Materials Processing
Microstructure and Mechanical Properties of YAG Laser Welded Spheroidal Graphite Cast Iron
2024 Volume 65 Issue 6 Pages 665-671
Details
2024 Volume 65 Issue 6 Pages 665-671
Cast iron welding improves the performance of joints using welding rods for cast iron. However, such rods are more than twice as expensive as those for mild steel. The authors found that when spheroidal graphite cast iron is melted and rapidly solidified, spherical graphite nodules aggregate and moved to the surface. If the spherical graphite nodules in the fusion zone moved to the surface, the carbon concentration may reduce, thereby suppressing the generation of ledeburite. In this study, by irradiating with YAG laser, I-type butt-welded joints were fabricated using L materials (with an average spherical graphite nodule diameter of 52 µm) specimen without groove width. Ferrite and pearlite matrices joints were undermatch joints. Applying PWHT (Post Weld Heat Treatment) restored the joint strength to more than 90% that of the base metal. The impact value as-welded joint dropped to less than 15% of that for the base metal. However, it could be recovered up to about 40% of that for the base metal by PWHT. Ferrite matrix joints after PWHT exhibited similar level impact values as the as-welded pearlite matrix joints. Moreover, the surface of spheroidal graphite cast iron with different numbers and size of spherical graphite nodules was irradiated with YAG laser. As the result, at irradiation feed rates above 100 cm/min L materials specimen exhibited a much lower hardness in the fusion zone than S materials (with an average spherical graphite nodule diameter of 27 µm) specimen. The spherical graphite nodules in the fusion zone moved to the surface during rapid melting by laser irradiation. In the S materials specimen studied, more ledeburite, martensite and retained austenite formed near the fusion boundary than with the L materials specimen.
This Paper was Originally Published in Japanese in J. Japan Foundry Engineering Society 95 (2023) 615–621.
While welding spheroidal graphite cast iron, welding defects such as porosity [1] and cracks [2] are formed. Porosity is caused by high-energy-density beams and tend to be formed in the fusion zone. Smaller porosity is formed by the diffusion of CO gas, which depends on the amount of carbon. Small defects are caused by the formation of chills in the fusion zone or hardening of the heat-affected zone, and are factors that reduce the toughness of the weld. Therefore, it is well known that spheroidal graphite cast iron is a low-weldability material. Cold welding, a common welding method for cast iron, employs arc, TIG (Tungsten Inert Gas), or MIG (Metal Inert Gas) welding without preheating or with localized low-temperature preheating. Various measures are taken to avoid cooling and hardening of the heat-affected zone in cast iron. For chilled cast iron, the fusion zone is diluted by welding with a Ni-based welding rod [3]. This further promotes graphitization and suppresses the formation of ledeburite. In addition, to avoid hardening of the heat-affected zone, preheating at 400 to 700°C [4, 5] is applied to suppress the martensitic transformation. Umetani et al. investigated the transformation process for ferrite matrix spheroidal graphite cast iron strengthened by a solid solution of Si subjected to thermal-cycle welding. As the number of spherical graphite nodules increased, the ferrite matrix spheroidal graphite cast iron promoted the formation of ferrite during the cooling process and suppressed the crystallization of ledeburite. The results showed that ledeburite can undergo complete graphitization during a short heat treatment [6]. Attempts have been made to reduce chill and heat-affected zone hardening of cast iron welds or reduce chill and martensite formation regions in order to improve joint performance. These attempts also reduce the welding heat input, minimize the width of the chill zone and heat affected zone, and suppress hardening of the heat affected zone. Efforts have been made to commercialize products welded using these methods [7]. However, welding rods for cast iron are more than twice as expensive as those for mild steel. We found that when a small piece of spheroidal graphite cast iron is subjected to a large current in a short period of time, it melts due to Joule heating, then rapidly solidifies, and the spherical graphite nodule aggregates and migrates to the surface [8]. The mechanism involved is unclear, although it is thought that the spherical graphite nodule aggregated due to surface tension and moved to the surface of the fusion zone, presumably as a result of the buoyancy of spherical graphite nodule, the poor wettability between spherical graphite nodule and the melt [9], and Marangoni convection [10] due to the carbon concentration distribution around spherical graphite nodule. It is expected that when the spherical graphite nodule in the fusion zone moves to the surface, the carbon concentration will decrease and the formation of ledeburite will be suppressed. This fusion bonding method is expected to improve the joint efficiency. In this study, butt welded joints of spheroidal graphite cast iron were fabricated by YAG laser irradiation without using a welding rod, and tensile and impact tests were carried out. Furthermore, spheroidal graphite cast iron was rapidly melted and solidified, and its structure and hardness were investigated.
In this study, 20-mm-thick FCD700 (100% pearlite) and FCD400 (20% ferrite) spheroidal graphite cast iron plates and 4-mm-thick FCD380 (100% ferrite) spheroidal graphite cast iron plates were used. Heat treatment was performed to adjust the microstructure in the specimens. In the present study, spheroidal graphite cast iron specimens with a small average spherical graphite nodule diameter (27 µm) and a large average spherical graphite nodule diameter (52 µm) are referred to as S specimens and L specimens, respectively. The details of materials are shown in Table 1. Figure 1 shows optical micrographs of the two types of ferrite matrix materials used. Specimens with different thicknesses (20 × 45 × 3 t and 20 × 45 × 10 t) were prepared from these materials. CO2 lasers have the ability to modify the microstructure and composition of near-surface regions without affecting the bulk, which is referred to as laser surface engineering [11, 12]. They are also capable of laser surface melting (LSM) and laser surface hardening (LSH), by rapidly heating the surface with and without melting, respectively. Tan et al. [13] reported that the surface hardness and abrasive wear resistance of spheroidal graphite cast iron can be significantly improved by LSH. Furthermore, there have been reports on mathematical modeling of the local melting of spherical graphite nodule and laser surface engineering to improve the wear resistance of austempered spheroidal graphite cast iron using a CO2 laser [14, 15]. On the other hand, YAG lasers melt bulk materials and are suitable for metal processing such as welding. In the present study, the surface of the specimen was melted by YAG laser irradiation, and the structure and hardness of the cross-section of the fusion zone were determined with an optical microscope and a Vickers hardness tester, respectively. The YAG laser processing machine used an MW2000 manufactured by Sumitomo Heavy Industries, Ltd., with a laser output power of 1.6 kW (continuous wave), a 5° tilt, a defocus distance of 10 mm outward from the focus, and a feed rate of 5 to 200 cm/min. Argon gas (20 l/min) was used as a shielding gas.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/F-M2024807tab01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Optical micrographs of ferrite matrix materials. S materials (a) and L materials (b).
For butt welding, I-type butt welding was performed using 3-mm-thick L specimens without groove processing. The welding conditions were a laser power of 1.6 kW, a defocusing distance of 10 mm and a feed rate of 30 cm/min. One welding pass was performed on each side. Tensile test specimens and impact test specimens with the dimensions shown in Fig. 2 were prepared from the welded parts. For tensile tests, a maximum of four tests were performed on each specimen at a crosshead speed of 8.3 × 10−3 mm/s using an Instron-type universal testing machine (IS-5000 manufactured by Shimadzu Corporation). For the impact test, a maximum of six tests were performed on each specimen using a Charpy impact tester (Tokyo Koki Co. Ltd.) with a capacity of 100 J. The test was conducted at room temperature, and the lifting angle of the striking hammer was 90°.
Specimens prepared from butt-welded parts for tensile test (a) and impact test (b). Gray filled patterns show weld area near center of specimen.
Figure 3 shows optical micrographs of cross-sections of 3-mm-thick ferrite matrix S and L specimens melted at a feed rate of 30 cm/min. The fusion zone was formed by heat conduction. This is the result of adjusting the laser defocus distance to suppress the heat energy density in the fusion zone. The fusion zone consists of primary dendrite, ledeburite, and retained austenite resulting from a pearlite transformation during cooling. The fusion depth in the S specimen was approximately 1.5 times that in the L specimen. Since the contact area between the fusion zone and spherical graphite nodule in the S specimen is larger than that in the L specimen, it is thought that the spherical graphite nodule fused into the fusion zone rapidly due to heating by laser irradiation. Furthermore, the laser beam absorption rate is much higher for liquid metal than for solid metal [16], so it is thought that the fusion depth was larger in the S specimen than in the L specimen. Figure 4 shows optical micrographs near the fusion boundary for the ferrite matrix materials. In the L specimen, ledeburite was formed around the spherical graphite nodule near the fusion boundary, and martensite was formed outside it. By contrast, the ledeburite, martensite and retained austenite that formed in the S specimen were distributed along the fusion boundary. This is because for the S specimen, the distance between spherical graphite nodules is small, and the whole matrix near the fusion boundary tends to transform into austenite. However, a large amount of ledeburite and martensite are formed in band-like areas around the fusion boundary, and it is thought that these do not contribute to the improvement of joint efficiency.
Optical micrographs of cross-sections of ferrite matrix materials irradiated at 30 cm/min. S (a) and L (b) specimens.
Optical micrographs near fusion boundary on base materials irradiated at 30 cm/min. S (a) and L (b) specimens. White dashed line indicates fusion boundary.
Figure 5 shows the tensile strength of the ferrite matrix and pearlite matrix base metals, the as-welded specimens, and the L specimens subjected to a PWHT (Post Weld Heat Treatment) at 873 K for 5.4 ks. The tensile strengths of the ferrite matrix and pearlite matrix base metal specimens were 431 MPa and 750 MPa, respectively. The joints at both matrices were undermatched joints. The tensile strength of the as-welded specimen was approximately 80% of that of the base metal. When PWHT was applied, the tensile strength of the joint almost recovered to the strength of the base metal. PWHT has been shown to be effective in making the tensile strength of the joint comparable to that of the base metal.
Tensile strength of base metal, as-welded specimen, and specimen after PWHT.
Figure 6 shows the results of a Charpy impact test at room temperature for ferrite matrix and pearlite matrix base metals, as-welded, and PWHT-treated L specimens. The impact values for ferrite matrix and pearlite matrix base metal specimens were 93 J/cm2 and 36 J/cm2, respectively. The impact value for the as-welded specimen decreased to less than 15% of that of the base metal. However, when PWHT was applied, the impact value recovered to approximately 40% of that of the base metal, and the ferrite matrix specimen showed an impact value comparable to that for the pearlite matrix base metal. All test specimens fractured at the weld boundary, and it is expected that the impact value for the joint will further improve if the microstructure at the weld boundary is improved.
Charpy impact values for base metal, as-welded specimen, and specimen after PWHT at room temperature.
Figure 7 shows optical micrographs obtained after tensile testing. The ferrite matrix specimen fractured near the center of the fusion zone. The as-welded fusion zone consists of primary dendrite transformed into pearlite, retained austenite and ledeburite. It is thought that a crack formed in the center of the fusion zone, and fracture occurred owing to segregation and defects. This is due to the release of constituent elements accompanying the growth of austenite dendrites during solidification of the fusion zone. In contrast, the pearlite matrix specimen fractured at the weld interface. Similarly to the ferrite matrix specimen, segregation and defects would have occurred in the center of the fusion zone, but the pearlite matrix specimen presumably fractured because of cracks in the hardened coarse-grain region of the weld interface. It was found that ferrite matrix and pearlite matrix specimens after PWHT fractured from the weld interface of the second bead through the fusion zone of the first bead. At the weld interface of the second bead, carbon diffused from the spherical graphite nodule into the matrix due to laser irradiation combined with the residual heat from the first bead. As a result, the carbon concentration in the matrix increased, owing to lowering of the melting point, which led to fusion [17], and a large amount of ledeburite was formed.
Optical micrographs around fracture after tensile test. As-welded (a) and (b) and PWHT (c) and (d) specimens.
Figure 8 shows optical micrographs obtained after impact testing. The ferrite matrix and pearlite matrix specimens fractured in the martensite region formed at the weld interface. The ferrite matrix specimen after PWHT shows that spherical graphite nodules deformed simultaneously with the ferrite matrix, and the impact value improved after recovery. In the pearlite matrix specimen, cracks grew only along the weld interface of the second bead. Furthermore, the crack penetrated the first bead from the bottom of the second bead. The recovered impact value of pearlite matrix specimen did not reach the impact value for the ferrite matrix specimen after PWHT.
Optical micrographs around fracture after impact test. As-welded (a) and (b) and PWHT (c) and (d) specimens. Area surrounded by solid line is deformed in Fig. 8(c).
Figures 9(a) and 9(b) show cross-sectional optical micrographs near the surface of the fusion zone of 3-mm-thick S and L specimens fused at a feed rate of 30 cm/min. It was difficult to confirm of the spherical graphite migration at the S specimen because the spherical graphite nodule diameter of original material was small. Spherical graphite nodule did not migrate to the surface of the fusion zone in the ferrite matrix and pearlite matrix specimens (Fig. 9(a)). On the other hand, in the ferrite matrix and pearlite matrix L specimens, spherical graphite nodule did migrate to the surface of the fusion zone (Fig. 9(b)). Thus, the floating spherical graphite nodule was spherical graphite nodule that migrated during the rapid fusing and subsequent solidification caused by laser irradiation. In addition, when the spherical graphite nodule diameter was small, it tended to fuse and disappear in the fusion zone. On the other hand, when the diameter of the spherical graphite nodules was large, some of the spherical graphite nodules presumably moved to the surface of the fusion zone while some of it fused, as a result of which the volume of the spherical graphite nodule decreased (Fig. 9(c)). Spherical graphite nodule moved to the surface of the fusion zone and remained undissolved in the fusion metal. Thus, it is thought that more spherical graphite nodule migrated out of the fusion zone than the spherical graphite nodule present on the surface of the fusion zone. Figure 10 shows the hardness of the fusion zones for 3-mm-thick ferrite matrix S specimens (○) and L specimens (◇) as a function of the feed rate during irradiation. When the feed rate was below 50 cm/min, the hardnesses of both the S and L specimens increased owing to the formation of martensite. On the other hand, when the feed rate exceeded 50 cm/min, the hardness of the S specimen decreased slightly. Furthermore, as the feed rate increased, the hardness of the L specimen decreased more significantly than that of the S specimen. This is thought to be due to a decrease in the carbon concentration in the fusion zone. At a feed rate of 30 cm/min, the fusion zone hardnesses were, respectively, 800 HV and 821 HV in the ferrite matrix and pearlite matrix L specimen joints. The former fractured near the center of the fusion zone in the tensile test, while the weld interface broke in the impact test. The latter fractured in the bonded portion during both the tensile and impact tests. The fusion zone is thought to have a higher hardness in the latter specimen, owing to the formation of martensite, which contains more carbon. It is thought that in the latter specimen, the coarse grains near the weld interface acquired a considerable amount of carbon during welding, and the carbon-rich martensite that was formed then caused fracture at the weld interface in both the tensile and impact tests. When subjected to PWHT, ferrite matrix and pearlite matrix L specimen joints both showed a tensile strength comparable to that of the base metal. On the other hand, both impact values recovered to only about 38% of that for the base metal. Although the welding residual stress was removed by PWHT, it is thought that the embrittlement factor remained for the weld interface structure due to the impact force.
Optical micrographs of cross-sections near surface of fusion zone. S (a) and L (b) specimens. Photo (c) is optical micrograph of ferrite matrix materials before melting. This photo shows the size comparison of spherical graphite nodules after melting.
Change in hardness of fusion zone of ferrite matrices S and L specimens as function of various feed rates of irradiation.
Ferrite matrix and pearlite matrix L specimens with a thickness of 10 mm were melted at a feed rate of 5 cm/min. Figure 11 shows the variation in hardness from the fusion zone to the heat-affected zone to the base metal. In the pearlite matrix (●), the hardness temporarily increased from the fusion boundary toward the base metal, and then decreased. This indicated that the heat-affected zone of the test piece was approximately 0.5 mm wide. On the other hand, the hardness of the ferrite matrix (◇) decreases from the fusion boundary toward the base metal, indicating that the heat-affected zone became narrower. TIG arc welding generates a heat-affected zone with a width of approximately 1 mm, but since laser irradiation has a more localized heat input than TIG arc welding, the heat-affected zone is narrower. Figure 12 shows the variation in the width of the heat-affected zone with feed rate for a 10-mm-thick pearlite matrix L specimen that was fused. The heat-affected zone gradually narrows as the feed rate increases, making it a more desirable structure for welds. It has been reported that S specimens containing a large amount of spherical graphite nodule are more suitable for arc welding [18, 19]. This contradicts the results obtained with laser fusing. In laser welding, L specimens are more suitable for welding owing to the structure of the heat-affected zone. Arc welding involves a molten pool, which distinguishes it from laser welding. In arc welding, there are four driving force-induced flow directions in the molten pool.
Change in hardness from fusion zone through heat-affected zone to base materials as function of distance from weld interface.
Change in width of heat-affected zone of pearlite matrix L specimens as function of feed rate of irradiation.
Figure 13 shows an overview of the driving force for the molten pool due to the arc heating. The driving forces are the drag force of the plasma airflow on the surface of the molten pool (a), the buoyancy force due to the density difference (b), the electromagnetic force due to the current flow (c), and the Marangoni force due to the surface-tension difference (d). However, (a) and (c) do not occur in laser welding. The macroscopic convection within the molten pool is determined by the delicate balance of these four driving forces. Furthermore, the formation of the molten pool is controlled through heat transport by convection. Therefore, the L and S specimens experience different driving forces for convection between their spherical graphite nodules, which may lead to a difference in heat transport between the spherical graphite nodules of L and S specimens during laser welding.
Overview of flow in molten pool by arc plasma [20]. Drag force (a), buoyancy (b), electromagnetic force (c) and Marangoni force (d).
The surfaces of spheroidal graphite cast iron specimens having a ferrite or pearlite matrix with different spherical graphite nodule diameters were irradiated using a YAG laser. Welded joints were fabricated, and tensile and impact tests were conducted on these joints. The authors performed melting by changing the feed rate during laser irradiation, and examined the structure and hardness of the melted part. Fusion was induced by changing the feed rate, and the structure and hardness of the fusion zone were investigated. The results are summarized below.
I would like to express my deepest gratitude to (now Emeritus) Professor Katsuhiko Kishitake of Kyushu Institute of Technology for his guidance during this research. I would also like to express my sincere thanks to Professor Emeritus Keisaku Ogi of Kyushu University and Dr. Takuro Umetani of Hinode, Ltd. for their helpful suggestions while writing this paper.
