The inside story ...
TWI Bulletin, May - June 2009
Cross section imagery using X-ray computer tomography
Mathew Amos is an Engineering Doctorate student based in the NDT department of TWI Ltd and isassociated with the Materials Science Centre at Manchester University. His primary area of research is in Digital Radiography and X-ray Computer Tomography. This article is generated from a Core Research Funded project for thepromotion of X-ray Computer Tomography in various industry sectors
Computed Tomography (CT) is a powerful non-destructive evaluation technique that uses a computer to produce 2-D cross-sectional and 3-D images of an object from digital X-ray images. As Mathew Amos explains characteristics of theinternal structure of an object such as dimensions, shape, internal defects, and density are readily available from CT images.
Fundamentally, tomography deals with the reconstruction of an object from its X-ray projections. The technique of tomography consists of passing a series of X-rays (in parallel, fan or cone formation) through an object, and measuring the decrease in intensity or attenuation in these X-rays by placing a detector on the receiving side of the object. These measurements of the X-ray attenuation are called projections and are collected at various angles and orientations (Fig.1). Each of the attenuation measurements is then digitised and stored in a computer where a specialised algorithm is then used to reconstruct the distribution of the X-ray attenuation to form a 3D volumetric model of the object under inspection.
Fig.1. 3D CT images showing slices taken through a mobile phone
General principles
X-ray attenuation is primarily a function of the X-ray energy used and the density and atomic number of the material being imaged. These attenuation coefficients are assigned grey values and correspond to the contrast in the finalimages. Thus the goal of CT imaging is to derive the unknown attenuation coefficients based only on the projection data acquired.
Consider the case of an X-ray passing through a sample with four attenuation coefficients as shown in Figure 2. The output intensity I is given by
Where x is the distance travelled by the X-ray. Taking the logarithm of Equation (1) and re-arranging, we obtain the projection function
The projection function represents the summation of the attenuation coefficients along a given X-ray path.
The calculation of these 1D projection functions produces what is known as the sinogram. The sinogram is then used to produce the 3D reconstruction of the object.
Fig.2. An X-ray beam passing through a specimen with four different linear attenuation coefficients
Three phases of CT reconstruction
1. CT Data acquisition
CT machines use various methods of acquiring the necessary projection data to produce a CT image. These methods are classified in terms of their scan geometry about the object. For example, in medicine, 2nd generation CT systems aregenerally used which operate a translate/rotate scan geometry about the object, whereas most industrial systems use a third generation configuration which involves use of cone beam projections and the object under study being placed ona rotating manipulator. This enables the system (Fig.3) to collect the many angles of X-ray attenuation data or X-ray projections needed to perform CT reconstruction.
Fig.3. The basic set-up of a CT system - the component rotates through 360° as the projection data is acquired using the digital detector
2. CT Data reconstruction
In this phase, specialised algorithms are used to reconstruct 2-D cross sectional images (Fig.4) from the attenuation projection functions (the sinogram).
The most common algorithm used for this process is the Filtered Back Projection method. Back projection is the mathematical process of obtaining the digital image from the projection data or sinogram, but if it is not filtered insome way, results in a very noisy image, hence the use of Filtered Back Projection.
Fig.4. The CT image is now digital in the form of a matrix of pixels, with each pixel assigned a specific grey level corresponding to the density of the material at that point
3. 3D Volume image formation
The step after reconstruction is the formation of the 3D volume of the object (Fig.5). For this phase a volume rendering software is required. The software enables the 3D model of the object to be produced allowing;segmentation of materials with different density, slicing, noise reduction and 3D measuring. 2D cross sectional slices are also displayed in the xy, xz and yz planes.
Fig.5. 3D CT image of a brass padlock
Advantages of X-ray CT
- Visualisation of the inspection results, providing easily interpreted quantitative information on the density and geometry of an object.
Compared to conventional radiography;
- CT imaging completely eliminates the superimposition of structures outside the area of interest.
- Improved probability of detection as the X-ray beam is more likely to be well orientated to any defects present.
- CT images are digital, they may be enhanced, analysed, compressed, archived, and the digital data can be used in performance calculations which enables comparisons with other NDT techniques.
Disadvantages
- Objects under examination must be small enough to be accommodated by the CT equipment available and must be fully penetrated by the X-ray energies produced by that particular system.
- Artefacts in the CT data, which limit a user's ability to quantitatively extract density, dimensional, or other data from an image.
Equipment
Micro-focus X-ray system
The X-ray Micro-focus system at TWI has a multi-metal target source with a 10 micron focal spot size capable of producing X-ray energies up to 225kV and 2mA (Fig.6). The manipulator has five axes of rotation and the detector is a CCD camera (900 x 900 pixels) with an image intensifier allowing real time radiography.
Fig.6. Inside the Micro-focus X-ray cabinet
The system is capable of CT imaging objects to a maximum 150mm in diameter. The tube ensures that using the highest magnification results in optimal edge sharpness and spatial resolutions approaching 15-20 microns. Projection image data is acquired in 16 bit format resulting in high contrast sensitivity. The CT system enables intensity distortions and pixel corrections to be performed resulting in higher image quality.
Results - Case studies
Case study 1 - Aluminium electron beam weld
Fig.7. 3D image of electron beam welded fuel-cooled heat exchanger
Component description
The component is a section of an electron beam weld found on a fuel cooled heat exchanger with the approximate dimensions of 72 x 8 x 20mm (Fig.7). The heat exchanger was due to be used on the Euro Fighter, but followingNDT trials was found to be defective. The NDT method originally used was 2-D film radiography, which found the weld to be porous. Computer Tomography inspection was applied as the client required more information in terms of the sizeand extent of the porosity and also the exact location of the porosity within the depth of the weld. All these measures are particularly difficult with 2-D radiography.
Inspection results
The results obtained highlight the capabilities of the micro-focus X-ray system at TWI Wales. Figures 8 and 9 are the 2D radiographs of the weld taken at different angles and Figures 10 and 11 show the CT images of the weld. Both techniques were able to detect the porosity present in the weld. However the CT images emphasise its advantage over 2D radiography.
Fig.8. 2-D radiograph of the heat exchanger
Fig.9. 2-D radiograph of the weld taken at a different angle
The post processing software allows the 3D volume image to be rotated and sliced in any direction, allowing the user to locate and analyse defects more accurately (Figure 10). The software also displays 2D cross sectional image stacks in three different orientations. Figure 11 shows a 2D slice in the sagittal direction with defects ranging in size from approximately 6mm to 200 micron.
Fig.10. 3D CT image sliced through weld
Fig.11. 2D cross sectional slice in the sagittal direction through weld
Combining the information in the 2-D slices with the 3-D representation allowed the client to visualize the full extent of the porosity present, including size and location within the weld.
It is clear from the results that CT imaging provides improved contrast over digital radiography and is an ideal inspection technique for locating and sizing volumetric defects such as weld porosity.
Case study 2 - Filter cap
Component description
This component is a plastic filter cap from an automotive petrol engined car (Fig.12), with approximate dimensions 49 x 100 x 100mm. Surface cracks had been identified in failed components and the client wanted to investigate if the crack had propagated from the die line. Micro-CT was chosen as the most suitable technique due to its ability to detect tight wall defects and inspect internal flaws that are not visible from the surface.
Fig.12. 3D CT image of the filter cap, showing the surface crack
The material properties and dimensions of this particular component dictated that a CT scan could be acquired using a low energy setting with minimal filtration resulting in increased contrast sensitivity in the final image. The scan results identified the crack propagation within the sample but further analysis found that unexpected porosity was also present (
Fig.13-14). These inspection results provided by TWI helped the client identify the cause and re-evaluate their manufacturing methods.
Figures 13 and 14 are from a selection of the images sent to the client after analysis. Figure 13 is the 3D volume of the component sliced through the area of interest. Figure 14 is a 2D cross sectional slice in the axial direction through the same area of interest.
Fig.13. 3D CT image slice, showing both the crack and the porosity
Fig.14. 2D CT cross section in the axial direction
Final comment
X-ray computed tomography is a radiographic inspection method that provides an ideal examination technique whenever the primary goal is to locate and size planar and volumetric detail in three dimensions.
As the method is X-ray based, it can be applied to both metallic and non metallic materials and is able to inspect very complex geometry components. Additionally internal detail of components can be reconstructed to an accuracy of microns producing inspection results that cannot be provided non-destructively by any other technique.
