Mass production of components with micro and nano scale surface features is known as micromoulding and is very
sensitive to a number of variables that can cause important changes in the surface geometry of the components. The
surface itself is regarded as a key element in determining the product's functionality and as such must be subject to
thorough quality control procedures. To that end, a number of surface measurement techniques have been employed
namely, White Light Interferometry (WLI) and Atomic Force Microscopy (AMF), whose resulting data is given in the
form of large and rather unmanageable Cartesian point clouds. This work uses Partial Differential Equations (PDEs) as
means for characterizing efficiently the surfaces associated with these data sets. This is carried out by solving the
Biharmonic equation subject to a set of boundary conditions describing outer surface contours extracted from the raw
measurement data. Design parameters are expressed as a function of the coefficients associated with the analytic solution
of the Biharmonic equation and are then compared against the design parameters describing an ideal surface profile.
Thus, the technique proposed here offers means for quality assessment using compressed data sets.
Micro injection moulding (micromoulding) technology has recently emerged as a viable manufacturing route for polymer, metal and ceramic components with micro-scale features and surface textures. With a cycle time for production of a single component of just a few seconds, the proces offers the capability for mass production of microscale devices at a low marginal cost. However, the extreme stresses, strain rates and temperature gradients characteristic of the process have the consequence that a slight fluctuation in material properties or moulding conditions can have a significant impact on the dimensional or structural properties of the resulting component and in-line process monitoring is highly desirable. This paper describes the development of an in-process, high speed 3-dimensional measurement system for evaluation of every component manufactured during the process. A high speed camera and microscope lens coupled with a linear stage are used to create a stack of images which are subsequently processed using extended depth of field techniques to form a virtual 3-dimensional contour of the component. This data can then be used to visually verify the quality of the moulding on-screen or standard machine vision algorithms can be employed to allow fully automated quality inspection and filtering of sub-standard products. Good results have been obtained for a range of materials and geometries and measurement accuracy has been verified through comparison with data obtained using a Wyko NT1100 white light interferometer.