Design of an experimental platform for nondestructive detection of the bonding area of building insulation layers using penetrating imaging radar

The bonding quality of external wall insulation layers is critical for ensuring building envelope safety, thermal performance, and overall energy efficiency. Conventional assessment techniques, including destructive testing, are often costly, time-consuming, and difficult to quantify. Conventional nondestructive assessment techniques, including ultrasonic testing and infrared thermography, are also subject to constraints in accuracy, efficiency, and applicability under complex conditions. Penetrating imaging radar offers rapid detection, strong resistance to interference, and the ability to penetrate nonmetallic materials, making it a promising approach. The objective of this study is to design, develop, and validate a laboratory-scale experimental platform based on penetrating imaging radar that enables nondestructive detection and quantitative evaluation of the bonding area of insulation layers. The platform is designed to facilitate controlled studies of various bonding configurations and provide a reliable methodology for accurate assessment of insulation integrity under different scenarios.

A novel experimental platform, characterized by the integration of a stepped-frequency radar system with a three-dimensional (3D) motorized scanning mechanism, was constructed. To realistically simulate external wall insulation systems, a simplified wall insulation model was developed, consisting of a concrete substrate, a polymer bonding mortar layer, and a graphite polystyrene insulation board. The radar system comprises a vector network analyzer, a wideband ridged horn antenna, coaxial cables, and a control computer, enabling automated and repeatable measurements. The 3D scanning system utilizes stepper motors and a gantry-style three-axis configuration to ensure precise positioning of the radar above the insulation model, thereby facilitating high-density spatial sampling. Data processing combines the range migration algorithm (RMA) for accurate spatial focusing with robust principal component analysis (RPCA) for effective clutter suppression. Subsequently, edge detection and binarization are applied to delineate the contours of mortar bonding, ensuring accurate representation of bonding distribution and enabling reliable quantitative evaluation.

A series of experiments were conducted using mortars of varying shapes and surface areas to replicate typical bonding conditions in external wall insulation layers. Radar scanning followed a predefined serpentine trajectory at a fixed height above the model surface. The raw radar images were severely defocused due to antenna radiation pattern effects, making direct geometric assessment difficult. After processing with the RMA- and RPCA-based methods, the spatial distribution and contours of the bonding mortar were reconstructed with high fidelity. The resulting images closely corresponded to the physical models. Quantitative analysis demonstrated that the relative errors between measured bonding areas and actual values were <3% across all experimental scenarios. These findings confirm the high accuracy, reliability, and repeatability of the experimental platform and the associated data processing workflow, even under challenging experimental conditions.

The experimental platform facilitates visualization and quantitative evaluation of insulation layer bonding areas. The integration of range migration focusing and clutter suppression based on RPCA enhances imaging clarity, target distinguishability, and measurement accuracy. The platform integrates concepts from electronics, nondestructive testing, and civil engineering, demonstrating significant interdisciplinary innovation and practical applicability. It provides an effective method for laboratory-scale assessment of insulation layer bonding quality and has potential applications in building envelope quality control, structural maintenance, and long-term reliability evaluation. Furthermore, the platform may serve as a tool for interdisciplinary experimental education, supporting research and talent development in related engineering and technical fields.