Rheo-mechanics modelling of 3D concrete printing constructability

Date
2019-12
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: Three industrial revolutions have occurred in the last 250 years that resulted in increased productivity for most economic sectors. This generally improved the standard of living for all individuals. Recent research has however found that the productivity of the USA’s construction sector regressed at 0.6 % per year over the last 55 years. This indicates that the construction industry has not yet experienced the benefits of industrialisation. The current industrial revolution, commonly referred to as Industry 4.0, presents technology such as 3D printing. Applying this technology in the construction sector yields 3D printing of concrete (3DPC), or digital construction, that holds tremendous potential for enhanced productivity. Early estimates indicate possible construction time savings of 50 % and waste savings of 30 %. Additionally, the realisation of geometrically-complex elements is possible without the need for formwork. Although a promising technology, it remains in the early stages of commercialisation and presents many challenges before mass adoption thereof. No material characterisation test currently exists that specifically appertains to 3DPC. Fresh state mechanical tests are mostly performed; however, they provide insignificant information on the appropriateness of a material for 3DPC. This process requires a material to be easily transported via pumping, but then also to possess sufficient strength after extrusion to support the weight of subsequent filament layers. The latter is commonly referred to as buildability in 3DPC terminology. Furthermore, limited constructability design guidelines are currently available; consequently, elements are typically printed at randomly chosen speeds and filament layer heights with the hope of a successful outcome i.e. the element does not collapse whilst being printed. The main aim of this research is thus to develop practicable analytical models based on rheological material properties that collectively contribute towards constructability design guidelines for 3DPC. This research initially presents the design and manufacture of an industrial-grade gantry type 3D concrete printer with a build volume of roughly 1 m3. The development of a high-performance, 3D printable, thixotropic concrete via the Fuller Thompson theory is explained. Progression and mastering of basic 3DPC technology is demonstrated by means of pictures from initial 3D prints conducted at Stellenbosch University to the latest X-project. Thereafter, a bi-linear thixotropy model that specifically appertains to 3DPC is developed. A current thixotropy model that accounts for structuration (Athix) is extended to account for re-flocculation after agitation (Rthix), which is a physical process. The material characterisation is solely conducted by the use of a rheometer. In addition to presenting insight into a material’s thixotropy behaviour, the bi-linear static yield shear stress evolution curve depicts the strength gain of a material after it has been extruded from the nozzle. The following chapter presents the development of a filament shape retention model that is based on the bi-linear thixotropy model. The model is simple and practicable, while numerical validation via finite element analysis depicts the conservatism of the model. Thereafter, an analytical buildability model is developed which is also based on the bi-linear thixotropy model. The model only accounts for material failure in the form of plastic yielding. In addition, the model accounts for various filament aspect ratios, which influence the apparent compressive strength of a material. The model is verified by experimental testing, yielding an 8.33 % under prediction of the total number of filament layers before failure. Finally, the three models are combined to yield a constructability design model for 3DPC. The model predicts the optimum print parameter combination, i.e. filament layer height and print speed, that successfully yields the entire specified print object in the least amount of time. In addition, a statistical design model is presented by incorporating material partial factors to reduce the model’s probability of failure to 10 %. In summary, this research contributes a statistically safe and time-optimised constructability design model for the 3DPC industry to facilitate commercialisation of the technology.
AFRIKAANSE OPSOMMING: Drie industriële rewolusies het plaasgevind in die afgelope 250 jaar wat gelei het tot verhoogde produktiwiteit vir die meeste ekonomiese sektore. Dit het die lewenstandaard vir alle indiwidue oor die algemeen verbeter. Onlangse navorsing het egter bevind dat die produktiwiteit van die VSA se konstruksiesektor die afgelope 55 jaar teen 0,6 % per jaar gedaal het. Dit dui daarop dat die konstruksiebedryf nog nie die voordele van industrialisasie ervaar het nie. Die huidige industriële rewolusie, algemeen bekend as “Industry 4.0”, bied tegnologie soos 3D-drukwerk. Die toepassing van hierdie tegnologie in die konstruksie sektor lewer 3D drukwerk van beton (3DDB), of digitale konstruksie, wat enorme potensiaal inhou vir verhoogde produktiwiteit. Vroeë skattings dui op moontlike konstruksietydbesparings van 50 % en afvalbesparings van 30 %. Daarbenewens is die realisering van geometriese-komplekse elemente moontlik sonder enige bekisting vereistes. Alhoewel ʼn belowende tegnologie, bly dit in die vroeë stadiums van kommersialisering en bied dit vele uitdagings voor massa-aanvaarding daarvan. Daar bestaan tans geen wesenlike materiaal karakteriseringstoets wat spesifiek op 3DDB betrekking het nie. Varstoestand meganiese toetse word meestal uitgevoer, alhoewel dit egter onbeduidende inligting oor die toepaslikheid van ʼn materiaal vir 3DDB bied. Hierdie proses vereis dat ʼn materiaal maklik vervoer moet kan word deur dit te pomp, maar dan ook voldoende sterkte besit na ekstrusie om die gewig van die daaropvolgende filamentlae te ondersteun. Laasgenoemde word algemeen na verwys as boubaarheid in 3DDB-terminologie. Verder bestaan daar beperkte ontwerpsriglyne vir oprigbaarheid; gevolglik word elemente tipies gedruk met lukraak gekose spoed en filament laaghoogtes met die hoop op ʼn suksesvolle uitkoms, d.w.s. die element tuimel nie inmekaar terwyl dit gedruk word nie. Die hoofdoel van hierdie navorsing is dus om uitvoerbare, analitiese modelle te ontwikkel gebaseer op reologiese materiaaleienskappe wat gesamentlik bydra tot oprigbaarheid ontwerpsriglyne vir 3DDB. Hierdie navorsing bied aanvanklik die ontwerp en vervaardiging van ʼn industriële graad “gantry”-tipe 3D beton drukker met ʼn bou volume van ongeveer 1 m3. Die ontwikkeling van ʼn hoëprestasie, 3D-drukbare, thiksotropiese beton met behulp van die Fuller Thompson-teorie word verduidelik. Vordering en bemeestering van basiese 3DDB-tegnologie word getoon deur middel van foto’s van aanvanklike 3D-drukke wat by die Universiteit Stellenbosch uitgevoer is tot die nuutste X-projek. Daarna word ʼn bi-lineêre thiksotropie model ontwikkel wat spesifiek op 3DDB betrekking het. ʼn Bestaande thiksotropie model wat voorsiening maak vir strukturering ( Athix) word uitgebrei om ook her-flokkulasie na steuring (Rthix) in ag te neem, wat ʼn fisiese proses is. Die materiaal karakterisering word uitsluitlik uitgevoer deur die gebruik van ʼn reometer. Benewens die aanduiding van ʼn materiaal se thiksotropie gedrag, toon die bi-lineêre statiese swigskuifspanning ewolusie kurwe die sterkte toename van ʼn materiaal nadat dit uit ʼn spuitstuk geëkstrueer is. Die volgende hoofstuk bied die ontwikkeling van ʼn filament vormhoubaarheidsmodel aan wat gebaseer is op die bi-lineêre thiksotropie model. Die model is eenvoudig en uitvoerbaar, terwyl numeriese validering via eindige elementontleding die konserwatisme van die model uitbeeld. Daarna word ʼn analitiese boubaarheidsmodel ontwikkel wat ook gebaseer is op die bi-lineêre thiksotropie model. Die model maak slegs voorsiening vir materiaal faling in die vorm van plastiese swig. Daarbenewens maak die model ook voorsiening vir verskeie filament-aspekverhoudings, wat die oënskynlike druksterkte van ʼn materiaal beïnvloed. Die model word geverifieer deur eksperimentele toetsing, wat ʼn 8.33 % onder-voorspelling van die totale aantal filamentlae voor faling lewer. Ten slotte word die drie modelle gekombineer om ʼn oprigbaarheidsmodel vir 3DDB te lewer. Die model voorspel die optimale drukparameter kombinasie, d.w.s. filament laaghoogte en drukspoed, wat die hele gespesifiseerde drukobjek in die minste tyd suksesvol sal oplewer. Daarbenewens word ʼn statistiese ontwerpsmodel voorgelê deur parsiële materiaalfaktore in te sluit om die model se falingswaarskynlikheid tot 10 % te verminder. Ter opsomming, dra hierdie navorsing ʼn statisties veilige en tyd-geöptimiseerde oprigbaarheidsmodel tot die 3DDB-industrie by om kommersialisering van dié tegnologie te fasiliteer.
Description
Thesis (PhD)--Stellenbosch University, 2019.
Keywords
Three-dimensional printing -- Concrete, Rheology, Mechanics of materials, Analytical modelling, Construction technology, UCTD
Citation