Applied Thermal Engineering 44 (2012) 100e107

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Predicting die life from die temperature for high pressure dies casting
aluminium alloy
Alastair Long a, *, David Thornhill a, Cecil Armstrong a, David Watson b
a
School of Mechanical & Aerospace Engineering, Queens University Belfast, Ashby Building, Stranmillis Road, Belfast BT9 5AH, Northern Ireland
b
Ryobi Aluminium Castings (UK) Ltd, 5 Meadowbank Road, Trooperslane Industrial Estate, Carrickfergus, Co. Antrim BT38 8YF, Northern Ireland

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this research was to determine the surface temperature of a high pressure die casting die
Received 13 September 2011 during casting conditions. This was achieved by instrumentation of an insert which was placed in the
Accepted 31 March 2012 shotplate region of the die. This research overcame the challenge of directly measuring the die surface
Available online 7 April 2012
temperature during a HPDC production casting cycle and shows that this is an effective method to
determine the die surface temperature during the casting cycle. The instrumentation results gave a peak
Keywords:
and minimum temperature of 500  C and 240  C respectively during steady state running conditions
High pressure die casting
with a molten aluminium casting temperature of 660  C. Stress analysis from the steady state measured
Surface temperature
Die life
temperature of the die surface was calculated through a simple FEA model and the resulting stress
Thermal fatigue fluctuation was applied to a fatigue equation for the die material, the predicted number of cycles for
Low cycle fatigue cracking to start was found to correlate well with observed die damage.
Heat check cracking Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction
High pressure die casting ‘HPDC’ is a process widely used to
manufacture non-ferrous castings for the automotive industry.
With HPDC the molten metal is forced into the die cavity under
pressure. Typified by high filling speeds and rapid solidification
rates, this casting process can produce shapes which are more
detailed than components manufactured using gravity or low
pressure die casting methods.
Non-ferrous alloys, mainly aluminium, magnesium and zinc are
most commonly cast using this process. HPDC is ideal for high
volume thin walled castings due to the fast cycle times, ranging
from seconds to several minutes depending on casting size and wall
thickness, therefore enabling production in excess of 60 castings
per hour from one casting machine.
The important material properties required for HPDC dies are,
resistance to thermal shock and to softening at elevated tempera-
tures [1]. The most commonly used die materials are H11 and H13
hot work tool steels. The longevity of the die is directly related to
the casting temperature of the molten metal, thermal gradients in
the die and the exposure frequency to the high metal temperature.
Due to the repeated temperature fluctuations on the die, it must be
* Corresponding author. Tel.: þ44 2890974277; fax: þ44 2890661729.
E-mail address: a.d.long@qub.ac.uk (A. Long).
dimensionally stable, have high hot yield strength, good resistance
to high temperature softening and good thermal conductivity [2,3].
HPDC accounts for almost 70% of aluminium components
manufactured today [4]. Many aluminium components for the
automotive industry are cast using this method, due to the high
productivity and near net shape production. Large components
such as gearbox housings and engine blocks are typical examples
where casting weight can be in excess of 15 kg. Due to the short
cycle times, the die is exposed to high temperature fluctuations
each and every casting cycle. The HPDC process involves rapid
temperature fluctuations on the surface of the die, resulting in
steep thermal gradients on and below the die surface [5]. It is very
difficult to measure the die surface temperature throughout the
casting cycle due to the difficulty of instrumentation on the surface
of the die.
1.1. Die temperature
With HPDC, the die surface is rapidly heated when the molten
aluminium is injection into the die cavity. After the casting solid-
ifies, it is ejected and then removed from the die. The die surface is
then cooled by the die-lube spray (mix of oil and water), used to
both lubricate and cool the surface of the die. For HPDC there is no
insulating diecoat layer between the casting and the die, which is
commonly used for gravity and low pressure permanent mould
casting. Therefore, the molten aluminium makes more direct

1359-4311/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2012.03.045
 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107 101

contact with the die surface, encouraging a fast solidification time

due to the high heat transfer coefficient between the die surface
and the casting [6e8]. As there is little insulation between the die
surface and the molten metal, the heat transfer coefficient between
the die and casting along with the peak die surface temperature are
much greater than seen with other casting methods.
Norwood et al. [9] reported, through instrumentation
measurements with thermocouples and temperature paints that
the surface temperature of the die reached between 400  C and
450  C approximately one second after metal injection and that the
die cooled to between 150  C and 200  C due to the die spray, giving
a temperature change of approximately 250  C throughout the
casting cycle.
Srivastava et al. [10] reported that for a typical thermal cycle
during the operation of an aluminium HPDC die, the surface of the
die reaches a peak temperature of 457  C and decrease to
a minimum of 107  C, giving a temperature fluctuation of 350  C
within 20 s.
Venkatasamy [11] measured die surface temperatures ranging
from 500  C to 120  C when a melt temperature of 660  C was used
for casting. Research has shown that as molten aluminium contacts
the die surface, a thin layer freezes almost instantaneously and as
a consequence, the die never reaches the temperature of the
molten metal [12]. Fig. 1. Image of the removable shotplate insert.
The average die running temperature is in the region of 200  C
to 300  C and during each casting cycle the die surface temperature
fluctuates from approximately 100  C to 500  C, resulting in cyclic Fig. 2 shows the location of the shotplate insert relative to the
high stress fluctuations, leading to heat check cracking [13]. Heat runner on slide 3. Slide 3 provides side details on the casting and
check cracking is a result of thermal fatigue and is the main life forms one boundary surface for the runner system. When the
limiting factor for the die associated with HPDC [14]. From more molten metal is injected into the die it will flow through the runner
accurate measurement of the die surface temperature, the system and therefore causes this part of the die to experience some
temperature fluctuations can be used to determine the stress of the highest temperature fluctuations.
fluctuation on the die surface. The stress fluctuations can then be Fig. 3 shows a photograph of the shotplate insert fitted to the
applied to a fatigue prediction model, the results of which can be shotplate. The outline of the runner system can also be seen. This
compared to actual die surface damage. would not be apparent on a new shotplate; it is a consequence of
the scouring caused by the flowing metal over several thousand
casting cycles.
2. Experimental Fig. 4 shows a sectioned view through the slide 3 assembly. This
also gives the three main steps required in order to remove the
2.1. Removable shotplate insert shotplate insert.

In order to better determine the surface temperature of the die Step 1 Remove the outer cover plate, by removing the M5 cap
during the casting cycle, a replaceable insert was designed which head screw holding it in place. This cover plate is put in
could be removed without causing too much disruption to place to ensure that if the cotter pin works loose, it will not
production. This insert was located in the shotplate,1 as this is the fall out and be squashed between the die halves, causing
region of the die which experiences the greatest temperature damage.
fluctuation and peak temperature during the casting cycle. Also by Step 2 Cotter pin is removed by screwing a slide hammer into the
positioning the insert in this location, there was no impact on the cotter pin’s M12 tapped hole, and tapping out the cotter pin.
casting and the die could run in production without affecting the
quality of the product.
The inserts, shown in Fig. 1, were machined from Böhler W302
[15] grade H13 tool steel, 45 HRC and were held in place by a cotter
pin which passes through the centre. The sides of the cylindrical
insert have a 5 draft, and the overall dimensions are as follows, H
48.5 mm, max B39.09 mm and min B30.6 mm.
The inserts were manufactured slightly too tall, so that excess
material could be removed by surface grinding until the inserts
were z0.05 mm above the shotplate surface. The height difference
was kept to a minimum to ensure that the insert did not foul on the
cast runner when the slide opened.

1
The shotplate is the die insert that faces the shot tube (where the metal enters
the die). It is this part of the die first exposed to the injected aluminium alloy and its
function is to direct the metal flow into the runner system. It is the most severely Fig. 2. Images showing the position of the runner system relative to the shotplate
stressed part of the die and is frequently replaced. insert on slide 3.
102 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107
Fig. 3. Photograph showing the fitted insert in the shotplate. The outline of the runner
system scoured on the face of the shotplate can be clearly seen.
Step 3 Remove shotplate insert by using a piece of 8 mm bar,
placed down the cotter pin channel below the insert to
lever it out.
The time taken to remove and replace the insert was approxi-
mately 5 minutes, resulting in a short down time for each insert
change.
2.2. Die instrumentation
The shotplate was instrumented so that the die surface
temperature could be determined.
For the instrumentation trial, temperature was measured using
an IOtech LogBook/360Ô data acquisition system [16]. All trials
were run with an acquisition rate of 60 Hz. To minimise interfer-
ence from the three-phase motors of the HPDC machine the
Logbook was powered by a 12 V DC battery.
2.2.1. Function of the shotplate insert
Trials were run where two thermocouples were placed in the
insert, approximately 0.2 mm from the cast surface. As it was not
possible to drill holes in a production die and there were risks with
having the thermocouples so close to the die surface, the shotplate
insert provided the ideal opportunity to directly measure the die
surface temperature. A number of casting cycles were run to
monitor the die surface temperature fluctuations, during die start
up and in steady state conditions.
Fig. 5 shows the instrumented shotplate insert with two type K
thermocouples welded approximately 0.2 mm from the die surface.
Due to the difficulty of running the thermocouple cables from
Fig. 4. Sectioned view through the shotplate insert, indicating how the insert is
removed.
Fig. 5. Instrumented shotplate insert showing the positions of the thermocouples.
the insert to outside the die, two thermocouples were used, in
case one cable was damaged during the test.
Fig. 6 shows the position of the thermocouples in the
shotplate insert. The first 2 mm of the thermocouple cable was
stripped and then positioned in the hole with approximately 1mm
protruding. The thermocouple tip was then fused to the die surface
with a TIG weld. The resulting weld was polished flat with 1200 grit
silicon carbide paper to leave a smooth surface. The distance
between the fused thermocouples wires and the join to the die
was approximately 0.2 mm from the surface of the insert. To
support the thermocouple wires within the insert the rear of
the thermocouple hole was filled with OMEGABONDÒ 400
cement [17].
2.3. Casting conditions
The instrumentation trial was run on a Toshiba 1650T cold
chamber die casting machine, with the die having a cycle time of
80 s. The die was run for several cycles to investigate the die
temperature during start-up and steady state conditions. The
material being cast was aluminium alloy (AlSi9Cu3Fe). The molten
metal temperature in the holding furnace was maintained at
660  C.
To provide rapid solidification of each casting the die contained
a multiplicity of cooling passages for liquid coolant. This included
a passage that passed to either side of the test insert. The coolant
Fig. 6. Image showing the thermocouple position in the shotplate insert. Left side,
before thermocouple fitted. Right side, thermocouple fixed in position, with weld on
the insert surface.
 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107 103

was kept running at all times, including during interruptions to

producing parts.
The die was made from H13 tool steel and at the start of each
casting cycle was sprayed with Chem Trend safety Lube 7714 for
approximately 2 s.

3. Results

3.1. Temperature measurement

For the trial, the die was started for one cycle, then paused and
run for four more cycles, then paused for a period of time so the die
could cool and then re-started to investigate the change in surface
temperature. These trials were run to investigate die surface
temperature during start-up and during steady state conditions.
The time scale on each of the subsequent graphs all relate back to
the start of the first casting cycle. Fig. 8. Graph showing the die coming to steady state running conditions.
Fig. 7 shows the measured temperature traces from the two
thermocouples placed in the shotplate insert from a cold start
location. As the heat flux is relatively high, this will cause steep
(23  C). For the first cycle, the peak temperature was measured at
temperature gradients within the body of the die, especially close
just 394  C, which then dropped very rapidly to less than 100  C
to the die surface; as much as 30  C/mm when the die first starts to
after 60 s. The die was shown to cool to 39  C after 285 s as the die
fill, though this drops back quite rapidly as the casting solidifies.
remained open. After which, the die was re-started and the peak
Inspection revealed the ends of the holes to be approximately
temperature during the first re-start cycle was 443  C.
0.2 mm from the surface of the insert. Calibration of the individual
Fig. 8 shows the shotplate surface temperature after the die was
thermocouples and data acquisition system using an Isotech Isocal-
restarted from cold. It can be seen that the die casting cycle was
6 thermocouple calibration unit, temperature readings are esti-
extended for several seconds at the end of cycle four which delayed
mated to be within 2  C. Verification of the consistency of the
the stabilization of the die temperature by several cycles. After
temperature readings was provided by comparing readings
approximately six more cycles (cycle 10) the temperatures have
between the thermocouples with a cold die which revealed that
stabilized.
they were all within 0.5  C of each other.
Fig. 9 shows cycle 12 in detail and it can be seen that maximum
Taking into account all of the above it has to be accepted that
and minimum temperatures are 500  C and 240  C, respectively.
absolute accuracy of the temperature in the first few seconds of die
Fig. 10 shows that the shotplate insert surface cools to less than
filling is probably quite poor; perhaps as bad as 10  C. However,
100  C, 180 s after the die was stopped. For these trials the die
the error should be reasonably consistent and as the temperature is
internal water cooling was continually switched on, therefore by
changing so rapidly (as much as 40  C/s) the error could equally be
switching the water off, the die would cool less rapidly and fewer
treated as a phase error of as little as 0.25 s. Later in the cycle the
castings would be scrapped during the subsequent restart if the die
rate of change is much reduced as in turn are the local temperature
was temporarily stopped during production.
gradients within the die; an error of 5  C is probably realistic for
this phase.
3.1.1. Experimental error
The purpose of the experiments was to record the thermal
cycling of the die. By the very nature of the experiments the ther- 3.2. Thermal image analysis
mocouples must be positioned in the path of maximum heat flow
as the heat is transferred from the aluminium into the die and then To illustrate that the two temperature measurements
to the water cooling passages within the die. This presents recorded in the shot plate insert were representative of the
a number of potential areas that can cause an error in the accuracy
of the instantaneous temperature recorded at any particular

Fig. 9. Graph showing cycle 12 after the die re-start when the die is at steady state
Fig. 7. Graph showing the surface temperature of the shotplate insert for die start up. running temperature.
104 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107
Fig. 10. Graph showing the insert temperature when the die was stopped.
whole shotplate and indeed the rest of the die cavity a FlirÔ SC640
[18] thermal imaging camera was utilised. Fig. 11 shows an image of
the shotplate insert during steady state running, just after the die
spray, to represent the coldest part of the casting cycle. The lower
part of the casting die can be seen and the shotplate insert is
indicated by the arrow. From this image it can be seen that the
shotplate insert is approximately 230  C and is slightly cooler than
the rest of the shotplate.
3.3. Numerical calculation
A common approach to estimating the peak die surface
temperature is proposed by Hattel [19]. The theory presumes that
the molten aluminium and the die can be treated as semi-infinite
bodies and that there is perfect contact between the two, thus
neglecting thermal resistance at the interface. A consequence of the
first assumption is that the predicted temperature is independent
of time, which clearly it is not. However, the method should provide
a reasonable estimate of the die surface temperature at the instant
the molten aluminium first makes contact; this approximates to the
peak cycle temperature.
Fig. 12 illustrates the predicted temperature variation between
domain 1 representing the molten aluminium and domain 2 rep-
resenting the die.
Fig. 11. Thermal image of the shotplate region of the die.
The temperature calculated at the interface between the two
domains is the Riemann temperature, Ts, given by:
b1 T1 þ b2 T2
Ts ¼ (1)
b1 þ b2
The Riemann temperature will give the peak die surface
temperature at the
pffiffiffifficasting/die
ffiffiffiffiffiffiffi interface.
where b ¼ kr Cp is called the mould material’s cooling
capacity.
The material properties for the molten aluminium alloy and the
die are as follows:
Aluminium at 660  C Die at 23  C
k1 ¼ 70 W/m K k2 ¼ 25 W/m K
r1 ¼ 2550 kg/m3 r2 ¼ 7830 kg/m3
Cp1 ¼ 1000 J/kg K Cp2 ¼ 460 J/kg K
The aluminium temperature is that recorded in the holding
furnace before being placed in the shot tube and injected into the
die.
Considering the first cycle from Fig. 7 when the die was started
from cold (23  C) the peak measured die surface temperature was
394  C. The corresponding Riemann temperature is 395.5  C.
Considering the assumptions made, this is a surprisingly close
estimate. For other cycles, however, the Riemann temperature
underestimated the peak cycle temperature; in one instant by as
much as 41  C.
3.4. Analysis of the measured die surface temperature
From the measured die surface temperatures, stress fluctuations
were calculated. These stress fluctuations were compared for
different operating conditions. Steady state running condition
stress fluctuation data was then applied to a fatigue prediction
equation which was compared to actual die heat check cracking
recorded from the shotplate.
3.4.1. Simple FE analysis to determine stress/strain from
temperature
ABAQUSÔ was used to calculate the resulting stress from an
applied temperature trace. A column of 100, 0.2 mm elements was
used to model the die surface using C3D8T elements (Coupled
TemperatureeDisplacement Elements). A transient analysis was
Fig. 12. Analytical solution for the two domains.
 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107
used to model the changing temperature trace with time. The
simple column of elements used to model the die is shown in
Fig. 13, the column of elements are restrained in the x and z axes, to
replicate the bulk of the surrounding die material around the
simplified model, but is free to move in the y axis, vertically.
This model was used to calculate the resulting stress and strain
from a temperature applied to the top surface of the model as
indicated. This simplified method was used due to the fast calcu-
lation time and ease of change of parameters, such as the applied
temperature trace.
3.5. Die surface temperature investigation
To determine the influence of the die surface temperature,
a comparison was made between a normal die start up to an
interrupted die start up (where the die was interrupted after the
first casting cycle). From this temperature measurement
a comparison between a normal run (die runs without stopping)
and an interrupted run (die does not run continuously and stops
often) was made.
The die start-up comparison is shown in Fig. 14, from which
the temperature data trace was analysed by ABAQUSÔ and shows
the temperature plot comparison of the normal and interrupted
die start-up, showing the temperature plot with time over 800 s.
It can be seen that for the normal start, the temperature fluc-
tuations are consistent and the peak temperature increases as the
die warms. For the interrupted die start, it can be seen that the die
surface temperature drops back to less than 50  C as the die was
stopped after the first casting cycle. When the die re-starts, it is
consistent with the normal start-up. The influence of the normal
and interrupted start-up is also seen in Fig. 15 where the resulting
stress vs. strain plots are shown.
Fig. 13. Column of elements used for the simplified stress analysis.
105
Fig. 14. Graph showing the start-up temperature curves from a normal and inter-
rupted die start-up.
From Fig. 15, a stressestrain comparison between the normal
and interrupted die start is shown, it can be seen that the inter-
rupted start shows a much greater fluctuation in the graph than the
normal start, indicating that the stress and strain fluctuations for
the interrupted start will be greater than that of the normal start.
Table 1 shows the stress and strain fluctuations for both the
normal and interrupted die start. The results show that the initial
stress and strain fluctuations are similar, however the greatest
change is seen when the die remains open for an extended time, as
the die surface temperature drops at cycle 5 for the interrupted die
start-up. When the die cools to room temperature it will experience
greater stress and strain fluctuations, þ57% stress and þ48% strain.
However the increased fluctuations are even more detrimental to
the die surface as by cooling the die surface to room temperature,
a great deal of tensile stress develops on the surface of the die. In
this case, the peak tensile stress will increase by 506% from 109 to
660 MPa, therefore greatly increasing the likelihood of the die
surface cracks growing. The more frequently the die is cooled to
room temperature, the more cracks on the surface of the die will
open and grow, resulting in more widespread and deeper cracking.
3.6. Fatigue prediction of cracking on the shotplate
The stress fluctuation at the shotplate insert from Fig. 15 was
determined during steady state production and found to be,
Ds ¼ 842 MPa from Table 1. This stress fluctuation was then placed
Stress strain plot of normal and interrupted die start up
8.0E+08
6.0E+08
4.0E+08
Stress Pa
2.0E+08
0.0E+00
-0.006 -0.005 -0.004 -0.003 -0.002 -0.001 0.000
-2.0E+08
-4.0E+08
-6.0E+08
-8.0E+08
Initial cycle
-1.0E+09
-1.2E+09
Total strain
Normal die start Interrupted die start
Fig. 15. Graph showing a stress strain plot of the normal and interrupted die start up.
106 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107
Table 1
Stress fluctuations for normal and interrupted die start-up.
Ds, MPa Cycle number
1 2 3 4 5 6
Normal 980 970 933 1116 942 924
Interrupted 953 1,046 949 938 1475
into Eq. (2) [20] to determine the number of cycles before die
cracking would occur.
Ds
¼ 1601  106 N 0:148
2
The stress fluctuation gave a result of 8311 cycles as the pre-
dicted number of cycles for cracking to start. This was compared to
photographs of the shot plate on two production dies (identified as
dies 3 and 7) taken at irregular intervals as production allowed. For
die 3, no cracking was apparent in the photograph recorded at 3151
cycles. The first indication of cracking was apparent in the photo-
graph taken at 7763 cycles (Fig. 16). The cracks are hard to pick out
from the photographs as they are partially obscured by machining
marks which are visible as evenly spaced light and dark lines which
are approximately horizontal in the centre of the picture. The
cracks are jagged dark lines that are approximately vertical in the
centre of the picture.
A comparison was made with the shotplate of a second die (die
7). Fig. 17 shows the die surface at 5368 cycles and Fig. 18 shows the
die surface after 7906 cycles. The outline of the shotplate insert is
shown in red with the location of the measured temperature on the
die surface shown by the white dots.
The approximately radial grey lines apparent in Fig. 17 are
not yet considered to be cracks. Fig. 18 shows the die surface after
7906 cycles and the light grey radial marks have been replaced
by darker lines suggesting that cracks are starting to form. This is
405 cycles, or 5%, earlier than predicted. Given that the fatigue
equation is based on visual observations of failure this is not
unreasonable.
4. Discussion
From the trials with the instrumented shotplate insert, it can be
seen that the temperature of the surface of the shotplate during the
casting cycle during steady running conditions is between 500  C
and 240  C, respectively. The initial die surface temperature during
the initial casting cycle was recorded at just 394  C when started
from cold.
Fig. 16. Image of the shotplate from die 3 after 7763 cycles with the area of heat
checking highlighted within the red outline. The white dot shows the location of one of
the two thermocouples. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
7 8
876 842
(2)
Fig. 17. Image of the shotplate from die 7 after 5368 cycles.
The trial also highlighted that the die will lose heat very quickly
if the internal cooling water is not switched off when the die is
stopped, the surface of the shotplate insert cooled to less than
100  C, 180 s after the die was stopped.
As the shotplate directs the molten metal into the die, it is fair to
assume that this region will experience the peak temperature
during the die fill. The rest of the die should experience a lower
maximum surface temperature and lower minimum surface
temperature during the die spray. Cooler regions of the die, further
from the shotplate will experience a lower average running
temperature as the molten metal cools as it passes through the die
and as the die lube will have a more noticeable reduction in die
surface temperature in these areas. Protruding regions of the die
may heat more that the shotplate, such as core pins which will be
surrounded by molten metal during the die fill.
If other areas of the die were instrumented using the same
method as the shotplate insert, a more complete picture of the die
surface temperature at different regions of the die could be
achieved.
The results from the numerical calculations to predict the peak
die surface temperature were within 10% of the measured
temperature.
The thermal image taken of the die showed that the tempera-
tures recorded by the thermocouples were representative of much
of the runner system during the die open part of the cycle.
The resulting stress calculations show that stress fluctuations
during an interrupted casting cycle were as much as 57% greater
than the standard cycle, reducing the die life when the casting
process is stopped often.
Fig. 18. Image of the shotplate from die 7 after 7906 cycles.
 A. Long et al. / Applied Thermal Engineering 44 (2012) 100e107
The stress fluctuation of the standard cycle when applied to the
fatigue life equation gave a predicted number of casting cycles
consistent with the onset of heat check cracking observed experi-
mentally on two different die casting shot plates.
The instrumentation results show that by using this method of
placing an instrumented insert into the production die, accurate
temperature traces of the complete die cycle can be measured
without interrupting/affecting production. This same method
should be suitable for instrumentation of any insert within the die
cavity to give further information of the temperature change
throughout the die cavity. It should be noted that HPDC dies
commonly have multiple inserts in their die construction.
5. Conclusions
The peak cycle surface temperature of the shotplate was
measured at just 394  C when started from cold. During steady
state running conditions, the shotplate surface temperature fluc-
tuated between 500  C and 240  C. When the die is stopped, heat is
rapidly lost. Therefore, to slow the rate of heat loss from the die,
internal water cooling should be switched off when the die is
stopped for a short period of time.
The stress analysis which compared the standard to the inter-
rupted die cycle indicates that frequently stopping and cooling the
die to room temperature, will result in more widespread and
deeper cracking on the die surface.
The comparison between the predicted results and the obser-
vations made from two shotplates shows that the fatigue life
prediction calculation for material W302 is consistent with the
production die.
Acknowledgements
This research was financially supported by Invest Northern
Ireland under Start Project ST247. The experimental work was
carried out at Ryobi Aluminium Castings (UK) Ltd. Technical support
and instrumentation was provided by Queen’s University Belfast.
References
[1] J.R. Davis, Tool Materials, ASM International, s.l., 1995, ISBN 0-87170-545-1.
[2] R. Ehrhardt, K.-D. Fuchs, B.M. Gehricke, J. Buckstegge, I. Schruff, Improving the
heat checking characteristics of ALSI H13 by modification of the chemical
composition, Die Casting Engineer. (2003) 30e37.
[3] D. Klobcar, et al., Thermo fatigue cracking of die casting dies, Engineering
Failure Analysis. 20 (2012).
107
[4] Bill Andresen, Die Casting Engineering, Marcel Dekker, New York, 2005, ISBN
0-8247-5935-4.
[5] D. Klobcar, J. Tusek, Thermal stresses in aluminium alloy die casting dies,
Compution Materials Science 43 (2008).
[6] Z. Guo, S. Xiong, B. Lui, M. Li, J. Allison, Determination of the heat
transfer coefficient at metal-die interface of high pressure die casting process
of AM50 alloy, International Journal of heat and Mass Transfer 51 (2008).
[7] M.S. Dargusch, A. Hamasaiid, G. Dour, D. Loulou, C.J. Davidson, D.H. St John,
The accurate determination of heat transfer coefficient and its evolution with
time during high pressure die casting of Ale9%Sie3%Cu and Mge9%Ale1%Zn
alloys, Advanced Engineering materials 9 (2007) 11.
[8] A.D. Long, et al., Determination of the heat transfer coefficient at the metaledie
interface for high pressure die cast AlSi9Cu3Fe, Applied Thermal Engineering 31
(2011) 17e18.
[9] A.J. Norwood, P.M. Dickens, R.C. Soar, R.A. Harris, Surface temperature of tools
during the high-pressure die casting of aluminium, Journal of Engineering
Manufacture. 221 (2007) B.
[10] A. Srivastava, V. Joshi, R. Shivpuri, R. Bhattacharya, S. Dixit, A multilayer
coating architecture to reduce heat checking of die surfaces, Surface and
Coating Technology, Elsevier, 2003.
[11] V.K. Venkatasamy, A Thesis, Analysis of In-cavity Thermal and Pressure
Characteristics in Aluminum Alloy Die Casting, The Ohio State University,
Ohio, 1996.
[12] Z.W. Chen, Skin solidification during high pressure die casting of
Ale11Sie2Cue1Fe alloy, Elsevier, Jounal of Material Science and Engineering.
A (2003) 348.
[13] A. Persson, S. Hogmark, J. Bergstorm, Temperature profiles and conditions of
thermal fatigue cracks in brass die casting dies, Journal of Materials Processing
Technology 152 (2004) 228e236.
[14] A. Persson, S. Hogmark, J. Bergstrom, Simulation and evaluation of thermal
fatigue cracking of hot work tool steels, International Journal of Fatigue 23
(2004).
[15] Bohler. Bohler steels; 06-01-2009. Available from: http://www.bohlersteels.
co.uk/buag_frontend/download/grade/139/1/W302DE.pdf [cited: 12 03
2012.].
[16] LogBook/360. IOtech. Iotech; 07-01-2009. Available from: http://iotech.
bentech-taiwan.com/iotech/logbook_series.pdf [cited 12 03 2012.].
[17] Omega; 12-12-2008. Available from: http://www.omega.com/manuals/
manualpdf/M1238.pdf [cited 12 03 2012].
[18] Flir; 12-03-2012. Available from: http://www.flir.com/uploadedFiles/
Thermography_APAC/Products/Product_Literture/SC640_Datasheet%20APAC.
pdf.
[19] J. Hattel, Fundamentals of Numerical Modelling of Casting Processes, Nar-
ayana Press, Denmark, 2005, ISBN 978-87-502-0969-0.
[20] Long AD. Extending life of high pressure die casting dies. Ph.D. thesis (Faculty
of Engineering and Physical Sciences, Queen’s University, Belfast); July 2010.
11706.
Nomenclature
b: Mould material cooling capacity, W/s K
Cp: Specific heat capacity, J/kg  C
k: Thermal conductivity, W/m K
N: Number of cycles
r: Density, kg/m3
T: Temperature,  C
Ts: Surface temperature,  C
Ds: Stress fluctuation, MPa