Magnesium Recovery and Addition Rate in Tundish-Treated Ductile Iron 

E. Jepsen

CWC Textron

Muskegon, MI

V. Popovski  

   Elkem Metals, Inc. 

Pittsburgh, PA

ABSTRACT

The calculations that are associated with such alloy additions are based on knowing the beginning and goal levels of the element and the recovery rate of that element from the alloy. These calculations are straightforward; Goodrich³ described tin as having “been a popular alloy addition because it does not promote chill, is effective in stabilizing pearlite on the skin of the casting and has a predictable, 100% recovery.”

This is not the case with more volatile elements like magnesium (Mg).  The recovery of Mg is far more problematic by the very nature of the element.  In describing the magnesium, Klein⁵ wrote, “It has limited solubility in iron, and a boiling point of 2025F (1170C), far below normal iron treatment temperatures of 2700-2750F (1482-1510C).”  This boiling and limited solubility cause a violent reaction that makes alloying iron with Mg a challenge.

The operating foundry recognizes that a certain amount of Mg-bearing alloy must be added to achieve the required level of nodularity in the final casting.  For cost and metallurgical reasons the foundry seeks to minimize and control the amount of Mg alloy (in this case magnesium ferrosilicon, or MgFeSi) needed for production of ductile iron.  The level of Mg in the final iron compared to the amount added in the form of MgFeSi is known as Mg recovery.  

Knowing and controlling Mg recovery is a critical factor in effectively operating a ductile iron foundry.  Karsay⁴ summarized in general terms what is in much of the literature when he wrote, “Recovery values vary with treatment batch size, master (treatment) alloy sizing, base iron chemical composition, treatment time, and with other influences.”   Responsible operators are mindful of these variables and adjust the amount of treatment alloy accordingly.   

However, by adjusting the amount of treatment alloy, another variable has been introduced not only to the final level of Mg in the casting but also to the recovery associated with the entire practice.

BACKGROUND

CWC Textron treats ductile iron in a tundish with MgFeSi.  The tundish uses a pocket for holding the alloy and the pocket location was adjusted in January 2006.  The addition rate of MgFeSi was periodically lowered as the foundry gained confidence in the improved practice. 

Figure 1 reflects 1212 heats of ductile iron.  The recovery formula was a straightforward one that accounts for variation in heat size and Mg content in the lot analysis of the treatment alloy:

Mg Recovery = actual residual Mg / [(% Mg in MgFeSi * lbs MgFeSi)/heat size]

It is impossible to account for every other variable in the ductile iron process, but a few other facts should be noted:

·         every heat was poured from the same 40-ton holding furnace

·         every heat was of the same grade of ductile iron

·         every heat was used to produce the same part

·         tap temperature and base sulfur showed no appreciable impact on     Mg recovery

The chart shows with a confidence interval of 55% that Mg recovery falls in relation with the amount of treatment alloy used.  Therefore, one cannot assume that each marginal unit of MgFeSi alloy will perform as well as the one added before it.  

CONCLUSION

The recovery of magnesium in tundish-treated ductile iron is subject to multiple influencing variables.  One of these is MgFeSi alloy addition rate.  Higher MgFeSi additions mean poorer overall Mg recovery.

At higher addition rates of MgFeSi it is hoped that all of the MgFeSi and cover material still fit in the pocket, but under these conditions it is also more likely that not all of the material actually fits in the pocket as is desired.  As such, it is difficult to say if the recovery is inferior because more alloy was added while fitting, or whether the alloy actually spilled out to where it did not belong.  It is therefore suggested that foundries need to keep the pocket clean so it retains enough volume for all of the MgFeSi and all the cover to fit into the pocket.

Even assuming the presence of a clean pocket and correct alloy addition practice, each marginal unit of MgFeSi is by definition closer to the the liquid iron and will have less of the needed pocket protection1 for good Mg recovery.  As such, each marginal unit will have inferior recovery than the whole and will drive down overall Mg recovery in the heat.  In short, given that a certain amount of treatment alloy results in a certain level of residual magnesium, the operator cannot assume that a given additional amount of alloy will achieve a linearly corresponding final level of magnesium in the casting.

This means that the operating foundry person must consider how to react to a drop in Mg recovery during the course of production.  The path of least resistance is to simply add more alloy because nobody can argue that the casting needs sufficient Mg to be nodular iron.  However, the price to pay for this expediency has been exposed as severe in that the added units of MgFeSi are not nearly as efficient as the regularly added units.  The correct response is to seek out and solve the cause of the inferior recovery and proceed with production at the prescribed operating levels.

REFERENCES

1. Elkem Metals, “The Sandwich Pocket Process” , Elkem Technical Information #11
2. Elkem Metals, “Factors Influencing the Recovery and Addition of Magnesium in Ductile Iron Ladle Treatment Processes”  Elkem Technical Information #23
3. Goodrich,George M., “Cast Iron Microstructure Anomalies and Their Causes” , AFS Transactions, pp. 679 (1997).
4. Karsay, S.I., “Ductile Iron Production Practices”, p. 58 (1979)
5. Klein, J., “Treatment Alloys and Materials”, Ductile Iron Handbook, p.148  (1992)