Fatigue cracks have been detected in the underframes, fuel tanks, and fuel-tank gussets of some diesel locomotives that are currently in service. This presents the risk of a fuel tank separating from its locomotive and has an associated cost for making repeated repairs. It is therefore an interest to both operators of locomotives and to the designers of future locomotive models to understand the root cause of the cracks and to determine an effective remedy. In order to accomplish these goals, a finite element analysis (FEA) was used to identify events in the service environment that are capable of driving crack growth at a frequency that roughly matches the observed interval between crack repairs. These events were then used in further FEA simulations to predict the efficacy of various proposed remedies. It was found that the fuel tank, engine, and alternator of a locomotive behave as structural elements that resist deformations in the underframe. This explains why locomotive models with a relatively large fuel tank when compared to their engine and alternator are particularly susceptible to developing cracks: they experience higher stresses on their fuel tank and supporting brackets. It was also found that the most effective remedies were those that mitigate stress concentration effects from the gusset's rectangular design.

目前在役的一些内燃机车的底架、油箱和油箱角撑板都发现了疲劳裂纹。这带来了油箱与其机车分离的风险，并且具有进行重复维修的相关成本。因此，机车运营商和未来机车模型的设计者都有兴趣了解裂缝的根本原因并确定有效的补救措施。为了实现这些目标，使用有限元分析 (FEA) 来识别服务环境中能够以大致匹配观察到的裂纹修复间隔的频率驱动裂纹扩展的事件。然后将这些事件用于进一步的 FEA 模拟，以预测各种建议的补救措施的有效性。发现油箱、发动机、机车的交流发电机和交流发电机的作用就像是抵抗底架变形的结构元件。这就解释了为什么与发动机和交流发电机相比，油箱相对较大的机车模型特别容易出现裂缝：它们的油箱和支撑支架承受更高的应力。还发现最有效的补救措施是减轻角撑板矩形设计的应力集中效应。