Ke Zhou,Samuel J. Grauer
Abstract:This text describes a method to simultaneously reconstruct flow states and determine particle properties from Lagrangian particle tracking (LPT) data. LPT is a popular measurement strategy for fluids in which particles in a flow are illuminated, imaged (typically with multiple cameras), localized in 3D, and then tracked across a series of frames. The resultant "tracks" are spatially sparse, and a reconstruction algorithm is commonly employed to determine dense Eulerian velocity and pressure fields that are consistent with the data as well as the equations governing fluid dynamics. Existing LPT reconstruction algorithms presume that the particles perfectly follow the flow, but this assumption breaks down for inertial particles, which can exhibit lag or ballistic motion and may impart significant momentum to the surrounding fluid. We report an LPT reconstruction strategy that incorporates the transport physics of both the carrier fluid and particle phases, which may be parameterized to account for unknown particle properties like size and density. Our method enables the reconstruction of unsteady flow states and determination of particle properties from LPT data and the coupled governing equations for both phases. We use a neural solver to represent flow states and data-constrained polynomials to represent the tracks (though we note that our technique is compatible with a variety of solvers). Numerical tests are performed to demonstrate the reconstruction of forced isotropic turbulence and a cone-cylinder shock structure from inertial tracks that exhibit significant lag, streamline crossing, and preferential sampling.
What problem does this paper attempt to address?
### Problems the paper attempts to solve
The paper aims to solve the problem of simultaneously reconstructing the flow - field state and determining particle properties from Lagrangian particle tracking (LPT) data. Specifically, existing LPT reconstruction algorithms assume that particles follow the fluid motion completely, but this assumption fails in the case of inertial particles. Inertial particles may exhibit lag or ballistic motion and have a significant momentum impact on the surrounding fluid. Therefore, the paper proposes a new LPT reconstruction strategy that combines the transport physical characteristics of the carrier fluid and the particle phase and can parameterize unknown particle properties (such as size and density), thereby achieving the reconstruction of unsteady flow - field states and the determination of particle properties.
### Key points
1. **Limitations of existing methods**:
- Existing LPT reconstruction algorithms assume that particles are ideal trackers, that is, they passively follow the motion of the surrounding fluid.
- For inertial particles, this assumption does not hold because inertial particles may exhibit lag or ballistic motion and may even have a significant impact on fluid behavior.
2. **Characteristics of the new method**:
- Combines the transport physical characteristics of the carrier fluid and the particle phase.
- Can parameterize unknown particle properties such as size and density.
- Represents the flow - field state using a neural network and represents particle trajectories using data - constrained polynomials.
- Determines the flow - field state, particle trajectories, and particle properties by minimizing physical residuals and boundary conditions.
3. **Application examples**:
- Through numerical tests, the application of this method in reconstructing forced isotropic turbulence and cone - cylinder shock structures is demonstrated.
- Inertial particle trajectories exhibit obvious lag, streamline crossing, and preferential sampling phenomena, and this method can effectively handle these complex situations.
### Formulas
- **Particle relaxation time**:
\[
\tau_p=\frac{\rho_p d_p^2}{18\mu_f}
\]
where \(\rho_p\) is the particle density, \(d_p\) is the particle diameter, and \(\mu_f\) is the fluid viscosity.
- **Stokes number**:
\[
\text{St}=\frac{\tau_p}{\tau_f}
\]
where \(\tau_f\) is the characteristic flow time scale.
- **Total loss function**:
\[
L_{\text{total}}(\theta_f,\Theta_p,\Psi)=\chi_1 L_{\text{flow}}^{\text{phys}}(\theta_f)+\chi_2 L_{\text{part}}^{\text{phys}}(\Theta_p,\Psi)+\chi_3 L_{\text{bound}}(\theta_f)
\]
- **Flow - physics loss**:
\[
L_{\text{flow}}^{\text{phys}}(\theta_f)=\frac{1}{d_f|V\times T|}\int_T\int_V\|e_f(x,t;\theta_f)\|^2_2\,dV\,dt
\]
- **Particle - physics loss**:
\[
L_{\text{part}}^{\text{phys}}(\Theta_p,\Psi)=\frac{1}{d_p n_p}\sum_{k = 1}^{n_p}\int_{T_k}\|e_p(t;\theta_p^k,\psi_k)\|^2_2\,dt
\]
- **Boundary loss** (e.g., no - slip wall):
\[
L_{\text{bound}}(\theta_f)=\frac{1}{d_u|A\times T|}\int_T\int_A\|u_f(x,t;\theta_f)\|^2_2\,dA\,dt
\]
Through these methods and