This paper proposed a process knowledge model for facilitating complicated industrial components' manufacturing by suggesting a combination of the manufacturing action elements under a given situation. The proposed knowledge model is built regarding the generalized manufacturing process as a sequence of the situations under which the experienced workers made their decisions and the actions which the workers took. Each situation and action can be broken down into a set of element situations and element actions. Based on the interviews with expert workers and the recorded manufacturing data, the pairs of situation element and action element are extracted, scored and stored in the knowledge model. The most efficient action suggested by the knowledge model is a combination of the action elements which are fuzzily inferred to be the most effective ones for the given situation. In this model, the effectiveness of each situation and action element pair is evaluated and scored based on the subsequent manufacturing steps and manufacturing milestones from the recorded data. The proposed knowledge model was proved to be effective when being applied to a series of 3D manufacturing data obtained in the shipyard.