top of page

Machine Learning - Driven Thermal Optimisations Comes to Open Market Slicing

AON3D, a Montreal - based maker of high temperature 3D printers, stated at RAPID + TCT 2023 that they would unveil a new machine learning driven thermal optimisation software, which they claim will lead to higher performance and more consistent parts in Material Extrusion (MEX).

The main technology is a simulation engine designed exclusively for MEX that gives quick and accurate predictions on heat flow behaviour across an object during printing. AON3D's hybrid modeling technique, unlike other software on the market today, combines the established accuracy and dependability of numerical approaches with the speed and convenience of use of statistical modeling.

The output can be utilised for parameter optimisation, pre-build QA/QC, post-build inspection, and other purposes during the build preparation process, providing various paths for cost savings and scrap rates, improve reliability, optimise performance, and drive better technical decision making.

The Missing Layer in Current Slicing

Build preparation software, also known as "slicers," provides varied levels of analysis and optimisation for processes such as toolpath behaviour, support placement, and other geometric elements. What they lack is a compleletly separate dimension of thermal and heat flow behaviour that should be adjusted, affecting final part performance, dimensional correctness, and other factors. Indeed, existing slicing approaches apply temperature and speed parameters in a "one size fits all" fashion, or necessitate arduous manual intervention over specific regions of layers.

Achieving Consistent, Higher - Performing Parts with Thermal Intelligence

AON3D's software will give customers with a full depiction of the temperatures and heat flows throughout a part starting with simply a G-code file. This allows you to easily observe how process decisions affect individual results and take corrective action before printing or establish inspection criteria after printing.

AON3D's program can also provide completely organised raw data for usage in other analytical software such as structural FEA or custom post-processors.

Data generated will be available at spatial resolutions as low as one third the width of an extruded polymer bead. In contrast to a layer-by-layer technique, this transient simulation caters for complicated part features and areas intrinsically, resulting in strong and dimensionally accurate parts.

AON3D is also working on composite and semi-crystalline polymers, which have historically been difficult to model and simulate, and include some of the most common options such as PEEK, Nylon, PAEK, PEKK.

Reducing the Downstream Quality Management Burden

By adding an entirely new type of quantitative data, AON3D's innovative software will enable engineers to make educated design and process optimisation decisions. Manufacturers can reach a new level of consistency and part confidence across machines and production runs by using the AON M2+ High Temperature 3D Printer and ReadyprintTM filaments, reducing the need for time-consuming testing and quality control.

For more information on AON3D, please contact our local Dynagraph representative.

9 views0 comments


bottom of page