HOME

Efficient design of tall tapered feeders

Toowoomba Foundry

Industry contacts:

• Dayalan Gunasegaram (dayalan@toowoombafoundry.com.au)

MISG moderators:

• Tony Roberts (aroberts@usq.edu.au)
• Mark Mcguiness (mark.mcguiness@mcs.vuw.ac.nz)

Figure 1a. A tall tapered feeder.	Figure 1b. A traditional feeder.

Tall Tapered Feeders (TTFs) are placed as near as possible to wide sections of the mould cavity so that hydrostatic pressure in the feeder acting on the top surface through the gaps between sand granules will push melt into the mould during shrinkage thereby improving feeding capabilities. If the feeder is successful, a pipe will be visible from the top and very little metal will remain in the feeder itself. At the present time however too much metal remains in the feeder and also in the channels through which the melt is poured; (the whole mould needs to be poured in less than 25 seconds).

Figure 2 shows an example of a TTF used in a casting at the Toowoomba Foundry. The TTF is shown as the lighter coloured cone shaped object in the centre of the picture. This picture shows the cope or upper box of the casting. In the lower centre right, the hole in which molten metal is added is shown, together with feeder lines connected the casting and the TTF. Indicative dimensions of a TTF feeder reservoir are shown in Figure 1a.

Figure 2. The cope, or upper box, of a casting.

The problem

While the TTFs almost always work with grey iron (GI) castings, the same is not true of ductile iron or spheroidal graphite (SG) castings which are becoming increasingly popular as they are stronger per weight. In the latter, when the casting section to be fed exceeds appoximately 25 mm, the TTFs become unreliable, resulting in shrinkage porosities in the castings of around 3%, compared with 1% with grey iron. In the SG iron castings, Mg is added to spherodise the graphite particles which solidify as flakes in GI in the absence of spherodising agents.

The difference in performance between the two compositions is believed to be due to the different solidification mechanisms undergone by GI and SG melts. While the GI castings tend to solidify more like pure metal (very short solidification range), SG castings form a skin around the melt (typical of medium to long solidification range alloys). This skin then prevents the atmospheric pressure from acting on the melt in the TTF, thereby retarding its performance. Since the skin takes some time to form, the TTF can still be successful if the casting section to be fed is small.

Also, a larger casting section entails a larger feeder diameter. However, since the success of TTFs depends on a vertical temperature gradient which can be lost due to temperature equalisation as the solidification time increases, thicker feeders feeding thicker sections may be less effective.

Factors involved

The design developed in one foundry need not work effectively in another, since there are several factors which are peculiar to each foundry. For example, the melt composition and mould variables such as mould rigidity may vary from foundry to foundry.

The following are some of the factors that influence the performance of TTFs:

• Melt composition
  • if there is less carbon percentage in the melt, the consequence will be increased shrinkage as there will be less graphite expansion;
  • melt composition also depends on the solidification mechanism (GI vs. SG).
• Melt temperature. There will be more feeding requirement if the temperature is higher because
  • the melt will contract more as it cools down, and
  • since the hotter melt will have expanded the mould more due to the excess heat it transfers, the casting cavity will be larger.
• Casting height and size of section to be fed. The feeder height should take into account the height of the casting, and the feeder diameter should be greater than the effective thickness of the section to be fed.
• The ability to maintain temperature gradients.
• Connection size. The size of the connection between the feeder and the casting should be large enough to allow the full feeding power of the feer to be successfully transferred into the casting being fed.
• The size of the sprue. The feeder and sprue compete to feed the casting. Therefore, the diameter of the top of the TTF should be greater than the diameter of the sprue, or else the sprue may feed the casting instead (possibly at the wrong location in the casting) and prevent the TTF from piping. Ingates to the feeders from the sprue are generally made thin so that they freeze quickly and cut off the influence of the sprue.
• Mould variables. If the mould is less rigid (e.g. due to increased sand moisture levels), then the casting cavity will expand more during the graphite expansion phase of solidification and will necessitate additional feed metal for soundness.

Possible investigations

• Magnetic stiring of the melt in the feeder. This may be technically too difficult and energy inefficient.
• Investigate the mechanisms of the cooling of a mould and a TTF so they can be better understood and parameterised.
• With some model of the cooling, investigate minimal volume (axi-symmetric) shapes of a feeder that can maintain hydrostatic pressure into the mould for a set time.
• Toowoomba Foundry are at present using a software package called maGma which seems to resolve the feeder at a relatively course level. Some higher resolution numerical solutions of the reservoir and cooling air/sand matrix may be useful.

Outcomes for Toowoomba Foundry

To expand the boundaries of feeder design so that TTFs can be used on thicker SG castings generally made in a commercial foundry. In particular, the following are the aims of the exercise with MISG 2001:

• Arrive at an effective thermal design for the top section of the TTF such that the melt in that top region remains molten for as long as possible, thereby keeping a path open for the atmospheric pressure to aid the feeding process in the making of 'large' SG castings. (Keep in mind the fact that the melt at the top of the feeder will be the coldest to begin with, due to flow patterns, thereby making the problem more challenging).
• To effectively calculate (or simulate) the heat convected from all sides of the TTF by the air that flows between sand granules, and the radiated heat, and heat conducted by sand granules.

References

Heine, R. W. 1982, Design Method for Tapered Riser Feeding of Ductile Iron Castings in Green Sand, American Foundrymens Society Transactions, pp. 147-158.

Heine, R. W. 1979, Risering Principles Applied to Ductile Iron Castings Made in Green Sand, American Foundrymens Society Transactions.

SG 500 data