Dataset of process-structure-property feature relationships for AlSi10Mg material fabricated using laser powder bed fusion additive manufacturing

This dataset reports microstructure and mechanical property features of AlSi10Mg manufactured using laser powder bed fusion over a wide range of processing conditions. Samples were fabricated with different combinations of laser power, scan speed, and hatch spacing to probe dense regimes as well as porous samples resulting from keyholing and lack of fusion. Pore and grain/sub-grain features for each processing set were quantified. Sample porosity was measured using Archimedes density measurements and X-ray computed tomography (XCT). XCT was also used to characterize the surface roughness of samples along with pore size and morphology. Electron backscatter diffraction (EBSD) was used to characterize the grain size and morphology while scanning electron microscope (SEM) imaging and was used to measure solidification cell size. Uniaxial tension tests were performed to ascertain yield and ultimate tensile strengths, elongation, and elastic modulus, and microhardness was measured using Vickers indentation.


a b s t r a c t
This dataset reports microstructure and mechanical property features of AlSi10Mg manufactured using laser powder bed fusion over a wide range of processing conditions.Samples were fabricated with different combinations of laser power, scan speed, and hatch spacing to probe dense regimes as well as porous samples resulting from keyholing and lack of fusion.Pore and grain/sub-grain features for each processing set were quantified.Sample porosity was measured using Archimedes density measurements and X-ray computed tomography (XCT).XCT was also used to characterize the surface roughness of samples along with pore size and morphology.Electron backscatter diffraction (EBSD) was used to characterize the grain size and morphology while scanning electron microscope (SEM) imaging and was used to measure solidification cell size.Uniaxial tension tests were performed to ascertain yield and ultimate tensile strengths, elongation, and elastic modulus, and microhardness was measured using Vickers indentation.

Value of the Data
• This dataset includes microstructural and property measurements that were obtained through comprehensive experimental investigation of the processing-structure-property (PSP) relationships of laser powder bed fusion (PBF-LB) additively manufactured AlSi10Mg fabricated using a wide range of processing conditions.The microstructure was characterized using X-ray computed tomography (XCT), scanning electron microscope (SEM) imaging, and electron backscatter diffraction (EBSD), and mechanical properties were characterized via mechanical testing and hardness measurements.• The data provided in this article expand on previous PSP investigations of PBF-LB AlSi10Mg by investigating a wide range of different processing conditions that have not been published in literature for this material.• The data can be used to support the optimization of PBF-LB AlSi10Mg such that the additively manufactured parts are defect-free with good mechanical properties.• The data can be used to develop, calibrate, and validate data-driven and Integrated Computational Materials Engineering (ICME) models to better predict PSP relationships in PBF-LB AlSi10Mg.

Objective
The objective of this dataset is to provide experimentally quantified PSP feature data for AlSi10Mg fabricated using PBF-LB additive manufacturing (AM), which can be used to develop, calibrate, and validate PSP models for this material system.

Data description
A total of 60 processing sets with different combinations of laser power (60-460 W), scan speed (250-30 0 0 mm/s), and hatch spacing (10 0 and 150 μm) were intentionally used to probe a wide range of processing regimes, including dense, lack of fusion (LoF), and keyholing.Uniaxial tensile samples, in accordance with ASTM E8/E8M [2] , were fabricated using a 3D System ProX DMP 320 PBF-LB AM system machine.As-fabricated samples were direct aged for six hours at 170 °C followed by five hours of furnace cooling in an Argon gas atmosphere.Microstructural characterization, using XCT and SEM imaging, and mechanical testing, though uniaxial tension testing and microhardness testing, were performed to quantify the microstructure and property of the material fabricated using different processing conditions.
The dataset includes eight processing parameters, nine microstructural features, and five mechanical property features, as shown in Table 1 .The processing conditions used in this study are given in Table 2 , and the statistical mean and standard deviation values of each PSP feature, measured for each processing parameter set, are summarized in Tables 3 through 9 .For each PSP feature given in Tables 3 through 9 , the experimental testing or characterization methods and the corresponding table index are listed in Table 1 .
The statistical mean and standard deviation values of all PSP features as a function of processing parameters, as well as the raw experimental data that were used, are available in the Zenodo data repository.A high-level overview of the data can be found as the "AlSi10Mg PSP feature table" Microsoft Excel spreadsheet in the repository.The raw experimental data for all PSP features can be found as individual files of tabular data (for surface roughness, pores, grain/sub-grain structures, and stress-strain behavior) and SEM/EBSD images (for grains and cells).

Process map design
Different combinations of laser power, P , scan speed, v , and hatch spacing, h , were used to probe dense and defect processing regimes.A constant layer thickness, l, of 60 μm was used.Similar to previous work on PBF-LB Ti-6Al-4 V [ 3 , 4 ], a variety of quantitative descriptors were used to define the processing condition, including volumetric energy densities and the melt pool instability factor.The volumetric energy density, VED, and modified volumetric energy density, MVED, were used to estimate the bounds of the LoF regime, which is caused by incomplete fusion of neighboring tracks, and the keyholing regime, which is caused by excessive energy input.Both VED and MVED metrics quantify the energy during fabrication and are defined as: where α and β are the laser absorptivity and thermal diffusivity for the powder, respectively, φ is the laser spot diameter, and C is a material constant.The difference between VED and MVED is that VED assumes that the laser penetration depth is the same as the layer thickness, which is a constant parameter, while MVED considers the penetration depth as a function of laser energy distribution [5] .
In addition to VED and MVED, the melt pool instability factor was used to describe the processing regime for beading-up or balling instability, which is caused by the over-elongation of the melt pool, such that the melt pool breaks up into small droplets due to surface tension.The melt pool instability factor is defined by the aspect ratio of the melt pool, which is related to laser power multiplied by scan speed ( P v ), defined as [6][7][8] : where L and W are the length and width of the melt pool along the fabrication substrate, respectively, is the laser absorptivity, κ and α are the conductivity and thermal diffusivity of the powder, respectively, T is the temperature difference between the material's melting point and room temperature, e is Euler's number, and C is a material constant.
All samples were designed with a dog-bone flat geometry according to ASTM E8/E8M [2] with the vertical build direction parallel to the tensile axis.An outer contour was applied using constant processing parameter (275 W, 800 mm/s) to minimize variations in surface roughness between samples.The interior bulk of each sample was fabricated using different processing parameters, which are given in Table 2 .A 245 °hatch rotation was applied between each subsequent layer.A constant layer thickness of 60 μm was used for all samples.

Internal pore characterization
XCT was used to characterize the internal pore structure of samples using a Phoenix v|tome|x L300 nano/micro-CT (General Electric, Inc., United States) machine with a voltage of 200 kV, current of 50 uA, and a voxel size of 10 μm.Samples were rotated 360 °during imaging to capture a total of 1400 2D images with 1 s of exposure time.The 2D images were then reconstructed to reveal the 3D volume information using software [11] (Phoenix datos|x, Waygate Technologies, Inc., United States).
XCT data were analyzed using image processing software, including Amira-Avizo [12] (v.2020.2,Thermo Fisher Scientific, Inc., United States) and MATLAB [13] (v.R2018b, The Math-Works, Inc., United States).Due to pore detection resolution limits of XCT [14] , internal pores with edge lengths smaller than three times the voxel sizes (30 μm) were removed from analysis.Quantitative features were extracted from XCT data, including size and morphological information of each pore (equivalent diameter and sphericity), surface roughness of the gauge region along the vertical build direction (average and RMS surface roughness), and pore volume fraction (XCT porosity).Archimedes density measurements were also performed to measure the porosity of each sample (Archimedes porosity) in accordance with ASTM B962 [15] as a comparison to XCT porosity.For all quantitative features, the statistical mean and standard deviation values were calculated for each processing parameter set.

Grain/sub-grain characterization
Grain and sub-grain solidification cells were characterized through imaging of the undeformed grip of small hatch spacing samples after tensile testing.The grip region was removed from the fractured sample using a low speed cutting saw, mounted into a 1.25-inch epoxy puck using an electro-hydraulic automatic mounting press (TechPress 3, Allied High Tech Products, Inc., United States), and prepared for imaging using a standard metallurgical grinding and polishing procedure with a final polish using 0.05 μm colloidal silica suspension.EBSD (Apreo 2, ThermoFisher, United States) was used with an accelerating voltage of 10 kV, a step size of 2 μm, and the Aztec software (Oxford Instruments, United Kingdom) to characterize grains from polished surfaces.Samples were then etched using Keller's etchant (1 wt% hydrofluoric acid, 2.5 wt% nitric acid, and 1.5 wt% hydrochloric acid, in de-ionized water) for 30 s to reveal sub-grain solidification cells, which were characterized using SEM (Verios G4 XHR, ThermoFisher, United States) imaging with an accelerating voltage of 10 kV.
EBSD images of grains and SEM images of cells were analyzed using MATLAB [13] to quantify the grain equivalent diameter (according to ASTM E2627 [16] ), grain circularity, grain eccentricity, and cell equivalent diameter.

Mechanical testing
Uniaxial tensile tests and Vickers microhardness measurements were performed to measure the mechanical properties of samples.Prior to tensile testing, specimens were painted with black and white speckle patterns to create trackable features for digital image correlation (DIC) to measure fields during tensile loading.An electromechanical load frame (Criterion Model 45, MTS Systems, Inc., United States) with maximum load of 10 kN was used to apply a quasi-static strain rate of 3 × 10 −4 /s.The gauge region of the samples was imaged using a digital camera (GRAS-50S5M-C, Point Grey Research, Inc., British Columbia) at a rate of 1 Hz using imaging software (VIC-Snap 9, Correlated Solutions, Inc., United States).The surface deformation and strain field were calculated using a DIC analysis software (VIC-2D 6, Correlated Solutions, Inc., United States).A cubic B spline interpolation algorithm, with a subset size of 29 pixels and a step size of 7 pixels, where the pixel size ranged from 15.7 to 25.9 μm, and a 24 mm vertical virtual extensometer were used to measure the axial strain.The ultimate tensile strength, yield strength, elongation to fracture, and elastic modulus were extracted from the stress-strain response of each sample.Three duplicate samples were tested for each processing parameter set to obtain statistical mean and standard deviation values of the mechanical properties.Vickers microhardness measurements (LM 110AT, LECO, Inc., United States) were performed according to ASTM E384 [17] .A total of 10 indents were made at dense locations within each sample using a load of 100 gf and a dwell time of 10 s.

Limitations
Flat tensile samples, all aligned with the vertical build direction, with constant thickness were studied to avoid additional impacts from sample geometry on part microstructure or mechanical properties.The characterization of grain/sub-grain features was only conducted on small hatch spacing samples.While 180 tensile samples were fabricated (3 duplicates for each processing parameter set), two samples failed during installation in the mechanical test frame, so mechanical test results from only 178 samples are included.

Ethics Statements
No human subjects, animal experiments and data collected from social media platforms were involved in this work.

Data Availability
Processing, microstructure, and mechanical property dataset for AlSi10Mg fabricated by laser powder bed fusion additive manufacturing (Original data) (Zenodo)

Table 1
Summary of processing, microstructural, and mechanical property features.

Table 2
Processing conditions used to fabricate samples.A constant layer thickness of 60 μm was used to fabricate all samples.

Table 3
Porosity calculated using X-ray computed tomography (XCT) and Archimedes density measurements.

Table 4
Pore equivalent diameter and sphericity measured using X-ray computed tomography.The numerical mean (mean) and volume weighted mean (Vol-mean) are presented.

Table 5
Arithmetic average (Ra) and root mean square average (RMS) surface roughness measured using X-ray computed tomography.

Table 6
Grain equivalent diameter, circularity, and eccentricity measured from electron backscatter diffraction data for small hatch spacing samples.The numerical mean (mean) and volume weighted mean (Vol-mean) are presented.

Table 7
Sub-grain cell equivalent diameter measured from scanning electron microscope images for small hatch spacing samples.

Table 9
Yield strength, ultimate tensile strength, elongation to fracture, and elastic modulus measured using uniaxial tension testing.