With the findings of this study surface defect densification defects and shape devia-tions can be avoided. Furthermore, the time for development of the rolling process can be shortened. Through a process analysis, the relationship between process input and process output was determined. The test gear is representative for automotive applications as well as suitable for fur-ther fatigue tests. An analogy process with cylindrical tools and cylindrical roller blanks was developed. The cylindrical rollers reflect the material properties and the geometrical kinematic contact conditions on the pitch circle and generate results of the rolling forces and the compression. In addition, a simulation model was developed based on the finite element method. In the gear rolling simulation a material model, which describes both the deformation and the compaction of powder metallurgical material, was developed and implemented. The process input variables are the process parameters, the tool geometry and the rolling blank material. Process parameters that were investigated were the number of revolutions, the number of reverse points and the rolling speed. The mechanical characteristics of the powder metallurgical material were measured. The rolling forces increase with the infeed. The increase of the rolling force is com-posed of a linear and a degressive proportion. The linear increase is caused by the elastic deformation of the tool shafts, tools and rolling blank. The degressive in-crease results from work hardening and compaction. The maximum rolling force is reached at the maximum infeed of the rolling tools. The maximum rolling force is dependent on the infeed, the blank material, the process parameters and the contact geometry. Infeed has the greatest impact on the rolling force and is followed by the carbon content and density. The surface of rolled powder metallurgical gears has an average surface roughness of about RZ = 1.0 – 1.5 µm. In extreme cases, surface defects called overlaps occur in the tooth root area. The intensity of surface defects is influenced by the combination of the occuring stress in the rolling contact and the durability of the blank material. Due to the small contact radii in the tooth root area high stresses occur. The surface can be improved by increasing the carbon content. The density and the sintering temperature have only a small effect on the surface. In addition, the load and the number of rollings worsen the surface. Densification is a type of plastic deformation which is caused by hydrostatic pressure. The hydrostatic pressure of the rolling contact causes a collapse of the pores and the powder metallurgical material compacts. The hydrostatic stress component is caused primarily by the normal force. The density distribution is neither in tooth face direction nor in profile direction homgeneous. In tooth face direction, the compression drops due to the missing axial support from the faces. In profile direction, the densifictaion rises up to a maximum compression in the field of rolling pitch diameter. Beginning from that maximum the compression drops in the direction of the tooth tip and tooth root. By adapting the stock for densification a uniform compression in profile direction can be achieved. In general, the compression can be improved by adjusting the rolling blank properties. Higher densities of the blank increase the densification depth. The radius of the contact showed a minor influence on the compression depth. The carbon content, which has an influence on the hardness of the material, does not affect the compression depth. During rolling reproducible profile errors typically occur. These errors are character-ized by an overmeasure on the tip and an overmeasure in the tooth root. Elastic deformations of the roll blank and incorrectly oriented material flow during compression are the cause for the profile deviations. For compensation of the deviations a method was developed. The deviations could be improved significantly by the correction of the tool. The reached profile quality was Q 7, the profile line deviation Q 3, and pitch deviations and roundness were in the range of Q 1.
展开▼
