Abstract
This paper develops a method to generate a numerically simulated surface to replace an actual rough surface, and then the contact performance of the mating surfaces is analyzed. First, we use a 3D surface profilometer to obtain the morphology information of an actual rough surface. Second, a numerically simulated rough surface is generated by using the Gaussian simulation theory, which correspond to the same surface morphology features as the actual rough surface. Third, the reverse engineering technology is used to generate the rough surface model and the interface contact models for the morphology features of the actual rough surface and the numerically simulated surface, respectively. Finally, we compare the contact stiffness and the contact area of the numerically simulated surface and the actual rough surface. The mean errors of the contact pressure for the numerically simulated surface and the actual rough surface are 30.31% (grinding rough surface) and 25.12% (milling rough surface), and the mean errors of the contact area percentage for different contact states are 28.46%, 33.85%, and 35.51% (grinding rough surface) and 27.37%, 21.37%, and 23.42% (milling rough surface), respectively. These results indicate that there are differences between the surface morphology of the numerically simulated surface and the actual rough surface. Therefore, in terms of surface morphology, the numerically simulated surface cannot be used to replace the actual rough surface. This paper lays a theoretical foundation for the accurate substitution of an actual rough surface.
Highlights
INTRODUCTIONBoth mating surfaces are rough.[1]. From a microcosmic perspective, using any processing method, it is difficult to make the surface of a part absolutely smooth
In reality, both mating surfaces are rough.[1]
The contact performance parameters of a mating surface were obtained through macroscopic tests, and it was difficult to guarantee the reliability of the test data
Summary
Both mating surfaces are rough.[1]. From a microcosmic perspective, using any processing method, it is difficult to make the surface of a part absolutely smooth. Rey et al.[15] combined a fast Fourier transform, translation process theory, and a Johnson translator system to develop a simulation method for generating rough surfaces. Manesh et al.[16] proposed a numerical simulation method for rough surfaces based on a stochastic process and time series model, and obtained a Gaussian distribution surface using the AR digital filter and the Johnson conversion system. Hyun et al.[21] used fractal geometry to establish microscopic contact models of two rough surfaces: a rigid surface and an elastoplastic disk. The contact performance of the two contact models is analyzed by a finite element method, and the substitutability of the numerically simulated surface for the actual machined surface is verified by comparing the contact surface stiffness and the contact area.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
