Abstract: Nylon bar is widely used in aerospace, automobile, machinery and other industrial fields due to its excellent comprehensive mechanical properties. In response to the urgent demand of internal defect detection for nylon bar, an ultrasonic water-immersion focus detection scheme is proposed, and a three degree of freedom scanning device with a rotary motion device is designed. Moreover, the system software is developed by multi-thread technology to realize automatic scanning and real-time imaging of nylon bar. The numerical model of ultrasonic propagation in water-nylon bar is established based on the finite element method. And, the effects of key parameters such as frequency and underwater sound distance on sound field distribution are analyzed. The built-up system has been applied in nylon bar detection with good stability. The ultrasonic B and C scan images obtained in detection can be used for localization and quantitatively analysis of internal defects of nylon bar. The accuracy of the system is verified by comparing with industrial CT detection results. The system realizes the automation and visualization of nylon bar detection, which provides an effective technical means for guaranteeing nylon bar quality.