For parts suppliers in the manufacturing industry, the process of preliminary production preparation and the subsequent calculation of offers are critical business activities. A vital part of production preparation is the design of fixtures and tooling necessary for many processes of metal forming. In order for a company to give quick responses to customer enquiries or changes in prior specifications, it would be highly beneficial with a degree of automation in this design process. This implies the development of a computer based system able to capture existing design procedures and associated knowledge for the classes of tooling required for the forming process.

In this work, an implementation for the rotary draw bending of aluminum tubing has been done to exemplify how to develop an automated design system. The system is based on heuristic knowledge developed over many years of practical experience, knowledge analytically derived from fundamental theory found in scientific literature, and rules based on empirical data from trial manufacturing. The system applies knowledge to a given specification that a skilled engineer otherwise would do manually. The system output can be used to evaluate whether a tube is producible.

The main idea behind the system is to use knowledge objects containing information on inputs, outputs, constraints and what software are used to implement the knowledge pieces. This approach makes the system highly flexible and allows for multiple types of knowledge that might overlap. When an offering calculation is wanted, the system is set to run applicable knowledge objects for presented input data. Other objects are run when an accurate calculation for detailing is wanted for a more detailed set of input data. The system is built on readily available commercial software packages connected with a simple Visual Basic .Net program.

When building a system of this kind, it is essential that the knowledge documentation and structure be such that the functions of the system can be easily understood by the users of the system and by future developers. Aspects of user friendliness, transparency and scalability are addressed in the summary of this paper.

For parts suppliers in the manufacturing industry the process of preliminary production preparation and subsequent calculation of offers are critical business activities. A vital part of production preparation is the design of fixtures and tooling necessary for many processes of metal forming. For a company to give quick responses to customer enquiries, or changes in prior specifications, it would be highly beneficial with a degree of automation in this design process. This implies the development of a computer based system able to capture existing design procedures and associated knowledge for the classes of tooling required for the forming process.

In this work we exemplify an automated design system for tooling by an implementation for rotary draw bending of aluminium tubing. The system is based on established design practice and heuristic knowledge developed over many years of practical experience. The system will evaluate whether a given specification is producible with existing materials and equipment, select suitable machine, determine process parameters and determine type and dimensions of components of form die, clamp die, follower or pressure die, wiper and mandrel. The system is built on readily available commercial software packages. When building a system of this kind it is essential that the knowledge documentation and structure is such that the functions of the system can be easily understood by the users of the system and by future developers. Aspects of user friendliness, transparency and scalability are addressed in the summary of this paper.