It is possible to make a nondestructive cut.

To take a cross-sectional view and see what’s inside, the object must be cut. Can you cut without destroying? Not a problem with computed tomography (CT). For many applications, this 3-D image investigation method opens new glimpses into the internal structure of an object without destroying it.

Anything small and concealed is suitable for microfocus computer tomography (µCT). µCT is a combination of CT and a microfocus x-ray tube that provides resolution in the micrometer range. Small, in this context, is relative. The “small” spectrum reaches from chip condensers in the millimeter range to human damage. Cast parts, made from aluminum, titanium or plastic; turbine blades, foams, sensors, coils, incandescent lamps, electronics components, valves, connectors, crimps, teeth, archaeological discoveries – these are just a few examples of the broad applications for µCT. The common precondition is that the object itself be rigid and not change during rotation. Further, it must be completely penetrable by the x-ray beam from every rotational angle.

When a 2-D image cannot provide ample information, a µCT investigation makes sense. This can occur when a component’s inner construction is so complex that the overlaying grayscale value structures of the projection image permit no conclusions regarding the third dimension of the inspection piece.

Figure 1 is a high resolution image of a PLCC obtained with µCT. The 25 µm-thick bond wires are easily recognizable, as is the excess die-attach adhesive at the edges of the chip. In this image, two ISO-grayscale value surfaces are extracted from the volume data and separated from each other according to color. Compared to the corresponding 2-D image (Figure 2), essential information can be gained from the projection image because the PLCC is constructed flat; as a result of the good contrast in the component structures, no depth information is concealed.



2-D investigations are sufficient in many cases. When 3-D investigation is required, however, some systems can perform both.

Virtual volume. µCT uses a volumetric method. From a series of 2-D projection images, taken from several rotational positions divided around 360°, the 3-D structure of the object is determined from its x-ray absorption. This principle is shown in Figure 3. A beam cone is emitted from a point-formed radiation source (focal point of the microfocus x-ray tube), fielded by the real-time detector and then evaluated. Between them, the inspection piece sits on a pivoting mount, the axis of which is aligned perpendicular to the central connecting line between radiation source and detector. A computer evaluates the 2-D images and geometric data, and then generates the depiction of a virtual volume. The higher the dynamic of the detector, the finer the volume image can be. Modern flat image detectors provide grayscale value resolution of 16 bits = 65.536 grayscale values. The volume depiction is assembled from “voxels” (volume elements); size is precisely known so that very accurate measurement can be conducted within these virtual volumes. A typical cube of such a virtual volume consists of 5123 voxels.


Voxel resolution depends on the geometric enlargement and the detector pixel resolution. The inspection object – or its pertinent area – must fit within the radiation beam cone during rotation and cannot come in contact with either the x-ray tube or the detector. The smaller an inspection object is, the larger its geometric enlargement can be, and therefore, its voxel resolution will be larger as well. With a typical µCT system, a 10-mm thick coil can be reconstructed with a voxel resolution of 20 µm. With smaller components and larger detectors with higher resolution, voxel resolutions of a few micrometers can be reached.

Dr. Udo E. Frank is responsible for µCT systems sales at Viscom AG (viscom.com); udoemil.frank@viscom.de.