Rock Deformation

The images on this page illustrate our unique rotary shear apparatus used in friction studies. The advantage of the rotary shear geometry is that the ring-shaped samples can be rotated around as many times as desired and so arbitrarily large slip can be attained at elevated confiing pressure. Both photographs and drawings are on this page to adequately show the machine. The labeled drawings show some of the functions of the various parts. Comparison with the photographs allow one to see what they really look like.

The apparatus is interfaced to a Hewlett Packard UNIX workstation and Hewlett Packard 3852 Data Acquisition and Control boxes. Experiments are conducted by entering commands into the computer, either manually or by a predetermined set of instructions in a "procedure file." Output is to a display screen, a digital plotter for real-time display of the history of the experiment, and to disk for later analysis. Some of the photographs show this equipment.

The middle part of the rotary shear apparatus, showing the components that produce the rotary motion. The small schematic diagram of the machine shown just below to the left, and the larger version of the diagram below that, show and label these features. The electro-hydraulic motor that provides the primary rotary motion is shown projecting toward the viewer at the bottom of the photo. The middle platen is just above the level of my waist, and the main thrust bearing rests on it. The rotary servo is at the level of my shoulder. The bottom of the pressure vessel is shown projecting down from the middle platen, and the piston with a slip-ring assembly for carrying the signals form the internal load and torque cells is seen just below that. The 1.0 GPa gas pressure intensifier and 1.0 GPa remotely operated valves are show in right part of the photo near the bottom and top respectively.

A three-part summary diagram that shows with pink dotted lines the relations between the overall machine, the sample assembly with internal displacement transducers, and the sample with its sliding jacket assembly. Enlargements of each of these three diagrams are below.

Diagram of the overall machine on the left and a photo of it on the right (prior to construction of a wooden platform that allows access to the top end).

The load, torque, rotary displacement, and axial displacement are all measured by transducers inside the pressure vessel, the load cell, the torque cell, the resolver, and the LVDT, respectively. Therefore, no correction needs to be made for friction at the pressure seals. Also, as shown in more detail in the diagram of the sample assembly below, nearly all of the elastic distortion of the machine is removed from the displacement measurements.

Both the axial and the rotary motion can occur under high speed servo control. In the case of the axial motion, this allows constant normal stress if the feedback signal for the servo is the axial force on the sample, measured with the internal load cell. In the case of rotary motion, the servo allows the stiffness of the machine to be electronically increased by using the resolver to provide the feedback signal.

More detailed photos of the machine. The wooden platform is just above my head in the left photo. The pore pressure system above the platform mostly obscures viewing the large pressure vessel. The right photo shows details of the gas pumping system, and the axial and rotary drive system. The spiral arrangement inside the top of the tie bars acts to guide hydraulic hoses that drive the high speed rotary servo system as they wrap around the machine as displacement occurs. When the rotary servo system is used only three revolutions can be made because of limited space for the hoses to wrap around. If the hoses are disconnected, as in the first photo on this page, there is no limit to the amount of rotation available, but rotary servo cannot be used.

A cross-section of the sample assembly shows 1) the resolver that measures the rotation of the bottom sample relative to the top sample, 2) the LVDT that measures the axial motion between the two sample halves, and 3) the alignment bearing.
The resolver body and the LVDT body are mounted on a tube that is attached near the top sample, and the inner rotating shaft of the resolver and moving core of the LVDT are mounted via the wedge assemblies near the bottom sample. The bellows allows accurate transmission of rotary motion to the resolver, while preventing axial motion of its rotating shaft that would ruin its bearings.

The alignment bearing keeps the to and bottom half of the sample running concentrically, while allowing axial motion. By acting on a small diameter shaft the alignment bearings contribute negligible torque to that measured by the internal torque cell.

Detail of the sample and the sliding jacket assembly used in rotary shear. The upper and lower sample rings may either touch each other directly in the case of experiments on initially bare surfaces, or they may be separated by a layer of crushed rock that simulates fault gouge.
The jacket assembly consists of O-rings and Teflon rings, against both the inside and outside boundaries of the sample. The O-ring prevents the gas that provides the confining pressure from getting to the sample. The four Teflon rings transmit the pressure from the O-ring to the sample to provide the confining pressure on the sample. If the O-ring was not isolated from the sample by the low-friction Teflon, the O-ring would be torn apart because friction between it and the sample would cause its top and bottom halves to move in different directions.

Not shown are ports that access the top and bottom sample rings to allow pressurized pore fluids to flow through them and across the sliding surface or gouge.

Dr. David Goldsby operating the rotary shear apparatus by entering commands to the HP computer. The machine itself is in a separate room in order to protect the operators from harm in case of a catastrophic leak or failure of the pressure vessel. One of the gas bottles that provides medium pressure gas to the gas pumping system in the next room is visible to the left, and some of the pressure gages are visible on the wall and panel. Above the pressure gages on the panel are shaft handles that run through the wall to the machine and remotely operate the 1.0 GPa gas valves. Just above those handles are a series of bubblers connected to plastic tubes that are used to detect the location of any gas leaks.

From left to right are David Goldsby, Giulio Di Toro, Terry Tullis, and Dr. Naoyuki Kato, looking at a plot of some experimental data on one of our 19" color monitors. The data is stored on CDs for ease of later access and because it is a stable storage medium for long-term data archiving. One of our IBM RS/6000 computers is on the shelf above the monitor.

Naoyuki Kato adjusting the reference voltage for the axial servo system. In the rack to the right at the top is a custom-made commercial interface box that deals with the analog and digital signals to and from the resolver and to the rotary servo. Below it is an electronics box that we made that controls the upper and lower pore pressure systems, including its two servo controllers, and two LVDT signal conditioners. Below that are the HP 3852 Data Acquisition and Control Interface units that contain the digital voltmeters, multiplexers, stepping motor controllers, digital I/O units, HPIB interface to the UNIX computer, etc.

In the rack to the left from the top is an overload failsafe circuit board, the signal conditioners for the 3 other LVDTs in the system, servo controllers and reference voltage generators for the axial and rotary servos. In the gray panels at the height of Naoyuik's right elbow are manual/computer controls for the gas booster and intensifier parts of the gas pumping system, for the electrohydraulic stepping motor and its brake/clutch system for gear changes. Below that is the HP UNIX computer that controls the machine and data acquisition.

Giulio Di Toro holding the sample grips into which quartzite samples have been epoxied. He is sitting in front of the digital plotted that provides areal-time graphic record of any of our 19 channels of data that we choose to plot to keep track of the progress of an experiment.

Terry standing on the platform at the level of the pressure vessel. The pressure vessel is encased with a blue-painted water-cooling jacket. The view of the pressure vessel is obscured by half of the flow-through pore pressure system. In the upper part of the view are two of the four remotely-air-operated valves that allow flow of pore fluid through the sample during sliding, changing its chemistry, etc. The half-inch diameter high-pressure piston generating the pore pressure is seen entering the bottom of the pore pressure intensifier just in front of the left part of the pressure vessel. The servo valve for the hydraulic cylinder loading the high-pressure piston is at the bottom edge of the photo, as is a solenoid valve that is part of a failsafe system that prevents the pore pressure from getting too high.