07/06/2004

Deep sea laboratory analyses long-term rock creep

Following a successful three year pilot study, the researchers are working on phase two with the deployment of a larger deep sea laboratory comprising of multiple sets of apparatus to be submerged at a depth of 2km for up to 10 years, 20km off the port of Catalina, Sicily.

Time-dependent properties of brittle rock deformation and fracture are important for understanding the long-term behaviour of the Earth's upper crust. By looking at rock creep scientists and environmentalists will discover more about how rocks fail and how earthquakes and faults develop. "The traditional way of investigating this has been to carry out brittle creep experiments on rocks held at constant differential stress in a laboratory and to measure the resulting strain as a function of time," comments Steve Boon, Senior Experimental Officer and one of the principle investigators, UCL.

"Previous results have identified three creep phases with deformation being distributed during the first two and localizing into a fault plane in the third. This project is the first of its kind and aims to extend the range of achievable strain rates through much longer-term experiments made possible by using the deep sea laboratory. Initial experimental observations suggest that there is a two-stage failure process which represents a transition from distributed microcrack damage up to some critical threshold where crack interaction leads to acceleration of failure on a localised fault plane."

It is difficult to set up this sort of experiment in a conventional laboratory on land because of the long timescales involved, and the risk of the experiment being disrupted by equipment failure or power cut, for example. Conducting the experiments at a depth of 2km mean the experiments can take place over a much longer time span and the risk of failure over time is reduced.

To simulate the conditions that rock is subjected to over a long period of time, the deep sea laboratory has been designed to load rock with the stresses it would experience in the natural world over thousands of years. The constant 20MPa pressure from the weight of the water is applied to a set of different size pistons in a pressure amplifier. This loads the jacketed water-saturated quartz rock sample, 40mm in diameter. Two sealed pressure housings alongside the sample each has an SLS 095 linear displacement sensor to measure the small deflections in displacement which are proportional to the strains experienced by the rock.

"We chose the SLS 095 sensor because it is capable of withstanding the extreme conditions of our project," says Steve. "The deflections to the actuators are as small as 10 microns and a certain level of sealing is required. SLS 095 is reliable and has a high level of sealing which will prevent the ingress of water, should the main pressure seal leak. The additional seals in the transducers protect the internal components. It has excellent resolution and high repeatability with little need for maintenance - a key issue as the apparatus will be immersed at 2km for up to 10 years and possibly beyond. I also have good experience of using them."

SLS 095 is a linear position sensor designed to provide maximum performance benefits within a compact body diameter of 9.5mm. The minimum size and weight make it suitable for robotics and small mechanisms. It is available with two mounting kits and the cable is integrally moulded to deliver excellent strain relief with secure sealing.

In initial experiments data was stored within the instrumentation package for up to six months - the lifetime of the battery pack. In the new set up a fibre optic telecommunication cable will communicate the data via a submersible telephone cable direct to Catalina in Sicily then to UCL. As the experiments are long term the sampling interval is every 45 to 60 seconds.

Information on the actuator displacement, strains experienced by the rock and the way in which rock behaves over a long period of time will help scientists in their understanding of failure processes in plate tectonics, earthquakes and faults.

"SLS 095 sensors will be used in the labs because we are extremely pleased with how well they are working with the rest of the apparatus and the results they have given us" says Steve.

This site uses cookies to improve our service, you can find out more here - for the best web experience please accept cookies

x