NO.

266
MARCH 2019

FOUNDRY PRACTICE
The authoritative magazine for foundry engineers
- SINCE 1932 -

| CRUCIBLES
Thermally-efficient crucible technology

| METALLURGICAL AND POURING CONTROL

SMARTT – defined hydrogen levels after aluminium
rotary degassing

| FILTRATION
Best practice filter application techniques
 EDITORIAL
Dear Readers,
Welcome to our 266TH EDITION of our in-house technical journal of
Foundry Practice. The journal, now in its 87th year, is designed to inform
foundrymen of Foseco’s latest technologies and application techniques
to ensure the ongoing advancement of our customer’s foundry practice.

This edition highlights a new technology in non ferrous metal treatment

and presents fundamental research from our R&D teams on thermally
efficient crucibles and best practice filter application techniques for
vertically parted moulding lines.

SMARTT – DEFINED HYDROGEN LEVELS AFTER ALUMINIUM ROTARY DEGASSING

The SMARTT degassing system is the latest advance in automated aluminium melt quality control. The SMARTT
degassing process takes ambient conditions into account when calculating the optimum treatment parameters to
facilitate the achievement of constant melt quality. Full data logging provides vital information for quality control.

THERMALLY EFFICIENT CRUCIBLE TECHNOLOGY: FUNDAMENTALS, MODELLING AND APPLICATIONS

FOR ENERGY SAVINGS
Calculating the energy savings from the deployment of thermally efficient crucibles in the foundry is extremely
difficult to do due to the number of influencing variables affecting performance and the problems of monitoring
the energy consumption of individual melting and holding furnaces. Through the development of mathematical
models, the benefits of using thermally efficient crucibles become clear and represent an important and underused
source of value for the foundryman.

BEST PRACTICE FILTER APPLICATION TECHNIQUES FOR VERTICALLY PARTED MOLDING MACHINES
The deployment of ceramic foam filters in gating systems of mass produced safety critical parts has become
standard practice in the modern foundry. However, the performance of the filter and the benefits achieved in
inclusion removal, turbulence control and lower scrap rates are strongly influenced by the correct placement of the
filter and the design of the filter print

JOHN SUTHERLAND
Intl Marketing Services Manager
 CONTENTS

03.

18.

13.

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system of any nature or transmitted in any
form or by any means, including photocopying and recording,

03. NON FERROUS FOUNDRIES without the written permission of the copyright holder.
CRUCIBLES
All statements, information and data contained herein are
published as a guide and although believed to be accurate and
Thermally-efficient crucible technology: reliable (having regard to the manufacturer’s practical experience)
Fundamentals, modeling, and applications for neither the manufacturer, licensor, seller nor publisher represents
or warrants, expressly or impliedly:
energy savings
Authors: Brian Pinto & Wenwu Shi (1) their accuracy/reliability
(2) that the use of the product(s) will not infringe third party rights
(3) that no further safety measures are required to meet local
legislation

13. NON FERROUS FOUNDRIES

METALLURGICAL AND POURING CONTROL
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 request.
SMARTT – defined hydrogen levels after
*Foseco, the logo, KALMIN and ENERTEK are Trade Marks of the
aluminium rotary degassing Vesuvius Group, registered in certain countries, used under licence.
Author: Ronny Simon
©Foseco International Ltd.

COMMENT
Editorial policy is to highlight the latest Foseco products and tech-

18. IRON FOUNDRIES

FILTRATION
nical developments. However, because of their newness, some de-
velopments may not be immediately available in your area.
Your local Foseco company or agent will be pleased to advise.

Best practice filter application techniques for PRINTED IN GERMANY

vertically parted molding machines
PUBLISHER:
Authors: Tony Midea, Andy Adams & Brian Dickinson Foseco International Limited
P.O. Box 5516
Tamworth
Staffordshire
England B78 3XQ
THERMALLY-EFFICIENT CRUCIBLE
TECHNOLOGY: FUNDAMENTALS,
MODELLING, AND APPLICATIONS
FOR ENERGY SAVINGS
Authors: Brian Pinto & Wenwu Shi, Foseco NAFTA

Multivariate mathematical models were created to simulate crucibles being used in aluminum foundry applications
with detailed materials characterization data as inputs. The aim was to investigate the effects of crucible geometry and
materials properties changes on the overall energy efficiency of the furnace toward melting and holding metal. Effects of
key thermal properties were also studied to understand their influence on energy efficiency and thermal stresses, another
key factor in understanding crucible behavior. Problems with evaluating these changes practically in foundries stems from
the inability to separate out extrinsic factors that also affect furnace efficiency, such as unique configurations, furnace
condition and, in some cases, poor operating practices. Since melting and holding metal in crucibles accounts for a large
portion of energy demand in the foundry industry, recent advancements in crucible technologies resulting from these
studies could significantly impact cost-efficiency and carbon footprint across the industry. In case studies of applications
such as aluminum melting and holding, considerable improvements in field performance have been reported.

Page 03
Foundry Practice Issue 266
 or fuel combustion generated inside desirable, whereas for holding, slow
the furnace chamber against the heat conduction is best. When a
outer crucible wall is directed to the crucible is used for both melting and
metal charge inside and subsequently holding applications within the same
melts it [11,12]. Literature studies furnace the challenge of creating a
reveal that recommended energy- universally efficient crucible becomes
saving measures are to optimize more apparent. To add to this
the furnace configuration and/or complexity, customer practices across
improve its melting rate [13-16] the industry are so variable that even
with little or no focus on crucibles. correlating a furnace’s efficiency to
If metal is molten, a well-insulated its own crucible becomes extremely
INTRODUCTION furnace expends only nominal energy difficult. For example, if a furnace
The energy used for melting and to keep it at a set temperature, has poor insulation, then the effect of
holding metal accounts for nearly compensating for heat losses to the changing to a high-thermal-efficiency
40% of the total energy costs in a environment. However, to get to this crucible will be completely clouded
typical foundry [1]. Metal casting point requires a tremendous amount by the gross inefficiency of the
industries are known for high energy of heat energy, not only to bring the furnace itself. This has been observed
demands, low energy efficiency metal to its liquidus temperature in many field tests. Although laws of
and high CO2 emissions [2-4]. On and melt it, but also to transmit thermodynamics predict improved
average, the energy consumed by that heat through a thick, high performance, it does not play out this
a foundry shop far exceeds that emissivity ceramic material having way in practice, making it very difficult
which it is predicted to use based on high specific heat capacity, all the to demonstrate an energy-saving
theoretical calculations [5-7]. This is while opposing the thermodynamic crucible to a customer. Therefore, a
due to inefficiencies associated with forces that favor carrying heat away better way to study and, to an extent,
the activities of metal melting and to the atmosphere. The crucible is prove the effects of a crucible on
casting; some are inherent to the a physical barrier between the heat thermal efficiency is to completely
process, while others are dependent source and the molten metal, so it normalize the environment. In
on the types of equipment used as plays a pivotal role in determining practice this is not possible; however,
well as specific practices. There are metal melting efficiency. Thermal using theoretical modeling based
opportunities to improve energy conductivity, specific heat capacity on finite element analysis methods
efficiency of a foundry operation, and geometry are the main factors, it can be done. This paper explores
significantly reducing environmental fixed quantities that govern heat how heat flow behavior and energy
impact while maintaining the sector’s transfer through a crucible. efficiency can be studied based solely
competitiveness in the process [8-10]. on changes made to the crucible
One of the most common methods This appears to provide convenient material properties and design in 2D
used to melt metals is with an electric- solutions for improving furnace and 3D computer models, keeping
resistance or fuel-fired furnace energy efficiency. However, if one the rest of the system constant. In
[11,12]. These furnaces contain considers the many aspects of doing so, the benefits of advanced
molten metal at high temperatures crucible and furnace use across the crucible technologies start to become
within large refractory crucibles. To industry, the solution becomes more clear.
melt, energy from resistive elements complex. For melting, fast heat
conduction through a crucible is very

Page 04
Thermally Efficient Crucible Technology
 EXPERIMENTAL resistance crucible furnace, from refractory, many properties were
which temperature and energy measured, to include: bulk density,
Finite element analysis (FEA) was consumption data were derived. porosity, specific gravity, modulus
performed using ABAQUS 6.11 For simulation in the computer of rupture (MOR), elastic (Young’s)
package with its heat transfer models, multiple crucible types modulus, thermal conductivity,
and temperature-displacement were considered, including both and specific heat capacity (Table
modules. A two-dimensional heat carbon- and ceramic (clay)-bonded I). Energy data collected from
flow model was created based varieties. As with any computer customer trials was done so using
on the model for a typical bowl- simulation, to develop the most a custom energy monitoring
shaped crucible (i.e. BU500) filled realistic model, reliable “real- device (FCTM-2, Foseco) capable
with 400 kg of molten aluminum. world” data are needed to describe of simultaneously monitoring
A three-dimensional model was the materials being tested. From energy usage and molten metal
based on a 100-kW electric- specimens of finished crucible throughput on the furnace.

Property Units Temperature (°C) Ref. ASTM standard

Bulk Density g/cm3 25 C830-00
Apparent Porosity % 25 C830-00
Apparent Specific Gravity - 25 C830-00
Modulus of Rupture MPa 25; 800; 1200 C78-02
Elastic Modulus GPa 25 - 1600 E1875-13
Thermal Conductivity W/m·K 200 - 1000 E1461-13
Specific Heat Capacity J/kg·K 200 - 1000 E1461-13
Table I. List of material property inputs for thermomechanical modeling of crucibles.

RESULTS AND region to absorb the heat, it can end On the underside of the crucible at
up superheated; heat can only be the center (Figure 1B) is its lowest
DISCUSSION dissipated by radiation or downward relative temperature because it heats
A two-dimensional axisymmetric conduction through the wall. This up the slowest. Within the aluminum,
model was constructed for the situation could lead to thermal the lowest temperature position is
express purpose of studying the shock cracks. Fortunately, the model in the top center (Figure 1B) due
effects changes to crucibles (i.e. is somewhat simplistic by assuming to its distance from the elements
geometry; refractory properties) have uniform heat flux; in an actual combined with surface radiation
on heat flow and aluminum melting furnace the heating elements are heat loss. However, since aluminum
efficiency. The model assumes a typically shorter than the crucible is thermal conductivity is much higher
continuous, uniform heat flux is tall, which results in reduced heating that refractory, the temperature
applied to the outside of a crucible of the upper wall. gradient in the metal is much smaller
(Figure 1). The model also assumes than within the crucible walls.
the crucible is partially filled with While this does alleviate
aluminum, allowing the inclusion superheating problems, it tends
of radiative heat transfer from a to create the opposite situation –
molten bath surface and the inside localized underheating, which leads
upper wall of the crucible. Figure to poor glaze protection, oxidation,
1B shows the nodal temperature and eventual thermal shock cracks
contours at 3970 s and 5470 s of anyway. The best practice is to use
the simulation, which demonstrate the furnace in a way that achieves
the temperature gradients within the a balance in these two phenomena;
aluminum and the crucible. Without fill levels should be as high as safely
metal against the crucible upper wall possible to avoid steep temperature
gradients along the crucible wall.

Page 05
Foundry Practice Issue 266
Figure 2 shows results of a heating
simulation focusing on the location
identified as the lowest aluminum
temperature position (‘x’ in Figure
1B) plotted versus time. As shown
in Figure 2A, each curve has three
distinct regions; temperatures rise
very quickly in the first region (I) due
to rapid heat conduction through
solid aluminum. On reaching the
solidus temperature (557°C) the
Figure 1. (A) Two-dimensional crucible model showing heat flux applied on the outside
slope decreases significantly due to surface. (B) Temperature profiles of crucible and molten metal in different time intervals with
the latent heat absorbed for fusion energy-efficient mix (3970 s and 5470 s).
(Hf = 398 kJ/kg), defining the
second region (II). On exceeding the
liquidus (613°C), the temperature
starts to rise quickly again (III). Figure
2A also shows seven different plots,
each of which represents the same
simulation but with a difference in
crucible material (A – F) with pure
graphite (G) as a reference. This
allows for the prediction of time
required to fully melt a specific
aluminum quantity as a function of
crucible composition (Figure 2B). The
process time ranged from 193 min to
234 min for refractory compositions Figure 2. (A) Temperature profiles of the coldest point inside (highlighted in Figure 1) crucible
with different compositions. Latent heat was set as 389 kJ/kg. Solidus temperature is 557oC
(best to worst) and 154 min for pure and the liquidus temperature is 613oC. (B) Estimated time for the molten metal to be heated
graphite. The use of pure graphite in at 750oC.
the model is solely as a theoretical
upper limit for the graphite-
containing refractory compositions For Material A, thermal conductivity and energy cost savings of $8.02 per
(A-F). The reason for differences in is low and specific heat capacity is metric ton.
the melt times for the refractory is high, resulting in the longest time
related to several key properties, required to melt the aluminum, and In addition to material properties,
which, through proper development consequently the highest energy geometric features of a crucible,
can be tailored to produce a more cost. Material B has the highest particularly shape and size, can be
thermally efficient material. The overall thermal conductivity but it highly influential over its energy
two most influential properties in also has a very high specific heat efficiency. Table III compares
this case are thermal conductivity capacity; therefore, the melt time simulations of two different crucible
(k) and specific heat capacity (c). A was only nine minutes less than configurations. One is a relatively
high thermal conductivity means Material A. Through R&D efforts small crucible with 181 kg capacity;
that heat transfer through a material to optimize these properties and the other is a much larger, crucible
is faster than through a material maximize efficiency, melt times were that can hold 816 kg of aluminum. By
with a low thermal conductivity. reduced via Materials C, D and E. altering the crucible geometry and re-
Conversely, a material with high Eventually, Material F was developed, running 2D melting time simulations,
specific heat capacity requires more with high thermal conductivity paired it becomes evident that increasing
absorbed energy to increase its with low specific heat capacity the crucible size has a significant
temperature than one with a low (branded as ENERTEK*). These effect. As shown earlier, a change
specific heat capacity. Table II lists properties, when entered in the to a more efficient crucible material
the thermal conductivity and specific thermal model predicted a 19.2% (from Material E to ENERTEK) alone
heat capacities for different crucible improvement in heating efficiency, results in a net energy cost reduction.
compositions. melt time reduction of 41 minutes

Page 06
Thermally Efficient Crucible Technology
When applied to the small 181 crucible may be beneficial to prevent As with most efforts to improve
kg crucible, the improvement is a wasted energy (keeping a large properties, there are limitations
modest 2.4% per MT. However, by crucible molten until the excess metal and trade-offs. Since crucibles
making the material substitution is completely consumed). For melting are subjected to a wide range of
and also increasing the crucible large quantities of aluminum, a large temperatures and the rate of change
size to 4x capacity, the energy cost crucible is more energy efficient on (T) can vary greatly, thermal stresses
per MT of aluminum melted drops a cost-per-kg basis, but it does take are inevitably generated within the
significantly from $8.02 to $3.23, a longer; time has associated costs as material during use. Cracking failure
61% reduction. This is because the well. and/or reduced longevity are both
mass ratio of crucible to aluminum effects of thermal stresses, since
changes significantly such that refractory materials possess limited
more total energy is used melting ductility. While seeking improved
the aluminum than heating up the thermal efficiency through material
crucible. The absolute masses of changes, the intensity of the residual
refractory and metal are higher in stresses could be unknowingly
the larger crucible; therefore, the increased such that the crucible
total time to melt increases to 351 simply cannot survive the application.
minutes, but the overall melt rate Fortunately, another useful feature
is increased from 0.91 kg/min to of the modeling software permits
2.32 kg/min, an increase of 154%. simulation of thermal stresses as
To melt the equivalent mass in a function of material properties,
the smaller crucible would take at crucible geometry, and temperature.
least 2.5 times as long to achieve, Along with measured mechanical
not including recharging and melt and physical properties data already
transfer time. It is true a smaller entered into the model, temperature
crucible can melt a lesser amount of profiles from actual heating cycles of
aluminum faster, so depending on various crucibles were also collected
the throughput of a foundry a smaller with a datalogger.

Material Thermal Conductivity Specific Heat Capacity Time to Total Cost

(W/m·K) (J/kg·K) Melt (min) Energy ($/MT)
at 200 C at 600 C at 200 C at 600 C
O O O O Use (kWh)
A 7.42 6.69 1200 1892 234 103.5 9.72
B 57.03 42.05 1169 1553 225 99.5 9.34
C 29.33 22.45 1330 1790 223 98.6 9.27
D 31.73 20.86 840 1384 216 95.5 8.97
E 27.92 23.41 891 1316 198 87.5 8.22
F (ENERTEK) 43.06 35.82 825 1133 193 85.3 8.02
Graphite 175 171 710 710 154 68.1 6.39
Table II. Physical properties of different crucible compositions with model-predicted total melting times, energy consumption, and associated
costs.

Material Thermal Conductivity Time to Melting Cost ($/MT)

(W/m·K) Melt (min) Rate
at 200 C
O
at 600OC (kg/min)
E 181 27.9 23.4 198 0.91 8.22
F (ENERTEK) 181 43.1 35.8 193 0.94 8.02
F (ENERTEK) 816 43.1 35.8 351 2.32 3.23
Table III. Comparison of melting time and energy cost for crucibles with different capacities.

Page 07
Foundry Practice Issue 266
Using this added information, thermal
stress states could be predicted using
the temperature-displacement model
in ABAQUS.

Figure 3 shows an example of the

information gained through the
computer model. A crucible made
from a traditional refractory (Material
E) experiences a maximum thermal Figure 3. (A) Comparisons of thermal stress for large crucibles with traditional and thermally
stress of 15 MPa during heating. efficient mix compositions. (B) Comparison of thermal conductivities for two different crucible
materials.
By changing the crucible to a
thermally efficient composition
(ENERTEK), the maximum thermal
stress is reduced significantly, to
8.8MPa. In this situation, efforts
to improve thermal efficiency also
lowered the thermal stress, but this is
not always the case. To illustrate this
point, consider the earlier assertion
that using a larger crucible is better
because thermal efficiency is much
higher. This is true but with an
increase in crucible diameter size, so
does the distance between the lowest
temperature location in the crucible
bottom (Figure 1B) and the heating
elements. This longer conduction
path through the crucible results in
a larger temperature gradient in the
crucible wall, which generates higher
thermal stresses. Figure 4. Predicted maximum thermal stress in crucibles with different dimensions. (A) 615
Shown in Figure 4, a 1055-mm-OD mm OD and 900 mm height, and (B) 1055 mm OD and 1100 mm height. Deformation scale
is 100.
crucible has a much higher thermal
stress (15.8 MPa) compared to one
with a 655-mm-OD (8.9 MPa). The several important features and rather discrete element blocks with a
high stress approaches the strength behaviors of an actual crucible finite size and location in the furnace.
of the crucible refractory itself. For furnace. The configuration and To better simulate this, an improved
this situation, to achieve high thermal position of the electric furnace three-dimensional model based on a
efficiency of large crucibles without heating elements is not well-defined typical electric resistance furnace was
exceeding the material design in the 2D model- a constant surface constructed.
stresses, it is necessary to utilize heat flux is not very realistic. This type
thermally efficient compositions of accuracy is very difficult to achieve
where high thermal conductivity helps since most crucible furnaces operate
to reduce temperature gradients and, around a temperature set point not
in so doing, thermal stress. unlike a thermostat. Thus, the heat
flux experienced by the crucible
Two-dimensional modeling allows exterior is more cyclic in nature,
the rapid calculation of energy with high and low temperatures
efficiency and the study of different bracketing the set point (Figure 5).
compositional effects; however, it is Furthermore, the heat source isn’t a
an oversimplification of a vastly more continuum around the crucible, but
complicated system, neglecting

Page 08
Thermally Efficient Crucible Technology
Figure 6A shows twelve (12) heating
panels distributed around a crucible.
Figure 6B shows the meshes used
for 3D modeling. Since symmetry
still exists within the furnace, one
30-degree segment was modeled
using dimensions scaled to an actual
furnace, taking into consideration the
crucible, aluminum, heating elements,
and insulation. As mentioned earlier,
the heat flux from the elements is
not constant. Figure 6C (black line)
shows the actual power consumed
by the furnace measured with a
data logger. By considering the
power factor, the input to the model
was calculated (red line) to closely Figure 5. Plots of temperature versus time on a 100-kW electric-resistance crucible furnace,
showing the cyclic nature of the heating and cooling (metal and chamber versus fixed set
simulate the actual case. point = 720°C).

The energy was input as body heat

This is because their distance to the ratio of energy used for heating and
flux into 11 rows of tubular elements.
crucible is larger in these areas, which melting the metal to the total energy
Six different heat transfer scenarios
reduces radiative heat transfer rates. expended (x 100%).
were considered for the model:
Like the two-dimensional model, a This exercise reveals that changing
1. Body heat flux input to heating
temperature relative minimum is at the crucible dimensions has an
elements that converts to
the bottom-center of the crucible, increasingly significant effect of
radiation.
where the differential can be as high reducing the mass of the crucible
2. Radiation heat from heating as 300°C. Figures 7C, 7D, and 7E while the volume of aluminum
elements projecting onto the show similar temperature contours (capacity) has increased. Although
crucible exterior. when the aluminum (coldest location) there is little change to the melting
is at 500°C, 600°C, and 700°C. time, the overall energy use is
3. Conduction heat transfer
Rather than repeating the studies reduced per kg of aluminum. For
between heating elements and
performed using the 2D model, it this system the maximum melt rate is
the block insulation.
was decided to use the 3D model increased 15% from 1.25 to 1.44 kg/
4. Conduction heat transfer to study other aspects of crucible min. For the same amount of energy
between the crucible and the geometry with respect to melt expenditure by the furnace, more
aluminum. time. Crucibles were modeled after of it is directed to the metal due to
5. Radiation heat transfer between designs comprised of high-efficiency the lower refractory mass to absorb
insulation and the outside of the refractory material (ENERTEK). Then, it. This increases the efficiency from
crucible. based on the geometric design 65.8% to 72.4%. Over the long-
changes, their energy consumption term this can add up to a significant
6. Radiation heat losses from the and theoretical efficiency were amount of savings. It should be noted
melt surface and top of the calculated and compared. The first that to perform the same simulation
crucible. was a standard crucible design but using data from a typical crucible
the subsequent models were that of material, a similar trend would be
Figures 7A and 7B show a similar shape but with increasingly observed, albeit to a lesser extent in
visualizations of the model with thinner wall cross-sections (larger the absence of the higher efficiency
colors representing component ID). Figure 8 shows a plot of the crucible material.
temperatures (red >> blue) at 1 hr lowest temperature location in the
and 2 hrs, respectively. In this time, melt (circle in Figure 7) for both
the heating elements reach very high crucibles as a function of time. Figure
temperatures, especially toward the 8B lists predicted characteristics
bottom and at the element edges. of both crucibles; ‘efficiency’ is the

Page 09
Foundry Practice Issue 266
Figure 6. (A) Photo showing the distribution of 12 elements (dodecagon). (B) Meshes showing the insulation panel, heating elements, crucible,
and aluminum melt (30o model with 39723 nodes and 35122 elements). (C) Energy consumption measured using an energy meter (kVA) for a
typical melting cycle and estimated input to the finite element model.

Figure 7. Simulated temperature profiles inside an electrical resistance furnace after (A) 1 h and (B) 2 h. Temperature of isolated crucible and
aluminum when nodal temperature (circle) is (C) 500oC, (D) 600oC, and (E) 700oC.

From these simulations it is clear manual sand casting from two near- while still in an industrial setting.
that by utilizing a thermally efficient identical electric resistance furnaces. Additionally, both furnaces were
crucible material coupled with a Furnace use was such that both only used one shift (8 hrs/day) and
lower mass/larger capacity design, were filled but only one was used then idled for the remainder of the
the melting of aluminum can be done at a time; therefore, one furnace time. This presented an opportunity
in a more energy-conscious manner. was always holding while the other to collect energy consumption during
The next logical step was to validate was being used to cast. What made many different modes of furnace
results produced by the simulations. this a particularly good trial site was operation.
An ENERTEK crucible with reduced that both furnaces were being used
mass and increased capacity was for the same operation by the same
manufactured for a special trial at operators, providing the best chance
a US foundry. The application was at minimizing uncontrolled variables

Page 10
Thermally Efficient Crucible Technology
Throughput of the furnace was or $2500 in electricity savings per furnace lid closed more- the benefits
accurately measured using a custom year (est. $0.08/kWh). This also of an energy-saving crucible would
crucible energy/throughput monitor translates to a reduction of 16,573 become more obvious. With
capable of constantly measuring kg of CO2 emissions per furnace per theoretical modeling it is possible to
energy use and able to keep track year. In a foundry that utilizes many eliminate these variables from the
of the amount of metal cast per furnaces, the total savings could be equation- to estimate differences in
day. This allowed for normalization quite substantial. energy efficiency directly influenced
of energy results to the quantity by changes made to crucible geometry
of aluminum cast. Based on an and composition, as well as gain
experiment spanning a six-month
SUMMARY AND insight as to the limits to which these
period where a standard competitor CONCLUSIONS features can be changed to support
crucible was compared to an energy- energy-saving initiatives. It is critically
Using traditional evaluation methods,
efficient ENERTEK crucible (Figure 9), important not to neglect considering
uncontrolled field trials, or simple
energy savings during casting was how changes to composition and/
energy comparisons, it has proven
on the order of 20% in favor of the or geometry will affect the stress
very difficult to justify changing to
energy-efficient crucible (764 kWh/ state of the crucible, particularly as a
an energy-efficient crucible. Almost
MT vs. 605 kWh/MT). function of temperature. Fortunately,
always the benefits are obscured
While holding the total energy use with a nominal amount of additional
in the presence of other foundry
was also reduced, by 14% (30.4 MWh information, these conditions can be
practice-related variables that detract
to 26.0 MWh). Extrapolating from simulated in a computer model as
from equipment efficiency. Were the
this study, it is estimated that for a well. With the ability to understand
foundry to eliminate or minimize
single furnace in constant operation, the characteristics and thermal
these issues; often it is something
the annual potential energy savings behavior of crucibles to a degree
simple like replacing deteriorated
could be as high as 26 MWh, that is relatively unexplored, new
insulation, keeping the
materials were developed that not
only showed high promise in the
theoretical realm, but also showed
definite improvements when applied
to an actual crucible in a real foundry
operation under close surveillance
where actual data collected was able
to validate the computer models.
Extrapolating this achievement
across an entire foundry’s operation
could have large implications with
respect to increased energy savings,
Figure 9. Energy consumption for two different type of crucibles, traditional and thermal
efficient mix with reduced ID used for (A) Casting furnace and (B) Holding furnace for a
minimizing carbon footprint and
6-month testing period. reducing overall costs of operation.

ENERTEK mix
Wall Thickness (43 (37 (31 (25
mm) mm) mm) mm)
Crucible Mass (kg) 173 157 132 111
Al Mass (kg) 353 366 379 403
Melt Time (min) 282 280 279 279
Melt Rate (kg/min) 1.25 1.30 1.36 1.44
Energy Use (kJ/kg) 1461 1400 1341 1264
Efficiency (%) 65.8 68.7 71.6 72.4
Figure. 8 (A) Temperate profiles for the standard crucible and crucible with increased ID. (B) Comparison of weight of crucible, weight of
Aluminum, and melt time, energy consumption, and theoretical efficiency as a function of refractory wall thickness.

Page 11
Foundry Practice Issue 266
These concepts are constantly being 11. Brown, John. Foseco non-ferrous
considered by foundry owners and foundryman’s handbook, 11th CONTACT
managers; with the help of these ed. (Woburn, MA: Butterworth-
and other evaluation tools they can Heinemann, 1999).
begin to understand that something
12. D.M. Stefanescu, ASM
as unassuming as a crucible can have
Handbook Vol. 15 Casting, (ASM
a significant impact on their bottom
International, 2008).
line.
13. A.O. Nieckele, M.F. Naccache,
M.S.P. Gomes. Appl. Therm. Eng.
31, 841 (2011). BRIAN PINTO
REFERENCES PRODUCT DEVELOPMENT
14. R.T. Bui, R. Ouellet. Metall.
MANAGER, N. AMERICA
1. K. Salonitis, B. Zeng, H.A. Mater. Trans. B 21, 487 (1990).
Mehrabi, M. Jolly, Procedia CIRP 15. Trinks, Willibald. Industrial
40, 24 (2016). brian.pinto@vesuvius.com
furnaces. Vol. 1. (John Wiley &
+1 412 505 6528
2. J.Y, Kwon, W. Choate, R. Sons, 2004).
Naranjo. “Advanced Melting 16. K. Pericleous, V. Bojarevics, G.
Technologies: Energy Saving Djambazov, R.A. Harding, M.
Concepts and Opportunities for Wickins, Appl. Math. Model. 30,
the Metal Casting Industry.” US 1262 (2006).
Department of Energy, Metal
Casting Portfolio 7 (2005).
3. P.Rohdin, P.Thollander, P. Solding,
Energy Policy 35, 672 (2007).
WENWU SHI
4. P. Thollander, S. Backlunk, A. SENIOR RESEARCH ENGINEER
Trianni, E. Cagno, Appl. Energy
111, 636 (2013).
wenwu.shi@vesuvius.com
5. Schifo, J. F., and J. T. Radia. +1 412 505 6539
“Theoretical/best practice
energy use in metalcasting
operations.” US Department of
Energy Industrial Technologies
Program, Report (2004).
6. M.R. Jolly, K. Salonitis, F.
Charnley, P. Ball, H. Mehrabi, E.
Pagone. Light Metals, ed. A.P.
Ratvik (New York, NY: Springer,
2017), p. 917.
7. T.E. Norgate, S. Jahanshahi, W. J.
Rankin, J. Cleaner Prod. 15, 838
(2007). DISCOVER MORE
8. R.M. Torielli, R.A. Abrahams,
R.W. Smillie, R.C. Voigt, China
If you are also interested in our
Foundry 8, 74 (2011) new ENERTEK ZnO crucibles,
9. S. Dalquist, T. Gutowski, please press this button.
Proceeding of IMECE. 62599,
(2004).
10. S. Fore, C. T. Mbohwa. J. eng.
design technol. 8, 314 (2010).

Page 12
Thermally Efficient Crucible Technology
SMARTT - DEFINED HYDROGEN
LEVELS AFTER ALUMINIUM
ROTARY DEGASSING
Author: Ronny Simon

The production of Aluminium castings globally is dominated by the automotive industry. To ensure that the correct casting
quality is achieved, a more effective and technically sound melt treatment is essential, followed by a well-designed and
controlled pouring practice. Automotive industry requests process reproducibility and so any melt treatment adopted must
be capable of achieving consistent levels of cleanliness and hydrogen control. Many quality management systems also
require a 100 % record of production data, so again a sophisticated melt treatment with data storage capabilities becomes
more attractive.

Page 13
Foundry Practice Issue 266
 INTRODUCTION Other factors such as ambient AMBIENT CONDITIONS
conditions and melt temperatures
Process control in general refers The melt forms an equilibrium
often vary in much wider ranges.
to the way in which foundries with the water in the surrounding
The influence on degassing is
maintain a tight control over the atmosphere; a warm and humid
usually underestimated or operators
various components and steps climate results in a much higher
change parameters based on their
involved in making castings. The hydrogen content in the melt than a
experiences. Variations in these
importance of process control is dry and cold climate (Figure 1).
starting conditions may cause
derived from the way in which a strict inconsistent results.
adherence to process control helps During rotary degassing the melt is
a foundry avert potentially costly in interaction with the atmosphere.
The hydrogen concentration in the
mistakes. Considering the fact, that The degassing simulation shows the
melt during degassing for various
process control requires a complete effect of different ambient conditions
ambient conditions and melt
monitoring of the various parameters, (Diagram 1).
temperatures has been calculated
any potential problem will be spotted using the Degassing Simulation for
early, before it becomes a significant Likewise, the use of forming gas –
the following widely common set
problem later. a N2-H2 mixed gas - for upgassing
of parameters (Table 1). Variations
procedures ends up with different
of the parameters illustrate the
The intelligent use of process hydrogen levels (Diagram 2).
influence on the degassing result
control technologies within the and the final hydrogen content in the
manufacturing process has beneficial melt after every single treatment.
effects far beyond the traditional
aspects of quality assurance:
• Increase throughput from existing ATL 1000 with 850 kg melt XSR 220 rotor
assets
AlSi7Mg 420 rpm
• Increase automation and reduce
human intervention 750 °C melt temperature 20 l/min inert gas
• Reduce rework, concessions and
50 % relative humidity 20 l/min forming gas with 20 % hydrogen
scrap
• Enhance production capability and 25 °C outside temperature 0,30 ml H2 / 100 g Al starting level
take on more work.
Table 1. Model simulation parameters

PARAMETERS
INFLUENCING
Water vapour
ROTARY TREATMENTS pressure (atm.)

In rotary degassing we differentiate

Hydrogen (ml/100g)

between factors that are almost

constant over longer periods of
time and variable factors. Alloy
composition, vessel geometry and
target melt quality are often well
known and do not change remarkably.
Usually several programs are set in
the PLC, defining treatment time,
rotor speed and gas flow rate. The
operator selects a program following Temperature (°C)

given instructions. The number of

programs is limited, the programs Figure 1. Influence of ambient conditions on hydrogen equilibrium
need to be changed manually in (0,005 atm = 5 °C / 50 % rH;
0,050 atm = 35 °C / 90 % rH)
case of process variations, and the
operator might choose the wrong
program.

Page 14
SMARTT
 Diagram 1. Degassing curves for different ambient conditions Diagram 2. Upgassing curves for different ambient conditions 50%rH / 25°C
0,50
30%rH / 15°C
0.30 0.50 85%rH / 45°C

85%rH / 45°C 0,45

50%rH / 25°C

ml hydrogen / 100 g aluminium

ml hydrogen / 100 g aluminium

0,40
30%rH / 15°C 0.40

Hydrogen content [ ml/100gAl ]

0,35
0.20
0.30
0,30

0,25

0.20
0,20
0.10
85%rH / 45°C
0,15
50%rH / 25°C
0.10
0,10 30%rH / 15°C

0,05
0.00
0.00 5 10 15
0 20
0 2 4 6 8 10 0,00 Times [ min ]
0 2 4 6 8 10 12 14 16 18 20

Treatment time [ min ] Treatment time [ min ]

Diagram 3. Degassing curves for different melt temperatures 0,50

Diagram 4. Upgassing curves for different melt temperatures
0.30 0.50
0,45
750 °C

ml hydrogen / 100 g aluminium

ml hydrogen / 100 g aluminium

700 °C 0,40
0.40
800 °C
0,35

Hydrogen content [ ml/100gAl ]

0.20
0,30
0.30

0,25

0.20
0,20
0.10 750 °C
0,15
700 °C
0.10 800 °C
0,10

0,05
0.00 0.00
0 2 4 6 8 10 0,00 0 5 10 15 20

Treatment time [ min ] Treatment

Times [ time
min ] [ min ]

MELT TEMPERATURE rotary degassing process just before Relative humidity and outside
each treatment. The target for the temperature are measured by a
The melt temperature influences the
optimization is a constant melt standard humidity meter, mounted
equilibrium with the atmosphere as
quality after each treatment. next to the control cabinet in the
well; melt at higher temperatures
area where the treatment takes
dissolves more hydrogen (Diagram 3).
The SMARTT software is installed on place. The actual readings are on-
a Windows PC, input and output of time transferred to SMARTT and
The variations in final results for use
data is carried out on a comfortable recorded over time.
of forming gas are even higher at
touch screen panel with a LAN
different melt temperatures (Diagram
connection to the SIEMENS PLC that A full report on SMARTT is given in
4).
finally controls the degassing unit. Foundry Practice 264 (2015).
A full description of the development
work of “Batch Degassing
Simulation” is given in Foundry
Practice 256 (2011).

SMARTT -
AN INNOVATIVE
PROCESS CONTROL
SMARTT is an acronym for self-
monitoring adaptive recalculation
treatment and an innovative
process control that analyses all
incoming parameters and calculates
the treatment parameters for the Figure 2. Schematic setting of SMARTT

Page 15
Foundry Practice Issue 266
 BU 600 with 530 kg melt 0,06 ml H2 / 100 g Al target
AlSi8Cu3 Standard optimization
750 °C melt temperature 240 s minimum time
XSR 190 rotor 500 s maximum time

Table 2. Process parameters for SMARTT degassing

Figure 3. Treatment parameters for different ambient conditions

PRACTICE OF the same hydrogen content in the provides a defined hydrogen content
melt at the end of each treatment. in treatment gas and ends in an
DEGASSING Foundry trials have shown that the equilibrium between treatment gas,
For different ambient conditions target was always reached regardless aluminum melt and atmosphere.
SMARTT calculates treatment of starting conditions.
parameters to reach a target
PRACTICE OF
Hydrogen
hydrogen content after each content
treatment. With increasing air [ml H2/100g Al] Stages during upgassing treatment

temperature and relative humidity, UPGASSING USING 0.32

the rotor speed and inert gas flow rate FORMING GAS 0.3

increases to compensate the higher 0.28

moisture content in atmosphere. Some applications in foundries 0.26

The optimization always starts at require a defined hydrogen content 0.24

Degassing Upgassing

minimum time, a time that allows such as in the casting of wheels. 0.22

sufficient oxide and inclusion removal It is common practice to run very 0.2

as well. If flow rate and rotor speed short treatment times to avoid too 0.18

are at its specific limit, the software much hydrogen removal; often oxide 0.16
2nd Stage

starts prolonging the treatment time removal is not sufficient. The use of 0.14

a N2-H2 mixed gas improves oxide 1st Stage

to reach the target (Table 2, Figure3). 0.12

A maximum treatment time limits removal due to longer treatment 0.1

temperature loss or melt shortage in times but the variations in hydrogen 0.08

the following casting step. at end of treatment are still high. 0.06

Variations in melt temperature 0.04

before degassing are compensated SMARTT now runs an inert gas 0.02

by SMARTT in a similar way. Finally, treatment followed by a two stage 0

0 50 100 150 200 250 300 350

every treatment is started with upgassing. The 1st stage runs with
different rotor speed, inert gas flow N2-H2 mixed gas only; during stage 2 Diagram 5. Stages of an upgassing
rate and treatment time to achieve a mix between N2-H2 and inert gas procedure

Page 16
SMARTT
Hydrogen transfer into melt becomes ATL 1000 with 850 kg melt 0,08 ml H2 / 100 g Al target for degassing
easier at higher temperatures which
reduces 1st stage time. In this way AlSi7Mg 0,15 ml H2 / 100 g Al final target
2nd stage is influenced as well; the 50 % relative humidity 360 s minimum time
effective hydrogen level in purge gas 25 °C outside temperature 600 s maximum time
gets lower. This value is exactly the
equilibrium between degassing the FDR 220 rotor 45 s dwell time (2nd stage)
melt, hydrogen pickup at melt surface Standard optimization 20 % hydrogen in N2-H2 mixed gas
and upgassing by N2-H2 mixed Table 3. Process parameters for SMARTT upgassing
gas. Under given conditions those
parameters keep the final hydrogen
Rotor Inert gas N2-H2 Time Effective
content in the melt at constant level;
[rpm] [l/min] [l/min] [s] H2 [%]
a dwell time of 30 – 45 s is sufficient
to get into that equilibrium.
Degassing 315 16 0 360 0
The mass flow controller for inert 720 °C 1st Stage 400 0 35 28 20
gas and N2-H2 mixed gas blends the 2nd Stage 400 26 9 45 5,3
correct effective hydrogen content Degassing 303 25 0 360 0
without operator involvement. The 740 °C 1 Stage
st
400 0 35 22 20
differences in effective hydrogen in
purge gas and resulted treatment 2nd Stage 400 28 7 45 3,8
times illustrate the complexity of Degassing 309 30 0 360 0
upgassing; it is obvious that a 760 °C 1st Stage 400 0 35 17 20
computer based simulation only 2nd Stage 400 30 5 45 2,8
can handle all variations in starting
Table 4. Treatment parameters for different temperatures for upgassing
conditions (Table 4).

The latest SMARTT version rotary degassing and predicts the best by a mathematical model. SMARTT,
communicates with either an treatment parameters for different based on the batch degasser
external temperature source or schemes. An integrated report system software, is an intelligent solution to
a handheld thermal couple. An stores all data per treatment in Excel handle data for rotary degassing.
external source can be a temperature format and enables the melt shop
reading that is already available from manager to run further analysis on
treatment crucible or ladle and sent the process. CONTACT
by ethernet or analogue signal to the
SMARTT software. Alternatively, the The use of SMARTT for degassing
operator uses a handheld thermal processes provides a melt on a
couple which is connected directly to constant hydrogen level independent
SMARTT and measures right before from inconsistent starting conditions
every rotary degassing; the reading is in a foundry. SMARTT enables the
used for optimization. foundry to always reach this in a
cost-effective way, there is no need RONNY SIMON
A report system is part of the SMARTT for compensating these variations NON-FERROUS
software package. All treatment data in overrunning the treatment which TECHNOLOGY MANAGER
are stored and available in Excel file wastes time, inert gas and graphite
format. consumables.
ronny.simon@vesuvius.com
In upgassing – often used in wheel +49 2861 83 504
SUMMARY foundries – even small changes in
SMARTT - innovative degassing environmental conditions or melt
control - offers a comfortable temperature have an enormous
interface to program all necessary impact on the hydrogen content DISCOVER MORE
treatment steps, it reads or measures after the treatment. These complex
the starting conditions before every relationships can only be managed
Find out more about our
SMARTT technology.
Page 17
Foundry Practice Issue 266
BEST PRACTICE FILTER APPLICATION
TECHNIQUES FOR VERTICALLY
PARTED MOLDING MACHINES
Author: Tony Midea, A. Adams, B. Dickinson

Vertically parted molding machines were introduced to the foundry industry in the 1960’s, and have since grown
to become the highest grossing method of producing iron casting tonnage. Ceramic foam filters were introduced
in the 1970’s and have matured to become a consistently performing device that is able to meet the production
demands of high speed, vertically parted molding machines, even those with the capability to produce up to
550 molds per hour. Countless filter application methods and techniques have been investigated by foundries,
equipment manufacturers and suppliers alike to develop optimum foam filter applications to meet the high
speed and precision placement requirements of the equipment. Some approaches have proven to be more
successful than others. This initial work focuses on the effect of filter placement in the gating system and the
print design itself on metal flow characteristics and casting quality.

Page 18
Best Practice Filter Application Techniques
 INTRODUCTION fluid dynamics software. Each the flow directly impinges on the
A standard 60x60x22mm of the two iron plate castings is filter itself, as shown in Figure 1. The
(2.36x2.36x0.866inch) square 203x355x19mm (8x14x0.75in) standard filter print is created in the
horizontal filter print was chosen as in dimension and approximately ram side of the mold, and adds about
the baseline configuration to begin 9.75kg (21.45lb) in weight. Total 9% to the gating system weight. The
the analysis. pour weight was approximately 25- gating system weighs 6.36kg (14lbs).
26kg (55-57lb), depending on the
Several modifications were made configuration. For the unfiltered At 0.3 seconds (Figure 2), the flow
to this filter print and runner system, the gating system weighed is just beginning to exit the filter, and
system such that the effect of these 5.82kg (12.8lb). The filter flow the filter print is not yet filled. The
design modifications on fluid flow was represented using 10ppi foam filter, acting as a flow discontinuity,
characteristics could be evaluated. In filtration pressure drop data for a removes a significant amount of
addition, a non-filtered system was 22mm (0.866in) thick filter. Fill time inertia from the flow, and reduces the
evaluated as well as a system with was approximately 11 seconds for all velocity of the metal to approximately
the filter location high in the mold configurations, representing a flow 0.3 to 0.4m/s (11.8 to 15.7in/s). The
to represent multiple casting cavity rate of approximately 2.3kg/s (5lb/s). non-filtered flow shows considerable
molding situations. air entrapment where the sprue
The first comparison is between a meets the runner bar, which increases
All fluid flow analyses were conducted configuration without a filter and a the potential for mold erosion.
using commercially available, configuration with a standard filter
first principles computational print with sprue designed such that

Fig. 1. Casting Configurations with No Filter (Left) and Standard Filter Print (Right)

Fig. 2. Flow Comparison for No Filter and Standard Filter Print Gating at 0.3 Seconds

Page 19
Foundry Practice Issue 266
Air entrapment continues at 0.5 profiles shows that there are significant the same filter print, but with the sprue
seconds (Figure 3) for the non-filtered differences in runner bar metal velocity. moved to the swing side of the pattern
configuration, while a small bubble of plate, as shown in Figure 5. This
air also appears just below the filter for The flow velocity is consistently higher change adds about 4% to the gating
the standard filter print design. Note the for the unfiltered gating system, as system weight, as compared to the
significant difference in flow velocities compared to the gating system with standard filter print design. The gating
between these two systems. the standard filter print located near the system weighs 6.62kg (14.6lb).
bottom of the mold.The next comparison
The runner bar is fully flooded at 0.9 is between the standard filter print
seconds (Figure 4), and the velocity configuration and a configuration with

Fig. 3. Flow Comparison for No Filter and Standard Filter Print Gating at 0.5 Seconds

Fig. 4. Runner Bar Side Centerline Flow Comparison for No Filter and Standard Filter Print Gating at 0.9 Seconds

Fig. 5. Casting Configurations with Standard Filter Print (Left) and with Cross-Over Sprue (Right)

Page 20
Best Practice Filter Application Techniques
With the standard sprue, the metal and most importantly, the flow begins design creates a strong eddy current
enters the filter print in a vertical to wash the filter horizontally, and immediately, and has the possibility
fashion, while for the cross-over sprue, begins forming a strong eddy current to move inclusions into the slag trap
the metal is directed horizontally. at the back of the filter print which during the entire filling cycle. The
This difference results in significantly could help to mechanically move standard filter print takes about 0.5
altered flow characteristics within the inclusions into the slag trap. seconds to create an eddy current,
filter print, clearly apparent in Figure and the current is smaller in size and
6 at 0.35 seconds into the fill. Until finally, at 0.65 seconds (Figure weaker in strength than for the cross-
7), both filter prints are fully flooded over design. Overall fill time between
For the standard gating, the flow and both slag traps exhibit eddy these designs is similar, and not
directly impinges onto the filter and current flow. affected by the flow differences within
begins to prime and flow into the filter. the filter print.
For the cross-over gating, the flow The comparative flow profiles within
impinges on the filter print back wall each filter print remain the same for
and does two things. First, the flow the rest of the filling process. The
begins to prime and enter the filter at main point to take away from these
the back of the filter print. Second, images is the fact that the cross-over

Fig. 6.
Flow Comparison for Standard
Filter Print Gating and Cross-Over
at 0.35 Seconds

Fig. 7.
Flow Comparison for
Standard Filter Print
Gating and Cross-Over
at 0.65 Seconds

Page 21
Foundry Practice Issue 266
Reviewing the flow at the vertically sectioned side Figure 9 shows two other designs that were also
centerline for the whole runner bar, the flow profiles are evaluated for this study, but the results will not be shown
very similar for the two configurations (Figure 8). explicitly here. Please reference the full 2018 Ductile
Iron Society paper of the same title as this article for the
detailed examination.

Fig. 8.
Runner Bar Side
Centerline Flow
Comparison for
Standard Filter
Print Gating and
Cross-Over at
0.9 Seconds

Fig. 9.
Casting
Configuration
with Cross-
Over Sprue and
with Well and
Configuration
with Filter at Top
of Sprue Gating
at 0.9 Seconds

Qualitative, comparative analyses, like the ones shown thus The quantitative evaluation is based upon these three
far in this paper, can provide powerful, convincing imagery main calculated objectives:
of gating system changes that positively or negatively
affect metal flow characteristics. 1) The air entrapment objective criterion calculates the
concentration of gas that has been trapped in the molten
Historically, comparative analyses between gating systems metal due to the collapse of air cavities. Higher values
have provided sufficient evidence to trial and implement indicate unfavorable flow conditions resulting in the
concepts and designs that improve metal flow and casting formation of small blowholes as well as defects due to
quality. However, an engineer is inclined to evaluate chemical reactions. The results are shown as the percentage
design concepts analytically, and to assign absolute values of gasses that has been dissolved in the molten metal.
with visuals. In effect, an engineer desires to combine a
quantitative analysis with a qualitative analysis. 2) The smooth filling objective criterion calculates the average
amount of metal front free surface area during filling, and is
This next section details how practical gating knowledge another measure of the potential for gas related inclusions.
was combined with the software program’s optimization It is calculated as an area, in millimeters.
and design of experiments (DOE) features such that all
five configurations could be simulated and quantitatively 3) The mold erosion criterion is calculated and recorded when
evaluated simultaneously. the metal flow impinging on a mold mesh cell exceeds a
certain velocity for a certain amount of time. This calculation
is complicated, and is properly explained in the full paper.

Page 22
Best Practice Filter Application Techniques
An initial, straight forward approach to evaluating For this objective, Designs 3 and 4 performed the
the various designs is to review how significantly the best, followed by Designs 2, 5 and 1.
configuration affects the individual criterion being
calculated. As an example, Figure 10 shows how each Design Description
configuration, or design, affected the calculation of the
air entrapment filling objective equation. (The red dashed 1. Configuration with no filter
line represents the average criterion result.)

Fig. 10. Main Effect for Air Entrapment Criterion 2. Configuration with standard filter print
Main Effects for Reduce Air Entrapment

3. Configuration with standard filter print,

cross-over sprue
Reduce Air Entrapment

4. Configuration with standard filter print,

cross-over sprue and well at the base

5. Configuration with filter near the top of the

mold

The most powerful part of the evaluation allows the For this analysis, there are three objectives, as discussed
engineer to review the effects of a design on multiple before, but now they can be evaluated simultaneously.
criteria at the same time (Figure 11). The designs are The ideal design would have the lowest calculated value
listed on the far right, and the calculated criteria are for each criterion. However, even if this is not the case,
located on the y-axis. Each calculated criterion is given a the individual results from each design can easily be
unique y-axis, and the values are shown with the criterion compared using this tool.
labeled at the top of the graph. The colored lines are used
to connect the criterion scores for each design. To find the best designs, the top red arrows can be
manipulated to remove the worst designs with the highest
Each design has a uniquely colored line. (Design 1 is calculated values. This is best demonstrated one objective
aqua, Design 2 is blue, Design 3 is red, Design 4 is orange at a time. To begin, Figure 12 shows the evaluation tool
and Design 5 is yellow.) with the “reduce air entrapment” arrow moved down
slightly to eliminate Design 1.

Fig. 11.
Parallel Coordinates Criteria Evaluation

Page 23
Foundry Practice Issue 266
Fig. 12. Parallel Coordinates Criteria Evaluation

Fig. 13. Parallel Coordinates Criteria Evaluation

Note, the line for Design 1 is eliminated,

and disappears from the chart. If the
“reduce mold erosion” arrow is pulled
down below the value of 4.27, the line
for Design 5 is eliminated, as shown in
Figure 13.

Based on these settings and criteria,

Design 2, Design 3 and Design 4 are
the best gating systems. A review of the
remaining criteria shows that there is still
Fig. 14. Parallel Coordinates Criteria Evaluation a large, relative separation in values for
the “reduce air entrapment” criterion,
so the “reduce air entrapment” arrow is
further lowered, thus eliminating the line
for Design 2, as shown in Figure 14.

Designs 3 and 4 are the best designs

based on this evaluation, and have
similar criteria values for all three
objectives. However, there are some small
differences that separate the designs.
By moving the “smooth filling” arrow
below the calculated value of 41,000, as
shown in Figure 14, the line for Design
3 is eliminated and Design 4 is revealed
as the best design of the five evaluated
(Figure 15) on the next page.

Page 24
Best Practice Filter Application Techniques
 Fig. 15.
Parallel Coordinates Criteria Evaluation

When considering all three criteria, Design 4, the cross-over Standard filter print with sprue on the ram side
filter print with a well, is clearly the best gating system. - Filter, acting as a flow discontinuity, removes significant
Design 3 is the second-best gating system, followed by inertia from the system (reduces velocity)
Designs 2, 5 and 1. These results are consistent with the - Creates small eddy current to move inclusions to the slag trap
conclusions from the qualitative evaluation. - 9% increase in gating system weight as compared to
unfiltered system
In general, the conclusions are as follows, starting with the - Recommended design if sprue must remain on ram side
best design based on this analysis.

S tandard filter print with sprue on the swing side *Reference: “Best Practice Filter Application Techniques for Vertically
Parted Molding Machines”, presented at the Ductile Iron Society Keith
and well at the bottom of the sprue
Millis Symposium, 26 October, 2018, Hilton Head, SC.
- Washes filter and quickly creates strong eddy current to
move inclusions to the slag trap
- Less risk of pushing inclusions directly through the filter CONTACT
- Minimal 2.5% increase in gating system weight, as
compared to same system without a well
- Recommended, preferred design

S tandard filter print with sprue on the swing side

but without the well TONY MIDEA ANDY ADAMS
- Washes filter and quickly creates strong eddy current to REGIONAL SIM MGR PRODUCT APPLICATION
move inclusions to the slag trap AMERICAS, JAPAN & KOR MANAGER
- Less risk of pushing inclusions directly through the filter
-  Minimal 4% increase in gating system weight, as tony.midea@vesuvius.com andy.adams@vesuvius.com
+1 440 863 2762 +1 440 863 2754
compared to standard filter print with sprue on the ram
side
- Recommended design if including a well is not possible
due to pattern plate real estate issues

BRIAN DICKINSON
PRODUCT MANAGER - FERROUS
FILTRATION
Check out our
filtration video. brian.dickinson@vesuvius.com
+1 440 863 2773

Page 25
Foundry Practice Issue 266
WE HELP TO SAVE
TONNES OF CO2 EVERY DAY.
By assisting with the construction of reliable wind turbines -
with the help of our products and experts.

Castings are indispensable in the building of wind turbines, that not only generate large volumes of
energy, but are also able to withstand the strongest of storms.

Foundries have relied on having a strong partner at their side for more than 100 years, with innovative solu-
tions, efficient technologies and products of the highest quality. Together with the expertise of experienced
foundry engineers - worldwide and also directly on your doorstep.

FOSECO. Your partner to build on.

FOLLOW US NOW ON
linkedin.com/company/foseco
 E E Y O U AT
S
i f a 2 0 1 9
g
HALL 12
STAND A01/A02

OUR HIGHLIGHTS
FEEDING SYSTEMS
• New, patented range of FEEDEX K VAK spot feeders
• FEEDEX SCK modular feeders
• FEEDEX NF1 range of exothermic feeders for Non Ferrous

FILTRATION
• HOLLOTEX Shroud for Steel application

COATINGS
• ICU - Intelligent Coating Unit coating control system
• INNER CLEANLINESS - Advances in coating
technology
• New range of DYCOTE SAFEGUARD products for Non Ferrous

CRUCIBLES
• New crucibles with a THERMACOAT external
coating layer
• ENERTEK ZnO highlighting the energy and cost
saving potential

MELT SHOP REFRACTORIES

• TRIAD Z range of no cement castables
• New generation of multi-life KALTEK cold-start
insulating ladle linings

METALLURGICAL &
POURING CONTROL
• Latest generation of SMARTT Melt Treatment
Stations and Degassing units for Non Ferrous
• Newly developed FERROLAB V thermal analysis
system

FOSECO. Your partner to build on.

Foseco International Limited ISSN 0266 9994

Printed in Germany
P.O. Box 5516 400121428
Tamworth
Staffordshire
England B78 3XQ
Registered in England No. 468147