Skip to main content

First article tomography inspection methodology advancement


Point cloud data are used as a link between different methods of digitising and the CAD representation. Certainly, they are also used as a base for comparison in first article inspection procedures. Since the direct output of a CT measuring device is a series of CT grey value images, the software tool has to calculate a dense point cloud on the surface of the measure part. Without the software tool, the CT measurement data cannot be imported into CAD systems; also, 3D measurements and comparisons to nominal CAD models are impossible. Features are: - Import of a series of grey value images - Accurate image segmentation (contour extraction) - Sub pixel operations - Handling of large datasets - Export of point clouds EMPA has developed a near productive software prototype for point cloud generation named Smart Points with following advantages: - Vast reduction of time and costs for point cloud generation (few minutes for processing, fully automated, user-friendly interface, fully reproducible) - Handling of huge CT data sets (no limits) - Improved accuracy using slice by slice optimised thresholding Arbitrary slice spacing possible (suitable for Smart CT data).
The software tools will be used for visualising deviations between the CT-scan (measured) data and the CAD (nominal) data. Deviations are shown slice-by-slice (according to the CT-scanning) or in any arbitrary plane. This way of showing results enables the user to quickly analyse the overall deviation from the CAD-model and to plan and quantify remedies for detected problems.
The software tools will be used for automating the analysis and the reporting of repetitive measurements of a batch of similar products. Visual Basic will be the platform for this automation, since Visual basic provides most performance, is the most flexible and easy to use. Writing COM-interfaces for all analysis and reporting functions make the underlying technology made available to a wide range of software platforms and languages.
New triangulation strategies based on CT-images have been developed. This has resulted in prototype software. The size of the resulting data is still too large so reduction algorithms have also been developed. The resulting algorithms are (semi-) multi-resolution versions of the existing marching cubes algorithm.
The huge amount of data generated in the total inspection process using CT-images is very difficult to present in a usable measuring report. The developed tools provide elegant solutions and are commercialised through integration in commercial inspection packages.The software tools will be used for reporting of the CT-scan measurements and subsequent comparisons with nominal data. Electronic and hardcopy reporting will be provided.
Two new methods for distance measurements on grey-values have already been developed: the four points method and the four intervals method. Methods for volume, area and outer dimension measurements directly on grey-values are implemented and tested. Tools for measuring geometric properties, through fitting of geometric entities directly on grey-values are under development.
The check of the minimum wall thickness of a cast or mould part is a standard procedure e.g. in first article inspection of an engine-head. Currently, it requires the destruction of the engine-head and a long-winded tactile measurement procedure. The software was developed by Daimler Chrysler, which automatically checks the CT-scans of a part for walls that are thinner than a given threshold. These algorithms are based upon the triangulation of the CT-data (STL-file). A special visualisation tool was also developed in order to obtain an easy and expressive visualisation of the critical walls (wall thickness smaller than threshold) of the part.
Different software tools for comparing CT-scan data and CAD-models are made in this project: - Registration algorithms of CT-scan data to CAD-models - Automatic calculation of deviations slice-by-slice - Inspection plans for automated subdivision of CT-scan data - Reporting of measurement results These software tools are specific to first article inspection with CT-scan data.
The consortium aims to accelerate and to optimise First Article Inspection by Computed Tomography (CT) especially in the case of very big complex parts. This will lead to a drastic reduction of mechanical complex parts development process chain, and time to market for innovative products. The result R5 covers all aspects of the project work, but as one single result to be exploited e.g. as a complete service.
Smart CT includes auxiliary software tools and a pre-processor which is logically located between CAD model and X-ray computer tomography. Smart CT pre-processes CAD models of industrial objects, (e.g. castings), before scanning. The results of pre-processing are optimised parameters that will be used for X-ray CT scanning. Optimised in the sense of reducing process time/costs of scanning.The SmartCT pre-processor tool provides an interface to: - CAD system for data acquisition based on STL data; - Computer tomography system (scanning parameters). The main advantages are: -Remarkable reduction of scanning time/cost (up to 50% depending on object geometry) at comparable accuracy; - Improved competitiveness of CT as a method for first article inspection; - Integrated in process chain.