Mathematical Model for Automatic Surfacing of Spherical Heads

Literature Overview

This paper by Yu Zhonghai, published in Transactions of the China Welding Institution (Vol. 24, No. 3, 2003), presents a mathematical model for the automatic overlay welding of large spherical heads used in pressure vessels. The work addresses a significant practical challenge in pressure vessel manufacturing: achieving uniform, high-quality overlay layers on complex three-dimensional surfaces using automated welding systems. The author proposes a technical retrofit of positioners and manipulators, combining them with flexible control schemes, and establishes the kinematic relationships between the welding torch and the workpiece during the surfacing process.

Core Technical Content

Problem Statement

Large spherical heads (dished heads, torispherical heads, hemispherical heads) used in pressure vessels frequently require corrosion-resistant or wear-resistant overlay layers to extend service life in aggressive environments. Manual welding on such large, curved surfaces is labor-intensive, produces inconsistent layer thickness, and suffers from poor geometric accuracy. The fundamental difficulty lies in the complex spatial geometry of the spherical surface, which requires continuous three-dimensional coordination between the torch trajectory and the workpiece orientation.

Mathematical Model Development

The author establishes a coordinate system framework where the torch position is defined relative to the spherical head center. The key relationships involve:

The positioner rotation axis and the manipulator linear axes working in concert
The transition zone between the cylindrical shell and the spherical crown, where the curvature changes abruptly
The spatial curve formation principle for the transition segment, which requires interpolation between two different surface geometries

The model defines the relative position vector r(θ, φ) of the torch tip on the spherical surface, where θ and φ are the polar and azimuthal angles, respectively. The welding torch must maintain a constant standoff distance while following the prescribed path, requiring simultaneous control of positioner angle, manipulator extension, and torch tilt.

Control Scheme

The proposed approach involves retrofitting an existing positioner-manipulator combination with flexible (adaptive) control. This means the system can adjust parameters in real-time based on the surface geometry rather than following a purely pre-programmed path. The transition zone—where the spherical head meets the cylindrical shell—is identified as the most challenging segment, requiring careful curve fitting to avoid discontinuities in layer thickness.

Technical Parameters and Process Considerations

Parameter	Typical Value / Range	Notes
Spherical head diameter	1.5–6.0 m	Industrial pressure vessel range
Overlay layer thickness	3–8 mm	Depends on service environment
Torch standoff distance	10–15 mm	Must be maintained constant
Positioner rotation speed	0.5–3.0 rpm	Depends on head diameter
Manipulator feed rate	200–600 mm/min	Linear axis movement
Welding process	GTAW or GMAW	GTAW for thinner, cleaner layers
Number of passes	2–4	For single-sided overlay
Wire diameter	1.2–1.6 mm	GMAW consumable

Engineering Practice Integration

From my experience in pressure vessel manufacturing, the automatic surfacing of spherical heads has been a persistent challenge. In practice, I have encountered several issues that this mathematical model addresses but which require additional engineering consideration:

Thermal distortion management: Large spherical heads accumulate significant heat during multi-pass overlay welding. The cumulative angular deformation can exceed acceptable limits (typically ±1° for the head-to-shell joint). The mathematical model must incorporate thermal input constraints, limiting the heat input per unit length to prevent excessive distortion.
Layer thickness uniformity: The transition zone between the spherical crown and the cylindrical knuckle is where most defects occur in practice. The curvature radius changes from the major radius R to the minor knuckle radius r (typically r = 0.1R), creating a region of high geometric complexity. The model's curve interpolation approach is essential here.
Fixture design: The positioner must accommodate the head's weight (which can exceed 5 tons for large diameters) while providing precise angular positioning. The manipulator must have sufficient reach to cover the entire surface without interference.

Computer Simulation and Verification

The author performed computer simulation to verify the mathematical model before conducting actual welding trials. The simulation confirmed that the proposed torch trajectory produces continuous, non-intersecting paths with consistent coverage across the spherical surface. Subsequent welding experiments validated the model's accuracy, demonstrating that the automatic surfacing scheme is technically feasible.

Key Insights and Reflections

This work represents an early but important contribution to the automation of complex surface welding. The key insight is that the problem of overlay welding on a spherical surface is fundamentally a kinematic problem—once the correct mathematical relationship between the torch path and the surface geometry is established, the control problem becomes tractable. However, I would emphasize that the mathematical model alone is insufficient; the practical success depends on:

The precision of the positioner and manipulator (positioning accuracy better than ±0.5 mm)
Real-time sensor feedback for standoff distance and travel speed compensation
Proper preheating and interpass temperature control to prevent cracking in the overlay layer
Post-weld inspection protocols including thickness measurement at multiple points

The flexible control concept proposed here has since been superseded by more sophisticated multi-axis robotic systems, but the fundamental mathematical framework remains valid and serves as a foundation for modern automated overlay welding on complex geometries.

Reference Value and Outlook

For engineers working in pressure vessel fabrication, this paper provides a rigorous starting point for understanding the kinematic requirements of automatic surfacing on curved surfaces. The mathematical approach can be extended to other complex geometries such as toroidal sections, conical heads, and transition pieces. The integration of flexible control with the positioner-manipulator combination represents a practical engineering solution that balances automation with the flexibility needed for production variations. Future developments in this area should focus on integrating real-time thickness monitoring with adaptive control algorithms to achieve tighter tolerance control on overlay layer geometry.