Quantum Solvers: Predictive Aeroacoustic & Aerodynamic modeling
Nis-Luca van Hülst, Theofanis Panagos, Greta Sophie Reese, Shahram Panahiyan, Tomohiro Hashizume · 2025
Read Quantum Solvers: Predictive Aeroacoustic & Aerodynamic modeling by Nis-Luca van Hülst, Theofanis Panagos, Greta Sophie Reese, Shahram Panahiyan, Tomohiro Hashizume (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-21683.txt.
Introduction
Five authors present a technical report on quantum methods for simulating airflow and noise around aircraft and vehicles. The report reproduces their winning entry to a 2024 industry quantum computing challenge run by Airbus and the BMW Group. It covers quantum-inspired methods and mixed methods that pair quantum and classic computing for hard simulation tasks. The materials are kept as submitted, as an archival record, with one added note pointing to the later published paper that grew out of the challenge entry.
Catalog record
| CREATOR | Nis-Luca van Hülst, Theofanis Panagos, Greta Sophie Reese, Shahram Panahiyan, Tomohiro Hashizume |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-21683
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Quantum Solvers: Predictive Aeroacoustic & Aerodynamic modeling Nis-Luca van Hülst, Theofanis Panagos, Greta Sophie Reese, Shahram Panahiyan, Tomohiro Hashizume Abstract I Preamble The following pages comprise our original Phase II submission for the Airbus x BMW Group Quantum Computing Challenge 2024 . To maintain the strict authenticity of the original competition entry, this content is presented in its exact handed-in format, commencing on the next page, aside from the specific citations to our own work, which were unpublished at the time of submission but now refer to the two following preprints: • Siegl et al. , “Tensor-Programmable Quantum Circuits for Solving Differential Equations” (2025) Siegl et al. 2025 • van Hülst et al. , “Quantum-Inspired Tensor-Network Fractional-Step Method for Incompressible Flow in Curvilinear Coordinates” (2025) van Hülst et al. 2025 Together, these publications substantiate the technical soundness and broader impact of the methods documented in this submission. II Submission Summary Recent decades have witnessed outstanding advances in technological and theoretical aspects of quantum computing Jaksch et al. 2000 ; Bluvstein et al. 2022 ; Bluvstein et al. 2024 ; Evered et al. 2023 . In particular, quantum computing platforms are advancing at such a pace that developing quantum algorithms and circuits for implementing problems have become a priority Bluvstein et al. 2024 ; AI and Collaborators 2024 . Motivated by this, in Phase 1 van Hülst et al. 2024 , which is an integral part of the manuscript and is assumed to have been read, we proposed a quantum-inspired method for solving the Navier-Stokes equations. We have shown that our method requires exponentially smaller computational resources \mathcal{O}(\log_{2}(N)) compared to conventional computational fluid dynamics (CFD) methods performed on N grid points. In this submission, we further extend our proposed methodology in two ways: I) We extend our method from Phase 1 to a quantum algorithm—Tensor-Programmable Variational Quantum Algorithm (TPVQA) Siegl et al. —by directly translating a general Matrix Product Operator (MPO) O Schollwoeck 2011 into a set of unitary quantum gates U_{O} Termanova et al. 2024 ( Figure 1 a.,b.) within a Variational Quantum Algorithm (VQA) setting Peruzzo et al. 2014 ; McClean et al. 2016 ; McArdle et al. 2019 ; Lubasch et al. 2020 ; Cerezo et al. 2021 . We test its performance by solving problem case 1—the point source propagation—on an (emulated) quantum computer. With this, we were able to reproduce the local pressure amplitudes which agreed perfectly with the analytical solution ( Figure 1 b.). This approach enables programming quantum algorithms using the flexible operator representation through MPOs, while only requiring to measure \mathcal{O}(1+\lceil\log_{2}\zeta\rceil) qubits, where \zeta is the maximum bond dimension of the MPOs. With this method, we further improve the scalability by polynomial in \log_{2}N . II) We extend our method from Phase 1 for solving problem case 2 —a flow around an immersed object—using curvilinear coordinate method ( Figure 1 c.,d.), allowing for an overall algorithmic complexity of \mathcal{O}(\chi^{3}) . Here \chi denotes the maximum bond dimensions of the fluid fields in the Navier-Stokes equations. We test its performance by simulating a flow around a cylinder (Fig. 1 c.), demonstrating that our Matrix Product State(MPS)-based method gives accurate results while allowing for strong compression. We conducted a benchmark with industrial-grade CFD solvers and obtained a relative error of less than 0.1\% with a compression factor >10 for the horizontal force exerted on the cylinder. We extended our method for generating grids of general shapes. In Fig. 1 d., we show a generated grid around NACA0012 aerofoil Ladson 1988 . These methods can readily be adapted to TPVQA for a a speed-up of \mathcal{O}(\mathrm{poly}(\chi)) . The integration of time evolution and grid generation is a focus of our future research. Our proposal offers an exponential reduction in required computational resources, hence enables larger-scale CFD simulations. It presents industrial opportunities such as faster optimization, design processes, and data analysis. Furthermore, it reduces the environmental impact of computing by reducing CO 2 emissions, thus contributing to sustainability. Considering that with just of order 100 s of error-corrected logical qubits, the computational ability of a quantum computer surpasses that of the world’s largest supercomputer, thus, further advancements in quantum computing technologies are anticipated Russell . This can be seen in the experimental realizations of error-corrected logical qubits Bluvstein et al. 2024 ; AI and Collaborators 2024 , indicating that we are at the dawn of the era of fully utilizing the exponential advantage of quantum computers. Figure 1: Case I: a. Translation of matrix-vector operation to a quantum circuit. In Phase 1, a field \ket{u} and an operator O in matrix-vector form (top) is compressed into MPSs and MPOs (yellow and green, bottom right). In phase 2 (bottom left), we translate of MPO to a quantum circuit U_{O} for use in the realm of TPVQA that solves for variational problem parameterized by \vec{\theta} . Here, the unitary U(\vec{\theta}) encodes the time evolution of \ket{u} when its optimal. b. Simulation of the 2D wave equation with TPVQA. (top) Pressure field at t=0.1 [s]. (bottom) Local pressure amplitudes along the indicated diagonal in the top panel (square) and their analytical predictions (solid). Case II: c. Flow around a cylinder. The steady-state flow around an immersed cylinder simulated with quantum-inspired method. d. Grid generation with the quantum-inspired approach. Body-fitted grid of NACA0012 aerofoil, generated with our quantum-inspired method. Details of the simulations in this figure are provided in the text, App. E , and App. F . III Detailed Explanation In this phase, we solve both problems from case 1 and case 2 of the challenge using a quantum-classical hybrid algorithm for case 1 and a quantum-inspired approach for case 2, by extending the proposed methodology in phase 1 van Hülst et al. 2024 for further efficiency and accuracy. Problem case 1: In this section, we study the propagation of point source using our newly developed Tensor-Programmable VQA (TPVQA) Siegl et al. 2025 , which leverages the flexible operator representation through MPOs integrated with a powerful VQA framework. We show that our approach yields a highly accurate evolution of a fluid field with measurement costs that do not increase with the discretization size. Problem case 2: In this section, we investigate the steady flow around an immersed cylinder. We use a quantum-inspired MPS-based approach that utilizes curvilinear coordinates method. We demonstrate that this approach provides an efficient implementation of this conventional and widely adopted CFD method. Subsequently, we present the results of our approach for the immersed cylinder and compare their accuracy with those obtained from industrial CFD software. This approach is readily translatable into a quantum algorithm through TPVQA. III.1 Problem case 1 Description of the problem.— We start by introducing the Linearized Euler Equations (LEE) for inviscid fluid in 2D with a point source F(x,y,t) and an absorbing boundary layer \gamma(x,y) (App. A ) Israeli and Orszag 1981 ; Mani 2010 ; Mani 2010 ; Perić and Abdel-Maksoud 2018 ; Carmigniani and Violeau 2018 as we solved in Phase 1 van Hülst et al. 2024 with quantum-inspired approach \displaystyle\partial_{t}p \displaystyle=-\bar{\rho}c^{2}\left(\partial_{x}u+\partial_{y}v\right)-\bar{u}\partial_{x}p+F(x,y,t)-\gamma(x,y)p, \displaystyle\partial_{t}u \displaystyle=-\partial_{x}p/\bar{\rho}-\bar{u}\partial_{x}u-\gamma(x,y)u, \displaystyle\partial_{t}v \displaystyle=-\partial_{y}p/\bar{\rho}-\bar{u}\partial_{x}v-\gamma(x,y)v. (1) Here, p=p(x,y,t)=p(t) [Pa] is the pressure field fluctuation 1 1 1 Here, for the sake of brevity, we use p(t) instead of p(x,y,t)