Assessment Of Single Point Uncertainty In Dimensional CT Measurements.pdf

Problem: Unlike traditional CMMs, CT measurements involve repeated scans of the same part, yielding slightly varying surface datasets even with identical acquisition parameters due to factors like noise and artefacts. This single point noise contributes to the overall uncertainty of CT-based dimensional measurements.

Method:

1. Simulation Model: The researchers developed a realistic simulation model of a CT system to generate synthetic measurement data.
2. Analysis: They analyze local variations within a series of 20 simulated (or actual) CT scans by comparing each scan's surface dataset against a nominal CAD model. This comparison calculates:
Single Point Noise: Standard deviation of repeated measurements at each point on the nominal surface in the direction of the surface normal.
Mean Systematic Offset: Average perpendicular distance between the extracted surface points and the nominal surface.
3. Validation: The method is validated using an existing uncertainty mapping technique to assess local CT measurement quality.

Key Findings:

The proposed method successfully identifies areas with increased uncertainty caused by artefacts and quantifies the single point uncertainty at each surface point.
It offers a locally resolved assessment of CT measurement quality, complementing global quality parameters.

Applications:

This method can enhance the accuracy and reliability of dimensional CT measurements in various industries by:

Better Understanding Uncertainty Sources: Identifying specific regions with high uncertainty allows for targeted improvement of acquisition parameters or data processing techniques.
Enhanced Quality Control: Locating areas of low measurement quality enables more precise analysis and interpretation of CT results, especially in complex geometries.