AI two-phase-flow simulation platform
Combines OpenFOAM customization, micro-scale DNS, engineering solvers and high-fidelity correlations to turn physics and test data into engineering rules.
Technology
The technical stack links simulation, hardware-in-the-loop validation, control optimization and real-condition testing across compute, storage and EV testing scenarios.
Combines OpenFOAM customization, micro-scale DNS, engineering solvers and high-fidelity correlations to turn physics and test data into engineering rules.
Prototype parameters
Performance envelope
Scans design variables, geometry, materials and boundary conditions to generate reduced-order performance envelopes for faster prototype iteration.
Illustrative control interface · Example values
Load
78%
Cooling
64%
Efficiency
92%
Uses efficiency maps, operating constraints, historical power, task queues, utilization and environment variables for cooling feed-forward and efficiency optimization.
Uses test benches, cold-plate samples, TTV thermal vehicles and automated reports to pull validation ahead of delivery.
System blueprint
Connect solar, storage and compute equipment to coordinate power supply and cooling.

Ecosystem
XJTU national two-phase-flow research platform
Joint thermal-management materials lab · Foshan Hebang New Materials
Advanced low-GWP refrigerant joint lab · Zhejiang Juhua
Full-lifecycle scenario test lab · Shenzhen Precise
Balanced flow allocation
Energy scheduling algorithm
Short/long-term compute-load forecast
We bring the demo kit and the test rig to you — rebuilt for your scenario from cold plate up to rack-level CDU.