The crucial issue in Knowledge-Based Engineering is representation of expert knowledge in a readable way for both a computer agent and human. Properly prepared knowledge (captured, assessed, formalized and structured) plays an important role for success of every knowledge based system; particularly in the discipline of machinery design. Many knowledge model properties are of great significance, in addition to formal structure and consistency, reusability, flexibility, interoperability of knowledge model are also becoming increasingly essential for the knowledge-based systems research community. This paper presents a hybrid knowledge representation for solving deductive engineering tasks in Knowledge Based Machinery Design. We propose two different methods for storing engineering knowledge; as a repository for declarative knowledge, an ontological knowledge based was applied; for procedural knowledge storage as a formal representation, a geometrical model was used. Apart from theoretical foundations the work is also concerned with presenting and discussing an industrial application of the above-mentioned knowledge representation. As a practical part an application for engineer supporting in design process of dies with non-cutting shape forming in automotive field is indicated. This part deals with problems, challenges and bottlenecks coming up during developing process that application and shows advantages and disadvantages of this knowledge representation. For carrying out the effort CATIA V5 as Computer Aided Design system was applied. To fulfill the knowledge model requirements, especially reusability, flexibility and inter-operability, Protégé as an ontology editor and knowledge-based framework with OKBC (Open Knowledge Base Connectivity) protocol was used.