Modeling and Evaluation of Boiler Pipe Tees Using Finite Element Analysis

Literature Overview

This paper by Zhang Qingyi from South China University of Technology, published in Chemical Engineering & Machinery (2011, Vol. 38, No. 1), presents a finite element analysis (FEA) approach to evaluating the stress distribution and remaining service life of tee fittings in boiler piping systems. The author employed ANSYS software to construct a detailed three-dimensional model of a boiler tee, incorporating the weld geometry, heat-affected zone (HAZ), and applied loading conditions. The model predictions were validated against actual field failure data from damaged tee welds, demonstrating that the FEA approach can accurately identify critical stress regions and predict remaining life. This work represents an early application of computational mechanics to piping component integrity assessment in the Chinese chemical engineering industry.

FEA Modeling Approach and Stress Analysis

The methodology employed in this study follows a systematic FEA workflow applicable to piping component assessment:

Modeling Parameter	Description	Relevance to Assessment
Geometry	3D solid model of tee including branch and run	Captures geometric stress concentrations at tee root
Material properties	Elastic, plastic, and creep properties of base and weld metals	Enables nonlinear stress-strain analysis
Weld modeling	Explicit HAZ and weld metal zones with distinct properties	Reflects microstructural heterogeneity
Boundary conditions	Internal pressure, thermal gradients, and external loads	Simulates realistic operating conditions
Mesh density	Refined mesh at weld root and HAZ	Resolves stress gradients accurately
Post-processing	Stress linearization through thickness per ASME methodology	Enables comparison with code allowable stresses

The key finding was that the high-stress region predicted by the FEA model along the weld HAZ range corresponded closely with the locations where actual tee failures were observed in service. This validation establishes confidence in using FEA as a predictive tool for identifying vulnerable regions in piping components. The author further employed ANSYS general post-processing to map stress through the wall thickness, perform stress linearization per ASME B31.3 Appendix G methodology, evaluate the component's working condition, and estimate the remaining number of load cycles before failure.

Stress Linearization and Code Compliance Evaluation

The stress linearization procedure is central to the engineering assessment methodology presented. In accordance with ASME B31.3 Section 344 and Appendix G, the total stress at any point in the component is decomposed into:

Membrane stress (Pm) — The average stress through the wall thickness, representing the global equilibrium component.
Bending stress (Pb) — The linearly varying component through the wall thickness, representing local bending effects.
Peak stress (PL) — The remaining non-linear component, typically associated with stress concentrations, residual stresses, and local geometric discontinuities.

The code allowable limits for these stress categories are:

Stress Category	Allowable Limit	Basis
Pm	S (allowable stress)	Plastic collapse prevention
Pm + Pb	1.5S	Elastic limit with safety factor
Pm + Pb + PL	3S (or 4S for certain cases)	Fatigue and stress concentration

By performing stress linearization on the FEA results, the author was able to determine whether the actual stress state exceeded these code limits and to estimate the fatigue life using S-N curves appropriate for the material and loading spectrum. The comparison between FEA predictions and actual failure locations provides strong evidence that stress concentration at the tee root weld, particularly in the HAZ where material properties may be degraded, is the primary driver of in-service failure.

Integration with Engineering Practice

This study demonstrates the practical value of FEA in piping integrity management programs. The following engineering applications can be identified:

Pre-fabrication stress analysis — FEA can be performed during the design phase to optimize tee geometry, welding procedure, and PWHT parameters to minimize peak stresses. For example, modifying the branch-to-run diameter ratio or fillet weld geometry can significantly reduce stress concentration factors.
In-service assessment — Existing tees with suspected damage can be evaluated using FEA with measured wall thickness data from ultrasonic testing, providing a quantitative basis for repair decisions.
Remaining life prediction — By incorporating the actual operating temperature and pressure history into the FEA model, engineers can estimate remaining fatigue and creep life, supporting risk-based inspection planning per API 580/581.
Repair design optimization — When field repairs are required per ASME PCC-2, FEA can evaluate the effectiveness of proposed repair configurations (saddles, reinforcement plates, sleeve welds) before implementation.

Study Insights and Reflections

This paper, while published in 2011, remains highly relevant to contemporary piping engineering practice. The integration of FEA with code-based stress assessment provides a powerful methodology for evaluating the integrity of critical piping components. However, several limitations must be acknowledged: FEA results are only as accurate as the input data (material properties, loading conditions, boundary conditions), and the validation against actual failures, while encouraging, represents a limited dataset. Modern practice should incorporate uncertainty quantification, probabilistic methods, and experimental validation through strain gauge measurements or digital image correlation on physical test specimens. Nevertheless, the fundamental approach — using computational mechanics to identify critical stress regions and evaluate remaining life — is sound and has been further developed in subsequent standards such as API 579/ASME FFS-1 and BS 7910. This study serves as an important reference for engineers transitioning from purely empirical to computational integrity assessment methodologies.