Mechanical Testing of 3D Metal Printed Stainless Steel Specimens

Additive manufacturing (AM) is a cutting-edge production method, which has come a long way since its first introduction in the ’80s. Back in the days its usage was very limited to stereolithography, and was only able to make weak structures, so it only worked for visualization. Four decades later it is one of the leading research fields in production areas, because of its flexibility and its ability to make almost any complex geometry. However, no matter how powerful it is, it is not omnipotent, there are certain size and shape restrictions even this method must apply to.


Introduction
Technological advances throughout the years led to various kinds of 3D printing processes. Some of them use wire or fibre as a base material, while some of them use powder, which can be polymer, ceramic, or even metallic based, or simply liquid. Some of the processes are the following [1]: Selecting the right 3D printing process is important and may depend on our needs. If our desired product is only for visualisation or prototyping, we should choose a method which works faster, but having poor mechanical properties. If the product must withstand a certain amount of force, in a working condition, we may choose a slower, but more solid-building method. Each additive manufacturing method has its own unique characteristics, but in their essence all of them are similar and that similarity lies within the layer-by-layer addition, which the process builds up the part from the building plate.
There are a lot of studies, which focused on the printing properties of different AM methods. They usually conduct measurements towards layer thickness and building orientation [2][3] [4]. The overall assessment is, that if we reduce the angle of the manufactured part, making it horizontally oriented [5], we can achieve ideal result, at least with thin-walled specimens. Round shapes are a bit trickier, but they share some of this beforementioned quality [6]. Round shapes however tend to build up less residual stress during their printing process, thus have better shape retention, while flat 3D printed pieces tend to warp, or in some extreme cases break, after or even during the process [7]. We wanted to experiment on these properties, while optimising our printing job for mechanical strength, printing time and warpage as well.

Materials and methods
To conclude our investigations, we selected Selective Laser Melting (SLM) as a manufacturing method.  The base material was a 17-4ph stainless steel powder bed, because of its excellent 3D printability and mechanical properties. The chemical composition is shown on  The powder was made using gas atomization technique, to ensure the spherical morphology, which is optimal for 3D printing. The reason for this, that the powder particles should "roll over" each other, when the feeding mechanism spreads the new layer. If there is an imperfect or checkered particle, it may stick on the feeder and create a drag line. On this drag line will be void, thus an imperfect layer will be introduced to the process.

Process parameters
The printing machine used was an Orlas Creator, SLM machine, with nitrogen shield gas. The process parameters are shown in Table II

Mechanical testing
The most efficient and cost-effective way to test 3D metal printed specimens, is to make tensile specimens. Its relatively fast to print, easily reproduceable, and the testing process is simple and fast, although require expensive hardware. The first batch of test specimens was made according to the ASTM E8 standard, however due to limited size, we had to differ from the given values. The results from a similar printing job are shown on Fig.: 2. The specimens were printed in a semi-vertical / semi-horizontal position, as Fig.:2 shows. In this way, the tensile testing was done in a parallel direction to the layers, thus granting us the most favourable results, while having an optimized printing job for printing time and warpage. If the one-dimensional pull force from tensile testing acts towards perpendicular to the layer setup, we will get weaker results [9].
During our experiments, we wanted to compare different thicknesses, thus altered the base tensile test specimens. 2 mm was set as basic, and there were 1 mm and 3 mm thick ones as well. Our goal here was to demonstrate, how these geometrical changes will affect the finished product in a simple testing environment. The tensile test results can be shown from Fig.: 3 to 5. The results, show that various thicknesses will produce different mechanical properties. The weakest values came out from the 1 mm thick specimens, mainly because these thin layers tend to build up huge amounts of residual stresses. This is another issue which occurs during 3D printing of metals, but it is for another study.
2 mm thick specimens had the most strain percentage values, while increasing the thickness to 3 mm lead to almost the same height in tensile strength, but with a more rigid breaking. According to this, we can assume that there is a lower and higher limit, where we can vary these values freely, without sacrificing much of mechanical strengths.

Horizontally modified tensile test specimens
The next step in our study was to make further modifications on the specimens. As we could see, printing 1 mm thick specimens yields in mostly poor results, we only carried on with 2 and 3 mm-s. The concept of this measurement is shown on Fig.: 6.  As the results show, in 0° orientation still 2 mm thick specimens give us the best tensile strength to strain values, however the tensile strength values have dropped with the change of orientation. What we also see, is that changing the specimens' layers total surface area, we will have significant changes in these properties. This way we can also assume that on these dimensions there must be a lower and higher limit, which we can change the values freely.

Summary
During our measurements, we can state that tensile testing of 3D metal printed parts is highly dependent on orientation and size, while there is a certain value in geometrical ratio which we do not go below or exceed to get the desired mechanical properties. This value is also depending on the residual stresses, which the process must deal with. Different types of specimens, however, require different handling, thus we need to be careful when comparing different printing orientations.