The Citrine smart materials platform is based on cutting-edge AI tools and smart data management infrastructure, and can be used for data-driven material and chemical development. The platform can predict the performance of materials under various formulations based on machine learning models and combined with the industry knowledge of user companies to help accelerate material research and development.
Material synthesis industry: 3D printing metal materials urgently need artificial intelligence to help break through performance limitations
Additive Manufacturing (AM) is commonly known as 3D printing. This technology can produce parts that are quite close to the final specifications, reducing material waste and welding and other subsequent processing. However, the limitations of printing materials make 3D printing not yet widely used in industrial fields. Take the HRL laboratory jointly owned by Boeing and General Motors as an example. The aerospace and automotive industries require alloys that are lightweight and have high strength at extreme temperatures. However, the strength of 3D printed aluminum alloys is not enough to meet the conditions. Since 3D printed metals will form hot cracking during the curing process, which affects the strength of the metal, the HRL laboratory hopes to find a nanoparticle that can nucleate microstructures to reduce the occurrence of hot cracks.
Case analysis of material intelligence R&D application: Take Citrine's assistance in finding aerospace-grade 3D printed aluminum alloy as an example
The Citrine smart materials platform is based on cutting-edge AI tools and smart data management infrastructure, and can be used for data-driven material and chemical development. The platform can predict the performance of materials under various formulations based on machine learning models, combined with the industry knowledge of user companies, and accelerate material research and development.
1. Core functions
The Citrine smart materials platform can quickly search for 11.5 million combinations of powders and nanoparticles. The platform finds combinations in batches for the properties of the target materials, identifies related data sets and data streams, and creates a data structure for material perception; then based on the data, it generates, refines, and validates the model. In addition, the platform supports the operation mode of man-machine integration in every link, and the functions involved mainly include:
Data management-The platform can extract, construct and store data during the development of materials and chemicals, avoid repeated tests, and quickly find relevant data sets.
Domain Knowledge Integration (DKI)-The platform will maximize the use of users' domain knowledge data to optimize the performance of machine learning models. Predict the performance of materials under various formulations through six steps: data import (Ingestion), data curation (Curation), data characterization (Featurization), design space generation, AI model generation and target material setting, and give experiments Suggest.
Dual interface operation-The platform has two different interfaces: one is a graphical web interface, which allows intuitive access to data and machine learning models. The second is a flexible Python client that can support data engineers and receipt scientists to automate workflows.
2. Application effect
In a collaborative project between HRL Labs and Citrine, the platform identified 100 candidate combinations of powders and nanoparticles to optimize the attributes that the HRL team needs to test for synthesis. The resulting material al7a77 is the first high-strength aluminum alloy powder raw material for ready-made AM machines registered with the Aluminum Association. At the same time, its material research and development work has achieved a number of achievements, including:
1. Time saving: Due to the material informatics method, the work of the experimental laboratory is reduced from several years to several days.
2. Performance improvement: The new alloy powder can still maintain its strength when used in ready-made 3D printing equipment.
3. Be the first to register: This research and development result becomes the first additive alloy registered by the Aluminum Association.
4. Entering the market for the first time: Two years after the publication of Citrine related research reports, NASA became the first commercialized customer.