This illustrated essay covers material introducing the concept of rheology in structural geology. There are links in the text to further information together with other parts of the learnstructure web resources.
START HERE
Rheology is concerned with relating the response of a material to the forces that act upon it. As you have seen from the virtual field trips, the ways in which rocks can respond to forces are highly variable. Not only do different rock types respond in different ways but also the same rock type can show various responses depending on the conditions under which the force is applied. As a general principal however, take it that rocks deform in the easiest way open to them. By that we mean that if it takes less force to break a rock than it does to distort it, the rock will break...
Geological we refer to forces as stresses (the force per unit area) and the response in terms of deformation is some form of strain.
Let us consider some examples.
In school physics classes you may have conducted a loading experiment on a spring, to demonstrate Hooke's Law. In the late 17th century, Robert Hooke found that the extension of a spring was simply proportional to the weight applied to it. Therefore, if you plot the change length of the spring (elongation = strain) against the weight (proportion to stress), you get a straight line (constant ration between stress and strain). This happens regardless of how stiff the spring is but springs with different stiffnesses generate lines with different slopes. Therefore the ratio between stress and strain is a material constant (elasticity, sometimes called Young's modulus). If the weight is removed from the spring the spring returns to its original length. This behaviour is termed elastic - and the strain induced is not permanent - it exists only which the load (stress) exists.
A classic way of viewing the mechanics of earthquake faulting comes from this. Consider a brick on a tabletop being dragged by a spring (often called a "slider-block"). In order for the brick to slide the pull on the spring must exceed the frictional resistance to sliding of the brick on the tabletop. You can pull the spring back and forth before the brick slips - this is elastic behaviour. But if the pull (stress) become high enough to overcome the friction then the brick slips and the spring recovers some (or all) of its strain.
Geologists know, from very accurate ground surveying, that analogous behaviour is shown by faults. For a fault to move the frictional resistance to sliding must be overcome. It does so dramatically - causing an earthquake, just as most slider-blocks suddenly slip. But the during the period between earthquakes the ground adjacent to the fault is sheared, rather like the spring being stretched. The earthquake releases this shearing and the ground springs back.
In a development from Hooke's law, when large loads are placed on springs, or loads are left on them a long time we find that the springs become permanently deformed. That is, when the loads were removed the springs did not return to their original form. If you increase loads (stress) you find that the strain increases too quickly for the simple elastic behaviour. Engineers refer to this as plastic behaviour and it leads to the permanent strain (some geologists refer to this as a non-recoverable strain). In geology this term is commonly equated to a rock acquiring a distortional strain (rather than simply breaking). For the geological example above, if the fault does not rupture (to make an earthquake) then the surrounding ground may become permanently distorted - in effect a shear zone.
Before leaving our earthquake and fault scenario behind, we might consider some further complexities that happen in nature. What might we do to our slider block system to make it slide more easily?
Clearly the best things to do all reduce the frictional resistance to sliding. One way might be to spread lubricant on the tabletop. Geologically this can happen in faults if the fault zone contains lots of weak material - such as fine-grained rock flour (most fault gouges - tectonic talcum powder!) where the grains can slide past each other. In some situations weak rock types can be smeared in fault zones (shales or rock salt). Another way of reducing frictional resistance is to "float" the slider block on water. Geologically this can happen by raising the pore fluid pressure, effectively jacking the wall-rocks apart.
The effect of reducing the frictional resistance in fault zones is to reduce its ability to support loads (stress). Our slider block will move at smaller pulls on the spring. So weak fault zones need less force to move - and it might be inferred - emit less energy when they slip. So, intuitively, weak fault zones should have fewer large earthquakes than stronger faults.
Let's look at our spring analogue a little more closely. If our spring is made from steel, at room temperature, it will behave elastically under a wide range of loading conditions. You have to increase the weight a lot to move from elastic to plastic behaviour. This is why steel springs are commonly used for spring balances. But if you heat up the steel the ability to deform elastically is gradually lost. It can deform plastically at lower stresses - rheology is temperature-dependent. This behaviour has of course been used since the stone age (it's why blacksmiths have forges). The same is true for rocks. Heat them up (move deeper in the crust) and they tend to deform plastically rather than elastically during tectonics. Obviously this also depends on how quickly you attempt to deform the rock (or steel bar). All solids (by definition) are able to transmit seismic waves (which are elastic) as short sharp shocks. But over longer periods they can flow (deform plastically - some people refer to this as "creep"). At the rates of most tectonic strains, rocks tend to deform plastically once the temperature has exceeded about 70% of the melting temperature (measured from absolute zero). But melting point is a function of the material (ice is lower than granite is lower than gabbro) so the composition of the rock is important. We will investigate these dual controls of temperature and composition in an analogue experiment in week 11.
If you look at the magnitude of earthquakes along a particular fault zone as a function of depth in the crust you commonly see a remarkable feature. The larger earthquakes tend to happen deeper. But below about 10-12 km there are very few earthquakes (apart from in subduction zones). So why does this happen?
The effect of burial - or going deeper in the crust - is to increase the confining pressure. This is equivalent of adding weights onto the slider block in our analogue above. This in turn increases the frictional resistance to sliding - thus requiring a greater pull on the spring to get a slip. When the block does slip it has the potential for a lot of energy release - a bigger earthquake. But breaking rocks becomes harder with depth. Breaking rocks generally increases volume (what was once solid rock is now the same volume of solid rock plus holes). This becomes increasingly difficult with depth as it acts against the confining pressure. So rock breaking (cataclasis) tends to be restricted to parts of the crust with relatively low confining pressures - the top 10-12 km.
At depths greater than 10-12 km, and temperatures greater than about 300 C, many common rock types deform on geological time scales in inelastic ways. They retain permanent strains and therefore deform plastically. Different rock-types respond differently, they have differing degrees of stiffness or viscosity. You will know from your experience of materials (e.g. chocolate or butter) that viscosity depends on temperature. The hotter the material the runnier it generally gets. In general as a package of different rock-types heat up together their viscosities become similar to each other.
But there's more to it than this. One type of ideal viscous material works like this. The greater the pull (stress) the faster it stretches (rate of strain). So stress is proportional to strain rate - not to strain (as for elastic materials). If you take the stress off the material ceases to strain (strain rate becomes zero) but the strain that it has accumulated is not recovered - it is permanently deformed. These simple materials, where stress is simply proportional to strain rate, are described as being linear-viscous (the stress/strain-rate plot is a straight line) or ideally Newtonian. However, many rock-types show more complex behaviours and are commonly referred to as being "non-Newtonian".
But just a minute - rocks can deform without breaking even at low temperatures.
This is true - although they can't do so very quickly! If rocks are given enough time to deform (the strain rate is low) then other mechanisms can kick in. Chemical solution and re-precipitation are particularly important in this regard. These processes (diffusional mass transfer) are strongly sensitive to surface area (caster sugar dissolves faster than a large sugar lump) as well as temperature. So fine-grained rocks are in general more prone to these processes than coarse grained parts of the same rock type. Of course different rock types are more soluble than others .. another potential cause of variety in structural geology.
While it is possible to measure viscosity of rock materials at the surface under laboratory conditions, in practice it is very difficult to extrapolate short-duration experiments to geological time-scales so as to make good estimates of their long-term viscosity. While it might be difficult to determine actual rock viscosities it is commonly much easier to deduce the relative viscosity (or stiffness) between different rock types. Geologically this qualitative approach is embraced by the concept of rock competence. Highly competent (stiff) layers, embedded in lower competent (runnier) material will form buckle folds and can tend to fault rather than flow. In layer extension it is the competent layers that boudinage with the less-competent matrix flowing around. Lower competent rock-types can distort more easily so commonly pick out a cleavage. Competent layers are commonly made up of limestone or sandstones with the incompetent layers made by shales (with a finer grain size and more mica or clay).
The rheology of rocks is enormously complex, to which the notes above only superficially allude. Not only does the nature of the rock (composition, grain-size) itself play a role but so too do external factors such as the confining pressure (depth) and temperature together with stress and strain rate. As if these variables are not enough, rocks can change behaviour with time during deformation. If you stretch the plastic collar to a four-pack of drinks cans you can find that after a small amount of elastic and then plastic behaviour it suddenly becomes softer, necking out to a thin strip. This is an example of strain softening. But the thin plastic collar is not much harder to stretch than before - so it has strain hardened. Eventually the material breaks. Rocks can show similar behaviour.
Take faults. These represent examples where deformation has become strongly focussed while the adjacent rocks are only weakly deformed. In these situations we might rightly assume that the fault itself is weaker than the surrounding material within which it lies. As each increment of slip is added to the rock volume it is the fault that deforms - it is easier to deform along the fault (slip) than to deform the surrounding rocks. If we assume that before the deformation began there were no faults present, the formation of a fault zone by breaking creates a weak zone that focusses further deformation. So breaking rock and making a fault makes the whole rock weaker. What is it that makes the fault weaker than its surroundings? Perhaps the fractures in the fault have elevated pore fluid pressures or the ground up fault gouge is easier to deform than the surrounding, coarser grained original rock. Both are possible explanations. A similar weakening can happen in mylonites in shear zones. Deformation can generate finer-grained material that is easier to deform than the original rock. In both cases the act of deforming the rock has changed its rheology to make it weaker. But it can go the other way. For example, DMT deformation in a fine-grained limestone may result in reprecipitation of calcite as coarse crystals in veins. These new grains will have less surface area per unit volume compared with the original rock - thus reducing the rate of dissolution and thus the rate of strain by this mechanism. So the rock becomes harder to deform as a consequence of the deformation (strain hardening).
Yes, complexity is the message.. but there are common underlying principles hopefully some of which you can now appreciate.
top of page