Selective Laser Sintering State-of-the-art

Laoui* * Catholic University of Leuven, Celestijnenlaan 300B, B-3001 Leuven, Belgium Abstract Selective Laser Sintering SLS is a rapid prototyping process that allows to progressively pr

Selective Laser Sintering: State-of-the-art

J.P Kruth*, T Laoui*

* Catholic University of Leuven, Celestijnenlaan 300B, B-3001 Leuven, Belgium

Abstract

Selective Laser Sintering (SLS) is a rapid prototyping process that allows to progressively produce (build) complex geometrical parts by fusing and solidifying successive layers of fine powder material deposited on top

of each other Fusion is induced by heat delivered by a laser beam that melts or sinters the powder grains together and to the previous solidified layer

The paper describes the state-of-the-art of SLS one decade after its first appearance (First SLS machine sold

in 1990) It deals with progresses in terms of SLS machines (hardware, optics, lasers,…), SLS processible materials (polymers, metals, ceramics, foundry sand, composites, cermets,…) and applications (prototypes, tooling, moulds and dies, functional parts,…)

Key words: Selective laser sintering, polymer, metal powders, cermets, simulation

1 INTRODUCTION

Selective Laser Sintering (SLS) is a Rapid

Prototyping (RP) process that fuses or sinters

powder particles together to generate solid parts The

fusing or sintering energy is provided by the heat of

a laser beam that scans across successive layers of

pre-deposited powder (Fig 1) The ‘green’ product

obtained after laser sintering might still be porous

Since this porosity is normally not desired (e.g

limited strength), a post-treatment is often applied

that mostly involves infiltration of the pores with a

polymer (e.g epoxy) or a metal (e.g Cu or bronze)

1),2) Alternatively, a furnace post-sintering or HIP

cycle may be applied to reach full density 3)

Fig 1: Schematic overview of SLS process

SLS might be considered the RP technology with

the highest potential, mainly because it can be

applied to produce parts in almost any (powder)

material: polymers, reinforced polymers, metals,

ceramics, composites, foundry sand, etc The process

may be used to sinter many off-the-shelf powder

materials not specially developed for SLS, even

though the two well-known vendors of SLS

equipment (DTM and EOS) offer special purpose

well flowing SLS powder materials Those powders

depict a fast sintering reaction tuned to the short laser/powder interaction time that results from the fast scanning of the laser beam across the powder layers The use of off-the-shelf powders adds a second major advantage over other rapid prototyping (RP) processes, like stereo-lithography (SL), laminated object manufacturing (LOM) or fused deposition modelling (FDM), that respectively require dedicated photo-polymers tuned to a specific light wavelength (SL), pre-shaped material foils generally provided with glue on one side (LOM) and calibrated polymer wire coils or sheet cartridges (FDM) 56),57) Generally, all powder materials may be processed on the same SLS machine, even though one of the two SLS vendors opted for marketing dedicated machines for the various materials: EOSint-P machine for polymers, EOSint-M for metals, EOSint-S for sands

Notwithstanding these potentials and the fact that SLS was, after stereo-lithography, one of the first RP processes (originally developed at University of Texas-Austin4)), SLS has till now been characterised

by a relatively small market penetration: see Table 1 However, in recent years it depicted one of the biggest growth (Table 1), that may be partly associated with the fact that RP applications are no longer limited to the production of visual polymer prototypes, but extend more and more to the production of functional prototypes, functional parts and rapid tools made from polymer, metal or sand This paper gives a survey of the state-of-the-art

of SLS, one decade after the first industrial applications emerged

2 SLS EQUIPMENT 2.1 Principle

A basic SLS machine is depicted in Fig 1 The SLS part is produced by depositing successive thin

l o o s e p o w d e r

s i n t e r e d p a t t e r n

l a s e r

e n e r g y b e a m

f o c u s e d a n d

d e f l e c t e d b e a m

C A D

p o w d e r d e p o s i t i o n

s y s t e m

b e a m d e f l e c t i o n s y s t e m

c o n t a i n e r

layers of powders (typically 0.1 to 0.3mm thickness)

in a building container Each layer of powder is first

sintered in accordance with the product’s geometry

coinciding with that layer or section of the product,

before a new layer of powder is deposited and

sintered

Table 1 : RP systems unit sales

Laminated Object Mfg

Selective Laser Sintering

2.2 Powder deposition

Depositing the powder layers is a critical issue in

the SLS process Several studies 5),6) were devoted to

develop suited powder deposition mechanisms to

deposit smooth thin layers of powder even if the

powder morphology does not yield good powder

flowability (e.g in case of irregular or very fine

powder particles that tend to coagulate rather than to

spread out regularly) Those mechanisms often use a

combination of specially shaped feeding slots,

free-rolling or counter-free-rolling cylinders, scrapers,

vibrators and other devices to deposit and pack or

densify powder in ever-thinner layers Ideally the

powder should reach the tap density when deposited

This is the highest achievable density of the powder

obtained by vibrating or tapping the powder without

enforcing any plastic deformation of powder grains

The best packing density is 74% for monosized

spherical powders Achieving this density or even

higher (using a careful selection of a particle size

distribution) may be favoured with vibrators or

pressing devices Some studies tried to levitate and

deposit thin powder layers with electrostatic charged

plates similar to those used to deposit toner in

photocopying devices 7),8) The development of those

deposition systems run in parallel with the

development of better flowing powders (e.g using

fluxing additives) and the trend to work with finer

powder and thinner layers (Present-day minimum

layer thickness is about 0.05 mm with typically

30µm powder)

2.3 Type of lasers

Commercial SLS machines (DTM and EOS) are

all equipped with CO2 lasers with maximum power

ratings between 50 and 200W The university of

Leuven developed two SLS machines equipped

respectively with Nd:YAG lasers of 300 and 500W

9), while the University of Connecticut and

University of Manchester used a 60W diode laser of

810nm 10),11) The University of Liverpool explored

the use of Q-switched Nd:YAG and short-pulse

Cu-vapour lasers 35) In future, other type of lasers might show up, like diode pumped solid state lasers and others

Fig 2: Absorption of light by some metals and polymers for different wavelengths

Optimally, the laser wavelength should be adapted to the powder material to be sintered, because the laser absorption coefficient greatly changes with the material and the frequency or wavelength of the laser light (Fig 2) 12), 13) From Fig

2, it could be anticipated that CO2 lasers (wave length 10.6µm) might perform better with polymer powders, while Nd:YAG (wavelength of 1.06µm) will be superior for metallic materials, since metals depict higher absorption at shorter wavelengths This

is confirmed by several studies of the University of Leuven in which Fe-Cu metal powder and WC-Co hardmetal or cermet powder were alternatively sintered with both laser types 14) Results proved that for the same amount of energy or at similar settings (laser power and scan speed), a Nd:YAG laser results in higher green part density, a larger sintering depth (allowing thicker layers to be sintered, hence reducing production time) and a higher yield (process efficiency) Few results are presented further in this paper (see Figs 5 and 6) The study also proved that, for those materials, the processing window of a YAG laser is larger than that of a CO2

0 0 0

0 0 5

0 1 0

0 1 5

0 2 0

0 2 5

0 3 0

W a v e l e n g t h ( m )  

0 1 0 2 0 5 1 2 1 0 2 0

N d : Y A G

L a s e r

C O

L a s e r

2

S t e e l

I r o n

C o p p e r 5

laser (Fig 5) This allows a wide variation of the

processing parameters (power and scan speed) or of

the powder composition (here mixture ratio between

Fe and Cu), while still ensuring good sintering

results (i.e good liquid phase sintering – see section

on SLS of metals) This enforces good process

controllability and reliability

2.4 Optics and scanners

The optics serve to bring the laser beam from the

laser source to the processing area and to focus it

onto the powder surface, while the scanner moves

the beam across the surface Scanning is normally

done with oscillating galvano mirrors that yield

higher scan speeds than those achievable with XY

tables translating mirrors and focussing devices

Nd:YAG lasers allow the use of fibre optics to guide

the beam The University of Leuven used such glass

fibre to guide a 500W CW Nd:YAG beam to a

focussing unit fixed to an XY table 9) The use of

such combination ‘fibre optics/table scanner’

eliminates the need for an expensive flat field lens

(needed with galvano scanners to convert the

spherical focussing plane into a flat plane coinciding

with the powder surface) Moreover, this solution

does not need any mirror and allows for higher

power (here 500W), whereas galvano mirrors can

not support over 300W (Above this power, mirrors

need powerful internal water cooling and their

inertia puts unrealistic requirements to the galvano

scanner) Using a fibre guiding a YAG beam has the

advantage to act as a beam integrator (i.e conversion

of Gaussian or other TEM power profile to a

constant power profile across the beam cross

section) It has the disadvantage to induce loss in the

beam quality, thus lower focability Using high

integrated powers with defocused beams, on the

other hand, allows to sinter wide tracks (10-15mm

wide) at once, thereby reducing the scan time for

sintering large areas However, the fibre/table

solution was abandoned in a later stage in favour of

a faster 300W YAG-beam galvano scanner, since

powers of 100-300W seemed sufficient for sintering

most materials (including refractory materials like

hardmetals, cermets and ceramics)

The University of Erlangen developed special

optics to split a laser beam in two, defocus one beam

and recombine the two beams concentrically into an

intensive focused spot surrounded by a larger less

intensive (defocused) spot 15) The large spot is used

to pre-heat the powder in front of the moving

intensive sintering spot

2.5 Processing chamber, atmosphere, powder

pre-heating

To achieve proper laser sintering, oxidation of

the powders should be avoided This is mainly

crucial for metals depicting very high affinity to O2,

but is not negligible for polymers and some types of

ceramics Therefore, SLS machines are generally

equipped with processing chambers working with a protective atmosphere (mostly N2 in case of polymers or metals, Ar for sensitive metals) Some machines have been designed 9),16) or modified 17) to operate under vacuum Those machines are aimed at sintering metals, although sintering in vacuum or low pressure might induce sublimation of the metal powder Today, the vacuum facility of the chamber

of the proprietary SLS machine of the University of Leuven is mainly used for O2-degassing of the powder (extracting O2 from the chamber or O2

trapped in the pores within the powder) After degassing, the chamber is filled with low pressure N2

or Ar, ensuring an O2-concentration below 0.5% Many SLS processing chambers are further equipped with powder pre-heating systems Pre-heating the powder prior to sintering yields several advantages including drying the powder (removing the adsorbed humidity), reducing the laser power and interaction time needed for sintering, and lowering the thermal gradient generated in this process thus reducing the subsequent thermal stresses and distortions

3 SLS OF POLYMERS

Polymer powders were the first and are still the most widely applied materials in SLS

Amorphous polymers, like polycarbonate

powders, are able to produce parts with very good dimensional accuracy, feature resolution and surface finish (depending on the grain size), but they are only partially consolidated As a consequence, these parts are only useful for applications that do not require part strength and durability Typical applications are SLS models for the manufacture of silicone rubber and cast epoxy moulds 18)

Semi-crystalline polymers, like nylons

(polyamide), on the contrary can be sintered to fully dense parts with mechanical properties which approximate those of injection moulded parts On the other hand, the total SLS process shrinkage of these semi-crystalline polymers is typically 3 - 4 %

19), which complicates production of accurate parts The good mechanical properties of these nylon based parts makes them particularly suited for high strength functional prototypes New grades of nylon powders (i.e Duraform PA1220)) even yield resolutions and a surface roughness close to those of polycarbonate (PC), making polyamide (PA) also suited for casting silicone rubber and epoxy moulds, even though parts with higher resolution and smoother surfaces can be produced from amorphous powders Those “Duraform” polyamides yield improved accuracy, processibility and recyclability,

as well as reduced nitrogen consumption 21) Other polymer-based materials available commercially are acrylic styrene (PMMA/PS) for

investment casting and an elastomer for rubber-like

applications

Table 2 gives an overview of the mechanical

properties of some SLS polymer materials (DTM)

4 SLS OF REINFORCED AND FILLED

POLYMERS

Polyamide powders can be relatively easily

reinforced with other materials in order to further

improve their mechanical and thermal properties

Several grades of glass fibre reinforced polyamide

powders are readily available on the market 22)

Polyamide coated copper powder (Cu-PA) is also

available for the production of plastic-metal

composite injection tools 23) This Cu-PA powder

mixture contains 70 wt% Cu (rest is PA) Compared

to plain PA parts, Cu-PA SLS parts are 3.5x heavier

(density of 3.45 g/cm3), 4x more thermally

conductive (1.28 W/m°C) and exhibit a similar

tensile strength (34 MPa) but a higher tensile

modulus (3.4 GPa) Applications include producing

Cu-PA inserts for building moulds Those Cu-PA

moulds can be used as laser sintered without need

for removal of the PA phase or without any

post-densification process To improve the surface finish

of the SLS Cu-PA inserts, the latter are first surface

coated with a resin (epoxy, acrylate, Imprex

Superseal) to fill the porosities and then finished

with a flexible cloth Molds containing such inserts

can deliver 200-400 parts in common plastics 23)

5 SLS OF METALS AND

HARDMETALS/CERMETS

The production of functional components from

metallic or ceramic powders or a combination of

those by SLS process is a very promising area

Industrial applications have already emerged

particularly in rapid tooling Other potential

applications include: one-of-a-kind complex metallic

or ceramic parts, prototyping of cutting tools,

moulds and inserts, EDM electrodes, etc The

following sections summarise the main

developments related to powders and laser sintering

mechanisms investigated so far in this rapidly

developing field

5.1 SLS of metals and cermets with polymer binder or infiltrant

DTM Corporation (Austin, USA) has developed

a process that applies polymer-coated steel powders (1080 steel, 316 or 420 stainless steel particles coated/mixed with a thermoplastic/thermoset material) for the SLS of metal parts During laser sintering, the polymer melts and acts as a binder for the steel particles A post treatment is necessary in which the polymer is burned out and the porous part

is infiltrated with copper or bronze 18),24) Over several years, DTM improved continuously their production process by reducing the number of post-processing cycles and their total duration For the third generation of RapidSteel called LaserForm

ST-100, composed of 60% 420 stainless steel and 40% bronze (89Cu-11Sn), DTM reported an infiltration time of 24h performed in one single step under pure nitrogen These developments resulted in improving several material properties of the final SLS parts such as strength, hardness, machinability, weldability, wear rate and thermal conductivity 24) Using this indirect SLS process, the University

of Texas at Austin sintered SiC particles coated with

a proprietary polymer binder to obtain a SiC preform with a typical density of 40Vol% 25) After polymer debinding (at 400°C) the SiC preform becomes quite fragile preventing further handling To improve the strength of the SiC preform, an additional firing step (1100°C, 2hrs, formation of SiO2 layer) was utilized followed by (a pressureless) infiltration (670°C) with a Mg-based (AZ91D) die casting alloy 25) EOS GmbH (Munich, Germany) avoids the use

of a polymer binder by directly sintering metal powders with a low melting point: i.e bronze-nickel based powders (EOS-Cu 3201 containing Cu-Sn, Cu-P and Ni particles) developed by Electrolux Co

2) After SLS, the part is infiltrated with epoxy resin

to fill in the porosities Hence the final part is a bronze-epoxy composite, rather than a plain metallic part and its mechanical and thermal properties are limited Infiltration with a metal like Cu or bronze is not possible in this case, since the green part would melt during infiltration Lately, EOS put into market

a new powder (EOS-DMLS Steel 50-V1 containing steel, Cu-P and Ni particles) yielding improved

Table 2: Overview of the mechanical properties of some SLS polymer materials (DTM).

Surface roughness Ra as

SLS processed (m) 7 12/8.5** value for DuraForm PA15

-mechanical properties The SLS part is about 70%

dense and thus can be used as such for inserts and

small mold components

5.2 SLS of metals and hardmetals/cermets by

liquid phase sintering

Many research institutes study the possibility of

directly laser sintering metal and ceramic powders

without use of any polymer component 16),27)-35)

Several approaches for binding the powder

particles together using laser beam energy have been

investigated at the University of Leuven including:

solid state vs liquid phase sintering, loose vs

pre-coated metal binder phase, mixed vs milled powders

30),36) Further research mainly focussed on liquid

phase sintering of uncoated powder mixtures The

basic material used in liquid phase sintering (LPS)

consists of a mixture of two metal powders (or a

metal and a ceramic powder) : a high melting point

metal/ceramic, called the structural material, and a

low melting point metal, called the binder Applying

heat to the system causes the binder to melt and to

flow into the pores formed by the non-molten

particles The classical stages of LPS are

schematically shown in Figure 3

The main advantage of liquid phase sintering is

the very fast initial binding This binding is based on

capillary forces, which can be very high: the reaction

speed in this stage is determined by the kinetics of

the solid-melt transformation This transformation is

several orders of magnitudes faster then solid-state

diffusion occurring in solid phase sintering

Once the binder metal is molten and spread out

into the solid lattice, the system cools down (because

the moving laser beam no longer feeds energy into

the material) and the situation is frozen Only the

first stage (rearrangement) of the LPS mechanism

takes place during laser sintering because of the very short laser-material interaction time (fraction of a second)

Early studies to sinter steel powder mixed up

with copper grains acting as binder material were performed at the universities of Aachen (Germany)

16), Leuven (Belgium) 27) and Erlangen (Germany) 28) The universities of Texas at Austin (USA) 38) and Leuven (Belgium) 32)-34) and some German Fraunhofer research centres 37) succeeded to laser

sinter hardmetals (i.e cemented carbides) and

cermets by SLS Figure 4a shows a typical

microstructure of a WC-9wt%Co powder mixture sintered by a CO2 laser showing a good bonding between the WC particles surrounded by a Co-binder layer and the presence of large pores (50-60%) After infiltration, Cu filled these open porosities(Fig 4b)

Such LPS green part has enough strength to withstand a post-processing cycle to bring the part to full density This post-processing consists of a furnace post-sintering to proceed with the next stages of LPS shown schematically in Fig 3 (solution reprecipitation and solid state sintering) or

an infiltration with a low melting point metal

A wide variety of powder combinations have been investigated in Leuven using LPS mechanism, including: Fe-Cu, Cu-coated Fe, Fe3C-Fe, Stainless Steel-Cu, Stainless Steel-CuFeCo-Co, Stainless Steel-CuP-Co, WC-Co, Co-coated WC, WC-Cu, WC-CuFeCo, TiC-Ni/Co/Mo, TiB2-Ni, ZrB2-Cu,

It is worth noting that finding the processing window yielding good sintering behaviour using LPS mechanism is not always evident and easy to obtain For instance in case of Stainless Steel-Cu powder mixture, the processing window turned out

1 R e a r r a n g e m e n t

S t r u c t u r a l e l e m e n t ( s o l i d p h a s e )

B i n d i n g e l e m e n t

P o r o s i t y

Fig 3: Different stages of Liquid Phase Sintering

C o

W C

P o ro s ity

Fig 4: SEM micrographs of WC-9wt%Co powder processed by SLS revealing (a) the distribution of WC particles within the Co binder matrix and (b) Cu filling the open pores after infiltration

to be quite narrow when this powder was sintered

with a CO2 laser and relatively larger using a

Nd:YAG laser as shown in Figure 5

The small difference in melting points between

the structural element (stainless steel) and the binder

element (Cu) and the high reflectivity of Cu particles

for laser light, particularly with CO2 lasers having a

longer wavelength (10.6 µm), induced mostly

simultaneous melting of both elements Owing to the

larger difference in their melting points, such

behaviour was not observed in the WC-Co powder

mixture and most of the SLS tests resulted in LPS

without much difficulty Further details about these

results can be found elsewhere 14),34),47) The type of

laser (and thus wavelength) was also found to have

an effect on the density of the SLS parts Parts

sintered with Nd:YAG laser gave a higher density at

similar processing parameters for both Stainless

Steel-Cu and WC-Co powder mixtures (Fig 6) 14)

It has also been shown that by optimising the

process parameters related to both laser (power, scan

speed, scan spacing) and powder (particle size,

powder composition, mixing vs milling of both

elements), the properties (such as surface quality,

density) of the sintered product were improved 34), 36)

For WC-9wt%Co material, it has been shown that the mechanically alloyed powders resulted in higher

green densities with better surface roughness compared to the layers prepared from the mixed powders 36) Furthermore, for a given scan speed, the surface quality of the sintered layers prepared from milled powders (as well as mixed powders) is improved with smaller particle size The main difficulty with the milled powder is to obtain a smooth and uniform layer during powder deposition For that, further research is underway to develop a deposition mechanism capable of depositing fine powders (mixed or milled) with particle sizes below

20 µm

5.3 SLS of metals by through melting

The Fraunhofer Institutes ILT and IPT (Aachen, Germany) applied a 300W Nd:YAG laser to completely melt metal powders (bronze, steel, stainless steel such as 316L) deposited in a standard way using a wiper (scraper) and producing directly dense parts (density > 95%) 39) Due to the melting nature of metals (tendency to form droplets to minimise their surface energy), a careful control of the process parameters is needed Moreover,

M e l t

L i q u i d P h a s e S i n t e r i n g

M e l t

F r a g i l e

N d : Y A G L a s e r

5

1 0

1 5

2 0

2 5

3 0

5

1 0

1 5

2 0

2 5

3 0

L i q u i d

P h a s e

S i n t e

-r i n g

S p e c i f i c I n d u c e d E n e r g y ( J / c m 2) S p e c i f i c I n d u c e d E n e r g y ( J / c m 2) Fig 5: Processing window for Stainless Steel-Cu sintered with both CO2 and Nd:YAG lasers as a

function of specific induced energy and vol% binder (Cu)

R e l a t i v e d e n s i t y f o r b o t h l a s e r s o u r c e s

2 0

2 5

3 0

3 5

4 0

4 5

5 0

5 5

0 1 2 5 2 5 0 3 7 5 5 0 0 6 2 5 7 5 0

P o w d e r 7 0 - 3 0 N d : Y A G

P o w d e r 7 5 - 2 5 N d : Y A G

P o w d e r 8 0 - 2 0 N d : Y A G

P o w d e r 7 0 - 3 0 C O 2

P o w d e r 7 5 - 2 5 C O 2

P o w d e r 8 0 - 2 0 C O 2

S p e c i f i c I n d u c e d E n e r g y ( J / c m 2)

Fig 6: Relative density of Stainless Steel-Cu SLS parts processed by CO2 and Nd:YAG lasers

overhangs with angles higher than 60° could not be

built with this process To maintain dimensional

control and improve accuracy and surface roughness

(Ra 50-80 µm), a final touch by high speed milling

of the lateral sides is utilised When this process was

used to sinter Al-30%Si, a maximum density of

90-95% was obtained 37)

EOS recently came to the market with a plain

steel powder that is laser sintered by through

melting The average particle size is 50µm, but an

enhanced steel powder with 20µm size is

announced

Osaka University (Japan) utilised a pulsed

Nd:YAG laser (50W mean power, 3kW maximum

peak power) to melt pure Ti spherical powders (200

µm and 50 µm average particle size) to produce

medical parts (dental crowns and bone models) 40)

For the coarse Ti powder, the SLS part delivered a

maximum relative density of 84% yielding a

maximum tensile strength of 70MPa Using fine Ti

powder (25 µm), a higher relative density

(maximum 93%) was achieved with a tensile

strength of 150MPa Due to the presence of

remaining porosity, the tensile strength of these SLS

parts is still lower than that of bulk pure Ti material

(275 - 481 MPa) 40)

5.4 Combination SLS/HIP

To produce fully dense metallic parts, the

University of Texas at Austin used a combination of

SLS and HIP processes 41),42) The objective of that

research project was to produce fully dense

functional metal parts using a SLS process

generating a porous core encapsulated in a fully

dense skin (i.e integrated SLS canning) followed by

a HIP treatment The can or skin material formed

directly by SLS (with density > 95%) around the

laser sintered metallic component (60-80% dense)

plays the role of an encapsulation material during

HIPing and becomes part of the final component

This process reduces the production time and costs

associated with the encapsulation and can/skin

removal after HIP The metals considered in that

research project were Inconel 625 Superalloy,

Stainless Steel (17-4 PH), Ti-alloys (Ti6Al4V) and

Molybdenum The University of Texas at Austin

claimed successful results in using the SLS/HIP

combination to process metallic parts 42)

6 SLS OF CERAMICS

The ILT and IPT Fraunhofer Institutes used also

the SLS process in an attempt to produce directly

ceramic parts without polymer binder material The

absence of any binder element makes the ceramic

laser sintered part very fragile and viable to

breakage Due to the short reaction time involved in

SLS, solid state sintering is not feasible To sinter

SiC powder material, a sufficient amount of laser

energy was supplied to induce high local

temperatures leading to a disintegration of part/surface of SiC particles into Si and C The free

Si then oxidises and forms SiO2, which plays a role

of a binder between the SiC particles 39),43) After laser sintering, the SiC parts could be infiltrated with

Si and reaction bonded to full density Zirconium silicates were also laser sintered by almost fully melting the powder particles forming large agglomerates 43),44) Similar to DTM’s polymer coated powder process, graphite coated with phenolic resin was also processed by SLS by melting only the polymer binder, which is burned out afterwards However, the resulting graphite part becomes very fragile 43)

7 SLS OF FOUNDRY SAND

Both commercial SLS machine vendors (DTM and EOS) offer sand powders that can be laser sintered in order to produce foundry sand moulds DTM, for instance, offers both Zr and Si sand: SandForm ZrII and Si released in 1997 45) Key characteristics include Shell Foundry Sand of given AFS grain fineness number (GFN# = 97 for Si and

99 for ZrII) and dimensional tolerances of 0.5mm SandForm Si, used predominantly for Al castings, is based on silica, which is prevalent in the market and has a low density SandForm ZrII can be used for both Al and Fe castings and its binder system matches silica Demonstrated applications are castings of power-train components, manifolds, automotive and heavy machinery parts

8 APPLICATIONS OF SLS

Fig 7: Polyamide SLS prototype (DTM) The largest application of SLS is still the production of rapid prototypes from plastic material (Fig 7) One of the advantages is the good strength

of SLS polymer prototypes as compared to

prototypes made by e.g stereolithography or ink jet

printing SLS allows to make functional prototypes

from traditional engineering thermoplastics (e.g

nylon) with properties nearly equal to injection

moulded parts

Fig 8: A plastic injection mould whose cavity

and core were made with the Rapidsteel 2 powder

Table 3 : Injection parameters for Rapid steel SLS

mould 46)

pressure

Injection temp.

Injection pressure

Cycle time

(350kN)

(60MPa)

40 sec

Polyamide

6.3% glass

(2MPa)

14 sec

Today, SLS of metals (with or without polymer

binders) mainly finds industrial applications for

‘rapid tooling’, i.e for fast production of moulds and

dies Fig 8 shows a plastic injection mould whose

cavity and core were made with the Rapidsteel 2

SLS process of DTM In comparison with

conventional tool making, the lead time for this

mould was reduced by 45% (22 days down to 12

days), while the cost dropped by 38% (inclusive

time and cost for CAD mould design, moulds

assembly, use and machining of standard mould

plates and accessories, etc.) 46) Typical injection

parameters and cycle times used with such mould for various plastic materials are reported in Table 3 Today, tool life times of 100,000 shots are reported with Rapidsteel moulds

Several WC-Co hardmetal injection moulds

were produced at the University of Leuven 32), 47) Figure 9 shows a number of 3-D green parts produced in Leuven, some of which representing mould inserts The parts are made from WC-9wt

%Co powder mixtures After laser sintering, the green part exhibits a density of about 40-47vol% A post-treatment is thus necessary to achieve full densification This is done by an infiltration process

in a furnace with a low melting point metal (e.g Cu

or bronze) under a controlled atmosphere (a mixture

of nitrogen and hydrogen)

Fig 9: Various WC-Co green parts made by SLS Further improvements in the SLS parts including

a minimum layer delamination (due to thermal stresses) and a reduction in the production costs of parts made of WC-Co powder mixture (due to the high price of Co), were achieved using alternative binders to replace partially or completely the Co phase A number of multi-component binder substitutes (to Co) considered in this study are alloys based on Fe-Cu/Ni-Co 48)

SLS may also be used to produce functional

metal/cermet parts or prototypes An example is

given in Fig 10 It represents a plain WC/Co drill bit head for drilling in stone material as it appears directly after SLS (i.e before Cu infiltration) The geometrical features look quite reasonable, however the final surface roughness and strength (even after

Fig 10: Photographs showing (a) Hilti TE-CX drill bit with full hardmetal head welded to a tool steel

and (b) a WC-9Co drill bit head made by SLS process

Cu infiltration) are still inadequate to carry out a

drilling test performance on a rock Further process

optimization (powder characteristics, laser sintering)

is needed in combination with a post-sintering or

HIPing treatment to produce strong fully dense

WC-Co drill bit heads

9 MODELLING AND SIMULATING THE SLS

PROCESS

The final properties of parts fabricated by SLS

are very much dependent on the process parameters

as well as on the powder characteristics As a result,

each powder/machine combination may require

extensive testing in order to identify the processing

window and to optimise the processing parameters

suitable to achieve the desired properties for a

specific application This can be expensive and time

consuming To gain a better understanding of the

physical phenomena that are taking place in this

relatively new process, more fundamental and

modelling work is needed Moreover, this will

reduce the number of experiments required to

optimise the processing parameters for a given

powder system Several research institutes have

taken initiatives to tackle this task using different

approaches The University of Leuven applied a

simple ray-tracing model to evaluate the laser

radiation penetration into a metal powder bed and

the resulting powder absorptance 49),50) At the

University of Texas and University of Leeds,

thermally-based finite element models have been

used to simulate SLS of amorphous polymers 51)-52)

Recently, the two-dimensional model developed at

the University of Leeds has been extended to the

study of crystalline polymers and metals in two- and

three-dimensions 53),54)

The ray tracing model of the university of

Leuven has been used to simulate laser sintering of

Fe-Cu and WC-Co powders using Nd:YAG and CO2

lasers The results include the evaluation of the total

energy incoupling (also called absorptance), the

optical penetration of the laser beam, and the

estimation of the sintering zone dimensions

(thickness and width of a single laser sintered track)

As an example of the results obtained so far, Fig.11

shows the calculated and measured sintering zone

dimensions when sintering Fe-Cu powder with a

Nd:YAG laser 55)

10 CONCLUSION

On the long term, selective laser sintering could

turn out to be one of the most successful rapid

prototyping processes, mainly because of its unique

ability to process nearly any material During just

one decade of existence, it has been demonstrated

that SLS may successfully be applied to produce

parts in a wide range of polymers (elastomers,

amorphous and semi-crystalline technical polymers),

metals, hardmetals/cermets, ceramics and sand Among others, it has been demonstrated that SLS is well suited to produce a variety of composite materials: glass reinforced polymers, metal/polymer composite (e.g Cu/PA), metal/metal composites (e.g Fe/Cu), cermets (e.g WC-Co) and others In many cases, standard off-the-shelf powder materials can be used, without the need to develop dedicated powders

It is expected that further development of the SLS process, equipment and of appropriate powder materials will further boost this rapid prototyping process in the years or decades to come

Fig 11: Thickness and width versus scan speed for Fe-30wt%Cu sintered track

11 REFERENCES

1) U Heymadi, K McAlea, Selective Laser Sintering of metal molds: the RapidTool process, Solid Freeform Fabrication Proceedings, pp

97-104 (1996)

2) U Behrendt, M Shellabear, The EOS Rapid Prototyping concept, Computers in Industry, 28,

pp 57-61 (1995)

3) S Das, N Harlan, J.J Beaman, D.L Bourell, Selective Laser Sintering of high performance high temperature materials, Solid Freeform Fabrication Proceedings, pp 89-95 (1996)

4) J.J Beaman, C.R Deckard, Solid freeform fabrication and selective powder sintering, 15th

NAMRC, North American Manufacturing Research Conference Proceedings, pp 636-640, (1987)

5) B Van der Schueren, J.-P Kruth, Powder deposition in selective metal powder sintering, Rapid Prototyping Journal, Vol I (3), 23-31 (1995)

6) B Lauwers, I Meyvaert, J.-P Kruth, Layer deposition in SLA and SLS, Prototyping Technology International ‘97, U.K & Intern Press, Dorking, pp 139-144, (1997)

7) J Bakkelund,, R Karlsen, Ø Bjørke, Fabrication

of Metal Objects Using Layer Manufacturing Technology and Powder Metallurgy Science, CIRP Annals, 46/1, pp 135-138, (1997)

Sintering Zone Dimension: Nd:YAG Laser - Power = 28.87 W

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Scan speed (mm/s)

Thickness Measurements Width Simulations Width Measurements

8) P Oborski, Design of a powder deposition

system and mechanical equipment to SLS,

Master Thesis at K.U Leuven, Dept Mechanical

Engineering, (1997)

9) B Van der Schueren, J.-P Kruth, Design aspects

of a ‘Selective Metal Powder Sintering’

apparatus, Proc of 11th Int Symp for

Electro-machining (ISEM-XI), pp 651-662, (1995)

10)T Manzur, C Roychoudhuri, H Marcus, SFF

using diode lasers, Proc Solid Freeform

Fabrication Symposium, pp 363-368, (1996)

11) L Li, K.L Ng, A Slocombe, Diode laser

sintering of compacted metallic powders for desk

top rapid prototyping, Proc of 7th Europ Conf

on Rapid Prototyping and Manufacturing, pp

281-296, (1998)

12)N.K Tolochko, T Laoui, Y.V Khloplov, S.E

Mozzharov, V.I Titov, M.B Ignatiev,

Absorptance of powder materials suitable for

laser sintering, Rapid Prototyping Journal, Vol

6(3), pp 155-160, (2000)

13)F Olsen, O Femming, Theoretical investigations

in the fundamental mechanisms of high intensity

laser light reflectivity, Proc of SPIE, Vol 1020,

High Power CO2 Laser systems and applications,

A Quenzer, pp 114-122, (1989)

14)J.P Kruth, P Peeters, T Smolderen, J Bonse, T

Laoui, L Froyen, Comparison of CO2 and

Nd-YAG lasers for use with Selective Laser

Sintering of Steel-Copper powders, International

Journal of CADCAM and Computer Graphics,

Vol 13, No 4-6, pp 95-110 (1998)

15)M Geiger, A Coremans, N Neubauer, F

Niebling, Advanced Rapid Prototyping by laser

beam sintering of metal prototypes - design and

development of an optimised laser beam delivery

system, R.-J Ahlers, G Reinhart (Eds.), Proc of

the Europto Series-Rapid Prototyping, Besançon,

France, pp 103-114, (1996)

16)Y Song, W Konig, Experimental study of the

basic process mechanism for direct selective

laser sintering of low-melting metallic powder,

Annals of the CIRP, vol 46/1, pp 127-130,

(1997)

17)A Coremans, Direktes Laserstrahlsintern von

Metallpulver – Prozessmodellierung,

System-technik, Eigenschaften laserstrahlgesinterter

metallischer Werkstücke, Ph.D., Univ Erlangen,

(1998)

18)K McAlea, et al., Materials and applications for

the Selective Laser Sintering Process, Proc 7th

Int Conf on Rapid Prototyping, pp 23-33,

(1997)

19)R Grimm, SLS and SLA, different technologies

for different applications, Prototyping

Technology International ‘97, UK & Intern

Press, Dorking, UK, pp 130-138, (1997)

20)B.M Schumacher, G.N Levy, Selective Laser

Sintering combined with systematic development

of new powders enabling innovative and prosperous redesign of process-layouts, Proc

12th Int Symp for Electro-Machining, pp

633-639, 1998)

21)S Seitz, New materials, 2nd Europ SLS users meeting, K.U Leuven, Belgium, pp 1-15, (1998)

22)S Seitz, et al., New Materials for New Applications in SLS, Proc LANE’97 , pp

623-628, (1997)

23)S Bruning, Copper Polyamide, 2nd Europ SLS users meeting, K.U Leuven, Belgium, pp 1-10, (1998)

24)K McAlea, DTM’s Selective Laser Sintering technology: new products, new benefits, new manufacturing horizons, Proc 8th Int Conf on Rapid Prototyping, pp 347-352, (2000) + Conf Presentation

25)M Wohlert, D Bourell, Rapid prototyping of Mg/SiC composites by a combined SLS and pressureless infiltration process, Proc Solid Freeform Fabrication Symposium, pp 79-87, (1996)

26)PM update, EOS expands its range of metal powders for rapid prototyping users, Metal Powder Report, p 6, (1998)

27)B Van der Schueren, J.P Kruth, Laser based selective metal powder sintering: a feasibility study, Proc 26th Int CIRP Sem on Manufacturing Systems, LANE’94, pp 793-802, (1994)

28)A Coremans, M Kauf, P Hoffman, “Laser assisted Rapid Tooling of molds and dies”, Proc

of 5th Europ Conf on Rapid Prototyping and Manuf., p 195 (1996)

29)S Das, N Harlan, J.J Beaman, D.L Bourell, Selective Laser Sintering of high performance high temperature metals, Proc Solid Freeform Fabrication Symposium, pp 89-95 (1996) 30)J.-P Kruth et al., Basic powder metallurgical aspects in selective metal powder sintering, CIRP Annals 45(1), pp 183-186, (1996)

31)D.L Bourell, H.L Marcus, J.J Beaman, Selective laser sintering of metal and ceramics, Int J Powder Met., Vol 28 No.4, pp 369-381 (1992)

32)J.-P Kruth, L Froyen, B Morren, J.E Bonse, Selective Laser Sintering of WC-Co ‘hard metal’ parts, Proc 8th Int Conf on Production Engineering (ICPE), Japan, pp 149-156, (1997) 33)T Laoui, L Froyen, J.-P Kruth, “Selective Laser Sintering of Hard Metal Powders”, Proc Rapid Prototyping and Manufacturing ’98 Conf., pp.

435-467, (1998)

34)T Laoui, J Bonse, L Froyen, J.-P Kruth, “ Influence of powder parameters on Selective Laser Sintering of Tungsten Carbide-Cobalt”, Proc 7th Europ Conf on Rapid Prototyping and Manufacturing, pp 271-279, (1998)