Metal additive manufacturing (AM) represents a true revolution in engineering, offering unprecedented design flexibility. In contrast to conventionally manufactured metal parts, AM processes introduce inherent defects in the printed alloys that cause a reduction of fatigue life and a large scatter in the fatigue properties. In the absence of large internal defects, most of the fatigue failures of the metal AM parts are driven by the rough surface. This study aims to fill existing gaps in the literature by addressing the lack of specialized models for the fatigue behaviour of metal AM surfaces in the as-built state (without any surface treatment applied after printing). Experimental and numerical campaigns were conducted on different alloys (AlSi10Mg, Ti6Al4V, and Co-Cr-Mo), providing comprehensive results and validating computational models to predict the fatigue limit and service life of as-built parts manufactured by laser-powder bed fusion (L-PBF) technology. To efficiently evaluate the surface quality, new 4-point bending (4PB) test setups and sample geometries are introduced. It is found that the orientation of the samples during printing greatly affects the surface quality, thus affecting the fatigue strength. In the context of Fracture Mechanics (FM), guidelines are provided for measuring critical anomalies and applying shielding models, considering the interaction of short cracks initiated from deep valleys of the surface. After validation of FM models, the focus shifts to prediction based on roughness parameters measured on 2D profiles extracted from surfaces measured by areal methods. Several measurement methods are compared to determine the most accurate. Extreme value statistics (EVS) of selected roughness parameters was employed to estimate an equivalent size of anomalies. As an alternative to FM, a point method based on the Theory of Critical Distances (TCD) is proposed, using finite element (FE) analysis of 2D models with true rough profiles measured by X-ray computer tomography (XCT) or optical areal microscopy. The strengths and weaknesses of the two approaches are explored, considering factors such as computational effort and prediction accuracy. In conclusion, this research stands out as pioneering in the development of predictive tools, offering an accurate methodology for estimating the strength and fatigue life of as-built AM parts.
La fabbricazione additiva dei metalli (AM) rappresenta una vera rivoluzione nell’ambito ingegneristico, offrendo una flessibilità di progettazione senza precedenti. Diversamente dalle tecnologie di produzione convenzionali, i processi additivi introducono difetti nel materiale che riducono la vita a fatica e sono causa di dispersione delle proprietà. In assenza di grandi difetti interni, le rotture a fatica delle parti AM metalliche sono causate per lo più dalla superficie rugosa. Questo studio si propone di colmare le lacune esistenti nella letteratura, affrontando la mancanza di modelli specializzati per le superfici nello stato as-built (senza trattamenti superficiali applicati dopo la stampa). Sono state condotte campagne sperimentali e numeriche su diverse leghe (AlSi10Mg, Ti6Al4V e Co-Cr-Mo), fornendo risultati approfonditi e convalidando i modelli computazionali per prevedere il limite di fatica e la vita utile delle parti as-built prodotte con tecnologia laser-powder bed fusion (L-PBF). Per valutare efficientemente la qualità della superficie, vengono introdotti nuovi setup di test 4-point bending (4PB) e nuove geometrie di campioni. È emerso che l’orientamento dei campioni durante la stampa influisce notevolmente sulla qualità della superficie, quindi incidendo sulla resistenza alla fatica. Nel contesto della Meccanica della Frattura (FM), vengono fornite linee guida per misurare anomalie critiche e applicare modelli di shielding, considerando l’interazione di cricche corte iniziate dalle valli profonde della superficie. Dopo la convalida dei modelli FM, l’attenzione si sposta sulla previsione basata su parametri di rugosità misurati su profili 2D estratti da superfici misurate con metodi areali. Diversi metodi di misura vengono confrontati per determinare il più accurato. Le statistiche dei valori estremi (EVS) di alcuni parametri di rugosità aiutano a stimare una dimensione equivalente delle anomalie. Come alternativa alla FM, viene proposto un metodo puntuale basato sulla Teoria della Distanze Critiche (TCD), che utilizza analisi a elementi finiti di modelli 2D con veri profili rugosi misurati con tomografia computerizzata a raggi X (XCT) o microscopi ottici areali. Vengono esplorati i punti di forza e di debolezza dei due approcci, considerando fattori come lo sforzo computazionale e l’accuratezza delle previsioni. In conclusione, questa ricerca si distingue come pionieristica nello sviluppo di strumenti predittivi, offrendo una metodologia precisa per stimare la resistenza e la vita a fatica di parti AM as-built.
Analysis of metal AM rough surfaces: impact on fatigue properties
BARRICELLI, LUCIA
2023/2024
Abstract
Metal additive manufacturing (AM) represents a true revolution in engineering, offering unprecedented design flexibility. In contrast to conventionally manufactured metal parts, AM processes introduce inherent defects in the printed alloys that cause a reduction of fatigue life and a large scatter in the fatigue properties. In the absence of large internal defects, most of the fatigue failures of the metal AM parts are driven by the rough surface. This study aims to fill existing gaps in the literature by addressing the lack of specialized models for the fatigue behaviour of metal AM surfaces in the as-built state (without any surface treatment applied after printing). Experimental and numerical campaigns were conducted on different alloys (AlSi10Mg, Ti6Al4V, and Co-Cr-Mo), providing comprehensive results and validating computational models to predict the fatigue limit and service life of as-built parts manufactured by laser-powder bed fusion (L-PBF) technology. To efficiently evaluate the surface quality, new 4-point bending (4PB) test setups and sample geometries are introduced. It is found that the orientation of the samples during printing greatly affects the surface quality, thus affecting the fatigue strength. In the context of Fracture Mechanics (FM), guidelines are provided for measuring critical anomalies and applying shielding models, considering the interaction of short cracks initiated from deep valleys of the surface. After validation of FM models, the focus shifts to prediction based on roughness parameters measured on 2D profiles extracted from surfaces measured by areal methods. Several measurement methods are compared to determine the most accurate. Extreme value statistics (EVS) of selected roughness parameters was employed to estimate an equivalent size of anomalies. As an alternative to FM, a point method based on the Theory of Critical Distances (TCD) is proposed, using finite element (FE) analysis of 2D models with true rough profiles measured by X-ray computer tomography (XCT) or optical areal microscopy. The strengths and weaknesses of the two approaches are explored, considering factors such as computational effort and prediction accuracy. In conclusion, this research stands out as pioneering in the development of predictive tools, offering an accurate methodology for estimating the strength and fatigue life of as-built AM parts.| File | Dimensione | Formato | |
|---|---|---|---|
|
Thesis_Barricelli.pdf
accessibile in internet per tutti
Descrizione: PhD thesis Lucia Barricelli
Dimensione
48.18 MB
Formato
Adobe PDF
|
48.18 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/216038