New project: increasing the quality of parts made by Laser Beam Melting

LBM is becoming an accepted fabrication process for metal parts. But it still suffers from problems that mean their quality is not guaranteed. Sirris has launched a new project aimed at improving control of the complex thermal process.

The additive manufacturing process LBM (Laser Beam Melting) is gradually becoming a production tool for complex and functional metal parts. The range of materials available today is meeting industrial needs. But in applications where the metallurgical quality and mechanical properties of the fabricated parts are critical, the technique does not provide sufficient guarantee on the properties obtained and the repeatability of the quality. 

When fabrication parameters are optimized, the properties obtained on specimens are close to those required for the materials used and very interesting for industrial companies. However, these properties cannot be guaranteed on real parts as the thermal conditions vary during fabrication and the mechanical and microstructural properties depend directly on the thermal history of each layer. 

The traceability of these conditions is also insufficient. They result from the complex heat balance between:

  • laser energy input on each part (fusion of metal powder, phase transformations, heating of previous layers)
  • heating of the support plate (reduction of thermal shock, internal stresses and distortions)
  • losses by conduction through the powder to the walls and plate
  • forced convection due to argon flushing
  • radiation

The operator has to choose the technical parameters (laser power, flushing speed and interval, plate temperature, etc.) according to the material and type of part (thin, solid, etc.), determined for standard conditions. As the cross-section of parts changes continuously and as different parts on the same fabrication plate interact, the actual conditions are different. There is not yet an efficient regulation system. The operator can make some corrections according to visual observation of the surface of the layers but the adjustments remain empirical. 

From the appearance of the layers and analysis of the parts made, it can be assumed that there is momentary overheating during fabrication due either to excess energy, or to insufficient dissipation of the accumulated heat. Conversely, delamination and rough layers are found with too low temperature.

These deviations cause local defects such as porosity, cracks by stress accumulation or variations of properties and even fabrication problems jeopardising the operations in process. 

Fabricators' current solutions integrate the visual or heat information obtained only at the powder surface and it is often based on a comparison with previous layers. 
Ideally, the temperature of each layer should be stabilized by means of changing parameters or regulation, but the first step is the measurement and modelling of these temperatures. 

The Qualam project, which has just started and is to continue for two years, aims to obtain better understanding of the thermal exchanges of LBM thanks to thermal modelling in real time coupled with measurement of the surface temperatures, to ensure better control of the process and optimization of the parameters according to the variations during fabrication. One step further towards the quality assurance and traceability required by industrial companies. 

The combination of the thermal meta-model and measurements of surface temperatures, free from the part in fabrication, will eventually provide the thermal history at all points, thanks to a 3D view of the thermal exchanges, in places where measurements are not accessible. 

Sirris will evaluate the state of the art of existing solutions and define the specifications. It will adapt equipment for the measurement system. Finally, on the basis of simple study parts, it will establish correlations between microstructures, mechanical properties and process parameters. The centre will also evaluate the improvements obtained in terms of quality. 

The partners for this project are the CRM, which will provide the machine's instrumentation, Cenaero which will develop the model and ULg-MMS which will characterise and analyse the physical properties of the printed parts.