wherein the temperature at which the metalpowder of the brazing paste is completely melted is above 1250 degrees C.

8. The method according to claims 1-7, wherein the heating includes the steps of:steadily raising the temperature from room temperature to a temperature at which thesolvents and/or binder evaporates; Keeping the temperature at which the solvents and/or binder evaporates until most orall of the binder an/or solvent is evaporated; elevating the temperature to the brazing temperature; and keeping the brazing temperature until the joints have been formed.

9. A brazed plate heat exchanger comprising a number of heat exchanger platesprovided with a pressed pattern of ridges and grooves adapted to keep the plates ona distance from one another under formation of interplate flow Channels, the heatexchanger further comprising port openings, wherein selective communicationbetween the interplate flow channels and the port openings is achieved by providingareas around the port openings of each plate on different heights, such that the areasaround the port openings of neighbouring plates either contact one another or do notcontact one another, hence blocking or allowing for communication between the portopening and the interplate channel, respectively, characterised in that the heatexchanger is at least partly brazed by a method according to claims 1 to 8.

10. The heat exchanger according to claim 9, wherein the areas around the portopenings that are contacting one another are provided with a brazing material havinga melting temperature lower than the brazing temperature.

11. The heat exchanger according to claim 10, wherein the brazing material providedat the areas around the port openings that are contacting one another is a brazingmaterial identical to the brazing material as defined in claim 1, however comprisingmore than 8% Si and/or up to 1,5% B.

12. A brazing alloy comprised in a brazing paste composed of a powder of said alloyand an organic binder and/or a solvent, said solvent being adapted to vaporize duringa brazing cycle, wherein said brazing alloy comprises: Trace amounts of C and S; 2-2,5%% Mo; 12-13% Ni; 17,5-18,5% Cr; 6-8.3% Si; 4.5-5.5% Mn; all percentages being given by weight, balance being iron (Fe).

13. The brazing alloy of claim 12, wherein the brazing alloy is a mechanical blend ofa stainless steel powder and a stainless alloy having a high percentage of Si and Mn.

14. The brazing alloy of claims 12 or 13, wherein the brazing alloy powder has aparticle size of less than 106 um.

15. The brazing alloy of claims 12 - 14, wherein the brazing alloy powder has aparticle size of less than 45 um.

16. The brazing alloy of claims 12-15, comprising:Trace amounts of C and S; 2,1-2,4% Mo; 12,2-12,8% Ni; 17,6-18,3% Cr; 6,2-6,8% Si; 4.6-5.3% Mn; all percentages being given by weight, balance being iron (Fe).

17. The brazing alloy of claims 12-16, comprising:Trace amounts of C and S; 2.1-2.4% Mo; 12,2%-12.4% Ni; 17,8%-18,1% Cr; 6.3%-6.7% Si; 4.7%-4.9% Mn all percentages being given by weight, balance being iron (Fe).

18. A brazing material comprising first alloy powder according to any of claims 12-17,characterised in that the first alloy powder has a first grit size differing from a grit sizeof a second alloy powder comprised in the brazing material.

19. The brazing material of claim 18, wherein the first grit size is smaller than 100pm.

20. The brazing material according to any of the claims 18 and 19, wherein thesecond grit size is smaller than 40 pm.

21. The brazing material according to any of the claims 18-20, wherein the first gritsize is smaller than 40 pm and the second grit size is smaller than 20 pm.