Professional Documents
Culture Documents
This Section excerpts only the most important parts, which every student must first
learn before going to the full Appendix mentioned above.
y = kx (51.1)
where k is the constant of proportionality between y and x. x being in the first degree
(power), the above will plot as a straight line, as seen in Fig.51.1.
51.2 Slope
k, which is the constant of proportionality, is in other words the slope of this line
which is the same at any point on the line (see Fig.51.2 in which the slope is plotted
against x). The dimension of k is the dimension of y divided by the dimension of x.
Slope is not the tangent of the angle , i.e. bc/ab, which students normally tend to
assume. If it were so, it would change with the scales to which y and x are plotted,
and besides, it would have been dimensionless irrespective of the dimensions of y and
x. However, if by tangent what is meant is the quantity on the y-axis represented by
the distance bc divided by the quantity on the x-axis represented by the distanceab,
then it correctly defines the slope k.
y = kx2(51.2)
which is a parabola (Fig.51.3), the slope of the curve at any point is the slope of the
tangent to the curve at that point, taken in the same manner as above. In this respect,
2
slope is the rate of change of y with respect to x which in this case is equal to 2kx.,
which linearly varies (directly proportional) with x(see Fig.51.4 in which the slope is
plotted against x)
Note that in a physical problem, if the variation were of this kind, (dy/dx) would be
called the tangent modulus, and (y/x) the secant modulus (Fig.51.5) at a point (x,y)
specified on the curve. A secant is the line joining the origin to a given point on a curve,
and (y/x) is the slope of such a line and hence the name secant modulus. The slope
of the tangent at the origin of the curve is called the initial tangent modulus. (Note that
the terms tangent and secant have both geometrical and trignometrical meanings.)
1
y
y=
xy = k (51.3)
Since x and y appear in the product, the relationship between them is not linear.
The variation of y with x is shown in Fig.51.6. The curve following Eq.(51.3) is a
rectangular hyperbola. In other words, while in direct proportion y increase linearly
with x, in inverse proportion y decreases hyperbolically with x. What is however
important to note in respect of inverse proportion is that the decrease in y with increase
in x is very fast in the initial ranges of values of x and very slow in the final ranges of
value of x.
x+y=k (51.4)
It may be indicated at this stage that xy = k will plot as a straight line (like Fig.51.7) in
logarithmic scales (Sec.51.8.1).
Fig.51.8 is plotted with three different indices of x (1,0 and -1), in which point (1,1)
is seen to be common to the three plots.
51.4 Logarithm
Logarithm of a quantity to any base is the index (or power) by which the base is to
be raised to obtain the quantity.
3
logab = 3
This means,
a3 = b (51.5)
In relation to the base and the logarithm, the quantity itself is called the antilogarithm.
In the above b is the antilogarithm, 3, the logarithm and a, the base.
If a has the dimension [L], b will have the dimension [L3]. Thus while the
antilogarithm, i.e., the quantity itself, and the base have dimensions, the logarithm (3)
has no dimension, it being merely the index of a. We further note that the dimensions
of the antilogarithm and the base are consistent with the value of the logarithm.
It is obvious on examining the logarithmic scale that the same distance or interval
represents increasingly smaller quantities as we go to the left and increasingly larger
4
quantities as we go to the right, from the middle. In other words, we are introducing a
wilfuldistortion by magnifyingsmaller differences and shrinking larger differences in
the representation of quantities on the logarithmic scale, which is necessary in
situations such as plotting the particle size in the grain size distribution curve, in order
to give significant representation to smaller particle sizes, even at the expense of
larger sizes, which in the absence of such distortion would have received very
insignificant representation, if plotted to arithmetic scale.
y = xa(51.6)
or y = x a(51.7)
(Fig.51.11), instead of fitting a curve through the points, we can obtain the value of the
power a by plotting a straight line through the points, provided both y and x are plotted
to log scales (i.e., their logarithms plotted to arithmetic scales). This follows from the
fact that if we take logarithm,
which means that log y will plot as a straight line against log x to the slopea (Fig.51.12).
In other words, the power a is obtained as the slope of the straight line fitted on a log-
log plot. (In the above, we are actually exploiting the convenience of fitting a straight
line by entering the points on the distorted log-log scales.)
xy = k (51.3)
we have a means of fitting a straight line provided the points are plotted on log-log
scales (Fig.51.13), in which the antilogarithm of either of the intercepts gives us the
unknown k.
y = ax(51.10)
or y = a-x(51.11)
Taking logarithm,
which means that log y will plot as astraight line against x to the slope loga (Fig.51.15).
Such a plot is called a semi-log plot, in which only one quantity is plotted to logarithmic
scale, the other quantity being plotted to arithmetic scale. The basea, which is the
unknown in the present case, is obtained as the antilogarithm of the slope of the
straight line fitted on the semi-log plot.
Apart from the uses of these scales, note that, in y = kx2 (Eq.51.2), if x is plotted to
the square scale, we get a straight line whose slope is k. The same applies to y = kx
and y = kx-1 (xy = k rectangular hyperbola) plotted respectively to square root and
where a and b are constants, is assumed in many problems in Soil Mechanics. In the
first place, the shape of this curve (Fig.51.23) follows by multiplying y =x and y
1
= (Fig.51.22). In order to determine the constantsa and b, the ultimate value of y
+
6
and the initial tangent modulus (Fig.51.23), one can manipulate the equation in the
following manner.
= a + bx (51.14)
If we plot the data (x/y) against x we get the straight line shown in Fig.51.24. It is seen
that a is the y-interceptand b is the slope of this straight line.
lim
( )
0
1 lim
from which = ( + ) =
1
Hence, =
The above shows that y ult is obtained as the inverse of the slope of the straight line.
1
=
+
1
At x = 0 = (51.16)
where (y/x) is the initial tangent modulus, which is obtained as the inverse of the y-
intercept a of the straight line in Fig.51.24.
(Note that (y/x) is the secant modulus. It becomes the initial tangent modulus, since
the secant coincides with the tangent at the origin.)
Eq.(51.15) reveals another possibility of plotting a straight line using inverse scales,
as in Fig.51.25 where a is the slope and b, the y-intercept. However, between
Figs.51.24 and 25, one would prefer the former because of the greater ambiguity in
plotting Fig.51.25 thanks to the fast convergence of the inverse scales.
7
2
However,M = - 2
2
2
In the above, elastic supports can take the place of springs. If were the criterion
of design all the three systems would be equally acceptable; on the other hand, if Y
were the criterion, the last two, if not all the three, might not have been acceptable.
Note: App. E cited (Kurian, 2005) presents many examples from geotechnical and
foundation engineering using the above concepts and scales.