FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 2

FOUNDRY PRACTICE
ISSUE

238

FIRST FOR FILTRATION: THE FOSECO STORY

MORE THAN JUST A FILTER

APPLICATION OF SEDEX CERAMIC FOAM FILTERS

ON VERTICALLY PARTED MOULDS SUCH AS DISAMATICS

FLOW MODIFICATION PROPERTIES OF CERAMIC FOAM FILTERS –

A SUMMARY OF RECENT WORK

BIBLIOGRAPHY
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 3

page Contents
.............................................

FOUNDRY
PRACTICE
1 FIRST
THE
FOR FILTRATION:
FOSECO STORY

ISSUE 238
March 2003

.............................................

5 MORE THAN JUST A FILTER

EDITED BY J.G. SUTHERLAND

.............................................

Front cover: A selection of filters for APPLICATION OF SEDEX CERAMIC FOAM

10
Aluminium, Copper, Iron and Steel FILTERS ON VERTICALLY PARTED MOULDS
applications SUCH AS DISAMATICS
BYK.C. TAYLOR
All rights reserved. No part of this &
publication may be reproduced, stored in a A. BAIER
retrieval system of any nature or transmitted
FOSECO GMBH
in any form or by any means, including
photocopying and recording, without the GERMANY
written permission of the copyright holder.

All statements, information and data

contained herein are published as a guide .............................................
and although believed to be accurate and
reliable (having regard to the manufacturer’s FLOW MODIFICATION PROPERTIES OF
practical experience) neither the
manufacturer, licensor, seller nor publisher
CERAMIC FOAM FILTERS –
A SUMMARY OF RECENT WORK

18
represents or warrants, expressly or
impliedly: BY B.R.ALQUIST
FOSECO METALLURGICAL, INC.;
(1) their accuracy/reliability USA
(2) that the use of the product(s) will not
&
infringe third party rights
F. CASSIO
(3) that no further safety measures are
required to meet local legislation FOSECO BRAZIL

The seller is not authorised to make .............................................

representations nor contract on behalf of
the manufacturer/licensor. All sales by the
manufacturer/seller are based on their
respective conditions of sale available on

27
request.
BIBLIOGRAPHY
FOSECO the logo, SEDEX, SIVEX, STELEX, and
KALPUR are Trade Marks of the Foseco
Group of Companies used under licence.

© Foseco International Ltd. 2003

.............................................
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 4

........................................................................................................
First for Filtration: the FOSECO story

It's a FOSECO "first" that has transformed modern

metal casting. From its introduction twenty one
years ago, the FOSECO filter has enabled castings to
be poured with greater productivity, to higher
quality standards, at lower scrap rates. Machining
costs have been reduced, and advances like thin
wall castings have been supported. FOSECO has
supplied more than four billion filters from its
manufacturing units across the world, and
continues to provide the application and methoding
support that helps foundries put filters to work to
deliver real added value. The new STELEX PrO
carbon bonded ceramic filter, to be launched at
GIFA 2003, is the latest development in a line of
Foseco filtration "firsts" that started in 1982…
First SEDEX

It was the year that the much-publicised but ill-

starred DeLorean sports car met its end, just a few
months after its launch. That year, 1982, another
significant and far more lasting development was
taking place in the automotive industry: FOSECO
were introducing the first SIVEX and SEDEX filters,
products that were to have an extraordinary impact,
not just in the automotive industry, but wherever
metal was poured.

From the FOSECO introduction of casting filtration

over twenty years ago, filters have become an
essential component in the production of quality
steel, iron and non-ferrous castings worldwide,
delivering dramatic cost savings and quality
superiority with a product that is easy to apply. At
the heart of its success is the FOSECO leadership in
all aspects of filtration technology, with ongoing
development and manufacture of filters for every
type of casting as well as a depth and breadth of
application expertise unrivalled elsewhere. With
comprehensive analytical resources, FOSECO
provides full technical support to help the
foundryman solve just about any metal pouring
problem. Additionally, manufacturing facilities across
the world provide reliability of supply with the
highest standards of product integrity and quality.
This is why FOSECO has supplied more foundry
filters than anyone else over the last two decades.
This is the success story that started twenty years
and four billion filters ago. It continues today.
Early days
The FOSECO SIVEX F filter for aluminium casting
was launched in 1982, after a development and
trialling process that had taken over two years. It
was quickly followed by SEDEX iron casting filter, a
product still going strong more than twenty years
later. STELEX S filters were introduced for steel
SEDEX now
casting a few years later, to be followed by the
STELEX ZR filter in 1991, a filter using zirconia
ceramic on a foam matrix. During this period,
FOSECO also introduced the patented KALPUR
system, integrating a ceramic foam filter in a sleeve,
the so called direct pouring of a mould and this is
enabling gating systems to be eliminated and
production times reduced.
In 1992 the SIVEX FC filter was introduced, solving
the problem of contamination through the break-up
of the materials used in the SIVEX F filter. The new
material used in the SIVEX FC filter is lighter than
aluminium so that it floats to the top and can easily
be removed on remelting. By 1996 the STELEX ZR
filter had become so widely accepted that the
original STELEX S product was phased out. Seven
years later the new STELEX PrO filter is further
1
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 5
.........................................................................................................

increasing filtration possibilities. Much lighter in

weight, this new uniquely bonded ceramic filter
reduces heat loss and enables steel to be poured at
lower temperatures, delivering energy savings. So
impressive have been the results that one customer
undertaking tests described STELEX PrO as ‘the
invisible filter’; pour rates were as fast as without
filtration, and metallurgical analysis confirmed the
casting’s integrity.

FOSECO filtration expertise has helped customers

producing every type of casting to add real value
through yield improvements, lower scrap, reject
and return rates. Downstream savings have also
been achieved from improved casting integrity, such
as less machining and heat treatment, and longer
cutting tool life, thanks to the dramatic reduction of
inclusions. Customers have even achieved higher
sales margins as a result of FOSECO filtration! A
manufacturer of excavation buckets cast in a hard
alloy had never been concerned about inclusions in
the castings. However, when FOSECO
SEDEX launch – GIFA 84
demonstrated that longer product life could be
achieved with filtration, the customer realised that a
higher price could be charged for his new,
‘premium’ product.

So as foundry filtration comes of age, it is important

to remember that although SEDEX and SIVEX filters
have been around for 21 years, only the name
remains the same. The products introduced all that
time ago bear no resemblance to the product today.
Ongoing research development and investment in
manufacturing mean that today’s products are
stronger, lighter and greatly improved in terms of
their level and consistency of performance. This is a
process which continues today.

Likewise, our unrivalled experience and expertise in

applying filtration technology is constantly
improving and evolving, enabling our customers to
produce stronger, better, more technically advanced SIVEX launch – GIFA 84

castings at lower cost.

FOSECO Filtration evolution
Success stories like this are typical of the ways in
which FOSECO is helping customers not just to 1982 SEDEX iron casting filters introduced
apply the latest filtration technology effectively, but
to derive significant added value from the products 1983 SIVEX F filters for non-ferrous casting introduced
that the company ‘first’ launched onto the market
twenty one years ago. 1984 SEDEX and SIVEX F filters first on show at GIFA

1988 STELEX S filters introduced for steel casting

1989 KALPUR direct pour sleeves with integral filters launched

1991 STELEX ZR filters launched

1992 SIVEX FC filters launched

1996 STELEX S filter production ceased

2003 STELEX PrO filters introduced

2
 FOUNDRY
PRACTICE 238
21/5/03
11:03 am
Page 6

Filtration advertisement from 1986

3
........................................................................................................
.........................................................................................................

FOUNDRY

4
PRACTICE 238
21/5/03
11:03 am
Page 7

Filtration advertisement from 2003

FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 8

More than just a filter
Introduction
The last 20 years has seen significant improvements
in foundry practice and process control.
Improvements in melting technology;
inoculation/nucleation and alloying processes;
temperature control and moulding and coremaking
practices have fundamentally improved the quality
and consistency of liquid metal in the foundry.
During the same period, the application of filtration
technology has continued to rise. How can it be
that despite advances in metal quality, the growth
in the number of filter applications shows no sign
of slowing down?
There are a number of reasons to explain these
seemingly contradictory statements:
........................................................................................................
Sources of Inclusions
"Dirty" metal is the most obvious source of
inclusion material. Unwanted materials may enter
the melt through any number of ways including the
addition of unshotblasted returns to the charge,
oxidation problems during metal transfer, the
generation of reaction products of alloying elements
and other melt treatment products and the erosion
of refractory materials and slag build-up due to
poor ladle maintenance.
However, often overlooked sources of problems occur
in the mould such as reoxidation due to excessive
turbulence; mould erosion due to high metal velocity,
poor mould strengths; and steam or gas problems in
the mould due to insufficient venting and/or metal
mould reactions.

❑ the ongoing development of filtration
technology to produce more effective and
efficient inclusion removal;
❑ the realisation by foundrymen that filters add as
much to casting production in the form of
process control and productivity enhancements
as they take away in the form of unwanted
non-metallic by-products of the casting process;
❑ rising casting quality standards;
❑ increasing casting performance requirements;
One of the benefits of using foam filters is in
assisting the foundryman in diagnosing the source
of scrap by separating the problems of dirty metal
from problems occurring behind the filter in the
gating system and mould cavity.
How Filters Work
Uni-dimensional products such as extruded (figure 1)
and pressed ceramics (figure 2) and steel or cloth
mesh only remove inclusions at the surface and
inclusions smaller than the minimum cell or hole size
are not retained and flow into the mould cavity
(figure 3).

❑ the pressure to reduce costs by reducing scrap

and/or improving yield;

❑ the drive to reduce casting weights, resulting in

thinner wall sections which are more sensitive to
inclusions and reoxidation defects;

❑ the rise in popularity of automatic pouring

leading to higher metal velocities and the
increased potential for turbulence and erosion
defects;

❑ the increased use of simulation and other Figure 1: Extruded filters

predictive techniques has shown the positive
benefits of filtration on mould-filling much more
clearly.

In short,the use of foam filters continues to grow

because the value added is greater than the cost of
their application and purchase.

It is not simply the elimination of scrap but the

improvement in the quality of casting production as
a whole. Whereas 20 years ago filters were used
only in an emergency, today they are fundamental
to the production of high quality automotive and
engineering components at optimum cost (1). Figure 2: Pressed filters
5
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 9
.........................................................................................................

Molten metal that flows past the filter cake and

into the passageways follows a tortuous path
through the body of the filter. The foam filtration
mechanism is based on multiple changes in metal
flow direction and reduction of flow speed, causing
smaller particles to be trapped in the internal filter
structure (figure 7).

Figure 3: Mechanism of uni-dimensional filter types

Figure 7: Deep bed filtration

The efficiency of foam filters in removing typical non-metallic inclusions arising from
the melting, alloying and moulding processes can be seen in Figure 8. In the first
picture, sand grains plug the filter pores. In the second image, slag accumulates in
the top of the filter. In the final image, a thin zone of magnesium sulphide can be
seen across the entire entrance face.

Figure 4: Foam filters

In contrast, foam filters (figure 4) are multi-

dimensional filters. The metal must flow through a
tortuous path before it enters the casting cavity

First, coarse inclusions, too large to enter the

passageways are trapped on the surface of the filter
(figure 5).

Figure 8: Removal of non-metallic inclusions

Finally, foam filters help to prevent the formation of reoxidation inclusion by

Figure 5: Surface filtration
promoting reduced-turbulent flow as the molten metal enters the casting cavity
As inclusions begin to accumulate on the filter face, a (figure 9).
"cake" of material is formed which only then permits
retention of some finer particles (figure 6).

Figure 6: Cake filtration Figure 9: Water model – demonstrating function of slag chamber and filter
6
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 10
Turbulence should be minimised for all alloys and
especially those that oxidise readily. Turbulence
leads to the entrapment of gas inclusions, folding in
of oxides and the creation of new oxides from the
exposure of clean metal to the atmosphere (3).
Filter selection, placement and running system
design are all fundamental to the minimisation of
reoxidation. Extensive water flow testing, x-ray
studies and fluid-flow analysis have proved
conclusively that foam filters, correctly applied,
have the greatest effect on preventing reoxidation.
Pressed filters have larger "dead zones" on the exit
face of the filter which reintroduce air bubbles into
the metal stream after the filter. The exit area of the
Pressed filter Extruded filter
Figure 10: Comparison of water flow through filter types
Pressed filter Extruded filter
Figure 11: Comparison of aerated water flow through filter types
........................................................................................................
water stream from a foam filter is much broader,
indicating a slower, less turbulent flow and there are
almost no air bubbles behind the filter (figure 10).
Repeating the same test with an aerated water flow
shows that both pressed and extruded products have
no effect on entrained air bubbles. In contrast, the
foam filter slows the flow of water so that the
entrained air can escape. This is the same mechanism
that reduces turbulence and gives a foam filter its
higher filtration efficiency (figure 11).
Foam filter
Foam filter
7
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 11
.........................................................................................................

Thus foam filters outperform other types in four

key areas:

❑ smaller, finer particles of inclusion material are

removed;

❑ the reduction in velocity and turbulence reduces

sand erosion:

❑ reduces the risk of reoxidation by entrapped

air due to post-filter turbulence;

❑ greater resistance to thermal breakage in daily

use.

Pouring without filter Benefits of Foam Filtration

The benefits of foam filtration are not only confined
to the reduction in scrap but are seen throughout
the foundry. Depending on the alloy and application
many of the following benefits are achievable:

Scrap control

❑ lower levels of scrap produced;

❑ improved ability to diagnose scrap problems due

to a clearer separation of metal and mould
factors;

❑ reduction in expensive machine scrap;

❑ running systems and downsprues fill smoothly

Pouring with filter
Figure 12: X-ray analysis shows the ability of filtration to
reduce turbulence in molten metal entering the
casting cavity
Further confirmation of the flow smoothing
properties of foam filters has been provided by real-
time x-ray (figure 12).
and remain full, reducing the probability of gas
bubbles entering the casting cavity;
❑ in the case of aluminium filtration non ceramic
foam filters eliminate the dangers of melt
contamination from hard ceramic particles and
iron pick up from steel mesh or sieves.

Attributes of Filters
The attributes of the three basic structures are
shown in the table below.

Attribute Foam Extruded Pressed

Dimensional Accuracy Moderate High High

Cold Strength Moderate Moderate High

Filtration Effectiveness High Moderate Moderate

Particle Retention Size Fine Medium Medium/Coarse

Effect on Flow Rate High Moderate Moderate

Turbulence Reduction High Minimal Minimal

Erosion Resistance High Moderate Moderate

Refractoriness High Moderate Good

8
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 12

Productivity Improvements
❑ lower tonnes shipped to tonnes melted ratio
(yield) due to the elimination of long running
systems designed to float out inclusion leads to
savings throughout the foundry;
❑ simplified gating systems since the filter itself will
reduce velocity and smooth the flow of liquid
metal;
❑ simplified pattern designs can allow a foundry to
reduce mould sizes or make more castings per
mould;
❑ higher yields mean that more moulds can be
poured per melted tonne;
........................................................................................................
References:
(1) SADON, P. HURDEBOURCQ, D. MORISSE, J-C.
BESVILLE, J-C. & TAYLOR, K. C. Industrial
experience in the filtration of cast iron at the
Peugeot Citroen Foundries, Proceedings of
Conaf 2001, 10th Foundry Congress, Sao
Paulo, Brazil, May 23rd to 25th 2001.
(2) KENDRICK, R. Energy saving using SIVEX FC
filters. FOSECO Foundry Practice, Issue 228,
September 1996.
(3) SANDFORD, P. & SIBLEY S.R. The application of
foam filters to optimize aluminum casting
production. FOSECO Foundry Practice, Issue 227,
April 1996.

❑ frequently possible to reduce pouring

temperatures;

❑ low density, non ceramic foam filters float to

the surface of molten aluminium on remelting
and can easily be skimmed off with the dross;

❑ lower levels of cash tied up in work-in-progress.

Casting Quality Improvements

❑ greater consistency in casting cleanliness;

❑ a reduction in the variability of metal flow of

hand-poured castings leading to increased
product consistency (2).

❑ improved mechanical properties due to greater

metallurgical consistency;

❑ increased fluidity of the metal improves casting

finish, appearance and surface details.

After Cast Operations

❑ improved machinability and reduced tool-wear;

❑ machining allowances can be reduced with

confidence due to a lower level of inclusions;

❑ reduction in inspection and expensive time-

consuming non-destructive testing operations;

❑ faster cycle times, especially where there are

major savings in after-cast operations increases
work schedule flexibility and foundry
competitiveness;

❑ non ceramic foam filters can be removed by

machining without damaging cutting or
machining tools.

9
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 13
.........................................................................................................
Application of SEDEX Ceramic Foam
Filters on vertically parted moulds
such as Disamatics
Synopsis
The paper seeks to describe the principles of
applying filters to vertically parted moulds such as
Disamatics. Key differences between horizontally
and vertically parted moulds are outlined. A simple
system to categorise Disamatic mould types into 3
categories is described and the effect of differing
ingate position on filtration application practice are
outlined. The calculation technique for determining
the minimum or controlling gating system cross-
sectional area for vertically parted moulds is
described. The optimum location of the SEDEX filter
in the mould is explained. Filter print designs for
Disamatics are given, their role in securing the filter
in the mould along with filter handling techniques
are described.
Introduction
The use of Disamatic moulding machines has grown
rapidly during the past three decades to such an
extent that there are currently many hundreds of
machines in daily operation world-wide. With their
extremely high productivity, hard mould and vertical
parting line the successful operation of Disamatics
requires considerable specialised knowledge and
technical skill. Gating, feeding, core insertion and
metal pouring technologies have all been specifically
adapted to obtain the maximum benefit from
Disamatic operation.
This also applies to filtration, a foundry technology,
which has also been adapted, ensuring that its
application is able to support the production of
quality cast components at the high levels of
productivity associated with the use of Disamatics.
1. Differences between horizontally and
vertically parted moulds:
There are a number of practical differences
between horizontally and vertically parted moulds;
these exert an influence on basic SEDEX application
principles and the runner system design:
❑ Moulds are more often bottom-filled, so more of
the iron to be poured is likely to be ABOVE the
ingates. This increases back pressure and results
in LOWER effective pouring heights.
❑ Disamatic moulds generally have SHORTER
pouring times, often in the range of 5 to 10
seconds, than horizontal moulds.
10
❑ Larger filter areas may have to be applied to
enable the quantity of iron to be poured in the
short TIME available.
❑ The mould cavities to be poured may often be at
2 OR MORE levels:
❑ To ensure simultaneous filling of the mould
cavities the gating system has to be carefully
CALCULATED and DESIGNED to ensure even,
rapid, non-turbulent filling of each mould level.
❑ Very limited time is available for insertion of the
filter into the mould print, so there is a special
need to devise rapid, accurate methods for
handling the filter; holding it, inserting and
securing it in the Disamatic mould.
❑ Access to the open mould is restricted
The design of good conventional gating systems for
Disamatic applications is a specialised subject in its
own right, it is therefore not surprising that the
design of running systems employing filters has
been slow to evolve.
2. Mould category, its effect on ingate
position and filtration practice:
There are THREE main categories of vertically parted
moulds which influence how filtration can be
applied, namely
Bottom gated mould cavities, these fall into
one of 3 sub-categories:
a) Single casting
b) Two or more castings at the same level
c) Two or more levels of castings
Categories 1a) and 1b): Moulds in these two
categories can be poured using much the same
filtration principles as those applied to horizontal
moulds, with the gating system control or choke
located at the beginning of the horizontal runner
bar(s), after this point the cross-section(s) can be
increased to minimise turbulence.
Category 1c): Moulds in this category are similar
to 1a and 1b but in order to keep the gating system
full during pouring it is necessary to locate the
controlling cross-sections (there may be more than
one) of the gating system at the junction of the
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 14

........................................................................................................
downsprue and horizontal runner bars. After this
point the cross-section of the runner bars and
ingates can be increased in an effort to slow the
flow of metal and minimise turbulent flow into the
mould cavity.

Side gated mould cavities, these fall into one

of 3 sub-categories:
a) Single casting
b) Two or more castings at the same level
c) Two or more levels of castings

Categories 2a) b) and c): Moulds gated in this

way require that the choke in the gating system is
located at the downsprue junction, otherwise the
pouring system cannot fill completely with liquid iron
and the filter will not be correctly primed, this may
permit slag and reaction products to pass through it.

Top gated mould cavities, these fall into one of

3 sub-categories:
a) Single casting
b) Two or more castings at the same level
c) Two or more levels of castings Figure 1: Effect of different mould cavity heights
on gating system dimensions
Categories 3a) b) and c): Moulds gated in this
Figure 1 is taken from a French paper, and shows
way require that the choke in the gating system is
the effect of the dimensional change of the falling
located at the ingates, otherwise the pouring
metal stream and influence of height on the
system cannot fill completely with liquid iron, the
calculation of the ideal gating system dimensions (1).
filter may not prime properly causing slag defects or
The use of a natural tapered downsprue means that
defects associated with gas and/or steam aspiration.
a vertical gating system can be kept completely full
This design is essentially a traditional, ingate
of metal; this reduces turbulence and the risk of gas
controlled or ‘pressurised’ gating system.
or steam aspiration.

Ideally foundries should try to plan their production

Tapered downsprue of 1-3 degrees results in:
methods in such a way as to place the majority of
castings in the categories 1a, b, and c for the best
❑ rapid filling
filtration results, the second best option is
categories 2a, b, and c. Use of these systems offer
❑ minimum turbulence
the founder the possibility of designing gating
systems to minimise mould filling turbulence.
❑ elimination of aspiration

3. Calculation of the controlling or minimum

❑ lower scrap.
gating system cross-sectional area for vertically
parted moulds:
Progressive ingate size reduction:
During pouring, as a stream of liquid metal falls it
accelerates; at any given point in the stream the
❑ controls and regulates the metal velocity
quantity of metal flowing must remain the same,
therefore the cross-sectional area of the stream
❑ compensates for differences in pouring height.
progressively decreases. In contrast with gating
systems designed for horizontally poured moulds,
Clearly there is no single Disamatic casting example
systems for vertically poured moulds can be
which completely summarises how to calculate the
designed to take advantage of natural fluid flow
gating system for use with a ceramic foam filter.
dynamics, resulting in a downsprue which is
There are however a number of general principles
gradually tapered from its top to the bottom.
which can be applied to most Disamatic gating
systems.

For a pattern plate with castings located at different

heights, it is necessary to calculate the effective
pouring height for EACH level of castings and then
to calculate the controlling cross-section for that
level using the standard gating equation:
11
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 15
.........................................................................................................

A = 22,6 x W
e x  x t x H
Friction Factor selection:
The value of the friction loss factor for SEDEX filtered Disamatic gating systems can
vary from 0.25 to 0.65, but is most commonly in the range 0.2 to 0.45. Selection of
the factor depends on the much the same variables which are applied to horizontally
parted moulds, mainly:
A = Controlling cross-sectional area or choke (cm2)
22,6 = Physical constant
❑ casting complexity,
W = Mould or spray weight (kg)
e = Density of iron – (g/cm3)
❑ casting geometry.
 = Friction factor
t = Pouring time (s)
The extra friction arising from the insertion of a foam filter should be compensated
H = Effective pouring height (cm) not mm
for by a reduction of 0,1 in the friction factor value used to perform the calculation
of the gating system dimensions.
The effective pouring height ‘H’ can be calculated
using one of the three simplified formulae:
4. Design of the gating system:
c Calculating the gating system ratios is only the beginning of the process; once that is
1) bottom gating H=h – –
2 done the system has to be designed to do the job intended. A Disamatic gating
2 system has to be calculated in steps from the bottom upwards; as it takes shape it is
2) centre line gating H=h – a
2c ‘assembled’ section by section beginning from the bottom of the mould, progressing
upwards, to give the final result.
3) top gating H–h

Figure 2 indicates the required pattern layout to make maximum use of a plate 650 x
h = parting line to top of mould (mm)
850 mm in size to produce 5 unventilated brake disks. The first judgement to be
a = parting line to top of mould cavity, including
made, bottom gating is not going to be possible, the 5 castings will have to be side
any feeder (mm)
gated; the spray can be classified as falling into category 2 and with two or more
c = overall depth of mould cavity (mm)
levels of castings sub-category ‘c’. It is proposed that castings in category ‘2c’ should
have the controlling gating system cross-section positioned just above and in front of
The controlling cross-sectional area:
the ingates, their section can be increased by approx. 10% to reduce metal velocity
The smallest effective cross-sectional area in any
and turbulence as the stream enters the mould cavity, which is how this system was
gating system is termed the choke, or the section
designed. The second judgement is the friction factor to be applied, with no filter a
which ‘controls’ or regulates the pouring time of the
relatively simple casting of this type would have a friction loss factor of about 0,5,
mould cavity.
however because the use of a foam filter will introduce extra friction losses, therefore
a value of 0,35 was selected.
In the case of a conventional system this controlling
cross-section is usually located at the ingates
The calculation is carried out in steps for each level of casting; this will result in
(pressurised or ingate controlled system) but
gating system dimensions which are suitable for each level of casting. The downsprue
where circumstances permit, in the case of bottom
dimensions for each lower level are added to those of the next level up to ensure
filling the controlling section may also be located in
there is sufficient cross-sectional area to supply metal to each level of mould cavities.
the downsprue (depressurised or downsprue
controlled system). In the case of a downsprue
controlled system it is possible to increase the cross-
sectional runner area behind the choke section by
10% and the ingate area by a further 10%, this
reduces velocity and turbulence.

The calculation of the section to achieve the

required pouring time has to be performed for each
level of castings. The areas are inversely
proportional; as the effective pouring height (H)
INCREASES, the cross-section of the ‘ingates’ and/or
downsprue (A) DECREASES.

Once the total area to achieve the pouring time has

been calculated then it is possible to calculate the
downsprue cross-section, runner bar cross-sections
and the operational casting ingate areas to be
applied to the specific casting(s) to be poured.

The Disamatic Nomogram

The Disamatic gating Nomogram can be very helpful
as a rapid means of determining the cross-sectional
area of the ingate or runner, its use is described in
detail in the Disamatic handbook (2). Figure 2: Basic plate layout for gating system calculation and design
12
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 16

........................................................................................................
The data for the lower pair of castings:
Runner Area Filter position
1268mm2 for DISA
❑ pouring time 6 sec DD1-50x50x15

❑ casting weight 7,8 kg

❑ allocation of runner weight 1,2 kg

❑ pouring height 485 mm

❑ effective pouring height 454 mm

❑ friction factor 0,35

Using the Disamatic Nomogram indicates that the

ingate area required to pour the casting in 6 seconds
is 199 mm2, this was confirmed by manual
calculation and also by the use of an ingate
computer calculation program. However in this case
the control area can be taken as the downsprue and
the ingate area can be increased by 10% of the
controlling section to give 219 mm2; these areas are
valid for both lower castings.
The data for the middle casting same pour time,
friction factor and casting weight as given above.
❑ pouring height 377 mm
❑ effective pouring height 347 mm
The required control section is 228 mm2, this area is
divided into two, half for each side of the casting,
to keep the gating system symmetrical. This 114
mm2 section must be added to the 199 mm2 for
the lower casting (because the downsprue at this
point has to supply 1,5 mould cavities), the ingate
cross-section for each half casting, 114 mm2, is
increased by 10% to give 125,5 mm2 per side.
The data for the UPPER two castings same pour time,
friction factor and casting weight, as given above;
❑ pouring height 216 mm
❑ effective pouring height 176 mm
Pouring time = 6 sec.
Runner system weight = 7,5kg
Figure 3: Finished layout
mould helps to ensure that the upper and lower cavities fill simultaneously and that
the gating system itself is completely filled with liquid metal during pouring. If the
flow rate in the bottom half of the mould is too high this will introduce turbulence
and increase the tendency to produce casting defects such as metal penetration,
erosion, pinholes and gas porosity.
One of the most important features of the design of a SEDEX gating system for
Disamatic application is that it should encourage rapid, even, smooth and non-
turbulent mould filling. The key dimensions of the gating system should be adjusted
according to the distance of the mould cavities from the top of the mould.
i.e. Use SMALLER cross-sections as the ferro-static pressure INCREASES and LARGER
cross-sections as the ferro-static pressure REDUCES.
Some recent simulation work appears to indicate that the introduction of a sloping
downsprue with sharp angular changes in direction may be beneficial in slowing the
flow of metal into the lower cavities. At the time of writing practical foundry
evaluation of this phenomena is still ongoing.
5. Optimum location of the SEDEX filter in the mould:
The choice of filter location in a Disamatic mould is often limited, in the case of a
multiple spray of castings for example, the filter will have to be positioned in the
UPPER half or quarter of the mould. However given the freedom to choose a
location, the upper half of the mould is the preferred option, the advantages are:

The calculated control section is 321 mm2, this area
is added to the previous two areas to give a total
downsprue cross-section of 634 mm2, which is
large enough to supply 2,5 mould cavities, the
ingate cross-section is 10% larger than the 321 mm2
or 353 mm2.
❑ reduction in the high initial iron velocity
❑ minimised risk of breakage
❑ reduction of ferrostatic pressure on the filter which reduces the risk of fluid slag
being pressed through the filter

Figure 3 shows the finished layout with a cross-

section of 634 + 634 mm2 or 1268 mm2 at the top
of the gating system, this is the section which is
able to supply all 5 mould cavities with liquid iron in
the required time of 6 seconds.

This progressive reduction in the cross-sectional area

of the gating system from top to bottom of the
13
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 17
.........................................................................................................

Figure 4: Filter positioned in the UPPER part of the mould Figure 5: Filter positioned in the LOWER part of the mould

This last point is illustrated by the results obtained from a large scale foundry trial in which the effect of filter
type and position in the mould on both foundry and machined casting scrap levels were evaluated. Figures 4 and
5 illustrate the layouts investigated and give the details of the gating systems employed for the trials, whilst
Tables 1 and 2 summarise the results obtained from the two series.

Locating the SEDEX filter at the bottom of the downsprue caused the casting defects to become finer and
more numerous as the filter structure became finer (Table 1). From these results it was concluded that it
would be better to locate the filter higher in the mould. This was evaluated in the second test series (Table 2),
the actual number of casting defects both in the foundry and after machining were significantly reduced.

No. of castings % Foundry No. of castings % Machined

Filter type
produced scrap machined scrap
SEDEX 10ppi 160 0,60 100 9,0
SEDEX 20ppi 170 2,85 100 2,0
SEDEX 30ppi 200 4,0 100 3,0
EXTRUDED 220 1,85 100 9,0
PRESSED 200 2,52 100 20,0

TABLE 1 Filter positioned at the BASE of the downsprue.

No. of castings % Foundry No. of castings % Machined

Filter type
produced scrap machined scrap
SEDEX 10ppi 90 4,7 80 2,5
SEDEX 15ppi 140 3,5 137 0
SEDEX 20ppi 100 1,0 97 0
EXTRUDED 120 13,6 102 2,9
PRESSED 140 9,6 122 1,7

TABLE 2 Filter positioned at the TOP of the downsprue.

Locating the controlling runner system cross-section just behind the filter allows the runner system to fill
completely and the filter to remain fully primed during the pouring sequence.
Cross-sectional areas behind the controlling cross-section can be increased in 10% steps from runner(s) to
ingate(s).

Increasing the area of the runner system section behind the controlling section helps to reduce turbulence and
erosion during mould filling. This aspect of filter use can often be as important to foundries as the filtration
effect itself.

The filter should NOT be inserted standing on its edge but placed horizontally in the mould;

❑ this provides maximum exit face support,

❑ the filter is much less sensitive to dimensional variation either of mould or filter,

❑ it is easier to apply thinner filtration products, provided these give an acceptable filtration result,

❑ there is less risk of the filter falling out of the mould cavity during blowing, movement and closing.
14
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 18
6. Filter print types and their role in securing
the filter in the mould:
It is absolutely essential that a carefully designed
and correctly dimensioned sand print is available
into which the filter can be accurately and securely
positioned. A suitable print has to be designed in
such a way that the filter:
❑ cannot move as the moulding line moves,
❑ cannot fall out as the mould is blown clean.
During pouring the filter print must also provide
adequate support for the filter to ensure that
during pouring:
❑ no breakage occurs,
❑ no unfiltered metal can bypass the filter.
FOSECO Disamatic filter prints are of three basic
types, each with features to suit specific operational
circumstances.
Parallel side wall prints:
The filter is positioned with its side walls at right
angles to the mould joint face, the complete filter
is inserted into the newly moulded exposed back-
face of the Disamatic mould. The filter is held in
position on the Disamatic Coresetter carrier plate
by means of air suction, which MUST be
automatically de-activated when the mould and
carrier plate come together, releasing the filter
which then remains secure in the mould.
Filter print type D1
As illustrated in Figure 6 this type of print is split in
two parts, the main print and a smaller section.
This arrangement gives support on all FOUR edges
of the exit face of the filter. A profile of the smaller
print portion MUST be accurately located on the
Disamatic Coresetter carrier plate to enable correct
positioning of the filter on the plate for insertion in
the mould.
D1 Sizes currently available:
❑ 50 x 50 x 15, and 22 mm thick filters
❑ 40 x 40 x 15 mm
Figure 6: Filter print D1
........................................................................................................
The D1 filter print produces two small tapered
fingers of sand on the rear sides of the print cavity
which hold the filter in its final position in the
mould, in a sort of finger and thumb grip, when
the Coresetter carrier plate is withdrawn. Because
these gripping fingers are under the centreline of
the print cavity, this also permits the use of filters
which are thinner than 22 mm.
Any sand grains which might be dislodged by the
filter as it is pressed into position fall safely into
the small profiled ridge created by the print,
behind the filter.
The D3 filter print illustrated in Figure 7 is basically
the same but it has TWO exits, for those cases
when a double downsprue is required.
Figure 7: Filter print D3
Diagonal Insertion:
Filter print DD1
With this type of print shown in Figure 8 the filter is
positioned in the mould diagonally, once in
position, slightly less than half of the filter is
exposed until the other mould half closes over it,
this locks it in position in the mould. For this
operation to be successfully performed it requires
an accurately dimensioned print, if this condition
cannot be met there is a potential risk of the filter
being crushed upon mould closure and filter pieces
finding their way into the castings.
Use of this print requires that a corresponding half
diamond or triangular shaped recess is also
produced in the Disamatic Coresetter carrier plate,
the filter is placed in this recess. The filter is treated
just like a small core, held in position on the carrier
Figure 8: Filter print DD1
15
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 19
.........................................................................................................
plate by means of air suction which is de-activated
when the carrier plate and the mould come
together, releasing the filter which then remains in
the mould.
To prevent the filter from being withdrawn out of
the mould print when the carrier plate moves away,
sand ridges at the TOP and BOTTOM of the print
grip the REAR EDGES of the filter and retain it in
the mould half until closure completes the
operation of securing the filter in the print cavity.
The print dimensions are such that when the filter is
positioned in the mould, it is pressed firmly into the
sand edges and, as with filter print D1, any sand
grains which might be dislodged by the filter edges
fall into the small profiled ridges created by the
print, safely behind the filter edge.
DD1 Sizes currently available:
❑ 40 x 40 x 15 mm thick filters
❑ 50 x 50 x 15 mm
❑ 50 x 50 x 22 mm
❑ 60 x 60 x 15 mm
Pouring Bush Location:
Filter print DPB3
With this type of print shown in Figure 9 the filter
can be positioned in the print by hand, in the top
surface of the mould.
The filter is basically positioned horizontally but it is
inclined at an angle of approximately 15 degrees;
this makes it easier to position the filter in the print
by hand using a finger and thumb action to insert it
and also helps to prevent the filter floating upwards
during pouring. This design of print is useful to
conduct filtration trials in cases where no Coresetter
is available, no cores have to be placed in the
mould or the foundry does not wish to use the
Coresetter for productivity reasons.
DPB3 Sizes currently available:
❑ 50 x 50 x 15 and 22 mm thick filters
❑ 50 x 75 x 22 mm only
Figure 9: Filter print DPB3
16
7 Placing the filter in the mould:
The filter can be inserted in the mould in a variety
of ways, as follows:
❑ pouring bush of the closed mould
❑ exposed mould face before closure.
❑ specifically designed carrier plate or mask
operating in combination with a Disamatic
Coresetter which inserts the filter into the open
mould half.
The third method is the most widely used insertion
technique so the subjects of filter prints and
Coresetter coremasks are inseparable from one
another. Accurate careful design of the coremask
and its correct alignment with the mould is essential
whether it is being used to position cores, filters
and sleeves or only filters.
Failure to achieve correct coremask and mould
alignment can result in broken filters as the
Coresetter tries to force the filter into the mould
print. Even if only slightly out of alignment the filter
may displace small quantities of moulding sand as it
is moved into position, this sand may fall into the
running system behind the filter and result in
casting defects.
8 Filter handling and carrier plate designs:
There are a number of carrier plate or Coremask
designs which are influenced by the type of filter
print employed, in all cases the mask must line up
exactly with both the pattern plate and mould.
A carrier plate for use in conjunction with filterprint
D1. This requires a profiled mask with a 5 mm deep
rectangular seat 52 x 23 mm, under which there is
either a flat steel plate or two support pins under
the filter which is held in the correct position by air
suction. This design avoids the need for any
supplementary location pins L and R of the filter.
For use in conjunction with filterprint D1 a carrier
plate equipped with an adjustable spring loaded clip
or small compressed air activated ‘fingers’ can be
used.
For use in conjunction with filterprint DD1 a carrier
plate must be produced with a right-angled
triangular pocket (the dimensions and shape of this
must correspond with the diagonal profile of the
filter); this pocket is cut into the carrier plate. The
filter is placed in this pocket and held in place by air
suction, ideally two air orifices being used; this
arrangement ensures that a closed filter edge is
always presented to one of the orifices. The filter is
released, just like a small core, once the coremask
closes onto the open mould half.
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 20
9 Economics of filtration:
Although the use of filtration is now widely
practised in the modern industrialised world,
unfortunately the economic benefits remain rather
poorly documented, foundries perhaps choosing to
remain silent concerning the undoubted economic
benefits which they obtain from the application of
filtration.These are clearly significant given that the
correct use of filters can:
❑ lower scrap levels,
❑ raise quality standards,
❑ improve yield levels,
❑ simplify methoding input,
❑ reduce inspection levels (3),
❑ reduce machining scrap levels (4),
❑ lower the amount of machine tool wear (5, 6),
In a well documented study of economic benefits
prepared by Sadon et al (7) machine shop savings in
excess of 281,000 (ca. $260,000) were shown on
an annual production of 2,7 million vertically
poured ductile iron crankshafts.
........................................................................................................
References:
(1) SCIAMA, G., and JOURNEE, F. Simulation du
remplissage par canaux verticaux degressifs.
Verifications experimentales in Fonderie
Fondeur d’aujourd’hui No 148 p 10 –32,
October 1995
(2) Gating section of the Disamatic application
handbook 1988
(3) DIETERLE; U. and BLECK J. Dross inclusions in
repetitive automotive ductile iron castings – an
example of foundry and supplier working in
partnership to solve a problem. FOSECO
Foundry Practice, Issue 229, February 1999
(4) DELANEY, I. N. SEDEX application to
DISAMATIC moulding machines. FOSECO
Foundry Practice, Issue 222, September 1991
(5) DELANEY, I. N. and WISE, M. L. H. Einfluss der
Filtration auf die Bearbeitkeit von
Bremsscheiben aus Gusseisen mit
Lamellengraphit. Giesserei 80 (1993) Heft 19,
Seite 663-668.
(6) SIMMONS; W. Aspects of foundry practice and
influence on the machinability of iron castings
FOSECO Foundry Practice, Issue 229,
February 1999
(7) SADON, P. HURDEBOURCQ, D. MORISSE, J-C.
BESVILLE, J-C. & TAYLOR, K. C. Industrial
experience in the filtration of cast iron at the
Peugeot Citroen Foundries, Proceedings of
Conaf 2001, 10th Foundry Congress, Sao
Paulo, Brazil, May 23rd to 25th 2001.
17
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 21
.........................................................................................................
Flow Modification Properties of
Ceramic Foam Filters –
A Summary of Recent Work
Abstract
The ability of ceramic foam filters to remove oxides
and other inclusions from the molten metal stream
as it fills steel castings is widely accepted and has
been confirmed repeatedly through post-pouring
evaluation of solidified castings. It has also been
known that the ceramic foam filter minimizes the
formation of reoxidation inclusions through metal
flow modification and reduction of turbulence from
the metal stream as it enters the casting.
However, until recently, the flow modification
capability of filters has only been implied by water
modeling and computer simulations of molten
metal velocity and turbulence. Through refinements
in the use of real-time X-ray techniques to study the
ferrous casting process, it has become possible to
see and record the effects of various gating and
filtration configurations on molten ferrous metal
flow and to confirm the validity of computer
simulations. This paper summarizes recent work in
the field of foam filtration knowledge.
Included in this paper are:
❑ A review of initial work in the use of real-time
x-ray procedures to examine the effect of
filtration and various gating arrangements on
simple plate castings
❑ A summary of recent work showing dynamic
computer simulations of fill profile characteristics
to determine fill times and optimum methoding
❑ Previously reported examples of real-time X-ray
images to confirm the validity of computer
simulations and the effect of filters on casting
filling characteristics and flow modification
❑ Comparisons of previously documented
computer simulation results to real-time X-ray
images of the same castings.
Introduction: The problem of
inclusions in steel castings
Non-metallic inclusions are a major cause of defects
in steel castings. Some, such as furnace slag, ladle
slag, flux and alloying residuals, are generated
outside the mould and are carried into the mould
cavity with the molten metal stream. Others
originate inside the running system and mould
18
cavity; these include mould sand, core and core
coating materials and reoxidation inclusions
generated when the molten metal is exposed to air
in the mould during filling.
Non-metallic inclusions compromise the physical
properties and surface appearance of the casting,
contribute to increased scrap rates or added cleaning
room expense and create greater potential for
customer returns because of machinability problems.
Both prevention and removal tactics are used to
reduce casting inclusions. Melting practices, ladle
design and gating system design can all be
optimized to minimize the occurrence of non-
metallic inclusions in steel castings. One of the most
efficient techniques is the use of ceramic foam
filters (although filtration is not a substitute for
good foundry practices.)
Ceramic Foam Filters
Filtration development history
Early efforts by foundries in the mid 1960's to
remove non-metallic impurities from liquid metal
included passing molten aluminum through a
packed bed of granulated refractory material. Since
then, a variety of other filtration media have been
tried, including steel wool, wire screens, woven
fiberglass fabrics, bonded aggregate, pressed
ceramic strainers and ceramic filters.
Ceramic foam filters and extruded ceramic cellular
filters were introduced in the late 1970's for
aluminum casting and, later, for copper-based
castings and iron castings at pouring temperatures
up to 1500ºC. Further developments in the mid
1980's led to the first ceramic filters suitable for the
filtration of some steel castings. Since the mid
1990's, improved manufacturing methods, the
development of lower-weight, higher-strength
ceramic materials and improved application
practices have made it practical to use ceramic foam
filters for many types of steel castings.
These new materials possess the high-temperature
strength, creep resistance, thermal shock resistance
and low heat capacity to prime effectively and
withstand the physical and thermal stresses of
flowing molten steel, plus the chemical
characteristics necessary for effective filtration.
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 22

How ceramic foam filters work:
Ceramic foam filters remove inclusions from the
molten metal stream and reduce the formation of
non-metallic inclusions in four ways:
❑ Coarse inclusions, too large to enter the
passageways, are trapped on the face of the filter
❑ As pouring continues, inclusions may begin to
accumulate on the filter face and form a "cake"
of material that filters out even finer particles.
❑ Molten metal that flows past the filter cake and
into the passageways follows a tortuous path
through the body of the filter. The filter walls
have a chemical attraction for the remaining
inclusions, causing small inclusions to be trapped
on the internal filter surfaces. It is estimated that
a 100 x 100 mm, 25 mm-thick filter has roughly
one square meter of total surface area.
❑ The filter prevents the formation of reoxidation
inclusions by smoothing the flow of the molten
material and reducing turbulence as the mould
cavity fills. This minimizes the amount of
molten metal surface exposed to oxygen in the
mould cavity.
........................................................................................................
Evidence of flow modification through
physical testing
A variety of physical testing procedures can be used
to gather empirical evidence of the ceramic foam
filter's effect on liquid metal flow.
Airflow Tests
Air-pressure-drop testing, routinely used during the
manufacture of ceramic foam filters to confirm that
production requirements are being met, is an
indication of the filter's ability to modify the flow of
metal during pouring.
Water Modeling
Water modeling tests also disclose the flow
modification properties of ceramic foam filters.
Visual evidence may be seen by holding a filter in
the path of a water stream and comparing the
appearance of the stream as it enters the filter and
as it exits the filter. Figure 1 illustrates the direct
impingement of a stream of tap water on a ceramic
foam filter. Turbulence is clearly visible in the water
entering the filter. The energy absorbing effect of
the filter is illustrated by the smooth flow of water
exiting the filter.

In the past, proof of the filter's ability to entrap

non-metallic inclusions was readily available by
examination and chemical analysis of the filter
portion of solidified gating and the castings
themselves. However, the ability of ceramic foam
filters to modify the flow of the molten metal to
reduce turbulence and prevent reoxidation
inclusions could only be presumed.

The Flow Modification Mechanism Figure 1: Ceramic foam filter reduces turbulence from
stream of water
Fluid flow through porous material
Fluid mechanics investigations have shown that the
steady state flow characteristics of a fluid through a
porous material are dependent on the density and
viscosity of the fluid, as well as the configuration
and porosity of the material itself. Requiring the fluid
stream to pass through the porous material (ceramic
foam filter) results in a reduction of the average
velocity and a drop in pressure as measured at the
entry face and again at the exit face of the filter.

These changes in average velocity and pressure

occur as energy in the metal stream is dissipated, Figure 2: Stream of water passing through strainer core
while the metal follows a tortuous path through the shows little reduction of turbulence
cells of the filter. The drops in pressure and velocity
For comparison purposes, the same demonstration
are the result of two factors. One is energy losses in
can be conducted using pressed strainer cores.
the fluid stream caused by viscous shear as the
Considerably less smoothing of the flow is seen. This
stream is separated by the cell walls in the filter, and
may be attributed to the "dead zones" on the entry
the other is inertial effects that result when the fluid
face of the strainer core that lead to splashing of the
is forced to change directions as it passes through
metal and to the lack of a tortuous pathway within
the filter labyrinth (Innocentini, Salvini, et al, 1999).
the body of the strainer core to create energy
absorbing inertial barriers (figure 2).

19
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 23
.........................................................................................................

Figure 3: Water modeling test apparatus used to determine permeability coefficients

A more scientific method of water modeling

incorporates the test apparatus seen in Figure 3.
It consists of a continuous circuit where water is
pumped from a reservoir, through the filter
mounted in a test chamber, and back to the
reservoir. A differential manometer measures the
pressure drop between the upstream and
downstream sides of the filter. This apparatus has
been used to define the flow coefficients necessary
to accurately model the flow of molten steel
through a filtered gating system and into the
casting cavity using computer simulation (Midea,
Alquist, Blackburn, 2002; Midea, 2001).

Computer Simulation
The use of computer-aided engineering software to
simulate molten metal flow profiles is another tool Figure 4a: Turbulence vectors resulting Figure 4b: Reduction of turbulence
that has been used to investigate the flow when no filtration is used in a vectors after metal passes
direct pouring application through filter in direct
modification capabilities of ceramic foam filters.
pouring application
Computer simulation enables steel foundries to
anticipate molten metal behavior as it flows
through the filter and into the casting cavity. It has
become a basic tool used by many foundries to
effect cost reductions and quality improvements in
the casting process (McMillin, 1999).

Simulation programs have evolved into powerful

tools that solve sophisticated equations, such as the
Navier-Stokes equations, to predict fluid flow
characteristics. Outputs, such as flow velocity and
pressure, can be analyzed to determine the relative
levels of turbulence as the metal flows through the
Figure 5: Computer simulation showing Figure 6: Computer simulation showing
gating and into the casting cavity. Figures 4a and
unfiltered metal entering plate casting filtered metal entering plate casting
4b illustrate the effect of a filter on metal flow as it
enters a valve casting. Figure 4a shows the Figure 5 shows the flow of metal into a plate casting in a direct pour application
turbulence that results when no filter is used and without filtration, whereas Figure 6 shows a direct pour application with a ceramic
the metal impacts directly on the casting wall. foam filter positioned at the base of the direct pour unit. It can be clearly seen that
Figure 4b shows the calming effect of the filter as the metal flows much more evenly and smoothly into the casting after it passes
the metal passes through it (Outten, 1996). through the filter.
20
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 24

A clear understanding of molten metal flow as it
passes through the filter allows steel casters to
better anticipate and avoid turbulence and prevent
reoxidation inclusions. It allows them to avoid
thermal gradients that result in shrinkage and other
casting defects. It also allows them to more
accurately predict fill times and flow characteristics
when selecting specific filters and is an important
issue in designing runner systems and maximizing
foundry efficiency.
Real-time X-Ray analysis of
molten metal flow
X-ray technology is an extremely useful tool to
verify filter flow characteristics. It has been widely
used in numerous casting research programs. A
detailed investigation, using real time x-ray
techniques to evaluate the ability of various filter
positions and orientations to modify metal flow,
was conducted in partnership with the Castings
Technology International of Sheffield, England.
All filters used in the x-ray investigations described
in this paper were 10 ppi (pores per linear inch)
zirconia ceramic foam filters.
Equipment description
The real-time x-ray equipment incorporates a Van
de Graaff x-ray source, a scintillator and a video
recorder as shown in Figure 7. The x-ray source
generates 2.4 M eV of power, which is directed at
the mould cavity positioned between the x-ray
source and the imaging system. The images are
captured on a phosphor screen and relayed to a
camera control using an angled mirror. These
images are video recorded for later review. The
camera is fitted with a close-up lens. This provides
an enhanced image 30.4 cm square. The mould is
positioned approximately 1.82 m away from the
x-ray source; with the scintillator screen another
45.7 cm away.
........................................................................................................
Figure 8: Real-time x-ray test setup
Remote controls are used to operate the bottom-pour
ladle inside the x-ray bunker. Activity within the
bunker is monitored with three remote video cameras.
Initial work – X-Ray analysis of
plate castings
Initial work, previously documented, involved the
pouring of plate castings - some with filters and
some without - to obtain visual confirmation of the
flow control effectiveness of filters in steel casting
(Wingfield, Delaney, Outten, 1999). In addition to
evaluating filter flow control effectiveness, these
tests were designed to compare the effects of
various gating positions and use of the direct
pouring process on flow characteristics as the metal
passes through the gating system.
A vertically parted plate casting was designed to
accommodate the needs of the X-ray process. Since
the thickness of the molding sand was critical to the
clarity of the x-ray images, it was decided to support
the mould within a specially designed steel frame
that reduced the amount of sand required. To
facilitate the various in-gate locations and filtration
arrangements, a core-assembled mould was
developed. Figure 9 shows the mould in the frame.

Figure 7: Schematic of real-time x-ray test apparatus

The mould is placed on a specially designed casting

tray that rests on a load cell. The load cell records
the weight of metal poured as a function of time.
With this arrangement, the volume of metal flowing
through the filter versus overall pouring time (flow
rate) can be determined (figure 8). Figure 9: Plate mould supported in steel frame for x-ray trial
21
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 25
.........................................................................................................

The dimensions of the casting pattern were
designed around the typical casting weight
produced with a 75 x 75 x 25 mm-thick zirconia
ceramic foam filter; this resulted in a plate 600 x
400 x 60 mm thick. The same plate size was used
for all tests.
can clearly be seen escaping from the mould as it fills
(figure 12). This exposure of large surface areas of
molten metal to air produces ideal conditions for the
formation of reoxidation inclusions.

Mould description:

Gating Positions:
Moulds were prepared with both top and side
gating. Some of the top-gated moulds were
prepared with a nominal 50 mm dia central sprue
for direct pouring; others were prepared with a
direct pouring unit that incorporated a 75 dia. x
25 mm-thick zirconia ceramic foam filter.
Figure 10: High speed video showing formation of
Two side-gating set-ups were tested. One reoxidation inclusions
incorporated a nominal 50 mm dia sprue and a
filter print sized to support a 75 x 75 x 25 mm-
thick zirconia ceramic foam filter for in-line
filtration. The other incorporated a direct pouring
unit, positioned on a side riser base, which
included a 75 dia. x 25 mm-thick zirconia ceramic
foam filter.

Filtration:
Four filtration options were evaluated; no filter,
pressed ceramic strainer cores, round ceramic foam
filters (direct pour applications) and square ceramic
foam filters (in-line applications). Where conventional Figure 11: X-ray view of unfiltered top-gated plate casting
gating incorporated a filter print, some castings were early in pouring sequence
poured without the filter and some were poured
with the filter in place for comparison purposes.

Trial Parameters:
Pouring for the tests was done with a 1-tonne
bottom-pour ladle having a 32 mm nozzle
diameter. Tests were made with carbon steel and
the pour weight was 150 kg. Initial ladle
temperatures were between 1600 and 1620ºC. By
the time the X-ray bunker was evacuated and
remote-controlled pouring began, the metal
temperature was between 1590 and 1600ºC.
Figure 12: X-ray view of unfiltered top-gated plate casting
In conjunction with these tests, high-speed slow- later in pouring sequence
motion videography was used to record the flow of
molten steel as it passed through the nozzle of a
bottom-pour ladle. As the metal stream contacts
oxygen in the air during pouring, inclusions,
indicated by the yellow colouration, can be seen
forming on the outer surface of the metal stream
(figure 10).

Trial Results:
In tests of the top-gated molding without the filter,
the metal was poured through the central sprue
directly into the mould cavity, filling the mould in
about eight seconds at a rate of 16.8 kg/sec (figure
11). In the slow-motion video of the pour, severe Figure 13: X-ray view of filtered top-gated plate casting
turbulence of the molten metal stream is evident as during priming of filter
air is entrained during pouring; large gas bubbles
22
FOUNDRY PRACTICE 238 21/5/03 11:04 am
Figure 14: X-ray view of filtered top-gated plate casting
showing reduced turbulence as metal enters
casting
Figure 15: X-ray view of unfiltered side-gated casting
(no filter in filter print) as metal "sprays" into
casting
Figure 16: X-ray view of unfiltered side-gated casting
showing "rolling" action of metal as casting fills
Figure 17: X-ray view of filtered side-gated casting showing
solid stream of metal as it enters casting
Page 26
........................................................................................................
In the same test set-up, this time with the direct
pouring unit in place, the mould filled in about nine
seconds at a rate of 15 kg/sec. The ceramic foam
filter was primed with liquid metal in less than 0.25
seconds (figure 13). The filter controlled the metal
flow, resulting in a head of metal above the filter in
which turbulence and trapped gasses could be
seen. The filter dissipated the turbulence, thus
helping to prevent entrained gasses from entering
the mould cavity.
Slow motion videography of the metal below the
filter clearly shows reduced turbulence in the metal
stream, and there is an absence of the large gas
bubbles seen rising to the surface in the unfiltered
test (figure 14).
When castings were poured with side gating that
incorporated a filter print to accommodate the in-
line ceramic foam filter, equally dramatic results
were obtained. When the mould was filled without
the filter in place, a wide "spray" of metal could be
seen entering the mould cavity (figure 15). As the
cavity filled, a "rolling" action was clearly seen as
the high-velocity incoming metal continued to scour
the walls of the cavity and continued to expose the
metal to reoxidation (figure 16).
With a 75 x 75 x 25 mm-thick ceramic foam filter
inserted in the filter print, the flow below the filter
was no longer a spray, but, rather, a solid stream
of metal (figure 17). This quieter flow filled the
cavity more smoothly, with very little rolling action
and with only the upper surface of the rising metal
exposed to reoxidation. This quieter filling
sequence allows the easy escape of mould gasses
and has a positive effect on the thermal
distribution at the mould wall. More uniform
mould filling may be expected as a result, with an
even distribution of heat throughout the mould
cavity providing ideal conditions for directional
solidification and sounder castings. Also, because
the metal stream remains intact and there is a
non-turbulent mould fill, the potential for cold-
lapping defects is reduced.
Later work – x-ray analysis to confirm
simulation results
While plate castings are ideal for x-ray studies
because of their relatively thin, constant cross
section, practical application of the results is limited.
To gather information that is applicable to typical
castings, an x-ray examination of a steel valve
production casting was conducted (Midea, Alquist,
Blackburn, 2001). As part of the test, the valve was
also modelled using computer simulation. Two
pouring configurations were prepared: a
conventional runner system incorporating a square
in-line zirconia ceramic foam filter; and a direct
pour application, having a round zirconia ceramic
foam filter, that eliminated the need for
conventional runner system components.
23
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 27
.........................................................................................................

Figure 18 shows a computer simulation of the

conventional running system incorporating a
pouring basin, downsprue, an in-line 75 x 75 x 25
mm-thick ceramic foam filter, runner bars and
ingates. Figure 19 shows a computer simulation of
the direct pour arrangement. In the direct pour
application, the unit incorporating an insulating
sleeve and 90 mm dia. x 25 mm ceramic foam filter
was positioned as a side riser on a suitable riser
base. The casting was poured directly through the
direct pour unit that also provides feed metal when
pouring is completed.

Mould description:
In actual production, the overall sand mould size for
the valve casting was 86 x 61 x 46 cm, too large for
optimum x-ray visualization. Thus, a specialized
proprietary investment casting process* was used to
form a highly refractory mould that was
exceptionally strong at both room and elevated
temperatures. Successive coats of slurry and stucco Figure 18: Computer simulation of valve casting rigged with in-line filter and conventional
gating system
were applied to a polystyrene pattern until the
desired thickness - adequate to safely contain the
metal to be poured - was developed. Each coat was
air dried at room temperature prior to application of
the next coat. The dried ceramic shell was then
fired to increase strength and burn away the
polystyrene pattern.

This process produced a relatively thin shell mould

of the valve, leading to clear images of the metal
fronts as they passed through the filter and flowed
into the mould cavity (figure 20).

Trial Parameters:
The valve was cast from ASTM A351.A351M-91a
Grade CF8M stainless steel, deoxidized with CaSi.
Castings were poured from a 907 kg bottom pour
ladle, with 408 kg of steel in the ladle. The pouring
temperature range was 1582-1593ºC.

Observations:
The predicted fill time for the direct poured casting
Figure 19: Computer simulation of valve casting rigged with direct pouring unit (including filter)
(total pour weight of 113 kg) using filter flow data on a side riser base
developed in the water modeling described
previously, is 12.3 seconds; the actual fill time from
foundry practice is known to be 12 seconds. There
is no foundry experience for the configuration of
this casting using a conventional runner system, but
the simulation for this arrangement (total pour
weight of 125 kg) predicted a fill time of 15
seconds, which was confirmed when the castings
were poured during the x-ray trials.

Figure 20: Specialized shell mould used to facilitate x-ray visualization of valve casting during pouring
24
FOUNDRY PRACTICE 238 21/5/03
Figure 21: X-ray view of direct-poured valve casting at
1.5 seconds into filling
Figure 22: X-ray view of direct-poured valve casting at
2.0 seconds into filling
Figure 23: X-ray view of in-line-filtered valve casting at
2.0 seconds into filling
Figure 24: X-ray view of in-line-filtered valve casting at
4.0 seconds into filling
11:04 am Page 28
........................................................................................................
Direct Poured Configuration
At 1.5 seconds, the metal has begun to enter the
casting cavity with minimum turbulence (figure 21).
At 2.0 seconds, the metal is beginning to back up
in the direct pouring unit. The level of the metal
predicted in the simulation generally matches the
x-ray results (figure 22).
In-line Configuration
At 2 seconds into the filling cycle, the computer
simulation does not perfectly predict the actual
filling profile, but there is close agreement with the
x-ray footage (figure 23).
At 4 seconds into the filling cycle, there is general
agreement between the simulation prediction and the
x-ray footage. This is important because one of the
most powerful applications of computer simulation is
the ability to predict the effect of the filter on flow
and thus, porosity/shrinkage (figure 24).
Poured weight vs. time comparisons
A comparison of the information obtained from the
load cell supporting the mould in the x-ray bunker
with the computer-predicted flow rate further
validates the simulation results. Both the measured
and predicted flow rates for the in-line
configuration were 7.7 kg/sec. (figure 25) and 9.1
kg/sec. for the direct pour configuration (figure 26).
Clearly, these results show that computer
simulations, when using accurate filter flow
coefficients, can closely predict flow rates and thus,
fill times.
Figure 25: Poured Weight vs. Time comparison graph, in-line filtration with conventional running system
Figure 26: Poured Weight vs. Time comparison graph, direct-pour unit (including filter)
25
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 29
.........................................................................................................

Summary/conclusion McMillin, T., "Optimize Casting Design and Quality

There is ample evidence that filtration is a valuable
tool in the removal of slag inclusions and the
prevention of reoxidation inclusions in steel
castings. Recent work in computerised simulation
and real-time x-ray examination of the flow
modification properties of ceramic foam filters will
make it possible to refine the selection and
application of ceramic foam filters.
Benefits to foundries include more precise filter size
selections, more accurate fill time predictions and
greater overall foundry efficiency through the
elimination of trial-and-error methoding.
with Process Simulation," Engineered Casting
solutions (Winter, 1999)
Midea, A.C., "Pressure Drop Characteristics of Iron
Filters, " AFS Transactions, 01-042 (2001)
Midea, A.C; Alquist, B.R.; Blackburn, C.; "Increasing
the Accuracy of Metal Flow Results in Steel
Castings," SFSA Transactions, (Oct., 2001)
Outten, J., "KALPUR for steel Direct Pouring System
Improves Yield, Productivity and Quality," Foundry
Practice, No. 227 (1996)

Acknowledgements Wingfield, B.; Delaney, I.; Outten, J.;"A Unique

The authors wish to thank Tony Midea for his work Look at the Benefits of Foam Filtration using X-ray
in directing the water modeling, the simulations Technology;" Presented at the Australian Foundry
and x-ray trials that are the basis for much of the 31st National Convention (1999)
original work reviewed here and for his guidance in
preparing this paper. * The REPLICAST process is a patented process
exclusive to the Castings Technology International,
References UK.
Innocentini, M.D.M.; Salvini, V.R.; Macedo, A.;
Pandolfelli, V.C.; "Prediction of Ceramic Foams
Permeability Using Ergun's Equation," Materials
Research, vol 21, No. 4, pp 283-289 (1999)

Innocentini, M.D.M.; Salvini, V.R.; Pandolfelli, V.C.;

Courey, J.R.; "Assessment of Forchheimer's
Equation to Predict the Permeability of Ceramic
Foams;" American Ceramic Society Journal, vol 82,
No. 7, pp 1945-1948 (1999)

Innocentini, M.D.M.; Salvini, V.R.; Pandolfelli, V.C.;

Courey, J.R.; "The Permeability of Ceramic Foams,"
The American Ceramic Society Bulletin (Sept., 1999)

Innocentini, M.D.M.; Pardo, A.R.F.; Salvini, V.R.;

Pandolfelli, V.C.; "How Accurate is Darcy's Law for
Refractories?," The American Ceramic Society
Bulletin (Nov., 1999)

26
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 30
Bibliography
In the 21 years since the introduction of foam filters to the foundry industry, FOSECO has not only been
responsible for major innovations in filtration technology but has led the field in the development of
application expertise. Typical examples are the many technical papers on this subject which we have written
and presented at conferences and congresses worldwide.
Review iron gating design, filter use to
optimize benefits; Adams A; Modern Casting
(USA), vol. 91, no. 3, pp. 34-36, Mar. 2001
Adams A;
Increasing the accuracy of metal flow results.
Pressure drop studies define new flow
coefficients for ceramic foam filters; Midea T;
Foundry Management and Technology (USA), vol.
129, no. 8, pp. 50-51,56-58, Aug. 2001 Midea T;
Industrial experience in the filtration of cast
iron at the Peugeot Citroen foundries; Sadon P,
Hurdebourcq D, Besville J-C, Taylor KC; Hommes et
Fonderie (France), vol. 314, pp. 24-30, 33-42,
May 2001 Sadon P, Hurdebourcq D, Besville J-C,
Taylor KC;
X-ray techniques provide real-time proof of
foam filter performance in steel castings;
Child N; Foundry Practice, vol. 233, pp. 12-16,
2000 Child N;
How STELEX filters and the KALPUR system
reduce your steel casting manufacturing costs;
Child N; Foundry Practice, vol. 229, pp. 18-20,
1999 Child N;
In praise of ceramic foam filters; Clark H;
Foundry Trade Journal (UK), vol. 173,
no. 3560, pp. 21-22, Nov. 1999 Clark H;
Dross inclusions in double poured castings
made of nodular iron - example of joint
problem solving by foundry and supplier;
Dieterle U, Bleck J; Giesserei (Germany), vol. 86,
no. 2, pp. 44-47, 15 Feb. 1999 Dieterle U, Bleck J;
SIVEX FC foam filters - more than just a filter;
Kendrick R, Wahl HP; Foundry Practice, vol. 229,
pp. 13-17, 1999 Kendrick R, Wahl HP;
Optimizing casting yield using computer
simulation; Midea AC, Christensen D, Outten JF,
Baker P; American Foundrymen's Society, Inc.,
Transactions of the American Foundrymen's Society,
Vol. 107 (USA), pp. 299-305, 1999 Midea AC,
Christensen D, Outten JF, Baker P;
KALPUR direct pour systems for mass
production iron castings; Sovakar A; Foundry
Practice, vol. 231, pp. 1-8, 1999 Sovakar A;
SIVEX FC - A 'Filter' with a ceramic foam
structure especially for aluminium repetition
castings; Wahl HP; Foundry Practice, vol. 232,
pp. 6-11,1999 Wahl HP;
A unique look at the benefits of foam filtration
using X-ray technology; Wingfield B; Australian
Foundry Institute, Casting the New Millenium: 31st
National Convention of the Australian Foundry
Institute Proceedings (Australia), pp. 18, Oct. 1999
Wingfield B;
........................................................................................................
Gray Iron Filtering; Besville JC; Hommes et
Fonderie, no. 186, pp. 7-17, June-July 1988
Besville JC;
Production experiences with improved
filtration of ductile iron castings; Mani PH, Fite S,
Rajan C, Rush L; American Foundrymen's Society,
Inc., Ductile Iron Society's 1998 Keith D. Millis
World Symposium on Ductile Iron Proceedings
(USA), pp. 252-271, Oct. 1998 Mani PH, Fite S,
Rajan C, Rush L;
STELEX ZR filters and KALPUR ST direct pour
units; Child N; Foundry Practice, vol. 228, pp. 1-2,
Sept. 1996 Child N;
Energy savings using SIVEX FC filters;
Kendrick R; Foundry Practice, vol. 228, pp. 13-14,
Sept. 1996 Kendrick R;
Waupaca breaks out of the mold with new
gating approach; Mathems R; Modern Casting
(USA), vol. 86, no. 10, pp. 47-48, Oct. 1996
Mathems R;
Reduce defects with direct pour technology;
Outten J; Foundry Management and Technology
(USA), vol. 124, no. 8, pp. 108-110, Aug. 1996
Outten J;
Optimization of Al casting productivity using
foam filter technology and application;
Sandford P, Sibley SR; Transactions of the American
Foundrymen's Society, Vol. 104, American
Foundrymen's Society, Inc., 1063-1068, 1996
Sandford P, Sibley SR;
Filtration of aluminium-a Weibull statistical
approach to the analysis of mechanical
properties; George RD, Keegan NJ; Foundry
Practice, vol. 226, pp. 14-17, July 1995 George RD,
Keegan NJ;
Aluminium casting productivity increases with
predictive techniques; Jeffs P; Foundry Practice,
vol. 226, pp. 1-2, July 1995 Jeffs P;
Optimising yield for iron castings through
direct pouring technique; Sovakar A;
Foundryman, vol. 88, no. 1, pp. 16-19, Jan. 1995
Sovakar A;
The Effect of Filtration of the Pressure
Tightness of Thin Walled Castings; Powell C;
Foundry Practice, vol. 225, pp. 15-16, June 1994
Powell C;
Advances in Production of High-Yield
Aluminum Castings Using Direct-Pour
Technology; Sandford P; American Foundrymen's
Society, Inc. (United States), pp. 817-824, 1994
Sandford P;
27
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 31
.........................................................................................................
Quality- and Cost Optimisation of a Safety
Critical Component Produced in Aluminium;
Wahl HP; Foundry Practice, vol. 225, pp. 7-8,
June 1994 Wahl HP;
Metal filtration: an easy way to produce
quality castings economically; Karkhanis SR,
Gadagkar VH; 41st Indian Foundry Congress
Transactions, Institute of Indian Foundrymen,
535-546, 1993 Karkhanis SR, Gadagkar VH;
Direct Pouring of Aluminum Improves Yields
and Quality; Pischel RP; Foundry Management and
Technology (USA), 121, (6), 40-42, 44, June 1993
Pischel RP;
Filtration of Liquid Metals; Clark H; Foundryman
(UK), vol. 84, no. 9, pp. 366-368, Oct. 1991 Clark H;
The Use of Ceramic Foam Filters for
Automotive Castings; Clark H, Kallisch W,
Matsuo H; IBF Publications (UK), pp. 34, 1991
Clark H, Kallisch W, Matsuo H;
Replacement of Conventional Runner Systems
for Steel Castings by Direct Pour, Filtration-
Feeder Units; Ecob CM, Moffat GL; Foundry
Practice, vol. 222, pp. 7-11, Sept. 1991 Ecob CM,
Moffat GL;
Ceramic Foam Filtration: a Cleaner Way to Cast
Aluminium; Karkhanis SR; Aluminium: Strategies
for the Nineties and Beyond. Vol. I, Aluminium
Association of India, 269-272, 1991 Karkhanis SR;
The Elimination of Conventional Running
Systems in Aluminium Sand and Gravity Die-
Casting Foundries; Sandford R, Kendrick R;
Foundryman, vol. 83, no. 6, pp. 262-267,
June 1990 Sandford R, Kendrick R;
Gating: the Foundryman's Dilemma, or Fifty
Years of Data and Still Asking "How?";
Wukovich N, Metevelis G; American Foundrymen's
Society, Inc., pp. 285-302, 1990 Wukovich N,
Metevelis G;
FOSECO - the Innovating Developer; Foundry
Trade Journal, vol. 164, no. 3411, pp. 296-297,
4 May 1990;
New Ways to Meet Foundries' Needs; Foundry
Focus, pp. 27, 29, Spring 1990;
New Procedure for the Inoculation of a Mold
With a Filter; Besville J-C; Hommes et Fonderie,
no. 198, pp. 13-21, Oct. 1989 Besville J-C;
Assessing the Efficiency of Ceramic Filters for
Ductile Iron Castings by the Use of Radio-
Active Isotopes; Boyce IS, Knights CF, Plowman
AM, Simmons W; Fonderie Fondeur Aujourd'hui,
no. 82, pp. 24-28, Feb. 1989 Boyce IS, Knights CF,
Plowman AM, Simmons W;
STELEX - an Efficient Filter for the Steel Casting
Industry; Foundry Practice, no. 217, pp. 11-14,
Apr. 1989;
Metals Filtration and Its Effect on the
Properties of Cast Material; Farre S; Gjuteriet,
vol. 78, no. 4, pp. 32, 34-35, 37-39, 42-44, 1988
Farre S;
28
Simplified Running Systems and Improved
Quality in Sand and Gravity Diecasting
Foundries; Sandford P; Foundryman, 81, (3), 110,
112, 114, 116, 118, Mar. 1988 Sandford P;
Gating design; Suschil AL, Plutshack LA; ASM
International, ASM Handbook. Vol. 15. Casting
(USA), 1988, pp. 589-597, 1988 Suschil AL,
Plutshack LA;
Influence of Metal Filtration on the Production
of High Integrity Cast Products; Simmons W,
Broome AJ; South East Asia Iron & Steel Institute,
pp. 20, 1987 Simmons W, Broome AJ;
Aluminum Alloy Filtering; Journal of Metals,
vol. 39, no. 9, pp. 47, Sept. 1987;
Castings Clean-Up, Filter Applications in
Aerospace; Foundry Trade Journal, vol. 161,
no. 3360, pp. 954, 19 Nov. 1987;
Inclusions in Cast Iron; Briggs J, Simmons W;
Foundry Practice., no. 212, pp. 11-15, Mar. 1986
Briggs J, Simmons W;
Metal Filtration; Mehrotra AK; 35th Annual
Convention of the Institute of Indian Foundrymen,
The Institute of Indian Foundrymen, 133-144, 1986
Mehrotra AK;
Machinability and Properties of Irons Filtered
With SEDEX (Ceramic Foam Filters); Menk W,
Simmons W, Wise MLH; Foundry Practice, vol. 212,
pp. 2-4, Mar. 1986 Menk W, Simmons W, Wise MLH;
Optimising Filter Performance; Plowman AM;
Foundry Practice, vol. 212, pp. 8-10, Mar. 1986
Plowman AM;
Filtering Liquid Metals for Enhanced Quality
and Productivity; Sibley S, Simmons W; Non
Ferrous Metals World, (1), 88-94, 1986 Sibley S,
Simmons W;
The Filtering of Molten Metal to Improve
Productivity, Yield, Quality and Properties;
Simmons W; Indian Foundry Journal, vol. 32, no. 8,
pp. 21-28, Aug. 1986 Simmons W;
The Removal of Non-Metallic Inclusions From
Molten Metal by Filtering; Simmons W; South
African Institute of Foundrymen, pp. 35, 1985
Simmons W;
Inclusion Control of Die-Casting Alloys; Miller
SK; Die Cast. Engineer, 28, (5), 54, 56, 58, Sept.-
Oct. 1984 Miller SK;
Effect of Filtration on the Castability of Brake
Discs Made With Lamellar Graphite; Delaney IN,
Aspinwall DK, Wise MLH; Giesserei, vol. 80, no. 19,
pp. 663-668, Oct. 1993 Delaney IN, Aspinwall DK,
Wise MLH;
SIVEX -- a New Cost-Effective Technique for
Cleaning Molten Aluminum; Sibley S; Foundry
Practice, vol. 20, pp. 5-6, Dec. 1980 Sibley S;
FOUNDRY PRACTICE 238 21/5/03 11:04 am Page 32

COMMENT
Editorial policy is to highlight the latest Foseco products and technical developments.
However, because of their newness, some developments may not be immediately available in your area.
Your local Foseco company or agent will be pleased to advise.
FOUNDRY PRACTICE 238 21/5/03 11:03 am Page 1

GIFA
16-21 JUNE 2003
Düsseldorf
Germany

Visit the FOSECO stand in Hall 12

Stand No. 12A05

www.foseco-at-gifa.com

Foseco International Limited

P.O. Box 5516
Tamworth
Design & Production: Warwicks UK Limited, Coventry, England.

Staffordshire
England B78 3XQ
Registered in England No. 468147

ISSN 0266 9994

Printed in England
40003365