Collapse of turbulence in optimised curved pipe flow

By combining local streamwise curvature with an oval cross-section, the researchers demonstrated a passive method to collapse turbulence and significantly reduce pressure loss in curved pipes.

The Gist

The "Smooth Sailing" Pipe: How to Calm a Storm Inside a Tube

Imagine you are trying to push a massive crowd of people through a long, winding hallway. If everyone is walking calmly in straight lines, the movement is efficient. But if the hallway suddenly turns a sharp corner, people start bumping into each other, swirling around, and creating chaotic "eddies." In the world of physics, this chaos is called turbulence.

Turbulence is a huge problem for engineers. It’s like trying to push honey through a straw while someone is shaking the straw—it takes a massive amount of energy (and money) to overcome that internal friction.

A team of researchers has just discovered a clever way to "calm the storm" inside curved pipes using nothing but clever geometry.

---

The Problem: The "Spinning Top" Effect

When fluid (like water or oil) hits a bend in a pipe, it doesn't just turn the corner; it starts to swirl. This creates something called Dean Vortices—think of them like two giant, invisible spinning tops rotating inside the pipe.

These "spinning tops" are troublemakers. They grab the chaotic energy of the turbulence and whip it around, making the friction even worse. Usually, if you want to stop turbulence, you have to use expensive machines or chemical additives. The researchers wanted to find a "passive" way—meaning a way to do it just by changing the shape of the pipe itself.

The Solution: The "Wide Turn" Strategy

The researchers used a supercomputer to play a game of "shape optimization." They asked the computer: "How can we change the shape of this 180-degree bend to make the fluid flow as smoothly as possible?"

The computer came back with a two-part "magic formula":

1. The "Hairpin Turn" (Increased Curvature)

Instead of a gentle, lazy curve, the computer designed a bend that gets much sharper in the middle (like a hairpin turn on a mountain road).

• Why it works: Surprisingly, a very sharp turn actually helps "squeeze" the turbulence out of certain parts of the flow, acting like a stabilizer that prevents the chaotic streaks from forming.

2. The "Oval Egg" (Cross-Sectional Change)

In a normal pipe, the hole is a perfect circle. The computer decided to squash the pipe into an oval shape right at the sharpest part of the bend.

• Why it works: Imagine trying to spin a top in a narrow hallway versus a wide ballroom. In the narrow hallway, you hit the walls and create chaos. In the wide ballroom, you have more room to move without causing a ruckus. By making the pipe wider in one direction (ovalizing it), the researchers "weakened" those giant spinning Dean Vortices. They gave the fluid more "breathing room," which prevented the swirls from gaining enough strength to keep the turbulence alive.

The Result: A Massive Energy Win

By combining the sharp turn with the oval shape, the researchers achieved something incredible: Relaminarization.

This is a fancy way of saying they turned a raging, chaotic storm of water back into a smooth, calm stream.

The numbers are impressive:

• Compared to a standard bend, this new shape reduced pressure loss (friction) by 53%.
• Compared to a perfectly straight pipe, it was still 36% more efficient than what we usually expect.

Why does this matter to you?

Every time you turn on a tap, or when a massive oil tanker moves fuel across the ocean, or when blood flows through your aorta (which is a natural curved pipe!), energy is being lost to turbulence.

If we can design pipes, heat exchangers, and even medical implants using these "optimized shapes," we can move fluids much more easily. It means less electricity used by pumps, lower costs for transporting goods, and a smaller carbon footprint for the entire planet.

In short: They found a way to make the "winding road" of fluid transport feel like a smooth, straight highway.

Technical Summary

Technical Summary: Collapse of Turbulence in Optimised Curved Pipe Flow

1. Problem Statement

Turbulence-induced friction is a primary driver of energy consumption in fluid transport and piping industries. While previous research has shown that curvature in pipes can induce partial relaminarisation (the transition from turbulent to laminar flow) at low Reynolds numbers ($Re_D \approx 3,000–6,000$), it has remained an open question whether turbulence can be suppressed at much higher Reynolds numbers ($Re_D \geq 10,000$) well beyond the linear stability limit.

The authors identify a critical tension in curved pipe flows: while streamwise curvature can have a stabilizing effect on turbulence, it also intensifies Dean vortices (secondary flows). These vortices redistribute turbulent kinetic energy (TKE) and generate cross-stream Reynolds stresses, which in turn feed back into the streamwise flow to regenerate turbulence. The central problem is determining if a geometric configuration exists that can leverage curvature to suppress turbulence without the secondary flow inadvertently sustaining it.

2. Methodology

The study employs a multi-fidelity approach combining automated optimization, high-fidelity simulations, and experimental validation:

• Shape Optimization: The authors used a free-form shape-optimization framework based on steady incompressible Reynolds-Averaged Navier–Stokes (RANS) equations with a $k–\omega$ SST closure model. The objective function minimized was the total entropy generation (a proxy for viscous dissipation and turbulence). The optimization allowed for simultaneous modification of the local centerline curvature ($\gamma$) and the cross-sectional geometry.
• Direct Numerical Simulations (DNS): To verify if the RANS-optimized design actually achieved relaminarisation, the authors performed DNS using the spectral-element solver Nek5000. Simulations were conducted at $Re_D = 10,000$ and $20,000$ to test the robustness of the effect at high Reynolds numbers.
• Experimental Validation: A physical test rig was constructed using vacuum-cast polyurethane resin. Pressure loss measurements were taken to validate the DNS results, comparing the optimized bend against both a baseline $180^\circ$ bend and a straight pipe of equal length.

3. Key Contributions & Physical Mechanisms

The paper identifies a dual-mechanism approach to achieve relaminarisation:

• Streamwise Curvature Effect: Increasing the local curvature ($\gamma$ up to $\approx 0.8$) introduces metric terms in the Reynolds-stress production equations that have a stabilizing effect on the streamwise Reynolds stress ($R_{ss}$), particularly near the inner wall.
• Cross-Sectional Modification (Ovalization): The optimization resulted in an oval cross-section elongated in the binormal direction ($\hat{y}$). This modification increases the lateral half-width ($a_y$), which reduces the binormal mean velocity gradient and weakens the centrifugal driving force of the Dean vortices.
• Disruption of the Regeneration Cycle: By weakening the Dean vortices, the production of cross-stream Reynolds stresses ($R_{yy}$ and $R_{rr}$) is significantly reduced. This prevents the feedback loop where secondary flows transport TKE and regenerate near-wall streaks, effectively collapsing the turbulence regeneration cycle.

4. Results

• Relaminarisation: DNS confirms that the optimized geometry (OPT) achieves near-full relaminarisation at both $Re_D = 10,000$ and $20,000$, a regime where turbulence is typically sustained in circular pipes.
• Pressure Loss Reduction:
  • The optimized bend reduces pressure loss by 53% compared to the baseline $180^\circ$ bend.
  • The optimized bend reduces pressure loss by 36% compared to a fully developed straight turbulent pipe of equal length.
• Secondary Flow Attenuation: The peak kinetic energy of the secondary flow in the optimized design is 50% lower than in the baseline design, despite the higher curvature.
• Robustness: The relaminarisation effect appears to be weakly dependent on the Reynolds number within the tested range, suggesting the mechanism is robust for high-speed industrial applications.

5. Significance

This work establishes a passive, mechanism-based route to turbulence control. It demonstrates that relaminarisation at high Reynolds numbers is possible if the geometry is engineered to decouple the stabilizing effects of curvature from the destabilizing effects of secondary flows. This provides a blueprint for designing more energy-efficient piping systems, heat exchangers, and biological conduits (like the aortic arch) by using shape optimization to minimize frictional energy losses.