Knowledge Transfer

TEMPEST’s university network is designed to make disciplinary depth more powerful by connecting it to complementary expertise. Knowledge moves in both directions as researchers and trainees learn how another field frames questions, produces evidence, and judges whether a model is useful.

The triple-closure problem cannot be addressed within one department or discipline. Particle and plasma physics, fluid dynamics, turbulence, applied mathematics, experiments, scientific computing, AI/ML, uncertainty quantification, astrophysics, and engineering each describe a different part of the causal chain. University-to-university knowledge transfer gives those communities a shared working language and creates teams that can follow a scientific question from a microscopic mechanism to a measurable application outcome.

TEMPEST will use common research questions rather than parallel work plans as the organizing unit. A team may begin with an experimental observation, identify the particle-scale or continuum information a model is missing, develop a structure-preserving representation, test it in high-resolution computation, and determine whether it changes a prediction in fusion energy, supernovae, or hypersonics. That process transfers questions, methods, assumptions, data, and standards of evidence—not only finished results.

What moves in both directions

• Across the university network: Deep disciplinary expertise, specialized experiments, theory and algorithms, courses and mentoring, research software, and long-horizon method development.
• Back to every participating team: New questions, shared benchmarks, cross-trained students, reusable methods, broader validation, and connections to applications outside a team’s original domain.

How TEMPEST will make the pathway active

1. Cross-disciplinary workshops: Focused workshops will bring researchers and trainees together around a defined closure, measurement, modeling, or application question. Participants will leave with shared definitions, testable hypotheses, and clear next steps.
2. Cross-institution research teams: Projects will deliberately combine complementary expertise and use joint milestones, shared data, common software interfaces, and regular working meetings to keep the pieces connected.
3. Student and postdoctoral visits: Short rotations and longer visits will place trainees inside another university group or experimental facility so they can learn its methods, tools, and standards through direct participation.
4. Shared mentoring and cross-training: Students and postdoctoral researchers will have access to mentors outside their home discipline. Short courses, tutorials, reading groups, and hands-on modules will help them build the vocabulary needed to work across fields.
5. Team retreats and synthesis sessions: Center retreats will create time to compare assumptions, reconcile terminology, identify gaps between projects, and decide which ideas are ready for a common benchmark or cross-application test.
6. Open research products: Where appropriate, teams will produce shared datasets, benchmark definitions, documented code, tutorials, white papers, and publications that allow knowledge to move beyond the original collaboration.

What the pathway supports

• Education mission: Knowledge transfer is part of the educational process. A TEMPEST trainee should be able to explain a scientific question to an experimentalist, mathematician, computational scientist, AI/ML researcher, and application expert—and understand what evidence each person needs. Cross-training does not replace disciplinary depth; it teaches students how to connect that depth to the rest of the Center.
• Research mission: The research benefit is faster, more reliable connection across scales. When teams share variables, uncertainty descriptions, data formats, software expectations, and validation criteria early, a closure developed at one scale is more likely to remain physically meaningful and usable at the next.

Examples of progress

• Co-advised students and dissertations
• Joint publications and proposals
• Shared benchmarks, data, and software
• Methods transferred between disciplines
• New cross-application research questions

University–national laboratory exchange connects the Center’s long-horizon research and training mission with mission-driven multiphysics, major facilities, high-performance computing, mature software, and rigorous verification and validation practice.

This pathway begins by aligning goals. Laboratory partners bring consequential scientific questions, established workflows, validation experience, data, facilities, and an understanding of where uncertainty limits a mission-relevant prediction. University teams bring new theory, mathematical structure, experimental ideas, algorithms, scientific machine learning, and the time and educational environment needed to explore methods whose value may develop over several years.

The exchange is bidirectional throughout the project. Laboratory scientists help define useful benchmarks, interfaces, evidence, and performance requirements before a method is finished. University researchers and trainees return tested models, documented algorithms, uncertainty information, and new physical understanding. Together, the teams determine whether an advance is credible and practical enough to enter a larger scientific workflow.

What moves in both directions

• Universities contribute: Fundamental theory, exploratory computation, new diagnostics and experiments, structure-preserving methods, students and postdoctoral researchers, and long-horizon development.
• National laboratories contribute: Mission questions, large-scale computation, national facilities, high-energy-density and plasma expertise, mature codes, validation practice, multidisciplinary teams, and operational constraints.

How TEMPEST will make the pathway active

1. Student internships: Students will work in laboratory teams on Center-aligned problems, gaining direct experience with mission questions, large-scale workflows, verification and validation, and collaborative scientific software.
2. Extended and embedded visits: Faculty, students, and postdoctoral researchers may spend longer periods working with a laboratory group when sustained engagement is needed to understand a code, facility, dataset, or scientific workflow.
3. Postdoctoral exchanges and co-mentoring: Joint mentoring and exchange opportunities will connect early-career researchers to both university and laboratory expectations and help build durable relationships across institutions.
4. Joint benchmarks and validation campaigns: Teams will co-design reference problems, comparison quantities, uncertainty descriptions, and evidence thresholds that allow a new closure or algorithm to be judged in progressively more demanding settings.
5. Workshops and partner roundtables: Focused workshops and recurring roundtables will identify shared research topics, compare current capabilities, review progress, and adjust the collaboration when partner needs or scientific evidence change.
6. Software and data exchange: Where agreements permit, teams will connect documented software components, shared repositories, curated datasets, reproducible workflows, and training materials so that an advance can be evaluated and reused.

What the pathway supports

• Education mission: Internships, long-term visits, and co-mentoring make laboratory knowledge part of a student’s education. Trainees learn how fundamental ideas are evaluated inside large multiphysics programs, how uncertainty and validation affect use, how teams maintain scientific software, and how to communicate across research, engineering, and mission contexts.
• Research mission: For TEMPEST, the laboratory connection is a demanding test of transfer. A model must preserve the right physics, remain stable and scalable, expose its uncertainty, and improve a meaningful quantity when it is placed inside a larger calculation. Laboratory feedback also returns new basic-science questions to the university teams.

Examples of progress

• PhD internships and extended visits
• Postdoctoral exchanges and joint mentoring
• Joint benchmarks and validation evidence
• Transferable software and data products
• New research questions informed by mission needs

University–industry exchange connects TEMPEST’s fundamental research to application requirements, engineering constraints, rapid design cycles, and the practical question of whether new scientific understanding improves a real decision.

Industry partners help the Center identify which uncertainties, missing physics, or computational limits matter most to a design or experimental program. Their perspective clarifies the quantities a predictive model must deliver, the fidelity and turnaround time that are useful, the evidence needed for confidence, and the constraints that determine whether a method can be adopted.

University teams contribute theory, experiments, advanced algorithms, scientific machine learning, validation methods, and a pipeline of students and postdoctoral researchers. TEMPEST’s role is to co-design fundamental questions with line of sight to use—not to replace a company’s engineering process. The strongest projects produce both new science and a clear path for evaluating whether that science changes an application-relevant outcome.

What moves in both directions

• Universities contribute: Fundamental explanations, experimental and computational methods, structure-preserving models, independent analysis, long-horizon exploration, and a cross-trained workforce.
• Industry contributes: Application requirements, engineering constraints, decision timelines, system context, validation opportunities, design feedback, and experience turning scientific capability into practice.

How TEMPEST will make the pathway active

1. Challenge-definition workshops: Center and industry teams will translate an application need into a researchable question with a measurable outcome, a defined missing influence, appropriate evidence, and a realistic path to evaluation.
2. Student internships and co-ops: Internships and longer co-op experiences will place students in industry settings where they can connect research methods to design decisions, teamwork, schedules, intellectual property, and professional communication.
   Extended technical exchanges: Longer visits by students, postdoctoral researchers, faculty, or industry scientists will support projects that require sustained access to specialized workflows, data, experiments, or design context.
3. Joint research and review cycles: Partner teams will review assumptions, intermediate results, uncertainty, performance, and usability throughout a project rather than waiting for a completed method to be handed over.
4. Benchmarks, software, and training: Where agreements allow, projects may produce precompetitive benchmarks, documented software components, tutorials, datasets, uncertainty reports, or training modules that can be evaluated and reused.
5. Retreats and partner engagement: Center retreats and partner sessions will bring industry perspectives into broader TEMPEST discussions and help researchers see where a method developed for one application may transfer to another.

What the pathway supports

1. Education mission: Internships and co-ops are educational experiences, not simply placements. Students learn to ask what decision a model supports, explain assumptions to colleagues with different expertise, work within engineering and data constraints, document results for reuse, and recognize when uncertainty makes a prediction unsuitable for action.
2. Research mission: Industry engagement keeps the Center’s basic science connected to consequential questions. Partner feedback can reveal which closure error changes a design, where a faster model is scientifically adequate, what evidence is missing, and whether a transferable method survives outside the conditions in which it was developed.

Examples of progress

• Student internships and co-op experiences
• Partner-defined research questions
• Validated methods with a line of sight to use
• Documented software, benchmarks, and training
• Workforce pathways into science and engineering