Surface-Streamline Aligned Coordinates

• Surface-streamline aligned coordinate systems are curvilinear coordinates with axes tangent and normal to vector field streamlines, facilitating reformulation of governing equations.
• They enable precise grid generation and analytical treatments in convection, boundary layer theory, and plasma simulations by conforming to complex surface and flow geometries.
• The coordinate system leverages scalar flux functions and numerical integration to manage nonorthogonality and singularities in domains with complex topology or X-point geometries.

A surface-streamline aligned coordinate system is a curvilinear coordinate system designed so that two of its axes are tangent and normal to the streamlines of vector fields (such as velocity or magnetic fields) on a given surface. These systems enable the reformulation of governing equations (e.g., convection-diffusion, Navier–Stokes) in a basis that is geometrically adapted to complex boundary layers, streamline topology, or flux-aligned surfaces. They are widely used in analytical and numerical treatments of convective transport, grid generation, boundary layer theory, and plasma simulations where precise fidelity to surface and flow geometry is essential.

1. General Construction Principles

Surface-streamline aligned coordinates are constructed by identifying a scalar stream- or flux-function whose level sets coincide with the desired streamlines or flux surfaces. The principal coordinate (often denoted $\tau$, $s$, $v$, or $\eta$) increases along the direction of the tangent vector to the streamlines, while the second coordinate ($C$, $n$, $u$, or $\zeta$) is constant on streamlines and varies transversely, labeling different streamlines or surfaces. For three-dimensional problems, a third coordinate ($y$, distance from the surface, or an angular variable) completes the system.

Key mappings and the theoretical framework for stream- and surface-aligned coordinates include:

• For a two-dimensional incompressible flow, the streamfunction $\psi(x, y)$ defines streamlines as $\psi = \text{const}$, and the transverse coordinate is constructed, possibly requiring an integrating factor for orthogonality if viscosity breaks the Cauchy–Riemann conditions (Leble et al., 2017).
• On a curved surface, such as a drop in shearing flow, surface streamline-aligned curvilinear coordinates can be constructed by solving the tangential velocity ODEs and parameterizing the surface in terms of streamline labels and distances along them (Narayanan et al., 24 Jan 2026).
• For magnetic field-aligned coordinates in plasma physics, one coordinate is along the field (streamline) while the transverse coordinates match arbitrary surface geometries (e.g., divertor plates), and nonorthogonality is explicitly retained and handled in the metric (Leddy et al., 2016).

The resulting coordinates may be orthogonal or nonorthogonal, as dictated by flow topology, surface geometry, or numerical convenience.

2. Mathematical Formulation and Metric Properties

The successful construction and application of streamline-aligned coordinates rely on careful consideration of the metric tensor, Jacobian, and basis vectors. The metric encodes both curvature and possible nonorthogonality, which is crucial for accurate differentiation and integration on the mapped domain.

For the nonorthogonal $(C, \tau, y)$ system on a sphere (Narayanan et al., 24 Jan 2026):

• $C$ labels surface streamlines (constant along each streamline).
• $\tau$ increases monotonically along a streamline.
• $y = r - 1$ describes radial distance from the surface.

The covariant basis vectors and metric on the surface $r=1$ are

$$e_C \propto \hat{\theta}, \qquad e_\tau \propto \cos\beta\,\hat{\theta} + \sin\beta\,\hat{\phi}, \qquad e_y = \hat{r}$$

where $\beta$ is the local skew angle. The metric tensor $g_{ij}$ and surface area element $\sqrt{\det g}$ are explicitly constructed, and the nonorthogonality manifests in off-diagonal metric terms and a $\sin \beta$ contribution to the Jacobian.

In two dimensions, orthogonality ($g_{\psi\phi} = 0$) may require an integrating factor $\mu(x, y)$, solving $\nabla\psi \cdot \nabla \mu + \mu\,\Delta\psi = 0$ to enforce exactness and maintain a well-defined transformation (Leble et al., 2017). For more complex geometries, e.g., with X-points, orthogonality in the chosen metric requires the Laplacian of the flux function to vanish at those singular points, possibly necessitating a modified "monitor metric" (Wiesenberger et al., 2018).

For field-aligned coordinates in tokamak simulations,

$(u, v, w)$

are constructed, with $v$ aligned to the magnetic field and $(u, w)$ forming an arbitrary surface normal to the field; the metric tensor captures all cross terms induced by relaxing poloidal-plane orthogonality and enables alignment with realistic surfaces such as divertor plates (Leddy et al., 2016).

3. Algorithmic Implementation and Numerical Grid Generation

Numerical construction of surface-streamline aligned coordinates typically proceeds via integration along vector fields and/or solution of elliptic PDEs for adaptivity and regularity.

The grid generation workflow, as in (Wiesenberger et al., 2016), includes:

• Definition of a domain via two level-sets of a scalar function $\psi(x, y)$.
• Construction of an initial (often orthogonal) grid via streamline integration (solving ODEs for vector fields tangent and normal to $\psi$ level sets).
• If enhanced control or adaptivity is required, solve an elliptic PDE with conductivity tensor $\chi$ (monitor metric) to redistribute grid density (e.g., concentrating cells near boundary layers), with boundary conditions set for periodicity and Dirichlet constraints as appropriate.
• Final coordinates $(u, v)$ are built via further streamline integration in the adapted frame for smoothness and geometric conformity.
• The Jacobian and metric are computed at all grid points for correct numerical stencils in PDE solvers.

For X-point geometries where $\nabla\psi=0$, block-structured grids with explicit topology must be used, and localized mesh refinement is applied to maintain accuracy where metric and cell sizes diverge (Wiesenberger et al., 2018).

3D implementations in plasma codes involve grid generators (e.g., Hypnotoad in BOUT++), which use prescribed surface shapes and local angles to compute nonorthogonality factors and construct the full metric offline (Leddy et al., 2016).

4. Reformulation of Governing Equations and Boundary Layer Analysis

Casting governing equations in streamline-aligned coordinates yields significant analytical and computational simplifications, especially in boundary layer and strong convection regimes.

In convective-diffusive mass transfer around drops (Narayanan et al., 24 Jan 2026):

• The convection-diffusion equation in $(C, \tau, y)$ becomes

$$u_r \partial_y \Theta + (u_\tau / k) \partial_\tau \Theta = (1/Pe) r^{-2} \partial_y (r^2 \partial_y\Theta)$$

In the thin $O(Pe^{-1/2})$ boundary layer, only normal diffusion survives, and all tangential derivatives simplify due to $u_C=0$, as $C$ labels streamlines. The similarity reduction then allows analytical or quasi-analytical solution of the boundary layer problem.

In boundary-layer turbulence, a locally streamline-normal $(s, n)$ frame reduces the RANS equation to a form where explicit curvature and acceleration terms enter as Christoffel-like coefficients, revealing the dominant balance of pressure gradient, Reynolds shear-flux, and curvature in the pre-separation region (Prakash et al., 2023).

Orthogonal transformations in free-convection problems allow for analytic relations between thermal gradients, velocity inclination, and the velocity modulus, with the integrating factor $\mu$ incorporating temperature field variation (Leble et al., 2017).

In tokamak edge plasma simulation, retaining exact alignment of one coordinate with the magnetic field allows the use of coarse discretization along the field and high-resolution adaptivity transverse to it, matching surface and field geometry while maintaining second-order accuracy in metric-induced finite-difference stencils (Leddy et al., 2016).

5. Topological Considerations and Handling of Singularities

Streamline topology—especially the presence of closed, open, spiraling, or separatrix surfaces—dictates the structure and mapping intervals for surface-streamline aligned coordinates.

• For 3D flows over spheres (drops), whether the parameter $A^2$ is positive or negative determines if surface streamlines are open (non-spiralling) or form closed spirals; integration limits and the treatment of coordinate intervals must reflect this topology (Narayanan et al., 24 Jan 2026).
• In domains with X-points (e.g., magnetic separatrices), the existence and smoothness of a streamline-aligned, orthogonal system depend on the vanishing of the Laplacian of the flux function at the singular point. Otherwise, a monitor metric is imposed locally, and block-structured grids with mesh refinement are required (Wiesenberger et al., 2018).
• For multiply-connected or complex with boundaries, domains may be cut into ring-patches or multi-block structures, gluing local coordinate maps to accommodate global topology (Wiesenberger et al., 2016).
• Nonorthogonality, as deliberately retained in some systems (e.g., for divertor-aligned grids), is handled by allowing the off-diagonal metric components and by constructing the coordinate system so that metric singularities are avoided except at controlled locations where refinement is concentrated (Leddy et al., 2016).

6. Applications, Performance, and Validation

Surface-streamline aligned coordinate systems are foundational in:

Application Domain	Physical Context	Key References
Boundary layer theory, scalar mass/heat transfer	Drops, thin films, complex flows	(Narayanan et al., 24 Jan 2026, Leble et al., 2017)
Turbulence and RANS analysis	Wall-bounded flows, pressure-gradient regions	(Prakash et al., 2023)
Tokamak edge plasma simulation	Magnetic field-aligned, divertor geometry	(Leddy et al., 2016, Wiesenberger et al., 2016)
Grid generation for PDEs on arbitrary domains	Elliptic grid, X-point structures	(Wiesenberger et al., 2016, Wiesenberger et al., 2018)

Validation of these coordinate systems is achieved via:

• Convergence studies with analytic manufactured solutions confirming expected order (typically second order) for numerical operators in complex non-orthogonal grids (Leddy et al., 2016).
• Direct measurement of grid quality metrics: maximum and minimum cell sizes, aspect ratios, and area conservation (Wiesenberger et al., 2016).
• Restoration of expected convergence in elliptic PDE solutions by localized mesh refinement at singularities (e.g., X-points) where metric or Jacobian diverges (Wiesenberger et al., 2018).
• Physical fidelity of boundary conditions and wall interactions in plasma and fluid simulations, especially at boundaries conforming to complex surface geometries (Leddy et al., 2016).

A key implication is that retaining perfect streamline (or field) alignment allows for significant coarsening along integral curves while maintaining physically essential resolution and accuracy perpendicular to those curves, with only minor computational overhead for additional metric calculations.

7. Limitations and Practical Considerations

Several critical considerations arise in the construction and use of surface-streamline aligned coordinates:

• The underlying scalar streamline/flux function $\psi$ must be single-valued and satisfy $\nabla\psi \ne 0$ except at controlled singularities, else the mapping degenerates (Wiesenberger et al., 2016).
• Orthogonality is not always possible, especially near X-points or other degeneracies—in those cases, a modified metric and/or block-structured, locally refined grids are necessary (Wiesenberger et al., 2018).
• For implementation, storage demands can be optimized by exploiting the tensor-product structure of the mesh, storing 1D metric stencils per direction (Wiesenberger et al., 2016).
• In viscous or boundary-layer-dominated regimes, the mapping and integrating factor may depend sensitively on temperature or other field gradients, requiring consistent closure equations (Leble et al., 2017).
• Choice of weight functions and monitor metrics in grid adaptation directly affects cell clustering, essential for resolving sharp layers or reaction zones (Wiesenberger et al., 2016).
• In practical large-scale simulations (e.g., BOUT++ tokamak codes), the additional cost of computing metric transformations and derivatives remains minor ($<1\%$ of runtime) relative to the gains in geometric fidelity and accuracy (Leddy et al., 2016).

Surface-streamline aligned coordinate systems form a unifying geometric and computational approach for expressing and solving physical transport phenomena where flow, field, or flux directions are dynamically or topologically complex. Their theoretical construction, metric formulation, and grid-generation algorithms are mature, robust, and central to contemporary computational physics and engineering (Narayanan et al., 24 Jan 2026, Wiesenberger et al., 2016, Leddy et al., 2016).