Digital Intelligence - Part 2

Digital Intelligence - The capability to Capture, Reconstruct and Analysis digitally the entire physical shape of a component very quickly and accurately

The 3D model forms an integral part of this link, as soon as the model comes into the design phase, it is possible to test a system or process by simulating it: control rules can be refined and a good range of settings and operating conditions can be defined.

In addition, the common model drives and helps to conceptualise the finish component and its function with the assembly, by using 3D visualisation and mix and match capabilities within the virtual products. Design and Analysis, virtual modelling processes or through low-cost 3D printing to refine the design. Simulation, the same model allows others to simulate user operational environments. Within Manufacturing, 3D printing at industrial scale, Direct digital fabrication methods enable manufacturing directly from the engineering model and reduces production times. Assembly, using immersive technology to improve labour intensive builds and re-thinking assembly. Operations and sustainment, learning from products while operational allows improvements. The use of data collected to improve the next concept.

Reverse Engineering of legacy and modified components.

Reverse engineering to create 3D models, facilitates these modern manufacturing methods accurately creating 3D CAD models of from the physical component. This is not just mean converting existing 2D drawing into an electronic 3D CAD model, as these drawing could be out of date and the approved component may have been subjected to several modification and concessions which are not reflexed on the original print.  For a component, which contains complex freeform surfaces along with geometric shapes can create additional challenges and can be a time consuming and frustrating process for design engineers.

2D drawing to 3D Model

To overcome some of the issues this creates, firstly users need to outline the capability process appropriate for the environment, which helps to demonstrate how the conversion from 2D can be achieved in a process driven and audit-able manner, whilst ensuring minimum manual input and mitigating risk. Detailed project plan and governance process should be created to help mitigate any assumptions. An audit-able process for the conversion of 2D drawing s and associated layouts should be formulated for the conversion in to a 3D CAD model. (see below workflow example diagram).

An example of audit-able process for the conversion of 2D drawings

Capturing data on the Physical component

Taking this further, using 3D scanning systems to capture accurate and high resolution data from the Physical component. Structured light systems are becoming the industry norm many provide unique blue light scanning and stereoscopic cameras system uses two (2) high resolution cameras as the method of data capture.

Traditionally systems use a structured light projector but in recent years this has been replaced with a blue LED and digital projector to reduce interference from ambient light and enable more complex projection patterns to be achieved. A fringe pattern is passed across the surface of the object and using triangulation principles the sensor can accurately measure the divergence & convergence of the pattern to create a high-density point cloud of surfaces this is then automatically converted into a polygon mesh, the so-called STL mesh.

The images are captured simultaneously by both cameras from different angles allowing fast digitizing of the surface, with a potential resolution of 0.031 mm, the sensors can capture large amounts of precise and accurate data. Using the integral software, individual scans are combined and a component alignment created all within the same software solution negating the need to export data in other 3rd party solutions, allowing the process to be fully traceable with the controlled workflows.

The capture point cloud / STL mesh model is then compared to the existing or newly created 3D CAD Model highlighting variations and deviations between the two models types

The above images show a mesh at very high resolution in all areas of strong curvatures such that fine details like the leading and the trailing edge are exactly reproduced

Components containing Freeform or lofted surfaces requires additional attention than a just a straight forward 2D to 3D conversion. The original data files that produced theses surfaces does not exist or un-available. NURBS (Non-Uniform Rational Basis Spline) surfacing is a method of creating highly accurate and rapid surfaces. The process splits the scanned STL mesh surfaces according to areas of high curvature, then divides patches created between the areas of curvature into smaller areas this helps to ensure geometric continuity throughout the surface. The NURBS surface will precisely follow the contours of the scan data, so for this reason it is essential to have high quality / accuracy scan data available. These created surfaces can be extended and trimmed as part of the created NURBS surface and individually compared to the original scanned data to ensure it conforms to shape and form of the original component surfaces. 

Geometric Features “Parametric features like holes, planes, etc.,” Using the senses integral software simple geometric features can be created using fitting algorithms, directly onto the scan data. These features can again be compared to the original scanned component and then exported individually in a host CAD system as an IGES file or parametrically exchanged directly into the CAD and so forming part of the component models history tree.

Once the Freeform and geometric features are exported into CAD as part of the history tree allowing features and surfaces to be edited freely by the host CAD system suite of editing tools to complete the finished model. Taking this further these models can be imported back into the senses software and component project file and compare and inspected against the original scan data to ensure all surfaces and features are within specified tolerances

In addition, the STL data model can be used for flow simulations and back to conceptualisation stage. It is a advantage that the simulations are based on data of real parts and not on theoretical or constructed parts.