Theory,
Spreadsheet and
References
Latest change 2025-07-05
In brief:
What I found on the internet and a few formulas I derived myself.
References to the spreadsheet I developed during my search.
Download the spreadsheet in Zip format. It can be opened in Excel, OpenOffice, LibreOffice, etc.
The fields with formulas are bold, for easy recognition.
The lines with green text give some differences or ratios between results which should differ only slightly. If you see large numbers there then most likely there is something wrong with the formulas or the numbers.
On this page you find an article from me about some aspects of the Period Time of a Foucault Pendulum.
Contents:
Terminology
In the order of the spreadsheet:
Constants
Pendulum Swing Times
Foucault Precession
Intrinsic Precession
Drive timing according to the Schumacher equation
Reverse Calculation of Q from Schumacher Equation (19)
Quality factor of the pendulum
Passage times for a certain radius
Radius of Passage at a certain time
Sensitivity of the Floor Unit adjustment
Energy in the system
Forces on the cable
Forces on the mounting point
Thermal expansion of the cable
Height of tip of bob at some distance from the center
Height from top for a given excursion of the cable
Dimensions and weight of a cylindrical Bob
Dimensions and parameters of the Coils
References
Articles and sites about theory and practice of several pendulums
Terminology.
In the literature about Foucault Pendulums often the following naming is used for the several parameters. I'll stick to this terminology as good as possible.
g for the accelleration of gravity on earth [9.81 m/s2] (can locally differ up to a few %, also dependent on altitude)
φ (fi) for the northern or southern latitude of the location of the pendulum. Southern latitude is sometimes given as negative.[degrees or radians]
L for the length of the pendulum, from the deflection point at the top to the center of mass of the bob. [m]
M for the mass of the bob.[kg] The mass of the cable is mostly neglected.
T for the period of the pendulum. [sec]
ω (omega) for the actual frequency of the pendulum. [rad/sec]
ω0 (omega-null) for the frequency at very small amplitude [rad/sec]
a for the amplitude in the desired direction, the major axis of the ellipse. [m]
b for the amplitude of the minor axis of the ellipse. [m]
ΩF (Omega-F) or FP for the Foucault Precession. [rad/sec]
ΩI (Omega-I) Ωe for the intrinsic precession due to the elliptical path. [rad/sec]
ΩA (Omega-A) for the siderical rotation of the earth.[rad/sec]
Θ (Theta) for the angle at maximal excursion.[rad]
Q for the quality factor of the pendulum [dimensionless]
E the energy in the system systeem [Joule] (alternates beteween potential energy and kinetic energy)
In the order of the spreadsheet
Be aware that some dimensions are given as projected on a horizontal plane below the bob.
Bringing them to the level of the bob's Center of Mass (COM) will change certain outcome. I've tried to take that into account.
Constants
The gravitational constant can be slightly different at your location.
For my location (vicinity of Arnhem, NL) I found 9.8123. source: https://upload.wikimedia.org/wikipedia/commons/c/ce/Valversnelling_in_Nederland.svg
The siderical day is defined as the rotation of the earth with respect to the far "fixed" stars. It is slightly shorter than the well known 24-hours day, because that one is related to the earth's rotation w.r.t. the sun. The earth also rotates around the sun in one year and so the difference is about 1/365 of a day.
Pendulum swing times
The exact period of a pendulum is hard to calculate because it also depends on the amplitude. Most formulas are approximations..
The most well known is:
T = 2 π √ (L/g) [sec] (1)
This is an approximation for very small angles. The error comes from the assumption that sin(Θ) = Θ, which is'nt.
I wrote a small article about a consequence of this difference. (2)
An amount of ellipticity also increases the swing time, but that effect seems to be very small. (Schumacher)
The theoretically exact solution with a lot of terms having a regular pattern: (3)
T = 2 π √ (L/g) * [ 1+ (1/2)2 sin2(Θ0/2) + ((1*3) / (2*4))2 sin4(Θ0/2 ) + ((1*3*5) / (2*4*6))2sin6(Θ0/2 ) + ((1*3*5*7) / (2*4*6*8))2 sin8(Θ0/2 ) + ...] [sec]
where Θ0 = arcsin (a/L)
Supposedly exact but converges with fewer terms which follow a (for me) unknown pattern: (4)
T = 2 π √ (L/g) * [ 1+ 1/16 Θ02 + 11 / 3072 Θ04 + 173 / 737280 Θ06 + 22931 / 1321205760 Θ08 + ...] [sec]
where Θ0 = arcsin (a/L)
An approximation which gives reasonal results without an infinite series of terms: (5)
T = 2 π √ (L/g) *[cos(θ/2)]^-{0.5*[cos(θ/2)]0.125}
The formula derived by Lima and Arun:
T = -2 π √ (L/g) * [ ln(a) / (1-a) ] [sec] where a = cos (Θ0/ 2) (6)
For the time being I use formula (2) , expanded over 3 terms.
As can be seen the period time T is slighly dependent on the amplitude. Read about possible consequences.
The Foucault Precession is
ΩF = ΩA / sin φ / [1 - (3/8 * a2 / L2)] [rad/sec] For NL at 52° that gives about 30.4 hours for 1 rotation. (7)
For small excursions the part between [ ] can be omitted, the difference is generally a few minutes. (8)
In my calculations I use the complete form.
Intrinsic Precession of the Ellipse
Is given by
Ωe = 3/8 * ϖ0 * ab / L2 [rad/sec] (9)
where ϖ0 = 2 * π / T [rad/sec]
Note: in my pendulum the value of b changes periodically amplitude and polarity.
Drive timing according to the Schumacher equation
Schumacher (download) describes in his article that if the drive pulse is given on a very special moment during the swing, the precession of the ellipse (not the ellipse itself) is perfectly suppressed. It turns out that his equation [19 in the article] cannot be solved analytically, the only way is to find an optimum by iteration (trial and error)
Change "Drive Position" such that the "Ratio Left / Right" becomes 1.0 as good as possible. "Drive Position" in samples after the zero crossing then gives the optimal drive moment.
Note that Schumacher's formula is quite sensitive for the Q and for the amplitude of the pendulum. You have to know the Q and keep the amplitude under control.
As the Q is mainly determined by the air friction it will be sensitive to the viscosity of the air, which is dependent on air pressure, humidity and temperature.
Reverse Calculation of Q from Schumacher Equation (19)
.
It does a reverse calculation of the pendulum's Q when one has found the optimal drive position.
I rewrote Schumacher's equation (19) to have only Q at the left side.
L, a and d are the parameters to be filled in.
alpha and delta are calculated as defined in the Schmacher article.
The nummerator and denominator are calculated separately to keep the formalas simple
The Q is the quotient of these.
The Qualityfactor
The Quality factor of a resonator (also e.g. a tuning circuit in a radio receiver) is defined as the ratio between the energy in the system and the energy loss per radian of the period.
With a Foucault pendulum it is nearly impossible to calculate the Q in advance, because the losses, mainly by air friction, are difficult to calculate.
There are however some methods to measure the Q of a practical pendulum.
1/ Measure the amplitude of the pendulum, then stop driving it and count the number of periods before the amplitude has reached half of its original value. Multiply that number by 4.53 and you have the Q.
2/ The same but count until the amplitude is 37% of the original value. Multiply by pi and you have the Q.
3/ Use both methods and take the average.
I used only method 2.
It must be noted that in many cases the air friction is not a contant fraction. The decay curve therefore is not an e-power function like e-t.
Q = π * τ / T [-] (10)
where τ is the time when the pendulum has reached 37% from the begin amplitude after stopping the drive.
Q is also defined as
Q = 2 π E / Eloss [-] (11)
where Eloss is the energy loss per radian, so
Eloss = 2 π E / Q [J] (12)
and this is the energy which has to be added during each period to keep the pendulum in motion.
So per half periode half of it.
Passage times for a certain radius
This is to calculate passing a certain distance in sample-ticks after the center passage.
I assumed here that the bob is making a perfect hamonic (sine-shaped) movement. That is not the case, but the deviation is a few promilles.
Radius of Passage calculated from passage time
Fill in a time in Sample ticks
I assumed here that the bob is making a perfect hamonic (sine-shaped) movement. That is not the case, but the deviation is a few promilles.
Sensitivity of the Floor Unit adjustment
Based on 0.1 mm displacement per rotation of the knob.
Energy in the system
The energy in the system can be calculated in two ways. From the difference in height that the bob passes between the zero crossing and the maximal amplitude, and from the velocity of the bob at the zero crossing.
The kinetic energy at the lowest point, that is at the highest velocity, is
Ev = 1/2 M (2 π / T)2 [J] (13)
This will give a slightly incorrect result because it assumes the motion of the pendulum to be perfectly harmonic (sine shaped) and that is not the case, there are odd harmonics in the actual motion.
The height reached at the maximal excursion is:
h = L - √ (L2 - a2) [m] (14)
and so the potential energy there is:
Ep = g * M * h [J] (15)
There must be: Ev = Ep = E (16)
Forces on the cable
The maximal force appears at the center passage, there the centripetal force adds to the weight of the bob.
Fmax = M * g + M * v2 / L [N] (17)
where v the velocity at center passage, approximated by 2 * π * a / T
The minimal force appears at the end of the swing
Fmin = M * g * cos(Θ0) [N] (18)
where Θ0 the angle at maximum deviation: arcsin (a / L)
Forces on the mounting point
It may be interrresting to have a look at these figures to judge the stability of your mounting point. The slightest movement here will degrade the performance of your pendulum.
Thermal expansion of the cable
This can be important with long pendulums because it may change the height of the bob above te floorunit. This will influence the effect of the drive coil and the amplitude of the sensed signals. However, the thermal expansion of the building itself will have an influence too, but is rarely known.
Height of the tip of the bob at some distance from center
The height ht of the bob-tip at a distance d from the center is:
ht = Lt * (1-cos(Θ)) where Θ = arcsin (dt/Lt) and Lt the length from the top to the tip of the bob.
Height from top for a given excursion of the cable
This was to calculate the height at which the magnet for the Hall Sensors had to be placed such that it makes a certain excursion.
Dimensions and weight of a cylindrical Bob
Design your own Bob.
Dimensions and parameters of a Coil
Number of wires parallel.
For the Drive Coil for the Chapel Pendulum I had available only to thin or to thick wire, and many bobines of the thin wire. So I decided to wind it with multiple wires, each to thin. This requires some extra calculation of what the final resistance would become.
Diameter of wire
I have ignored the thickness of the insulation layer.
Fill Factor.
Here it is the ratio between the total cross section of the copper wires and the available space as defined by the coil former dimensions.
If all the space is perfectly filled with round copper wire, each side-aside you may expect a fill factor of π / 4 = 0.78. In practice you will never realize that.
If the outcome is much smaller there is room for more windings. If it is larger the windings will overflow the available space.
References:
(1) can be found in any textbook about Foucault pendulums.
(2) and (3) https://en.wikipedia.org/wiki/Pendulum_(mechanics)#Arbitrary-amplitude_period
(4) was found in a link which is now dead
(5) was found in Lima and Arun (download)
(6), (7) without the term between [ ] found in many publications about the Foucault pendulum.
The term between [ ] was found in Haringx and Suchtelen (download)
(8) -up are commonly known in kinematics and mechanics.