There is More to a Pore: 3D Analysis for Additive Manufacturing

– Moving Beyond the Relative Density Metric –

Design and manufacturing teams producing metal parts by additive manufacturing need to consider the influence of porosity on the performance of their parts in service. In some cases, the presence of porosity is not critical to performance or is intentionally designed into the part to perform a specific function. In load-bearing applications, the stressed members should approach full density to mitigate part failure during service.

The usual Archimedes method is a low cost and straightforward approach for measuring the bulk relative density of a printed part. However, one drawback is the lack of information obtained regarding the size, shape, and spatial arrangement of pores that make up the sample’s total porosity. Such information is useful to diagnose the root causes of void formation during the printing process or to assess the influence of undesired porosity on the mechanical properties of the part.

The parts described in this document were printed and scanned as part of ongoing studies [1,2] at the Multi-Scale Additive Manufacturing (MSAM) Lab on campus at the University of Waterloo. The MSAM lab is a pioneer in making microscale computed tomography (μCT) analysis a central tool in additive manufacturing development, and Expanse has been privileged to contribute our CT expertise to their projects.

Multi-scale Additive Manufacturing Lab
Inconel 625 Test Specimen Additive Manufacturing

3D rendering of an Inconel 625 test specimen, printed using a laser powder bed fusion process at MSAM (left). Transparent renderings are shown at (middle) and (right). The color bar indicates the volume of each pore in cubic micrometers. The specimen has a diameter of 5 mm and a height of 7.5 mm.

Microscale X-ray computed tomography (μCT) is a powerful tool to extract information on pore size, shape, and spatial arrangement in additively manufactured parts. A µCT scan produces a 3D dataset in which the full volume of the part, or a specific region of the part, is represented as discrete volume elements (voxels). The intensity value for each voxel describes the local X-ray absorption. In this post, we discuss methods to visualize and characterize porosity to extract information useful to those designing and developing new materials, parts, and processes for additive manufacturing.

Process Parameter Mapping Graphic

Minimum intensity projections onto XY and YZ planes (top left and right) showing regions of lower density (pores). This test specimen was printed in Inconel 625 at MSAM using a laser powder bed fusion process. The relative density of the specimen is plotted along the build direction (bottom right) and the cross section having the lowest relative density value (highest porosity) is shown (bottom left) for the location marked with an x.

Finding Patterns in the Pore Structure

Often, characterization of porosity, pore size, and pore shape is performed by sectioning a printed part and imaging the polished 2D cross section in an optical or electron microscope. While metallographic methods are well established and fairly straightforward, the analysis is confined to a single plane. The primary disadvantage is that volumetric information and patterns in the features of interest that extend outside the image plane are not accessible.

A complete set of volumetric information about the printed part can be obtained using μCT. The 3D pore structure can be visualized quickly and intuitively from the raw data in a volume rendering that accentuates regions of lower density. This method has the advantage of clearly distinguishing between specimens that have the same bulk relative density values, but which have stark differences in pore size and their spatial arrangement. The utility of the volume rendering is that it allows the engineer to quickly identify the different classes of porosity that exist in the as-printed part.

Special Recipe Nickel-Iron Alloy DMLS

Volumetric images highlighting the pore structure are shown for two different specimens printed with different print settings. Relative densities are measured through μCT at 99.8% (Specimen 1) and 99.9% (Specimen 2). These test specimens were printed in a special recipe nickel–iron alloy, using a laser powder bed fusion process at MSAM.

Each class of porosity has a different formation mechanism associated with one or more parameters of the additive manufacturing process. See this article by Marc Saunders for more information on linking process parameters to the different classes of porosity that can form under different operating conditions. The information available in a volume rendering can be extended to quantify the different classes of porosity. An example is to compare spatial arrangement of porosity by plotting the pore volume against the distance from each pore voxel to the outer surface of the part. This provides a metric to understand the relative contributions of porosity located in the skin of the part compared to porosity that is concentrated within the interior of the part. After refining the applicable process parameters, the effectiveness of the process change can be measured in terms of the reduction in pore volume for each porosity class.

Nickel-Iron Alloy Pore Volume Histogram

Pore volume as a function of distance from part surface for two different specimens printed with different print settings. Relative densities are measured through μCT at 99.8% and 99.9% for Specimen 1 and Specimen 2, respectively. These test specimens were printed in a special recipe nickel–iron alloy, using a laser powder bed fusion process at MSAM.

Characterizing Pore Size and Shape

The size and shape of individual pores can be described in many ways, as best suits the needs of the application. A key advantage of μCT analysis is that many different types of measurements can be obtained from a single data set. Example metrics to describe the size of pores, without fitting primitive shapes, include pore volume and Feret diameter (caliper length). Alternatively, primitive shapes such as ellipsoids may be fitted to each pore and the major and minor axis lengths, among other properties, can be summarized. To visualize this information, 3D renders of the pores can be color mapped using the metric of your choosing.

Example metrics for pore morphology include the ratio of volume to surface area, sphericity, and aspect ratio. A description of pore shape is significant in terms of the stress concentration factor that exists at the pore location. Initial investigation into electron beam melting additively manufactured parts has suggested that pores located near the part surface, and which have a higher aspect ratio, are more likely to be a crack initiating defect.

3D rendering colored by pore volume

Small region of interest from an Inconel 625 sample extracted from the approximate location enclosed in the red box at (left). The pores shown at (right) are color mapped to their volume in cubic micrometers.

Inconel 625 3D sphericity histogram

Sphericicy of pores in Inconel 625 sample

Special Application to Parts with Functional Porosity

Cellular structures and porous metal filters are two practical examples where the internal void structure is both complex and critical to performance. A method to create internal open or closed cells was recently developed at the University of Waterloo’s MSAM Lab using a modified binder jetting process [2]. A set of porous titanium specimens were printed as part of a development project to create cellular structures with graded pores arranged in a controlled manner. To provide a full volumetric analysis, µCT was required to identify and describe the location-specific pore size distributions, as well as the geometric characteristics of the cells embedded within the part.

With µCT, the pore structure was characterized in terms of its mean and maximum pore diameters, and clear visualizations to describe how these parameters vary throughout the specimen were created. This information can be used to compare the design and as-printed pore size gradients. In addition, the pore network information extracted from the porous structure can be used for fluid transport studies to predict performance characteristics of the as-printed pore structure.

Binder Jetted Ti Controlled Porosity Specimen

3D rendering of a porous metal test specimen, printed using a modified binder jetting process at the MSAM laboratory. 

Binder Jetted Ti Controlled Porosity Specimen Max Pore Diameter

3D rendering of the internal pore structure, with individual pores color mapped by their max-projected pore diameter. This visualization provides a clear view of the 3D spatial arrangement of pores within the part in a way that can only be obtained through the use of X-ray computed tomography.

Mean Pore Diameter Binder Jet Titanium

Analysis of the mean pore diameter in the porous titanium test specimen. Here, mean pore diameter is calculated as a function of individual print layers, highlighting the ability of X-ray computed tomography to inform development of novel materials with functional porosity..

Putting it all Together

What questions would you ask if you had access to nearly limitless information about your printed part? Any or all of the above visualization and quantification techniques can be used to accelerate the development of your new additive manufacturing materials, parts, and processes. To manufacture functional parts, µCT can be used to shed light on what pore sizes, shapes, and arrangements make up the bulk porosity value. This information can be used to guide process parameter selection including laser power and scanning velocity, as examples in laser powder bed fusion processes, to mitigate the severity of each class of porosity encountered.

Going Further

Expanse Microtechnologies Inc. offers a full suite of porosity analysis techniques to meet your development needs.

Contact Craig Metcalfe at craig.metcalfe@expansemicro.com to learn more about bringing µCT into your own process parameter mapping efforts.

References

  1. Contact Mihaela Vlasea to learn more about this study.
  2. Esmat Sheydaeian (2017). Systems, materials, and methodologies for multi-material hybrid additive manufacturing functionally graded structures. UWSpace. http://hdl.handle.net/10012/12556
2018-07-03T20:00:15+00:00
Bitnami