Scanned Orbital Welding: Thermal Modeling and Lumped Adaptive Control

1999 ◽  
Vol 121 (4) ◽  
pp. 393-399
Author(s):  
H. Sfetsos ◽  
J. Angelis ◽  
C. Doumanidis

Scan orbital welding of cylindrical vessel, flange, and piping parts is performed by their rapid revolution under a radially or axially translated heat source, with its power modulated so as to implement a specified thermal distribution. Thus, the plasma-arc welding torch sweeps the stainless steel surface to generate a desirable temperature field and the concomitant material features. A numerical simulation of the thermal field is developed for off-line analysis. On this basis, a lumped thermal regulator of the heat-affected zone, employing infrared temperature feedback at a single spot, as well as standard PI, gain scheduling, and self-tuning control algorithms is tested. The thermal model is also employed for real-time torch efficiency identification and compensation. The numerical reference model serves as the basis for an in-process adaptive thermal control system to regulate the temperature field, using thermal feedback from the infrared pyrometer. A distributed-parameter control strategy, with guidance of the torch motion and power by a new weighted attraction strategy to randomly sampled points, is tested on scan-welded flanges. The regulator is validated computationally and experimentally, and its applicability to other scanned processing of materials is considered.

1996 ◽  
Vol 118 (4) ◽  
pp. 571-578 ◽  
Author(s):  
C. Doumanidis ◽  
N. Fourligkas

In thermal manufacturing processes performed by a localized, sequentially moving heat source, simultaneous regulation of multiple thermal quality characteristics requires real-time control of the temperature field developed through the distributed heat input on the part surface. Such control of the thermal field to a desired distribution employs infrared sensing and feedback of the surface temperature hill, to modulate the torch power and motion in-process. The torch trajectory is guided in real time by an efficient optimization algorithm based on the concept of moving complexes. This distributed-parameter control strategy is developed using a numerical simulation model of thermal processing, and its performance is evaluated experimentally in heat treatment of thin stainless steel plates. The thermal controller is applied to the new scan welding process, in which it drives the torch in a reciprocating motion along the weld, yielding a uniform and smooth temperature field, and thus a favorable material structure and mechanical properties. Application of such thermal control to various other material processing methods is also investigated.


Sign in / Sign up

Export Citation Format

Share Document