Comparing point clouds and CAD models on the Cloud


The power efficiency of water turbines is strongly related to the shape of their blades. The CAD-model and the models used for calculating the power efficiency do not exactly represent the blade shapes produced. A turbine and its blade last decades. During their lifetime the blade shape is worn by silt and sand. Under some circumstances cavitation can severely damage the blade surface. The application experiment thus serves two purposes:

  • Comparing the produced turbine blades with their nominal shape (CAD-shape).
  • Checking older turbines and their blades for wear and deciding on the need for repairs and upgrades.

Description of the infrastructure

  • Local test infrastructure at SINTEF
  • Arctur HPC Cluster and Cloud using the CloudFlow services built on SINTEF’s TINIA infrastructure
  • Stellba supply CAD-models and point clouds
  • Test and usage scenarios, validating the experiment by University of Nottingham.

Detailed description of the demonstrator

In the case of comparison of measurement and CAD-models for water turbine blades the software can be built around atomic geometric services. These can be combined in the proper way through the workflow control mechanisms of CloudFlow. This is in some way returning to the idea of product specific CAD/CAM-solutions that dominated the early 1980s. The manufacturing industry can once again tailor their CAD/CAM and metrology solution by combining Cloud services best suited for the challenges.
Innovation and novelty (business perspective): The Cloud opens up the possibility to break down the technology wall between measurement and CAD/CAM. This is achieved by tailoring niche services in the Cloud addressing specific product classes where the volume is too small for commercial software to be profitable when run on client PC or servers. By targeting specific product classes the functionality can be reduced to a necessary minimum, avoiding comprehensive and often complicated user interfaces, and at the same time addressing the global manufacturing market for said product classes.

Expected results (technical perspective): The result of the application experiment will be a Cloud service with minimal requirements to the hard- and software of the client computer. The solution will include the following atomic service

  • Import the nominal CAD-model of the turbine blade, use of JOTNE’s Express Datamanager already used by SINTEF’s atomic CAD-functionality
  • Import the point cloud
  • Register the point cloud with respect to the turbine blade
  • Compare point cloud and measurement, use of SINTEF’s atomic CAD functionality.
  • Create an error map for the CAD-model
  • Visualize the shape difference.

Evaluate performance indicators (simplicity of use and quality of error map for CAD-model) – they should show simplicity of use, correct difference between measurement and CAD-model.



Leader: SINTEF (Norway).
Participants: University of Nottingham (UK), Arctur računalniški inženiring d.o.o.(Slovenia), Stellba Hydro GmbH & Co KG (Germany)

Machinery & equipment

Modelling & Simulation