International Journal of Machine Tools & Manufacture ] (]]]]) ]]]–]]] Rapid tooling analysis of Stereolithography injection mould tooling Sadegh Rahmati , Phill Dickens 1 Manufacturing Engineering Group, Mechanical Engineering Department, University of Imam Hussain, Tehran, Iran Abstract Increasing competition in global markets is exerting intense pressure on companies to trim their product cycles continuously. As delivery times and costs of tools are on a downward trend, the modern tool manufacturer is under pressure to produce tools quickly, accurately and at a lower cost. Reducing the time to produce prototypes is a key to speed up the development of new products. Rapid tooling (RT) with particular regard to injection mould fabrication using rapid prototyping (RP) technology of Stereolithography (SL) may lead to savings in cost and time. In this paper, SL is used to directly build rapid injection mould tools for short run production. SL tools have been evaluated to analyse the maximum number of successful injections and quality of performance. SL epoxy tools were able to resist the injection pressure and temperature and 500 injections were achieved. The tool failure mechanisms during injection are investigated and tool failure either occurs due to excessive flexural stresses, or because of excessive shear stresses. r 2006 Elsevier Ltd. All rights reserved. Keywords: Rapid prototyping; Stereolithography; Rapid tooling; Injection moulding 1. Introduction The design to production time for new components continues to decrease so that the long lead-time in producing tooling conventionally becomes more of a barrier in responding to customer demand [1,2]. Increase in design capabilities, product variety, demand for shor- tened lead-time and decrease in production quantities are the major driving forces in the development of rapid tooling technologies, where tooling time and cost are significantly reduced [3–5]. At the same time, SL tooling techniques are improving and are becoming increasingly popular among manufacturers [6–8]. The development of the SL injection moulding tools at the University of Nottingham, has taken place along two fronts. The first was to provide material data for tool design under extreme conditions of stress and temperature, and obtaining data from different tests, which resemble real situations [9]. The second development was a theoretical and analytical analysis of SL tools during the injection process [10]. It has shown that SL injection mould tooling can be used successfully in low to medium numbers, and up to 500 parts have been produced with one tool [11]. The rest of the paper addresses the following: Section 2 outlines the experimental procedure, Section 3 focuses on the injection pressure analysis, Section 4 deals with tool temperatures studies and testing mechanical properties of the epoxy resin on tensile and impact strength, Section 5 concentrates on the failure mechanisms during injection by looking at flexural stresses, shear stresses, crack propaga- tion and fatigue using SEM observations, and Section 6 is the results’ summary. 2. Experimental method In constructing SL injection moulding tools, epoxy insert shells were fabricated directly from CAD data on an SL machine (SLA 250). These inserts were then fitted into steel mould bases through steel frames, and back-filled with an aluminium powder/aluminium chip/epoxy resin mixture (Fig. 1). The back-filled mixture added strength to the inserts and allowed heat to be conducted away from the mould. The modular steel mould bases were two standard base plates machined with a cylindrical pocket to fit the steel frames and the inserts [12]. The SL tools were then tested in a 50 ton Battenfeld production moulding machine ARTICLE IN PRESS www.elsevier.com/locate/ijmactool 0890-6955/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijmachtools.2006.09.022 Corresponding author. E-mail address: [email protected] (S. Rahmati). 1 Professor & Head of Rapid Manufacturing Research Group, Loughborough University, UK. Please cite this article as: S. Rahmati, P. Dickens, Rapid tooling analysis of Stereolithography injection mould tooling, International Journal of Machine Tools & Manufacture (2006), doi:10.1016/j.ijmachtools.2006.09.022
Transcript
ARTICLE IN PRESS
0890-6955$ - se
doi101016jijm
CorrespondE-mail addr
1Professor amp
Loughborough
Please cite th
Machine Too
International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]
wwwelseviercomlocateijmactool
Rapid tooling analysis of Stereolithography injection mould tooling
Sadegh Rahmati Phill Dickens1
Manufacturing Engineering Group Mechanical Engineering Department University of Imam Hussain Tehran Iran
Abstract
Increasing competition in global markets is exerting intense pressure on companies to trim their product cycles continuously As
delivery times and costs of tools are on a downward trend the modern tool manufacturer is under pressure to produce tools quickly
accurately and at a lower cost Reducing the time to produce prototypes is a key to speed up the development of new products Rapid
tooling (RT) with particular regard to injection mould fabrication using rapid prototyping (RP) technology of Stereolithography (SL)
may lead to savings in cost and time In this paper SL is used to directly build rapid injection mould tools for short run production SL
tools have been evaluated to analyse the maximum number of successful injections and quality of performance SL epoxy tools were able
to resist the injection pressure and temperature and 500 injections were achieved The tool failure mechanisms during injection are
investigated and tool failure either occurs due to excessive flexural stresses or because of excessive shear stresses
The design to production time for new componentscontinues to decrease so that the long lead-time inproducing tooling conventionally becomes more of abarrier in responding to customer demand [12] Increasein design capabilities product variety demand for shor-tened lead-time and decrease in production quantities arethe major driving forces in the development of rapidtooling technologies where tooling time and cost aresignificantly reduced [3ndash5] At the same time SL toolingtechniques are improving and are becoming increasinglypopular among manufacturers [6ndash8] The development ofthe SL injection moulding tools at the University ofNottingham has taken place along two fronts
The first was to provide material data for tool designunder extreme conditions of stress and temperature andobtaining data from different tests which resemble realsituations [9] The second development was a theoreticaland analytical analysis of SL tools during the injectionprocess [10] It has shown that SL injection mould tooling
e front matter r 2006 Elsevier Ltd All rights reserved
achtools200609022
ing author
ess rahmatirapidtoolpartcom (S Rahmati)
Head of Rapid Manufacturing Research Group
University UK
is article as S Rahmati P Dickens Rapid tooling analysis
ls amp Manufacture (2006) doi101016jijmachtools200609022
can be used successfully in low to medium numbers and upto 500 parts have been produced with one tool [11] Therest of the paper addresses the following Section 2 outlinesthe experimental procedure Section 3 focuses on theinjection pressure analysis Section 4 deals with tooltemperatures studies and testing mechanical properties ofthe epoxy resin on tensile and impact strength Section 5concentrates on the failure mechanisms during injection bylooking at flexural stresses shear stresses crack propaga-tion and fatigue using SEM observations and Section 6 isthe resultsrsquo summary
2 Experimental method
In constructing SL injection moulding tools epoxy insertshells were fabricated directly from CAD data on an SLmachine (SLA 250) These inserts were then fitted into steelmould bases through steel frames and back-filled with analuminium powderaluminium chipepoxy resin mixture(Fig 1) The back-filled mixture added strength to theinserts and allowed heat to be conducted away from themould The modular steel mould bases were two standardbase plates machined with a cylindrical pocket to fit thesteel frames and the inserts [12] The SL tools were thentested in a 50 ton Battenfeld production moulding machine
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Spring amp Gating system
Cavity Steel Frame
Core Steel Frame
Ejector Pins
StereolithographyCavity Insert
Back Fill material
StereolithographyCore Insect
Fig 1 Cross sectional view of the SL injection moulding tool inserts
Fig 2 The moulding after being ejected from the SL tool
SprueCorner ejectors
Middle ejectorSprue bush
Moulding
Ejector base
Load cell
Core side
Cavity side
Middle ejector
Load cell wires
Fig 3 The location of the ejectors and the load cells
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]2
to produce parts from polypropylene (PP) and Acryloni-trile Butadiene Styrene (ABS) (Fig 2)
During the moulding process the temperature andpressure of the cavity were monitored and the melttemperature was controlled using different thermocouplesto ensure that the conditions within the cavity were asuniform as possible Fractured samples of both the mouldsand the mouldings were examined using either an opticalmicroscope or a Scanning Electron Microscope (SEM)Fractured cubes were used to investigate the failure crosssections and fractured surfaces Failed cubes embeddedinto the moulding material were mounted using a castingmaterial and cut and polished in order to be examinedusing an optical microscope However fractured surfacesof the cubes and the core were investigated using SEMwhich led to the interpretation of the failure mechanism ofthe SL tooling
3 Injection pressure analysis
When the melt enters the cavity it moves radially awayfrom the centre and hits the blocks Then the flow movesin three directions upwards and around the cubes until thethree flow fronts meet at the back symmetrically The flowloses pressure and heat as it moves away from the centreand in addition to this pressure loss the flow movingupwards faces additional loss due to the bends There aretwo main forces acting on the blocks one due to the shearstress acting on the base the other is the bending stresstrying to tip over the blocks In general at any instantwhere the injection pressure is higher than the toolstrength failure is feasible To avoid this care is taken toinject at a temperature where the tool has sufficientstrength This criterion has led to a well-defined cyclewhere injection always takes place when the tool tempera-ture has dropped to 45 1C where the materialrsquos strength isable to resist the injection pressure
When a shot is made plastic is pushed into the cavityand as a result pressure is exerted on the tool The pressureprofile is investigated using load cells placed at the bottomof the ejectors (Fig 3) The pressure exerted on the ejectors
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
will be transferred to the load cells placed at the other endof the ejectors Three ejector pins out of five one in themiddle and two in the corners were selected for measuringthe pressure All load cells were wired to a Data Loggerand operated by a PC The voltage changes during
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 3
injection were recorded and converted into pressure Theresults are plotted in Fig 4 where the maximum injectionpressure of 1650 psi (114Mpa) on the middle ejectordrops to about 1300 psi (9MPa) on the corner ejectors
4 Temperature and material studies of SL epoxy tool
A number of thermocouples were inserted in differentlocations of the core and cavity to continuously monitorthe temperature in real time In particular thermocoupleswere inserted on features which are more vulnerable toheat such as cubes which were heated from five directionsA PC temperature logger was used to monitor and recordthe data and analyse the changes in temperature duringinjection Fig 5 is typical real time temperature behaviourwhere the cycle time is long enough to cool the mould to45 1C before starting a new injection
Standard specimens were built for tensile strength shearstrength according to ISO527 and impact strength testaccording to ISO 179 Tests were carried out using aminimum of 10 specimens tested at each temperaturewhere samples were heated by built-in heating chamberand tested using Instron 1195 machine at different
0
200
400
600
800
1000
1200
1400
1600
1800
0 3 6 9
Pressure Profile inside the SL
Pre
ssu
re (
psi
)
Tim
Fig 4 The pressure profile inside
0
20
40
60
80
100
120
0 100 200 300
Tem
per
atu
re (
Deg
C)
Time (se
Fig 5 Temperature variation inside the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
temperatures The average impact strength result wasdetermined to be 284 kJm2 and the average result atvarious temperatures is plotted in Fig 6 Epoxy tensile andshear test results are shown in Fig 7 Although themaximum tensile and shear strength is at 20 1C there isrelatively little elongation and elasticity at this temperaturewhich means that the impact resistance is minimum(Fig 6)Thermosetting materials such as epoxy possess a
relatively wide glass transition temperature With anincrease in epoxyrsquos tool temperature the tensile strengthof the tool decreases while its impact resistance increases(Figs 6 and 7) These properties work in favour of the SLtooling technique Due to the absence of water cooling andthe short freeze time of 15 s stress and warpage wereintroduced to the part in particular at hot spots where heatwas dissipated at a lower rate However increasing thefreeze time from 15 to 35 s has minimised the warpageIt is clear that there is an optimum mould temperature to
get an acceptable moulding without damaging the SL toolThis temperature is a trade off between the tensile andshear strength on one hand and toughness strength on theother hand For example injection at tool temperatures of
12 15 18 21 24
Corner Pressure 1Middle PressureCorner Pressure 3
Epoxy Cavity in Three Locations
e (sec)
the cavity at three locations
400 500 600
Core1Core2Core3Core4Cavity1Cavity2Cavity3Cavity4
c)
epoxy tool during successive cycles
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
20
30
40
50
60
70
80
90
100
10 20 30 40 50 60 70 80 90
Impa
ct S
tren
gth
(kJ
m2 )
Impact Strength
Temperature (Deg C)
Fig 6 Impact strength of filled epoxy versus temperature
0
10
20
30
40
50
60
70
10 20 30 40 50 60 70 80 90 100 110
Ten
sile
Str
eng
th (
MP
a)
0
10
20
30
40
50
60
70
Tensile (MPa)
Shear (MPa)
Temperature (degC)S
hea
r S
tren
gth
(M
Pa)
Fig 7 Maximum Tensile and Shear strength of epoxy SL5170 versus temperature
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]4
25 1C would probably cause failure because of the toolrsquoslack of toughness However at 45 1C the tool has survivedand more than 500 successful shots were made Using anair jet to cool the mould has proved to be successful andhas doubled the rate of production by reducing the cycletime from 43 to 2min
5 Failure mechanism analysis
When the plastic is injected into the cavity there is asudden pressure rise within the cavity which is the highestpressure reached during the moulding cycle (Fig 4) Thispressure exerts a force on the core features which maycause tool fracture if the ultimate tensile strength orflexural strength of the material is exceeded Fig 8 showsthe various scenarios which may arise during plasticinjection In (a) there is no failure in (b) there is a flexuralfailure and in (c) there is a shear failure Flexural stress canlead to instant failure or alternatively to crack propaga-tion and fatigue failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
The majority of failures observed during this investiga-tion were due to flexural stresses During flexural failurethe injection pressure overcomes the toolrsquos flexural strengthso that the feature rotates about its pivot point andultimately breaks off (Fig 8(b)) This may occur if theinjection pressure is beyond the flexural strength of the SLtool but flexural failure is usually due to the history of theloading The flexural stress for a cantilever beam with auniform force F acting on it is given in [13]
s frac146F h
at2 (1)
where h a and t are the cube height width and depthrespectively as shown in Fig 9 Table 1 shows thetheoretical calculations of flexural stresses for the SL cubesversus their flexural strength Using this equation it can beseen from Table 1 that only the largest cube should survivethe injection pressure and the rest of the cubes fail
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
cube moulding
(a) (b) (c)
Flow direction
Melt pressureMelt pressurepivoting point
Fig 8 Schematic view of different scenarios which may occur during injection (a) No failure (b) Flexural failure (c) Shear failure
Plastic flow
X
YZ
t Flexural stress in X direction
Neutral AxisY
Y
h
a
Fig 9 Schematic view of the cubersquos stress parameters and the
approaching flow
Table 1
Flexural stresses exerted on the SL cubes
Moment of
inertia I
(m4)
Moment M
(Nm)
Flexural
stress s(Mpa)
Flexural
strength at
40 1C (Mpa)
Cube 1 108 1012 1687 4685 650
Cube 2 625 1012 1687 6746 650
Cube 3 320 1012 1687 10541 650
Cube 4 135 1012 1687 18740 650
Fig 10 Moulding showing the attached plastic of crack before failure
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 5
However in practice the SL tools have produced hundredsof parts prior to failure so that the theoretical model in Eq(1) overestimates the flexural stresses There are tworeasons for this discrepancy First the flexural stressformula assumes a minimum beam aspect ratio of 10 whilethis ratio here is four [14] Secondly the injection pressureexerted on the cubes during injection was taken to be thepressure at front of the cubes but in reality this pressure ispartly counteracted by the melt pressure behind the cubes
In the case of the smallest cube the net pressure is foundto be 153 psi which gives a flexural stress of 15Mpa usingEq (1) which is less than 27Mpa (flexural strength of thetool at 40 1C) This suggests that all of the SL cubes shouldsurvive the injection pressure A better theoretical methodfor calculating the flexural stresses would be through the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
application of computational fluid dynamics (CFD) andfinite element method (FEM) which will combine the fluidand stress analysis to model the SL tool
52 Crack propagation and fatigue
Flexural stresses can also induce a lsquolsquofatiguersquorsquo typeprocess spanning a number of moulding cycles In thissituation the cube pivots as in Fig 8(b) without beingfractured but a crack is initiated at the intersection betweenthe face of the cube in tension due to flexural stresses andthe core face perpendicular to it During subsequent cyclesthe crack propagates through the base of the cubeeventually resulting in failure Failure analysis of theSEM images has revealed that the crack propagatesthrough the cubes prior to the ultimate failure Micro-scopic pictures of mouldings numbered sequentiallyindicate that the crack has started well before the ultimateflexural failure Fig 10 is a picture taken of the crosssection of a moulding before the actual failure happenedwhere subsequent injection mouldings have exhibiteda lsquolsquopositiversquorsquo flaw corresponding to the inverse of thecrack generated Fig 11 shows the flexural failureof a similar cube to that seen in Fig 10 after a numberof shotsCrack initiation in SL tools occurs predominately at
stress concentrations such as sharp corners or at stairsteppings (an inherent property of SL parts) Crack
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Fig 11 Flexural failure as a result of crack propagation
Fig 12 SEM observation revealing striation marking on the fractured
surface
Fig 13 SL cube being sheared off during injection moulding process
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]6
formation may also result from flaws or microscopicdefects created during photo-polymerisation process dueto material discontinuities [15] Sharp corners stairstepping voids or flaws are a cause or source of crackinitiation Fatigue failure can be minimised by introducingfillets at the sharp corners in order to reduce the stressconcentration and crack propagation Evidence of thecrack failure as shown in Fig 12 can be seen on thefracture surface in the form of lsquolsquostriationsrsquorsquo where each oneof these marks represents crack growth At the tip of thecrack and in a small region near the tip the yield strengthof the material is exceeded In this region plasticdeformation occurs and the stresses are limited by yielding[17] After each cycle the crack grows in the same manneruntil a critical crack length is reached At this point thecrack tip can increase in velocity and spread all the wayacross the cube resulting in failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
During shear failure the feature is sheared off in thedirection of the melt flow Fig 13 shows the cross sectionof a sheared SL cube Notice that the SL cube has beenpushed across by the flow of plastic The shear stress at apoint in a section is given by [18]
t frac14VQ
Ia (2)
where V is the shear force at the given section Q is the firstmoment of the area about the neutral axis I is the momentof inertia of the cube section with respect to the neutralaxis and a is the width of the cross-section As the shearstress calculation results show in Table 2 the maximumshear stresses produced in the SL tool during operation arebelow the shear strength of the SL tool Moreover the SLtool can survive at injection temperatures beyond 40 1C asshown in the last column of the Table 2 Fig 14 shows themaximum shear stresses at various points of the cube baseversus the average shear stress The plot of the maximumshear stresses at various points results in a parabolic curve
6 Conclusions
SL tools have been successfully tested where failureswere observed after 500 shots SL tool failure mechanismshave been investigated and different scenarios have beendemonstrated Using a thermoplastic with a meltingtemperature of 200ndash300 1C in epoxy SL tooling which hasa Glass transition temperature (Tg) of about 60ndash90 1Cseems unrealistic or impossible However the key point tothe success of this technique is the very low thermalconductivity of the SL tool and the short injection time(Fig 15) These two factors are the key to the success of theSL injection mould tooling which are overlooked bymany
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
[11] S Rahmati PM Dickens C Wykes Pressure effects in stereo-
lithography injection moulding tools Seventh European Conference
on Rapid Prototyping and Manufacturing Aachen Germany 1998
pp 471ndash480
[12] G Menges P Mohren How to make injection molds Hanser
Munich ISBN0-02-947570-8 1986
[13] GC Ives JA Mead MM Riley in RP Brown (Ed)Handbook
of Plastics Test Methods second ed London ISBN0-7114-5618-6
1981
[14] RA Douglas Introduction to Solid Mechanics Sir Isaac Pitman amp
Sons Ltd London 1989
[15] RW Hertzberg JA Manson Fatigue of Engineering Plastics
Academic New York 1980
[17] JW Dally FR William Experimental Stress Analysis 3rd ed
McGraw-Hill ISBN 0-07-015218-7 1991
[18] F Cheng Statistics and Strength of Materials 2nd ed McGraw-Hill
ISBN 0-07-115666-6 1997
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Spring amp Gating system
Cavity Steel Frame
Core Steel Frame
Ejector Pins
StereolithographyCavity Insert
Back Fill material
StereolithographyCore Insect
Fig 1 Cross sectional view of the SL injection moulding tool inserts
Fig 2 The moulding after being ejected from the SL tool
SprueCorner ejectors
Middle ejectorSprue bush
Moulding
Ejector base
Load cell
Core side
Cavity side
Middle ejector
Load cell wires
Fig 3 The location of the ejectors and the load cells
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]2
to produce parts from polypropylene (PP) and Acryloni-trile Butadiene Styrene (ABS) (Fig 2)
During the moulding process the temperature andpressure of the cavity were monitored and the melttemperature was controlled using different thermocouplesto ensure that the conditions within the cavity were asuniform as possible Fractured samples of both the mouldsand the mouldings were examined using either an opticalmicroscope or a Scanning Electron Microscope (SEM)Fractured cubes were used to investigate the failure crosssections and fractured surfaces Failed cubes embeddedinto the moulding material were mounted using a castingmaterial and cut and polished in order to be examinedusing an optical microscope However fractured surfacesof the cubes and the core were investigated using SEMwhich led to the interpretation of the failure mechanism ofthe SL tooling
3 Injection pressure analysis
When the melt enters the cavity it moves radially awayfrom the centre and hits the blocks Then the flow movesin three directions upwards and around the cubes until thethree flow fronts meet at the back symmetrically The flowloses pressure and heat as it moves away from the centreand in addition to this pressure loss the flow movingupwards faces additional loss due to the bends There aretwo main forces acting on the blocks one due to the shearstress acting on the base the other is the bending stresstrying to tip over the blocks In general at any instantwhere the injection pressure is higher than the toolstrength failure is feasible To avoid this care is taken toinject at a temperature where the tool has sufficientstrength This criterion has led to a well-defined cyclewhere injection always takes place when the tool tempera-ture has dropped to 45 1C where the materialrsquos strength isable to resist the injection pressure
When a shot is made plastic is pushed into the cavityand as a result pressure is exerted on the tool The pressureprofile is investigated using load cells placed at the bottomof the ejectors (Fig 3) The pressure exerted on the ejectors
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
will be transferred to the load cells placed at the other endof the ejectors Three ejector pins out of five one in themiddle and two in the corners were selected for measuringthe pressure All load cells were wired to a Data Loggerand operated by a PC The voltage changes during
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 3
injection were recorded and converted into pressure Theresults are plotted in Fig 4 where the maximum injectionpressure of 1650 psi (114Mpa) on the middle ejectordrops to about 1300 psi (9MPa) on the corner ejectors
4 Temperature and material studies of SL epoxy tool
A number of thermocouples were inserted in differentlocations of the core and cavity to continuously monitorthe temperature in real time In particular thermocoupleswere inserted on features which are more vulnerable toheat such as cubes which were heated from five directionsA PC temperature logger was used to monitor and recordthe data and analyse the changes in temperature duringinjection Fig 5 is typical real time temperature behaviourwhere the cycle time is long enough to cool the mould to45 1C before starting a new injection
Standard specimens were built for tensile strength shearstrength according to ISO527 and impact strength testaccording to ISO 179 Tests were carried out using aminimum of 10 specimens tested at each temperaturewhere samples were heated by built-in heating chamberand tested using Instron 1195 machine at different
0
200
400
600
800
1000
1200
1400
1600
1800
0 3 6 9
Pressure Profile inside the SL
Pre
ssu
re (
psi
)
Tim
Fig 4 The pressure profile inside
0
20
40
60
80
100
120
0 100 200 300
Tem
per
atu
re (
Deg
C)
Time (se
Fig 5 Temperature variation inside the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
temperatures The average impact strength result wasdetermined to be 284 kJm2 and the average result atvarious temperatures is plotted in Fig 6 Epoxy tensile andshear test results are shown in Fig 7 Although themaximum tensile and shear strength is at 20 1C there isrelatively little elongation and elasticity at this temperaturewhich means that the impact resistance is minimum(Fig 6)Thermosetting materials such as epoxy possess a
relatively wide glass transition temperature With anincrease in epoxyrsquos tool temperature the tensile strengthof the tool decreases while its impact resistance increases(Figs 6 and 7) These properties work in favour of the SLtooling technique Due to the absence of water cooling andthe short freeze time of 15 s stress and warpage wereintroduced to the part in particular at hot spots where heatwas dissipated at a lower rate However increasing thefreeze time from 15 to 35 s has minimised the warpageIt is clear that there is an optimum mould temperature to
get an acceptable moulding without damaging the SL toolThis temperature is a trade off between the tensile andshear strength on one hand and toughness strength on theother hand For example injection at tool temperatures of
12 15 18 21 24
Corner Pressure 1Middle PressureCorner Pressure 3
Epoxy Cavity in Three Locations
e (sec)
the cavity at three locations
400 500 600
Core1Core2Core3Core4Cavity1Cavity2Cavity3Cavity4
c)
epoxy tool during successive cycles
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
20
30
40
50
60
70
80
90
100
10 20 30 40 50 60 70 80 90
Impa
ct S
tren
gth
(kJ
m2 )
Impact Strength
Temperature (Deg C)
Fig 6 Impact strength of filled epoxy versus temperature
0
10
20
30
40
50
60
70
10 20 30 40 50 60 70 80 90 100 110
Ten
sile
Str
eng
th (
MP
a)
0
10
20
30
40
50
60
70
Tensile (MPa)
Shear (MPa)
Temperature (degC)S
hea
r S
tren
gth
(M
Pa)
Fig 7 Maximum Tensile and Shear strength of epoxy SL5170 versus temperature
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]4
25 1C would probably cause failure because of the toolrsquoslack of toughness However at 45 1C the tool has survivedand more than 500 successful shots were made Using anair jet to cool the mould has proved to be successful andhas doubled the rate of production by reducing the cycletime from 43 to 2min
5 Failure mechanism analysis
When the plastic is injected into the cavity there is asudden pressure rise within the cavity which is the highestpressure reached during the moulding cycle (Fig 4) Thispressure exerts a force on the core features which maycause tool fracture if the ultimate tensile strength orflexural strength of the material is exceeded Fig 8 showsthe various scenarios which may arise during plasticinjection In (a) there is no failure in (b) there is a flexuralfailure and in (c) there is a shear failure Flexural stress canlead to instant failure or alternatively to crack propaga-tion and fatigue failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
The majority of failures observed during this investiga-tion were due to flexural stresses During flexural failurethe injection pressure overcomes the toolrsquos flexural strengthso that the feature rotates about its pivot point andultimately breaks off (Fig 8(b)) This may occur if theinjection pressure is beyond the flexural strength of the SLtool but flexural failure is usually due to the history of theloading The flexural stress for a cantilever beam with auniform force F acting on it is given in [13]
s frac146F h
at2 (1)
where h a and t are the cube height width and depthrespectively as shown in Fig 9 Table 1 shows thetheoretical calculations of flexural stresses for the SL cubesversus their flexural strength Using this equation it can beseen from Table 1 that only the largest cube should survivethe injection pressure and the rest of the cubes fail
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
cube moulding
(a) (b) (c)
Flow direction
Melt pressureMelt pressurepivoting point
Fig 8 Schematic view of different scenarios which may occur during injection (a) No failure (b) Flexural failure (c) Shear failure
Plastic flow
X
YZ
t Flexural stress in X direction
Neutral AxisY
Y
h
a
Fig 9 Schematic view of the cubersquos stress parameters and the
approaching flow
Table 1
Flexural stresses exerted on the SL cubes
Moment of
inertia I
(m4)
Moment M
(Nm)
Flexural
stress s(Mpa)
Flexural
strength at
40 1C (Mpa)
Cube 1 108 1012 1687 4685 650
Cube 2 625 1012 1687 6746 650
Cube 3 320 1012 1687 10541 650
Cube 4 135 1012 1687 18740 650
Fig 10 Moulding showing the attached plastic of crack before failure
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 5
However in practice the SL tools have produced hundredsof parts prior to failure so that the theoretical model in Eq(1) overestimates the flexural stresses There are tworeasons for this discrepancy First the flexural stressformula assumes a minimum beam aspect ratio of 10 whilethis ratio here is four [14] Secondly the injection pressureexerted on the cubes during injection was taken to be thepressure at front of the cubes but in reality this pressure ispartly counteracted by the melt pressure behind the cubes
In the case of the smallest cube the net pressure is foundto be 153 psi which gives a flexural stress of 15Mpa usingEq (1) which is less than 27Mpa (flexural strength of thetool at 40 1C) This suggests that all of the SL cubes shouldsurvive the injection pressure A better theoretical methodfor calculating the flexural stresses would be through the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
application of computational fluid dynamics (CFD) andfinite element method (FEM) which will combine the fluidand stress analysis to model the SL tool
52 Crack propagation and fatigue
Flexural stresses can also induce a lsquolsquofatiguersquorsquo typeprocess spanning a number of moulding cycles In thissituation the cube pivots as in Fig 8(b) without beingfractured but a crack is initiated at the intersection betweenthe face of the cube in tension due to flexural stresses andthe core face perpendicular to it During subsequent cyclesthe crack propagates through the base of the cubeeventually resulting in failure Failure analysis of theSEM images has revealed that the crack propagatesthrough the cubes prior to the ultimate failure Micro-scopic pictures of mouldings numbered sequentiallyindicate that the crack has started well before the ultimateflexural failure Fig 10 is a picture taken of the crosssection of a moulding before the actual failure happenedwhere subsequent injection mouldings have exhibiteda lsquolsquopositiversquorsquo flaw corresponding to the inverse of thecrack generated Fig 11 shows the flexural failureof a similar cube to that seen in Fig 10 after a numberof shotsCrack initiation in SL tools occurs predominately at
stress concentrations such as sharp corners or at stairsteppings (an inherent property of SL parts) Crack
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Fig 11 Flexural failure as a result of crack propagation
Fig 12 SEM observation revealing striation marking on the fractured
surface
Fig 13 SL cube being sheared off during injection moulding process
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]6
formation may also result from flaws or microscopicdefects created during photo-polymerisation process dueto material discontinuities [15] Sharp corners stairstepping voids or flaws are a cause or source of crackinitiation Fatigue failure can be minimised by introducingfillets at the sharp corners in order to reduce the stressconcentration and crack propagation Evidence of thecrack failure as shown in Fig 12 can be seen on thefracture surface in the form of lsquolsquostriationsrsquorsquo where each oneof these marks represents crack growth At the tip of thecrack and in a small region near the tip the yield strengthof the material is exceeded In this region plasticdeformation occurs and the stresses are limited by yielding[17] After each cycle the crack grows in the same manneruntil a critical crack length is reached At this point thecrack tip can increase in velocity and spread all the wayacross the cube resulting in failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
During shear failure the feature is sheared off in thedirection of the melt flow Fig 13 shows the cross sectionof a sheared SL cube Notice that the SL cube has beenpushed across by the flow of plastic The shear stress at apoint in a section is given by [18]
t frac14VQ
Ia (2)
where V is the shear force at the given section Q is the firstmoment of the area about the neutral axis I is the momentof inertia of the cube section with respect to the neutralaxis and a is the width of the cross-section As the shearstress calculation results show in Table 2 the maximumshear stresses produced in the SL tool during operation arebelow the shear strength of the SL tool Moreover the SLtool can survive at injection temperatures beyond 40 1C asshown in the last column of the Table 2 Fig 14 shows themaximum shear stresses at various points of the cube baseversus the average shear stress The plot of the maximumshear stresses at various points results in a parabolic curve
6 Conclusions
SL tools have been successfully tested where failureswere observed after 500 shots SL tool failure mechanismshave been investigated and different scenarios have beendemonstrated Using a thermoplastic with a meltingtemperature of 200ndash300 1C in epoxy SL tooling which hasa Glass transition temperature (Tg) of about 60ndash90 1Cseems unrealistic or impossible However the key point tothe success of this technique is the very low thermalconductivity of the SL tool and the short injection time(Fig 15) These two factors are the key to the success of theSL injection mould tooling which are overlooked bymany
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
[11] S Rahmati PM Dickens C Wykes Pressure effects in stereo-
lithography injection moulding tools Seventh European Conference
on Rapid Prototyping and Manufacturing Aachen Germany 1998
pp 471ndash480
[12] G Menges P Mohren How to make injection molds Hanser
Munich ISBN0-02-947570-8 1986
[13] GC Ives JA Mead MM Riley in RP Brown (Ed)Handbook
of Plastics Test Methods second ed London ISBN0-7114-5618-6
1981
[14] RA Douglas Introduction to Solid Mechanics Sir Isaac Pitman amp
Sons Ltd London 1989
[15] RW Hertzberg JA Manson Fatigue of Engineering Plastics
Academic New York 1980
[17] JW Dally FR William Experimental Stress Analysis 3rd ed
McGraw-Hill ISBN 0-07-015218-7 1991
[18] F Cheng Statistics and Strength of Materials 2nd ed McGraw-Hill
ISBN 0-07-115666-6 1997
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 3
injection were recorded and converted into pressure Theresults are plotted in Fig 4 where the maximum injectionpressure of 1650 psi (114Mpa) on the middle ejectordrops to about 1300 psi (9MPa) on the corner ejectors
4 Temperature and material studies of SL epoxy tool
A number of thermocouples were inserted in differentlocations of the core and cavity to continuously monitorthe temperature in real time In particular thermocoupleswere inserted on features which are more vulnerable toheat such as cubes which were heated from five directionsA PC temperature logger was used to monitor and recordthe data and analyse the changes in temperature duringinjection Fig 5 is typical real time temperature behaviourwhere the cycle time is long enough to cool the mould to45 1C before starting a new injection
Standard specimens were built for tensile strength shearstrength according to ISO527 and impact strength testaccording to ISO 179 Tests were carried out using aminimum of 10 specimens tested at each temperaturewhere samples were heated by built-in heating chamberand tested using Instron 1195 machine at different
0
200
400
600
800
1000
1200
1400
1600
1800
0 3 6 9
Pressure Profile inside the SL
Pre
ssu
re (
psi
)
Tim
Fig 4 The pressure profile inside
0
20
40
60
80
100
120
0 100 200 300
Tem
per
atu
re (
Deg
C)
Time (se
Fig 5 Temperature variation inside the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
temperatures The average impact strength result wasdetermined to be 284 kJm2 and the average result atvarious temperatures is plotted in Fig 6 Epoxy tensile andshear test results are shown in Fig 7 Although themaximum tensile and shear strength is at 20 1C there isrelatively little elongation and elasticity at this temperaturewhich means that the impact resistance is minimum(Fig 6)Thermosetting materials such as epoxy possess a
relatively wide glass transition temperature With anincrease in epoxyrsquos tool temperature the tensile strengthof the tool decreases while its impact resistance increases(Figs 6 and 7) These properties work in favour of the SLtooling technique Due to the absence of water cooling andthe short freeze time of 15 s stress and warpage wereintroduced to the part in particular at hot spots where heatwas dissipated at a lower rate However increasing thefreeze time from 15 to 35 s has minimised the warpageIt is clear that there is an optimum mould temperature to
get an acceptable moulding without damaging the SL toolThis temperature is a trade off between the tensile andshear strength on one hand and toughness strength on theother hand For example injection at tool temperatures of
12 15 18 21 24
Corner Pressure 1Middle PressureCorner Pressure 3
Epoxy Cavity in Three Locations
e (sec)
the cavity at three locations
400 500 600
Core1Core2Core3Core4Cavity1Cavity2Cavity3Cavity4
c)
epoxy tool during successive cycles
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
20
30
40
50
60
70
80
90
100
10 20 30 40 50 60 70 80 90
Impa
ct S
tren
gth
(kJ
m2 )
Impact Strength
Temperature (Deg C)
Fig 6 Impact strength of filled epoxy versus temperature
0
10
20
30
40
50
60
70
10 20 30 40 50 60 70 80 90 100 110
Ten
sile
Str
eng
th (
MP
a)
0
10
20
30
40
50
60
70
Tensile (MPa)
Shear (MPa)
Temperature (degC)S
hea
r S
tren
gth
(M
Pa)
Fig 7 Maximum Tensile and Shear strength of epoxy SL5170 versus temperature
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]4
25 1C would probably cause failure because of the toolrsquoslack of toughness However at 45 1C the tool has survivedand more than 500 successful shots were made Using anair jet to cool the mould has proved to be successful andhas doubled the rate of production by reducing the cycletime from 43 to 2min
5 Failure mechanism analysis
When the plastic is injected into the cavity there is asudden pressure rise within the cavity which is the highestpressure reached during the moulding cycle (Fig 4) Thispressure exerts a force on the core features which maycause tool fracture if the ultimate tensile strength orflexural strength of the material is exceeded Fig 8 showsthe various scenarios which may arise during plasticinjection In (a) there is no failure in (b) there is a flexuralfailure and in (c) there is a shear failure Flexural stress canlead to instant failure or alternatively to crack propaga-tion and fatigue failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
The majority of failures observed during this investiga-tion were due to flexural stresses During flexural failurethe injection pressure overcomes the toolrsquos flexural strengthso that the feature rotates about its pivot point andultimately breaks off (Fig 8(b)) This may occur if theinjection pressure is beyond the flexural strength of the SLtool but flexural failure is usually due to the history of theloading The flexural stress for a cantilever beam with auniform force F acting on it is given in [13]
s frac146F h
at2 (1)
where h a and t are the cube height width and depthrespectively as shown in Fig 9 Table 1 shows thetheoretical calculations of flexural stresses for the SL cubesversus their flexural strength Using this equation it can beseen from Table 1 that only the largest cube should survivethe injection pressure and the rest of the cubes fail
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
cube moulding
(a) (b) (c)
Flow direction
Melt pressureMelt pressurepivoting point
Fig 8 Schematic view of different scenarios which may occur during injection (a) No failure (b) Flexural failure (c) Shear failure
Plastic flow
X
YZ
t Flexural stress in X direction
Neutral AxisY
Y
h
a
Fig 9 Schematic view of the cubersquos stress parameters and the
approaching flow
Table 1
Flexural stresses exerted on the SL cubes
Moment of
inertia I
(m4)
Moment M
(Nm)
Flexural
stress s(Mpa)
Flexural
strength at
40 1C (Mpa)
Cube 1 108 1012 1687 4685 650
Cube 2 625 1012 1687 6746 650
Cube 3 320 1012 1687 10541 650
Cube 4 135 1012 1687 18740 650
Fig 10 Moulding showing the attached plastic of crack before failure
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 5
However in practice the SL tools have produced hundredsof parts prior to failure so that the theoretical model in Eq(1) overestimates the flexural stresses There are tworeasons for this discrepancy First the flexural stressformula assumes a minimum beam aspect ratio of 10 whilethis ratio here is four [14] Secondly the injection pressureexerted on the cubes during injection was taken to be thepressure at front of the cubes but in reality this pressure ispartly counteracted by the melt pressure behind the cubes
In the case of the smallest cube the net pressure is foundto be 153 psi which gives a flexural stress of 15Mpa usingEq (1) which is less than 27Mpa (flexural strength of thetool at 40 1C) This suggests that all of the SL cubes shouldsurvive the injection pressure A better theoretical methodfor calculating the flexural stresses would be through the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
application of computational fluid dynamics (CFD) andfinite element method (FEM) which will combine the fluidand stress analysis to model the SL tool
52 Crack propagation and fatigue
Flexural stresses can also induce a lsquolsquofatiguersquorsquo typeprocess spanning a number of moulding cycles In thissituation the cube pivots as in Fig 8(b) without beingfractured but a crack is initiated at the intersection betweenthe face of the cube in tension due to flexural stresses andthe core face perpendicular to it During subsequent cyclesthe crack propagates through the base of the cubeeventually resulting in failure Failure analysis of theSEM images has revealed that the crack propagatesthrough the cubes prior to the ultimate failure Micro-scopic pictures of mouldings numbered sequentiallyindicate that the crack has started well before the ultimateflexural failure Fig 10 is a picture taken of the crosssection of a moulding before the actual failure happenedwhere subsequent injection mouldings have exhibiteda lsquolsquopositiversquorsquo flaw corresponding to the inverse of thecrack generated Fig 11 shows the flexural failureof a similar cube to that seen in Fig 10 after a numberof shotsCrack initiation in SL tools occurs predominately at
stress concentrations such as sharp corners or at stairsteppings (an inherent property of SL parts) Crack
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Fig 11 Flexural failure as a result of crack propagation
Fig 12 SEM observation revealing striation marking on the fractured
surface
Fig 13 SL cube being sheared off during injection moulding process
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]6
formation may also result from flaws or microscopicdefects created during photo-polymerisation process dueto material discontinuities [15] Sharp corners stairstepping voids or flaws are a cause or source of crackinitiation Fatigue failure can be minimised by introducingfillets at the sharp corners in order to reduce the stressconcentration and crack propagation Evidence of thecrack failure as shown in Fig 12 can be seen on thefracture surface in the form of lsquolsquostriationsrsquorsquo where each oneof these marks represents crack growth At the tip of thecrack and in a small region near the tip the yield strengthof the material is exceeded In this region plasticdeformation occurs and the stresses are limited by yielding[17] After each cycle the crack grows in the same manneruntil a critical crack length is reached At this point thecrack tip can increase in velocity and spread all the wayacross the cube resulting in failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
During shear failure the feature is sheared off in thedirection of the melt flow Fig 13 shows the cross sectionof a sheared SL cube Notice that the SL cube has beenpushed across by the flow of plastic The shear stress at apoint in a section is given by [18]
t frac14VQ
Ia (2)
where V is the shear force at the given section Q is the firstmoment of the area about the neutral axis I is the momentof inertia of the cube section with respect to the neutralaxis and a is the width of the cross-section As the shearstress calculation results show in Table 2 the maximumshear stresses produced in the SL tool during operation arebelow the shear strength of the SL tool Moreover the SLtool can survive at injection temperatures beyond 40 1C asshown in the last column of the Table 2 Fig 14 shows themaximum shear stresses at various points of the cube baseversus the average shear stress The plot of the maximumshear stresses at various points results in a parabolic curve
6 Conclusions
SL tools have been successfully tested where failureswere observed after 500 shots SL tool failure mechanismshave been investigated and different scenarios have beendemonstrated Using a thermoplastic with a meltingtemperature of 200ndash300 1C in epoxy SL tooling which hasa Glass transition temperature (Tg) of about 60ndash90 1Cseems unrealistic or impossible However the key point tothe success of this technique is the very low thermalconductivity of the SL tool and the short injection time(Fig 15) These two factors are the key to the success of theSL injection mould tooling which are overlooked bymany
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
[11] S Rahmati PM Dickens C Wykes Pressure effects in stereo-
lithography injection moulding tools Seventh European Conference
on Rapid Prototyping and Manufacturing Aachen Germany 1998
pp 471ndash480
[12] G Menges P Mohren How to make injection molds Hanser
Munich ISBN0-02-947570-8 1986
[13] GC Ives JA Mead MM Riley in RP Brown (Ed)Handbook
of Plastics Test Methods second ed London ISBN0-7114-5618-6
1981
[14] RA Douglas Introduction to Solid Mechanics Sir Isaac Pitman amp
Sons Ltd London 1989
[15] RW Hertzberg JA Manson Fatigue of Engineering Plastics
Academic New York 1980
[17] JW Dally FR William Experimental Stress Analysis 3rd ed
McGraw-Hill ISBN 0-07-015218-7 1991
[18] F Cheng Statistics and Strength of Materials 2nd ed McGraw-Hill
ISBN 0-07-115666-6 1997
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
20
30
40
50
60
70
80
90
100
10 20 30 40 50 60 70 80 90
Impa
ct S
tren
gth
(kJ
m2 )
Impact Strength
Temperature (Deg C)
Fig 6 Impact strength of filled epoxy versus temperature
0
10
20
30
40
50
60
70
10 20 30 40 50 60 70 80 90 100 110
Ten
sile
Str
eng
th (
MP
a)
0
10
20
30
40
50
60
70
Tensile (MPa)
Shear (MPa)
Temperature (degC)S
hea
r S
tren
gth
(M
Pa)
Fig 7 Maximum Tensile and Shear strength of epoxy SL5170 versus temperature
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]4
25 1C would probably cause failure because of the toolrsquoslack of toughness However at 45 1C the tool has survivedand more than 500 successful shots were made Using anair jet to cool the mould has proved to be successful andhas doubled the rate of production by reducing the cycletime from 43 to 2min
5 Failure mechanism analysis
When the plastic is injected into the cavity there is asudden pressure rise within the cavity which is the highestpressure reached during the moulding cycle (Fig 4) Thispressure exerts a force on the core features which maycause tool fracture if the ultimate tensile strength orflexural strength of the material is exceeded Fig 8 showsthe various scenarios which may arise during plasticinjection In (a) there is no failure in (b) there is a flexuralfailure and in (c) there is a shear failure Flexural stress canlead to instant failure or alternatively to crack propaga-tion and fatigue failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
The majority of failures observed during this investiga-tion were due to flexural stresses During flexural failurethe injection pressure overcomes the toolrsquos flexural strengthso that the feature rotates about its pivot point andultimately breaks off (Fig 8(b)) This may occur if theinjection pressure is beyond the flexural strength of the SLtool but flexural failure is usually due to the history of theloading The flexural stress for a cantilever beam with auniform force F acting on it is given in [13]
s frac146F h
at2 (1)
where h a and t are the cube height width and depthrespectively as shown in Fig 9 Table 1 shows thetheoretical calculations of flexural stresses for the SL cubesversus their flexural strength Using this equation it can beseen from Table 1 that only the largest cube should survivethe injection pressure and the rest of the cubes fail
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
cube moulding
(a) (b) (c)
Flow direction
Melt pressureMelt pressurepivoting point
Fig 8 Schematic view of different scenarios which may occur during injection (a) No failure (b) Flexural failure (c) Shear failure
Plastic flow
X
YZ
t Flexural stress in X direction
Neutral AxisY
Y
h
a
Fig 9 Schematic view of the cubersquos stress parameters and the
approaching flow
Table 1
Flexural stresses exerted on the SL cubes
Moment of
inertia I
(m4)
Moment M
(Nm)
Flexural
stress s(Mpa)
Flexural
strength at
40 1C (Mpa)
Cube 1 108 1012 1687 4685 650
Cube 2 625 1012 1687 6746 650
Cube 3 320 1012 1687 10541 650
Cube 4 135 1012 1687 18740 650
Fig 10 Moulding showing the attached plastic of crack before failure
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 5
However in practice the SL tools have produced hundredsof parts prior to failure so that the theoretical model in Eq(1) overestimates the flexural stresses There are tworeasons for this discrepancy First the flexural stressformula assumes a minimum beam aspect ratio of 10 whilethis ratio here is four [14] Secondly the injection pressureexerted on the cubes during injection was taken to be thepressure at front of the cubes but in reality this pressure ispartly counteracted by the melt pressure behind the cubes
In the case of the smallest cube the net pressure is foundto be 153 psi which gives a flexural stress of 15Mpa usingEq (1) which is less than 27Mpa (flexural strength of thetool at 40 1C) This suggests that all of the SL cubes shouldsurvive the injection pressure A better theoretical methodfor calculating the flexural stresses would be through the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
application of computational fluid dynamics (CFD) andfinite element method (FEM) which will combine the fluidand stress analysis to model the SL tool
52 Crack propagation and fatigue
Flexural stresses can also induce a lsquolsquofatiguersquorsquo typeprocess spanning a number of moulding cycles In thissituation the cube pivots as in Fig 8(b) without beingfractured but a crack is initiated at the intersection betweenthe face of the cube in tension due to flexural stresses andthe core face perpendicular to it During subsequent cyclesthe crack propagates through the base of the cubeeventually resulting in failure Failure analysis of theSEM images has revealed that the crack propagatesthrough the cubes prior to the ultimate failure Micro-scopic pictures of mouldings numbered sequentiallyindicate that the crack has started well before the ultimateflexural failure Fig 10 is a picture taken of the crosssection of a moulding before the actual failure happenedwhere subsequent injection mouldings have exhibiteda lsquolsquopositiversquorsquo flaw corresponding to the inverse of thecrack generated Fig 11 shows the flexural failureof a similar cube to that seen in Fig 10 after a numberof shotsCrack initiation in SL tools occurs predominately at
stress concentrations such as sharp corners or at stairsteppings (an inherent property of SL parts) Crack
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Fig 11 Flexural failure as a result of crack propagation
Fig 12 SEM observation revealing striation marking on the fractured
surface
Fig 13 SL cube being sheared off during injection moulding process
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]6
formation may also result from flaws or microscopicdefects created during photo-polymerisation process dueto material discontinuities [15] Sharp corners stairstepping voids or flaws are a cause or source of crackinitiation Fatigue failure can be minimised by introducingfillets at the sharp corners in order to reduce the stressconcentration and crack propagation Evidence of thecrack failure as shown in Fig 12 can be seen on thefracture surface in the form of lsquolsquostriationsrsquorsquo where each oneof these marks represents crack growth At the tip of thecrack and in a small region near the tip the yield strengthof the material is exceeded In this region plasticdeformation occurs and the stresses are limited by yielding[17] After each cycle the crack grows in the same manneruntil a critical crack length is reached At this point thecrack tip can increase in velocity and spread all the wayacross the cube resulting in failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
During shear failure the feature is sheared off in thedirection of the melt flow Fig 13 shows the cross sectionof a sheared SL cube Notice that the SL cube has beenpushed across by the flow of plastic The shear stress at apoint in a section is given by [18]
t frac14VQ
Ia (2)
where V is the shear force at the given section Q is the firstmoment of the area about the neutral axis I is the momentof inertia of the cube section with respect to the neutralaxis and a is the width of the cross-section As the shearstress calculation results show in Table 2 the maximumshear stresses produced in the SL tool during operation arebelow the shear strength of the SL tool Moreover the SLtool can survive at injection temperatures beyond 40 1C asshown in the last column of the Table 2 Fig 14 shows themaximum shear stresses at various points of the cube baseversus the average shear stress The plot of the maximumshear stresses at various points results in a parabolic curve
6 Conclusions
SL tools have been successfully tested where failureswere observed after 500 shots SL tool failure mechanismshave been investigated and different scenarios have beendemonstrated Using a thermoplastic with a meltingtemperature of 200ndash300 1C in epoxy SL tooling which hasa Glass transition temperature (Tg) of about 60ndash90 1Cseems unrealistic or impossible However the key point tothe success of this technique is the very low thermalconductivity of the SL tool and the short injection time(Fig 15) These two factors are the key to the success of theSL injection mould tooling which are overlooked bymany
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
[11] S Rahmati PM Dickens C Wykes Pressure effects in stereo-
lithography injection moulding tools Seventh European Conference
on Rapid Prototyping and Manufacturing Aachen Germany 1998
pp 471ndash480
[12] G Menges P Mohren How to make injection molds Hanser
Munich ISBN0-02-947570-8 1986
[13] GC Ives JA Mead MM Riley in RP Brown (Ed)Handbook
of Plastics Test Methods second ed London ISBN0-7114-5618-6
1981
[14] RA Douglas Introduction to Solid Mechanics Sir Isaac Pitman amp
Sons Ltd London 1989
[15] RW Hertzberg JA Manson Fatigue of Engineering Plastics
Academic New York 1980
[17] JW Dally FR William Experimental Stress Analysis 3rd ed
McGraw-Hill ISBN 0-07-015218-7 1991
[18] F Cheng Statistics and Strength of Materials 2nd ed McGraw-Hill
ISBN 0-07-115666-6 1997
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
cube moulding
(a) (b) (c)
Flow direction
Melt pressureMelt pressurepivoting point
Fig 8 Schematic view of different scenarios which may occur during injection (a) No failure (b) Flexural failure (c) Shear failure
Plastic flow
X
YZ
t Flexural stress in X direction
Neutral AxisY
Y
h
a
Fig 9 Schematic view of the cubersquos stress parameters and the
approaching flow
Table 1
Flexural stresses exerted on the SL cubes
Moment of
inertia I
(m4)
Moment M
(Nm)
Flexural
stress s(Mpa)
Flexural
strength at
40 1C (Mpa)
Cube 1 108 1012 1687 4685 650
Cube 2 625 1012 1687 6746 650
Cube 3 320 1012 1687 10541 650
Cube 4 135 1012 1687 18740 650
Fig 10 Moulding showing the attached plastic of crack before failure
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 5
However in practice the SL tools have produced hundredsof parts prior to failure so that the theoretical model in Eq(1) overestimates the flexural stresses There are tworeasons for this discrepancy First the flexural stressformula assumes a minimum beam aspect ratio of 10 whilethis ratio here is four [14] Secondly the injection pressureexerted on the cubes during injection was taken to be thepressure at front of the cubes but in reality this pressure ispartly counteracted by the melt pressure behind the cubes
In the case of the smallest cube the net pressure is foundto be 153 psi which gives a flexural stress of 15Mpa usingEq (1) which is less than 27Mpa (flexural strength of thetool at 40 1C) This suggests that all of the SL cubes shouldsurvive the injection pressure A better theoretical methodfor calculating the flexural stresses would be through the
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
application of computational fluid dynamics (CFD) andfinite element method (FEM) which will combine the fluidand stress analysis to model the SL tool
52 Crack propagation and fatigue
Flexural stresses can also induce a lsquolsquofatiguersquorsquo typeprocess spanning a number of moulding cycles In thissituation the cube pivots as in Fig 8(b) without beingfractured but a crack is initiated at the intersection betweenthe face of the cube in tension due to flexural stresses andthe core face perpendicular to it During subsequent cyclesthe crack propagates through the base of the cubeeventually resulting in failure Failure analysis of theSEM images has revealed that the crack propagatesthrough the cubes prior to the ultimate failure Micro-scopic pictures of mouldings numbered sequentiallyindicate that the crack has started well before the ultimateflexural failure Fig 10 is a picture taken of the crosssection of a moulding before the actual failure happenedwhere subsequent injection mouldings have exhibiteda lsquolsquopositiversquorsquo flaw corresponding to the inverse of thecrack generated Fig 11 shows the flexural failureof a similar cube to that seen in Fig 10 after a numberof shotsCrack initiation in SL tools occurs predominately at
stress concentrations such as sharp corners or at stairsteppings (an inherent property of SL parts) Crack
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Fig 11 Flexural failure as a result of crack propagation
Fig 12 SEM observation revealing striation marking on the fractured
surface
Fig 13 SL cube being sheared off during injection moulding process
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]6
formation may also result from flaws or microscopicdefects created during photo-polymerisation process dueto material discontinuities [15] Sharp corners stairstepping voids or flaws are a cause or source of crackinitiation Fatigue failure can be minimised by introducingfillets at the sharp corners in order to reduce the stressconcentration and crack propagation Evidence of thecrack failure as shown in Fig 12 can be seen on thefracture surface in the form of lsquolsquostriationsrsquorsquo where each oneof these marks represents crack growth At the tip of thecrack and in a small region near the tip the yield strengthof the material is exceeded In this region plasticdeformation occurs and the stresses are limited by yielding[17] After each cycle the crack grows in the same manneruntil a critical crack length is reached At this point thecrack tip can increase in velocity and spread all the wayacross the cube resulting in failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
During shear failure the feature is sheared off in thedirection of the melt flow Fig 13 shows the cross sectionof a sheared SL cube Notice that the SL cube has beenpushed across by the flow of plastic The shear stress at apoint in a section is given by [18]
t frac14VQ
Ia (2)
where V is the shear force at the given section Q is the firstmoment of the area about the neutral axis I is the momentof inertia of the cube section with respect to the neutralaxis and a is the width of the cross-section As the shearstress calculation results show in Table 2 the maximumshear stresses produced in the SL tool during operation arebelow the shear strength of the SL tool Moreover the SLtool can survive at injection temperatures beyond 40 1C asshown in the last column of the Table 2 Fig 14 shows themaximum shear stresses at various points of the cube baseversus the average shear stress The plot of the maximumshear stresses at various points results in a parabolic curve
6 Conclusions
SL tools have been successfully tested where failureswere observed after 500 shots SL tool failure mechanismshave been investigated and different scenarios have beendemonstrated Using a thermoplastic with a meltingtemperature of 200ndash300 1C in epoxy SL tooling which hasa Glass transition temperature (Tg) of about 60ndash90 1Cseems unrealistic or impossible However the key point tothe success of this technique is the very low thermalconductivity of the SL tool and the short injection time(Fig 15) These two factors are the key to the success of theSL injection mould tooling which are overlooked bymany
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
[11] S Rahmati PM Dickens C Wykes Pressure effects in stereo-
lithography injection moulding tools Seventh European Conference
on Rapid Prototyping and Manufacturing Aachen Germany 1998
pp 471ndash480
[12] G Menges P Mohren How to make injection molds Hanser
Munich ISBN0-02-947570-8 1986
[13] GC Ives JA Mead MM Riley in RP Brown (Ed)Handbook
of Plastics Test Methods second ed London ISBN0-7114-5618-6
1981
[14] RA Douglas Introduction to Solid Mechanics Sir Isaac Pitman amp
Sons Ltd London 1989
[15] RW Hertzberg JA Manson Fatigue of Engineering Plastics
Academic New York 1980
[17] JW Dally FR William Experimental Stress Analysis 3rd ed
McGraw-Hill ISBN 0-07-015218-7 1991
[18] F Cheng Statistics and Strength of Materials 2nd ed McGraw-Hill
ISBN 0-07-115666-6 1997
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Fig 11 Flexural failure as a result of crack propagation
Fig 12 SEM observation revealing striation marking on the fractured
surface
Fig 13 SL cube being sheared off during injection moulding process
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]6
formation may also result from flaws or microscopicdefects created during photo-polymerisation process dueto material discontinuities [15] Sharp corners stairstepping voids or flaws are a cause or source of crackinitiation Fatigue failure can be minimised by introducingfillets at the sharp corners in order to reduce the stressconcentration and crack propagation Evidence of thecrack failure as shown in Fig 12 can be seen on thefracture surface in the form of lsquolsquostriationsrsquorsquo where each oneof these marks represents crack growth At the tip of thecrack and in a small region near the tip the yield strengthof the material is exceeded In this region plasticdeformation occurs and the stresses are limited by yielding[17] After each cycle the crack grows in the same manneruntil a critical crack length is reached At this point thecrack tip can increase in velocity and spread all the wayacross the cube resulting in failure
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
During shear failure the feature is sheared off in thedirection of the melt flow Fig 13 shows the cross sectionof a sheared SL cube Notice that the SL cube has beenpushed across by the flow of plastic The shear stress at apoint in a section is given by [18]
t frac14VQ
Ia (2)
where V is the shear force at the given section Q is the firstmoment of the area about the neutral axis I is the momentof inertia of the cube section with respect to the neutralaxis and a is the width of the cross-section As the shearstress calculation results show in Table 2 the maximumshear stresses produced in the SL tool during operation arebelow the shear strength of the SL tool Moreover the SLtool can survive at injection temperatures beyond 40 1C asshown in the last column of the Table 2 Fig 14 shows themaximum shear stresses at various points of the cube baseversus the average shear stress The plot of the maximumshear stresses at various points results in a parabolic curve
6 Conclusions
SL tools have been successfully tested where failureswere observed after 500 shots SL tool failure mechanismshave been investigated and different scenarios have beendemonstrated Using a thermoplastic with a meltingtemperature of 200ndash300 1C in epoxy SL tooling which hasa Glass transition temperature (Tg) of about 60ndash90 1Cseems unrealistic or impossible However the key point tothe success of this technique is the very low thermalconductivity of the SL tool and the short injection time(Fig 15) These two factors are the key to the success of theSL injection mould tooling which are overlooked bymany
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
[11] S Rahmati PM Dickens C Wykes Pressure effects in stereo-
lithography injection moulding tools Seventh European Conference
on Rapid Prototyping and Manufacturing Aachen Germany 1998
pp 471ndash480
[12] G Menges P Mohren How to make injection molds Hanser
Munich ISBN0-02-947570-8 1986
[13] GC Ives JA Mead MM Riley in RP Brown (Ed)Handbook
of Plastics Test Methods second ed London ISBN0-7114-5618-6
1981
[14] RA Douglas Introduction to Solid Mechanics Sir Isaac Pitman amp
Sons Ltd London 1989
[15] RW Hertzberg JA Manson Fatigue of Engineering Plastics
Academic New York 1980
[17] JW Dally FR William Experimental Stress Analysis 3rd ed
McGraw-Hill ISBN 0-07-015218-7 1991
[18] F Cheng Statistics and Strength of Materials 2nd ed McGraw-Hill
ISBN 0-07-115666-6 1997
of Stereolithography injection mould tooling International Journal of
ARTICLE IN PRESS
Table 2
Shear stresses acting on the SL cubes
Shear area AS (mm2) Shear force V (N) Shear stress tave (Mpa) Shear strength at 40 1C (Mpa) TMAX (1C)
Cube 1 36 42164 1171 243 653
Cube 2 30 42164 1405 243 615
Cube 3 24 42164 1757 243 559
Cube 4 18 42164 2342 243 464
14
NA
12
shear stress at 14 fron NA
shear stress at NA
average shear stress1171 MPa
1318 MPa
1757 MPa
1318 MPa
0
0
Fig 14 Distribution of the shear stresses across the largest cube base
0
200
400
600
800
1000
1200
1400
1600
1800
0 10 20 30 40 50 60 70 80 90 100
Pre
ssu
re (
psi
)
010
20
30
4050
607080
90
100
110120
Tem
per
atu
re (
Deg
C)Pressure
Temperature
Time (sec)
Fig 15 Plot of temperature and pressure versus time during injection cycle
S Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]] 7
Although epoxy has a very low tensile or shear strengthat high temperatures during the first few seconds ofinjection in which the maximum pressure is exerted on thetool the heat has not been able to penetrate Therefore thetool strength is still maintained and low conductivity of theepoxy works in favour of the process initially It can beconcluded that the tool must be cooled down in eachcycle to as low as 40ndash50 1C before the next injection ismade Tool cooling can be achieved either through freeconvection which takes 4ndash5min or through forcedconvection by means of an air jet which reduces the cycletime to 1 2min The results of the work can be summarisedas follows
Please cite this article as S Rahmati P Dickens Rapid tooling analysis
More than 500 parts were produced using the epoxy SLcore and cavity using external air jet to cool the tool to45 1C
Tool failure during injection is independent of the
plastic temperature
Failure during injection may occur either at low tool
temperature when tool toughness is not sufficient or athigh tool temperature (above epoxy Tg)
As experience and theoretical calculations confirm
flexural stresses during the injection process are themost probable cause of failure Reducing the featuresaspect ratio of tool decreases the chances of flexuralfailure
tereolithography injection mould tooling International Journal of
ARTICLE IN PRESSS Rahmati P Dickens International Journal of Machine Tools amp Manufacture ] (]]]]) ]]]ndash]]]8
Shear stress failure during injection is less likely thanflexural failure in particular when the SL tool is warmedto over 40 1C prior to injection
References
[1] D Chen F Cheng Integration of product and process development
using rapid prototyping and work cell simulation technology Journal
of Industrial Technology 16 (1) (2000)
[2] JA McDonald CJ Ryall DH Wimpenny Rapid Prototyping
Casebook Professional Engineering Publishing UK 2001
[3] MA Evans RI Campbell A comparative evaluation of industrial
design models produced using rapid prototyping and workshop-
based fabrication techniques Rapid Prototyping Journal 9 (5) (2003)
[4] A Venus S Crommert Manufacturing of Injection Molds with SLS
Rapid Tooling Rapid Prototyping vol 2 (2) Dearborn USA 1996
[5] Y Li M Keefe EP Gargiulo Studies in Direct Tooling by
Stereolithography Sixth European Conference on Rapid Prototyping
and Manufacturing Nottingham UK July 1997 ISBN0-9519759-7-
8 pp 253ndash266
[6] P Decelles M Barritt Direct AIM Prototype Tooling 3D Systems
1996 PN 7027511-25-96
[7] T Greaves (Delphi-GM) Case study using stereolithography to
