How to Design a
Proper Feeding System for Iron Castings
A properly designed feeding system for iron castings (both gray
and ductile) requires an understanding of how these alloys
differ from others, such as steel. If these differences are not
considered, feeding systems may be less than adequate and
casting quality will suffer. In many cases, feeders designed
essentially for steel castings lead to production defects when
applied to iron. Such misunderstandings lead to suggested
solutions that often worsen the situation.
Understanding design properties specific to iron, when applied
in conjunction with simulation software, can lower scrap rates
and result in quality castings. Additionally, running computer
simulations prior to initial production can help avoid weeks and
months of defective castings in mere minutes.
Follow the Rules
The biggest difference between iron and other alloys is the
expansion of iron during graphite precipitation in
solidification. This difference means iron castings can become
self-feeding after the onset of expansion in most situations, so
no further feeding is required.
Appropriately, a feeding system for iron castings should provide
metal only for the contraction of the liquid alloy and
solidifying iron prior to expansion. Once the expansion begins,
a well-designed feeding system should contain the pressure so
the casting is self-feeding during the remainder of
solidification. This principle is in direct contrast to other
alloys, such as steel, where there is no expansion and feed
metal must be supplied during most or all of solidification.
Another major difference between iron and other alloys has to do
with the mechanism involved in piping. Iron alloys (particularly
ductile iron) do not readily form a solid skin during
solidification. For feeders to pipe effectively, atmospheric
pressure must be able to collapse the weak plastic skin when the
internal pressure drops. After a feeder punctures the skin, the
internal pressure then is equalized within a feeding zone (the
area within a casting where liquid metal can flow from one point
to another in response to expansion pressures). Only one feeder
should be used for each feeding zone. If multiple feeders are
placed on the same zone, one feeder will begin piping while the
others will not. Often, porosity will be seen at contact points
of the non-piping feeders.
Iron’s requirement for a single feeder within a single zone is
the design rule that is violated most often. When porosity is
found at a feeder contact point, the tendency of many engineers
is to add more feeders; this is exactly the wrong approach to
take and will worsen the situation.
The ductile iron control arm, shown in Figure 1a, is an example
of an iron casting with an incorrectly designed feeding system.
The metalcasting facility originally approached the feeding
design for this iron casting by placing two symmetrical feeders,
shown in Figure 1b.
This approach was understandable, because the feeders were
attached to the heaviest sections of the casting. During initial
production, porosity occurred at one feeder contact on a
consistent basis, as shown in Figure 2. The porosity was not
always at the same contact, but on a large majority of castings,
one contact showed evidence of porosity while the other did not.
As a result, the metalcasting facility could not produce a
quality casting with this pattern design.
Design with Data
To correctly design feeder systems for iron castings, it is
necessary to determine the location and size of the casting’s
feed zones. Understanding the transfer modulus (MTR), a
calculation relating to metal flow within a casting, can help
determine if a casting has one or more feeding zones.
If metal cannot flow from one location to another, each feeding
zone may require its own feeder (but no more than one). The
casting modulus (MC) represents the ratio of volume to surface
area in various areas of the casting. The modulus is used to
estimate the order of solidification by allowing engineers to
estimate the progress of solidification in a casting. In iron
castings, the modulus is used to estimate when expansion will
begin and is expressed as a percentage of complete
Modern software programs can simulate solidification in a few
minutes, and the resulting data then can be converted to casting
modulus values. A casting with a higher modulus (heavy section
castings) will begin to expand earlier and will undergo more
expansion than a casting with a low modulus (light section
castings). The point when expansion begins is referred to as the
shrinkage time point.
Knowing the shrinkage time point allows the calculation of an
equivalent modulus value that corresponds to the modulus at
which contraction of the iron stops and expansion begins. This
modulus value is known as the MTR, because it defines the areas
of the casting where liquid metal transfer is possible. The
calculation of MTR is:
MTR = SQR (ST/100) * MC
By plotting the MTR in a casting simulation, one can determine
whether the entire casting is a single feed zone (the modulus
transfer is continuous throughout the casting) or contains
multiple zones (modulus transfer is discontinuous). The number
of feeding zones then determines the number of required feeders,
using one feeder per zone.
The value of transfer modulus can be understood as representing
the casting modulus value below which feeding from risers is no
longer effective and the iron becomes self-feeding due to
expansion. The expansion pressure must be controlled, which
means, assuming the mold is rigid enough, all contacts with the
casting (gates and riser contacts) should be solid enough to
ensure the expansion pressure is contained within the casting
after the onset of the graphite expansion. This also means the
modulus of the feeder contact neck should be equal to transfer
modulus to ensure the feeding of the liquid contraction will be
able to occur and also that the expansion pressure will be
contained within the casting due to freezing of the feeder
contact at the correct point in solidification.
To resolve this problem in the ductile iron control arm example,
the casting was analyzed to determine feeding requirements.
First, a solidification simulation of the casting without gating
or feeders was performed. The results of this simulation are
shown in the plot of solidification time (in minutes) in Figure
The data from the simulation was converted to modulus data so
the feeding calculations could be performed. It is tempting to
conclude the original feeder design was correct, because the two
areas of high modulus value in the casting were in close
proximity to the feeder contacts in the original design.
However, it is necessary to analyze this casting further to
determine the shrinkage time and MTR to understand the location
and size of the feeding zone(s). Analysis of the iron
characteristics indicates the value of the MTR is 0.254 in.
(0.645 cm). Creating a plot of this value within the casting
will indicate the location of feed zone(s), shown in Figure 4.
The entire casting is actually a single feed zone. The areas of
higher modulus are connected by a section in which the modulus
is above the transfer modulus value, thus allowing liquid
transport for feeding throughout the casting. Only a single
feeder should be used to avoid potential porosity at a
The computer simulation in this case took 16 minutes to perform,
and after calculating the shrinkage time and transfer modulus,
the plot was created in 5 minutes. After about 20 minutes of
analysis, the correct feeder design was determined. Had this
been done before the original pattern equipment was created,
several months dealing with defective castings could have been
avoided. The associated costs were far greater than the upfront
investment for the simulation software and training to perform
The pattern was revised to reflect a single feeder, shown in
Figures 5a-b. The feeder in this case is not connected to an
area of high modulus. In iron castings, the location of the
feeder is not as critical as with steel castings because of the
expansion pressure’s effects throughout the casting after
graphite precipitation begins.
No porosity was discovered at the casting’s contact area with
the single feeder, shown in Figure 6, or elsewhere. A simple,
quick analysis of the casting produced the correct feeder design
that resulted in a sound casting. Computer simulation offers
engineers an effective tool to design a production process,
thereby avoiding potential costs involved with defective
Home Page | More Articles