- 學位論文
薄壁球墨鑄鐵表層異常組織之成因及對策以及對於熱疲勞性質之影響
Effect of the Abnormal Skin Structure on Thermal Fatigue Property of Thin-Wall Ductile Cast Iron
摘要
本研究是針對薄壁球墨鑄件，探討一些關鍵冶金及製程參數對於鑄件表層所形成之異常組織(如: 石墨退化、缺石墨區、全波來鐵層或全肥粒鐵層、碳化物等)的影響，並探討可以降低或消除該表層異常組織之方法。本研究所探討之冶金及製程參數包含: 澆鑄溫度、殘留鎂量、接種方法、造模材料、塗模劑種類等。最後，本研究亦深入分析異常表層組織之形成機制。 田口實驗分析結果顯示，對於高溫澆鑄(約1450˚C)之試片(2mm、3mm、6mm)而言，消除表層異常組織之最佳之鑄造條件為: 呋喃模之硬化劑/呋喃樹酯比例為60-80%、塗模劑為石墨、處理條件為球化劑1.2%+澆斗接種0.6%；另一方面，對於低溫澆鑄(約1400˚C)之相同試片而言，最佳之鑄造條件為: 呋喃模硬化劑/呋喃樹酯比例為80%、塗模劑為SiO2、處理條件為球化劑1.2%+澆斗接種0.6%。另一方面，由田口實驗分析結果可知，針對塗模劑種類而言，在所探討之三種不同塗模劑中，對於降低表層異常石墨組織之傾向，在高溫澆鑄時(約1450˚C)以石墨最佳，在低溫澆鑄時(約1400˚C)以SiO2最佳，而三者中以MgO稍差，惟，差異不大。但針對整體表層異常組織面積量之統計分析結果得知，使用MgO塗模劑對於降低表層異常組織的效果最優，其次為SiO2塗模劑及石墨塗模劑，而未使用塗模劑之效果最差。 針對2mm(固定球墨數約1380±100#/mm2)、3mm(固定球墨數約920±100#/mm2)及6mm(固定球墨數約654±58#/mm2)之試片而言，試片EGH-80%: (球化劑1.2%+接種劑0.6%、塗模劑為石墨、澆鑄溫度為1450˚C及硬化劑/呋喃樹酯比例為80%)具有最佳的耐熱疲勞性質，且耐熱疲勞性質隨表層石墨異常組織面積之增加而降低；另外，針對最佳參數組合之E爐次試片，在相同之澆鑄溫度條件下，熱疲勞壽命隨著鑄件厚度之增加而增加。另一方面，在相同鑄件厚度，但不同澆鑄溫度之條件下，高溫澆鑄之熱疲勞壽命均較低溫澆鑄者高。 試片在進行RT~800°C之熱疲勞試驗時，於第一回加熱過程中，部分肥粒鐵與大部分波來鐵會變態為沃斯田鐵，並於水淬火過程中變態為麻田散鐵；在第二回加熱過程中，部分肥粒鐵與部分麻田散鐵會變態為沃斯田鐵及回火麻田散鐵，在水淬火後，沃斯田鐵會變態為麻田散鐵，因此，最終形成肥粒鐵、麻田散鐵以及回火麻田散鐵之共存組織，且在表層異常石墨周圍幾乎為麻田散鐵。隨著熱循環次數的增加，表層之異常石墨中的C會持續沿著晶界擴散至晶界周圍的沃斯田鐵，且C會累積在晶界上或擴散至鄰近之非石墨區，使得冷卻後在異常石墨周圍形成較緻密的麻田散鐵(或回火麻田散鐵)。由於基地之回火麻田散鐵在熱循環過程中會逐漸析出C並沉積於晶界上而形成二次石墨，另一方面，基地之C含量逐漸降低而轉變為肥粒鐵。且在晶界所形成的二次石墨會逐漸與表層異常石墨串聯，又由於石墨與基地(肥粒鐵)之強度差異極大，在每一熱循環之冷卻過程中所產生的拉應力會逐漸累積而引發裂紋發生，且裂紋依但產生會沿著異常石墨及晶界上之二次石墨生長，直至破裂為止。
並列摘要
The primary purposes of this research are threefold: (1) to study the effects of metallurgical and processing parameters on the formation of the abnormal structure on the skin of the thin-wall ductile iron castings. The parameters examined include the hardener/resin ratio, type of coatings, the pouring temperature, and the addition amounts of nodularizer and inoculant, (2) to explore the mechanisms for the formation of the abnormal skin structure, and (3) to correlate the skin abnormal structure in terms of degraded area with the thermal fatigue life of the thin-wall ductile iron castings. Taguchi’s analysis results indicate that for castings (2mm, 3mm and 6mm in section thickness) poured at around 1450oC, the optimal casting conditions regarding the prevention of the abnormal skin structure are: the hardener/resin ratio of 60-80%, graphite coating, and 1.2% nodularizer+0.6% inoculant. On the other hand, for castings poured at around 1400oC, the optimal casting conditions are: the hardener/resin ratio of 80%, SiO2 coating, and 1.2% nodularizer+0.6% inoculant. In addition, among the three types of coating (graphite, SiO2 and MgO) investigated, graphite coating exhibits the most effective when poured at a relatively high pouring temperature (1450oC), while SiO2 is the most favorite coating for a relatively low pouring temperature (1400oC). Nevertheless, no significant difference in effectiveness on the prevention of the abnormal skin structure was observed among these three coatings, but MgO coating shows the least variations in the degraded area of the abnormal skin structure. Furthermore, significant improvement was obtained when a coating was applied as compared to the case without coating. Regarding the thermal fatigue property, the thermal fatigue life increases with decreasing the degraded area of the abnormal skin structure. For all the castings with three different section thicknesses, namely, 2mm (with fixed nodule count of 1380±100#/mm2), 3mm (with fixed nodule count of 920±100#/mm2) and 6mm (with fixed nodule count of 654±58#/mm2), specimen EGH-80% (1.2% nodularizer+0.6% inoculant, graphite coating, pouring temperature of 1450oC, and hardener/resin ratio of 80%) exhibited the best thermal fatigue property. In addition, increased thermal fatigue life was registered for castings poured at a relatively higher temperature, i.e., 1450oC, and with a thicker section, i.e., 6mm. When the fatigue specimens were subjected to thermal cycles between room temperature and 800oC, phase transformations took place and the microstructures varied accordingly. In the first thermal cycle, most of pearlite and part of ferrite in the as-cast condition transformed to austenite during the heating and holding stage, and then the austenite formed transformed into martensite during the subsequent cooling cycle (water quench). As a result, martensite and un-transformed ferrite were present after the first thermal cycle. In addition, the transformed martensite can be observed to be distributed mainly in areas immediately adjacent to graphite nodules and also in the vicinity of grain boundaries where the original pearlite phase was present. In the second thermal cycle, part of ferrite and martensite again transformed to austenite during the heating and holding stage, while at the same time the un-transformed martensite were tempered. In the subsequent cooling cycle, austenite once again transformed into martensite, and therefore, martensite, tempered martensite and un-transformed ferrite coexist in the microstructure. The afore-mentioned pattern of phase transformation continued to operate in the following thermal cycles. However, at certain stage in the course of cyclic thermal fatigue test, minute graphite particles, the so-called secondary graphite, started to precipitate at grain boundaries, owing to the repeated tempering of both martensite and tempered martensite. When the martensite was heated to the pre-set 800oC, but failed to transform to austenite due to the temperature being not high enough for phase transformation to take place, the martensite was tempered at a relatively high temperature (800oC) instead, causing the precipitation of secondary graphite along the grain boundaries. With the progress of the thermal cycles, the precipitation of the secondary graphite also continued, which resulted in a gradual reduction in the dissolved carbon content in the matrix. Consequently, the volume fraction of the ferrite phase increases, while at the same time the amounts of martensite and tempered martensite decrease. Under the influence of residual tensile stress developed in each thermal cycle, cracks tended to be initiated at the graphite-matrix interface and/or in the vicinity of the precipitated secondary graphite particles, and once cracks occurred, they propagated along the secondary graphite particles at the grain boundaries and also the interface between the degenerate graphite phase and the matrix.