Evaluating measurement techniques: establishing a testing framework for residual stress in selective laser melted Ti-6Al-4V

Anderson, Lucas Steven (2017-12)

Thesis (MEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Residual stress (RS) plays an important role in the mechanical performance of components. Due to the manufacturing process involved in Selective Laser Melting (SLM), high RS is generated within the produced components. These stresses can increase component failure rates either during the manufacturing phase or in service. An investigation was performed into the capabilities of various stress measurement techniques for the application of measuring the RS distribution in SLM produced Ti-6Al-4V (Ti64). This investigation will be used as the basis for creating a testing framework for further studies involving the RS distribution in SLM produced Ti64. The stress measurement techniques were identified and reviewed with respect to the following: stress scale measurable, the stress tensor produced, measurement type, measurement penetration into SLM produced Ti64 and the achievable stress resolution in Ti64. Three techniques were selected for further evaluation, namely: neutron diffraction (ND), X-ray diffraction (XRD) and stress relaxation coupled with Digital Image Correlation (DIC). SLM produced Ti64 specimens built with nine combinations of build layer thickness and exposure strategy were used as test specimens. ND was used to resolve the macro-stress distribution along a plane running through the depth of the tested specimens and XRD was used to measure both near surface stress and, combined with electro-polishing, the stress distribution through individual build layers. The development of a technique – using focused ion beam (FIB) micro-milling and DIC displacement mapping – for the measurement of the residual stress at the layer scale, was also initiated. ND was capable of performing volumetric stress distribution measurements through the full depth of the specimens. Long testing durations and limited accessibility limits its application to RS measurements in SLM produced Ti64. A reduced analysis domain should be used in future testing to allow for more stress orientations to be scanned. The XRD technique, coupled with electro-polishing, was capable of resolving the in-plane stress distribution through individual build layers. The use of the sin2ψ method simplifies the calculation of the stress components. Care should be taken when interpreting the results obtained at the surface as high surface roughness can lead to erroneous stress results. Due to equipment failure the FIB-DIC technique could not be investigated fully. A validation test showed that the technique was able to resolve the in-plane strain components resulting from stress relaxation to a depth of ~20 μm. Further work on this method will include testing on SLM produced Ti64 specimens. The influence of build layer thickness and exposure strategy on RS was also investigated. An increase in the build layer thickness resulted in a decrease in the stress component magnitude and gradient at both the component scale and at the layer scale. The exposure strategy influences the homogeneity of the stress distribution. A uni-directional exposure strategy produces an approximately uni-axial stress distribution at the component scale, whereas the use of two or more laser vector directions results in an approximately bi-axial stress distribution at the component scale. The stress distribution at the layer scale remains uni-axial regardless of the exposure strategy used.

AFRIKAANSE OPSOMMING: Residuele spanning (RS) speel 'n belangrike rol in die meganiese gedrag van komponente. Die vervaardigingsproses van Selektiewe Lasersmelting (SLM) genereer hoë interne RS in komponente. Hierdie belastings kan die voorkoms van komponentfaling tydens die vervaardigingsfase of tydens gebruik verhoog. ‘n Ondersoek is uitgevoer in die doeltreffendheid van verskeie spanningmeetingstegnieke vir RS in SLM vervaardigde Ti-6Al-4V (Ti64). Hierdie ondersoek sal gebruik word as ‘n basis vir die skep van 'n toetsraamwerk vir verdere studies van RS-verspreiding in SLM vervaardigde Ti64 komponente. Die spanningsmetingstegnieke wat geïdentifiseer was is hersien met betrekking tot: die meetbare spanning skaal, die spanningstensor wat gebou kan word, die tipe metings wat uitgevoer is, hul metingspenetrasie in SLM vervaardigde Ti64 en laastens, die haalbare spanningsresolusie in komponente. Drie tegnieke was gekies vir verdere evaluering, naamlik: neutron diffraksie (ND), X-straal diffraksie (XD) en materiaal ontspanning in samewerking met digitale beeldkorrelasie (DBK). SLM vervaardigde Ti64-monsters is met nege kombinasies van laagdikte en blootstellingstrategieë gebou. ND is gebruik om die makrospanningvelde te aniliseer op die vlak wat deur die diepte van die getoetsmonsters loop. XD is gebruik om beide naby-oppervlakspanning te meet en, gekombineer met elektropolering, om die spanningsverdeling van individuele boulae te meet. Die ontwikkeling van 'n tegniek vir die meting van die RS op die skaal van die geboude lae is ondersoek met behulp van gefokusde ioonbundel (FIB) mikrograaf en DBK-verplasingsmetings. Die studie het gevind dat ND in staat was om volumetriese spanningsverdelingsmetings deur die volle diepte van die monsters te neem. Lang toetstye en beperkte toeganklikheid beperk ND se toepassing op RS-metings in SLM vervaardig Ti64. Daar was bevind dat 'n verminderde analise gebied gebruik moet word in toekomstige toetse om voorsiening te maak vir ʼn groter aantal geskandeerde spanningsoriëntasies. Die XD-tegniek, tesame met elektropolering, kon die in-vlak spanningsverdeling verkry deur individuele boulae. Die gebruik van die sin2ψ-metode vereenvoudig die berekening van die spanningskomponente. Voldoende sorg moet geneem word met die interpretering van resultate wat op die oppervlak verkry word omdat hoë oppervlakrofheid kan lei tot foutiewe spanningslesings. As gevolg van toerustingonderbreking, kon die FIB-DBK-tegniek nie volledig ondersoek word nie. 'n Validasietoets het getoon dat hierdie tegniek die spanningskomponente van spanningsverslapping in die boulae tot 'n diepte van~20 μm kon bereken. Verdere werk op hierdie metode sal die toetsing van SLM vervaardig Ti64-monsters insluit. Die invloed van die boulaagdikte en blootstellingstrategie op RS is ook ondersoek. 'n Verhoging van die boulaagdikte het daartoe gelei dat die spanningskomponent se grootte en gradiënt op beide die komponentskaal en die boulaag skaal verminder is. Daar was verder bevind dat die blootstellingstrategie beïnvloed die homogeniteit van die spanningsveld. 'n Enkelrigtingsblootstellingstrategie veroorsaak 'n ongeveer eenassige spanningsverdeling op die komponentskaal, terwyl die gebruik van twee of meer laservektorrigtings 'n ongeveer twee-assige spanningsverdeling op die komponentskaal tot gevolg bring. Die spanningsverdeling op die boulaagskaal bly eenassig ongeag die blootstellingstrategie wat gebruik word

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/102816
This item appears in the following collections: