Automated RFID-Integration in Additive Manufactoring
Integrating Sensors into Components Manufactured by Metal Powder Bed Fusion
Research by Maximilian Binder, Christian Dirnhofer, Philipp Kindermann, Max Horn, Matthias Schmitt, Christine Anstaett, Georg Schlick, Christian Seidel and Gunter Reinhart at Fraunhofer IGCV has led to the development of the first RFID-Ring prototype based on automated RFID-integration in additive manufacturing.
In the interaction of structural weakness and measurement quality, there is a break even point at which a switch from conventional production to production using LPBF is beneficial.
The in-situ integration of sensors into additive manufactured (AM) components is becoming increasingly important and moving into the focus of current research. Manual sensor integration during the metal- and laser-based powder bed fusion (LPBF) process is usually associated with long process downtimes, the loss of the shielding gas atmosphere and a change in component temperature.
The resulting effects are e.g. changed mechanical properties. Therefore, one way to avoid these effects is presented by automated sensor integration. A procedure for the design of kinematics to meet the requirements is presented and experimentally validated. As a result, a functional automation unit in an LPBF plant is presented.
Current fast-growing digital trends, such as condition monitoring and predictive maintenance, increase the pressure on the manufacturing industry to be able to produce components whose condition is monitored by sensors. In this way, maintenance intervals can be minimized and components proactively replaced even before their technical failure. This development involves the need to increase sensor integration for self-monitoring, as well as new requirements for the sensors to be positioned closer to the measured variable, and thus necessitates new solutions.
Additive manufacturing processes, such as the widely used metal-based LPBF process, offer the possibility to integrate sensors at any point in the component and allow cavity designs corresponding to the sensor geometry. These advantages are seen not only by research groups but also by representatives from industry, who expect extensive future developments and potential for AM.
LPBF sensor integration is particularly advantageous for complex component geometries with measuring points that are difficult to access. Fig. 1 shows the theory that, with the help of LPBF, sensors can be better placed at their desired measuring points than with conventional methods and thus the measuring quality of the sensor must increase.
On the one hand, there are, for example, adhesive strain gauges that do not cause a structural weakening of a component, but also do not achieve a very high measurement quality. On the other hand, complex structures exist where conventional sensor integration requires either high structural weakening or poor measurement quality.
Figure 1 therefore shows the interaction of structural weakness and measurement quality. According to this qualitative chart, there is a breakeven point at which a switch from conventional production to production using LPBF is beneficial. Currently, three different approaches are known as a way of integrating sensors during the LPBF process: The manual insertion of sensors, automated insertion or direct manufacturing of sensors, for example implemented via a multi-material mechanism. However, the maturity levels of these approaches still differ widely (Table 1).
On the one hand, manual sensor integration offers significant disadvantages compared to an automated solution, as the integration times are much higher, and losses in component quality and reproducibility are to be expected. An automated sensor integration solution, on the other hand, is inferior to the direct manufacturing of sensors via a multi-material LPBF process. Though, this concept still has limited maturity and more research is required before it can be tested to what extent direct sensor production is even possible.
Sensor Integration in LPBF
Manual sensor integration
Manual sensor integration has already been considered by several research groups: Havermann and Maier worked on the integration of Fiber Bragg Grating sensors into metallic LPBF components [7,8]. Mathew also focused on the integration of fiber optic sensors, with the consideration of special surface coatings, to ensure better material connection to the component [9]. Hossain, with his focus on electron beam melting, was able to integrate piezoceramics [10]. Preliminary work by the authors includes the integration of Pt100 sensors [11], RFID tags [12] and an electric motor [13]. In all preparatory work, the process chamber door was opened in order to integrate the sensors. Stoll proved for the material 1.4404 that process interruptions associated with the opening of a process chamber door have significant disadvantages: reduction in strength and changes in the modulus of elasticity are the result [14].
Automated Sensor Integration
With regard to automated sensor integration during the LPBF, predominantly theoretical preliminary work has been done so far: Sehrt has already described a concept for automated sensor integration in a publication, but has not yet implemented it [15].
Conceptually, it was possible to present the implementation of an automation unit within the framework of a test setup in the previous work of the present authors [16].
Manufacturing of Sensors Using LPBF
There are already some approaches for the manufacture of conductor paths (and therefore simple sensors) which are insulated from the rest of the component by the use of ceramics, using LPBF. The production of these tracks is seen as the first step towards the production of sensors, since sensors such as thermocouples, strain gauges or antennas are virtually made entirely of conductive tracks. A great challenge is to build up an insulating material. Initial approaches, for example by Koopmann, are to combine steel (X38CrMoV5-3) and ceramics (mixture with a weight ratio of 80 pct ZrO2 and 20 pct Al2O3). In that work, suitable parameters were found in order to build up insulating layers in the direction of build-up. If this approach is combined with multi-material mechanisms such as those of Anstaett and Aerosint in the future, the feasibility of manufacturing conductive paths seems likely.
There are already some approaches for integrating sensors into components manufactured by metal powder bed fusion. At a specific time during the LPBF process, layerwise manufacturing is interrupted in order to insert the desired type of sensor. Therefore, a cavity is built up, which is then filled with metal powder, as a result of the process.
To be able to integrate a sensor into the cavity, the following steps have to be fulfilled by an automation unit (see Fig. 2):
- Cleaning of the cavity
- Sensor selection and pick-up from a storage location
- Placement of the sensor in the desired cavity
- Realization of sensor communication
- Process restoration for LPBF resumption (e.g. layer homogenization)