黃銅退火雙晶與高碳麻田散鐵相變雙晶之奈米顯微組織研究

Abstract

退火雙晶經常發生在f.c.c金屬或合金中，它是由於在再結晶成長的過程中，偶發的堆疊錯誤所造成的。因此，退火雙晶對材料機械性質的影響甚是重要。我們需要更加地了解退火雙晶的晶界與一般晶粒的晶界對強度的貢獻是否有差異。本實驗的硬度量測顯示雙晶晶界的平均硬度值比一般晶粒晶界的平均硬度值僅稍微低了5Hv。而且從TEM的圖片也可以看出雙晶晶界除了有橫滑移的現象產生外，也可以阻擋差排的滑移，並在晶界上產生階梯狀的凸起。此階梯狀的凸起可視為差排在雙晶晶界上分解的證據。由於差排在雙晶晶界上分解是能量不利的，且分解後的差排可留在晶界上合併釋放，因此致使雙晶晶界的存在可維持一定的強度和極優異的韌性。所以，雙晶的密度也是影響機械性質的另一個重要因素。我們做一連串不同溫度的熱處理去量測雙晶的密度。發現雙晶的密度跟晶粒的大小成正比。這個結果符合Pande所提出的經驗式：N=Kt ln(D/ D0)。本實驗以C2600黃銅(α-brass)為材料量測到Kt值接近0.3。 相變雙晶通常是為了維持麻田散鐵相變時所需的非均質晶格不變應變而形成的。板片狀麻田散鐵最特殊的特徵就是在中心由非常高密度的相變雙晶所組成的中脊面區域。中脊面從以前就被視為是相變最先產生的區域，但相關文獻卻不太普及。因此，本實驗利用DSC實驗及TEM觀察thin plate和lenticular麻田散鐵內部中脊面的細部特徵以及它最初的形貌。由DSC實驗和試片在液態氮中經由不同時間的深冷處理觀察到thin plate麻田散鐵比lenticular麻田散鐵先形成。因此，我們可以推斷thin plate麻田散鐵可以轉變為lenticular麻田散鐵。TEM照片顯示thin plate麻田散鐵內佈滿了跨越整個板片的雙晶，而lenticular麻田散鐵則是由中脊區、雙晶擴展區和非雙晶區所組成。所以，前者是藉由雙晶來達到晶格不變應變，而後者是同時藉由雙晶和滑移兩種形式來達到麻田散鐵相變所需維持的晶格不變應變。此外，在回火實驗中觀察到由許多密集雙晶所構成的中脊區域提供了碳化物最有利的析出位置。且由回火後麻田散鐵內的基地在中脊區域有一小角度的旋轉，可推測中脊區為一高應力集中區，回火後會在此產生應力釋放使得中脊區的兩旁晶格有一個小角度的旋轉。高碳高鉻的不鏽鋼合金在600˚C回火0.5 ~ 2小時先產生M3C碳化物而後再形成M23C6碳化物。實驗發現前者和麻田散鐵基地維持Bagaryatsky方位關係，後者則接近Kurdjumov-Sashs方位關係。

Annealing twins usually form as a consequence of growth accidents or are presumed to form on stacking faults during the recrystallization of fcc metals and alloys. Therefore, the effects of annealing twins on mechanical properties are very important. It is desirable to determine the difference in strength contributions between general grain and twin boundaries. The results of hardness measurements have shown that the hardness of the twin boundary is a little lower (about 5Hv) than that of the general grain boundary. TEM micrographs indicated that slip lines can penetrate twin boundaries by cross-slip, or if obstructed, form ledges at the twin boundaries. The observations of the ledges at twin boundaries provided evidence for the dislocation dissociations. The energetically unfavorable dissociated reactions and the coalescent partial dislocations released at the twin boundary contribute to the maintained strength and excellent ductility of the twin boundaries. Additionally, a series of annealing treatments at different temperatures were carried out to measure twin density. The results show that annealing twin density depends on grain size. Pande's experiential equation for calculating the annealing twin density (N=Kt ln(D/ D0)) agrees well with our experimental results. The material depending value (Kt value) in Pande's experiential equation for C2600 brass was measured at about 0.3. Transformation twins usually form in the high carbonic martensite transformation to maintain the inhomogeneous lattice - invariant. A typical characteristic of lenticular martensites is the appearance of an obvious high density twinned region (i.e., a midrib region). The midrib is considered to be the region where martensite transformation starts. In this work, it has been found that thin-plate and lenticular martensites co-existed in the specimens of Fe-1C-17Cr stainless steel. The substructures of thin plate martensites and lenticular martensites were examined using TEM, focusing on the details of the midrib region. The results of the DSC experiment and the course of the isothermal holding in the liquid nitrogen (-196˚C) indicated that the thin plate martensite formed first and lenticular martensite later. These results provide evidence to suggest that thin plate martensite can be transformed into lenticular martensite. Transmission electron microscopy revealed that thin plate martensite is composed of a set of internal transformation {112} twins crossing through the interior plate, while the lenticular martensite contained three subzones: the midrib region, extended twinned region, and untwinned region. The results obtained from TEM observations suggest that the transformations of thin plate martensite and lenticular martensite are initiated at the same midrib region. During the growth, the former keeps the lattice-invariant deformation mode of twinning, whereas the latter combines both twinning and slip modes. Additionally, the result of tempering experiments indicated that the midrib region of the martensite contained a large amount of twinned boundaries, which is the preferential position for carbide precipitations. TEM observation showed that tempering treatment resulted in the release of stress at the midrib region, i.e., the stress-concentrated region, and caused the martensite crystal to rotate slightly. TEM results indicated that M3C type carbide was dominant after tempering at 600˚C for 0.5 hours, but M23C6 type carbides was frequent after tempering for 1 and 2 hours. Analysis of diffraction patterns revealed that in this Fe-1C-17Cr alloy, Bagaryatsky OR was found between ferrite and M3C carbide, and Kurdjumov-Sashs OR was found between ferrite and M23C6 carbide.