Chantal Labrecque & Martin Gagné
Rio Tinto Iron & Titanium, Sorel, Québec, Canada
DIS Presentation June 2001
Introduction
This research project has been initiated to develop a simple technique
to achieve a carbide free structure in a thin-wall test casting having
the attributes of a production part. Once this first objective
completed, the second part consisted in Correlating the microstructure
and the mechanical properties to the chemical composition and the
cooling conditions. The following summarizes the principal observations
and conclusions of this research work.
Experimental Procedures
The experimental casting is shown in Fig. 1. It consists in four
sections each having four vertical plates of different thicknesses: two
plates of 3, one of 6 and one of 10 mm. Because of the orientation of
the plates, a draft angle is introduced to facilitate the mold-pattern
separation. Thus the section thickness is thinner at the top and bottom
compared to the center. Two alumina foam filters are also inserted in
the mold.
The charge is melted in an induction furnace (350 lb), heated to 1530°C
and Mg treated with 2%-FeSiMg5% using the plunging technique. The
inoculation is carried out in two steps. The first one is achieved by
adding 0,75%-FeSi75% in the metal stream during the metal transfer to
the ladle. The second inoculation step takes place in the pouring box
where 0,5% Si64/Bi1/RE1 is mixed with the liquid metal before flowing
into the mold. The filling time is approximately 6-7 seconds.
Results
Melts are produced within the following chemical compositions: 3.6 to 4.0% C; 2.35 to 3.0% Si. The Mn content is kept below 0.1% except for one test for which 0.36% Mn is added. The % Pearlite, the Nodularity and the Nodule Count distribution across the vertical central plane are measured in the 3 and 10-mm plates of each casting. The typical microstructure of the centre of a 3-mm plate is presented in Fig. 2. The nodularity level achieved in the 3-mm plates is always higher than 90% and the pearlite content varies from 25 to 55%. The residual Bi content in the casting is 15 ppm. This small amount does not affect the nodularity in the 3 and 10-mm plates. The difference of nodule count (same vertical position) between the 3 and 10-mm plates lies in the range of 600 and 700 Nod. /mm2. The nodule count distribution in different 3-mm plates is presented in
Fig. 3. Nodule count increases with the reduction of the plate thickness toward the edges. This is expected because of the increased cooling rate associated with the reduction of section thickness. However, it is also observed that the nodule density increases in the center of the plates. Such a phenomenon is not seen in the 10-mm plates. The physical parameter that controls Nodule count is the cooling rate; however at the center this high cooling rate is ascribable to the solidified iron surrounding the liquid iron acting as a heat sink. This is similar to the inverse chill phenomenon but in this case, no carbide occurs.
Figure 4: Mechanical Properties vs. Section Size and
Filter Effect. |
Figure 5: Mechanical Properties vs. %Si. |
The mechanical properties were measured in the 3 and 10-mm plates. The results are correlated to the section size, and the filtering effect (Fig. 4), and to the silicon content (Fig. 5). These are compared to those of the ASTM A536 standard (continuous line). Figure 4 shows that the properties of the 10-mm plates are closer to those of the standard than the 3-mm plates. For equivalent UTS values, the thin wall iron casting exhibits lower ductility. This is ascribable to the higher nodule count and to the lower ferrite content of the 3-mm plates when compared to the 10-mm plates. One of the experimental castings was produced without a filter. The mechanical properties of these plates (Fig. 4) lay below the general distribution of the properties. This is indicative of the benefit of filtering to reduce the occurrence of defects that decrease the mechanical properties. Finally, in Figure 5, the relation between silicon content and mechanical properties is presented. It appears that the lower the silicon content, the better are the mechanical properties even for thin wall casting (compare to Fig. 4 to identify the 3 mm plates). Thus using very high silicon concentration may not be necessary to produce a good quality thin-wall ductile iron casting. The inoculation technique has to be very efficient and the pouring temperature and rate high. In these experiments, the pouring rate was more than 3 lb/s.
Conclusions
- Significant microstructural variations exist within a 3-mm casting. This is ascribable to different cooling rates as well as different filling conditions. Designers have to be aware of the range of properties that may be found in such castings.
- High cooling rates (not %Si) control the nodule density and distribution.
- For ~ 3-mm casting, % Si controls the Position of max. Nodule Count, % Pearlite.
- Filtering improves the consistency of static mechanical properties (UTS and %El)
- It was possible to produce a good quality thin-wall D.I. casting having a %Si lower than usually recommended. It might be beneficial to keep the %Si as low as possible in order to get better mechanical properties in thin wall Ductile Iron castings.
|
