In an attempt to achieve superior efficiency in coal fired power plants, Grade 91 steel has been highly utilized for piping and structural components. This steel is utilized for its good mechanical properties and creep resistance, even at high temperatures. Recent issues have developed however, in which premature failure due to creep can shorten the life of the steel. This is often attributed to improper post weld heat treatments in which the A1 temperature of the steel is exceeded, causing creep strength and other properties to suffer. This study aims to investigate and characterize Grade 91 weld metal's response to tempering as well as observe the effect of alloying elements on A1 temperature in an effort to mitigate the issue of premature failure.
Post weld heat treatments were applied to Grade 91 welds to investigate its response to tempering. Heat treatments of 730°C, 760°C, and 790°C were applied for 2 and 4 hour increments. Following heat treatment, hardness maps were created across the cross section of the weld. PWHT at 730°C revealed partially tempered martensite throughout the weld metal and CGHAZ, resulting in a structure that was too hard for service. At PWHTs of 760°C and 790°C soft regions existed throughout the weld metal and HAZ. Soft regions in the FGHAZ are thought to occur from overtempering. Soft regions in the weld metal are most likely owed to ferrite that has transformed due to the A1 temperature being exceeded.
Following heat treatment metallographic characterization techniques were used to observe the effects of PWHT on the welds. Using light optical and scanning electron microscopy, alpha and delta ferrite was observed throughout the microstructure. Charpy Impact testing was performed on heat treated samples as well. The heat treatments of 730°C provided the worst results with energy absorption of 15-30 ft-lbs. depending on area. Heat treatments of 760°C and 790°C gave impact test results of around 70 ft-lbs.
A design of experiment (DoE) was used to determine the affects of alloying elements on the A1 temperature of Grade 91 weld metal. A model based approach was initially utilized in which the compositional range of elements as set by the AWS A5.5 code was used for the DoE range and the A1 temperature was set as the DoE response. The A1was determined from thermodynamic simulations.
A model based equation for predicting the A1 temperature in Grade 91 weld metal based on composition was developed. Using the model based DoE the elements Mn, Ni, Cr, Mo, Cu, V, Nb, C, and Si were found to have a significant effect on the A1 temperature and were screened out to be used as factors in the development of a measurement based DoE.
The measurement based DoE included 36 compositions for testing. Test samples were produced in the form of small buttons which were evaluated using SS-DTA and dilatometry techniques to determine their A1 temperature. Upon completion of the measurement based DOE, a predictive equation will be developed for the A1 temperature in Grade 91 weld metal.