The Federal Railroad Administration's Office of Research and Development is conducting research into passenger locomotive fuel tank crashworthiness. A series of impact tests is being conducted to measure fuel tank deformation under two types of dynamic loading conditions -blunt and raking impacts. This program is intended to result in a better understanding of design features that improve the puncture resistance of passenger locomotive fuel tanks. One reason for performing this program is to aid in development of appropriate standards for puncture resistance to be applied to alternativelydesigned fuel tanks, such as on diesel multiple unit (DMU) passenger rail equipment. This paper describes the results of the third blunt impact test of retired F-40 locomotive fuel tanks.The test setup was designed for the Transportation Technology Center (TTC) in Pueblo, Colorado, to impart blunt impacts to the bottom of each fuel tank specimen. The specimens tested to date are from FRA-owned retired F-40 passenger locomotives. To conduct the test, each tank was emptied of fluid and mounted on a crash wall with the bottom surface exposed. A rail cart modified with a "rigid" indenter measuring 12 inches by 12 inches, was released to impact the bottom of fuel tank at a target impact speed. The first two tests, conducted on October 8 and 9, 2013, were designed to impact the center of two different tank designs. Tests were conducted at impact speeds of 4.5 and 6.2 mph and caused maximum residual dents of 5 inches and 1.5 inches, respectively. On August 20, 2014 the test of fuel tank 234 was conducted to impact the tank off-center between two baffles. Forcedeformation measurements were collected for each tank during the three tests. The series of tests provide information regarding the influence of tank design on puncture resistance.In the test of tank 234, the target impact speed was 12.5 mph, and the actual impact occurred at 11.2 mph. The test resulted in a residual dent depth of approximately 9 inches, and buckling of several internal baffles. The impact did not result in puncture of the tank. Following the test, the tank was cut open to permit examination of the baffles. This examination revealed a different baffle geometry than was modeled based on pre-test measurements.Finite element analysis (FEA) was used to predict the behavior of the tank during the test. The FE model of the tank required several material properties to be defined in order to capture puncture behavior. The combination of metal plasticity, ductile failure, and element removal would permit the model to simulate puncture for this tank. Following the test, the tank was cut open, revealing a different baffle arrangement than had been initially thought. The post-test FE model was then updated to include the actual baffle arrangement of tank 234. With the actual baffle arrangement included in the model, the FE results are in fairly good agreement with the test. Additional changes to the ductile failure criterion were also made in the post-test model.The o...