Xiao Shang - Research projects

This full-sized DED system is equipped with a powerful 1kw IPG laser, two powder feeding hoppers, and a user-friendly interface. Materials available for printing includes 316L stainless steel, PH-17-4 stainless steel, Ni superalloys and more. It also supports in-situ monitoring with a CCD camera and a high-speed IR camera. Upgrading work is ongoing for fully autonomous closed-loop control.

DED processing parameter optimization

Associated with LEMAM @ UofT

Optimizing the processing parameter for DED, and in fact for AM in general is chanllenging but at the same time crucial. Being able to fast and accurately identifying the optimal processing parameters reduces the cost, materials waste, and more importantly, improves the build quality.
We use machine learning and computer vision methods to study the melt pool and melt track variations under various processing parameters in order to create a framework for the esay identification of optimal processing parameters.

Application-specific materials-by-design is a long-standing challenge due to the need for capturing the complex process-microstructure-property (P-S-P) relations. The efficient identification of microstructures inversely from target mechanical properties is intractable because of the high dimensionality of the design space. So far, there has only been limited preliminary investigations on establishing surrogation P-S-P models with over-simplified material representations forwardly (as opposite to inverse design). For the arguably more important inverse design problem, there have been even fewer reports. In this work, we provide an end-to-end framework that tackles both forward prediction and inverse exploration to streamline materials-by-design. Using advanced deep-learning and genetic algorithm, our framework exhibits promising potential in cutting down the time needed from target mechanical properties directly to desired material microstructure.

Our poster won a poster prize award from Digital Discovery at the 2023 Accelerate Conference by the Acceleration Consortium!

Please check out the published paper:
Shang, X., Liu, Z., Zhang, J., Lyu, T., & Zou, Y. (2023). Tailoring the mechanical properties of 3D microstructures: A deep learning and genetic algorithm inverse optimization framework. Materials Today. https://doi.org/10.1016/j.mattod.2023.09.007

Image source:
(a) Shinohara, Y., Handbook of Advanced Ceramics
(b) Fan, X., et al., Thermoelectric magnetohydrodynamic control of melt pool flow during laser directed energy deposition additive manufacturing. https://doi.org/10.1016/j.addma.2023.103587
(c) Todaro, C., et al., Grain structure control during metal 3D printing by high-intensity ultrasound. https://doi.org/10.1038/s41467-019-13874-z

Functionally graded materials (FGMs) are a class of materials whose compositions and/or microstructures vary along one or multiple spatial directions, thus lead to varying material properties. Compared with traditional composite materials, FGMs agglomerate contradictory properties such as high strength and toughness, good wear resistance, good heat resistance, high stiffness to weight ratio, which are well sough after in modern industries such as the aerospace, automotive, and biomedical sectors.
In this project, I combine the flexibility of DED with external magnetic and ultrasonic vibration fields to locally tailor the microstructure of the builds. This way, FGMs with locally varying mechanical properties can be achieved.

Bistable Auxetic Metamaterials (BAMs) are a class of monolithic perforated periodic structures with negative Poisson’s ratio. Under tension, a BAM can expand and reach a second state of equilibrium through a globally large shape transformation that is ensured by the flexibility of its elastomeric base material. However, if made from a rigid polymer, or metal, BAM ceases to function due to the inevitable rupture of its ligaments. The goal of this work is to extend the unique functionality of the original kirigami architecture of BAM to a rigid solid base material. We use experiments and numerical simulations to assess performance, bistability, and durability of rigid BAMs at 10,000 cycles. Geometric maps are presented to elucidate the role of the main descriptors of the BAM architecture. The proposed design enables the realization of BAM from a large palette of materials, including elastic-perfectly plastic materials and potentially brittle materials.

To find our more about this work:
Shang, X., Liu, L., Rafsanjani, A. et al. Durable bistable auxetics made of rigid solids. Journal of Materials Research 33, 300–308 (2018). https://doi.org/10.1557/jmr.2017.417. Full-text paper
Featured in the MRS Bulletin