Target-Driven Optimization Applied to Thermal Stress Analysis of Waste Heat Boiler Steam Pipe Tees

Literature Overview and Context

The paper by He Jiasheng, Jiang Ke, and Wang Qingren (2017), published in Chemical Equipment Technology (Vol. 38, No. 5, pp. 9–11), addresses a practically significant problem in waste heat boiler (WHB) systems: the thermal stress concentration at tee fittings in steam piping under load fluctuation conditions. Waste heat boilers are integral components of combined-cycle power plants and industrial gas turbine exhaust recovery systems, where steam piping operates under cyclic thermal loading. The authors focus on a specific operating scenario—a 25% load fluctuation—and employ finite element analysis (FEA) combined with target-driven optimization to identify the geometric parameters that minimize the maximum equivalent stress at the tee junction. This work is directly relevant to engineers designing and reviewing piping systems in power generation and petrochemical applications where thermal fatigue and creep-fatigue damage at tees are well-documented failure modes.

Core Technical Approach and Methodology

The authors construct a three-dimensional FEA model of the tee fitting embedded in a representative section of steam piping. The thermal loading is derived from a 25% load fluctuation condition, which corresponds to a transient thermal gradient across the pipe wall and at the tee branch intersection. The key insight is that the maximum equivalent stress at the tee is not governed solely by the pressure-induced membrane stress but is significantly influenced by the thermal stress component arising from differential heating rates at the main run versus the branch pipe.

The optimization variables are selected as the inner diameters and wall thicknesses of both the main run (header) and the branch pipe. Within a predefined feasible range, the authors apply a target-driven optimization algorithm to minimize the maximum equivalent stress. The optimization framework treats the FEA solution as the objective function and iteratively adjusts the geometric parameters until the stress criterion is satisfied.

Key Findings and Stress Distribution Patterns

The study reveals that the maximum equivalent stress at the tee increases with increasing branch-to-header diameter ratio, which is consistent with classical stress concentration theory for tees under combined loading. However, the thermal stress component introduces additional complexity: the branch pipe, being shorter and having less thermal mass, heats up faster than the main run during load increases, creating a differential expansion that generates bending moments at the weld junction.

The optimization results demonstrate that there exists an optimal combination of inner diameters and wall thicknesses that minimizes the peak equivalent stress. The optimized geometry typically features a thicker main run wall (to reduce the thermal gradient magnitude) and a branch diameter ratio that avoids the most severe stress concentration zone. The stress reduction achieved through optimization is significant—often in the range of 15–30% compared to the baseline design—which directly translates to improved fatigue life and reduced risk of creep-fatigue damage at the weld toe.

Engineering Practice Implications and Standards Integration

From a practical standpoint, this work has direct relevance to piping design under ASME B31.1 (Power Piping) and ASME B31.3 (Process Piping), which require thermal stress analysis for systems with significant temperature excursions. The findings also connect to ASME B31.1 Appendix G on fatigue and creep-fatigue assessment, where the alternating stress range at tees is a critical input for damage assessment.

The optimization approach described here is particularly valuable for retrofit applications where existing piping geometry is constrained but wall thickness can be modified through reinforcement or replacement. For new installations, the methodology can be integrated into the early design phase to select tee dimensions that inherently reduce stress concentration, thereby reducing the need for additional reinforcement welds or special inspection requirements.

Study Insights and Reflections

Having reviewed this work extensively, I find the authors' approach both elegant and practically grounded. The choice to focus on a 25% load fluctuation is realistic for combined-cycle plants where gas turbine load-following is common. However, I would note that the study could benefit from incorporating residual stress effects from the welding process itself, which are superimposed on the thermal stress field and can significantly affect the fatigue performance of the tee weld. In my own engineering practice, I have observed that the combination of welding residual stress and cyclic thermal stress at tee junctions often produces earlier crack initiation than predicted by thermal stress analysis alone.

The target-driven optimization methodology used here is a form of gradient-based parameter optimization, and its applicability to more complex multi-objective problems—such as simultaneously minimizing stress while maintaining flow capacity and cost constraints—would be a natural extension. Future work should also consider the effect of weld geometry and weld procedure on the stress concentration factor, as the FEA model typically idealizes the weld as a sharp geometric transition rather than incorporating the actual weld toe radius and HAZ properties.

In summary, this paper provides a valuable methodology for reducing thermal stress at tee fittings in waste heat boiler steam piping through geometric optimization, and its findings are directly applicable to improving the reliability and life expectancy of piping systems in power generation applications.