Thermal monitoring instrumentation for metal additive manufacturing - PYRAM

Lead Research Organisation: Cranfield University
Department Name: Sch of Aerospace, Transport & Manufact

Abstract

The aim of this proposal is to develop a temperature measurement instrument for use in a wire based additive manufacturing (AM) processes. The instrument will use a lensed fibre-optic cable fed to a camera-based design, allowing it to operate in different deposition environments and will be compatible with a variety of metals. The system will provide real-time temperature images of the AM melt pool that will improve the effectiveness of the additive manufacturing processes. In turn, this will result in high-quality AM parts and improved productivity via quality control enhancement.

Metal AM of large components will have a major impact on the production of specialist components due to its inherent cost and material savings as well as offering a route to easily changing the design and allowing component customisation. Metal components are formed by feeding a wire into a welding arc, or laser, which is then moved to deposit molten metal in predetermined positions and a structure is built by doing this in repeating layers. Structures built using this technique can have excellent material properties, but due to variations in the temperature of the melt pool, the internal metal structure can sometimes be irregular, which causes variations in the final mechanical properties of the component. Temperature measurement of the melt pool surface addresses these variations by ensuring the production of; a constant material internal structure, repeatable layer dimensions and component temperature heating/ cooling cycles, thereby ensuring good component quality control. If the metal surface temperature can be specified, measured and controlled, then this guarantees the mechanical properties of the component are within the required specifications.

Thermal camera measurements of melt pools is challenging as the temperatures can be in excess of 2400 deg. C. Also, the intense light from the arc can blind the camera or degrade measurement accuracy. Commercial thermal cameras used for AM processes tend to have large, fixed lenses which makes installation of the lensed camera difficult in the limited space around typical AM torches as they need to be line-of-sight with the melt pool, to view it clearly. To overcome these challenges, a novel optical fibre two-wavelength camera instrument, tailored for the wire-based AM process operating over the range 800-2400 deg. C, will be developed, which is not blinded by the intense arc light and has compact and flexible imaging optics. This allows the camera head to be used in restricted spaces but the camera instrumentation itself can be located some metres away from the AM processing tool, on the robot arm. This instrument could also be usefully used with other AM welding processes with some adaption, and applications where physical access is very restricted e.g. gas turbine engines. The two-wavelength design uses an optical-fibre bundle and a camera together with special filters to block the unwanted light but transmit two image "colours". These two colours are then imaged on the same camera sensor separately. The temperature of the images is determined by the ratio of the two light signals. This ensures a wide operating temperature range, without requiring special knowledge of the thermal properties of the melt pool itself. The instrument design overcomes the challenges presented by the intense light and restricted access. Custom software will produce a real-time temperature map of the melt-pool, and allow the instrument to be then used with the process software controlling the AM machine. This will allow power feedback control of the welding arc and hence limit significant variations in the melt pool temperatures. The research will develop a state-of-the-art instrument addressing one of the major challenges facing metal AM processes and provides a route to fabricating reproducible and specification compliant components.

Publications

10 25 50