EmSim - Method for determining adaptive monitoring limits

Process monitoring (PM) in machining has evolved from monitoring the condition of tools to a comprehensive method of ensuring production quality. It has a key role to play in automation and in ensuring error-free processes. The adaptability of PM systems to changing processes, tools and machines is becoming increasingly important for companies, as the demand for product individualization and variant diversity is constantly growing. However, complex parameterization procedures and the limited transferability of current methods to single-part production remain major challenges.

Existing monitoring systems are often designed for series production and use direct or indirect sensor signals that require complex calibration. The development of model-based simulation approaches and machine learning methods offer potential for making adaptive monitoring limits more efficient. However, solutions that take into account dimensional tolerances, component geometries and machining operations in single-part production are lacking. The development of such adaptive methods could revolutionize process planning and quality assurance.

Objective of the project

The main objective of the project is to research a method for the adaptive generation of process monitoring limits in process planning for machining production. Dimensional tolerances and prediction quality are to be taken into account in order to improve manufacturing quality, particularly in the production of individual parts. To this end, simulation-based approaches are being developed that can efficiently and precisely determine limit values and at the same time be adapted to critical and non-critical process cases.

The sub-goals include the development of an efficient simulation and current modeling for limit design, the investigation of critical process cases for adaptive limit formation and the investigation of systematic simulation errors to ensure highly accurate material removal simulation. Furthermore, we aim to model relevant process signals for different machining operations, as well as to adapt the simulation-based PM limits based on the prediction quality. On this basis, we combine all individual elements into a comprehensive method that can be seamlessly integrated into practice.

Advantages

• Process reliability - reduction of false alarms and increased sensitivity to process errors
• Flexibility - adaptation of monitoring to individual parts and processing operations
• Efficiency - automatic and precise determination of limit values in process planning
• Quality improvement - consideration of dimensional tolerances and component geometries
• Productivity - Reduced parameterization effort and faster setup of new processes
• Cost-effectiveness - minimization of rejects and optimized use of resources

Procedure

In the EmSim research project, we are developing a method for the adaptive generation of process monitoring limits in process planning. Dimensional tolerances from the design and the NC code from CAM planning serve as the starting point. The first step involves the further development of an efficient material removal simulation for calculating local engagement conditions.

On this basis, we develop numerical-empirical current models that generate rough tolerance bands for monitoring limits. We examine critical process cases such as drilling and turning grooving operations separately in order to develop specific models for different tolerance specifications. In parallel, we investigate systematic errors of high-resolution simulations and evaluate their potential for process-stable operations.

Finally, we integrate all elements into an overall system and validate this through machining tests with an active monitoring system. An accompanying database records process and quality data in order to continuously improve the models and monitoring limits.

Are you also interested in a cooperation project?

Please contact Martin Winkler via e-mail to winkler@ifw.uni-hannover.de or by calling +49 511 762 4991.