Steven Dutch, Professor Emeritus, Natural and Applied Sciences, Universityof Wisconsin - Green Bay

Elastic material deforms under stress but returns to its original size and shape whenthe stress is released. There is no permanent deformation. Some elastic strain, like in arubber band, can be large, but in rocks it is usually small enough to be consideredinfinitesimal.

Many elastic materials obey **Hooke's Law** behavior: the deformation
isproportional to the stress. This is why spring balances work: twice the weight
results intwice the deformation.

For materials, Hooke's Law is written as: **Stress = E Strain**. Alternatively,
therelationship is sometimes written **E = Stress/Strain**. This is the reverse
of the waythe law is written in most physics texts. In physics, we can often
apply the stress in acontrolled way and we are interested in predicting the
behavior of the spring, forexample, how it oscillates. In materials science
and geology, we often know the strain andwant to know what stress produced it.
The two versions are equivalent; the only differenceis which side the constant
is written on. The constant **E** is called **Young'sModulus**. Because
strain is dimensionless, **Young's Modulus has the units of pressure**or
stress, i.e. pascals.

If strain = 1, stress = E. Thus, Young's Modulus can be considered the stress it wouldtake (theoretically only!) to result in 100 percent stretching or compression. In reality,most rocks fracture or flow when deformation exceeds a few percent, that is, at stresses afew percent of Young's Modulus.

The seismic P- and S-wave velocities in rocks are proportional to the square root of E.

For most crystalline rocks, E ranges from 50-150 Gpa, averaging about 100.
If we take100 Gpa as an average, and consider one bar (100,000 pa) of stress,
we have: 10^{5} = 10^{11}Strain, or Strain = 10^{-6}.
Thus, rocks typically deform elastically by **10 ^{-6} per bar ofstress**.
This is a useful quantity to remember. Elastic strain in rocks is tiny - eventen
kilobars typically results in only one percent deformation - if the rock doesn't
failfirst.

When a material is flattened, it tends to bulge out at right angles to the compressiondirection. If it's stretched, it tends to constrict. Poisson's Ratio is defined at theratio of the transverse strain (at right angles to the stress) compared to thelongitudinal strain (in the direction of the stress).

ex=strain in x-direction

ey=strain in y-direction

Poisson's Ratio = ey/ex

Note that the ratio is that of *strains*, not *dimensions*. We
would notexpect a thin rod to bulge or constrict as much as a thick cylinder.

For most rocks, Poisson's Ratio, usually represented by the Greek letter
nu (ν), averagesabout 1/4 to 1/3. Materials
with ratios greater than 1/2 actually *increase* involume when compressed.
Such materials are called *dilatant*. Many unconsolidatedmaterials are
dilatant. Rocks can become dilatant just before failure because microcracksincrease
the volume of the rock. There are a few weird synthetic foams with *negative*Poisson's
Ratios. These materials are light froths whose bubble walls collapse inwardunder
compression.

Poisson's Ratio describes transverse strain, so it obviously has a connection
withshear. The **Shear Modulus**, usually abbreviated G, plays the same role
in describingshear as Young's Modulus does in describing longitudinal strain.
It is defined by **G =shear stress/shear strain**.

G can be calculated in terms of E and v: **G = E/2(1 + **
ν**). **Since v ranges from 1/4
to1/3 for most rocks, G is about 0.4 E.

The bulk modulus, K, is the ratio of hydrostatic stress to the resulting
volume change,or **K = pressure/volume change**.

It's easy to show the relationship between K, E, and Poisson's ratio (ν). Consider theeffects of pressure P acting on a unit cube equally along the x- y- and z-axes. Thepressure along the x-axis will cause the cube to contract longitudinally by an amount P/e.However, it will also bulge to the side by an amount vP/e, in both the y- andz-directions. The net volume change just due to the component in the x-direction is (1 -2ν)P/e. The minus sign reflects the fact that the bulging counteracts the volume decreasedue to compression. Similarly, compression along the y- and z- axes produces similarvolume changes. The total volume change is thus 3(1 - 2ν)P/e.

Since K = P/volume change, thus **K = E/(3(1 - 2**ν**))**.
Since v ranges from 1/4 to1/3 for most rocks, K ranges from 2/3E to E.

Physically, K can be considered the stress it would take to result in 100 per centvolume change, except that's physically impossible and elastic strain rarely exceeds a fewpercent anyway.

If ν = 1/2, then K becomes infinite - the material is absolutelyincompressible. Obviously real solids cannot be utterly incompressible andtherefore cannot have ν = 1/2.

There are really only two independent quantities, so if we know any twoquantities E, v, G and K, we can calculate any others. The relations are shownbelow. Find the two known parameters and read across to find the other two.

Known: |
E = |
ν = |
G = |
K = |

E, ν |
E | ν | (E/2)/(1 + ν) | (E/3)/(1 - 2ν)) |

E, G |
E | (E/2G) - 1 | G | (E/3)/(3 - E/g)) |

E, K |
E | (1 - E/3K)/2 | E/(3 - E/3K) | K |

G, ν |
2G(1 + ν) | ν | G | (2/3)G(1 + ν)/(1 - 2ν) |

G, K |
12G^{2}/(3K + 4G) |
(2G - 3K)/(3K + 4G) | G | K |

K, ν |
3K(1 - 2ν) | ν | (3/2)K(1 - 2ν)/(1 + ν) | K |

Viscous materials deform steadily under stress. Purely viscous materials like liquidsdeform under even the smallest stress. Rocks may behave like viscous materials under hightemperature and pressure.

**Viscosity** is defined by **N = (shear stress)/(shear strain rate)**.
Shearstress has the units of force and strain rate has the units 1/time. Thus
the parameter Nhas the units force times time or kg/(m-sec). In SI terms the
units are pascal-seconds.Older literature uses the unit *poise*; one pascal-second
equals ten poises.

Few if any physical parameters show such a tremendous range as viscosity

Material | Viscosity Pa-Sec |
---|---|

Hydrogen gas, 15 degrees K | 0.0000006 |

Air, 0 degrees C | 0.000017 |

Liquid Hydrogen | 0.00001 |

Water 0 degrees C | 0.0018 |

Water 100 degrees C | 0.0003 |

Heavy Machine Oil 15 C | 0.66 |

Glycerin, 20 degrees C | 1.5 |

Honey 20 degrees C | 1.6 |

Basalt Lava | 1000 |

Rhyolite Lava | 1,000,000 |

Rock Salt | 10^{14} |

Marble | 10^{16} |

Granite, Quartzite | 10^{18} - 10^{20} |

Asthenosphere | 10^{19} - 10^{20} |

Deep Mantle | 10^{21} - 10^{22} |

Shallow Mantle | 10^{23} - 10^{24} |

Viscosity is very dependent on temperature. If it seems that water out of a boilingteakettle splashes more and soaks through clothing more quickly than cold water, that's noillusion. The viscosity of water at 100 C is only one-third as much as room temperatureand one sixth what it is at 0 C. The viscosity of glycerine drops from 6700 at -40 C to.63 at 30 C, a factor of 10,000 in only 70 degrees.

Note that, strictly speaking, solid rocks aren't viscous. The figures given
reflecttheir flow rates at the temperatures and pressures typically found during
crustaldeformation, and they are *highly* approximate and *extremely*
dependent ontemperature and pressure. Variations by several orders of magnitude
areperfectly possible and commonly seen.

Plastic material does not flow until a threshold stress has been exceeded. Plasticflow involves many different mechanisms at the atomic level and there are many differentequations for different plastic flow mechanisms. Plastic flow therefore does not lenditself to neat physical parameters the way elastic and viscous deformation do.

One of the most common forms of plastic flow is **Power-Law Creep**, given
by theformula: **Strain Rate = C (Stress) ^{n} exp(-Q/RT)**

Let's take each part of the formula in turn:

**C**is a scaling constant.**n**means that the strain rate increases much faster than stress. Typically n is about 3 but can range from a bit less than 2 to 8. Recall that with viscous deformation stress is proportional to strain rate (n=1). With power-law creep it's faster: the effective viscosity drops with stress.**Q**is the activation energy required to get crystal dislocations moving. It's typically 100-300 kilojoules per mole, sometimes up to 500.**R**is the Universal Gas Constant that turns up everywhere in physical chemistry. In SI units it equals 8.3144 joules/mole-degree Kelvin.**T**is the temperature in degrees Kelvin. As T increases, Q/RT decreases and thus exp(-Q/RT) increases, though much more slowly than exp(T). At very large T, Q/RT approaches zero and the exponential term approaches 1. This does not happen, though, at geologically realistic temperatures.

The function Exp(-Q/RT) increases approximately linearly at first, then very slowly asymptotically approaches 1. The curves here use geologically realistic values of Q. At high temperatures the temperature dependence is weak because rocks are ductile enough to deform as fast as strain can be applied. Another way to look at it is that ambient thermal energy is nearly enough to overcome the activation energy barrier. |

Power-law creep is given by Strain Rate = C (Stress)^{n} exp(-Q/RT).
If n = 1 and Q=0,then we just have Strain Rate = C (Stress), or in other words
viscous flow. Q = 0 means ittakes no energy to dislocate atoms; that is, the
material deforms under even the slighteststress. In this case C is just 1/viscosity.
However, you can't simply look up values of Cand equate 1/c to viscosity because
the other terms in real power-law creep can beextremely large.

Some everyday materials obey power-law creep or some similar behavior.

- Mayonnaise in a jar flows very sluggishly. You can tilt the jar on its side without it spilling. On the other hand, it offers no resistance to a knife and spreads easily on bread. A slight increase in shear stress (applied by the knife) results in a huge increase in strain rate or, effectively, a great decrease in viscosity.
- Ball-point pen ink has to be viscous enough to stay in place if the pen is put upside-down in a pocket, yet flow easily through a small orifice as the ball rotates. This is another material whose viscosity drops as shear stress is applied.

Materials such as these are commonly described as *shear-thinning* fluids
becausethey become less viscous with increasing shear stress. However, it's
easy to see theybehave in a manner similar to power-law creep.

**Viscoelastic** Combines elastic and viscous behavior. Models of glacio-isostasyfrequently
assume a viscoelastic earth: the crust flexes elastically and the underlyingmantle
flows viscously.

Return to Course Syllabus

Return to Techniques Manual Index

Return to Professor Dutch's Home Page*Created
23 February 1999, Last Update 26 February 1999*