Material Information

Mudflow a review
McQuilkin, Stephen Joseph
Publication Date:
Physical Description:
ix, 181 leaves : illustrations ; 29 cm


Subjects / Keywords:
Mudflows ( lcsh )
Mudflows ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 178-181).
General Note:
Submitted in partial fulfillment of the requirements for the degree of Master of Science, Department of Civil Engineering.
Statement of Responsibility:
by Stephen Joseph McQuilkin.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
17882396 ( OCLC )
LD1190.E53 1986M .M36 ( lcc )

Full Text
Stephen Joseph McQuilkin
B.S., Northeastern University, 1979
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Civil Engineering

This thesis for the Master of Science degree by
Stephen Joseph McQuilkin
has been approved for the
Department of
Civil Engineering

McQuilkin, Stephen Joseph (M.S., Civil Engineering)
Mudflow: A Review
Thesis directed by Professor Nien Y. Chang
The mudflow is a natural geomorphic process, one which has
been occurring throughout geologic history. Until recently, damage
resulting from mudflow was minimal, due, primarily to the fact that
it was limited to remote locations such as steep mountainsides and
narrow ravines. However, increased development and construction over
the years, within the floodplain and on alluvial fans has left many
areas susceptable to loss of life and property, due to the impact
force, mass transporting and inundating properties of the viscous
mudflow. Man has further aggravated the situation by stripping and
altering the natural landscape, steepening or undercutting slopes,
deforestation, and by concentrating storm runoff in artificial or
man-made drainageways.
Most of the research and documentation on the subject has
been conducted only over the last 20 years. The mudflow is often
overlooked or obscured by other mass movements, such as landslides or
floods. Included in this paper is an organized approach to the better
understanding and future research of mudflow. Definitions are set
forth in an attempt to distinguish mudflow from other types of mass
movement and to lay the groundwork for a geomorphologic
classification system. A review of some of the factors which affect
mudflow are discussed along with some of the more common mitigation
measures, both active and passive. The three stages of mudflow -

i v.
source, flow, and deposition are examined herein, however, the
focus of the paper is on the source or formation of the mudflow.
Factors which affect the formation are discussed, along with some of
the more common methods of anaylsis. Case histories for three
seperate mudflow locations are presented at the conclusion. A
parametric study is also presented, using the results of previous
researchers, in an attempt to correlate a range of values for eleven
different mudflow-related variables.
This study does not yield precise equations or solutions to
the mudflow problem, but rather an ordered approach to the future
study and research of this very important geomorphologic process.

I. INTRODUCTION.................................................. 1
Definitions and Classification............................ 3
Magnitude of the Mudflow Problem........................... 10
Mitigation Measures........................................ 14
Passive Mitigation Measures............................. 15
Active Mitigation Measures............................... 28
Early Warning System..................................... 34
Purpose and Scope of the Study............................. 37
General.................................................... 38
Stage II The Flow........................................ 41
Stage III Deposition..................................-. 52
Mudflow Models............................................. 53
General................................................... 63
Geologic Formation....................................... 65
Climatological and Environmental Factors............... 68
Human Factors.............................................. 72
The Effects of Fire and Deforestation...................... 74

IV. MUDFLOW STAGE I FORMATION............................... 80
Conditions Which Lead to the Formation of Mudflow......... 81
Shear Strength........................................... 81
Drainage Conditions...................................... 91
Methods of Analysis..................................... 102
Proposed Mudflow Failure Mechanism........................ 112
V. ANALYSIS OF PREVIOUS MUDFLOWS............................. 120
Introduction.............................................. 120
Case Histories........................................... 122
Wrightwood, Southern California May, 1941............ 122
Wasatch Front Range, Utah 1983........................ 132
Xaio-Jiang River, Hunnan Province, China................ 138
Analysis of Mudflow Parameters............................ 149
IV. SUMMARY AND CONCLUSIONS..................................... 166
Conclusions.............................................. 166
Areas Which Require Additional Research................... 171
BIBLIOGRAPHY...................................................... 178

1.1 Classification of Mudflow-Related Mass Movements.............. 4
1.2 Damage to Building Sites Versus Revised Building Codes_____ 19
2.1 Hyperconcentrated Sediment Flow as a Function
of Concentration.........................*................ 43
2.2 Typical Fluid Viscosities.................................. 47
3.1 Debris Production for Individual Debris Basins During
the 1978 and 1980 Storms.................................. 79
5.1 Characteristics of, the Dachao River Tributaries............ 142
5.2 Characteristics of the Jiang-Jia Ravine..................... 145
5.3 Mudflow Parameters...................................... 150-152

1.1 Typical Section, Alluvial Flood Plain....................... 8
1.2 Passive Mitigation Measures.............................. 21-23
1.3 Terracing Repair Method................................... 25
1.4 Rock Buttress, Stabilization Berm........................... 27
1.5 Control of Surface and Subsurface Drainage.................. 29
1.6 Typical Debris Basin........................................ 31
1.7 Typical Debris Fence........................................ 33
1.8 Typical Debris Fence Detail................................. 33
2.1 Classification of Hyperconcentrated Sediment Flow........... 42
2.2 Fluid Models................................................ 46
2.3 Mudflow Wave Effect......................................... 51
2.4 Cross-section Showing Deposition Process in
Jiang-Jia Ravine............................................ 51
2.5 Debris Flow Models........................................ 54
2.6 Process of Stoppage of Forefront of Debris Flow............. 56
2.7 Temporal Variation in the Edges of Debris Flow.............. 56
2.8 Contour Map of Debris Cone.................................. 58
2.9 Zoning of Potential Hazard Area............................. 58
2.11 Position of Debris Flow With Respect to Time................ 61
3.1 Debris Production Rates During 1978 nad 1980................ 78
4.1 Hvorslev Failure Surface.................................... 83
4.2 Hvorslev Failure Envelope................................. 83

4.3 Average Grain Size Distribution Curve From 65 Soil
Samples Taken From Utah's 1983 Mountain Slide Areas........ 86
4.4 Typical Schematic Cross-section of Tip No. 7 at
Aberfan, Wales.............................................. 88
4.5 Range of Slope Angles for the Occurrence of Mudflow.... 90
4.6 Drained and Undrained Loading............................ 93
4.7 Effects of Water on the Formation of Debris Flow....... 96
4.8 Forces Acting on an Earth Slope............................ 98
4.9 Idealized Cases of Seepage in Slope Stability Analysis... 107
4.10 Forces Acting on an Earth Slope........................... 110
4.11 Graphical Solution to Debris Flow Stability Analysis.... 110
4.12 Typical Mudflow Source Area............................113-115
5.1 Wrightwood Upper Source Area.............................. 123
5.2 Longitudinal Section Through a Mud Surge.................. 126
5.3 Grain Size Distributions for Wrightwood Mudflow........... 128
5.4 Plot of Median Grain Size Versus Distance From Source----- 128
5.5 Annual Average Precipitation for Utah's North
Central Zone............................................... 134
5.6 Mudflow Grain Size Distributions.......................... 155
5.7 Viscosity Versus Slope Angle.............................. 158
5.8 Depth of Flow Versus Slope Angle.......................... 161
5.9 Velocity Versus Slope Angle............................... 163
5.10 Depth of Flow Versus Viscosity............................ 164

By January 25th, 1969, an accumulation of between 14 and 18
inches of rain had fallen on the Topanga Canyon region of the Santa
Monica Mountains in Southern California. Topanga Creek was flooding
well above its banks.. At 3 o'clock in the morning, the residents of
the area reported hearing a loud rumbling noise, later found to be
the roar of a flow of viscous mud. The mud impacted the house at 529
North Creek Trail with such force and velocity so as to, literally,
crush the wood frame dwelling, killing three of its occupants, and
sweeping it off to the swelling waters of the nearby creek. Earlier
that evening, close by, in Mandeville Canyon, six firemen were
trapped in a house while attempting to rescue a man who was killed
while in bed, when a mudflow crashed through the back of his house.
Over the course of the storm, from January 18th to 25th, a dozen
fatalities were reported, due directly to injuries sustained from
high velocity impact of mud and debris to residential dwellings.
Hundreds of injuries and millions of dollars in damage were caused by
water laden mud, trees and boulders, sweeping down steep, unstable
hillsides. A month later, torrential rains again converged on
Southern California, causing similar mudflow events. Eight
fatalities were reported in the San Gabriel Mountains alone. Homes

were destroyed and many major highways were inundated by mud and
Until recent years, many natural disasters, such as the
mudflow, have gone relatively unnoticed. Little public attention has
been focused on this subject, and hence, research and funding has
been limited. However, the mudflow is not a new phenomenom.
Geologic exploration of many alluvial fan areas suggests that
mudflows have been occurring throughout geologic history. As man
continues to develop previously undisturbed lands, his presence is
constantly imposing upon areas of delicate natural balance.
For instance, as the availability of residential building sites
continues to diminish in the Los Angeles County area, developers dare
to construct buildings on steep hillsides and within narrow ravines.
Uncontrolled timber harvesting in the Pacific Northwest, combined
with excess seaonal rainfall, leaves the mountainsides susceptable to
erosion, mudslides and mudflow. The under-cutting of the toe of a
steep slope during the construction of a highway or railroad, can
easily upset the stability of the slope enough to initiate a flow or
a slide.
Unfortunetly it often takes a major catastrophe, such as the
debris flows in the Santa Monica Mountains, to bring the problem to
the attention of society, just as modern day earthquake research was
not expanded or accelerated until such occurrences as the 1964
Anchorage, Alaska, the 1971 San Fernando, California, or the 1973
Niigata, Japan earthquakes.
But what can we, as engineers, do to help control this
problem? Perhaps there is no way to completely stop or alleviate the

occurrence of mudflow it is indeed a natural process. We must
first understand the problem before conclusions can be drawn and
solutions recommended. This thesis includes a logical and practical
approach to the study of mud and debris flow. First, definitions are
established to help distinguish the mudflow from other forms of mass
movement and sediment transport. Next, some of the more common
mitigation measures are discussed. These include both active and
passive mitigation methods. Factors which affect the formation of
mudflow are also discussed. The mudflow actually consists of three
distinct processes the formation or source, the actual flow of mud,
and the deposition process. Each of these processes are discussed in
Chapter II; however, the focus of this study is on the source area.
Since this part of the flow relates best to the field of geotechnical
engineering, the fundmentals of soil mechanics are used to disect the
problem, and to suggest theories for the driving forces and the
failure mechanism behind this process. Case histories are then
presented along with an analysis of mudflow-related parameters.
1.2 Definitions and Classification
Many terminologies have been used in the past to define a
mudflow event. On the other hand, the term "mudflow" has been used
interchangeably, to describe anything from liquefaction to landslide.
It is only proper, therfore, that we begin this study with a brief
review of the fundamental definitions and classification of some of

Classification of Mudflow-Related Mass Movements
1. Slope Failure a. Falls Through the air, Large rock or soil Gravitational, little or
free fal1. fragments. no shear displacement.
b. Topples Overturning moment. Large fragments, etc. Often caused by pore
fluids in cracks or adjacent fragments.
c. Slides Gravitational, slip Any type of material. Shear strain
surface. displacement,
- Rotational Concave failure surface.
- Tranlational Planar failure surface.
d. Lateral Spreads Lateral extension, Predominantly fine Shear or tensile
liquefaction. grained. fractures.
2. Floods a. Clearwater Hydraulic (Newtonian) Water N.A.
b. Mudflood High concentration of fine grained silts, etc.
3. Flows a. Dry Resembles viscous fluid. Fine-grained loess, Slip surfaces of differ-
sand, silt. ential movement or shear.
b. Wet Fluidizing effect of Instability due to excess
- Mudflow water. Any type of material. pore water, steep slopes.
- Debris Flows Boulders, trees, ice.

the more common types of mass movements and interrelated forms of
sediment transport.
Table 1.1 summarizes the three basic categories of mass
movement which are often associated with mudflow: 1) slope failure,
2) floods, and 3) flows. Although these three processes are
dependant upon many different factors, four basic criteria may be
used to distinguish between them. These are the type of movement,
the material characteristics, the failure mechanism, and the
transporting mechanism. A brief definition of each of these three
types of movement is presented herein (27) and (33).
1. Slope Failure This process is defined as the downward
and outward movement of slope forming materials such as rock, soils
or fills, either natural or manmade. A slope failure constitutes a
group of slope movements wherein a shear failure occurs along a
specific surface or plane, and may be further subdivided into the
following four categories:
a) Falls: A mass of large fragments (rock, debris, boulders,
etc.) which detach from a steep slope or cliff, and descend rapidly
through the air by free fall, leaping, bounding or rolling. Little
or no shear displacement takes place and the driving forces are
predominantly gravitational.
b) Topples: The forward rotation of a soil or rock fragment
about some pivot point, due to the action of gravity. This form of
movement is often caused by pore fluids in cracks or in adjacent
c) Slides: This form of slope failure consists of shear
strain and displacement along one or several surfaces. This type of

failure may be instantaneous or progressive. It may propagate from
an area of local failure. The displaced mass may slide beyond the
original ground. The most common type of slide is the rotational
slide, where movement takes place along an internal slip surface
which is often spoon shaped. In a translational slide, the mass
progresses down and out along a planar surface with little or no
rotary movement.
d) Lateral Spreads: This from of slope failure involves
lateral extension accompanied by shear or tensile fractures. This
may occur either as an overall extension without a recognized shear
surface or zone of plastic flow, or as fracturing and extension of
coherent material such as the liquefaction or plastic flow of
subjacent material.
2. Floods When one thinks of a flood, what is usually
brought to mind is innundation, either by river or coastal flooding,
caused by excessive precipitation and storm runoff. But the term
"flood" can also apply to a hydraulic process by which sediment is
transported in a natural ravine or manmade channel. For the purposes
of the present study, we will define two types of floods, the dear-
water flood, and the mudflood. Although both share the same type of
transporting mechanism and hydraulic features, they differ with
regard to flow velocity, particle transport, sedimentation, scouring
and deposition characteristics. Both types of flow involve Newtonian
fluids, and hydraulic properties are governed by open channel flow
(such as Manning's equation). The concentration of solids and the
viscosity of flow are often used to differentiate between a clear
water flood and a mudflood.

The term "clearwater flood" generally refers to the flow of
clear water, or water carrying a minimal amount of sediment, usually
fine grained. The velocity is usually much greater for a clearwater
flood than for a mudflood, while the viscosity is nearly that of
water. The flow may be either laminar or turbulent, depending on the
characteristics of the channel. Very little erosion, particle
transport or deposition takes place in this type of flood. Solids
are suspended and transported by the fluid.
In a mudflood, the water carries heavy loads of sediment and
course debris (often as high as 50 percent by volume). Mudfloods
typically occur in drainage channels or on alluvial fan areas located
in mountainous regions. A typical alluvial flood plain is
illustrated in Figure 1.1. The velocity is usually much slower than
that of a clearwater flood, and the viscosity is usually much higher.
Mud and debris are incorporated within the fluid, thus forming a
composite fluid material, with composite fluid properties. Channel
scour is quite common and larger particulates tend to deposit as the
velocity decreases.
' Transformation from a clearwater flood to a mudflood during
the course of a storm event is also quite possible. For instance, as
a clearwater flood progresses, the erosive forces and caving in of
channel sideslopes often add enough sediment and debris to increase
the concentration of solids to the level of a mudflood, or even a
mudflow. The reverse process may also occur. As precipitation
continues to collect, the concentration of a mudflood may become
diluted, while the deposition of sediment takes place, thus reducing
the concentration to that of a clearwater flood.

Figure 1.1 Typical Section, Alluvial Flood Plain
Source:. Delineation of Landslide, Flash Flood
and Debris.Flow Hazards in Utah (Specialty
Conference Papres, Utah State University, 1984).

3. Flows Flows may be defined as movements of earth
materials (with or without water) with a velocity distribution
resembling that of a viscous fluid. They may contain and transport a
variety of materials, self supported by the density and viscosity of
the flowing mass. Flows may occur either in the wet or in the dry,
with a wide range of particle velocities and distributions. Various
terminologies have been used to describe the types of materials
transported by a flow. The term "debris flow" has been used to
describe a flow containing large boulders, trees and assorted
fragments, suspended in the viscous fluid. These types of flows
commonly occur in mountainous regions, where excess snowpack arid
precipitation tend to erode the loose alluvial soils in steep narrow
ravines. The term "debris avalanche" is often used interchangably
with debris flow, especially in mountain or glacier areas.
The type of flow with which this paper is particularly
concerned is the mudflow. Mudflows have attained notoriety in the
past twenty years, due to extensive damages and losses incurred from
recent events throughout the world. The fluidizing effect of water
is generally an intergral part of the process. Once in motion, a
small stream of water heavily laden with soil has a tremendous
transporting power. As more material is added to the stream, from
erosion, channel scour or undercutting of the channel banks, the size
and power of the flow increases. These flows commonly follow
pre-existing drainageways, but may also occur on open hillsides and
slopes. The consistency is often described as that of wet concrete.
The mudflow actually consists of three distinct stages: 1) the
formation or source, 2) the actual flow of mud, and 3) the deposition

process. Each of these components is discussed in further detail in
Chapter III.
1.3 Magnitude of the Mudflow Problem
The mudflow problem is certainly not isolated to the United
States. Probably the most catastrophic of all flows ever to occur
were the debris avalanches of 1962 and 1970 on the slopes of Mt.
Huscarin, in the Andes Mountains, of Peru. In January of 1962, a
glacier high on the north peak of Mt. Huscarin failed and formed an
ice avalanche. The ice avalanche, in turn, gathered and incorporated
a combination of ice, soil, water and rock, thus forming a debris
avalanche. The debris ,.swept through the valley villages below,
killing 4,000 to 5,000 of the area's inhabitants. Eight years later,
in 1970, a similar event occurred, triggered by an earthquake off of
the coast of Peru. This debris avalanche descended into the same
valley at speeds of up to 200 miles per hour, causing more than
18,000 fatalities. The ridge which protected the towns of Ranrahica
and Yuanguay from the previous flow of 1962, was overtopped in 1970.
Both towns along with 15,000 to 17,000 of its inhabitants were buried
and destroyed.
In China, the mudflow is also well documented. Here, the
flows are termed "rock glaciers" because they usually occur in alpine
regions such as the Tibetan Plateau, the Northeast Plateau, and the
mountainous regions of the Szechuan and Hunan Provinces. They are
also common to the southern mountains of Taiwan. Melting snow at

elevations of 6,000 meters (above sea level), combined with excess
precipitation during the rainy season, increase the level of
subterranean water, causing great volumes of ice, rock, water and
soil to flow down drainageways at speeds of up to 10 meters per
second. These flows deposit debris in valley villages, blocking
highways, destroying bridges and buildings, and uprooting forests.
The debris blocks drainageways, bridges and culverts, forming natural
dams and levies, backing up flood waters and inundating farmlands and
villages. The vast deposition of enormous boulders has earned the
title "the sea of rocks" by the natives.
Japan has also suffered extensive damage and loss of life
from mudflow and other'slope movements. Although some of these slope
failures have been triggered by seismic activity, most are a direct
result of the heavy rains which occur during the typhoon season.
It is estimated that there exists more than 60,000 channels or
ravines in Japan with the potential for mudflow occurrence. A total
of 2,674 fatalities were recorded in Japan due to natural disasters
over the period from 1967 to 1976. These include debris flows, slope
failures and floods.. Of the total, 36 percent (954) were directly
attributed to mud and debris flows. Hence, the mudflow has been
widely studied and documented by Japanese scientists.
Mudflows which have occurred in North America have not
resulted in such large losses of life, and have not, therefore, been
as widely publicized. This is due, primarily, to the fact that most
flows have occurred in more remote, non-populated areas. Strict
zoning ordinances have discouraged residential construction in
alluvial fan areas or areas of high geologic risk. However, mudflows

have occurred, on occasion, in some populated areas. These include:
o Virginia, 1969 Probably the worst natural disaster in
Central Virginia's recorded history occurred when flooding and debris
flows resulted from Hurricane Camille. A substantial portion of the
total 150 deaths were attributed to mud and debris flows.
o Saint Jean-Vianney, Canada, 1971 40 homes were destroyed
and 31 persons were carried to their deaths, as a debris flow swept
through this Quebec town.
o Buffalo Creek Dam, Saunders, West Virginia, 1972 Heavy
rains led to the failure of three coal refuse impoundments. The
resulting debris flow, consisting of water, coal wastes and sludge*,
travelled 15 miles downstream, killing 125 people and leaving 4,000
Mud and debris flows are-also quite common in the alpine
regions of Utah, Colorado, Oregon and Washington. Wet Pacific storm
fronts which move over the higher elevations of the Rocky Mountains
produce a snowpack tens of feet thick over the course of the winter
months. Warmer spring temperatures, combined with increased seasonal
precipitation, combine to generate excess pore water pressures within
the, already steep and unstable, soil masses, making them susceptabe
to erosion, slope failure and, eventually, mudflow.
The spring months of 1983 brought widespread damage to the
State of Utah, in the form of landslides, flooding and debris flows.
Literally thousands of mudflows occurred along the Wasatch Front
Range, in the central part of the State, causing more than 250
million dollars in damages. Much of the damage was the result of
flooding of buildings and utilities, deposition of mud into storm

drains and sanitary sewers, and total inundation of farmlands. These
devestating flows were so extensive that 22 of Utah's 28 counties
Were declared national disaster areas. Worst hit were the areas in
and around Salt Lake City, Farmington, Bountiful and Provo.
Nowhere in the United States is the mudflow problem so
prevalent as in Southern California. In heavily populated areas such
as Los Angeles County, buildable land is fast becoming a valuable and
diminishing resource. Rapid population growth forces developers to
construct residential dwellings upon steep hillsides and within
narrow ravines. In this area, debris flows during rainstorms present
a greater risk of death and injury than all other kinds of slope
failures combined.
High intensity rainstorms have occurred in Southern
California with an almost recurring pattern or frequency as far back
as records have been kept (1880). The storms of 1938, 1952, 1958,
1962 1969, 1978 and 1980 have been most severe. Records indicate
that approximately 6 to 12 fatalities occur per each event, with mud
and debris flows being the primary cause. During the period' from
1962 to 1971, 23 people died in the Greater Los Angeles area as a
direct result of being buried or struck by debris flows.
Of recent years, the storms of 1978 and 1980 have been
particularly devastating. Storm related losses, within the city of
Los Angeles alone, amounted to an estimated 100 million dollars in
1978 and'140 million dollars in 1980. In 1978, ten fatalities were
attributed to debris flow. The cities of Monterey Park, Laguna Beach
and Malibu, along the Pacific Coast Highway, and the Santa Monica,
San Gabriel and Santa Anna Mountains were the worst hit.

In mid-Febuary, 1980, a series of six storms swept through
Southern California, bringing nearly 13 inches of rain to Los
Angeles, and more than double that amount to the mountainous regions
outside of the City. The estimated total loss within the six
southern counties affected by landslides, mud and debris flows, to
homes, businesses, agricultural lands and public facilities, is
estimated at 500 million dollars. In the small community of Hidden
Springs, in the San Gabriel Mountains, ten people perished, during
incidents of debris flooding. The effects of forest fire and flash
flood sequence is particularly significant in the formation of
mudflow in this arid region of the country. It is discussed in more
detail in subsequent chapters of this paper.
1.4 Mitigation Measures
Mudflow mitigation measures are generally grouped into two
basic categories passive and active measures. Passive measures may
be further divided into either preventive or protective measures.
Preventive passive measures may be defined as those steps which may
be taken to reduce or eliminate.the potential occurrence of mudflow.
These may include slope improvements, dewatering or revegatation.
Protective passive measures do not attempt to stop or hinder the flow
of mud, but instead try to keep structures from being constructed
within the potential path of mudflow (floodplain). This may be
accomplished by a regulating authority, building or grading codes, or
by government insurance programs. An active mitigation measure is

one which attempts to contain, store, divert or reroute the flow, and
hence the potential energy and destructive force stored within.
Active measures are also known as structural measures. Another
method of protection, which does not actually mitigate mudflow
damages, but can serve to reduce the number of fatalities and
personal injuries, involves the use of an early warning system or
network. This method may take the form of meteorological prediction
or forecasting, geological mapping and survey, or the use of
geotechnical instrumentation devices to monitor the stability of a
particular slope. Each of these mitigation measures is briefly
described below.
1.4.1 Passive Mitigation Measures
History has shown that, demographically, man has almost
always settled in areas close to water. Whether it is a river,
seaport or an oasis, water is a necessary means of survival and
civilized growth. Development and construction within floodplain
areas was therefore, not uncommon, particularly in agricultural
regions. As time went on, growth continued within the flood plain,
with the recognition and understanding that occasional disasters were
inevitable, and could even be considered periodic. Some persons were
willing to take the risk that they could survive or outlive a
catastrophic event, at least for a period of time. However, as
urbanization occurred, the potential for a major disaster in some of

the more densely populated floodplain areas became more real. Indeed
such disasters did eventually occur.
In the United States many social programs have since been
established to assist people, when in need during a legitimate
disaster. Programs such as the National Flood Insurance Program
(NFIP), adopted in 1968, or the Federal Disaster Protection Act of
1973, were established to assist communities during a time of
disaster. Since that time, the government has learned that it is in
its own best interest to discourage construction and development
within areas prone to flooding or geologic hazards, such as the
mudflow. When a community participates in such a program, flood
insurance becomes available, in the event of a disaster, provided
that the community adopts a policy for flood plain regulation. This
type of program emphasizes the passive approach of keeping people
away from floods and flows, rather than the more expensive method of
keeping floods and flows away from the people. The Federal Emergency
Management Agency (FEMA) is charged with the responsability of
administering the National Flood Insurance Program.
FEMA considers events such as floods and flows to be both
predictable and preventable. They will provide financial assistance
to a community, if the community takes the passive floodplain
precautions outlined above. However, the occurrence of a landslide
or a slope failure is considered to be an act of God, and is covered
under such a program. This is why so much time and effort has been
devoted to establishing distinct limits, definitions and
classification of these related forms of mass movement. For
instance, suppose a mudflow were to occur, during a particular storm

event, in a given location. Damages sustained by a participating
community would be covered by the Program. Yet, if a mudslide were
to occur in the same location, under the same circumstances, that
same community would not be covered. The difference between the two
processes is merely the concentration of solids in the fluid, and the
type and of failure and transporting mechanism.
Another passive mitigation method makes use of building codes
and grading ordinances to regulate development within the floodplain.
Building codes and grading ordinances are usually enforced at the
city or county level. They set forth criteria and standards which
must be met by the builder or developer. Restrictions are usually
placed on the steepness of slopes, surface drainage, embankment
materials and compaction. In most locations, at least one
exploratory boring, along with .a geology report, is required at each
building site. A licensed engineer or geologist is usually required
to certify the design, thus accepting the liability for the
structure. Building sites must be inspected and approved. This
protects the structure and its future occupants from the potential
hazards of mudflow.
The City of Los Angeles Department of Building and Safety
provides probably the best example of how to control mudflow damages
through the use of building codes. The City of Los Angeles and the
Los Angeles County Flood Control District keep excellent records
regarding slope failure, flooding and mudflow activity, and the
damages associated with each. As Los Angeles developed into one of
the country's major metropolitan areas during the 1950's and 1960's,
storm related damages brought on stricter grading and building codes.

Prior to 1952 there were no grading codes, no soils engineering, no
exploratory borings, nor soils reports. Grading and construction
were left to the discretion of the builder or developer. After
extensive damages were sustained from the storms of 1952, however,
the first grading codes were enacted. It required that the soils or
civil engineer determine any geologic hazards for each specific
building site. In 1958, the code was further expanded to require
that geologic reports be submitted to the City for each and every
project. The storms of 1962 brought about even stricter building
Table 1.2 (30) serves to demonstrate the effectiveness of the
code in mitigating mudflow damages. A summary of the reduction in
damages to building sites is given as the code was instituted from
1952 to 1969. Note that the predictable failure percentage decreased
from 10.4 to 1.3 to 0.15 over this period. It is estimated that
overall damages associated with various forms of mass movement and
sediment transport decreased from 49 to U million dollars, over the
period from 1952 to 1978.
Another very basic method of mitigating mudflows and floods,
which is very common to all, is simply by spanning the channel or
alluvial fan area with a bridge or structure. Bridges are commonly
used to separate a highway or railroad from an area of conflict, such
as a floodplain. Note that a culvert is not, by definition, a
passive mitigation measure, since a culvert is used to divert or
channelize a flow underneath an embankment. Remember that a passive
mitigation measure is one which attempts to keep structures out of
the path of the flow of mud, and not vice-versa. The number of
bridge piers should be kept to a minimum, to avoid clogging of mud

TABLE 1.2 Damage to Building Sites Verses Revised Building Codes
Pre 1952 1952 1963 1963 1969
No grading code, no soils Semiadequate grading code, very New modern grading code, soils
engineering, no engineering limited geology, but no status engineering and engineering
geology. and no responsability. geology required during design, soils engineering and engineering geology required during construction, design engineer, soils engineer, and engineering geologist all share legal responsability.
Approximately 10,000 Approximately 27,000 Approximately 11,000
sites constructed. sites constructed. sites constructed.
Approximately $3,300,000 Approximately $2,767,000 Approximately $182,400a
damages. damages. damages.
Approximately 1,040 Approximately 350 Approximately 17
sites damaged. sites damaged. sites damaged.
An average of $330 per An average of $100 per An average of $7 per
site for the total site for the total site for the total
number produced: number produced: number produced:
$3,300,000 $2,767,000 $80,000
10,000 sites 27,000 sites 11,000 sites
Predictable failure Predictable failure Predictable failure
percentage: 10.4% percentage: 1.3% percentage: 0.15%
1,040 damaged 350 damaged 17 damaged
10,000 total sites 37,000 total sites 11,000 total sites
a0ver $100,000 of the $182,000 was incurred on projects where grading was in operation and no residences
were involved; thus less than $80,000 occurred on sites constructed since 1963.
Source: James E. Slosson and James P. Krohn, Southern California Landslides of 1978 and 1980 (Proc. Storms,
Floods and Debris Flows in Southern California and Arizona in 1978 and 1980, California Institute of
Technology, 1980) p. 297

and debris in the channel. Foundations should be designed to
withstand the scouring effect of a highly concentrated flow. Land
bridges are sometimes used to span an unstable area, where removal of
slope, material is not possible. They do not, necessarily, span an
actual mudflow channel or ravine. Land bridges are supported by deep
seated foundations and are designed to withstand the lateral forces
exerted by flow or slide. Figures 1.2a 1.2e depict some of the key
elements of protective passive mitigation.
Preventive passive mitigation measures may be described as
those measures which attempt to prevent the occurrence of mudflow, by
improving or stabilizing the mudflow source area. The most common
method involves revegetation of the slope by planting trees, shrubs
and native grasses. This helps to create a root network which can
reinforce the loose soil much in the same way that reinforcing steel
is used in concrete to resist tensile stresses. In arid regions,
roots extend deep into the soil in search of water, especially during
the dry season. In natural streambeds or drainageways, however,
water is often readily available at or near the surface, at least a
greater percentage of the year. Here the roots extend only a shallow
distance into the soil, and provide only minimal protection. To be
effective, vegetation should meet three criteria: It should grow
fast, be relatively fire retardant, and should be deep rooted.
Grasses and surface growth also help to protect the slope by
providing a natural cover to reduce the amount of infiltration of
storm runoff. Although the root network is shallow, it is well
integrated, and helps to reduce the effects of erosion. Terracing of
slopes is a method often used in China to stabilize a steep, unstable

Figure 1.2 Passive Mitigation Measures
Source: Delineation of Landslide, Flash Flood and
Debris Flow Hazards in Utah (Specialty Conference
Papers, Utah State University, 1984).

* '-r;
r :. ; /.
Figure 1.2 Passive Mitigation Measures


slope. The U.S. Forest Service and Soil Conservation Service has
conducted extensive research and field experimentation oh the
stabilizing effects of vegetation to control erosion and mudflow.
Probably the most effective preventive measure involves some
form of regrading of the slope. If the slope angle can be flattened
(oftentimes by just a few degrees), the stability of the slope can be
greatly improved. This method usually requires a substantial amount
of earthwork, and a great deal of additional working room at the toe
of the slope. As is most the case, right of way is limited at the
toe and will not allow sufficient room for the slope improvements.
In many instances, construction of a structure, highway, railroadr
etc. actually requires that the slope be undercut or steepened. In
these cases, a retaining type structure is most often employed. These
structures may include gravity walls, cantilevered walls, tie-back
walls, crib or bin walls, sheet piles, gabions, geotextile fabric
walls, or reinforced earth. They must be designed to withstand very
large lateral earth and surcharge pressures, and work best if
properly drained. The main drawback to using a retaining wall to
stabilize an earth slope is that the slope must be undercut
substantially in order to construct the footing or foundation of the
wall. This excavation can lead to instability or even failure of the
slope, thus defeating the purpose of the wall altogether. Figure 1.3
shows a typical retaining/terracing mitigation concept.
In some instances, earth buttresses, berms or shear keys are
an inexpensive method of stabilizing a slope. Earth buttresses are
composed of rock or earthfill, placed at the toe of the slope to

Figure 1.3 Terracing Repair Method
Height of Slope Varies
Masonry Block Hal 1
(typi ca 1)
Original Slope
(varies in steepness)
(8" x 8" x 12 blocks
#4 steel reinforcement
'32" on center, grated centers)
Block Wall
Granular Backfi
Drainage Line
Excavate for Footing
Block Hall
Drainage to
Storm Sewer
Source: Allen E. Chin* Landslide and Debris Flow Mitigation Measures
for Existing Developments (Specialty Conference Papers, Utah State
Uni vers i ty, 1984), p.16.

provide additional weight and resisting force. Shear keys are
essentially prisms of compacted fill, placed to support isolated
areas of an unstable slope. Both of these methods have proven
effective, but again are often subject to right-of-way restrictions.
Along Interstate 7Q in Vail, Colorado, heavy rock buttresses have
been used to control areas prone to mudflow and slope failure.
Figures 1.4a 1.4b demonstrate how a rock buttress or berm may be
used to stabilize a slope.
Hillside benches are used to alleviate the shear stress on
long, steep slopes, particularly when the toe of the slope is
disturbed. The City of Los Angeles Department of Building and Safety
requires that all slopes and excavations be constructed with benches.
In some cases, a very important project may warrant that the
unstable slope material be removed or excavated. Depending upon the
situation, this could mean the unloading of just the head of the
slope or could possibly mean the removal of the entire slope,
altogether. Unloading of the head of the slope can sometimes prove
to be one of the least expensive yet most efficient remedial
measures. However, access to the head of an unstable slope is
usually extremely difficult. Material removed from the head of the
slope can often be used at the toe for added stability. Removal of
an entire slope can be a very expensive proposition, and is only used
as a last resort, and only if the importance of the project proves it
to be cost effective.
Water plays a significant role in the stability of an earth
slope. Uncontrolled surface runoff is the primary cause of erosion
and undercutting of slopes. Surface runoff can be controlled by

b) Stabilization Berm Used to Correct Landslide
Figure 1.4
Source: Transportation Research Board, Landslides Analysis and
Control, (National Academy of Sciences, Washington, D.C.,' Special
Report 176, 1978), p.184.

diversion, regrading, or by linings such as riprap or geotextile
fabrics. A product known as "grass-crete" or "grass-pavers" uses a
combination of patterned precast concrete block and natural grass to
serve as a lining or stabilizing material. Tension cracks, which
sometimes develop within a cohesive soil, can be sealed or grouted to
prevent the penetration of surface water. Relatively impermeable
materials such as clay, concrete or asphalt are used, on occasion, to
pave the surface of a slope, thus preventing the infiltration of
Techniques have also been developed to aid in the control of
subsurface water. Horizontal drains and tiles are an effective means-
of lowering the water table, provided that it is relatively close to
the ground surface. Longitudinal drains may take the form of
perforated pipe, or of a combination of a stone filter and a
geotextile fabric filter. Vertical sand drains are also used for the
purposes of dewatering and dissipation of excess pore water
pressures. Figures 1.5a 1.5b show methods of surface and
subsurface water control.
1.4.2 Active Mitigation Measures
Active, or structural mitigation measures are those measures
which store, divert or route the flow of mud away from residences,
structures or areas of development. They are usually the only
feasible means of protecting an existing structure from the path of a

a) Surface Drainage of Slope By Diversion
Ditch and Interceptor Drain.
b) Horizontal and Vertical Drains to Lower
Groundwater in Natural Slopes.
Figure 1.5 Control of Surface and Subsurface Drainage.
Source: Transportation Research Board, Landslides Analysis
and Control, (National Academy of Sciences, Washington, D.C.
Special Report 176, 1978), p.T76.

potential mudflow (other than moving the structure altogether).
Active measures may take many different forms.
One method, which has been used with some success in
California and in China, is that of the debris or sedimentation
basin. During a given storm, within a particular watershed, the
potential total volume of solids to be deposited can be estimated by
knowing or being able to predict a) the quantity of storm runoff, and
b) the average concentration of solids in the flow. By routing or
directing the flow through a debris basin (or series of basins), the
mud will deposit in a Controlled manner. To make optimal usage of
the storage volume of the basin, spillways and outlet structures are-
designed to release the storm water while retaining only the solids.
Proper sizing of the basin involves some knowledge of the
settlement/deposition characteristics of the solids, much the same as
in the design of settlement basins for a water treatment facility.
Steel rails are often used at the downstream face of the basin to
catch large inclusions such as trees and boulders while allowing the
water and mud to escape (see Figure 1.6). Although the debris basin
is a reliable method of collecting mud and debris, maintenance can
often be a problem and the cost can be prohibitive. It is not
uncommon for mudflows to yield hundreds of thousands of cubic yards
of deposited solids in a single event.
Diversion channels are used to collect and to transport
water-laden mud around or away from residential or populated areas.
These channels are usually trapazoidal in shape, and constructed of
concrete. They must be designed with sufficient capacity to avoid
the potential clogging of the channel with mud and debris, and with

Figure 1.6 Typical Debris Basin
i :
Structural Defense for Arresting and Separating Debris Flows
(Diagrammatic Sketch)
Fence 6-8 ft (1.8-2.A m) high with vertical bars 18 in. (50 cm) apart
Source: Allen E. Chin, Landslide, and Debris Flow Hazard Mitigation Measures
Measures for Existing Development. (Specialty Conference Papers. Utah State
University, 1984), p.18.~ .

enough freeboard to avoid overtopping the channel banks. The channel
lining must also be designed to withstand the tremendous scouring
effects-of the flowing mud.
Deflecting type barricades and walls have been used, with
some success, to protect structures from the impact of mudflow. By
placing a deflection wall uphill (upstream) of a structure, at an
angle of less than 90 degrees to the channel, the flow tends to
divert to one side of the structure. This also has the effect of
reducing the lateral component of force of the mud against the wall.
Walls which are place perpendicular to the channel, for the express
purpose of dissipating the flow "head-on", are not recommended. To
overcome the force of the flow, these devices must be quite large and
heavily reinforced. A typical deflection wall is shown in Figure
1.7. Oftentimes the barrier is designed too low, and is totally
overcome and inundated by the mud. Large earthen levies have also
been used to deflect and divert mudflow away from highways and
A debris fence (Figure 1.8) may be used to retard the flow of
debris down a slope, while at the same time catching some of the
larger fragments. It also serves to break up the flowing mass. Some
researchers claim that this allows entrapped air to escape from the
flow. It is speculated that entrapped air reduces friction and forms
a cushion within the flow, upon which the mud gains such terrific
velocity and momentum. Proper placement of the flow fence is
necessary if it is to be effective. If placed too near the toe of
the slope or confluence of drainage channels, the fence may be
overcome by the maximum concentrated force of the flow.

= iuoaot rnnr.F of DEBRIS FLOW
F = F COS 0
a) Force Diagram for
Deflection Wall.
b) Typical Deflection
Wall, "A"-type.
Figure 1.7
Figure 1.8 Typical Debris Fence Detail.
Source: Robert A. Hollingsworth and G.S. Kovacs, California Soil
SIuriips and Debris FIOws, (Specialty Conference Pap?rs, Utah State
University, 1984), p.Tl and p.13.

Energy dissipaters and check dams have also been used to
reduce the gravitational energy stored within the flow, and to
protect the streambed from undercutting, scour and erosion.
1.4.3 Early Warning System
During the early morning hours of January 25, 1969, a total
of twelve fatalities were caused by eight major debris flows in the
Santa Monica Mountains of Southern California. All of the victims
were inside residential structures that were destroyed by the impact
or flooding of debris. Eight of the twelve victims were in their
bedrooms, where they were crushed beneath collapsing walls or buried
by muddy debris. In one instance, the mother of two children had
reportedly awakened and was on her way to their bedroom when the
house was leveled and carried away to a flooding stream. Another
adult was trapped and killed when returning to his house after,
evacuating his family, when it was struck by a high velocity flow.
It has been observed (by those fortunate enough to escape
injury when their homes were damaged or destroyed) that if an adult
of the household was alert to the problems outside of the house,
almost invariably, the approaching hazard was recognized in time to
evacuate its occupants. The value of advance warning, therefore,
seems clear.
The purpose of an early warning system is to predict or to
provide a means of warning in the event of a sudden change that could
be indicative of an impending earth movement. Advance warning may

take on a variety of forms. For instance, meteorological forecasting
has progressed to the point where life saving predictions can often
be made using relatively standard data. The failure of a slope is
most often caused by an increase in pore water pressures, due
primarily, to excess precipitation. Slope failure and mudflow are
often the result of a period of sustained precipitation, followed by
a high intensity outburst. If accurate meteorological records have
been kept for a region, a correlation could be developed between
rainfall accumulation and intensity, and the occurrence of mudflow.
Failure of a particular slope and mudflow occurrence could be
forecast based upon a range of precipitation amounts. Campbell (5)
used rain gauge data, along with conventional slope stability
analysis, to determine the amount or combinations of precipitation
intensities at which mudflow coaid be expected to form. A knowledge
of the soil runoff and infiltration characteristics is necessary in
this type of analysis.
Mapping is another method which is valuable in the
delineation, of mudflow hazard zones, and may be helpful in the
development of an early warning system. Maps are available which
depict topography, geology, soils and other special terrain and
cultural features. Many mudflow hazard zones have already been
mapped by previous researchers. The United States Geologic Survey
(USGS) publishes topographic surveys, usually at 1:20,000 scale, for
most of the continental United States. The U.S. Department of
Agriculture, Soil Conservation Service has published soil surveys for
most of the country since 1899. These surveys depict the regional
soil types and properties, and describe their suitability for

agriculture and engineering. Aerial photography and satellite
imagery are particularly effective in the before and after
comparisons of slope failure or mudflow locations.
Field instrumentation can also be used to identify and
monitor potential slope failures. Surface surveying of a slope using
conventional methods such as monument survey or horizontal movement
extenseometers is the most common method of detecting movement
trends. The development of the inclinometer has been the most
important contribution to the analysis and detection of landslide
movement in the last 20 years. Pore pressure and ground water
measurements such as observation wells and piezometers are often used -
to monitor hillside slopes. These devices are not well suited to
mudflow, however, which often occurs suddenly and spontaneously.
These devices require constant monitoring to be effective, and are
better suited for special cases. They may be used to monitor a large
hillside which threatens a heavily populated community below, or
where a construction project is taking place adjacent to a steep
slope. It would be quite an expensive proposition to install and to
monitor field instrumentation devices for every area of potential
slope movement.
Another problem associated with an instrumented warning
system is the determination of allowable tolerances. At what point
would the instruments be programmed to set off the warning? Should
each instrument be monitored individually or should they be monitored
as a group being dependant upon one another? No computer or
advance warning system could ever take the place of engineering

judgement in the evaluation of data from a field instrumentation
1.5 Purpose and Scope of Study
The purpose of this study is to develop a better
understanding of the mudflow phenomenom, in order that a logical and
organized approach to mudflow research may continue in the future.
The purpose of this present study is not to develop precise solutions
or equations, but rather to assimilate the data and observations of
previous researchers, in an attempt to form some correlations and
range of values between the many mudflow-related parameters, and to
distinguish mudflow from other forms of mass movement and sediment
Although all three phases of mudflow (source, flow, and
deposition) are discussed herein, the focus of this paper is on the
phase which relates best to the geotechnical field the source, or
formation of mudflow. The following chapter discusses the three
phases of mudflow. Subsequent chapters cover the factors which
affect the formation of mudflow, and then the formation of mudflow in
further detail. The study concludes with some case histories and a
review of some of the variables associated with mudflow, and some of
the relationships between them.

2.1 General
The mudflow is a very complex form of earth movement. At the
source area it behaves as a relatively rigid slab of earth, but
quickly changes to a viscous fluid-like material upon failure. It is
not well understood exactly why or when this transformation takes
place. The conditions of resistance to downslope movement change
from sliding friction to viscous stresses, and different types of
analyses must be used to evaluate each of the three distinct phases
of the mudlfow process.
The first stage may be defined as the source, or formation of
mudflow, and is governed predominantly by the laws of slope
stability. It is the aspect of mudflow which relates best to the
geotechnical field, and the one with which this paper is primarily
concerned. When a steep unstable slope, comprised of loose weak
soils with little or no ground cover, is subjected to intense
precipitation, the saturation of the soil skeleton and the
corresponding reduction in effective stress cause the soil to slide
or fail. This failure takes place because the reduced shear strength
of the soil is overcome by the weight of the water-soaked embankment,

which increases the driving force necessary to induce a shear
However, the failure mechanism associated with mudflow
formation cannot be treated as the classical slope stability problem.
The mudflow is not subject to static conditions or defined boundary
conditions, as such, but is constantly undergoing changes in
geometry, pore pressure distribution, and shear strength (i.e.
remolded or residual values). The seepage flow of groundwater along
the bedding plane causes a wide variation of pressure gradients. The
slope geometry is constantly undergoing changes as mud, which has
already failed upslope, erodes and undercuts the toe of the subjacent
slope material as it passes through.
Once the slope has failed, and the transformation from soil
solid to viscous fluid has occurred, the second stage, the actual
flow of mud, takes place. Depending on the concentration of solids
within the flowing mass, it may behave as a visco-plastic
(non-Newtonian) fluid, or as an ideal fluid, similar to clear water,
and is governed by the concepts of open channel hydraulics. Mudflows
possess a combination of density and strength that will support and
transport tremendous volumes and enormous inclusions. The ability to
support large objects such as trees and boulders, stems from the
matrix viscosity (a velocity-dependent strength), and the shearing
resistance of the mass (a velocity-independant strength). The
fluidizing effect of water (and sometimes air) allows the flow to
reach avalanche speeds. The erosive and scouring effects of the
flowing mud may remove tons of material from the channel bottom, thus
adding even more potential energy to the viscous flow. This stage of

mudflow does not lend well to geotechnical analysis, but is best
studied using the fundamentals of fluid mechanics and open channel
hydraulics. It may be modeled hydraulically, as a one dimensional,
non-steady state, composite flow of soil and water.
The third and final stage of mudflow involves the process of
deposition. As the soil-laden floodwaters reach flatter slope
gradients, the matrix viscosity tends to decrease and loose energy,
allowing the solid constituents to settle out. The deposition
process may be modeled as a two dimensional, non-steady state,
composite flow of soil and water. As the concentration of solids
decreases, the flow separates back to its original composition of mud
and clear water flooding. The post deposition process may be modeled
as two dimensional, non-steady state, flow of water. These alluvial
debris deposits may become hundreds of feet thick over the years.
Oftentimes, the soil profile may resemble that of a glaciated till.
It is essential that a method for delineating the limits of debris
fan flooding and deposition be developed, in order that passive
measures of mitigating mudlfow damage (i.e. discouraging development
within the floodplain) be instituted.
As previously stated, this paper is primarily concerned with
the first stage of mudlfow the source or formation. The following
chapter is devoted to a more detailed study of this portion of the
problem. Included herein is a brief summary of the second and third
stages the flow and depostion processes, in an effort to present a
complete overview of the mudflow problem. Also included is a brief
summary of some of the more common mudflow models. To understand
fully the formation stage, it is helpful first to review some of the

fundemental hydraulic features, and to see how the three stages of
mudflow interact with one another.
2.2 Stage II The Flow
The second stage of mudflow is best studied using the
principles of fluid mechanics or open channel hydraulics. In these
fields, mudflows or debris flows are sometimes referred to as
"hyperconcentrated sediment flows". This term refers to a broad
spectrum of earth movements ranging from slope failure to mudflood.
The definitions set forth in Chapter I should be used to distinguish
between these various types of movement.
The concentration of solids (either by weight or by volume)
is often used to differentiate between landslides, mudflows and
mudfloods. O'Brien and Julian (26) suggest the values shown in Table
2.1 and Figure 2.1 to distinguish between four different types of
earth movement and sediment transport. The sediment concentration
serves as a good indicator of the viscosity, turbulence, dispersive
and yield stresses associated with each type of movement. Landslides
contain the highest concentration of solids, but do not exhibit any
fluid-like properties and should not be analyzed as such. Mudfloods
and Clearwater floods are unable to resist shear stresses without
undergoing some form of motion, nor do they exhibit any appreciable
yield strength. These types of flow behave, instead, as Newtonian
fluids, and are best modeled using the conventional methods of open
channel flow (i.e. Manning's equation).

Figure 2.1 Classification of hyperconcentroted
Sediment Flow
Source: J.S. O'Brien and P.Y. Julian, Physical Properties and
Mechanics of Hyperconcentrated Sediment Flow, (Specialty
Conference Papers, Utah State University, 1984).

Table 2.1 Hyperconcentrated Sediment Flow as a Function.
of Concentration
Source: J.S. O'Brien and P.Y. Julien, Physical Properties and
Mechanics of Hyperconcentrated Sediment Flow/ (Specialty-Conference
Papers, Utah State University, 1984).
Concentration Concentration
by Volume by Weight
Flow Type C V C w Flow Characteristics
.53-.90 75-.96 Will not flow, failure by block sliding
Landslides .50-.53 .73-.75 Block sliding failure with internal deformation during the slide, slow creep prior to failure
.48-.50 .72-.73 Flow evident, slow creep sustained mud flow, plastic deformation under
Hud Flows its own weight, cohesive, will not spread on level surface
l/l 1 ** 00 .69-.72 Begins spreading, cohesive
.40-.45 .65-.69 Mixes easily, shows fluid properties in deformation; spreads on horizontal surface but maintains
- a inclined fluid surface, large particle settling, waves appear but dissipate rapidly
.35-.40 .59-.65 Marked settling, spreading nearly complete on horizontal surface,
Mud Flood. liquid surface two phases appear, waves travel substantial distance
.30-.35 .54-.59 Separation of water on surface, two phases, waves travel easily, most sand and gravel has settled out
t .20-.30 .41-.54 Distinct wave action, fluid surface all particles resting on bottom in quiescent fluid condition
Water <.20 .41 Water flood with bed and suspended
Flood loads
^This information is qualitative guideline in which the concentration
refers to the fluid matrix consisting of silts, clays and fine sands.
The concentration by weight is computed using 2.72 as the specific
gravity for the sediment as measured in the laboratory.

In mudflows, the sediment concentration is sufficient to
maintain a density and strength which will support and transport
large inclusions, in a quiescent condition without settling. These
inclusions tend to stay near the surface of the flow, due to the
fluid matrix density and small settling velocities (in the absence of
turbulence). Debris flows are distinguished from mudflows when more
than 50 percent of the sediment is coarser than sand. The fluid
matrix exhibits a resistance to motion or high yield strength
corresponding to high viscosities.
Hyperconcentrated sediment flows are governed by, among other
factors, complex processes of energy dissipation, including viscous
and turbulent stresses, the interaction of water and sediment, the
exchange of sediment particles with the channel boundaries, and the
dispersive stress (collisions between suspended particles). Cohesive
strength associated with clay minerals, also affects the fluid
From elementary fluid mechanics, we recall that the force
within a fluid that is required to cause a given deformation is
proportional to the rate at which the body deforms. The viscosity of
a fluid is a measure of the fluid's resistance to shear or angular
deformation, expressed by the relationship,

where: T = shearing stress
jjl = coefficient of viscosity
= velocity gradient, rate of deformation
The distinction between a solid and a fluid lies in the
manner in which each can resist shearing stresses. An ideal fluid
may be defined as one in which there is no friction (i.e. its
viscosity equals zero). A fluid for which the constant of
proportionality (viscosity) does not change with the rate of
deformation, is said to be a Newtonian fluid. At the other end of
the spectrum is the ideal elastic material which is not dependant
upon the rate of deformation (in a fluid mechanics sense), and is
controlled by another type of proportionality constant the elastic
modulus. Figure 2.2 demonstrates the range of fluid/solid types of
movement with which we are presently concerned. The mudflow exhibits
neither ideal fluid nor ideal elastic properties, and lies somewhere
in between these two extremes. The Bingham or visco-plastic model
bes't simulates the flow of mud. The Bingham fluid model combines an
initial yield stress, K, and a linear stress-strain relationship to
relate the fluid's rate of deformation to the shear stress.
In mudflows, the large concentrations of fine sediment (which
along with the fluid and water are commonly referred to as the fluid
matrix) tend to alter such properties as the viscosity, density and
turbulence from that of a Clearwater flow. This has the effect of
increasing the lift, drag and buoyancy forces, giving the fluid the
ability to support such large objects. Debris flow viscosities in

Shear Strength
Figure 2.2 Fluid Models.

Table 2.2 Typical Fluid Viscosities
Newtonian Fluid
Viscosity (poises)
Water, 20C. 0.01
Machine Oil, light, 15.6C. 1.14
Machine Oil, Heavy, 15.6C. 6.61
Glycerine, 14.3C. 13.87
Glycerine, 2.8C. 42.20
Asphalt, 47C. 10.4x10
Pitch, 15C. 13.0 x 10
Non-Newtonian Fluid Viscosity (poises)
Experimental Kaolin Slurry 0.30
Oil Well Drilling Muds O.Ox to 3.0
Ketchup 0.83
Mustard 2.94
Mayonnaise 6.33
Wet Cement Paste 24
Wet Cement Mortar 34
Natural Mudflows (Wrightwood):
Muddy Water between surges 1.0
Mudflows (range of 45 measurements) 400 to 1,000
Basaltic Lava (25 % Crystal Slurry) 6,500
Source: Russell H. Campbell, Soil Slips, Debris Flows, and
Rainstorms in the Santa Monica Mountains and Vicinity,
Southern California, (Washington, D.C., Geological Survey
Professional Paper 851), p.26.

excess of 1,000 poises have been measured in the laboratory. The
viscosities of some common fluids are Table 2.2.
This relationship is valid when the shear stress, T, is
greater than the yield stress and is linearly proportional to the
deformation. The Bingham model can be used to describe the motion of
mudflow in a smooth prismatic channel for partially turbulent or
translational flows without energy loss due to boundary roughness.
The term, C (C*U/(jy)^> is a composite of the dispersive and the
turbulent stresses as defined below,
The general form of the Bingham fluid equation is,
where: T = shear stress
K = yield stress
u = viscosity of the fluid matrix
C = variable, function of depth and
j^r = deformation in the y-direction
Turbulent stress: T^ = p K Y (gy)
2 u2 ,dUv2
Dispersive stress: Td = a^ X2 Dg (gy;)2

where: p = density of the fluid (water)
K = Von Karman constant
Dg = representative grain diameter
= constant
X = linear concentration = f (C0> Cy)
C = concentration by volume
CQ = maximum static concentration by volume
The turbulent stresses assist in suspending particles into the flow
by exchanging momentum from the fluid to the sediment particles. The
dispersive stresses impart momentum transfer between the particles.
Increasing particle concentration has the effect of dampening the
turbulence. Energy is absorbed from the body of the flow and
expanded by increasing the sediment particle velocity and the height
of suspension'.
It has been reported by most observers that mud tends to flow
in bursts or waves. These waves may come as frequently as every few
minutes or as far apart as several hours. Sharp and Nobles (28)
noted that the frequency of bursts at Wrightwood, California, was
proportional to the time of day and the corresponding temperature.
As the warmer mid-afternoon temperatures caused the dense snowpack to
melt, the time interval between bursts decreased, but as the cooler
evening temperatures arrived, the mud gradually became frozen and the
frequency of bursts decreased, eventually stopping all together.
Jian, Jianmo, Chen and Defu (23) depict the bursting effects
of mudflows in the Jiang-Jia Ravine in Southwest China as shown in

Figure 2.3. The three main features of the flow are described as the
head, the body and the tail. The wave effect of the mudflow is
caused by the depth of the flowing mass trying to overcome the yield
or the shear strength of the fluid. In order to maintain the flow,
the shear stress due to gravity must be greater than the fluid's
resistance to shear. As the first wave of mud moves down the rough
dry channel, it tends to adhere to the channel surface due to
increased shearing resistance, and is brought to rest, somewhat
prematurely. As successive waves move down the channel, the effects
of the boundary roughness have been partially negated, due. to the
"paving" action of the preceeding wave. Successive waves have a
tendency, therefore, to travel further down the channel, and with
higher velocities. The velocity of the flow has been approximated
empirically by the following relation,
V = 13.5 ($)0,062 (^r)0,025 (ghi)0,5 (2.5)
where: V = velocity of the mudflow surge ( /sec)
d = median grain diameter (cm)
h = depth of the surge (cm)
C = concentration of viscous grains
smaller than 0.005 mm. (%)
Y = density of the mudflow (^m/cm)
g = acceleration due to gravity ( /sec2)
i = inclination of the channel bed

a) Longitudinal view
Flih| ft it I*
|aV| }
b) Plan view

... x
u. Ji? 7_T

Figure 2.3 Mudflow Wave Effect
Figure 2.4 Cross-Section Showing Deposition Process
in Jiang-Jia Ravine
Source: Li Jian, Yuan Jianmo, Bi Cheng and Luo Defu, The Main Features
of the Mudflow in Jiang-Jia Ravine, (Berlin, Zeitschrift fur \
Geomorphologic, N.F. Bd. 27, Heft 3, 1983), p.333 and 377.

2.3 Stage III Deposition
One of the more difficult problems associated with the third
stage, the deposition of mudflow, is the delineation of the limits of
the hazardous zone. In order to institute an effective passive
measure for mitigating the damage due to mudflow, such as the
National Flood Insurance Program, the limits of the hazardous
deposition area must first be defined. This area of deposition,
commonly referred to as the debris fan or cone, is an area where
certain types of development or construction should be discouraged.
In evaluating the limits of debris fan flooding and mudflow
runout distance, the following factors should be considered: 1) the
volume of mud or debris material in the flow, 2) the gradient of the
flow path, 3) the geometry width, depth, sides!opes of the
channel and of the fan area itself, and 4) the composition and the
concentration of the flowing mass.
L Jian, Jianmo, Chen and Defu (23) have observed the erosion
and depostion characteristics of the mudflows in the Jiang-Jia
Ravine. Survey cross-sections taken at a reference station near the
mouth of the ravine over a period of twenty years (from November,
1966 to June, 1977), reveal the massive deposition process which has
taken place. The annual output of debris in this particular channel
is on the order of 4 to 5 million cubic meters per year, including
boulders with a diameter of up to 7 meters. It is not-uncommon for a
layer of mud, 5 to 6 meters deep, to be deposited in the lower
portion of the basin during the course of a single mudflow event.

This ravine undergoes one of the most dramatic geomorphic processes
of deposition in the world, thus emphasizing the role which mudflow
has played in modeling the earth's surface.
2.4 Mudflow Models
Bagnold (2), and later Takahashi (31), investigated the flow
of granular material in debris flows. A uniform layer of loose
granular material (with negligible cohesive strength) is modeled as
shown in Figure 2.5. The layer has a depth, d, and makes an angle,
0, with the bedding plane. As the layer first becomes saturated by
the parallel seepage flow of depth, hQ, the applied tangential
stress, "T, increases. Even though there is no increase in pore water
pressure at this point, movement of the upper surface of the bedding
material,, a^, will commence as T becomes larger than the internal
resisting stress, T^. Increased pore spaces in the layer will result
in greater fluidity and a more rapid flow. The shear stress, T, and
the resisting stress, TL, may be defined by the relations,
Shear stress, T = g sin.8 [C*(a-p) a^ + p(a^+hQ)] (2.6)
Resisting Stress, TL = g cos 0 [C*(a-p) aL] tan 6

b) -Model of Ouui-StMdy Propagation of Debrie Flow
Figure 2.5 Debris Flow Models from Bagno!d (1956) and
Takahashi (1978), a) non-Stationary Bed
b) Stationary Bed
Source: T. Takahashi, Mechanical Characteristics of. Debris
Flow, CProc. A.S.C.E. Journal of the Hydraulics Division,
7of7 4, No. HY8, August, 1978), pp. 1160,1162.

where: o = density of the granular material
p = density of the fluid
C* = grain concentration by volume
= the angle of internal friction
By equating these two equations and assuming a factor of safety of
1.0, we find the critical slope angle for the occurrence of debris
flow to be,
tan 0 = C* ^g tan <(> (2.8)
C* (a-p)+. (l+h0/(J)
A flow which occurs on a slope flatter than 0 is said to have a
temporal stationary bed, while a flow which occurs on a very steep
slope (greater than 0) is said to have a non-stationary bed.
Takahashi (32) has also conducted extensive research
regarding the delineation of hazardous zones of debris flow runout.
The profile of a debris flow deposition area is modeled as shown in
Figure 2.5. Here the conservation of momentum is equated between
section I and section II, at time, t, and time t + A t. A
relationship is then developed for the runout distance as a function
of the viscosity, change in bedding angle, concentration of solids,
and the densities of the fluid and of the sediment.
As the mud travels down the channel, it flows over previously
deposited material, thus adding to the height and surface area of the
debris fan. As the height increases, it eventually backs up, causing
the deposition process to occur in the channel itself. The formation

Figure 2.6 Process of Stoppage of Forefront
of Debris Flow.
Figure 2.7 Temporal Variation in the Edges of
Debris Flow.
Source: T. Takahashi, Estimation of Potential Debris Flows
and Thei r Hazardous Zones; Soft Countermeasures for a Pi saster,
(Journal of Natural Disaster Science, Vol. 3, No. 1, 1981).

of the debris cone occurs when the width and the slope widen and
flatten out at the downstream end.
Figure 2.7 shows the results of many experiments in which the
temporal variations of debris flow runout are modeled. The number on
each curve indicates the time, in seconds, which have elapsed since
the arrival of the forefront of the debris flow wave, at the mouth of
the channel. Figure 2.8 shows the arrangement of the finished
contours after the depostion of the debris cone (shown in Figure 2.7
It has been observed that both the thickness of the deposit and the
mean grain diameter of the sediment both decrease as the distance
outward from the channel mouth increases. Using the mudflow front
discharge, the particle size distribution and the total sediment
yield, etc., one can delineate the hazardous zone of debris flow
runout by approximating an effective diameter, x^, as shown in
Figure 2.9.
Deleon and Jeppson (14) proposed a theoretical model based on
one dimensional Saint-Venant equations of continuity and motion,
together with the Chezy equation for defining friction slope. This
was found to be suitable for describing open channel debris flow in
the laminar range with Reynolds number equal to 500 or less, for
steady state flows. The advantage of this model over other models is
that it does not depend upon coefficients that are difficult to
obtain or vary with flow conditions.
A computer program, based on this model, was developed at the
Utah Water Research Laboratory. Based upon steady state, but
non-uniform flow throughout the channel, the program first generates
depths, velocities, flow rates, areas, etc. of the debris flow at

Figure 2.8 Contour Map of a Debris Cone
Figure 2.9 Zoning of a Potential Hazard Area
Source: T. Takahashi, Estimation of Potential Debris Flows and Their
Hazardous Zones; Soft Countermeasures for a Disaster, (Journal of
Natural Disaster Science, Vol. 3, No. 1, 1981).

specified stations. The quantities are then compared to those of an
equivalent water flow, to contrast the differences between the two.
They are then calculated as a function of time, in order to determine
the velocity and the position of the surge front at each station.
Figures 2.10 and 2.11 show the results of this program as applied to
the Rudd Creek, Utah mudflow of 1983.
Jeyapalan, Duncan and Seed (21) investigated the flow
characteristics of mine tailings such as the Buffalo Creek, West
Virginia flow of 1972. The behavior of tailings materials can be
represented, with reasonable accuracy, by the. Bingham plastic
rheological model. The analysis is applicable to the flow of
tailings on horizontal and sloping planes and in prismatic valleys.
The analysis can be performed using dimensionless charts in the case
of flow on planes, and by the use of a computer program in the case
of flow in prismatic valleys. Computer programs are also available
for analyzing the potential inundation zones likely to result from
turbulent flows of fluid tailings. Flume experiments indicated that
the model is reliable for viscous fluids. Application of the
analysis procedures to field cases also showed good agreement.
However,/ the available laboratory test data for the Bingham
parameters are limited. There is a need for reliable and practical
proceedures for measuring the properties which characterize the flow
of mine tailings materials.
Johnson's model (4) is based upon the Navier-Stokes equations
and assumes that the debris flow behaves as a Bingham fluid. A
Poisson equation is obtained which describes the fluid movement in
channels as a function of the velocity, Bingham viscosity, fluid unit

339 10/3 1913 2130 2698 3223 3/83 4300' 4838 33/3 3913 8430
Figure 2.10- Plan and Profile of Rudd Creek Computer Model
Source: R.W. Jeppson, Mechanisms Associated with Utah*s 1983
Slides and Debris Flows, (Specialty Conference Papers, Utah
State University, 1984).

1 1 1 1 1
1 53 | | | | | | | | | | | | | 8 1075 1613 2150 2688 3225 3763 4: | 00 4t | 38 52 1 1 1 1 75 5913 6450
Figure 2.11 Position of Debris Flow with Respect to Time.
soyrce: R.W. Jeppson, Mechanisms Associated With Utah's 1983 Slides .and Debris Flows, (Specialty
Conference Papers, Utah State University, 1984). -----------------------

weight, slope angle and shear strength. Johnson modified the
Coulomb-viscous model to include the flow of central "plug" or
"raft" of relatively rigid debris, moving at a uniform velocity. The
velocity of adjacent material decreases parabolically outward from
the edges of the plug to the channel boundaries. As the flow reaches
flatter gradients, and the velocity tends to decrease, the thickness
of the flow increases. Plug flow is modeled for three basic channel
shapes: semi-circular, "V"-shaped and rectangular. Infinitely wide
channels are also analyzed. Velocity distributions which closely
resemble the theoretical forms have been measured for many
experimental flows. The assumption of a Coulomb-viscous model
explains many features of debris flow deposits including the tendency
for surfaces of debris flow deposits to be smoother than the surfaces
over which they travel, and the ability of debris flow to transport
such large inclusions.

3.1 General
The mudflow is a very complex form of earth movement. Its
study is not limited to geotechnical engineering alone, but involves
the fields of meteorology, hydrology and hydraulics, geology and
geomorphology. Indeed, each of these fields of research suggests
its own theories as to what the controlling factors are, which affect
the formation of mudflow. Some of these factos may be summarized as
Geologic Formation Many geologic processes and stages of
development have combined to shape the earth's upper crust, leaving
some portions of the landscape conducive to mudflow. These areas are
usually in the form of canyons and ravines, steep mountain swales or
alluvial fans. Some of the factors resulting from geologic formation
include the composition and structure of the soil, the effects of
weathering, the lithology and stratigraphy, faulting, jointing,
foliation and bedding surfaces of the parent bedrock materials.
Environmental and Climatological Factors These factors
include the hydrological or meteorological characteristics of the

region, the micro-climate including the regional precipitation in the
form of rainfall or snowpack, the altitude, and the prevailing winds.
Evaporation, infiltration, slope exposure, and the effects of frost
action, all have profound effects on the slope forming soils and
their stability. Natural earth processes such as earthquake,
volcanic eruption, hurricane, typhoon, tsunami, and flash flooding
also fall into the category of environmental factors.
Human Factors These factors include all activities caused
by human activity, civilization and development. They include the
construction of highways and railway embankments, quarrying and
mining, stripping of landscape and removal of ground cover, the.
surcharging and overloading or slopes, construction of underground
utilities, and vibrational and dynamic loadings.
Fire and Deforestation Another factor which affects the
formation of mudflow (particularly in arid regiohs such as the
southwestern United States) are the effects of fire and high
temperature, and the corresponding loss of vegetative cover. Forest
fire disintegrates and weakens the root network and ground cover
provided by the natural landscape. A watershed which has suffered a
forest fire yields a much higher percentage of sediment during a
flood or a flow than one which has not. It has also been observed
that the intense heat generated by forest fire, has a significant
effect on the strength and soil properties of the upper soil
The factors which affect the formation of mudflow are herein
discussed in greater detail.

3.2 Geologic Formation
It is the source area of the mudflow with which we are
concerned when we discuss the geologic formation. What natural
environmental and climatic processes, such as erosion, wind or even
time, have occurred to form these unstable slope materials? The
present day landscape topography, geologic structure and
composition is the result of millions of years of development and
modification. The science that deals with landscape development is
known as geomorphology. The topographic features found in the upper
horizons of the earth's crust are relatively young in age, but their
parent materials, from which they came, are. actually quite old.
Heavy sediment transport is a part of the long term geologic process
which downcuts the mountains at a rate of approximately one meter per
thousand years (while tectonic processes uplift them at a much higher
The many physical and chemical ways by which the original
composition and structure may be modified include: the action of
wind, water and ice; weathering, mass wasting and erosion; and
volcanic activity. Climatological, and environmental factors are
discussed in greater detail in Part 3 of this chapter.
Geomorphologists have conducted extensive research on the
classification of distinct landforms, which are susceptable to slope
failure and mudflow. Geologic mapping and aerial photography has
aided greatly in the identification and recognition of these source
areas. Because of the nature of their development and the stage of

their evolution, experience indicates that certain landforms are more
susceptable to slope failure than are others.
.Conditions for mudflow occurrences are usually quite
favorable in mountainous regions where very steep slopes and narrow
ravines are often found. Precipitation can be extreme, and ground
cover is often very shallow or even nonexistant. Residual soils
which have formed over the years, due to the actions of glaciers,
water or wind, cannot maintain as steep an angle of repose as the
underlying bedrock, and hence, a delicate balance exists with regards
to the stability of these residual soils.
Mudflows are quite common in glacial and glaciofluviai
deposits. Glacial deposits, as the name implies, are those
formations having been transported by the actions of glaciers. The
forward motion of a glacier is normally stopped during the spring
months or during periods' of warmer temperatures. Melted runoff
causes rock fragments to wash or to settle out, thus producing a
disordered landscape. Glacial deposits may vary in size and
composition from boulders to clay sized particles. Where the clays
often contain sand or silt pockets or lenses, high pore water
pressures can develop, causing the soil mass to disintegrate into a
mixture of sand, silt and chunks of clay. The area where these
materials are deposited are called moraines. Moraine areas are
characterized by deranged drainage patterns, and irregular, untilled
hilly terrain. Mudflows, soil slumps and debris flows occur
frequently in cut slopes in moraine deposits, due to the many
undrained depressions and seepage zones within the soil mass.

Manmade or artificial fills often fail when constructed upon moraine
Mudflows are also common to colluvial soil deposits formed
from sedimentary rocks, depending on the steepness of the bedrock
slope and the depth of the loose, weathered, residual materials.
Colluvial soils are derived from the underlying parent materials by
weathering and gravitational creep. Alternate swelling and shrinkage
of clays, due to changes in moisture, contribute to the breakup of
parent rock material and the downslope creep of the detrius.
Sedimentary rock is the most widely spread form of all surface rock.
Shales, sandstone and limestone residuals are also conducive to the
formation fo mudflow. Certain shales in South America and India
experience a rapid disintegration and become residual soils rather
rapidly. Some igneous residuals, such as basaltic lavas and other
ancient volcanic deposits, are susceptable to flows, especially when
interbedded with a softer layer.
Alluvial fan areas common to the western United States
consist of very loose, unstable soils which have been deposited from
a previous flow and which may oftentimes contribute to future flow
events. Alluvial fans are deposits of unconsolidated rock and soil,
which collect in a fan-like shape at the base of a steep ravine or
mountain swale. They may be formed by a slope failure, mudslide,
mudflow or flood. When storm water concentrates and flows over the
fan area, the loosely compacted alluvial soils are easily eroded and
add to the flowing mass.
Mudflows are also fairly common in glaciated clays and marine
deposits. In the Norfolk Coast of England, mudflows frequently occur

in either fissured or intact clay beds, which are interbedded with
layers of fine sand. Trapped or perched water within these layers
can create increased pore water pressures. These flows are
supposedly caused by erosion of the laminated sand layers.
Reportedly, mudflows have also occurred in varved clays in Canada.
The cause is said to be the loosening of an underlying varve by the
percolation of groundwater. In North Kent, England, mudflows have
also occurred in stiff unfissured London Clay.
3.3 Climatological and Environmental Factors
As previously mentioned, the occurrence of slope failure is
highly dependant upon the geologic formation. The development of
landforms, in turn, is greatly affected by climatological and
environmental factors. When subjected to different variations of
climate, precipitation, surface hydrology, etc., the parent rock
material may yield substantially different soil types, with different
soil properties.
Probably the most important of these factors is that of
climate, and its effects on surface hydrology. The climate of an
area, as expressed in its various components of weather, is the
ultimate driving force influencing slope movement. Some of the data
available from weather stations for investigation include: the
precipitation in the form of rainfall and snowfall, temperature,
prevailing wind speed and direction, relative humidity, and
barometric pressure.

Of these data, precipitation is the most widely studied. A
survey conducted by the Federal Highway Administration indicates that
water -is the controlling factor in over 90 percent of all slope
failures along highways of the Interstate System. Mudflow occurs
when the ground saturates with water during a rainstorm, causing the
pore pressures to increase and the effective stresses to decrease
within the soil mass. Furthermore, the additional weight of the
water at the head of the slope, tends to increase the driving force
of failure. Many researchers have tried to correlate slope movements
with rainfall. Campbell (5) defined two conditions related to
precipitation which must usually be met in order that mudflow shouftf
form: 1) that the antecedant moisture of the soil is too high to
accept any additional water in the form of precipitation, storm
runoff and infiltration, and 2) the occurrence of a high intensity
rainfall outburst. In contrast, Crozier (10) found the correlation
between precipitation and slope movement difficult to justify. Rain
gauges are oftentimes located too far away from the study area to
give a true representation of the actual precipitation at the site.
Precipitation amounts, even within a limited study area, are subject
to great variation, and exact interpretations cannot always be made.
A knowledge of the regional meteorology is important in the
prediction and correlation of precipitation. Considerable research
has been conducted on the effects of weather on flooding or mudflow
formation; Statistical analysis may be used to relate long range
patterns, multiple storm sequences and the frequency of return, based
upon historical data. The mudflows of California and Utah were the
result of abnormally warm, wet Pacific weather fronts. The origin of

these fronts have been traced back to vast heat reservoirs in the
Pacific Ocean, commonly referred to as sea surface temperatures
(3ST).- Coastal and orographic influences, and spatial and seasonal
variations must also be considered in forecasting or estimating
precipitation amounts. The development of satellite imagery and
isohyetal analysis has aided greatly in the determination of rainfall
and snowfall amounts.
Once the precipitation amounts have been determined, its
effect upon the surface and groundwater hydrology must be analyzed.
Surface hydrology is concerned with the areal and temporal
distribution of the rainfal1/snowfal1, the effects of infiltration
and evaporation, the determination of storm runoff quantities, and
the concentration and routing of flows. Surface runoff may reduce
the stability of a slope by erosion of the toe, by increasing the
pore presures within the upper soil horizons, or by increasing the
surcharge weight at the head of the slope. Increased groundwater may
lead to the formation of a perched water table or underground
springs, which oftentimes, go undetected by conventional monitoring
or exploration methods. Subsurface water can also generate increased
pore pressures, and can create a weak, unstable layer within the
overall slope. Moisture also expedites the breaking and fracturing
of rock materials by the continual, or seasonal freezing and thawing
action. This process is quite common is talus rock slopes.
The effects of snowfall and snowpack should also be
considered. Not only does snowpack represent a potential storage of
runoff, but the melting process normally takes place during the
warmer spring months, often simultaneously with the rainy season. If

deep enough, snow can act as a surcharge load upon the slope, thus
increasing the shear stress imposed upon the soil. When thoroughly
compacted, snow and ice can prevent the dissipation of excess pore
water pressures at the surface. Snowpack is usually measured as a
percentage of normal snow depth. For instance, the snowpack of 1983,
in Utah's Wasatch Front Range, which, upon melting, generated
extensive flooding and debris flows, was recorded as being over 200
percent of normal depth.
Temperature is yet another climatic factor which must be
considered. It is, among other factors, a function of the altitude.
On average, for every 500 foot gain in elevation above mean sea
level, the temperature drops by about 1 degree. The transition from
rainfall to snowfall, and from snowpack to snowmelt is a function of
the temperature. Again, it was the drastic change in temperature in
Utah, in 1983, which caused the abrupt snowmelt to occur, triggering
flash flooding and debris flows.
Exposure to sunlight is also another factor which can affect
the long range stability of a slope. Some observers have reported
that north-facing slopes are more stable than south-facing slopes,
even when comprised of similar soil and rock materials. North-facing
slopes are snow covered a greater percentage of the year, experience
fewer days of freeze-thaw, retain their soil moisture longer, and
usually have a better vegetative cover. All of these factors result
in less active erosion and allow steeper slopes to be maintained.
Sudden and natural catastrophic occurrences, which trigger
slope failures or flows, should also be classified as environmental
factors. These may include earthquakes, volcanoes, hurricanes,

tsunamis, and flash floods. Hurricane Camille caused numerous flows
and floods in Western Virginia in 1969. It was an earthquake which
triggered the devastating debris flows at Mt. Huscarin in South
America in 1969 and 1970. Many mudflows in Japan have been
attributed to typhoon flooding, as well as seismic action.
3.4 Human Factors
The mudflow is, basically, a natural phenomenon one which
has been occurring throughout geologic history, and one which will
continue to occur, with or without the intervention of man.
Historically, most mudflows have occurred in remote, nonpopulated
areas, such as mountainous ravines or canyons, or on steep hillsides.
Until more recently, mudflow neither affected nor was affected by
man. However, as civilization and development continue to expand
into new areas, and buildable land becomes more scarce, human factors
have a greater impact.
One way in which man affects the natural environment is by
altering the groundwater table. By raising or lowering the
water table, the shear strength of the soil may be reduced, due to
the changes in intergranular forces (effective stress), due to
variations in water content and pore water pressure. The piezometric
level can be lowered significantly by the installation and pumping of
wells, while irrigation tends to raise the water table. Impounded
water may also raise the local water table and reduce the stability
of the slope. Water may become impounded by the construction of

highway embankment or dam across a natural drainageway. Rapid
drawdown of a water table may cause slope instability if the slope is
not permeable enough to dissipate the excess pore water pressures as
quickly as the water table is lowered. Prior to drawdown, the water
table has a stabilizing effect on the slope, and when lowered, can
reduce the ability of the slope to resist failure.
Another way in which development may affect slope stability
is merely by urbanization of an otherwise natural drainage basin.
Urbanization has the effect of increasing storm water runoff by:
1) decreasing the time required for storm water to concentrate (due
to the construction of storm drains, improved channels, gutters^-
etc.), and 2) increasing the amount of impervious area with the basin
(construction of parking lots, buildings, etc.). When improperly
routed, this concentrated runoff may flow' across and erode loose
soils, or may undercut the toe of of slopes or channel banks, thus
causing or contributing to slope failure or mudflow. Water can be
diverted to potentially unstable slopes due to the construction of a
highway or railroad embankment. Culverts or bridges are normally
placed where an embankment crosses a natural drainageway, however
these structures are often undersized and become clogged with debris.
This forces the mud-laden stormwater to seek alternate crossings,
which may be at undesirable locations of fill material or loose slope
Other ways in which man and his civilization may affect the
formation of mudflow include:
o By overloading or surcharging adjacent to the slope. This could be

the result of embankment fills; stockpiling ore, rock, mine
tailings, or waste piles; the weight of buildings, trains, or even
reservoi rs.
o The removal of lateral support may also increase the shear stresses
exerted upon the slope. This could result from undercutting the
toe of the slope, or the removal of retaining walls or sheetpiling,
o Quarrying or mining could have the effect of removing or disturbing
the foundation soils from beneath the slope, thus causing local
ground collapse. The effects of "piping" from leaking underground
drains or utilities can also create voids and cause settlement in
the subgrade. Leaking water lines, sewer canals, or reservoirs can
also generate excess pore water pressures within the soil, and will
also tend to increase the weight and, hence, the driving force at
the head of the slope.
o The stripping of landscape or the removal of ground cover will
leave the surface soils susceptable to erosion, and will expedite
the weathering process of the underlying rock,
o Transitory earth stresses in the form of vibration and dynamic
loading may trigger slope failure at the head of a mudflow source
area. These stresses may result from blasting, heavy machinery,
train or truck traffic, or pile driving.
3.5 The Effects of Fire and Deforestation
Fire and deforestation also have a significant effect upon
the formation of mudflow and mudflood. Semi-arid, mountainous

regions, commonly found in the southwest United States are subject to
extreme variations in seasonal climate and precipitation. Extended
periods of draught can leave the mountain-desert vegetation
(sometimes known as chapparall) dried and parched to the point where
even the slighest spark can set a raging fire. Fanned by winds,
which are often caused by the change in pressure and temperature due
to the intense heat, the fire can spread and incinerate an entire
drainage basin in but a matter of hours.
Case histories in Southern California have shown that erosion
and total sediment yield may be increased by as much as fifty times
for the first year following a significant burn. High temperature
fires affect the formation of mudflow in the following different
1) By loosening the outer few inches of the soil profile.
This causes the concentration and redistribution of the coarser
grained soils (layering), thus leaving the soil mantle vulnerable to
2) The intense heat generated by forest fire (as high as
2000 F.) causes a chemical transition to occur, forming a relatively
impermeable, wax-like, water repellant zone, just a few inches below
the ground surface. This transition occurs in what is known as
hydrophobic, or non-wettable soils, and tends to inhibit the
infiltration of rainfall into the subgrade. Essentially, it is
caused by ammonium hydroxide and other organically derived compounds,
which accumulate below chapparall vegetation, and tend to vaporize
when exposed to high temperature. The vapor condenses in a cooler
zone of concentration, a few inches below the ground surface,

forming an impervious layer. Rainfall penetrates the surface layer
and reduces its shear strength. Excess water migrates downslope,
carrying away the weakened material in the form of mudflow.
3) Fire destroys the natural vegetative cover, which protects
and reinforces the underlying soil, leaving the ground surface
unprotected from concentrated storm runoff. The character and
density of the vegetation has a profound influence on the way in
which storm runoff is dissipated. Ground cover prevents the
raindrops from hitting the soil at a high velocity, allowing a
greater amount of moisture to enter the soil, rather than to run off
the surface and cause erosion. Once the moisture is absorbed by the
soil, a part of it is taken up by the vegetation, which returns most
of it back to the atmosphere through the process of transpiration.
Ground cover also imparts strength to the soil through its
interlocking roots, which prevent down-slope movement of the soil.
4) Although it has not necessarily been proven, it has been
reported by some observers that intense heat tends to at least
change, if not reduce the shear strength properties of a soil to that
of a residual value.
Davis (13) investigated the relationship between forest fire,
sediment yield and mudflooding in the Los Angeles County area. Los
Angeles County comprises a land area of approximately 4,000 square
miles, of which 47 percent is mountainous, the rest being alluvial
valleys and coastal plains. Much of the deposited debris resulting
from the storms of 1978 and 1980 can be directly attributed to burned
watersheds caused by four major fires. The Mill Fire of 1975, the
Village Fire of 1975, the Middle Fork Fire of 1977 and the

Pinecrest Fire of 1979 consumed 208 square miles of watershed above
urban areas, dams and debris basins owned and operated by the Los
Angeles County Flood Control District. Of the total 2,250,000 cubic
yards of debris deposited in the District basins in 1978 and 1980,
900,000 cubic yards can be directly attributed to watershed burns.
Figure 3.1 and Table 3.1 show the relationship between percent of
burned watershed and debris production for 59 different drainage
basins. In most cases, a basin which had previously suffered a major
burn, produced a higher yield of debris upon the occurrence of

Debris Production mVkm
1 o6
1 o4
1 o3
1 o2
Basin Area km2
Figure 3.1 Debris Production Rates During 1978 and 1980
Source: J. Daniel Davis, Rare and Unusual Postfire Flood Events
Experienced in LoS AngeleFDuring 1978, (California Institute of
technologyi Proc. Storms, Floods and Debris Flows in Southern
California and Arizona in 1978 and 1980), p. 250.
1 I r I I I I I
i i i I ii n
IS $

o 0
$>£ O D

g D

' > i i i i 111
' i i i 11
j_________i______i -i i i i

Mo. Debria Basin
Sq Mila Sq Kn
Debris Production
Cu Yd/ cu H/ In Teres Cu Yd/ Cu H/ tn Terns
Sq Hlla Sq R* of Mean Sq Mile Sq Kb of Mean
Year Percent
I Aliso 2.77 7.17 9,700 2,860 3.25 6,600 1,950 2.21 1970 66
2 Auburn 0.19 0.49 16,100 4.750 3.82 88,900 26,240 21.08 1977 100
2 Sailay 0.60 1.55 22.500 6.640 6.28 145,200 42,860 40. S3 1977 100
4 Big Dalton 2.62 6.79 31,100 9,180 5.18 38,800 11,450 6.47 1960 100
5 Blanchard 0.50 1.29 73,200 21.610 9.15 14,500 4,280 1.83 1975 88
Blue Gua 0.19 0.49 100,600 29,700 12.58 13,400 3,960 1.68 1975 100
7 Bradbury 0.68 1.76 12,100 3,570 1.37 25,500 7,530 2.89 1958 100
8 Brand 1.03 2.67 51,600 15,230 15.75 22,000 6,490 6.71 1927 66
9 Cartar 0.12 0.31 200 60 0.12 88,900 26,240 53.88 1977 100
10 Child* 0.31 0.80 13.900 4.100 2.39 7,700 2,370 13.26 1927 100
11 Cook* 0.58 1.50 102,200 30,170 23.35 22,000 6,490 5.03 1975 100
12 Dear 0. 59 1.53 37,600 11,100 9.27 11,200 3,310 2.76 1964 100
13 Dunaauir 0.64 . 2.18 98.700 29,140 27.33 22,800 6,730 6.31 1975 100
14 Bagla 0.48 1.24 60,300 17,800 6.54 10,300 3.040 1.12 1975 100
15 Englewild 0.40 1.04 2,600 830 0.28 30,900 9,120 3.06 1968 100
16 TmLz Oak* 0.21 0.54 12,400 3,660 1.66 150 44 0.02 1935 100
17 fern 0.30 0.78 14,600. 4.310 1.91 7,100 2,100 0.93 1962 48
18 Gould 0.47 1.22 8,400 2,480 1.39 7,100 2,100 1.18 1959 82
19 Kalla 1.06 2.75 43,600 12.870 5.82 18,400 5,430 2.45 1933 85
20 Harrow 0.43 1.11 5,100 1,510 0.63 7,500 2.210 0.92 1968 100
21 Hay 0.20 0.52 21,700 6.410 6.73 0 0 0 19S9 100
22 Hillcreat 0.35 0.91 16,000 4,720 4.67 3,800 1,120 1.11 1964 100
22 Hog 0.30 0.78 12,800 3,780 10.10 8,600 2.540 6.79 1962 100
24 Hook Eaat 0.18 0.47 11,900 3,510 1.49 13,200 3,900 1.65 1968 TO
25 Hook Vaat 0.17 0.44 10,800 3,190 1.35 21,200 6,260 2.65 1968 98
26 Kinnaloa 0.10 0.52 16,500 4,670 1.94 6,900 2,630 1.05 No Recorded
27 Klnnaloa West 0.16 0.41 22.300 6.500 2.62 20.300 5.990 2.39 Bum
29 Lannan 0. 23 0. 63 12. 900 3.7 90 l. 71 3S.900 lo.eoo 4.79
29 Las Floras 0.45 1.17 17,400 5,140 6.05 74.400 21,990 25.09 1979 90
30 La Tuna 5.34 13.93 32.200 9,510 11.42 14,300 4, 220 5.07 1955 61
31 Llneklln 3.69 9. 56 10,200 3,010 2.89 S.800 1,710 1.64 1970 95
32 Lincoln 0. SO 1.29 11.000 3,250 3.23 5, 700 1.680 1.67 1935 98
33 Little Dalton 3. 31 8.57 22,400 6,610 3.73 30,200 8.920 5.03 1960 89
34 Haddock 0 35 0.65 1.700 500 0.38 21,800 6,440 4.82 1958 95
35 May Mo. 1 0 ''0 1.81 10,400 3,070 2.04 3, 200 940 0.63 1966 100
36 Hay No. 2 0.09 0.23 6,600 1.950 1.30 6.200 1,830 1.22 1966 100
37 Morgan 0.60 1.55 7.100 2.100 3.47 9,700 290 4.75 No Recorded
38 Nichols 0.35 0.91 6.700 1,980 1.36 29.200 8,620 5.93 Burn No Recorded
39 Pickens 1. 50 3.88 09,900 26,540 18.09 15.800 4.660 3.18 Burn 1975 86
40 Rowley 0.27. 0.70 174,600 5,160 32.00 27.700 8,180 5.07 1975 58
41 Rubio 1.26 3.26 5.400 1,590 1.42 84,557 24,960 22.20 1979 100
42 Ruby Lower 0.28 0.73 180 53 0.07 11,800 3,483 4.64 1953 88
43 Santa Anita 1. 70 4.40 23,800 7,030 3.97 12,400 3,660 2.07 1969 23
44 Sawpit 2. 7B 7.20 7,900 2,330 1.32 13,200 . 3,990 2.20 1958 60
45 Schoolhousa 0. 20 0.73 5,700 1,680 0.81 0 0 0 1962 100
46 Shielda 0.23 0.60 117,200 34,600 15.35 49,400 14,580 6.47 1975 87
47 Sierra Hadra villa 1.46 3.70 8,000 2,360 1. 33 64.200 18,950 10.70 1978 S3
48 Snover 0.23 0.60 7, 300 2.150 0.97 18,000 5,310 2.40 1933 97
49 Spinks 0.44 1.14 7,900 2,3 30 1.64 21,300 6,290 4.42 1953 17
SO Stetson 0.29 0.75 S, 300 1.S60 5.70 4,800 1.420 5.16 1962 100
51 Sullivan 2. 38 6.16 9,900 2,630 5.20 14,800 4,370 8.64 1979 45
S3 Sunset Upper 0.44 1.14 43.000 12,690 9. 38 31,600 9.330 6.89 1964 99
53 TurnbuJ1 0.99 2.56 2,900 B30 4.55 4,300 1,270 6.99 1967 66
54 Ward 0.12 0.31 148,100 43.720 37.60 38,900 11,480 9.88 1975 100
55 Neat Ravine 0.25 0.65 8,300 2,450 1.18 6,100 1,800 0.87 1935 100
56 Wildwood 0.65 1.68 23,000 6,790 3.44 12,700 3.750 1.90 1957 76
57 Wilson 2.58 6.68 S, 500 1,620 1.34 2,058 610 0.50 1966 35
58 Winery 0.18 0.47 2,700 800 0. J6 20 6 0 1962 36
59 Zachau 0. 35 0.91 141,800 41,860 18.91 26,400 7,790 3.52 1975 98
Table 3.1 Debris Production for Individual Debris asins During the
1978 and 1980 Storms.
Source: J. Daniel Davis, Rare and UnusaVPostfire Flood Events
Experienced in Los Angeles During 1978, (California Institute of
Technology, Proc. Storms, Floods and Debris Flows in Southern
California and Arizona in 1978 and 1980), pp. 248-249.

The mudflow is formed by a variety of processes, both natural
and manmade, which combine either to reduce the shear strength or to
increase the shear stresses imposed upon the slope forming material.
Seldom, if ever, can a slope failure be attributed to one single
factor. The processes which lead to a mudflow actually begin with
the formation of the slope itself. The effects of erosion,
weathering, and tectonic movement all contribute to the formation of
the potentially unstable slope materials and geometry. Eventually
the strength of the soil is exceeded by some event or triggering
mechanism, and the material is set in motion. Sowers and Sowers (33)
point out that "calling the triggering factor which leads to the
failure of a slope (which may already be set in motion) the cause of
the failure, is like calling the match that lights the fuse which
detonates the dynamite which destroys the building, the cause of the
Mudflow is most often formed by a reduction in shear strength,
an increase in shear stress or by a change in drainage conditions.
These factors are discussed herein, in greater detail.

4.1 .Conditions Which Lead to the Formation Of Mudflow
4.1.1 Shear Strength
One of the properties which controls the stability of an
earth slope and is a major factor in the formation of a mudflow, is
the shear strength of the soil. The angle of internal friction, ,
and the cohesion, c, are the two parameters commonly used to define
the soil shear strength. The two components of stress to which an
element of soil may be subjected are the normal stress, o, and the
shear or tangential stress, T. Using the Mohr-Coulomb failure
criteria, the stresses which constitute failure along a plane in an
element of soil may be defined by the following relation:
T = c + a tan$ (4.1)
The Mohr-Coulomb failure theory states that failure will occur along
a plane on which the ratio of the shear to normal stress reaches a
critical and limiting value, defined by the Mohr-Coulomb failure
criteria. The shear stress along the plane of failure, T^, is
called the shear strength. The cohesion, c, is the result of
either overconsolidation or of electro-chemical or valence bonding
among fine-grained particles, such as clay or silt minerals. The
cohesion appears to yield a component of strength to the soil, even
when the normal stress, a, equals zero. The cohesive strength of a
slope forming soil is often ignored in order to simplify the analysis
and to yield a more conservative factor of safety, since this

strength is often lost as the soil becomes saturated or is subjected
to large strains. Although this may be an appropriate assumption
in the conservative design of a slope or embankment, the cohesion
may play an important role in the stability (or failure) of a slope,
and should be included in our present study. Soil samples recovered
from deposited mudflow materials, often indicate the presence of a
substantial percentage of fine-grained cohesive soils.
The problem of correctly assessing the shearing resistance of
a cohesive soil, and the change in strength which occurs with time,
is one of the more difficult and important questions which the
geotechnical engineer faces. Hvorslev (27) modified the Mohr-Coulomb
failure theory to include the effective cohesion, cg, as a function
of the void ratio, e. Experimental results show that cohesion is
directly related to the void ratio for both normally and
over-consolidated clays. Hvorslev also found that the shear strength
of a soil is time-dependent, upon the manner in which the load is
applied. The Mohr-Coulomb failure criteria may be used to analyze
the state of stress or strain of a given soil element, at a
particular point in time, however, it does not take into account
changes in loading history. Since soils are not perfectly elastic
materials, their behavior is dependant not only on the initial and
final states of stress, but also on the way in which the state of
stress is changed and on the previous history of loading. The
effective stress path method serves to trace the series of stress
increments which lead to the failure of a particular soil. The
Horvslev surface (see Figure 4.1) is a three dimensional plot in
stress Space of the variables q, p, and e,

Figure 4.1 Hvorslev Failure Surface
a) Void -Ratio- Effective -Stress Diagram at Failure
i i
i i
Figure 4.2 Hvorslev Failure Envelope
Source: William H. Perl off and William Baron, Soil Mechanics
Principals and Applications^ (John Wiley a Sons, 1976) p.356.

where: q =
P =
e =
l/j2 [( lh (trl+ff2+ff3^
void ratio
volume of voids
total volume
The cohesion may be expressed as a function of the equivalent
consolidation pressure, cfe, the consolidation pressure corresponding
to the void ratio of the soil at virgin compression (see Figure 4.2).
Fine grained soils are also very susceptable to strength loss
due to remolding. This property, known as sensitivity, is especially
pronounced in cohesive materials. Disasterous earthflows have
occurred in highly sensitive clays found in the coastal regions of
Norway, Canada, and Japan. The process is spontaneous, yet
iterative, in that the initiation of slope movement causes remolding
of the soil, which in turn causes a reduction in strength. As the
soil weakens, it may undergo a greater degree of movement,
corresponding to a greater degree of remolding, until enventually, it
fails and flows like a dense viscous fluid. In Ullensaker, Norway,
in 1953, 200,000 cubic meters of sensitive clay liquified and flowed
away in one such earthflow occurrence.
Quantitatively speaking, sensitivity is the ratio of the
undisturbed shear strength to the remolded shear strength at the same
water content. When a sensitive clay is remolded, the clay particles
are reorientated and become more dispersed, reducing the attraction
between the particles and destroying all chemical bonds, thus
decreasing the strength of the overall soil structure. In soils with
soluble chemical bond, the soil may lose strength as it becomes

saturated with rain water, and the salt concentration and ion
exchange is reduced by leaching.
Mudflow may occur in almost any type of soils. Usually the
soil possesses some degree of cohesion, but is not necessarily pure
clay (i.e. <{> = 0). Figure 4.3 shows the mean and standard
deviation of grain size distribution for 65 soil samples taken from
mudflow deposits from the 1983 Utah mudflows. The distribution is of
a fairly well graded soil with particle diameters as large as 3/4
inch, but with a substantial portion (approximately 20%) of
fine-grained particles. Analysis of other mudflow deposits (see
Chapter V.) reveals that the mud, usually, consists of a

cohesive strength, Cg, prior to loading, the rapid precipitation
which commonly accompanies mudflow, tends to mobilize the cohesive
strength rather quickly.
Even if the mudflow material does not consist of a pure clay,
it may still undergo a reduction' in shear strength upon failure.
This is due to a variety of reasons such as the physical breakdown of
the soil particle bonding, due to the shearing action of the failed
mass. As an overconsolidated soil is sheared, the particles tend to
realign somewhat parallel to the failure surface, thus offering a
reduced resistance to failure. This reduced strength, known as the
residual strength,^, causes the slope to be less stable at flatter
angles of repose, and also requires a flatter gradient for the
deposition process to occur. This strength reduction is common in
areas where a pre-existing failure surface, such as a landslide, 1s
present. Loss of strength may also be attributed to chemical changes

Figure 4.3 Average Grain Size Distributions Curve from 65 Soil Samples Taken
from Utah's 1983 Mountain Slide Areas.
Source: R.W. Jeppson, Mechanisms Associated With Utah's 1983 Slides and Debris
Flows, (Specialty Conference' Papers, Utah State University, .1983).

or to the loss of cementing agents in the soil structure due to
In Aberfan, Wales, in 1966 (27), 140,000 cubic yards of
debris from a mine tailings embankment underwent failure and flowed
into the nearby city, killing 144 persons. A typical section of the
Aberfan slide is shown in Figure 4.4. The slope consisted of
excavated mine waste, predominantly fragments of shale rock mixed
with mine tailings and debris, piled to a height of 200 feet, at an
average angle of 36. Bishop found the drained shear strength
parameters of undisturbed samples taken from the slope to be ' =
39.5 and c = 0.
The slope is underlain by a glacial deposit which had
undergone movement in the past, and had formed a pre-existing slip
surface. Underlying the glacial drift is a water bearing sandstone
which produced an underground spring, causing periodic erosion and
undercutting of the toe of the slope. Bishop found the slope to be
unstable and that it had been moving for a number of years prior to
its failure. The.movement was caused by the continual erosion of the
toe, and was driven by the dumping and surcharging of materials at
the head of the slope.
The failure was triggered by heavy rains, which caused the
pore pressures in the sandstone layer to build. Increased water
pressure in the toe of the slope reactivated the sliding movement
along the preexisting slip surface. However, pore water pressure in
the slope itself was minimal. The outward movement of the toe
created large shearing strains, which caused a localized failure and
a flow to occur. As the flowing mass was removed from the toe, a

Figure 4.4 Typical Schematic Cross-section of Tip No. 7 at
Aberfan, Wales.
Source: William H. Perloff and William Baron, Soil Mechanics
Principals and Applications, (John Wiley & Sons, 1976) p.579.

secondary failure of the overall slope occurred. The secondary slide
cut into the sandstone layer and released the impounded water from
it. The water mixed with the mine tailings, the glacial deposits and
the debris, to form a debris flow.
The factor of safety was originally calculated to be 1.45,
based upon the shear strength of the slope materials. However shear
tests conducted on samples taken along the slip surface indicated an
residual angle of shearing resistance, r, of 18. In order to
maintain equilibrium (F.S. = 1.0) it would require an angle of
internal friction of 29. This loss of strength was attributed to
the physical breakdown of the particles due to shear and chemical
changes. The interface between the bedding plane and the residual
soil often represents a boundary of loose, weak soil where shear
failure may occur.
The natural slope of the bedding plane is also an important
factor controlling the formation of mudflow. Mudflow is most likely
to occur within a limited range of slope angles. Campbell (5)
estimated this angle to range from 26 to 45 degrees from horizontal
(Figure 4.5). Slopes in the Southern California region are generally
stable for angles less than 26 under most precipitation conditions.
Where the slope angle is greater than 45p, the soil cover is commonly
quite thin due to wind, erosion, previous slides, etc. In these
locations bedrock is predominant at the surface, and the thicker
residual deposits which are necessary for the formation of mudflow
are not prevalent. These slopes are more prone to rockfalls than to
mudflow. In Chapter V. the range of slope angles for the cases

Figure 4.5 Range of Slope Angles for the Occurrence of Debris Flow
Source: Russell H. Campbell, Soil Slips,
Debris Flows, and Rainstorms in the Santa
Monica Mountains arid Vicinity, Southern
California, (Washington, D.C., Geological
Survey Professional Paper 851, 1975), p. 12.
ai Ratio

Q Percent
Likelihood of
slab soil slips
Common effects
on flow velocity

studied is found to be 15 to 50, much in accordance with the range
reported by Campbel1.
4.1.2 Drainage Conditions
Application of the Hvorslev hypothesis or the stress path
method requires a knowledge of the insitu drainage conditions. A
load applied to a soil sample is carried by either: 1) the soil
skeleton itself, through intergranular contact, or 2) by the pore
fluid (commonly water). In soil mechanics, water is assumed as an
incompressible material, which does not undergo any volume change
(within the range of stresses relevant to our present study) when a
load is applied. The principle of effective stress, as proposed by
Terzaghi in 1923, defines the effective stress, o, as that portion of
the total normal stress not carried by the pore fluid or,
d* = d u (4.2)
where u is the pore water pressure.
In saturated soils, the effective stress is not directly
measurable, and is, therefore, a somewhat intangible quantity
(however, it can be determined, indirectly, from the total stress and
the pore water pressure). In most practical instances, soils contain
some amount of water, and is at least partially saturated. The
effective stress concept is, therefore, the most applicable method
throughout the field of geotechnical engineering. The relationship