When it comes to the latest aero engine components, today’s production technology is already challenged when it comes to meeting the requirements concerning accuracy, component safety and efficiency in series production processes.
In particular, the thin-walled and highly twisted blades of the latest blisk designs require an extremely stable milling process and high degree of process planning to avoid vibration of the blade during milling to avoid unacceptable surface defects, making this the most critical process in blisk production. Despite all these challenges being met for current geometries, future blisk designs are still evolving and will soon exceed the capabilities of today’s production systems.
Smart sensor technology
The usual strategy for creating a milling process for a new blisk geometry is mostly
based on trial and error, making it inefficient, time-consuming and dependent on the worker’s experience. Additionally, this makes the cost and duration of this phase even less predictable.
To avoid all these uncertainties, the approach followed by IPT, in cooperation with Ericsson, is based on real-time live data collected from smart sensors and the machine control system. In combination with workpiece simulations, this data gives an exclusive insight in what is really happening at the contact point between the tip of the cutting tool and the component surface.
To enable the monitoring of this most important process area, the workpiece itself is equipped with a wireless smart sensor that detects the current process stability status. This smart sensor captures workpiece vibrations using a miniature acceleration sensor directly attached to the component’s surface as shown above (Fig.1). Battery powered evaluation electronics and a wireless 5G transmission module enclosed in an IP68 waterproof housing make it possible to have the whole unit inside the milling machine and attached to the workpiece during the entire milling process of the blisk. The sensor system transmits vibration levels in a frequency range up to 10kHz to a receiver system outside of the machine for subsequent data analytics and determination of process stability.
5G for production
Wireless transmission technologies are widespread across many application fields. In industry, these applications are mostly restricted to non-critical uses with no safety issues, low data rates and no hard time constraints. For manufacturing and process control of complex products such as blisks, the transmission requirements are much higher. Timing, reliability and high data transmission rates are crucial when it comes to wireless closed-loop control systems and high precision process monitoring.
Where most standard technologies like Wi-Fi, Bluetooth, ZigBee, etc. fail, the future mobile communication standard 5G offers a latency of 1ms or less in combination with up to 10Gbit/s throughput. In its final stage, once a global infrastructure has been deployed and multi-level device support implemented, 5G will be a global network with the aforementioned characteristics, representing a universal solution for connected applications on shop floor level, logistics, tracking and combinations of these. Making this technology available and exploring industrial, reliable ultra-low latency use cases in a real production environment is the main goal of IPT’s and Ericsson’s cooperation.
The blisk milling project is the first of multiple trials currently being explored on the 5G equipped IPT shopfloor in Aachen, using the only precommercial 5G system in a production environment worldwide provided by Ericsson. With this unique combination of 5G equipment and comprehensive manufacturing machinery, Fraunhofer IPT is capable of providing partners from various industries and research fields with a unique testbed (Fig.2) to realise numerous industrial wireless applications.
Adaptive process control
After wirelessly acquiring the vibration information directly from the process, evaluating and processing are the next logical steps towards an optimised and stable process design, where the blade is not energised at resonance frequencies (eigenmodes). With a continuous extraction of the machine coordinate data from the control system of the machine, the sensor data can now be interlinked with the location on the part’s surface where it was captured.
By using this spatially resolved vibration data, in combination with the simulation data of the blade, a precisely tailored control strategy can be derived. Taking the current process stability from the sensor, the tool location from the machine control and the simulation data as input, a control algorithm can calculate the optimum spindle speed with respect to the real-life workpiece status and its constantly changing behaviour. Feeding back this information to the machine control finally closes the loop. With the simulation data as background information on the part’s behaviour, critical frequencies can effectively be avoided in an online closed loop control as they get detected.
Digital twin and Connected blisk
While process monitoring and control is important during manufacturing, documentation remains a continuous task throughout the whole lifetime of a part, ranging from CAD/CAM data, manufacturing data and metrology data to operational data and maintenance data. This documentation is required by law, causes an enormous workload and, in most cases, is done manually by the worker being in charge of the current processes. Applying data connectivity to all those stages enables the next important milestone towards the most central aspect of Industry 4.0; integrated and centralised data in combination with its availability to all networked client systems and users.
The ‘Digital Twin’ (Fig.3) of a component, an assembly or an entire aero engine combines all this data. The automation of this data collection requires highly integrated sensor devices with advanced pre-processing capabilities and a universal wireless data interface. 5G embodies exactly this interface, a global network suitable for all kinds of data acquisition systems ranging from low latency sensor and control applications to high volume long term statistical data.
For future aero engines, this will be achieved by highly-integrated devices that combine sensors, pre-processing, data-storage, communication modules and efficient energy supply systems. Devices like this will be highly ruggedised to resist the extreme environmental conditions inside an aero engine, which allow them to be embedded into the component throughout its entire lifetime.
The digital twin will allow a better understanding of what happened during the manufacturing of a part. In case of a defect or failure, the reason for this can be traced to its origin in order to eliminate the problem and prevent it in future. During operation, the collected data will be essential to determine the component’s health or wear status and schedule the next maintenance tasks accordingly. Finally, the collection of the resulting maintenance data will be an important input for the next generation of blisk design, based on the new information and insights empowered by 5G.
Authors: Paul Becker (1), Niels König (1), Sascha Gierlings (1), Tommy Venek (1), Thomas Bergs (1, 2), Robert Schmitt (1, 2)
(1) Fraunhofer Institute for Production Technology IPT, Aachen, Germany
(2) Laboratory for Machine Tools and Production Engineering (WZL) of RWTH Aachen University, Germany