In-Process Monitoring of Laser Sintering
Lead Research Organisation:
CARDIFF UNIVERSITY
Department Name: Sch of Engineering
Abstract
Metal powder bed fusion, most commonly known as metal 3D printing, allows the manufacture of complex, optimized and lightweight components otherwise impossible to manufacture with traditional methods. This method of manufacture is more prone to generate defects such as cracks and voids. This project plans to utilize Acoustic Emission to monitor the mechanical energy released by cracks and voids during the build process. The project aims to develop a detailed understanding of the acoustic signatures of developing defects and to work towards the implementation of a process control system capable of detecting critical flaws and stop individual builds, which will greatly reduce material waste and aid with part redesign. This project directly aligns with two EPSRC themes control Engineering and Manufacturing Technologies.
People |
ORCID iD |
| Rhys Pullin (Primary Supervisor) | |
| Stephen Ball (Student) |
| Description | The possibility of locating where a crack has occurred in a 3D-metal print to within an acceptable degree of accuracy has been proven. In a print made of many small parts, this will allow the printing of the defective part to be stopped without effecting the others. A key discovery was the identifying of key differences in the shapes of individual build plates using the Polytec 3D scanning vibrometer. This meant that whilst in some plates the signal would enter the heater plate at the centre and head outwards towards the waveguides, for others it would go through the edges and straight down into the waveguides and to the sensors. This meant that for seemingly similar plates there could be very different location accuracy. To overcome this, the heater plate and waveguide designs were altered so that the waveguides would be secured directly to the build plate. This significantly simplified the wavepaths for signals. Additionally, it was found that due to the tuning nature of the waveguides and the huge number of reflections occurring in the small plate, that any meaningful characterisation of defects from acoustic emission is unlikely. |
| Exploitation Route | There are still steps to be taken in the charcaterisation of signals. It has been found that charcaterisation ofdefects by acoustic emission alone is unlikely. However, if combined with other monitoring methods, such as thermal or visual methods it may still be possible. |
| Sectors | Aerospace, Defence and Marine,Manufacturing, including Industrial Biotechology |
| Title | Delta T Data |
| Description | Data was collected from multiple build plates, at various levels of securedness, in order to create Delta T maps of the plate. This is a selection of matrices that contain the difference in time it takes to reach a pair of sensors from each point on the build plate. This can then be used to locate in-process signals with a high degree of accuracy. By loosening some bolts, the effect of the plate not being secured properly could be seen in location results. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2018 |
| Provided To Others? | No |
| Impact | This data gave a greater understanding of the physical factors that can affect the ability to locate defects by acoustic emission. |
| Title | Vibrometry Data |
| Description | A Polytec 3D Scanning Laser Vibrometer was used to collect data on how signals propagate through the the various parts of the build plate assembly (build plate, heater plate and waveguides). Data was collected for 4 waveguide designs, 3 build plates, and for each build plate attached to the heater plate. The data includes the x,y and z velocities of a dense grid of points across the surface. This can then be manipulated to give data on displacement, frequency and wavenumber. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2019 |
| Provided To Others? | No |
| Impact | This data revealed alot about how propagate and disperse through the structure. The most notable detail revealed was that for different build plates, the signal would travel into the heater plate and towards the waveguides in different positions. This could have significant effect on the location accuracy. Beacause of this the design was altered so that the signals would travel directly from the build plate and into the waveguides. |
| Description | Speaking of Science 2018 |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Postgraduate students |
| Results and Impact | In May 2018 I took part in Cardiff University's Speaking of Science, an event in which PhD researchers present to others from across the University. The main challenge of the event was to convey the scope and importance of my work to an audience with no or very little specialised knowledge in the area. |
| Year(s) Of Engagement Activity | 2018 |