Magnetic Resonance Imaging of Concrete

This project is being undertaken by Emanuele Marfisi and Chris Burgoyne of Cambridge University Engineering Dept, in collaboration with Prof. Laurie Hall and Dr Gao Amin of the Herchel Smith Laboratory for Medicinal Chemistry.

The concrete must be made from materials that do not distort the MRI signal. These images show beakers of water containing pieces of granite, on the left, and white limestone, on the right. The granite contains some magnetic components that distort the signal, while the limestone is non-magnetic and therefore suitable for MRI scanning. White Portland Cement, normally used for architectural purposes, was also found to be suitable.

Once the concrete has hardened, the sample can be loaded to produce fractures and then rescanned. The image on the left, which shows the same sample as the image immediately above, shows the cylinder after it had been loaded in compression. The internal fractures can be clearly seen, as can the distinction between filamentary fractures and wider ones. These images can then be superimposed to show the relative position of the cracks and the pieces of aggregate.  A 3D movie (5 Mbyte file) constructed from a similar data set for a confined concrete sample is available.

Tests have been carried out on cylinders with and without external confinement. These four images show transverse and longitudinal slices through both types of cylinder. It has long been known that confinement affects the fracture patterns. MRI allows us to see these differences clearly and to carry out measurements on those fracture patterns. These images are shown inverted - water (and hence cracks) shows as black; solid concrete shows as white.

The structure and fracture images can be combined.  These images show pull-out specimens, in which an aramid FRP rod was cast into a concrete cylinder and then withdrawn.  The top row of images show a section where the AFRP was bonded to the concrete.  The lower row shows sections where the rod and the concrete were debonded by a plastic sleeve.  The images on the left show the structure soon after casting - the white square is a registration marker.  The central images were obtained after loading - the fracture surface can clearly be seen.  The images on the right show the previous images rotated and overlapped - in many cases the fractures skirt the aggreagte pieces, but in some cases the aggregate itself is fractured.