Driving Development in the Fast Machining of Complex Parts

Summary

Analysis by McKinsey & Co finds that the additive manufacturing (AM) industry is now worth an estimated €13.4 billion with a 22% annual growth rate. But the report also finds that, despite the benefits offered by AM, there are still significant barriers preventing mass uptake. Here, Ross Turnbull, Director of Product Engineering and Business Development at ASIC company Swindon Silicon Systems discusses these challenges, and how smarter sensors can help overcome them. 
 
In an industrial landscape where customers are seeking more tailored product solutions, AM technologies can prove particularly advantageous. Any design can be produced on a single piece of machinery, with no need to create specific moulds or tooling for a single production run. As a result, manufacturers benefit from increased flexibility in design, enabling customisation of goods for product differentiation and improved customer-manufacturer relationships.
 
AM also provides a quicker route from digital design to physical prototype, enabling a fast design to manufacture cycle. In applications where the product must fit within a limited space envelope or among other components, fit and function can be confirmed at this stage without the need for costly investment into full-blown production processes. And thanks to the unique layer-by-layer assembly approach of AM, it’s possible to machine even the most complex of component geometries in a single step. 
 

Understanding common challenges

Despite the potential offered by additive manufacturing, there are still challenges with successfully implementing and using the technology. Research by Deloitte suggested that reliability continues to be a concern for wider implementation of AM. This includes a lack of process stability, part quality and reproducibility. For large scale production processes, particularly for industries like the aerospace and automotive sectors, such discrepancies in product quality could be disastrous for the manufacturer. Quality issues can lead to increased cost, delays and product recalls.
 
Problems such as poor layer adhesion, cracks and warping are common. As a result, there’s a need for manufacturers to have better visibility over their processes. Monitoring process parameters can help quickly spot and rectify defects before they turn into major issues in component structure and integrity.
 
Optical sensors are ideal for this, offering manufacturers vision into interior and exterior properties. Mono or colour CCD and CMOS optical sensors, for example, can be useful to identify part defects. Laser scanners can be used to analyse layer geometry. 
 
For processes like extrusion deposition where the build material must be melted, any fluctuation in temperature can significantly hinder the quality of the finished products. Here, IR cameras may be used to visualise temperature gradients and identify problem areas.
 

Building better sensors

Though significant progress has been made in optical image sensors, there are still limitations when it comes to AM applications. Developing optical sensors with the performance required to make AM successful can be a challenge. It’s imperative that the data collected by the sensor is reliable, with minimal noise for the most accurate signal and decision-making.
 
So, how can we achieve this? First, we’ll need to delve a little deeper into the optical sensor. Most sensors will combine the sensing element with one or more ICs acting as the sensor interface. The ICs are responsible for the necessary signal conversion and processing tasks, used to translate light intensities into electronic signals that can be interpreted and actioned by other devices.
 
But relying on standard ICs can be problematic, as these off-the-shelf solutions are unlikely to offer optimal performance. Signal processing issues can result in poor-quality data, making closed-loop control and autonomous operation difficult. Furthermore, performance and reliability issues mean that overcoming the issues of process stability and repeatability — highlighted earlier by Deloitte’s research — is impossible. As a result, this standardised approach is unlikely to overcome the industrial challenges currently posed by AM.  
 
Instead, the adoption of a custom IC solution should be considered, in the form of an Application Specific IC (ASIC). An ASIC is a completely bespoke chip customised for the application requirements.
 
An ASIC designed for a particular AM process, for example, can be specifically designed to overcome the aforementioned challenges. This includes investment into processing power for scenarios where accurate decision-making must be done in real time. Or, for components where high precision takes priority, the performance of the ASIC can be optimised for this too with sensor-specific protocols for higher quality images.
 
Full control over the design means that the ASIC often has a much smaller footprint than other comparable solutions, making the sensor easy to integrate into existing machinery. And with a customisable choice of wired and wireless communication protocols available, ASICs can support both new and legacy equipment.
 
Additive manufacturing offers the potential to transform large-scale production processes, enabling machining of complex components in a variety of materials in just a single step. But if we’re to scale up production, more sophisticated sensing technologies and quality assurance steps will be a necessity. Customised ICs provide the advantage here, offering a custom solution to fit the specific needs of any AM process sensor.