I. Introduction
Composition (in weight %) of the low-carbon, low-alloy P265GH-(“killed”) steel is C0.2; Si0.4; Mn 0.8-1.4; P0.025; S0.015; N0.012; Cu0.3; Mo0.08; Ni0.3; Cr0.3; V0.02; Nb0.2; Ti0.03. It is to be emphasized that chemical composition shown here corresponds to the maximum element content found in the specification for the P265GH-steel (Material No. 1.0425, according to DIN EN 10028-2). However, the actual composition of the hot-rolled P265GH-steel plates used in the manufacturing can be somewhat different, and vary (although within the compositional range listed above) from batch to batch. Materials entering fabrication process (machining) were supplied as plates in the annealed conditions (910 °C; +20 °C, -15 °C with the soaking time sufficient to attain the target temperature over the whole section, and subsequent cooling in still air). (Note: All certificates of the materials used are now available from the supplier). At this point it is also important to mention some “critical” temperatures for the P265GH-steel, viz. A1724 °C and A3860 °C.
Brazing of the steel parts (“Plate” and “U-shape”) using oxygen-free copper (OFC) as a filler metal was conducted in vacuum (~ 5 x 10-2 mbar) “graphite” furnace at 1120 °C.
In the present MEMO we will discuss results of metallographic examination of a number of P265GH-steel/Cu/P265GH-steel brazements taken from an “edge” as well as a “central” part of two different brazed products. Sample labeled A1 Mat-Tech designation MT 12-154) was taken from the central area of the product exhibiting after brazing distortion, whereas sample labeled B1 (Mat-Tech designation MT 12-156) was cut from the central area of the product where no apparent distortion was observed. Similarly, sample labeled A2 (Mat-Tech designation MT 12-155) is taken somewhere close to the “edge of the product” exhibiting distortion, and sample B2 (Mat-Tech designation MT 12-157) was taken from the “edge of the product” where no apparent distortion was visible.
After standard metallographic preparation cross-section of the joints were examined by optical microscopy. In order to reveal microstructural features in the low carbon, low-alloy P265GH-steel parts, the polished metallic surfaces were etched in 2% Amyl Nital. The use of amyl alcohol (instead of ethyl alcohol) minimizes the tendency towards the formation of etch pits and roughening of ferrite.
Additionally, in order to gain insight into interaction between liquid copper (filler metal) and P265GH-steel upon brazing, cross-sections of the brazement (sample MT 12-136 cut from brazed product at Mat-Tech) were also studied by Scanning Electron Microscopy (SEM) and Electron Probe Microanalysis (EPMA).
II. Macrostructural constituents
Ferrite
Apparently, Ferrite is the major constituent of the P265GH-steel. Ferrite is essentially iron containing (at room temperature) less than 0.005 wt. % of Carbon as well as some alloying elements such as Mn, Si, etc. An example of ferrite is seen in Fig. 1.
Fig. 1: Microstructure of equiaxed (polygonal) ferrite and pearlite (“dark” contrast) in the P265GH-steel/Cu/P265GH-steel joint labeled A1, Mat-Tech designation MT 12-154 (Upper part of the sample is “U-shape” and the bottom part is “Plate”). Some epitaxial ferrite is also present (arrows). The sample was taken from the “central” part of the brazed product exhibiting some apparent distortion. (Bright-field optical micrograph. 2% Amyl Nital etch.)
In most low-carbon steels, ferrite appears as an equiaxed morphology, which is also called polygonal ferrite. Ferrite can also appear in an elongated morphology in steels that have been rolled (especially after cold rolling). However, when this steel is annealed, these ferrite grains recrystallize into new equiaxed grains. Another form of ferrite is epitaxial ferrite, formed when a steel is heated into the two-phase ferrite + austenite region. During cooling, new ferrite can grow epitaxially on the existing ferrite grains. In our case, it is conceivable that the epitaxial ferrite can form in the P265GH-steel upon brazing since the standard cooling schedule involves soaking at 830 °C (within the two-phase ferrite + austenite region). An example of epitaxial ferrite can be seen (arrow) in Fig. 1.
Caveat 1: Because its low carbon content, ferrite is soft and easily deformed. This means that special care must be taken to avoid an excessive cold work (shear) during sample preparation (cutting, coarse grinding, etc) which may result in an artifact microstructure (e.g. elongated grains, broken carbides at sheared edge, etc.)
Cementite
When the carbon content of the steel exceeds the carbon solubility limit in ferrite, the excess carbon appears in the form of iron carbide Fe3C (cementite). Cementite is a hard phase and is usually not desired (in its pure form as large inclusions) in most low-carbon steels because it is hard and brittle and is detrimental to formability. No separate cementite precipitates are present within the microstructure of the P265GH-steel parts after brazing. Instead, the cementite phase is revealed in a different morphology, a constituent called pearlite (“dark” contrast in Fig. 1). Note that two-phase structure of the eutectoid is not resolved at this magnification.
Pearlite
In the P265GH-steel, the excess carbon appears in a distinctive lamellar morphology of pearlite, which is a two-phase constituent with alternating plates of cementite and ferrite. This morphology is classic for eutectoid transformation under slow (near-equilibrium) cooling. Appearance of pearlite in low-carbon, low-alloy P265GH-steel is seen in Fig. 1, where “dark” patches of pearlite appear at the ferrite grain boundaries. However, the lamellar morphology of the pearlite is not apparent in this micrograph, because the interlamellar spacing of the ferrite and cementite plates is not resolved at this magnification. At higher-magnification two-phase lamellar structure of the pearlite becomes visible (Fig. 2). Again, the P265GH-steel/Cu/P265GH-steel joint labeled A1 (Mat-Tech designation MT 12-154) is taken as an example.
Fig. 2: Microstructure of the P265GH-steel/Cu/P265GH-steel joint labeled A1 (MT 12-154) in the vicinity of the brazing seam. Two-phase lamellar structure of the pearlite patches is clear visible. The sample was taken from the “central” part of the brazed product exhibiting some distortion. (Bright-field optical micrograph. 2% Amyl Nital etch.)
Martensite
When low-carbon steel are moderately alloyed and rapidly cooled from the austenite phase field or the two-phase ferrite + austenite field, a constituent called martensite is formed. However, martensite is not found in the P265GH-steel after the brazing procedure; in fact in most cases of low-carbon steels it is highly undesirable.
Austenite
Other than dual-phase special high-strength steels, we should not see any retained austenite in low carbon steels.
At this point it seems worthwhile to make a few general comments about sampling for metallographic examination of rolled products.
III. Orientation of view
Before the sample preparation of any metallographic material can begin, we must consider which view (longitudinal, transverse, or planar) will yield the desired information. Fig. 3 shows a sketch representing these three views.
Fig. 3: Sketch showing the three planes of rolling. RD indicates rolling direction.
The longitudinal view will definitely reveal the most information about the material that has been rolled. This view, which is perpendicular to the surface and parallel to the rolling direction, more clearly illustrates the degree of grain elongation. For example, elongated ferrite
grains in the in the bottom steel plate of the P265GH-steel/Cu/P265GH-steel joint are shown in Fig. 4. These parts of the micrographs represent the longitudinal views of the polished and
Fig. 4: Microstructure of the P265GH-steel/Cu/P265GH-steel joints taken from central area of two different brazed products: a) exhibiting some distortion after brazing (labeled as A1; Mat-Tech designation MT 12-154) and b) without distortion (labeled as B1; MT 12-156). In both cases, elongated bands of pearlite (“dark” contrast) and ferrite grains are clearly visible in the longitudinal view of the bottom steel components (“Plate(s)”). Note that similar pattern can also be discerned in the microstructure of the upper steel part (“U-shape”) shown in (b). (Bright-field optical micrographs. 2% Amyl Nital etch.)
etched specimens of the hot-rolled and annealed P265GH-steel after machining (with subsequent stress-relieve) and further processing according with the brazing schedule. The ferrite grains are equiaxed in morphology but extend in long bands, as does the pearlite (“dark” constituent). This morphology of ferrite and pearlite is called banding.
At this point it seems worthwhile to elucidate a little bit on the “banding” in rolled products. In general, large steel ingots which have significant casting segregation frequently show alternate layers of ferrite and pearlite arranged in bands. This phenomenon of “banding” arises after heavy deformation usually by rolling that spreads out the regions in the cast structure that were either rich or poor in alloying additions. Since the diffusion coefficient of interstitial carbon is very high, the banding is unlikely to result from the casting segregation of carbon itself, but comes from those slowly diffusion alloy additions which effect austenite to ferrite transition temperature. Regions which have low concentration of austenite stabilizing elements such as Manganese will, on slow cooling, transform first to ferrite and reject their carbon content to the regions richer in Manganese that still have the austenite structure. Further cooling concentrates the carbon content in these Mn-rich regions until they transform to the eutectoid pearlite structure. This “banding” is reduced transverse ductility and fracture toughness of the alloy.
Somewhat similar pattern (long bands of pearlite) can also be discerned in the microstructure of the upper steel plates shown in Fig. 4b and Fig. 5, which is indicative for a longitudinal view of the rolled product.
Fig. 5: Microstructure of the P265GH-steel/Cu/P265GH-steel joints taken from the “edges” of two different brazed products: a) exhibiting some distortion after brazing (labeled as A2; MT 12-155) and b) without distortion (labeled as B2; MT 12-157). Elongated bands of pearlite (“dark” contrast) and ferrite grains are clearly visible in the longitudinal view of the upper steel part (“Plate”) in (a) and the bottom steel part (“Plate”) in (b). Note that similar pattern can also be discerned in the microstructure of the upper steel part (“U-shape”) shown in (b). (Bright-field optical micrographs. 2% Amyl Nital etch.)
On the other hand, the upper part in Fig. 4a (“U-shape”) and bottom part in Fig. 5a (“U-shape”) afford a much different view of the same plate. Perhaps, this is more close to a planar view, where the polishing is nearly parallel (?) to the bands of pearlite and ferrite.
Caveat 2: The rolling direction must be known for each P265GH-steel part within the brazing assembly so that proper information is presented in the polished specimen.
IV. Interaction between liquid copper (filler metal) and P265GH-steel upon brazing
Before we start to elucidate reaction phenomena occurring during fabrication of the P265GH-steel/Cu/P265GH-steel joints, one important observation has to be mentioned. Given geometry of the steel parts to be joined and the manner in which filler metal (Cu) is introduced in the joint, it was rather unexpected to find in the final brazement (often) continuous Cu-based layer between the steel components. This can be appreciated just looking at a serious of micrographs presented in the Fig. 6. One sees that continuous Cu-based interlayer is invariably present between the steel components in all samples examined. As a matter of fact, it turned out that here we are dealing with a certain type of a so-called “butt”-type joints with some “fillets” on either side of the brazing seam. It was found that the Cu-based interlayer varies in thickness and in some cases can be very irregular and even discontinues (Fig. 7).
It is not clear how this interlayer is formed during the brazing process, if no clearance (gap) between the initial components was (deliberately) provided (and maintained!). In principle, it is possible that liquid copper is spread along the interface between the steel parts of the brazing assembly given that the mating surfaces always have a certain roughness. Under condition of the brazing process liquid copper wets steel surfaces well. Moreover, the liquid filler metal can also penetrate along grain boundaries of ferrite. The result of this transport mechanism operation can be seen in Fig. 8. In the vicinity of the brazing seam most of the grain boundaries (of equiaxial ferrite) are “decorated” by the Cu-rich phase (“white” contrast on back-scattered electron micrographs). This type of “grain boundary penetration” can still be found inside of the steel component at distance of about 100 μm from the filler metal/steel interface. Another interesting observation is to be mentioned here. It was detected (in terms of EPMA) that Manganese which is present (up to 1.4 %) in the initial P265GH-steel is leaching from the base metal into the liquid Cu-rich melt. Most likely, this can be attributed to the high thermodynamic affinity of Mn towards Cu. The latter is manifested by the complete mutual solubility of these elements in the liquid as well as in the solid state.
Fig. 6: Microstructure of two different P265GH-steel/Cu/P265GH-steel joints in a) and b) samples were taken from the central part and at the “edge” of the product exhibiting some distortion after brazing (A1; MT 12-154 and A2; MT 12-155). In c) and d) samples were cut from the central part and at the “edge” of the brazed product without any apparent distortion (B1; MT 12-156 and B2; MT 12-157). Note that continuous Cu-based layer is invariably present between the steel components. (Bright-field optical micrographs. 2% Amyl Nital etch.)
a.
b.
c.
d.
Fig. 7: Microstructure of two different P265GH-steel/Cu/P265GH-steel joints showing different appearance of the Cu-based interlayer: in a) and b) samples were taken from the central part and at the “edge” of the product exhibiting some distortion after brazing (A1; MT 12-154 and A2; MT 12-155). In c) and d) samples were cut from the central part and at the “edge” of the brazed product without any apparent distortion (B1; MT 12-156 and B2; MT 12-157).
Fig. 8: Back-scattered Electron Images (BEI) of the P265GH-steel/Cu/P265GH-steel joint (Mat-Tech designation MT 12-136) in the vicinity of the brazing seam: a) general view and b) magnified area of the interaction zone close to the steel/filler metal interface. Aggregation of Cu-based material at grain boundaries of ferrite is clear visible. Note that precipitation of Fe-rich “dendrite-shape” particles occurs within the Cu-based interlayer upon cooling.
In this context, it also relevant to remind that in systems exhibiting isotropy of solid-liquid interfacial energy (like, for example, BBC-ferrite / liquid Cu), any region of liquid phase will aggregate at grain boundaries junctions and take up a uniform shape controlled by a unique dihedral angle, θ. The dihedral angle may vary between the limits 0° and 180°. There are two critical dihedral angles at which quite different characteristic microstructures are produced:
- if θ is greater than 60° the liquid will form discrete droplets along grain edges (boundaries) and at some grain corners
- if θ has a value between 0° and 60° most, if not all, of the grain edges and grain corners are replaced by continuous channels of liquid
- when θ = 0° (the condition of complete wetting) and if sufficient liquid phase is present, all high-angle grain boundaries disappear and the grains are separated from each other by thin film of liquid
Looking now again at the micrographs given in Fig. 8, it is easy to get convinced that we are dealing here with the second “limiting case”, although the present situation might be rather close to the “condition of complete wetting”, i.e. the corresponding dihedral angle has very low value.
Let us return to Fig. 7d. The Cu-based interlayer in this joint is quite uniform and of about 30 μm thick. Even taking into account all the possibilities explained above, there still remains a question: ”How such a thick continuous Cu-based layer can be formed within the brazing assembly?”
