22S5003_Atch 2_TO 0001 SOO.pdf

Technical Operations VI (TOPS VI) Task Order Description/Statement of Objectives

TOPS Effort #: 0001

Date: 17 September 2021

TO Title: Material Informatics for Materials Discovery, Optimization and Performance Prediction II

Period of Performance: 51 months after mail date (48 months for technical effort plus 3 additional months for completing the final report).

SOO Paragraphs:

1.0 Background: The purpose of the program is to leverage emerging machine learning and data analytics approaches to develop new techniques for materials discovery, material microstructure/property optimization and performance prediction. As these machine learning and data analytics techniques have been employed to revolutionize different economic sectors including advertising, finance, and biotechnology there is significant potential to improve and accelerate materials development, design and sustainment by applying these approaches to the broader field of materials science and engineering. Although material informatics or the use of machine learning and data analytics is not entirely new, there are a wide array of problems and applications in materials that could be positively impacted by the application of current state-of-the-art emerging approaches in machine learning and data analytics. These innovative approaches will cover analysis, evaluation, modeling and simulation methods.

2.0 Scope: The task focuses on applying machine learning and data analytics approaches to develop approaches to demonstrate their efficacy for materials discovery, material optimization, and performance prediction. Specifically, this effort seeks to employ machine learning to explore novel chemistries and compositions for new materials. Emphasis on employing machine learning tools would be on physics and/or chemistry-derived approaches rather than the curve fitting.

Additionally, data analytics approaches will be applied with sufficient fidelity to effectively characterize and quantify existing materials and their underlying microstructure and conversely, predict materials properties with confidence capturing realistic spatial and temporal materials microstructure morphology distribution. To assess fidelity of the data analytics and machine learning tools emphasis should be on the accuracy and confidence on capturing the local materials phenomena (such as statistical anomaly in microstructure that may be perceived as defects or localized damage or crack) on the non-local response such as materials or structural element overall macroscopic strength or performance. In this aspect, evolving graph theory based Bayesian learning principles/algorithms or similar approaches might possess value. Finally, with appropriate experiments and or simulation, machine learning will be applied to understand the relationships between the quantified material structure and the resulting material performance.

Applicable material classes include ceramics, polymers, metals and biological materials.

Demonstration of the developed approaches should be performed with key performance goals related to meet notional requirements for specified applications relevant to the United States Air Force.

3.0 Objectives: This effort will consist of three tasks described as follows:

3.1 Task 1: Materials Discovery. The objective here is to apply advanced data mining tools coupled with knowledge of process modeling to develop new novel materials and material solutions. Specifically, the objective is to develop a digital model-based framework for materials characterization and discovery to allow for library-based property description that includes statistically defined materials constitutive relations, simulations and inverse design of novel materials with defined properties governed by morphology.

3.2 Task 2: Materials Optimization. The objective here is to apply innovative and efficient machine learning and data analytics approaches to suggest optimization of chemistry and topology to optimize materials for specific performance goals. For example, tools can be applied to improve processes for additive manufacturing which presents the challenge of a high-dimension process space that is difficult to optimize and establish a robust product-specific processing path and similarly guiding process protocols for complex multi-phase, multi-constituent materials morphology optimization. By leveraging advances in autonomous robotics, artificial intelligence, data science, and high-throughput in-situ techniques, a system may be developed that designs, executes, and analyzes process optimization experiments on-the-fly. The goal is to allow machine-controlled, real-time adjustment of manufacturing and local design to deliver performance and macroscopic dimensions within specifications.

3.3 Task 3: Materials Performance Prediction. The objective here are to develop algorithms that can link distributed microstructure at multiple scales and phases to the underlying response functions that dictate overall material behavior. Much work has been done to characterize and model structure-dependent response; however, with the advent of new experimental techniques that enable acquisition of terabytes of microstructure and response data there is a lack of tools that can reasonably interrogate these data to determine meaningful relationships. Specifically, material structure can be quite complex and important physics dictate behavior on multiple scales. Relationships between the complex structures on multiples scales and the response are not always immediately apparent.

4.0 Description of Work:

4.1 Task 1: Materials Discovery. The contractor shall apply advanced data mining coupled with knowledge of process modeling to develop new novel materials and material solutions.

Specifically, the a digital model-based framework for materials characterization and discovery will be developed to allow for library-based property description, simulations and inverse design of novel materials with defined properties governed by morphology. Approaches may consider innovative, efficient data mining approaches and use of processing modeling tools to suggest new novel chemistries or architectures with potentially desirable properties.

4.2 Task 2: Materials Optimization. The contractor shall employ machine learning and data analytics approaches to suggest optimized chemistries and/or topologies at pertinent scales to optimize materials for specific performance goals. By leveraging advances in autonomous robotics, artificial intelligence, data science, and high-throughput in-situ techniques, a system should be developed that designs, executes, and analyzes process optimization experiments on-the-fly. Additionally, the contractor shall utilize optimization concepts, (e.g. agent-based and evolutionary algorithms), metaheuristics, and multi-model predictive analytics to quantify the relative quality of paths through configuration space.

4.3 Task 3: Materials Performance Prediction. The contractor shall develop tools and approaches for performance prediction as a function of the underlying material chemistry and/or topology.

Current state-of-the-art imaging and in situ characterization tools should be employed and be integrated to the predictive tools to generate relevant data for application of the machine learning approaches. Additionally, robust simulation of microstructure-sensitive performance can also be utilized for the developed analytic approaches. Algorithms and methods should be developed link distributed microstructure at multiple scales to the underlying response functions that dictate overall material behavior. Further, the contractor should quantify the fidelity of performance prediction tools in linking the local material anomaly (microstructure defects, damage, etc.) to macroscopic response.

5.0 Deliverables

5.1 CDRLs as listed below:

CDRL Name Frequency

Data Item A001 Final Report End of T.O

Data Item A002 Funds and Man-Hour Expenditure Report Monthly

Data Item A003 Contract Funds Status Report Quarterly

Data Item A004 Technical Status Report Quarterly

Data Item A005 Presentation Material As Required

Data Item A006 Data Management Plan, Continuously

Data Item A010 Technical Report-Study End of T.O

Data Item A011 Computer Software Product End Items As Generated

Data Item A014 Computer Software Product/Source Code As Generated

Data Item A018 Interface Design Description

As Generated

Data Item A019 Software Design Description

As Generated

5.2 Software. The contractor shall deliver any software developed under this program including the source code (A011, A014, A019 (if applicable)).

5.3 Hardware. The contractor shall deliver any hardware developed under this program included any testing devised developed to meet the overall program requirements.

5.4 Test Samples, Prototypes and Residual Materials. The contractor shall return to the Government all test samples, prototypes and residual materials not consumed in testing at the conclusion of the technical effort.

5.5 Data. The contractor shall deliver all relevant data including raw machine output along with machine calibration parameters related to any and all experimentation, demonstration or simulations that result from the program (A011, A014, A019 (if applicable)).

6.0 Operations Security Requirements: The contractor shall participate in all activities associated with the disciplines of the organization’s Industrial Security, Information Security, Personnel Security, Operations Security (OPSEC), and Antiterrorism programs, following appropriate measures in each program as required for this particular contract. Security measures are required to reduce program vulnerability from successful adversary collection, exploitation of critical information, and violations of export control requirements. The prime contractor shall ensure all subcontractors, if applicable, conform to these requirements as required by the prime contractor.

Program Protection Plan (PPP) – Any potential critical program information (CPI) generated as part of this effort will be reviewed to determine the need for a PPP or to be included as part of an existing PPP.

Start Date: Mail Date

Funding Profile:

FY21 FY22 FY23 FY24 Total

$300,000 $300,000 $200,000 $200,000 $1,000,000

Place of Review, Inspection, and Acceptance: Destination

Estimated Level of Effort: 12 person years