Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

2.

5 The measurement of gz

• The pendulum (absolute)


• Weight drop (absolute)
• Spring-mass (relative)

The gravitational acceleration can be measured absolutely or relative


changes with respect to a base station can be measured. The accuracy of the
former tends to be less than the precision of the latter. Usually a map of the
Bouguer anomalies in some area defines the geologic target and the actual
values of gz are not needed for the interpretation. For larger scale surveys or
to ‘connect’ several different surveys it may be necessary to establish absolute
g values for the base stations. Two common absolute gravimeters are the
pendulum and the weight drop.

The pendulum

For small θ the force balance


equation is:
F = m g sinθ ≈ mg x/L
= m ∂2x/∂t2

So we get the equation:


∂2x/∂t2 = g x/L
the solution for which is:
x = A sin (2πt/T) , where T = 2π( L/g ).5 is the period of the pendulum
So, g = 4π2 L/ T2
and the sensitivity for measuring g comes from:
dg/dt = -8π2 L/ T3 = -2g/T
∴ dg/g = -2 dT/T

A resolution of 0.1 mgal, 10-4/103, corresponds to dT/T of 5 in 108. If T


was very large, like 100 seconds, then T would have to be measured to an
accuracy of 10-6 seconds. This is just possible but to make a few
measurements of T to get some measure of the accuracy would take some
time. Worse, the length of the pendulum would probably be impractical to
get such a long period. Despite these problems intricate pendulums are used
which have an absolute accuracy of 0.1 mgal.

The weight drop

For free fall in a vacuum, the distance,


∆z, traversed in time t is :
∆z = 0.5 g t2
from which g is obtained via:
g = 2 ∆z / t2

Again the sensitivity to g comes from:


dg/dt = -2g/t
or dg/g = -2 dt/t

This now presents the same problem as the pendulum: only if ∆z is big
will t be big enough so that it can be measured to one part in 108. With clever
laser interferometers and accurate clocks it is possible to build almost portable
weight drop absolute gravimeters with an accuracy of 0.1 mgals.

Spring-mass systems: relative g measurement

The spring force = kz


the weight = mg

At equilibrium then, z = mg/k


and so g = k dz/m and as g changes
from g to g+∆g, z changes to z+∆z then
∆g /g = ∆z /z

Again z must be big to get sensitivity in g. An interesting way to look


at this problem is to note that if the spring is long, the period of the system,
considered as an oscillator will also be big. The period of such a mass- spring
oscillator is:
T = 2π(m/k)1/2 or m/k = T2 / 4π2

and if this is substituted into 2.5.1 above we get:


∆g = 4π2 ∆z / T2
which looks a bit like the pendulum expression except it’s relating ∆z to the
change in g. This expression still relates ∆g to ∆z but in terms of the period
of the system rather than the spring constant and mass. In such a system high
sensitivity means a large value of ∆z/∆g which means that T must be big. For
the simple hanging mass system shown above this implies that m has to be big
and k small. This is difficult to achieve in a practical portable system.

Lucian Lacoste in 1934 invented a spring-mass system that gave a huge


T in a small package.

He found that he could arrange this configuration to have a very long period.
In force balance terms it means that in a given g field the system is in
equilibrium over a range of θ values. If, for example, both positions shown
are stable then a change in g will lead to a large change in θ.
An excellent derivation of this is presented in Telford et al. If the force
in the spring is k(s - L0), which basically means that the spring cannot contract
to zero length, then it may be shown that:
∆g = k’(L0/s) ∆s
where k’ = (k/m) (a/b) (a/s) (all of which are fixed properties of the design)
So we can get any ∆s we want by juggling the fixed quantities,
especially L0. Lacoste invented a spring with a tiny L0 - a close to ‘zero
length’ spring. He achieved the desired result of a big change in ∆s for a
small change in g in a small package. The working sensitivity of a Lacoste-
Romberg gravimeter is ± 0.01 mgal and there are some specialized units that
reach ± 0.003 mgal (the D meter ).

You might also like