Inoclulating iron through filters.

Inside This Story:

* In-mold inoculation and ceramic foam filtration have been practical methods to obtain efficient iron castings.

* Recent investigations have examined combining the two methods at one point in the mold by inserting inoculating pellets in the filter centers.

* This article details how well an inoculant-filter assembly performs under various mold and heat conditions with molten iron.

Since it was first introduced to the metalcasting industry in the 1960s, the method of in-mold inoculation has helped produce castings with stable microstructures. Throughout the years, this technique has been utilized by placing inoculants, such as sintered, blocks, cast blocks, pellets or sized chunks, inside a mold, and it has become a common practice. To be effective, the inoculant must suppress the formation of carbides and Type D & E graphite.

But to produce sound castings day-in and day-out, ferrous metalcasting facilities should strive to cast components with not only the correct microstructure, but also with clean metal.

Like in-mold inoculation, ceramic foam filtration is not new to the industry and has gained acceptance for effectively removing impurities and modifying fluid flow. To obtain sufficient cast iron applications, the filter must provide high thermal shock resistance, excellent corrosion resistance and good dimensional stability at a low cost.

With the proven benefits, recent developments have focused on combining the two techniques--placing the inoculant inside the filter, which then is placed within the gating system.

Using an in-filter inoculation system provides several technical and economic advantages. It practically eliminates chill resulting from inoculant fading and improves mechanical properties and machinability of irons while simultaneously providing effective filtration. Additionally, in-mold inoculation will increase the consistency of the inoculating process for each mold, which would reduce microstructure differences between castings.

The in-filter inoculating system was further investigated when an automotive metalcasting facility was experiencing chill and hard spots on a component. The firm's current stream inoculation unit lacked the desired reliability, so the facility looked at options for a secondary or safety inoculant. Other metalcasting firms had encountered similar situations, as well. Through preliminary tests, it was discovered that an inoculant filter could function as the primary inoculant and replace the need for stream inoculation entirely.

As a result, further investigations were performed to discover the effectiveness of a combination of in-mold inoculation pellets with a filter for thin-wall gray iron casting applications.

The studies involved two phases: evaluating the dissolution behavior of the inoculant pellet embedded into a ceramic foam filter; and evaluating how effective the inoculant-filter assembly was in nucleating and cleansing the molten metal. This article examines the methods taken to investigate inoculant-filter behavior and what was discovered from each study.

Fully Dissolved

In the first part of the investigations, eight 0.03-lb. (15-g) inoculant pellets were produced and examined to see how well they dissolved when immersed in molten iron.

Each 1-in. diameter pellet was made from ferrosilicon and included 5.5-6.5% zirconium (Zr) to indicate dissolution. The pellet was embedded in the central cavity of a 2.36-in (60-mm) diameter ceramic filter (Fig. 1). This inoculant-filter assembly was placed in a nobake pouring cup and pig mold assembly (Fig. 2), which was cast with iron ranging from 2,327-2,570F (1,275-1,4100. Each inoculant variation was tested three to five times at different pour temperatures and pour time lengths.

[FIGURES 1-2 OMITTED]

The pig castings then were weighed to determine how much iron passed through the inoculant-filter assembly.

Two of the pig castings were drilled on their tops at five separate points where 0.04 lbs. (20 g) of turnings were collected at each point and analyzed for Zr content. It was found that Zr content in iron after inoculation increased by a factor of 10, and 64-70% of the Zr available in the inoculant pellet was recovered in both castings (Table 1). Often, all particles within the inoculant pellet dissolve at similar rates.

Next, the filters were visually examined for any inoculant material that did not dissolve, and this revealed differences in the pellets' behaviors. The first group of four inoculant pellets dissolved in most cases, but the second group of inoculant pellets dissolved in all cases, even at the lowest temperature (2,328F/1,275C) and with the least amount of iron poured (34 lbs./15.4 kg). Further, the filters were analyzed for non-metallic inclusions trapped in the filters. It was found that none of the filters contained any residual ferrosilicon material. The samples where inoculants did not fully dissolve only contained small particles of gray iron and refractory material.

Although this first investigation proved that the pellets dissolved adequately, more studies had to be performed to better evaluate inoculant-filter properties. Therefore, two inoculant pellets that completely dissolved were selected for another investigation.

Fluid-Flow Analyses

A nobake test mold with five sequential wedge cavities (Fig. 3) was designed for further study to examine how metal flow characteristics affect the inoculant-filter assembly. In addition, one end of each cavity was designed so that iron flow completely stopped immediately after filling. The inoculant pellets were inserted into a 2.16 x 2.16 x 0.86-in. (55 x 55 x 22-mm) filter (Fig. 4), which was placed in the mold.

[FIGURES 3-4 OMITTED]

Iron was poured at a flow velocity of 1-3 lbs./in./sec, at temperatures of 2,462F (1,350C), 2,552F (1,400C) and 2,642F (1,450C) for a first set of eight castings and at temperatures of 2,424F (1,329C) and 2,680F (1,471C) for a second set of eight castings.

After cooling, three of the wedges of each casting were taken from the cavities and drilled in three positions at the end opposite to the crossover segments. These drill turnings were chemically analyzed for residual Zr content. An analytical model was created with the expectations that temperature and flow rate affect dissolution behavior. However, differences between the analytical model and the experimental data showed only partial correlations (Figs. 5-6).

[FIGURES 5-6 OMITTED]

Nevertheless, the general nature of how well the pellets dissolve follows a positive correlation with the flow rate of molten iron and proves that the Zr dissolution rate and recovery can be predicted.

Keep It Clean

Although the inoculant-filter assemblies showed that inoculants will properly dissolve, it still needed to be proven that the assemblies provide clean and well-inoculated metal for a casting. The second part of the investigations included several criteria to focus on the iron's chill tendency, microstructure (graphite morphology and metallic structure), cleanliness and the integrity of the filter. Two individual inoculant compositions of ferrosilicon were utilized in this phase: one was enriched with Zr, the other with strontium (Sr). The inoculant-filter assembly was the same as the one used in the flow analyses.

Another nobake mold was utilized, but this was designed with two identical casting cavities located parallel to each other. Each casting cavity contained both a test plate and a chill wedge and had a separate filter print. The inoculant-filter assembly was placed in the filter print of one test plate while an identical filter without the inoculant pellet was installed in the same location of the other plate (Fig. 7). Each test cavity contained a chill wedge, and the iron used to cast the molds was poured at 2,700F (1,482C).

[FIGURE 7 OMITTED]

The chill tendency of the iron was studied after the castings cooled by measuring the depths of each chill. It was found that the zirconium-containing pellets showed the most consistent chill reduction (Tables 2-3).

The microstructure of the polished sample was then examined for each chill wedge. In all cases, the inoculated iron significantly reduced the formation of carbides at a depth of 0.2 in. (5 mm), increased the formation of Type A graphite (which enhances both the mechanical properties and the machinability of the gray iron casting) and reduced the percentage of Type D and E graphite (which can indicate that the iron is poorly inoculated in a particular area of the casting). The inoculant also helped create a pearlitic matrix. This confirms that tile inoculant filter assembly can effectively reduce the formation of iron carbides (which can accelerate tool wear and reduce the machinability of the casting) while promoting the formation of Type A graphite in a pearlitic matrix.

After this process, the focus was shifted to the test casting within the mold itself. To determine the cleanliness of the casting, the cope surfaces of the test plates were machined to reveal any subsurface defects and inclusions. Each inclusion was removed from the test plate and examined. The two plates then were studied to identify the nature and morphology of inclusions.

In all cases, the areas that appeared to be subsurface inclusions after machining were actually voids or porosity in the casting. All of the defects examined contained no extraneous material indicating that the filter prevented any particles of the pellet or extraneous inclusions in the molten iron from entering the cast plate.

The spent filters next were examined for each test casting produced, but they were found mostly to be free from inoculant materials. The material observed in several samples was evaluated and found to be reoxidation products normally associated with the filtration of gray iron, but none of the samples contained any residual ferrosilicon material within the filter area.

Filtered Out

These investigations display examples of how in-mold inoculation can benefit an iron casting. The inoculant pellet properly dissolves and, along with the filter, helps remove non-metallic material from the metal. Although further investigations should be conducted to prove the findings here, in-mold inoculation and more recently, filtration, are two procedures that have benefited metalcasting facilities for a number of years. By combining the two into one method, it helps eliminate a step in the casting process without sacrificing quality. This can lower production costs while accelerating production cycle times and help metalcasting facilities achieve their ultimate goal.

This article was adapted from a paper (04-092) presented at the 2004 Metalcasting Congress.

For More Information

"Improving Inoculation of Ductile Iron," S. Lekakh and C.R. Loper Jr., 2003 AFS Transactions, Paper No. 03-103.

"Filtering Basics: Who, What, Where, Why & How," K. Adams, E.J. Williams, S. Kannan, MODERN CASTING, March 2002, p. 19-21.

Yury S. Lerner is the Foundry Educational Foundation Key Professor at the Univ. of Northern Iowa, Cedar Falls, Iowa. Don Craig is a product manager and Leonard Aubrey is the vice president of new product development at the Selee Corp., Hendersonville, N.C. Thomas Margaria is a research group leader and Roland Siclari is a the manager of technical services and product development for Pechiney Electrometallurgy, Paris.