Atte Antikainen: Modelling Residual Stresses in Metal Alloys during Rapid Solidification in 3D Printing

Modelling Residual Stresses in Metal Alloys during Rapid Solidification in 3D Printing

By Atte Antikainen

Overview

One focus area of my research has been to evaluate the tendency of metals produced by laser powder bed fusion (3D printing) to form residual stresses, which in the worst case can lead to cracking of the component.

Residual stresses formed during 3D printing are a complex phenomenon influenced by many material properties, such as the coefficient of thermal expansion, melting point, strength, toughness, and thermal conductivity.

Research & Innovation

With the support of the Walter Ahlström Foundation, I learned Python programming for solving thermodynamic problems and developed a practical tool for evaluating thermal shrinkage in 3D-printable metals.

The key innovation was to apply modelling originally developed for phase diagram calculations to assess residual stress tendencies. Using the CALPHAD method, temperature-dependent coefficients of thermal expansion can be calculated for different alloys.

However, traditional calculations do not account for chemical inhomogeneity, which is typical in any solidification process, including metal 3D printing. The CALPHAD method can also be used to calculate segregation, i.e. this inhomogeneity.

My approach combines these two functionalities into a single computational framework, enabling the calculation of temperature- and segregation-dependent thermal expansion coefficients for a system solidifying from the melt.

The results can be used both qualitatively and quantitatively to compare the tendency of different materials to form residual stresses during cooling. The lower the residual stresses, the easier the material is to 3D print.

Impact

The support from the Walter Ahlström Foundation played a crucial role in the development of this computational tool, as the original projects where the need was identified had already ended. The incentive grant encouraged the development of the idea into a functional and usable solution.

I have used this computational tool and its components in other publications related to my doctoral research, as well as in material development projects beyond my dissertation.

I have also presented the implementation of this method at a customer event of the software developer, where the solution received positive feedback.

Visuals & Publication

The animation illustrates a layer-by-layer solidifying material, showing the melting point, molar volume (from which the coefficient of thermal expansion can be derived), and phase fractions for each layer.

The image presents the coefficient of thermal expansion calculated from the animation data as a 3D representation as a function of temperature and segregation.

The full article is available here:
https://doi.org/10.3390/met1211189