1020碳鋼氮化鋁與碳化鈮–氮化鋁複合析出物析出行為與高溫抗沃斯田鐵晶粒粗化之研究

Abstract

滲碳處理是強化齒輪與螺栓等部件疲勞性質與耐磨耗特性的重要處理，使用微合金元素析出物能夠在高溫滲碳時細化沃斯田鐵晶粒。鋼材純化時會利用鋁進行脫氧作業，使得鋼材中通常會包含微量的鋁，若可利用鋼材中本就存在的鋁，使其於齒輪鋼常用製程中析出，並於最後滲碳時達到抗粗晶效果，則可省去添加其他微合金元素以及額外熱處理製程的成本。本研究使用1020碳鋼，於第四章至第六章中設計一高、一低鋁氮含量的兩個材料進行實驗，透過膨脹儀進行”熱軋後等溫析出實驗”，以及模擬實際部件製造階段的”熱軋後等速冷卻+再回溫析出實驗”，以了解AlN在沃斯田鐵為主或肥粒鐵為主基地中的析出行為，並再透過”偽滲碳沃斯田鐵晶粒度實驗”了解先前實驗所析出之AlN如何影響後續沃斯田鐵晶粒成長情況。於第七章中則額外添加鈮，並設計一等鋁氮比、一過鋁氮比的兩個材料進行實驗，並同樣透過”熱軋後等速冷卻+再回溫析出實驗”了解Nb(C, N)與AlN的析出行為，並再透過”偽滲碳沃斯田鐵晶粒度實驗”了解所析出之Nb(C, N)與AlN如何影響後續沃斯田鐵晶粒成長情況。 研究結果顯示AlN於沃斯田鐵基地中析出困難，主要藉由依附其他已存在基地中的Ti或MnS顆粒異質成核析出，僅能在高溫較長時間持溫下才能有機會生成獨立之AlN析出物，且析出尺寸較大(數百奈米至微米等級)、析出密度低(TEM中投影距離約數個微米)。相比之下，AlN於肥粒鐵基地中析出容易且機制多元，除了同樣可依附於已存在顆粒析出之外，亦可直接同質成核析出於肥粒鐵中，或是析出於肥粒鐵晶界、晶粒中差排、沃斯田鐵與肥粒鐵晶界等位置，其中最主要的析出方式是直接同質成核析出於肥粒鐵中，且析出尺寸較小(約30~60奈米)、析出密度較高。此析出結果亦反應於後續沃斯田鐵化組織上，由於沃斯田鐵基地析出困難，需要較多析出時間達到較好的細化晶粒效果，但析出情況較差，使後續回溫時較容易出現高比例的異常晶粒成長區域，而細化區域的晶粒尺寸約5~6微米。對比沃斯田鐵基地析出，肥粒鐵基地析出容易，使得較長析出時間反而可能使析出物粗化，降低細化晶粒效果，但較佳的析出情況使其有機會完全抑制異常晶粒成長，使晶粒尺寸細化至約2~3微米，顯示不增加額外齒輪鋼製程且僅使用AlN析出物仍可於後續高溫維持細小晶粒的可能性。 添加鈮之後由於Nb(C, N)具有較佳熱穩定性，因此無論鋁氮比、冷速快慢與回溫處理與否，900度沃斯田鐵化時均可達到完全細化晶粒效果，相比於無添加鈮材料仍可能於900度發生異常晶粒成長，其抗粗晶效果具有顯著的提升。而當沃斯田鐵化溫度提升至1000度，等鋁氮比鋼種雖然皆發生異常晶粒成長，但仍具一定晶粒細化效果，但細化效果受到750度回溫時析出物粗化影響而有所下降；過鋁氮比鋼種則已完全無細化晶粒效果，顯示過鋁氮比除了影響AlN熱穩定性，亦會影響Nb(C, N)於高溫的熱穩定性。後續晶體結構分析中亦發現Nb(C, N)與AlN至少具有兩種方位關係可相互生長，顯示二者間的強烈連結性，亦可呼應二者析出行為可能互相影響。

Carburizing is an important treatment for enhancing the fatigue resistance and wear properties of components such as gears and bolts. The use of microalloying precipitates can effectively refine austenite grains during high-temperature carburization. During steel refining, aluminum is commonly used as a deoxidizer, resulting in trace amounts of Al remaining in the steel. If this residual Al can be utilized to form AlN precipitates during conventional gear steel processing and effectively suppress grain coarsening during final carburization, it may eliminate the need for additional microalloying elements and extra heat treatment steps, thereby reducing processing costs. This study used 1020 carbon steel as base to design several materials. In Chapter 4 to 6, two material compositions were designed: one with high Al and N content and the other with low Al and N content. By using a dilatometer, a "Hot rolling – Isothermal heat treatment experiment" and a "Hot rolling – Constant rate cooling and reheating heat treatment experiment", simulating practical manufacturing conditions, were conducted to understand the behavior of AlN precipitation in both austenite-dominated and ferrite-dominated matrices. Additionally, a "pseudo-carburizing experiment" was performed to examine how AlN formed in previous treatments influences subsequent austenite grain growth. In Chapter 7, extra niobium was added, and two material compositions were designed: one with nearly equal atomic ratio of Al and N and the other with an over Al/N ratio. a "Constant rate cooling and reheating heat treatment experiment" were also conducted to understand the precipitation behavior of Nb(C, N) and AlN. Additionally, a "pseudo-carburizing experiment" was performed to examine how Nb(C, N) and AlN influences subsequent austenite grain growth. The results revealed that AlN precipitation in the austenite matrix is challenging, resulting in preferential heterogeneous nucleation on pre-existing Ti or MnS particles in the matrix. Independent AlN precipitates were only observed after prolonged holding at high temperature, exhibiting large sizes (hundreds of nanometers to the micron scale) and low densities (interparticle distances of several micrometers in TEM). In contrast, AlN precipitated more easily and with diverse mechanisms in the ferrite matrix. Precipitation behaviors included homogeneous nucleation within ferrite grains, heterogeneous nucleation on existing particles, at ferrite grain boundaries, dislocations, and austenite-ferrite boundaries. The primary mechanism was homogeneous nucleation within ferrite grains, producing smaller particles (30~60 nm) with higher densities. These precipitation behaviors directly influenced subsequent austenite microstructures. Due to the difficulty of AlN precipitation in the austenite matrix, longer holding times were required to achieve better grain refinement. However, the poor precipitation conditions resulted in a higher proportion of AGG regions during reheating, with refined grain sizes of approximately 5~6 μm in non-AGG areas. In comparison, precipitation in the ferrite matrix was more effective, though extended holding times could lead to precipitate coarsening and reduced grain refinement efficiency. Nonetheless, the superior precipitation behavior in ferrite enabled complete suppression of AGG under certain conditions, yielding fully refined grains with sizes as small as 2~3 μm. Upon Nb addition (~0.02 wt %), the high thermal stability of Nb(C, N) enables complete grain refinement during austenitization at 900 °C, regardless of Al/N ratio, cooling rate, or prior reheating treatment. This represents a marked improvement over Nb-free steels, which readily exhibit AGG under the same conditions. When the austenitization temperature is increased to 1000 °C, alloys with an approximately stoichiometric Al/N ratio still display a measurable refinement effect, although AGG occurs and the effectiveness is diminished by precipitate coarsening that develops during the preceding 750 °C reheating step. By contrast, alloys with an excess Al/N ratio show no grain-refinement capability at 1000 °C, indicating that an over-stoichiometric Al/N ratio not only reduces the thermal stability of AlN but also lowers that of Nb(C, N). Crystallographic analyses further reveal that Nb(C, N) and AlN can grow with at least two distinct orientation relationships, underscoring a strong mutual affinity that likely interacts their precipitation behaviors at elevated temperatures.