研究目的
To develop and optimize a fiber optic hydrogen sensor using palladium foils attached to fiber Bragg gratings for transformer monitoring, and to investigate the differences in response time between gas and oil environments.
研究成果
The developed sensors show high sensitivity and low variability in hydrogen detection, but response times are much longer in oil than in gas. The primary factor for the slow response in oil is the slowed absorption of hydrogen on the palladium surface, not diffusion through palladium or oil. This indicates a need for further optimization for practical applications in transformer monitoring.
研究不足
The response time in oil is significantly slower than in gas, potentially due to surface impurities or environmental factors like oil slowing hydrogen absorption. The theoretical models may not fully account for real-world conditions such as surface cleanliness or exact foil properties.
1：Experimental Design and Method Selection:
The study involved designing and optimizing a fiber optic sensor with palladium foils and FBGs. Two custom experimental setups were used to test sensor response in hydrogen gas mixtures and dissolved hydrogen in oil. Theoretical models for hydrogen diffusion and absorption in palladium were applied.
2：Sample Selection and Data Sources:
Fifteen sensors were manufactured in parallel using a vacuum bagging process. Data were collected from these sensors exposed to 5% H2 in N2 gas and the same gas bubbled through transformer oil.
3：List of Experimental Equipment and Materials:
Equipment includes self-built test setups for gas and oil measurements, a commercial DGA system (Myrkos-Version: Field package, Morgan Schaffer), palladium foils (100 μm thick), fiber Bragg gratings, and transformer oil (BS 148 class-I, Savita Oil Tech. Ltd).
4：Experimental Procedures and Operational Workflow:
Sensors were tested at 90°C. Hydrogen concentration was pre-mixed and introduced into measurement chambers. Response curves were recorded, and hydrogen uptake in oil was measured with the DGA system.
5：Data Analysis Methods:
Data analysis involved calculating response times, standard deviations, and comparing with theoretical models (e.g., diffusion equation, Arrhenius law, Sieverts's law). Numerical solutions and exponential decay approximations were used.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
