In brief:
On 2024-08-08 I did a few experiments to get some feeling about the influence of crosswind on the pendulum.
Conclusion: A systemic air flow, even if hardly noticable, easily kills the performance of a (short) foucault pendulum.
In the course of november this conclusion was confirmed again when the house heating was switched on. Read more......
Tip: Use a canldle or so and check near the location of your pendulum that it has a steady flame. If the flame flickers or wavers your pendulum may be in trouble.
Experiments with crosswind.
Fig 1. The first experiment with a normal household fan on the lowest speed and from some distance was complete disastrous.
Within a minute an ellipticity over 50% had developed, causing loss of Center Pass detection and Drive synchronisation.
At the start of the test the pendulum was swinging in the NW - SE direction and the fan is roughly at NE.
(In this picture North is roughly in the upper right direction. Think a line from the center of the coil set to the black conical base of a floor lamp )
Fig 2. Later on I repeated the experiment with a much smaller fan, which resulted in a stall in the N-S direction with almost no ellipticity.
I have no means of measuring such a low air velocity, but in the vicinity of the bob it was noticable with a wet hand.
Fig 3. Two times an experiment with the smaller fan.
Legend:
Yellow staircase: time of day in hour steps. bottom = 00:00.
White: amplitude of minor axis. center = 0, pos = CCW, neg = CW
Light blue: Precession angle. Scale +180 deg ... -180 deg.
The first experiment started at 17:25 when the pendulum had an amount of ellipticity CCW.
Within 30 minutes the ellipticity went CW, then CCW again and after 2 hours approached zero, as normal for this foucault angle N-S.
From 20:00 the pendulum plane stalled at N-S until the fan was switched off at 21:30.
From then on the pendulum behaved as usual, we can see that on the almost linear course of the precession and the sine shaped amplitude of the minor axis.
There is an anomaly visible in the time line (staircase) from 15:00 to 17:00. This is now known as a software bug which has been repaired.
The next day at 16:30 I repeated the experiment with the fan pointing in a slightly different direction
We see the precession go retrograde for a short time, then accellerates to the N-S direction and stalls there until the fan is switched off at 23:30.
Data source: 2024-08-08.log and 2024-08-09.log.
Experience november 2024
After the experiments with the magnetic fields were finished in october I wanted to do a series of tests with different drive timing (See the Schumacher Priniple) but there was a problem. Even with the drive timing I had used from the summer on the pendulum now ran very irregular. After several experiments I found the origin of the problem: the house heating had been switched on.
My pendulum is located under a vide, next to the stairs to the upper floor. During the cold season I never have to use the heaters upstairs, because enough warmth is going up and the temperature there is lower, but very acceptable. On the ground floor my house has floor heating, which is known to hardly give rise to airflow. Yet enough cold air fell through the vide to disturb the pendulum. I found out that that was the cause by placing a windshield around the floor unit.
Fig 4. The windshield is improvised with thin cardboard, tape and folded edges for some rigidness.
Fig 5. The performance of the pendulum before and after placing the windshield. Legend as in fig 3.
Source files: 2024-11-21 through 2024-11-25. The shield was placed on 11-23 around 14:00, with a relaunch of the bob.
The conclusion is that a systemic air flow, even if hardly noticeble, will have a serious effect on the performance of a foucault pendulum.
Such an airflow can be expected at locations with some form of airconditioning or other mechanical ventilation, but also unnoticible airdraught resulting from local temperature differences can adversely influence a foucault pendulum.