Abstract
Keywords: Valve lift, Ramp curve, Valve train, Chain, Multi-body Dynamics
In recent years, the number of engines that use a slide-assembled camshaft to enable the camshaft holder to be integrated with the engine head is increasing. The journal diameter must be made larger than the exterior of the cam, causing the journal clearance to increase. Consequently, the cam is liable to deform, and the phenomenon whereby the valve lift was lower than the design value occurred. Also, there were random variations of lift between each valve despite the fact that all cylinders used the same cam profile. Particularly, the reduction of the ramp curve resulted in slapping noise and valve jump, which adversely affects noise, durability, and performance. It was considered that the reduction of lift and random variations of lift are intimately related to the camshaft and the dynamic behavior of valve train, however the cause was unknown.
In order to elucidate the cause of these phenomena, a study based on simulation was carried out. In the conventional method, prediction using a single valve train was commonly carried out, but it was impossible to predict the random variations between each valve.
Accordingly, multi-body dynamics was employed to establish unique simulation technology that could take into account the behavior of not only the valve train, but also the camshaft, rocker arm shaft, and chain as well. As a result, it was possible to simultaneously predict the dynamic behavior of the valve train for all cylinders.
Also, in order to accurately predict the behavior in an actual engine, it is important to adjust the tappet clearance and the valve train rigidity. Accordingly, in this research the rocker arm unit rigidity test and the tappet clearance setting were reproduced in the simulation. Regarding the former, an actual rigidity measurement was constructed in the simulation, and an accurate value of valve train rigidity was determined. Regarding the latter, the tappet clearance was set with the camshaft fixed via the chain in the same phase as in an actual engine. As a result, it was possible to reproduce the random variations of the lift that occur in an actual engine. The validation was carried out with attention focused on the ramp curve of the valve lift of inline 4 cylinder engine. As a result of the validation, it was found that the simulation accurately reproduced the ramp curve in an actual engine, thus confirming that the accuracy of the simulation was adequate.
It was thus found by using the established simulation that the causes of the reduction and random variations of lift lay in the tappet clearance setting method and the deformation of the camshaft.
The clearance of the No.4 cylinder, which was set to a small clearance condition increased compared to the other cylinders when the engine was run. In addition, it became clear that whereas on the No.1 cylinder side the chain pulls down the camshaft, the motion on the No.4 cylinder side was not regulated, and the lift decreased further.