An Application for Predicting Environmental Conditions within an Additive Manufacturing Facility
Environmental factors such as temperature and humidity can significantly affect the quality of the metal powder used in power bed fusion additive manufacturing, leading to variations in the integrity of finished components. It is therefore important to monitor and minimize the exposure of metal powder to adverse atmospheric conditions. Closed loop climate control systems can be employed to keep room temperature and humidity within a predefined range, however these systems can be expensive to install and operate for large factory areas where multiple machines may be in use. This additional expense can represent a significant barrier to manufacturers who are interested in adopting metal powder additive manufacturing technology. To reduce this barrier, predictive modeling of the climate within an additive manufacturing facility can be employed to complement (or even as an alternative to) full facility climate control. By simulating the evolution of temperature and humidity within a factory, it is possible to predict where climate control systems may be required and proactively mitigate against environmental effects, for example by altering production schedules to account for climate conditions.
This presentation explores the potential value of such modeling by showcasing an application that can simulate the environmental conditions within the DRAMA facility, located at the Manufacturing Technology Centre (MTC) in Coventry, UK. DRAMA is a three year collaborative research project focused on building a stronger additive manufacturing supply chain for the aerospace industry, and the DRAMA facility is a reconfigurable learning factory that suppliers can use to test and validate their process chain. A time-dependent model of nonisothermal flow within this factory has been developed in COMSOL Multiphysics®, for the purpose of predicting the temperature and humidity within the facility. The model is embedded within a COMSOL® application, simplifying model customization and enabling it to read data from temperature and humidity sensors for use in initial and boundary conditions. By validating the predictions of this model against sensor data from the facility when is operational, it is anticipated that the value for using predictive modeling in this area will be demonstrated. Ultimately, it is hoped that the application will help schedule manufacturing operations within the facility and permit greater control of powder and part quality, enabling it to become a key component of a digital twin of the facility.
