Ž .Materials and Design 20 1999 115]121

Processing of titanium net shapes by SLSrHIP

Suman DasU, Martin Wohlert, Joseph J. Beaman, David L. BourellLaboratory for Freeform Fabrication, The Uni ersity of Texas at Austin, Austin, TX 78712-1063, USA

Abstract

Ž .SLSrHIP is a hybrid direct laser fabrication method that combines the strengths of selective laser sintering SLS and hotŽ .isostatic pressing HIP . SLS can produce complex shaped metal components with an integral, gas impermeable skin. These

components can then be directly post-processed to full density by containerless HIP. SLSrHIP is envisioned as a rapid, low-costreplacement for conventional metal can HIP processing. The advantages of freeform fabrication combined with in situ HIPencapsulation include the ability to perform containerless HIP, no adverse container]powder interactions, reduced pre-processingtime, and fewer post-processing steps compared to conventional HIP of canned parts. SLSrHIP is currently being developed forInconelW 625 superalloy and Ti-6Al-4V. This paper focuses on microstructure and mechanical properties of SLS processed andHIP post-processed Ti-6Al-4V. SLSrHIP technology for Ti-6Al-4V was demonstrated by fabricating a subscale AIM-9 missileguidance section housing to specification. This work is funded jointly by DARPA and ONR under contract N00014-95-C-0139titled ‘Low Cost Metal Processing Using SLSrHIP’. Q 1999 Published by Elsevier Science Ltd. All rights reserved.

Keywords: Selective laser sintering; Hot isostatic pressing; Direct fabrication; Titanium

1. Introduction

Over the last 10 years, SFF technologies worldwidehave attained a state of maturity. A variety of SFFtechnologies are commercially available to producecomplex shaped three-dimensional parts and tooling ina variety of materials including plastics, paper, po-lymers, wax, sand, ceramics and metals. The next majoradvance in SFF research and development is takingplace in the area of direct fabrication processes, espe-cially for low-volume production of functional metal,cermet and ceramic components or tooling. Direct fab-rication implies layerwise shaping and consolidation of

Ž .feedstock e.g. powder, wire, ingot, paste or melt tocomplex shapes having full or near full density withoutthe use of intermediate binders, furnace densificationcycles, or secondary infiltration steps. These methodstypically require minimal thermomechanical post-processing or machining to obtain desired geometry,

U Corresponding author. Tel.: q1-512-471-5838; fax: q1-512-471-87827; e-mail: das@sffoffice.me.utexas.edu

structure and properties. There is a growing demandamongst industrial and government users for directSFF processes that enable production of small lot or‘one of a kind’ functional metal, cermet and ceramiccomponents. A number of direct fabrication methods

w xare under development 1]7 . The materials systemsinvestigated by these methods include steels, nickelbase superalloys, titanium and its alloys, refractorymetals, bronze-nickel and cermets.

Ž .Selective laser sintering SLS is a SFF techniquethat creates three-dimensional freeform objects directlyfrom their CAD models. An object is created by se-quentially fusing thin layers of a powder with a scan-ning laser beam. Each scanned layer represents a crosssection of the object’s mathematically sliced CADmodel. In direct SLS, a high energy laser beam directly

Ž .fuses a metal or cermet powder to high density )90% ,preferably with minimal or no post-processing require-ments.

As part of ongoing direct SLS research efforts at theUniversity of Texas, a hybrid net shape manufacturingtechnique called SLSrHIP is being developed for high

w xperformance metal components 8,9 . SLSrHIP com-

0261-3069r99r$ - see front matter Q 1999 Published by Elsevier Science Ltd. All rights reserved.Ž .PII: S 0 2 6 1 - 3 0 6 9 9 9 0 0 0 1 7 - 5

( )S. Das et al. r Materials and Design 20 1999 115]121116

bines the freeform shaping capability of SLS with thefull densification capability of HIP to produce netshape, high value metal components at significantlyreduced costs and shorter lead times. Such componentsinclude fighter aircraft turbine engine hardware, navaland submarine components. Since direct SLS had pre-viously demonstrated the potential to consolidate metalpowders to net shapes with densities in excess of 80%

w xtheoretical density 10 , SLSrHIP is envisioned as anatural combination of SLS and containerless HIP toobtain full density metal parts.

The working principle of SLSrHIP is to consolidatethe interior of a component to 65% or higher densityand to fabricate an integral, fully dense, gas imperme-

w xable skin or ‘can’ at the part boundaries in situ 11 . Acomponent is produced by selectively consolidating ametal powder with a laser beam layer by layer. Unlikeother direct laser fabrication processes that strive tofuse powder to full density across each layer, inSLSrHIP the laser beam fuses the metal powder onlyat the part boundaries to form an integral, gas im-permeable skin or ‘can’ to a density exceeding 92%theoretical density. This is the fractional density atwhich the porosity typically changes from intercon-

w xnected or surface connected to closed 12 . The powderin the interior of each layer cross-section can either beleft unprocessed so as to provide approximately 65%packing density or be optionally laser sintered to anintermediate density typically exceeding 80% of theo-retical density. Thus, the net shape PrM part is shaped,canned, evacuated and sealed in situ. The encapsulatedpart is then post-processed by containerless HIP to fulldensity. A final machining step may be applied if neces-sary.

SLSrHIP has several advantages over conventionalHIP methods. Since an integral skin or ‘can’ is formedof the same material as the part, a secondary canningstep is not necessary. The integrally canned, net shape

component produced by SLS is directly post-processedby containerless HIP to full density. Adverse con-tainer]powder interactions are eliminated and post-HIP container removal is not required. Tooling andpre-processing steps associated with container fabrica-tion and filling are also eliminated. SLSrHIP enablesproduction of complex shapes at reduced cost andshorter lead-times.

Under the ongoing research program, SLSrHIPprocess development is being undertaken for two mate-

w xrials. Based on a survey 8 of several naval installa-tions, the materials selected for technology develop-ment and demonstration are the alloys InconelW 625and Ti-6Al-4V. The component selected for SLSrHIPtechnology demonstration of Ti-6Al-4V is the AIM-9Sidewinder missile guidance section housing, shown inFig. 1.

2. SLSrrrrrHIP processing

To demonstrate the feasibility of constructing inte-grally canned shapes by SLS suitable for HIP post-processing, a number of cylindrical specimens, 0.5inches diameter, 2 inches long with a 0.100-inch skinwall thickness were produced by SLS. A typical inte-

Fig. 1. Guidance section housing base for the Sidewinder missile,baseline component for Ti-6Al-4V SLSrHIP technology demonstra-tion.

Fig. 2. Typical SLS processed, Ti-6Al-4V integrally canned cylinderical specimen.

( )S. Das et al. r Materials and Design 20 1999 115]121 117

grally canned cylindrical specimen is shown in Fig. 2.These specimens were post-processed to full density bya HIP cycle consisting of 4 h at 9258C and 103 MPaŽ .15 000 psi .

3. Microstructure and mechanical properties

To compare the microstructure of SLSrHIPprocessed Ti-6Al-4V against conventionally canned and

HIPed Ti-6Al-4V, a reference specimen from the samelot of powder as that used to fabricate cylindricalspecimens was encapsulated in glass and post-processedusing the same HIP cycle. The results, shown in Fig. 3,indicated that the microstructure of SLSrHIPprocessed Ti-6Al-4V in the bulk is the same as that ofconventionally canned and HIPed Ti-6Al-4V. The skinregion of a SLSrHIP processed Ti-6Al-4V part exhibitsa microstructure different from that seen in the bulk.This is expected because during SLS, the skin region is

Ž . Ž .Fig. 3. Etched microstructure of glass encapsulation HIP Ti-6Al-4V a and SLSrHIP processed Ti-6Al-4V b .

( )S. Das et al. r Materials and Design 20 1999 115]121118

Fig. 4. Can]core interface.

Table 1Ti-6Al-4V hardness values

Material Processing Hardness

Ti-6Al-4V SLS processed skin 36 HRCŽ .Ti-6Al-4V Cast grade C-5 39 HRC

Ti-6Al-4V SLSrHIP processed core 35]37 HRCTi-6Al-4V Glass encapsulated HIP 34]36.5 HRC

completely melted by the laser and undergoes rapidsolidification, whereas the interior of the part is leftunprocessed, and undergoes densification during HIPpost-processing. The micrograph in Fig. 4 shows theinterface between the skin and the core. The differencein microstructures is clearly evident with the skin re-gion exhibiting a high aspect ratio Widmanstatten typemicrostructure while the core exhibits a coarser mi-crostructure that results from the coalescence of priorparticle grain boundaries.

Hardness was evaluated and compared against con-Ž .ventionally processed material Table 1 . SLS process-

ing of the skin to full density resulted in materialhaving hardness values comparable to cast material.Hardness values for the SLSrHIP processed coreagreed well with values for conventionally canned andHIPed material and did not differ much from thehardness values for the SLS processed skin region.

Tensile specimens conforming to ASTM E 8, roundŽtension test specimen small-size specimens, 0.25 inch

.nominal diameter were machined from integrallycanned Ti-6Al-4V cylinders fabricated by SLS andpost-processed to full density by HIP. The results indi-

Table 2Ti-6Al-4V tensile properties

Processing Tensile ElongationŽ .strength %

Ž .ksi

SLSrHIP 162 5Sidewinder specification 111 5

w xAnnealed 13 130 10w xSolution treated 14 150 8

Ž .Cast grade C-5 130 6

cated that the tensile strength specification for theSidewinder housing demonstration component was ex-ceeded while the minimum elongation was barely metŽ .Table 2 . Future testing will establish correlationsbetween tensile properties and oxygen levels inSLSrHIP processed parts.

4. Chemical analysis

Oxygen levels in commercially available titaniumfeedstock and finished product are critical from perfor-mance considerations. The oxygen levels can vary from0.08 to 0.40% by weight depending on the particularapplication. The strength typically increases with in-creasing oxygen content at the expense of ductility.Lower oxygen levels improve ductility, fracture tough-ness, stress-corrosion resistance and resistance to crackgrowth.

Oxygen and nitrogen levels in SLSrHIP processedTi-6Al-4V specimens were determined by ASTM E

( )S. Das et al. r Materials and Design 20 1999 115]121 119

w x1409-97 inert gas fusion technique 15 . The levels werecompared against levels in the starting powder as wellas against specifications for the Sidewinder housingcomponent. The results of analyses conducted by twoanalytical labs are shown in Table 3. All three SLSrHIPspecimens were taken from the core region of eachpart. The results indicated that the oxygen specificationon the Sidewinder housing was barely met. However, it

Ž TMis to be noted that both starting powders PREP and.Argon atomized have 0.19% by weight oxygen and the

powder used to fabricate each of the three SLSrHIPspecimens had been recycled at least six times. Thisindicates that on average 100 ppm oxygen pickup oc-curred during each SLS run. This is an extremelyencouraging result since it corresponds to the typicaloxygen pickup observed by Crucible Research in con-

w xventional metal can Ti-6Al-4V HIP processing 16 .To lower oxygen levels in SLSrHIP processed Ti-

6Al-4V as well as to increase powder recyclability, ELIŽ .Extra Low Interstitials grade Ti-6Al-4V powder with0.099% by weight or lower oxygen may have to be used.In addition, protocols will be established for mixingrecycled powder with virgin powder so as to maintainoxygen levels well within specification. During the es-tablishment of such protocols, tracking chronologicalchanges in oxygen levels of parts produced sequentiallyfrom the same lot of powder will be an importantcriterion. ELI grade Ti-6Al-4V powder will be tested infuture SLS trials.

5. Technology demonstration

A subscale version of the guidance section housingfor the AIM-9 Sidewinder missile was processed bySLS. This demonstration component, shown in Fig. 5was built at 90% scale. The typical length scale on thecomponent shown is 3 inches. Although this compo-nent was selected for SLSrHIP technology demonstra-tion, metallographic evaluation of the ‘as SLS

Table 3Oxygen and nitrogen levels in Ti-6Al-4V

Specimen Oxygen NitrogenŽ . Ž .wt.% wt.%

Sidewinder specification 0.25 max ]Ž .Ti-6Al-4V grade C-5 0.25 max

TMPREP Ti-6Al-4V 0.19 0.01Ar atomized Ti-6Al-4V 0.196 0.02

Ž .SLSrHIP 040898 Lab A 0.250 0.019Ž .SLSrHIP 040898 Lab B 0.255 0.037Ž .SLSrHIP 042998 Lab A 0.2485 0.073Ž .SLSrHIP 042998 Lab B 0.200 0.016Ž .SLSrHIP 050198 Lab A 0.251 0.038Ž .SLSrHIP 050198 Lab B 0.167 0.040

Fig. 5. SLS processed Ti-6Al-4V demonstration component.

processed’ component revealed full density, as shownin the micrographs of Fig. 6. This microstructure isconsistent with cast Ti-6Al-4V.

6. Conclusions

Feasibility of fabricating integrally canned complexshapes by SLS for containerless HIP post-processinghas been demonstrated for Ti-6Al-4V. Simple cylindri-cal shapes were fabricated and consolidated to fulldensity by HIP. Microstructure and mechanical proper-

Ž .ties hardness and tensile reveal that materialprocessed by the SLSrHIP technique is equivalent toconventionally processed material. Oxygen levels inSLSrHIP processed Ti-6Al-4V meet specification onthe component chosen for technology demonstration.A simplified version of a Ti-6Al-4V demonstrationcomponent was successfully fabricated.

Future work will focus on demonstrating SLSrHIPtechnology for InconelW 625. Further characterizationof structurerproperty relationships as well as optimiza-

( )S. Das et al. r Materials and Design 20 1999 115]121120

Fig. 6. Microstructure of fully dense, SLS processed Ti-6Al-4V demonstration component.

tion of processing parameters for build speed, skinthickness, oxygen content and HIP cycle time will alsobe undertaken.

Acknowledgements

The authors wish to acknowledge research funding

provided by DARPArONR under contract N00014-95-C-0139 titled ‘Low Cost Metal Processing UsingSLSrHIP’.

References

w x1 Das S et al. Direct SLS processing for production of cermetcomposite turbine sealing components. Mater Manu Process

Ž .1998:13 3 .

( )S. Das et al. r Materials and Design 20 1999 115]121 121

w x2 Griffith ML et al. Freeform fabrication of metallic componentsŽ TM .using laser engineered net shaping LENS . In: Solid

Freeform Fabrication Symposium Proceedings, 1996.w x3 Mazumder J et al. The direct metal deposition of H13 tool

steel for 3-D components. JOM, May 1997.w x4 Lewis GK et al. Properties of near-net shape metallic compo-

nents produced by the directed light fabrication process. In:Solid Freeform Fabrication Symposium Proceedings, 1997.

w x5 Fessler JR et al. Laser deposition of metals for shape deposi-tion manufacturing. In: Solid Freeform Fabrication SymposiumProceedings, 1996.

w x TM6 Arcella FG et al. Materials characterization of lasercasttitanium. Adv Powder Metall Particulate Mater, 1996.

w x7 Das S et al. Direct selective laser sintering of high performancemetals for containerless HIP. Adv Powder Metall ParticulateMater, 1997.

w x8 Das S et al. Direct selective laser sintering and containerlesshot isostatic pressing for high performance metal components.In: Solid Freeform Fabrication Symposium Proceedings, 1997.

w x9 Knight R et al. Metal processing using selective laser sintering

and hot isostatic pressing. In: Solid Freeform Fabrication Sym-posium Proceedings, 1996.

w x10 Zong G et al. Direct selective laser sintering of high tempera-ture materials. In: Solid Freeform Fabrication SymposiumProceedings, 1992.

w x11 Freitag DW et al. Laser directed fabrication of full densitymetal articles using hot isostatic pressing. US Patent 5,640,667,1997.

w x12 German RM. Powder metallurgy science. 1994; Chap 7: 57,MPIF.

w x13 American Society for Testing and Materials, ASTMB. 348-95astandard specification for titanium and titanium alloy bars andbillets, 1995.

w x14 ASM. Metals Handbook, vol. 3. 9 ed. Properties and selection.ASM International, 1993:388.

w x15 American Society for Testing and Materials. ASTM E 1409-97standard test method for determination of oxygen in titaniumand titanium alloys by the inert gas fusion technique, 1997.

w x16 Yolton CF. Crucible research. Personal communication, 1998.