Theoretical and Experimental Damage Prediction in Cold and Semi‐Hot Bulk Forming of Ductile Steels

Abstract

Bulk forging is among the most important manufacturing methods in metal forming, due to its wide applicability from some ounces to several tons of steel in a high diversity of shapes and forming conditions. Economical constraints demand for further optimisation and cost‐effective production. This requires the application of suitable finite elements simulation software, in order to support the already digitalised construction processes. Ductile damage is one of the most severe problems to arise during the production sequences, not only in cold but also in semi‐hot forging operations. Mathematical approaches exist for the modelling and simulation of ductile fracture in steel. In this paper some widespread used damage models are introduced and discussed. Their damage prediction quality has been verified by experiments, the tensile test and the collar specimen upsetting with several different steels under cold and semi‐hot forging conditions. The methods for the experimental fracture detection are introduced as well. In cold forging the passive ultrasonic testing with integrated statistical filtering algorithms is used. As this method is not applicable to semi‐hot forging experiments, optical fracture detection by means of a high‐speed camera is used instead. A very interesting material behaviour of the steels tested has been identified in the semi‐hot upsetting of collar specimen. For every steel a distinct temperature crossover interval exists, in which the forging process abruptly changes from damaged to undamaged state. This interval amounts to some degrees Celsius only for each of the seven materials investigated. Among the damage models proposed, the Model of Effective Stresses by Lemaitre is chosen for the application to a cold and a semi‐hot forging operation. These industrial processes of an axle end (cold) and a journal bearing (semi‐hot) are susceptible to damage for reasons to be discussed in this paper. It will be shown that the internal fracture of the axle end (chevrons) and the surface fissures of the journal bearing can be predicted with high accuracy. Moreover, the application of the damage model in the finite element software MSC.SuperForm 2004 offers a promising approach for process optimisation. Several possibilities could be tested for their suitability of reducing the calculated damage: geometry variation of the forming tools, process annealing, different materials. The use of damage models in finite element simulation can be regarded as a further step towards an optimal process design.