⚛️ quantum physics

Quantum lattice Boltzmann method via density-matrix encoding for fluid simulation with wall boundary conditions

This paper proposes a quantum lattice Boltzmann method utilizing density-matrix encoding and Kraus operators to effectively implement solid wall, inlet, and outlet boundary conditions, thereby enabling accurate quantum simulations of complex two-dimensional fluid flows.

Original authors: Hao Su, Boyuan Wang, Yue Yang

Published 2026-09-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Hao Su, Boyuan Wang, Yue Yang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine trying to predict how air flows over a new airplane wing or how blood moves through a complex network of vessels. For decades, scientists have relied on powerful classical computers to solve the mathematical equations that govern these fluid motions. These simulations are essential for engineering and science, but they hit a wall when the problems become too large or the shapes too intricate. The calculations required can be so immense that even the fastest supercomputers struggle to finish them in a reasonable time. This has led researchers to look toward a different kind of computing power: quantum computing. Unlike traditional computers that process information in a straight line of zeros and ones, quantum machines use the strange rules of quantum mechanics to handle many possibilities at once. In theory, this could allow them to solve fluid dynamics problems exponentially faster. However, there is a major hurdle. The laws of quantum mechanics are very strict about how information can be manipulated, requiring processes to be reversible and perfectly balanced. The messy, real-world behavior of fluids, especially when they hit a solid wall or enter a pipe, often involves irreversible changes that seem to break these strict quantum rules.

A team of researchers at Peking University has now proposed a way to bridge this gap, offering a new method to simulate fluid flow on a quantum computer that can handle solid walls and complex shapes. Their work focuses on a technique called the lattice Boltzmann method, which models fluids not as a continuous substance, but as a vast collection of tiny particles moving and colliding on a grid. While this approach is already popular on classical computers, adapting it for quantum machines has been difficult, particularly when it comes to boundaries. In the real world, when a fluid hits a solid wall, it stops moving at the surface and slides along it; this is known as the no-slip condition. Previous attempts to simulate this on a quantum computer were limited to simple, repeating patterns or periodic boundaries, effectively ignoring the complex, solid shapes found in real engineering problems. The researchers realized that to make this work, they needed a more flexible way to represent the state of their fluid particles, one that could accommodate the "messy" rules of boundaries without breaking the quantum system.

To solve this, the team developed a new algorithm that uses a mathematical tool called a density matrix. Think of this as a way to describe the fluid that is more general and forgiving than the standard methods used in quantum physics. This flexibility allowed them to design a specific set of steps that mimic what happens when fluid hits a wall. Instead of trying to force the wall into the quantum rules, they introduced a clever intermediate step before the fluid particles move to their next position. In this step, the algorithm swaps the directions of the particles that are about to hit a wall with particles that are already at the wall, effectively bouncing them back in place before the movement happens. This "component exchange" ensures that the fluid stops at the wall exactly as it should, without needing to change the fundamental rules of how the fluid moves. They also figured out how to set the speed of the fluid entering and leaving the simulation area by preparing special helper bits of information and swapping them into the main system, much like setting the initial conditions for a race.

The researchers tested their new method by running simulations on a classical computer that mimics a quantum device. They started with simple, known flows, such as fluid moving through a channel where the speed decreases over time, and confirmed that their quantum-inspired algorithm matched the correct mathematical answers. They then moved on to more challenging scenarios, including fluid flowing past a backward-facing step, where the channel suddenly widens, creating a swirling eddy behind the step. They also simulated fluid flowing around a cylinder and even around the complex shapes of the letters "P," "K," and "U." In every case, the results were accurate. For the backward-facing step, the simulation correctly predicted the length of the swirling zone behind the step, matching the results of established classical software. When they simulated flow around the letters, the algorithm successfully captured the intricate swirls and vortices forming in the gaps between the letters and trailing behind them, proving that the method can handle arbitrary, complex geometries.

The study demonstrates that it is possible to simulate realistic fluid boundaries on a quantum framework, a significant step forward for the field. The researchers validated their approach by comparing their results against known analytical solutions and high-fidelity classical simulations, finding close agreement in velocity profiles and flow structures. However, they are careful to note that this work is currently a simulation of a quantum algorithm, not a run on actual quantum hardware. The method still requires a significant number of steps to ensure the particles remain independent and the simulation stays stable, which could be a challenge for future real-world quantum devices. Despite these limitations, the work provides a clear blueprint for how to handle the most difficult part of fluid simulation—solid walls and complex shapes—within the strict constraints of quantum computing. By showing that these boundaries can be enforced naturally and accurately, the team has opened the door for quantum computers to eventually tackle the large-scale, complex fluid problems that are currently out of reach for classical machines.

Technical Summary: Quantum Lattice Boltzmann Method via Density-Matrix Encoding for Fluid Simulation with Wall Boundary Conditions

Problem Statement
While the Quantum Lattice Boltzmann Method (QLBM) offers a promising paradigm for efficient fluid dynamics simulations by leveraging quantum parallelism and exponential memory efficiency, its practical application has been hindered by two primary challenges: the incompatibility of non-linear collision terms with unitary quantum operations and, more critically, the difficulty of imposing realistic boundary conditions. Previous QLBM frameworks, such as the ensemble-based approach by Wang et al. [37], were largely restricted to periodic boundary conditions. The inherent unitary and reversible nature of quantum computing conflicts with the non-linear, irreversible nature of the Navier-Stokes equations and the specific requirements of solid wall boundaries (e.g., no-slip conditions) and inlet/outlet velocity profiles, particularly for arbitrary geometries.

Methodology
The authors propose a generalized QLBM framework that addresses these limitations through three core methodological advancements:

Density-Matrix Encoding: Moving beyond pure-state encoding, the authors encode the probability distribution of the lattice gas ensemble into a density matrix ( $\varrho$ ). This formulation allows for the representation of non-unitary operations via Kraus operators. By satisfying the conditions of Completely Positive Trace-Preserving (CP-TP) maps, the algorithm can implement stochastic processes (such as collision and boundary reflection) that are mathematically equivalent to the classical LBM but compatible with quantum hardware through the Stinespring dilation theorem.
Component Exchange for Wall Boundaries: To enforce no-slip wall conditions on arbitrary geometries without modifying the streaming operator, the authors introduce a "component exchange" step immediately preceding the streaming step.
- Grid points are classified into boundary, adjacent, boundary-inner, and fluid-inner categories.
- For adjacent points near a wall, the algorithm splits the particle state into branches. Branches streaming toward the fluid proceed normally, while those streaming toward the wall are reflected back to the same grid point with reversed velocity configurations.
- This is achieved via a pre-computed permutation (SWAP operations) that exchanges velocity components before transport occurs, naturally satisfying the halfway bounce-back rule.
Inlet/Outlet Conditions via Ancilla Registers: Inlet and outlet velocity profiles are imposed using controlled SWAP operations. An ancilla register is prepared in the target velocity distribution state. A controlled SWAP exchanges the velocity register of the target grid points with the ancilla register, effectively overwriting the local distribution with the prescribed profile. The ancilla is then traced out.

The overall algorithm iterates through a sequence of: (1) Component Exchange, (2) Streaming, (3) Collision, (4) an H-step (to restore independence of velocity distributions and maintain thermodynamic equilibrium), and (5) Inlet/Outlet enforcement.

Key Contributions

General Boundary Scheme for Static Walls: The paper presents a QLBM algorithm capable of handling static solid wall boundaries of arbitrary geometry alongside inlet and outlet conditions, overcoming the periodic boundary restriction of prior ensemble-based QLBM. The authors explicitly note that moving boundary problems remain entirely unaddressed in the current framework and would require special treatment not yet implemented.
Non-Unitary Implementation: It demonstrates how density-matrix encoding and Kraus operators can be utilized to implement necessary non-unitary fluid dynamics operations (collision and boundary reflection) on quantum devices.
Algorithmic Design: The specific design of the "component exchange" step allows the standard streaming circuit to remain unmodified, simplifying the circuit structure for complex geometries.
Complexity Analysis: The authors provide a resource analysis indicating that the inclusion of boundary treatments does not alter the asymptotic gate complexity of the original QLBM, as the streaming step remains the dominant term ( $O(2^q n^2)$ ).

Simulation Results
The algorithm was validated using a quantum emulator on a classical computer across several 2D benchmark flows:

Decaying Poiseuille Flow: The simulation accurately recovered the time-decaying trigonometric velocity profiles and enforced no-slip conditions at the walls, showing convergence with grid refinement.
Backward-Facing Step Flow: The method successfully captured flow separation and recirculation zones. The dimensionless recirculation length ( $L_r/h$ ) showed close agreement with OpenFOAM reference solutions (1.375 vs. 1.1563), with relative errors decreasing as the Reynolds number increased.
Flow Past a Cylinder: The algorithm handled curved boundaries effectively, capturing low-velocity regions behind the cylinder with relative errors of 12.53% to 17.34% compared to reference data.
Complex Geometry ("PKU" Obstacle): The method was tested on a flow past a complex shape (the letters "PKU"), successfully capturing streamlines and vortex formation in the gaps and wake, demonstrating extensibility to non-canonical geometries.

Significance and Claims
The paper claims that this work represents a significant step toward the practical deployment of QLBM for real-world Computational Fluid Dynamics (CFD). By resolving the boundary condition bottleneck for static walls, the proposed method extends the applicability of quantum fluid simulation from idealized periodic domains to problems involving solid walls and complex geometries. The authors emphasize that their approach maintains the theoretical advantages of quantum computing (parallelism and memory efficiency) while introducing a flexible circuit design capable of handling the non-unitary requirements of physical boundaries.

However, the authors remain modest regarding immediate practical deployment. They acknowledge that the H-step introduces significant complexity and circuit depth, which may currently undermine quantum speedup for moderate-sized problems on near-term hardware. Furthermore, the current framework addresses only static boundaries, leaving moving or deformable boundaries for future investigation. The work is presented as a foundational algorithmic extension rather than a fully realized hardware demonstration.

More like this