1. GET THE WEAR OUT
Extend Asset Life
Steve Wilson, Cold Jet, Loveland, OH
Most would agree that the mold is the heart of the
molding process. From those that make them to those
who use them, we all want them to last for as long as
they were designed. A lot of work goes into striving
toward that goal: good design, proper metallurgy
selection, configuration, coatings, etc. With all that
having been taken into account, it’s not surprising
molders often want a warranty that the mold will last
the desired, if not quoted asset life.
So what’s the problem with offering a mold
warranty? Why don’t molds always last as long as they
were supposed to? Reality is, many facets of a molds
life can affect is longevity, from the design itself to
how well it is maintained.
How well the mold will be maintained is one of
the dilemmas for mold builder’s face when asked to
offer a warranty. Molders are the ones responsible to
properly care for and maintain the mold. How and how
often molders choose to maintain their molds can
affect mold life.
Today, high-dollar and often complex molds, are
run and maintained in varying degrees of skilled
molding shops and tool rooms. For some, mold
maintenance is an afterthought,rather than a priority.
A fairly recent industry poll conducted by the
American Mold Builders Association (AMBA), cited
that the #2 issue molders deal with is finding ways to
improve their “Operational Excellence” (lean
manufacturing, waste reduction, zero defects, higher
throughput,continuousimprovement, scrap reduction,
efficiency improvement, etc.).
Mold maintenance has a significant contribution
toward “Operational Excellence” and the life of the
mold. One of the keys to mold maintenance is mold
cleaning. According to Tooling Docs, Ashland, OH,
60-70% of the time spent on mold maintenance is
mold cleaning.
The most commonly used methods for cleaning
plastic injection molds today have been around for
years, and involve manual cleaning which can be
abrasive and contribute to mold wear.
Tooling Docs notes that there are four known
contributors to mold wear that can shorten the asset
life of a mold, see Fig. 1. Note “Cleaning Techniques”
in Fig. 1 below in the “Maintenance Practices”
quadrant. The method a molder chooses to clean the
mold you built can contribute to “Mold Wear”.
Fig. 1-Charting Mold Wear, ToolingDocs, Ashland,
OH
We all acknowledge that mold cleaning remains a
critical component of producing quality products.
Cleaning mold cavities and vents of resin off-gasses,
cured material and mold release agents can prevent a
variety of molding problems: burn, sticking parts,
short shots,plate-out, contamination, blemishes, flash
and actual mold damage.
Fig. 2 notes actual mold damage that can occur
when vents are not properly routinely cleaned. When

2. air is trapped inside the mold cavity, it begins to jet or
diesel. The result, mold damage.
Fig. 2- Mold Damage, Trapped Air (dirty vent)
So why don’t molders clean molds more
frequently? Mold cleaning is often postponed for a
variety of reasons: traditional manual methods cause
extended downtime because the process is
burdensome, the mold can be hot,it often involves the
use of chemicals containing VOC’s which employees
don’t like using and leave a residue that can cause
further cleaning issues, wipes, media which can be
abrasive, and brushes. Often, traditional manual
methods wear away critical mold tolerances of parting
lines, rolling over edges, shut-offs and vents. Over
time, traditional manual cleaning methods reduce the
asset life of the mold. With regard to traditional mold
cleaning, one molder in Cincinnati, OH noted, “To
clean it is to destroy it”, Tom Mendel of Performance
Plastics.
Fact, we have to clean our molds, but why do we
have to do it in a way that prematurely wears out the
mold? So what if you could clean yourmolds not only
more often, faster, cheaper but in a non-abrasive,
sustainable, environmentally responsible manner?
What if mold cleaning could be better managed
providing a win-win for everybody, facilitating a
platform for mold warranty programs. It’s a win for
the mold builder because clean molds run better,
providing more consistent parts and longer life.
There is something that can be done about the
common denominator to premature mold wear -
molders need to abandon old-schoolmethods of mold
cleaning.
The case for dry ice mold cleaning – a gateway to
easing the pain of offering a mold warranty. ANTEC
papers have been presented on various benefits to
cleaning molds with dry ice: faster, better, cleaner,
environmentally friendly, worker safety,etc., but let’s
focus on one of the relevant issue for mold builders:
non-abrasive mold cleaning and making it easier so
that they will do it routinely.
Many studies have been done and are being done
with regard to non-abrasive mold cleaning with dry
ice. One such study comes from Kettering University
(formerly GMI) in Flint, MI.
They recorded the following: Methodology and
Test Variables for a dry ice cleaning study:regarding
various metal mold substrates -“Materials were tested
with the Wand at an angle of 90 degrees and 45
degrees from the sample surface,each angle was tested
at 1” and 3” from the surface of the sample, and each
material was tested for 30 seconds and 60 seconds
durations for all variables. Materials were subjected to
both full sized pellets, shaved pellets,or shaved block.
Variables that remained constant are the following: air
pressure at 75 psi, dry ice flow rate at maximum level
(level 6).”
Using such cleaning methodology or techniques
and test variables, the metal samples would be
subjected to the worse case scenarios for dry ice
cleaning. Effective dry ice mold cleaning always
involves the fastest possible traverse rate while
maintaining an effective clean. Typical dry ice
cleaning does not involve stationary cleaning in one
spot as tested.
Quoting from the university study: “the picture
below shows a picture of the blasted and un-blasted
area of the die steel. Die steel after blasting for 60
seconds at a distance of 1 inch. Right side of the green
lines shows blasted area”, see Fig. 3:

3. Fig.3- Kettering University Study
The university testing concluded the following:
“The sample had no noticeable damage after blasting
with shape block ice at a distance of one inch, angle of
90 degrees and for either 30 second or 60 second
durations.”
Another independent study drew a similar
conclusion. The picture in Fig. 4 is the micrographic
examination of the metallurgical structure of
Martensitic Stainless Steel (440C) after dry ice
cleaning. Noted was the unaltered carbon particles at
the core and at the surface, before and after dry ice
cleaning.
Fig. 4 Core: Before/After Surface: Before/After
In addition, studies,such as “Dry Ice Blasting for
the Conservation Cleaning of Metals”, Rozemarijn
van der Molen, Ineke Joosten, Tonny Beentjes and
Luc Megens, have also shown that cleaning with dry
ice does not damage most industrial substrates.
Because the particles are relatively soft. The hardness
of dry pellets was found to be 1.5-2.0 Mohs,which is
soft compared to other forms of blast media (See Fig.
5, Krieg 2008). The dry ice particles have little
hardness and are therefore non-abrasive to any
substrate harderthan dry ice
Mohs Hardness Scale for Minerals
1 – Talc
2 – Gypsum, Dry Ice (Fingernail, Baking Soda ~
2.5)
3 – Calcite (a penny)
4 – Fluorite (Corn Cob ~ 4.5)
5 – Apatite (Glass Beads & Nut Shells ~ 5.5)
6 – Orthoclase, Feldspar, Spectrolite (Steel File ~
6.5-7.5)
7 – Quartz, Amethyst,Citrine, Agate (Garnet ~ 7.5)
8 – Topaz, Beryl, Emerald, Aquamarine
9 – Corundum, Ruby, Sapphire (Alum. Oxide ~ 8.5)
10- Diamond
Fig. 5 – Mohs Hardness Scale for Minerals
From a process viewpoint, one of the reasons dry
ice cleaning is so successful in non-abrasive mold
cleaning on more delicate mold substrates and
complex geometries, is the process flexibility. A very
effective way to control the Kinetic Energy
contribution to dry ice cleaning (KE = 1/2M x V2), is
changing particle size and pressure/velocity machine
setting.Cleaning is even gentler when 0.3 mm dry ice
MicroParticles are used in lieu of the traditional
3.0mm pellets. See Figs. 6 - 9.
Fig. 6- MicroParticles – thin, hard contaminates
(off- gasses)

4. Fig. 7- Traditional 3MM Pellets – thick, brittle
contaminates (screw cleaning)
Fig. 8 – 3.0 mm dry ice Fig. 9 – 0.3 mm shaved ice
Because of their smaller size, MicroParticles of
dry ice require less air to attain full acceleration for
effective cleaning. Their smaller mass and less air
bring a gentle cleaning to delicate substrate
metallurgies ormold features.The less air requirement
also equates in much quieteroperating dry ice cleaning
solutions.
Multiple studies have concluded that dry ice
cleaning is an effective, non-abrasive way to clean
injection molds without wearing away the asset life of
a mold
Many are also aware that dry ice cleaning has a
thermal effect while cleaning. It is very cold, -109
degrees F (-79.5 degrees C). This inherently low
temperature give the dry ice cleaning process a unique
thermodynamically induced surface mechanisms that
affect the adhesion of the contaminate to the substrate
(by embrittling it, causing it to shrink, creating rapid
micro-cracking and causing the bond between the
contaminant and the mold to fail). (co-efficient of
thermal expansion & contraction dissimilar materials)
This unique property of dry ice does aid in the removal
of the contaminant from the mold.
A common question from mold builders and
molders examining the dry ice cleaning process is,will
the thermal effect of the dry ice cleaning process
impart a thermal stress on the mold during cleaning?
It has been proven that this Thermal Effect of dry
ice cleaning does not damage the tool. First, this
thermal effect rapidly disappears on impact when the
dry ice strikes the contaminant. Upon impact, the dry
ice sublimates, another aspect of dry ice cleaning.
Volumetrically, dry ice will expand around 800 times
its size during sublimation, blowing the contaminant
off the mold from the inside out. At that moment, the
thermal property of dry ice is gone.
Secondly, Mark Krieg in his work, Analyse der
Effekte beim Trockeneisstrahlen (2008), showed that
the contribution of the Thermal Effect is minimal
toward the overall cleaning effect. On an ambient
mold substrate, the Thermal Effect is contributing
approximately 10% of the cleaning. This contribution
can be as high as 50% when blasting on an object at a
temperature of 500 degree C. Result, hotter molds
clean faster and easier.
A similar study was conducted by James Snide,
C02 Pellet Cleaning – A preliminary Evaluation,
Materials & Process Associates, Inc. (Oct. 12, 1992),
to measure any thermal stress during dry ice cleaning.
This study showed that the temperature decrease
occurs on the surface only, so that there is no chance
of thermal stress occurring to the substrate metal.
To illustrate this principle, an experiment was
performed where thermocouples were imbedded into
a steel substrate at varying depths (flush with the
surface to 2mm deep), see Fig. 10.

5. Fig. 10 Temperature Response of Thermocouples
placed at Various Depths in the Substrate
A dry ice particle stream was constantly swept
across the test specimen for 30 seconds (a relatively
long time for this process) and the thermocouple
recorded the changing temperatures at the various
depths. As shown in Fig. 11, the surface-mounted
thermocouple recorded a temperature drop each time
the particle stream passed directly upon it (50 degree
C in about 5 seconds). In contrast, the thermocouples
imbedded at various depths in the substrate recorded a
slow gradual drop in temperature corresponding to the
overall test plate temperature drop. The thermocouple
2mm deep only dropped 10 degree C after 30 seconds.
This curve illustrates that the Thermal Effect occurs
only at the surface where the contaminate is bonded to
the mold substrate and has no detrimental effect on the
mold.
Fig. 11 Temperature Response of Thermocouples
Placed At Various Depths in the Substrate
James Snide also conducted anotherstudy on the
thermal effect of dry ice cleaning in the rubber
industry,a thermoset application, where the molds are
hot. In this study the molders were operating at 149 +
degree C (300+ degree F) and were cleaned with –
78.3 degree C dry ice particles. He noted, “The
temperature differences between the hot mold and cold
dry ice will not cause cracking”.
He notes, there are two reasons for this
phenomenon. First, as seen in Fig. 11 above, the
temperature gradient occurs at the surface.Second, the
thermal stresses involved are much less than those
encountered during normal heat treat.
He went on to note in his study,the thermal stress
due to a temperature differential can be estimated
using the following equation where “sy” is stress (psi),
“DT” is temperature gradient (degree F), “a” is
coefficient of expansion and “g” is Poisson’s Ratio.
The corresponding parameter values are:
And the thermal stress (psi)is:
Where the temperature differential will be 135
degrees F / 57.2 degree C (Based on Fig. 10). He
noted, “This temperature gradient leads to a low
tensile stress of30,240 psi / 2085 bar. Even if the mold
temperature was brought down to the temperature of
the ice (an unrealistic extreme), the temperature
gradient would be -109F – 350F which gives 459 F /
2237.2 C, for which the corresponding tensile stress is
102,800 psi/ 7088 bar. This calculated stress is below
the yield point of steel in the hardened condition.
Again, these thermal stresses would be far less than
those encountered during normal heat treatment,
where the temperature differentials would exceed 500
degree F / 260 degree C.”
He concluded,“even at high impact velocities and
direct ‘head-on’ impact angles, the kinetic effect of
solid C02 particles is minimal when compared to other
media (grit, sand, PMB, etc.). This is due to the
relative lack ofhardness ofthe dry ice particles and the
almost instantaneous phase change to a gas on impact,
which effectively provides an almost nonexistent
coefficient of restitution in the impact equation.
Because dry ice blasting is considered non-abrasive
and relies on the thermal effect discussed above, the
process may be applied to a wide range of materials

6. without damage. Soft metals such as brass,beryllium,
and aluminum cladding can be dry ice cleaned for the
removal of coatings or contaminates without creating
surface stresses (pinging,pitting or roughness).”
Bottom line, an effect way to extend the asset life
of the tool, and to facilitate mold warranties, is to
better manage the mold cleaning techniques. Dry ice
cleaning is a vehicle to taking some of the heat of
offering mold warranties.
Some tool builders already prefer to have their
molds cleaned with the dry ice cleaning process,
especially those who manufacture extremely high
tolerance molds for the LSR molding segment, like
M.R. Mold and Roembke Manufacturing.
Others, who coat molds with DLC’s, like
Oerlikon Sulzer, also recommend the use of dry ice
cleaning for cleaning coated mold components for the
safe removal of off-gasses without causing damage.
They note in their presentations,“There is no wear or
damage to the tooling and all the spray turns to a gas.”
Here’s to molders being as intentional with their
mold maintenance practices as they have been with
their scientific molding practices. Let’s get the wear
out of mold cleaning and eliminate this hurdle to
offering mold warranties.