Knowledge Base /Experiments & Facilities

🇩🇰 Hartmann Flow & the 1937 Mercury Duct Experiments

Julius Hartmann and Freimut Lazarus's landmark 1937 theoretical and experimental investigation of liquid-mercury duct flow under a uniform transverse magnetic field.

In 1937, Danish physicist Julius Hartmann and his collaborator Freimut Lazarus published Hg-Dynamics I & II, establishing the theoretical and experimental foundation of liquid-metal duct magnetohydrodynamics. Their work provided the first quantitative measurements of Lorentz-force pressure drops, velocity-profile flattening, and magnetic turbulence suppression in a conducting fluid.

Experimental Apparatus & Operating Regime

Hartmann and Lazarus pumped liquid mercury ($\mathrm{Hg}$) at room temperature through electrically insulating rectangular channels and circular pipes placed between the poles of a DC electromagnet:

  • Benchmark duct geometry: Rectangular and square cross-sections of millimetre scale (for example, the widely cited K 33 square duct of $1.14 \times 1.16\text{ mm}^2$ cross-section and $140\text{ mm}$ length) to achieve high aspect-ratio uniformity and measurable pressure gradients.
  • Dimensionless range: Low magnetic Reynolds number ($\mathrm{Rm} \ll 1$, quasi-static limit) with Hartmann numbers up to $\mathrm{Ha} \approx 18\text{–}20$, where $$\mathrm{Ha} = B_0 a \sqrt{\frac{\sigma}{\rho \nu}}$$ measures the ratio of electromagnetic to viscous forces ($a$ is the channel half-width, $\sigma$ electrical conductivity, $\rho$ density, and $\nu$ kinematic viscosity).

Governing Physics & Analytical Solution

$$\rho \nu \frac{\mathrm{d}^2 u}{\mathrm{d}y^2} + \sigma B_0^2 \left(\frac{E_z}{B_0} - u\right) = \frac{\mathrm{d}p}{\mathrm{d}x}$$$$u(y) = u_0 \left(1 - \frac{\cosh(\mathrm{Ha}\,y/a)}{\cosh(\mathrm{Ha})}\right)$$
  • Core flattening: Induced currents $\mathbf{J} = \sigma(\mathbf{E} + \mathbf{u}\times\mathbf{B})$ oppose the core velocity and close through thin wall boundary layers, turning the parabolic Poiseuille profile into a flat plug flow as $\mathrm{Ha}$ increases.
  • Hartmann boundary layers: Along walls perpendicular to $\mathbf{B}_0$, viscous and Lorentz forces balance inside thin Hartmann layers of thickness $\delta_{\mathrm{Ha}} \sim a/\mathrm{Ha}$.

Key Experimental Findings & Legacy

  1. Magnetic Pressure Drop: At high $\mathrm{Ha}$, the friction factor increases linearly with $\mathrm{Ha}$ due to Joule dissipation in the Hartmann layers, matching Hg-Dynamics I in the laminar regime.
  2. Turbulence Suppression (Relaminarization): In initially turbulent mercury runs, increasing $B_0$ damped velocity fluctuations and shifted the pressure-drop curve back toward the laminar Hartmann prediction—later extended to higher $\mathrm{Ha}$ and $\mathrm{Re}$ by Murgatroyd (1953) and Shercliff (1953).
  3. Code Benchmark: Hartmann flow remains the primary analytical verification test for every modern CFD/MHD solver.

References

  1. Hartmann, J. (1937). Hg-Dynamics I: Theory of the Laminar Flow of an Electrically Conductive Liquid in a Homogeneous Magnetic Field. Kgl. Danske Videnskab. Selskab, Mat.-Fys. Medd. 15(6).
  2. Hartmann, J. & Lazarus, F. (1937). Hg-Dynamics II: Experimental Investigations on the Flow of Mercury in a Homogeneous Magnetic Field. Kgl. Danske Videnskab. Selskab, Mat.-Fys. Medd. 15(7).
  3. Shercliff, J. A. (1953). Steady motion of conducting fluids in pipes under transverse magnetic fields. Mathematical Proceedings of the Cambridge Philosophical Society 49(1), 136–144. doi:10.1017/S0305004100028139
  4. Murgatroyd, W. (1953). Experiments on magneto-hydrodynamic channel flow. Philosophical Magazine 44(359), 1348–1354. doi:10.1080/14786441208520404