行政院國家科學委員會補助專題研究計畫 ▓ 成 果 報 告
□期中進度報告
對流效應對球墨鑄鐵單向凝固的影響
The effect of flow convection on directional solidification of ductile irons
計畫類別：▓個別型計畫
計畫編號：NSC 92－2216－E －011 －028－
執行期間： 92 年 08 月 01 日至 93 年 07 月 31 日 計畫主持人：雷添壽
計畫參與人員：張文雄、楊棟賢、王信義、邱銘祿、廖宏彬
成果報告類型(依經費核定清單規定繳交)：▓精簡報告
本成果報告包括以下應繳交之附件：
▓出席國際學術會議心得報告及發表之論文各一份
處理方式：得立即公開查詢
執行單位：國立台灣科技大學機械工程系
中 華 民 國 92 年 10 月 29 日
行政院國家科學委員會專題研究計畫成果報告
對流效應對球墨鑄鐵單向凝固的影響究
The effect of flow convection on directional solidification of ductile irons
計畫編號：NSC 92-2216-E-011-028- 執行期限：92 年 8 月 1 日至 92 年 7 月 31 日 計畫主持人：雷添壽 國立台灣科技大學機械系 計畫參與人員：張文雄、楊棟賢、王信義、邱銘錄、廖宏彬
摘要
本研究報告是探討凝固過程中因溫度差異及重力而引起的流體對流效應對球墨鑄鐵材 質的影響。研究過程是先建構一套凝固鋒面分別往上及徃下的凝固設備，它是將冷激水套 安置在垂直 Furan 鑄模的中間段，以建構出上、下兩區段的單向凝固狀態。實驗結果發現，
在下區段的冷卻速率比上區段來的緩和，這主要是在下區段的流體對流較強所致。對流的 混合效果會使流體溫度較為均勻並降低偏析性很強的元素如鉬等的偏析程度。本報告將分 析經由 SEM 及 EBSD 等設備的觀察結果，以探討對流效應對單向凝固球墨鑄鐵的顯微結 構、球墨品質、元素偏析以及碳化物成長方向的影響。
關鍵詞：流體對流、單向凝固、球墨鑄鐵。
Abstract
This research report is aimed to study the effect of flow convection due to the coupling result of temperature gradient and gravity during the directional solidification of ductile irons.
A vertical cylindrical casting set-up has been designed to provide two zones of solidification simultaneous with two moving fronts directionally upward and directionally downward respectively. This two zones directional solidification casting was composed with a cylindrical Furan sand mold with a water chilled center-holed-block at the middle of the casting.
Experimentally, it was observed that the cooling rates at different locations of the downward zone was much more even than that of the upward one and this is considered as the result of more even temperature distribution along the downward front moving zone due to a stronger fluid convection. Fluid convection tends to mix up the liquid and resulted in a more even temperature distribution and a less degree of segregation of element molybdenum which has a strong tendency to segregate to the carbide in ductile iron. The characteristic of matrix microstructure, the quality of nodular graphite, element segregation and growing direction of carbide of this directionally solidified ductile iron also had been investigated with SEM and EBSD technique.
Key words：Fluid convection, Directional solidification, Ductile iron
Introduction
The solidification of commercial ductile iron is a rather complicated system, and the solidified microstructure of ductile iron depends on many factors such as cooling rate, iron chemistry and foundry practices. The effect of bottom chilling in a directional solidification on the segregations of alloying elements of ductile iron castings had been investigated [1]. In this paper the effect of fluid convection on segregations is the main goal of investigation. A directional solidification system consists of a vertical cylindrical casting set-up has been designed to provide two zones of solidification simultaneous with two moving fronts directionally upward and directionally downward respectively. This two zones directional solidification casting was composed with a cylindrical Furan sand mold with a water chilled center-holed-block at the middle of the casting. Theoretically, due to the effect of gravity and the density difference between hot and cold liquid the solidification front moving upward will experience less even no fluid convection relative to its downward counterpart.
Experimental
The instrumental set-up for the two zones directional solidification of casting is shown as in Fig.1(a) which consists of two sections of Furan sand mold and sandwiched with a chilled copper mold. This solidification set-up provides two zones with different solidification direction: an upward zone above the chilled mold and a downward zone below the chilled mold. The cooling curves of the casting at different locations were measured from the embedded K-type thermal couples and recorded in a PC. After pouring from top, exothermic compound was added to cover the casting and mold to prevent the heat loss from the top. Since the heat of casting extracted by the copper mold is much faster than the loss to the furan sand mold, this set-up provides a satisfactory result for directional heat flow. Fig. 1(b) and 1(c) show the separated chilling copper molds which set above the bottom section and the casting cylinders respectively. For the two castings the one with a smooth middle section was chilled with the copper mold.
The cast iron was melted with an induction furnace in a charge of 45 kg and treated with nodulizing materials and inoculants before pouring. Chemical compositions of three ductile irons analyzed from chilled specimens are shown in Table 1. It was not alloyed in iron A, but for iron B being alloyed with elements of Cr and Mo. In iron C, the rare earth elements (Ce+La+Nd) were added together with Mg bearing nodulizing materials with a target value of 0.08%. For these three irons, the Iron A was easy for machining, but Iron C was real hard for machining. Casting blocks were sectioned vertically and then prepared for further examinations with OM, FE-SEM of JSM-6500 and EPMA of JXA-8600SX.
Results and discussions
Fig. 2(a) shows a set of cooling curves recorded at different locations of the ductile iron A with a diameter of 50 mm and a length about 250 mm, the cooling curve marked with 0 is the temperature history at the center point of the chilling block, the curves marked with – and + are the ones located on the downward zone and upward zone respectively. This typical cooling curves show that the liquid temperatures of the downward zone reached to same level before the start of solidification as shown in Fig. 2(b) more clearly, but the upward zone had a temperature gradient existed as shown in Fig. 2(c). This difference in the temperature distribution of downward and upward zone could only be contributed by the coupling effect of gravity and liquid convection, relatively a stronger liquid convection existed in the downward zone which mixing up the liquid and resulted in a much more even temperature distribution than the upward ward zone.
Microstructures and Segregation:
The microstructures and its mapping with Mo element of casting iron B at upward and downward solidification zone are shown in Fig. 3, and the chemical analysis of these structures with an EPMA of JEOL Super probe JXA-8600SX is listed in Table 2. The large chunks are cementite with a Mo around 1~4 at.%, but the bright phase with a finer structure having a Mo content greater than 20 at.% especially on the center area of section 9 which near the location of cooling curve marked with “+2” of Fig. 2. Several publications indicated the existing of Mo-rich carbides in ductile irons with Mo as the alloying element [2,3].
From the data shown in Table 2, the atomic ratio of carbon in this Mo-rich compounds are around 50% indicated these compound could be carbides with MC type structure. Table 2 also shows that the concentration of Mo of the carbide in the center area of section 3 which near the location of cooling curve marked with “-2” of Fig. 2 is in a lower level than its counter location part of section 9, and this difference could be the result of a stronger fluid convection in the downward solidifying zone. Section No.6 is the one within copper mold having a lowest segregation of Mo due to a fast cooling rate.
The Mo-rich fine carbides with some faceted faces as shown in Fig. 3 (a) only have a thickness of about 1 micro-meter, and this could be the last drop of liquid enriched in Mo which solidified in a very fast cooling rate.
From Table 2, it is interesting to note that the segregation tendency of element P is same as Mo and segregate together with the Mo, however the element Cr shows different tendency of segregation which segregate into the larger carbides, the atomic ratio of carbon in these carbides are around 25% indicated these compound could be carbides with M3C type structure or cementite structure.
Fig. 4 shows the microstructures of ductile iron B from section which near the chilling copper mold with a FE-SEM of JSM-6500. The microstructure of the matrix mainly is pearlite structure resulted from the eutectoid reaction, but in a higher magnification, there is a different structure from pearlite as shown in Fig.4 (b). This structure only occupies a small area and located between carbides; from its feature this structure more like the ausferrite structure resulted from a transformation of austempering during a continuous cooling stage [4]. The ausferrite shown in Fig.4 (b) is not quite as sharp as that transformed with an isothermal austempering [5].
It is interesting to note that in there several areas in Fig. 4 (a) can be identified out this ausferrite structure.
Growth of Carbides:
Microstructures as shown in Fig. 5 is a eutectic cementite oriented longitudinally which grew in a location about 2 mm from the Furan mold and about 20 mm from the top surface of the copper mold. The bottom side of the picture is the Furan mold and the left side of the picture is the copper mold. It seems difficult to tell the growing direction of the edgewise and sidewise growth of the cementite, but it does show out the branches within the sidewise growth, in theory it means that different degrees of under cooling were encountered during the growth [6]. The Oxford Inca Crystal System was used to study the orientation of the growing direction and it turned out the surface being family of {4 1 1} plane. However, much more investigation will be done in the future.
Quality of graphite nodules:
Fig. 6 shows the microstructures of graphite nodules at three sections. Section No.9, No.6 and No.3 are corresponding to the same sections discussed previously as in Table 2 about topics of chemical segregation. The graphite nodules shown in left hand column are smaller and in good quality, but in the right hand column are relatively huge and exploded. Section No.9 and No. 3 are roughly located at the middle of the upper and lower solidification zone respectively, in practice it is not easy to explain this, since the local solidification time were only about several minutes as shown in Fig.2. It would be necessary to analyze the characteristics of graphite nodules with image-analyzer quantitatively.
Summary
Some results are summarized as following:
1. Fluid convection tends to mix up the liquid and resulted in a more even temperature distribution and a less degree of segregation of element molybdenum which has a strong tendency to segregate to the carbide in ductile iron.
2. The microstructure of matrix mainly is pearlite but with some small ausferrite structures existed between the carbides.
3. Most graphite nodules were in good quality but some large and exploded nodules were appeared in areas with short local solidification times.
4. EBSD technique was used to investigate the growing direction of sidewise branches of carbide with a family plane of {4 1 1}.
References
[1] LEI Tien Shou and Chang Wen Shiung, Proceeding of the 8th Asian Foundry Congress, AFC-8, 17-20 October 2003, Bangkok, Thailand (2003), p.18.
[2] B. Black, G.. Burger, R. Logan, R. Perrin and R. Gundlach, SAE Transactions, Vol. 111, Section 5 (2002), p. 976.
[3] P.C. Liu and C.R. Loper, Jr., AFS trans., Vol. 89 (1981), pp.131-140.
[4] B.V. Kovacs Sr., Modern Casting, Vol. 80 (1990), p.38.
[5] D.C. Wen and T.S. Lei, Materials Transactions, JIM, Vol. 40, No. 9 (1999), p. 980.
[6] M.C. Flemings, Solidification Processing, McGraw-Hill (1974), p.187.
計畫成果自評
一、本研究得到的成果及預期貢獻為:
1. 研究裝置成功地建構區分成上、下兩個區段可以同時分別往上及往下單向凝固的設
備，並有效地用以研究流體對流效應對球墨鑄鐵在單向凝固特性的影響。
2. 以 EPMA 分析上、下兩個區段凝固區中元素在凝固過程的再分佈特性，進一步瞭解流
效應影響元素偏析的方式，這將對鑄件凝固過程發生逆冷硬的機制提供分析的基礎。
3. 以 EBSD 分析碳化物成長方向與凝固方向之間的關係，以及碳放物成長時主枝晶與分
枝晶之間的方位關係，。
二、本研究中人員的訓練:
1. 積極指導博碩士研究生直接參與研究，並鼓勵學生出席參與 2003 年 10 月在泰國曼谷
舉行的國際 AFC-8 的研討會，以及國內 92、93 年度所舉行的鑄造學術研討會的論文發表。
2. 積極進行國際論文發表，並參加 2004 年 9 月在匈牙利米什克爾次大學所舉行的第 4 屆
凝固與重力國際研討會的論為發表。
Table 1 Chemical compositions of ductile irons from chilled specimens.
Elements [mass %]
Castings
C Si Mn P Mg Cr Mo R.E. Fe
Iron A 3.70 2.40 0.20 0.03 0.03 0.04 --- 0.0 bal.
Iron B 3.67 2.44 0.48 0.04 0.05 0.73 0.52 0.0 bal.
Iron C 3.68 2.23 0.48 0.04 0.02 0.73 0.53 0.08* bal.
*This value is the target value of the casting.
Table 2 Chemical analysis from various location and structures of the casting iron B.
Elements [atomic %]
Specimen number of
sections C Si P Cr Mn Fe Mo
No.9-center-Mo+ 59.198 0.406 0.251 1.81 0.217 17.525 20.594 No.9- center -Mo+ 45.753 0.310 0.437 1.623 0.271 12.232 39.375 No.9- center -Mo+ 31.968 8.898 1.273 0.76 0.408 32.816 23.877 No.9- center -Mo- 24.641 0 0 3.189 0.901 69.984 1.285 No.9-rim-Mo+ 36.811 7.301 1.105 0.855 0.438 34.292 19.20 No.9-rim-Mo- 30.057 0.028 0.023 3.568 0.906 64.415 1.003 No.6-center-carbide 23.725 1.663 --- 1.465 0.734 71.299 1.114 No.6-rim-matrix 19.092 4.186 --- 0.404 0.314 75.959 0.045 No.6-rim-matrix 26.242 0.024 --- 1.158 0.470 71.921 0.185 No.6-rim-carbide 19.522 2.121 0.008 0.685 0.433 77.071 0.161 No.3- center -Mo+ 12.031 5.659 0.738 0.434 0.715 76.869 3.554 No.3- center -Mo+ 11.227 4.497 1.009 0.918 0.686 66.804 14.86 No.3- center -Mo- 19.435 0.016 0.011 2.052 0.711 77.141 0.635 No.3-rim-Mo+ 15.970 4.429 0.416 1.146 0.612 65.242 12.186
No.3-rim-Mo- 20.512 0 0.032 3.166 1.146 71.506 3.638
(a) drawings
(b) bottom section with chilling coppers
(c) casting cylinders Fig.1 Experimental set-up for directional solidification and its photos
Fig. 2(a) Typical cooling curves at different locations of castings.
Fig. 2(b) Typical cooling curves at bottom section of casting.
Fig. 2(c) Typical cooling curves at top section of casting.
(a) No9-center-BEI image (b) No9-center-Mo mapping
(c) No3-center-BEI image (d) No3-center-Mo mapping Fig. 3 BEI image and Mo mapping of carbides at two zones form EPMA.
(a) No7-center-1000x (b) No7-center-10000x
Fig. 4 SEI microstructures of ductile iron B from section near the chilling copper mold.
(a) No.7-rim-longitudinal (b) Detail of left one with EBSD spot Fig. 5 Microstructures show branches of sidewise grow of carbides.
No.09-V-00R-420X No9-V-00R2-420X
No.06-V-00R-420X No6-V-00R2-420X
No.03-V-00R-420X 100 um
No3-V-00R2-420X 100 um
Fig. 6 Microstructures show the quality of graphite nodules at three sections, the graphite nodules shown in left hand column are in good quality, but in the right hand column are relatively huge and exploded.
