DTU Mechanical Engineering has received EUR 1.1 million from Innovation Fund Denmark for a new project to identify how a 3D-printed part will behave during production and when it is subsequently subjected to load during operation. Professor Jesper Hattel is coordinating the department’s efforts.
Digital tools help control production
As part of the project, the Danish Technological Institute is establishing an entire production line for 3D metal printing, while DTU Mechanical Engineering is responsible for developing what is called the digital twin, i.e. the digital representation of the product, and how it behaves during production and afterwards in operation.
The project will give Danish industry access to knowledge and know-how about the production of 3D metal printing, but more than that, it will also enable industry to adapt the digital models that DTU researchers are developing for many other types of production. So the perspectives for the manufacturing industry actually go far beyond 3D printing in metal.
“The philosophy behind the digital twinning of products and systems is a very important part of Industry 4.0,” says Professor Jesper Hattel.
“It is a fundamental and general concept that we are able to establish digital twins for products and systems, and thereby understand how the products are manufactured and behave when in operation in order to ultimately improve production and the products’ properties and performance.”
The digital models need to be fed with data gathered from, for example, sensors, which during production measure various process parameters such as temperatures etc. Scanners will also gather accurate data about the 3D-printed part’s dimensions, and thus whether it lives up to the geometrical tolerances which have been specified beforehand for the parts.
“If during the production process you can scan your way to establishing that the part will bend, then you can feed this information back to the computer model, which will then say something about what needs to be done to avoid it,” says Jesper Hattel.
At the end of the day, it means that you can adapt and adjust production the moment errors occur and the product deviates from the requirements which have been set in advance.
3D metal printing in production and operation
“The existing computer simulation models of how 3D-printed metal parts behave in production and subsequently are still incomplete, but with this project, we will have the opportunity to arrive at an entirely new understanding of how the entire process chain affects the properties of the parts, and how it needs to be described numerically—and this requires the competencies of all three sections at the department,” says Professor Hattel.
The project will focus specifically on the form of 3D metal printing known as selective laser melting (SLM), in which layer upon layer of metal powder is distributed on a metal substrate before being laser welded from above to produce the finished part. This is the same gradual build-up of material known from 3D printing in, for example, plastic, which means that you can build parts with complex internal structures.
The process of 3D printing, or additive manufacturing, is in itself digital in nature, because the actual printing process is completely computer-controlled from a CAD drawing in 3D with a complex plan for how the laser fuses the metal powder, the so-called scanning strategy.
The part is printed on a metal substrate, to which it is secured with small supports that cause thermal stresses which are then released in the form of deformations when the part is cut free.
“What we see here,” says Professor Hattel, “is the deformation of the part due to thermal variations during the printing process, and this is one of the biggest problems with 3D metal printing. However, we can compensate for it by creating a new scanning strategy which takes this information into account. We’ve produced a computer simulation of how the process unfolds, and using this, we can then suggest another scanning strategy so that the thermal field becomes more homogeneous. Thus, by means of computer simulations, we can predict the scanning strategy so we don’t suffer excessive thermal deformations, but so far only for simple parts like that shown.”
The new models which the research arrives at will ultimately end up as simulation models which can be used generally in production to predict how the 3D-printed metal parts will behave under different conditions during manufacture and afterwards in operation.