COULOMB AKD E_hRTH PRESSURE

by H. &. GOLDER
Lcmdon

There are two approaches to a natural problem. They are the approach of the pure
scientist and that of the engineer.
The pure scientist is interested 6nly in the truth. For him there is only one answer---
the right one-no matter how long it takes to get it. For the engineer, on the other hand,
there are many possible answers, all of which are compromisesbetween truth and time, for
the engineer must have an answer now ; his answer must be sufficient for a given purpose,
even if not true. For this reason an engineer must make assumptions-assumptions which
in some cases he knows to be not strictly correct-but which will enable him to arrive at an
answer which is sufficiently true for the immediate purpose.
Mistakes are not made when an engineer makes his assumptions. Mistakes are made
when other engineers forget the assumptions which have been made, and the assumptions
on which a theory is based are only too easily forgotten. How many engineers, for instance,
could write down the assumptions on which the Theory
of Bending is based, and how many would apply the
theory to a short, deep beam without realizing that this
would be an error ?
One of the theories which is most maltreated in this
respect is the Wedge theory of Earth Pressure. This
theory was first propounded bv Coulomb, but so much
has since been attributed to him which he never said,
and so m,uch bar; 1, :en forgotten which he did say, that
it is good to gc . . k to his original work and-review the
part which is terest to students of Soil Mechanics.
Charles A .gli.-,tin Coulomb was born at AngoulCme
in France on Illne 14, 1736. He was a descendant of a
distinguished family of Montpelier. He would have c
liked to have devoted his whole attention to the pursuit t
of science, but he found it more convenient to enter the E

military profession as an engineer. He served in FIG. 1

Martinique and North ;\merica before being stationed
permanently near Paris in 1781, when he devoted himself entirely to the investigation of
the laws of electricity and magnetism and of the forces of friction and resistance of various
liinds.
He was a profound and ingenious natural philosopher and his work was very diverse.
Among his many essavs and mkmoires are works on the force of torsion, improvements to the
compass, a type of &ving bell, observations on windmills, electricity and magnetism, the
friction of pivots, the circulation of sap in trees, observations on the daily labour performed
by men, and the cohesion and viscosity of fluids. He invented the torsion balance and pro-
pounded the inverse square law both for electric and magnetic charges. His experiments
provided much of the evidence on which the science of electrostatics has been based.
At one time he tried unsuccessfully to rent a windmill in order the better to study the
sails and mechanism.
At the Revolution he was expelled from Paris as one of the nobility and he went to live
on a smaIl estate which he owned near Blois, where he made observations on plant physi-
ology. He was later recalled to Paris to take part in the new determination of weights and
measures which had been decreed by the revolutionary government.
Dr. Thomas Young, in his Life of Coulomb (Young 1855), says that his moral
character was said to have been as correct as his mathematical investigations. At an early

Downloaded by [ UNIVERSITY OF SHEFFIELD] on [16/09/16]. Copyright ICE Publishing, all rights reserved.
 COULOMB AND E,iKTH PRESSURE 67

period of life he gained the grateful acknowledgments of the inhabitants of Brittany for
his disinterested exertions in preventing the execution of some public works which threatened
to be ruinous to the province. His manners were serious but gentle, and sometimes diversi-
fied by a mild gaiety which made him very amiable in society. His disposition was generous
and benevolent, but notwithstanding all his modesty, he could exhibit sufficient spirit when
he was called upon to repel an unjust attack.
He died on August 23, 1806. He was a Lt.-Col. of Engineers, Chevalier of the Order of
St. Louis, Member of the Legion of Honour, and of the Academy of Science.
In 1773 he presented to the Academic des Sciences the essay which contained his earth
pressure theory. This was his earliest work and was entitled Essai sur une application des
regles de Maximis et Minimis & quelques ProblPmes de Statique, relatifs a 1Architecture.
In addition to the work on earth pressure, the essay contained notable contributions to the
theory of strength of materials and of the arch. (See Hamilton 1936.) According to Young
this was the first time that maxima and minima had been used as criteria &in statical
problems.
At this time (1773) Coulomb was not a member of the Academic, to which he was elected
in 1786, and the essay was published in 1776 in volume vii of Memoires par divers Savans
Etrangers. As a humble outsider, Coulomb in his introduction, used a metaphor, which
in view of the science which was to grow up from his work, is singularly apt if not prophetic.
He said :-.
Ce Memoire, compost depuis quelques anners, netoit dabord destine qua mon usage
particulier, dam les differens travaux dont je suis charge par mon etat ; si jose le presenter
B cette Academic, cest quelle accueille toujours avec bonte le plus foible essai, lorsquil a
lutilite pour objet. Dailleurs les Sciences sont des monumens consacres au bien public ;
chaque citoyen leur doit LIIItribut proportionne a ses talens. Tandis que les grands hommes,
port& au sommet de ledifice, tracent et elevent les eta, 2s superieurs, les artistes ordinaires
repandus dans les etages inferieurs, ou caches dans lobscurite des fondemens, doivent seule-
ment chercher a perfectionner ce que des mains plus habiles ont tree.
To-day, 175 years after it was written, I lobscuritis des fondemens is slowly being
dispelled.
Coulomb stated that he had assumed the resistance due to friction.to be proportional
to the pressure, as Monsieur Amontons had found, and the total cohesion to be pro-
portional to the area of the surface of rupture. This is the basis for the familiar Coulombs
Law : s=c+n tan 4. He also said that it is best to carry out tests on the material which
is to be used, and although he was talking of stone and brick, the advice would be accepted
as sound by anyone who knows anything of soils. He pointed out that friction and cohesion
are not active forces like gravity, which always exercises its complete effect, but reactions
which are measured by the limits of their resistance. When one says, for example, that for
a certain polished wood, the friction on a horizontal surface of a body weighing nine pounds,
is three pounds, one means that any force less than three pounds will not disturb its state
of repose.
In the sections on strength of materials and the resistance of pillars of masonry, Coulomb
proved incidentally, that in a compression test on a purely cohesive material, failure occurs
on a plane of 45 to the direction of compressive stress and that the compressive strength is
twice the shear strength. He made one of his few errors here in assuming that the shear
strength and the tensile strength are equal. He was led into this error by too few tests on
stone, but he was the first to include the compression side of the beam in his analysis of
bending forces and thus broke away from an error which had persisted from the time of
Galilei.
It is interesting to note that Coulomb did not use an angle of friction in his calculations
for earth pressure, but a coefficient of friction which he called n. in modem nomenclature
n=cot # where $ is the angle of internal friction, but Coulombs approach was necessarily

Downloaded by [ UNIVERSITY OF SHEFFIELD] on [16/09/16]. Copyright ICE Publishing, all rights reserved.
 68 H. Q. GOLDER
free from the assumption that 4 is. equal to the angle of repose, an assumption which
led to much confusion in the nineteenth century.
The angle of repose is approximately equal to the lower limit of the angle of internal
friction which can rise to much higher values.
The cohesion Coulomb called 6, which was assumed to be a constant.
By considering the statics of a wedge of earth such as CBK (Fig. l),* he obtained an
expression for the ,horizontal force A which would keep the wedge in equilibrium when the
full shearing resistance was mobilized along BK.
He then showed that force A could increase to a value A, which is what we now call
the passive resistance, and that equilibrium would be maintained until this value was exceeded
when the wedge would move up the plane BK. He pointed out that the expression for A
became infinite if x=a/n, which meant that the plane of failure was inclined at # to the
vertical. The frictional reaction across the plane BK would in this case be horizontal and
therefore capable of supplying the necessary reaction to prevent the wedge from moving
up the plane however great A might become. He further showed that when both friction
and cohesion were zero his expression for A reduced to a fluid pressure, and that A was always
less than this fluid pressure and A was always greater.
Coulomb then stated that it was necessary to find the surface of rupture for which A
is a maximum and A is a minimum, but he would concern himself only with the former. He
realized that the surface of rupture might be curved but assumed that a straight line would
give the maximum pressure since this conforms with experience in that when a retaining
wall fails due to the pressure of the earth behind it, the mass of earth which slips approaches
very closely to a triangle.
According to Collin (1846), Perronet (1788) was the first to observe, or at least to record,
that in high clay banks failure takes place on a curved surface. (See also Skempton 1946.)
Coulomb then differentiated his expression for A, equated the result to zero, and so
obtained a value for X, the distance from the top of the wall to the top of the plane of rupture.

[or a =45+$/2]
Substituting this value of x in the expression for A he expressed the maximum value
of A, i.e., the active pressure, as A =ma--GEa, where m and 1 are coefficients depending on
the coefficient of friction n. He pointed out that the cohesion has no influence on the
position of the plane of failure and that for a purely cohesive material the plane is inclined
at 45.
This expression for the active pressure on the back of a vertical wall, with no surcharge
and no wall friction, can easily be reduced to the more familiar expression :-
1 1 -si?z fj 1 -sin +
2 wh------ -2chp
1+sin + cos 4
by a simple trigonometrical transformation if n is written equal to cot 4. This was first done
by Woltmann (1794) who omitted the cohesion term. This basic earth pressure equation
was restated by Francais (1820), and by several other authors in the nineteenth century,
and by Bell in 1915.
Coulomb nowhere stated that the distribution of pressure with depth is triangular, but
he implied this in his calculation of the overturning moment on the wall. In his expression
for A, viz. : A=ma--%a, he treated a as a variable, when dA=da(2ma--61), which he
said expressed the difference between the pressures on surfaces CB and CB.

*Fig. 1 is taken from the original diagram in the essay, but the lettering has been altered to avoid a
slight confusion which existed.

Downloaded by [ UNIVERSITY OF SHEFFIELD] on [16/09/16]. Copyright ICE Publishing, all rights reserved.
 COULOMB AND EARTH PRESSURE

Et puisque la verticale CB ne peut pas porter une moindre force que A, la ligne BB
_ ne pourra point etre supposee pressee dune moindre force que dA : The element of moment
of the force dA is therefore (b--~)&wz---6Z)da (see Fig. 1) and the total overturning
nbJ 61b=*
moment about the base is- -- In this statement the assumption of hydrostatic pres-
3 2
sure distribution is implicit.
The statement quoted above is not correct, but it is difficult to see this because of the
negatives and inequalities which the sentence contains. Logically it should have finished
I -
. . . la ligne BB ne pourra point &tre supposSe pressee dune plus grande force que dA.
For the force on CB+ the force on BB=A+dA. The force on CB cannot be less than A,
i.e., is >A. Therefore, the force on BB <dA, i.e., the force on BB cannot be greater
than dA.
Given his statement that the force on BB cannot be less than dA, Coulombs assumption
that it must equal dA is reasonable, since it gives the maximum value of the overturning
moment and-also since it is difficult to conceive any reason why the force should be greater
than dA.
Had he not made this slip, however, and had he stated that the force on BB could be
less than dA, it is improbable that he would have made the .assumption that the whole of
the increment of force dA must act on the increment of length BB, if only-for the reason
that this would not lead to the maximum overturning moment.
It is perhaps wrong to call this an assumption of Coulombs, but it is at least a premise
which has been insufficiently examined since, and as such affords a good example of the
forgotten assumptions referred to earlier.
However this may be, whether Coulomb knowingly assumed that the pressure distribu-
tion was hydrostatic, or was misled by this simple but rather subtle error to a belief that
this was statically necessary, the fact remains that, although as he himself proved, the
differential of the total pressure, i.e., the rate at which total pressure increases, is directly
proportional to the height of the wall, this does not mean that the distribution of pressure is
hydrostatic. The distribution of pressure cannot be determined from the wedge theory,
but is dependent on the way in which the wall moves.
Coulomb then gave examples to show that the formula obtained gave results in accord-
ance with the then accepted practice. The friction was taken as equal to that in earth,
.which when left to itself, takes up an angle of slope of 45, and the cohesion was taken as
zero, as is the case for newly deposited earth. It is interesting that Rankine (1862) also
recommended that cohesion should be taken as zero, but for the opposite reason, namely,
that weathering in time destroys the cohesion. Coulombs reason is probably the better.
Coulomb claimed that the thickness of walls built by Marshal Vauban, the celebrated
French military engineer, agreed sufficiently well with those given by his method. He
admitted, However, that Vauban also built counterforts to his walls, but said that this increase
in strength could not be regarded as unnecessary in fortifications which must not collapse
at the first cannon shot. It is interesting to remember that at this time probably all retaining
walls were part of some system of fortifications, and that only with the coming of the railways
some sixty years later, were retaining walls used on a large scale in civil engineering works.
The critical height of a vertical bank of cohesive earth was next determined by putting

A =0 in the formula. The result a=: is the same as the more readily recognizable form

h=4 1+sifi 4
- In the same way, stated Coulomb, if the height of the excavation is given,
W cos iQ

one can determine the angle ofslope such that the soil will support itself by its own cohesion.

*Note : There are several misprints in the form&z in the original.

Downloaded by [ UNIVERSITY OF SHEFFIELD] on [16/09/16]. Copyright ICE Publishing, all rights reserved.
 70 H. (2. G0cDE.R

He did not solve this problem, however, and it was left to Francais (l&20), who deduded
the expression, based on a plane slip surface :
4c sin /3 tos +
ff__-
in I-cos (fl-#
where fi.= the critical slope for height H.
This was restated by CuImann (1866) to whom it is generally, but incorrectly, ascribed.
The question of a surcharge P was then discussed and simply solved, after which Coulomb
went on to deal with the question of wall friction. He realized that wall friction reduces thl:
total pressure and also increases the stability of the wall by augmenting the moment of its
weight about the toe. He gave a value of x corrected for the effect of the wall friction, thii
value to be substituted in the expression for 9. He did not include the effect of wall adhesion.
The effect of wall friction was first included in the trigonometrical form of the formul:l
by Xayniel (1808).
Including the effect of wall friction Coulomb deduced the rule of thumb r=l3/lOO
where c is the thickness of the wall and 11is its height, thus a wall 3 ft. thick would support
a bank 20 ft. high. In the same wav a formula &uld be obtained for a 1)attered wall, but
the thicknesses given were much smailer than those usually employed in practice. Couloml)
gave three reasons for this, which were
1 The friction of soil on mnsonr> is not 3; great as that of soil on soil.
2. \iater often percolates into the soil, or IJetwecn the soil and the masonry and set Y
up hydrostatic prcssurcxh. Although to ob\-iate this, drains are built in behind the wall,
these drains become l~locked citllcr 1,~ the> se)il in snspcnsion in the water or 1)~ ice, and arc%
therefore usc~lcr;s.
3. Humidity c-ll~ng~~ not only the \veight of t!lc soil, bllt also its friction. Couloml)
attested to having seen soapy soils which would stand at an angle ot 45 when dry, but
which when wet would hardly stand at an angle of 60 to 70 (presumably to the vertical).
Therefore, in order to u3(~ the dimensions a< given 1)~ the formula, it was necessary that no
water should penetrate the soil whosc~ prcssnrc has hccn calculated, or if it did that it should
canse no increase in volume. This increase in \-olumc which humidity causes in soils (atI
esample is the cracks which dronght causes in the surface of the gronnd) produces a prcssul-(L
on walls which crnly csperience can dcterminc.
Coulomb then made some remarks on the effect of frost on walls. He said that it
increases the pressure due to the increasr in \-olumc~on freezing, that it causes blocking ot
the drains, and that an!- wall subjected to frost action before the mortar is dry will lose thus
greater part of its strength and 1)~ incapal)lc of resistanccl.
In spite of all these remarks, Coulomb was still of the opinion that it was safe in an!-
soil to construct a wall on a batter of I 6, with tlto thickness at the top equal to l/7 of the.
height.
In section 15 Couloml) set out to tind the surface uf greatest prcssurt, whether curvetl
or plane. After obtaining a very long formula containing a large number of coefficients,
in a manner which anticipated in a remarkable way the development uf the method of
slices for the circular arc method of analysis of the stability of earth slopes, he cancludetl
that the development of these remarks and the application of the preceding formula required
an especial piece of work which would take him too far from the simplicity which he had
prescribed for himself in the memoire. He hoped, however, to be able to treat this question
some other time, as it depended in part on the principles which he had just explained and as
there were still several very curious problems to 1~ solved.
After this delightful nnderstatement, Coulomb left the problem of earth pressure and
turned his attention to the stability of the arch.

Downloaded by [ UNIVERSITY OF SHEFFIELD] on [16/09/16]. Copyright ICE Publishing, all rights reserved.
 COULOMU AND EARTH PRESSURE 71

Summarizing Coulombs contribution then :--

1. He stated the law of shear resistance for soils
s=c+n tan 4.
2. He stressed the importance of testing the material dealt with.
3. He stated the formula for the pressure on a vertical wall for soils having both friction
and cohesion
. PA _f& z&2 ,/&1_1sln
1 fszn + cos 4
4. He showed that the plane of failure is inclined at an angle of 45+#2 to the horizontal.
5. He introduced the effect of w-all friction.
6. He stated the critical height for a vertical bank of earth
H =A! l+sind
c
W cos +

7. He stressed the importance of drainage!

8. He introduced the conception of limits, i.e., active and passive earth pressure.
9. He assfcnled that the surface of failure would be a plane, but stated that it might be
curved.
10. He did not assume that the tangent of the angle of repose was always equal to the
coefficient of friction of the soil, although he used a particular case in which this was so,
as an example.
11. He did not prove that pressure distribution must be hydrostatic, although an un-
fortunate error masks this point.
This, then, is Coulombs contribution to the problem of earth pressure.
This essay, together with the work of Franqais and Collin, laid down ali the basic ideas
necessary for stability problems in all soils, including those with cohesion. Yet after this
period many retrograde steps were taken. Cohesion was ignored, no tests as good
as those of Collin were carried out, the curved surface of rupture was forgotten, and the
false . angle of repose assumption reigned supreme. A vast structure of elaborate theo-
retical deductions and elegant but useless constructions was built up on this false foundation,
both in France and England, until the increasing number of large scale failures of earthworks,
in the second decade of the present century, forced engineers once again to look at and study
the soil they were using. This work of Coulombs, even though written in 1773, is more
important than anything written from 1850 to 1910.
. Coulombs essay concluded with the charming little engraving reproduced on the cover
of this journal.

13IBLIOGRAPHY
BELL, A. L. (1915). The Lateral Pressure and Resistance of Clay and the Supporting Power of Clay
Foundations. Min. Proc. Inst. Civ. Eng., cxcix.
COLLIN, A. (1846). Recherches Experimentales sur les Glissements Spontan& des Terrains Argileux.
P&S.
COULOMB, C. A. (1776). Essai sur une Application des R&gles de Maximis et Minimis & quelques
Problemes de Statique, relatifs & 1Architecture. Mem. Div. Sav. Acad&nie des Sciences.
CULMANN, C. (1866). Graph&he Statik. ~tivich.
FRANFAIS. (1820). Recherches sur la Pou&e des Terres sur la Forme et les Dimensions des Murs de
Revetment et sur les Talus dExcavation. . Paris.
HAMILTON,S. B. (1936). Charles Augustin Coulomb. Trans. Newcomen Society, Vol. XVII. London.
MAYNIEL, K. (1808). Trait6 analytique et experimental de la Pow&e des Terres. Paris.
PJZRRONET, J. R. (1788). Oeuvres de Perronet-Fran$oisAmbroise Didot.
RANKINE, W. J. M. (1862). Civil Engineering. Griffin, London.
SKEYFTON, A. W. (1946). Alexander Collin 1808-90, A Pioneer in Soil Mechanics. Transactions;
Newcomen Society, 1946. Not yet published.
WOLTMANN, R. (1794). Hydraulische Architectur, vol. III.
YOUNG, T. (1855). Life of Coulomb. In Miscellaneous Works, vol. 11. London.

Downloaded by [ UNIVERSITY OF SHEFFIELD] on [16/09/16]. Copyright ICE Publishing, all rights reserved.