The purpose of MAShES was to develop a system for real time (RT) control and high-speed monitoring for laser processing that brings into play:- Reliable monitoring of temperature distribution, - The 3D seam profile and 2D melt pool geometry, - The surface texture dynamics, and...
The purpose of MAShES was to develop a system for real time (RT) control and high-speed monitoring for laser processing that brings into play:
- Reliable monitoring of temperature distribution,
- The 3D seam profile and 2D melt pool geometry,
- The surface texture dynamics, and process speed
in a unified and compact embedded solution.
MAShES RT control can act simultaneously on multiple process variables, including laser power and modulation, process speed, and spot size. MAShES was designed under a modular approach, customizable for different laser processing applications. Scenarios of high added value and impact have been selected for demonstration: laser metal deposition and laser welding.
Overall, MAShES developed a breakthrough compact multimodal imaging system for RT closed-loop control of laser processing. It was built on a novel multispectral optics and multisensor arrangement in the VIS-MWIR spectrum. Temperature, geometry, and speed, was imaged accurately and reliably. The system featured embedded RT process control together with the capability of cognitive readjustment and process quality diagnosis. Therefore, MAShES addressed the development of a novel intelligent and self-adaptive system for continuous and autonomous process control.
The use of MAShES system will allow the harmonization of high performance and quality with cost effective productivity, enabling at process level, reconfigurable, adaptive, and evolving factories. End-users would be capable to deal with highly dynamic operations in a productive way.
The following activities were carried out during MAShES project:
• Validation of spectral behaviour of MAShES multispectral optical system, characterizing the transmission through the SWIR and MWIR paths. This characterization allowed to identify an error on the manufacturing process of some of the coatings, and to obtaining new ones to be integrated in the MAShES multispectral optical system.
• Temperature determination with the MAShES multispectral optical system, using a blackbody as a reference for the temperature determination of a grey body.
• Laboratory validation of MAShES embedded real time (RT) control system for laser welding and LMD. Two prototypes of the multispectral optics were integrated in two different laser processing heads for laser welding and LMD validation. The embedded RT monitoring and control system show its ability to detect different defects in laser welding and to assure a controlled LMD process.
• Multispectral optical system. The needed optics (dichroic mirrors, lenses, …) and optomechanical parts were manufactured and assembled. Commercial bandpass filters were used for band separation. Table 1 shows the configuration implemented for the 3 prototypes evaluated in the project.
• Embedded electronics for real time (RT) for multimodal monitoring and control. The embedded electronics designed in the first period was manufactured. RT monitoring and control functions were developed and implemented: multimodal fusion, melt pool geometry determination, cooling rate measurement, close-loop control, and convolutional neural network for fuzzy control.
• Simplified embedded electronics for RT control of LMD process, using only one high velocity SWIR/MWIR sensor. It was an answer to the requirements by the laser processing CNC machines for a more compact RT control system. The optical configuration used for this simplified system is shown in Table 2.
• Cognitive control system and interfaces, for laser welding and LMD. A proper approach regarding the architecture of the cognitive system was achieved for quality diagnosis and RT control.
• Demonstration of MAShES system for laser welding. Laser welding trials were carried out to characterize the performance of the MAShES system based on different tests on the Jeep Renegade front door (right side) made in Melfi plant.
• Demonstration of MAShES system for LMD, applied for the reconstruction of moulds used in production.
• Benchmarking related to the tests performed for the evaluation of MAShES system and its sub-developments and provided a comprehensive evaluation of the different pilot cases organized and executed during the project’s period. The assessment was conducted based on specific Key Performance Indicators (KPIs)
• Dissemination of the project have been addressed increasing the presentations in fairs, specialized conferences, and workshops as well as the number of scientific publications.
• The exploitation plan was further developed, with the aid of an innovation workshop, where the business plan for the exploitable results was detailed. Three exploitable results were identified:
o Embedded electronics and control system for LMD, that is being commercialized by NIT with the support of AIMEN.
o Cognitive control system and interfaces, that will be exploited by ILT.
o MAShES system, to be commercialized by Permanova, for robotized laser processing systems, and by Prima for laser processing CNC machines.
A novel open source middleware based on ROS has been developed that aims to facilitate deployment and virtualisation of novel approaches to multimodal monitoring of laser processing by allowing a smooth integration of sensors and equipment in any laser robot cell.
A novel multisensor board capable to image melt pool at high speed in different spectral bands from visible to middle wavelength infrared (0.4-5 µm wavelenght). This board enables high speed spectral imaging of the melt pool for multimodal monitoring of laser processing.
A novel optics has been designed for coaxial multispectral imaging of melt pool in laser processing applications that can be integrated in existing laser heads.
Overall, the MAShES project has enabled for the first time coaxial multispectral imaging (from the visible to the MWIR) for laser processing real time control, and allowing accurate temperature determination (> 500 ºC).
More info: http://www.mashesproject.eu.