Search

The authors wish to explain the similarity between some figures in the above paper (hereafter called the JPD paper) and in their other publication, Lee S, Jang Y H and Kim W 2008 Effects of nanosized contact spots on thermal contact resistance J. Appl. Phys.103 074308 (hereafter called the JAP paper), and to explain the differences between the two papers, which are not explicitly stated in the JPD paper. The main objective of the JAP paper is to calculate the thermal contact resistance of the nanosized contact spots in multiscale contact. During the process of multiscale analysis, the thermal conductivity varies, especially below the phonon mean free path. The JPD paper deals with the electrical contact resistance in the multiscale contact distribution with an assumption of constant electrical resistivity, which is known as a different kind of physics in a larger characteristic length scale. There are similar figures in the JPD paper and the JAP paper: figures 6, 7 and 8 in the JPD paper and figures 3, 4 and 5 in the JAP paper. Two research works were performed on the basis of a specific microcontact distribution. In the JAP paper, the scale of the contact distribution is in the range of the phonon mean free path of Si, which is a very small size of contact distribution. In the JPD paper, the scale of contact distribution is in the continuum scale, which is larger than the phonon mean free path. In addition, due to the characteristics of a fractal surface which repeatedly generates a similar shape of contact distribution in the different length scales, the shape of contact distribution looks similar, but the total sizes of domain in the JPD and JAP figures are different. The projected areas L x L of fractal surface of the JAP paper and JPD paper are 10 mm x 10 mm and 10 mm x 10 mm, respectively. The length scale is already stated in the JAP paper, but not in the JPD paper. Thus, we have to state that the figures were adapted from the JAP paper without clear attribution and that credit should be given to the original source of the images. Most significantly, the two most important features of the two works are based on totally different physics and the calculation methods for contact resistance. The text of section 4.3 of the JPD paper should be modified. The left column of page 5 and the first two lines of the right column should be replaced by the following text. The projected area of a self-affine fractal surface is plotted in figure 6. In order to find the microcontact spot distribution from the fractal surface, numerical simulations are applied to the model of multiscale rough surface. Several contact models are constructed according to the resolution of 1/24 to 1/28. Figure 7 shows a typical finite element model in which the rough elastic surface is in contact with a perfectly flat rigid surface. Due to the fine meshes in the contact interface, strong mesh gradations are required. The contact simulation enforcing the impenetrability constraint on the mating surfaces is performed by a commercial finite element package, Abaqus/HKS ([22] of the JPD paper). The resulting microcontact spot distributions for different resolution of the rough surface are obtained in figure 8. As reported by Borri-Brunetto et al ([23] of the JPD paper), the actual contact area decreases and the number of contact spots increases as the resolution of the discretization escalates. A more detailed description for the model can be obtained in the JAP paper. [ABSTRACT FROM AUTHOR]

For contact between rough surfaces of conductors in which a clustered contact spot distribution is dominant through a multiscale process, electrical contact resistance (ECR) is analysed using a smoothed version of Greenwood's model (Jang and Barber 2003 J. Appl. Phys. 94 7215), which is extended to estimate the statistical distribution of contact spots considering the size and the location simultaneously. The application of this statistical method to a contact spot distribution, generated by the finite element method using a fractal surface defined by the random midpoint displacement algorithm, identifies the effect of the clustered contact distribution on ECR, showing that including a finer scale in the fractal contact surface causes the predicted resistance to approach a finite limit. It is also confirmed that the results are close to that of Barber's analogy (Barber 2003 Proc. R. Soc. Lond. A 459 53) regarding incremental stiffness and conductance for elastic contact. [ABSTRACT FROM AUTHOR]