The machining of biocomposites remains among the important challenges that face the economic world for the development of these eco-friendly materials in order to achieve their final transformation and meet industrial standards. The main issues of biocomposite machining result from the multiscale cellulosic structure of the natural plant fibrous reinforcement that induces high fiber deformation, high sensitivity to process parameters, and high dependence on environmental conditions (humidity and temperature) during cutting operations. Moreover, plant fibrous reinforcements show in general a random structure that is intimately related to the random distribution of the technical fiber size. Indeed, when observing a biocomposite surface made of unidirectional flax fibers and polylactic-acid (PLA) matrix, it can be noticed from Figure 1 that the fibrous reinforcement can be in the form of separated elementary fibers (from 10 µm to 20 µm in diameter) or technical fibers that are a bundle of several elementary fibers. These technical fibers have a variable diameter (depending on the number of elementary fibers) that ranges from 50 µm to 150 µm. In some cases, the plant reinforcement is in the form of fiber yarns that can reach around 1mm in diameter. Therefore, the plant fibrous reinforcement is present at different scale levels in the biocomposite structure with different mechanical properties at each of these scales. Surface characterization of biocomposites for machining analysis should then consider this multiscale configuration in order to discriminate the effect of each material or process parameter. In this paper, the machined surfaces of flax/PLA biocomposites performed with the milling process have been analyzed using a white light interferometer at different scales as shown in Figure 2. Energetic analysis and SEM observations of the machined surfaces have been performed to compare with the results of the multiscale topographic measurements. At each measurement scale, the 3D mean arithmetic roughness parameter (Sa) has been calculated to assess the effect of process parameters. Results demonstrate the significant impact of the analysis scale on the topographic behavior of the machined surfaces in the function of the process parameters. The variation of the analysis scale allows the involvement or the non-involvement of several machining-induced mechanisms such as single fiber deformation, fiber bundle deformation, matrix plowing, interface damages, and milling-induced waviness. Consequently, the discrimination of the process parameter effect (that is responsible for each machining-induced mechanism) is not similar when varying the topographic measurement scale. By comparing with the microscopic SEM observations and the energetic analysis, the results of this work lead to identifying the relevant analysis scale for an efficient machinability qualification of unidirectional flax fibers reinforced PLA biocomposites.

Water spray due to a tire rolling on a wet road surface has been studied to mitigate the reduced visibility [1, 2] or, in recent years, estimate the road surface wetness [3, 4, 5]. Previous research, conducted mainly with trucks [1, 2], has enabled to understand the effect of the vehicle speed and the tire tread sculpture on the amount of water spray. Despite the fact that the road surface is inseparable from the tire in a contact study, less is known about the link between the road surface texture and water spray. This paper aims at responding to this gap. The study involves a passenger car and water spray is characterized by the vibration of wheel arch liners due to impacts of water droplets. The instrumentation of the car, consisting of equipping the front wheel arch with accelerometers, is presented. Tests are performed on six road surfaces, representing various textures, and at different vehicle speeds and water depths simulating the road wetness during and after a precipitation. 3D topographical maps (squares with 100 mm sides) of the test surfaces are measured by means of a portable device. Signals recorded by the accelerometers are filtered and processed. The surface topography is analyzed using the MountainsMap software from which geometrical and functional parameters are calculated. Results, showing that the acceleration amplitude increases with increasing vehicle speeds and increasing water depths on the road surface (Fig. 1), are consistent with previous studies. They also highlight the fact that current parameters used in the road field to characterize the road surface texture (the so-called Mean Texture Depth and Mean Profile Depth) are not relevant enough to explain the observed difference between the test surfaces (Fig. 2). Analysis of parameters characterizing the orientation of the texture and the void distribution better reflect the role of road surface texture with respect to water evacuation (by drainage or squeezing) from the tire-road contact area. With this new finding, it is possible to establish a relationship between water spray and influencing factors (water depth, vehicle speed, road surface texture). Discussions are made in terms of application of this relationship to an on-board estimation of the tire-road friction to improve the vehicle stability and the driver safety.

Plastic components are produced in large volumes in fast and highly automated production lines. However, the injection moulds are individually manufactured and manually polished to reach the required surface finish to produce highly glossy plastic components with surface roughness levels in the nm-range. Today’s ever-increasing demands for high quality products, shorter lead times and reduced costs, push manufacturers to find alternative finishing solutions. The inspection and validation of the mould- and component surfaces are based on master plaques developed for each specific purpose. But there is a need to make these evaluations more objective, i.e. to quantify the perceived quality of the plastic components to specify standardised and measurable parameters for the mould surface, since the moulds are manufactured by different suppliers using various types of manufacturing procedures leading to varying surface structures. This study was based on nine different part geometries, nine colours and five types of polymers in different combinations, in total 180 samples, and nine mould surfaces. The components were visually inspected and graded at the company based on an internal evaluation procedure based on the total appearance of the component. All component- and mould surfaces were measured with a coherence scanning interferometer using a 10x Mirau interferometer objective giving a measurement area of 0.86mm x 0.86 mm. A measurement strategy was developed for quality control of components, recommending measurements ’every other mm’ to catch local surface variations as milling marks or holes. Perceived quality was related to surface texture parameters, separated for longer and shorter wavelengths. However, the mould surface was replicated to a varying degree by the different plastic materials, and in some cases other features (as small holes or scratches) rather than the mould surface texture constituted the dominant surface structure of the components, see figure 1. Defects were divided into four classes, circular (in- & outwardly, i.e. holes and peaks), and elongated (in- & outwardly, i.e. scratches and ridges). Yet, visual evaluations are not consistent but are ‘adjusted’ to the context (as part type and placement according to other components). The long-term goal is to develop a surface metrology strategy for production of injection moulded plastic parts, including the surface quality of moulds.

One important key factor for smart and sustainable production is more intensive use of smart and customized metrology systems. The wood industry has not yet benefited from this technology, but a combination of “intelligent monitoring and multi-objective optimization approaches should pave the way for controlling the sawing process so higher surface quality and cost-efficient machining is achieved” [1]. Advanced techniques for, for example, saw pattern optimization, timber grading and traceability are commercially available, but there is a lack of methods for measuring and characterizing sawn wood surfaces to provide feedback on tool wear and tool stability, as well as to provide a framework for the industry to define functional surface quality of wood products. Over-compensation of cutting parameters is one example where a better understanding of resulting wood topography combined with smart online metrology systems could contribute to a more sustainable production - a production increase corresponding to 75,000 m3 sawn timber per mm reduced saw blade thickness is realistic, which would be further improved if dimensional stability could be increased [2]. This study includes 36 boards (spruce trees) taken from the production line in a sawmill (circular sawing); 12 boards just after a saw blade shift, 12 boards mid-time, and 12 boards before a blade shift (i.e. sawn with a worn blade). All boards were scanned at site with board scanners (accuracy on the mm level), as well as with a laser scanning system (micro-epsilon scanCONTROL 2900-100 with a quoted resolution of 1280 pixels/profile) giving an accuracy on the um level. The areal surface parameters defined in ISO 25178-2 [3] are used to characterize the measured surface topography, especially the surface parameters root mean square height (Sq), reduced peak height (Spk) and reduced pit depth (Svk). These parameters have been used by other researchers to detect ‘fuzziness’ (Rpk), to indicate tool wear (Rk), and to define anatomical features (Rvk) [4-6]. The aim of this paper was to find surface features that could be used as indicators for tool changes instead of today's’ scheduled ones. Besides, other surface deviations such as feed roller marks were of interest. The final goal is to gain a better understanding of sawn wood topography at the micro-mm range in order to utilize the potential of surface metrology for higher process control and to facilitate new manufacturing techniques for sustainable engineered wood products.

Cellular metallic materials are a new, relatively young group of materials since around 1940. They are basically characterized by a lower mass density compared to the solid material. In other words, open structures are present in a given envelope volume of a body made of this material. This property has considerable application potential, making specific applications in medicine, filter technology and the chemical industry, among others, a possibility for the first time. In this respect, the manufacturing technology of the structures is a necessary prerequisite for process-reliable application. However, the basis for this is the evaluation of the suitability for production. For this purpose, a holistic system for the objective evaluation of highly heterogeneous surfaces was developed taking cellular metallic materials as an example. On the one hand, it is based on an optical method to evaluate cross-sectional images (Figure 1) that characterize the unprocessed state of the material and are generated by µCT or micrographs [1]. The processing condition is recorded using a tactile method. This is based on the use of a newly developed cutting-edge stylus tip in conjunction with a contour measurement instrument. As a result, specimens were analyzed that had been modified by milling, EDM and grinding operations (Figure 2). As a result, the influences of different machining conditions are analyzed in order to substantiate the suitability of the measuring method. In addition, one-dimensional parameters are introduced that can be used to evaluate and compare the surface condition and are based on the Abbott-Firestone curve. This information is a necessary prerequisite for interpretation of simulations of the surface state [2].

Roll-to-roll printing (R2R) is a critical process for the production of large area flexible electronic devices e.g. flexible photovoltaics. R2R production proceeds by the process of printing or depositing of multiple layers of conductive tracks in a controlled and positionally defined manner on meter wide flexible polymer substrates. R2R has revolutionised flexible device production if the print operating variables and web (substrate) handing are well controlled and optimised. R2R manufacture has developed capabilities from conventional paper through to high value components. Such devices have high demands on the geometric precision (spatial positioning and track thickness) to tight manufacturing specification <1um film thickness[1]. Slot-die printing technologies are commonly used in R2R to deposit conducting inks tracks to a range flexible substrates. In slot-die printing, the uniformity of printed films is directly linked to the operating variables such as ink pump flow-rate, speed of the web, ink drying and print gap [2,3]. Imperfections in R2R operation lead to unwanted topography variation, thickness/width variation or defects in thin film surfaces. This results in low productivity and large amount of waste. Unfortunately, current R2R practice is largely limited to post process metrology of the printed film surface and track dimensions, this can result in excessive wastage and increased processing costs. Real time film surface monitoring giving reliable quantitative information to process controllers is the key to successful R2R closed loop processes. The present paper presents the development of a dynamic surface assessment approach based on in plane and out of plane surface metrology facilitating successful R2R closed loop processes.

The angle resolving scattered light technique is a powerful tool for inline roughness measurement of fine machined surfaces. The key features of the sensor are that it is robust against vibrations in the direction of the optical axis, that a high measuring speed can be achieved and that extremely fine roughness structures down to the nanometre range can be detected. In many applications it has been shown that the shape of the measured angular distribution correlates very well with the tribological behaviour of functional surfaces. However, it is not trivial to establish a direct relationship between the angle-resolved data from scattered light sensors and the spatially resolved surface topography. The reason for this is the non-coherent light source used (and thus the lack of phase information), which makes it difficult to obtain direct depth information about the surface topography. This makes it difficult to compare the data from the angle resolving scattered light sensor with other surface measuring devices. To solve this problem, we present a method for the reconstruction of surface profiles from angle-resolved scattered light measurements. The mapping from the given scattered light data to surface topography is not unique, and the reconstruction of said surface topography can be classified as a non-linear inverse problem. Two key elements are necessary to solve the non-linear inverse problem. Firstly, an accurate and fast simulation of the measurement process is needed. For this purpose, we have modelled an optical digital twin of a commercially available angle resolving scattered light sensor with all the important optical features. The digital twin is programmed in C++ language with a Python interface. A virtual "real-time measurement" can be carried out through parallelisation. Secondly, a simplified model of the surface profile is needed. And for this purpose, we propose a model based on a sparse approximation of a certain type of surface topography. In previous studies, we have successfully used a Fourier based sparse approximation for turned profiles. In this work, we have generalized the sparse approximation technique using machine learning algorithms from the domain of sparse dictionary learning. We will present several examples of surface reconstruction from in-process scattering light measurements and compare the reconstructed profiles with the profiles measured by a tactile measurement system.

Editors of surface texture analysis software need to validate their calculation, ideally with an implementation-independent manner. Today, there is no universal reference for parameters, and especially for material ratio parameters, Rk parameters, and volume parameters that are based on the Abbott curve. At Digital Surf, we developed validation, several years ago, that are based on mathematics. Softgauges are created using mathematical signals and parameters are expressed by mathematical formulae. By injecting the softgauge equation through the parameter equation, it is possible to obtain a mathematical expression of the “true result”. Then, softgauges are created as files with enough resolution and used to test software calculated. This method makes it possible to confirm that a calculation algorithm is correct, but it also provides a way to check the sensitivity of a particular parameter or a particular algorithm to influence factors such as noise, outliers, slope, etc. For this study, two families of softgauges were designed to test material ratio parameters. The two main goals of this study were, 1) to obtain reference values for all parameters based on the Abbott curve, and 2) to have a method that is independent from any implementation.