#corrosionmodeling

Quantum Computing: A New Take on Global Corrosion Modeling

Corrosion Modeling

Worldwide, corrosion costs $2.5 trillion annually. It threatens defense, automotive, and aerospace infrastructure's performance and structural integrity beyond financial losses. Even though standard computational methods have improved corrosion inhibitor design, they often fail to understand the atomistic principles of materials like niobium and magnesium alloys. Due to the closely connected electrical states involved in corrosion, exact simulations on standard supercomputers are mathematically unfeasible.

However, Boeing Research & Technology, HRL Laboratories, and the University of Technology Sydney found that quantum computing may provide the breakthrough. The study team created a hybrid classical-quantum technique to accurately simulate complex corrosion processes. Quantum hardware is used to model tightly linked chemical systems to generate corrosion-resistant materials.

Problem with Niobium-Magnesium Alloys

The study examines niobium-rich refractory alloys for jet engines and magnesium-rich alloys for lightweight aircraft constructions. Due to its strength-to-weight ratio, magnesium is valuable yet corrodes quickly in water. This process's exothermic Hydrogen Evolution Reaction (HER) yields reactive, short-lived intermediates that are difficult to observe experimentally.

High-temperature niobium alloys have high melting points but limited oxidation resistance. To improve durability, scientists must understand alloy matrix oxygen diffusion. Current computational barrier in calculating corrosion rates and oxygen diffusivity is the team's quantum-based ground-state energy estimation (GSEE) of these materials.

Computer Architecture Hybrid

The proposed method uses a Quantum Benchmarking Graph (QBG) to break down complex material simulations into subroutines. This hybrid technique computes initial geometries and multi-configurational electronic structures using high-fidelity quantum calculations and classical pre-processing.

Qubitized Quantum Phase Estimation (QPE), which recovers Hamiltonian spectrum information, is part of the procedure. To accelerate this procedure, the researchers used the pyLIQTR computational tool for the first time in a materials science challenge. This software estimates the hardware resources needed to simulate future quantum machines.

Hardware Needs Variable

Depressingly, the study shows how much quantum technology is needed for industrial corrosion models. Simulations of magnesium and niobium require 2,292 to 38,598 logical qubits, according to the resource study. Instead of noisy intermediate-scale quantum (NISQ) gear, fault-tolerant, error-corrected quantum devices are needed.

Computational complexity is massive. In the largest supercell models, the number of T-gates, indicating quantum circuit depth, ranges from 1.04×1013 to 1.96×1016. These high estimations are necessary for “utility-scale” findings that are better than traditional methods, according to the researchers.

First/Second Quantization

One of the paper's key technical insights is the quantum computer's encoding. Second quantization was compared to plane wave basis first quantization. They found that first quantization boosts basis functions without increasing qubit counts since it is more space-efficient. Second quantization for magnesium alloys performed worse than first quantization even in “worst-case” conditions. Despite hefty elements like tungsten and hafnium, second quantization is competitive at lower energy cutoffs for niobium alloys.

Forward-looking

DARPA-funded technology allows materials to be produced “from first principles” on a computer before being fabricated in a lab. The researchers believe that tensor factorizations or enhanced pseudopotentials could reduce the number of logical qubits needed.