Inside the mechanics of a pipeline’s weakest turn

In gas and oil pipelines, curved bends called pipeline elbows are among the most failure-prone components. As sand and other solid particles travel through the line, they cannot follow the fluid smoothly around the bend and instead strike the outer wall, causing severe erosion. Existing studies attempting to model this process often fail to reflect real operating conditions. Ahmed et al. developed an experimental approach that accurately models particle motion within pipeline elbows.

The method integrates particle-tracking algorithms, high-speed imaging, and a rapidly eroding paint to directly correlate particle trajectories with surface wear. The results show that erosion is driven by particle inertia: High-inertia particles strike the outer wall and lose measurable velocity, indicating kinetic energy transfer through inelastic impacts. The study also finds that flow-directing components called guide vanes can delay erosion onset by redistributing particles more evenly and reducing peak wall impacts.

“This dual-diagnostic approach gives us mechanistic insight that simulations alone cannot validate,” author Yan Zhang said. “Guide vanes also require no active control or exotic materials, making it a practical and scalable mitigation solution.”

The study used a lab-scale setup with a transparent 4-inch glass elbow, supplied by an air fan and controlled sand injection. A high-speed camera captured particle motion, and their algorithm tracked about 10,000 trajectories per test to quantify velocities and spatial distributions. The elbow’s interior was coated with a uniform paint layer that wore away under particle impact, providing a low-cost proxy for metal erosion.

The team will expand their study to a broader parameter space, varying particle size, flow velocity, concentration, and elbow geometry to generalize the findings.

Source: “Experimental analysis of particle-laden flow and erosion mitigation in a 90-degree elbow using guide vanes,” by Imtiaj Nahin Ahmed, Dathi Rajasekhar, Trung B. Le, Hong Pan, Zhibin Lin, and Yan Zhang, International Journal of Fluid Engineering (2026). The article can be accessed at https://doi.org/10.1063/5.0291484 .