Effect of polishing disc compliance on the pressure distribution and material removal
Arunachalam, Adhithya Plato Sidharth
Date of Issue2019-04-22
School of Mechanical and Aerospace Engineering
Polishing is a key manufacturing process which has become inevitable in many industries such as shipbuilding, aerospace, automobile and bio-medical sectors. The compliant polishing tools (CPTs) are commonly used in industries, and there are different variants of the compliant polishing tools such as disc, wheel or belt types. Initially in this dissertation, in order to reduce the ambiguity involved in tool selection, two multi-criteria decision-making methods (MCDMs) methods are considered for a case study involving the selection of optimal polishing tools at different polishing stages. The CPTs are usually preferred to the rigid tools for the reasons like controlled material removal, better conformability to the polishing surface and better blending of finished and unfinished surfaces. Although these CPTs are being used commonly in industries for a long time, the material removal behaviour of these tools has not been investigated intensively. To study the underlying relationship between the compliance and contact pressure which influences the material removal profile, contact pressure distribution experiments are carried out for tools with different compliance using pressure films. The static pressure distribution showed variations in the contact area and pressure that could be correlated to material removal profile. Polishing experiments are carried out using a robotic polishing setup with force control, at different load values for both ‘Hard’ (low compliance) and ‘Soft’ (high compliance) rubber backing polishing tools. In order to predict the material removal profile, the finite element analysis simulation is carried out using ‘dynamic-implicit’ in ABAQUS® for the grainless disc, followed by executing the UMESHMOTION FORTRAN subroutine in ‘General-static’. In order to model the material removal profile in compliant tools with coated abrasive discs along with grains, the first step is to model the three-dimensional surface topography of the abrasive disc. Hence, an abrasive disc with prismatic shaped abrasive grains is considered. Initially, abrasive discs with grit sizes 60 and120 grit sizes are scanned to obtained the peak height and spatial distribution, which are found to follow a normal distribution. In this modelling technique, the protrusion height is determined not only by the position of the grain in the resin but also by the orientation of the prismatic grains. Using MATLAB® code, iterations are carried by randomly orienting the grains in the three mutually perpendicular axes, till the required statistical parameters are achieved. Later, the surface fitting is performed for the randomly distributed grains, and different 3D surface parameters such as amplitude, functional, spatial and hybrid parameters of the simulated surface are compared with the measured surface. Using the same methodology three-dimensional surface topology is simulated for the irregularly shaped grains too. To model the material removal profile considering the grains, two different approaches are proposed. In the first approach, using the grain distribution information, the abrasive disc with randomly oriented grains is modelled using python script in ABAQUS®. Using the Johnson-Cook damage model, the simulated material removal profile for three rotations are considered, and the effect of compliance on the material removal is discussed. In the second approach, a semi-analytical model is proposed to simulate the material removal profile. The randomly oriented grains are modelled as conically shaped grains with spherical tips. Using the average dynamic contact pressure distribution from the FEA results, the normal force acting on top of each grain is obtained, and the equivalent indentation depth is computed. Later, using relevant scaling, the complete material removal profile is obtained and compared with the experimental measurements for the hard backing pad. From the above-mentioned studies, it is evident that tool compliance plays a predominant role in material removal. Presently different tools are used for different compliance, and each time the tool is changed according to required material removal. A novel active compliant tool is designed to achieve the variable compliance within a single tool. Initially, different designs are proposed, and finally, the rubber tool design with retractable stiffeners is considered based on the stiffeners position in the disk (completely retracted to completely inserted), a range of compliance can be achieved in an analogue sense. Experimental trials are carried out to test the tool at three different positions of the stiffeners, and the material removal profile for different cases is compared. The material removal depth changed with respect to the stiffener positions where maximum material removal depth (with lower contact area) is noticed for the completely inserted stiffeners condition and vice-versa for completely retracted compliant state. Hence, this tool could be considered for high material removal operations like in case of weld seam and also during the blending operations where lesser material removal is needed.