Speaker
Description
The surface condition of stainless steel strongly influences corrosion resistance, aesthetics, and the need for post-weld cleaning. In particular, oxide formation during thermal processing modifies surface emissivity and radiative heat exchange, which can bias temperature readings and, at the same time, provide an opportunity for non-contact diagnostics. This work investigates the feasibility of determining the oxidation state of stainless-steel surfaces (oxidized vs. non-oxidized and intermediate conditions) during and after welding-like thermal loading using infrared thermography combined with digital image analysis.
An experimental setup is developed to reproduce a controlled welding scenario with a moving heat source and synchronized thermographic recording. During thermal excitation, the interaction between the heated zone and the surrounding atmosphere is modified by forced gas flow delivered near the heat-affected region. Three classes of process atmospheres are considered: (i) inert shielding gas, (ii) active gas mixtures, and (iii) oxidizing atmospheres, enabling controlled promotion or suppression of oxide formation both during the pass and in the cooling stage. Additionally, stainless-steel specimens with deliberately varied initial surface conditions, including mechanically cleaned surfaces, pre-oxidized surfaces, and surfaces with different roughness levels are used for post-process assessment development.
Thermographic sequences are analyzed in two complementary ways. First, transient thermal descriptors are extracted (e.g., heating/cooling rates, peak temperature evolution along the travel direction, spatial gradients near the heat-affected zone, and time constants derived from local cooling curves). Second, digital image-processing features are computed directly from the infrared imagery (texture and spatio-temporal pattern metrics, tracking of isotherm shapes, and statistical descriptors of emissivity-related contrast changes). These features are fused to classify the surface state and to localize regions with different oxidation levels in a map-based form. Where possible, the diagnostic procedure is designed to be robust to uncertainties in absolute emissivity by emphasizing relative, dynamic, and spatial signatures that arise from oxide growth and surface transformations.
The expected outcome is a thermography-driven methodology that enables in-process and post-process assessment of surface oxidation state without contact and without interrupting the welding operation. The study also aims to quantify how shielding conditions and initial surface preparation influence thermographic observables, and to outline practical guidelines for integrating such monitoring into welding and thermal-processing workflows. The proposed approach can support real-time quality control, reduce unnecessary post-weld treatment, and improve repeatability in applications where surface integrity of stainless steel is critical.
