Lifter D.O.E. (design of experiment)
Data supplied by Willy Oscar Guns & Paul Keijzer

The most comprehensive data collection from Paul Keijzer

Please note that you need an Excel viewer to read this file.

Measuring the lifting force with mercury contacts.

Here you can see a massive lifter made out of thick copper plates. Total weight 196g.

The copper wire has a diameter of 0.1 mm, the copper plate is 1 mm thick.
Height of the copper plate = 40 mm, with a rounded edge of 5 mm.
The copper wire is separated by an air gap of 20 mm.

You can see the mercury contacts with the iron needles.
The lifter has been balanced with a nylon wire and is hung to the digital balance (FX-300 Max=310g d=0.001g).
A variable high voltage power supply unit (0-30 kV) was used. In this picture you can see the whole instruments' setup.
This picture shows a detail of the lifter with mercury contacts.


First experimental results.

U in kV F in g I in microA P in W S in m2
0 0 0 0 0.024
5 0.007 1 0.005 0.024
10 0.03 9 0.09 0.024
15 0.47 88 1.32 0.024
20 1.04 250 5 0.024
25 1.57 420 10.5 0.024
30 2.2 750 22.5 0.024


Rel. humidity = 68%
Ambient temp = 18C









Influence of the distance of the wire to the plate of a lifter element.

The lifter elements are made of copper (height 2.4 cm and 3.4 cm)

A = height of the plate
C = distance of the wire from the plate
B = current in µA
Unless stated, values above row B, are force in g.

A C C C C C A C C C C C
24 mm 60 mm 50 mm 40 mm 30 mm 20 mm 34 mm 60 mm 50 mm 40 mm 30 mm 20 mm
0 kV 0 g 0 0 0 0 0 kV 0 0 0 0 0
5 kV 0 0 0.002 0.001 0.002 5V 0 0.001 0 0.001 0.001
10 kV 0.011 0.027 0.015 0.027 0.007 10 V 0.006 0.013 0.05 0.008 0.004
15 kV 0.069 0.13 0.22 0.298 0.235 15 V 0.015 0.19 0.112 0.299 0.22
20 kV 0.24 0.3 0.48 0.56 0.612 20 V 0.026 0.37 0.35 0.59 0.547
25 kV 0.52 0.54 0.86 1.06 0.97 25 V 0.11 0.65 0.76 1.07 0.9
30 kV 0.85 0.86 1.4 1.6 1.4 30 V 0.36 1.13 1.33 1.59 1.2
B B B B B B B B B B B B
0 µA 0 0 0 0 0 0 0 0 0 0 0
5 0 0 0 0.5 0.5 5 0.5 0 0 0.5 0.5
10 1 1 1 2 1 10 1 1 1 1 4.5
15 5 6 11 19 38 15 5 7 6.5 18 30
20 19 40 30 55 135 20 17 23 27 60 140
25 60 90 75 120 300 25 46 65 80 130 330
30 130 170 165 220 560 30 90 115 240 250 520






Influence of the metal of the plate.

For related experiment go to Experiment 8.

U in kV Lead copper Al foil Al plate Average for Al
kV F in g F in g F in g F in g F in g
0 0 0 0 0 0
10 0.059 0.027 0.06 0.026 0.043
15 0.314 0.298 0.227 0.303 0.265
20 0.605 0.56 0.43 0.626 0.528
25 0.99 1.06 0.68 1.05 0.865
30 1.55 1.6 0.86 0.95 0.905


Distance 30 mm
Height 3.4 cm




Picture showing one element of the lead lifter.

Conclusion:
Lifting force for Cu, Pb is approximately the same and both higher than that for Aluminium. Lifting force for Al is the lowest, at 30 kV the lift is less for Al foil than for Al plate. Force somehow depends either on thickness or mass of metal.



Influence of the area S of the plate.

kV 0.0068 m2 0.0048 m2 0.0080 m2
0 F in g F in g F in g
5 0.001 0.002 0.002
10 0.004 0.007 0.010
15 0.22 0.235 0.157
20 0.547 0.612 0.347
25 0.9 0.97 0.523
30 1.2 1.4 0.733




Conclusion:
There is a maximum plate area above which thrust starts decreasing. This is due to the foil depth, which interacts with external accelerated ions as shown below.





Influence of wire polarity on current.



In this experiment, a lifter V1.0, with 20 mm airgap, 100 µm wire and a variable voltage supply of 10 KV, was setup. Two separate experiments were run, one with wire positive and one with wire negative, relative to the foil. In each test run, voltage was varied from zero to 10 kV, and current (Ip for positive current, In for negative current) plotted for each data point. The resultant curve shown above was plotted. Curves Ypos, Yneg show best fit curves generated by Excel.

Conclusion:
Negative voltages on the wire generate more current, more than twice as much.



Influence of wire polarity on thrust & efficiency



Legend for curves shown:
P = power input (watts)
F = Force generated (Newtons)
a = calculated alpha = kF/(id)
i = current (amps)
k = ion mobility
F/P = force to power ratio
_p/n = ratio of values positive to negative

The ratios of measured Force, efficiency, power, loss factor for air movement and current. Values measured for positive voltage were divided by values generated by a negative voltage. A positive voltage has an efficiency of about 2 times better mostly because current is 50% lower.



The efficiency is more or less independent of voltage at least in the range 7400 V to 10 kV. Just before breakdown it decreases. Just above the starting voltage it seems to be slightly higher.



The above readings from Paul Keijzer indicate that the flat part of the efficiency doesn't match Evgenij Barsukov's ion equations. It should be hyperbolic.

However, Paul did another measurement with the same lifter and 18 µm diameter wire, where it matched better. F/P is the calculated efficiency based on the measurements (just below 1 g/W ) accompanied by the theoretical hyperbola. In order to improve the efficiency I tried to decrease the air velocity by putting a piece of cardboard on top of the lifter. Now this is a very strange but maybe exciting result. F/P_C is the efficiency with the cardboard on top. The force and current were much lower than without the board, but decreasing the voltage seemed to boost the efficiency F/P skyhigh: 152 g/W at 5000 V, 0.19 g and 0.33 µA.

I considered the possibility of measurement inaccuracy in the denominator, but that would still not explain the high efficiency. For even lower voltage and power the outcome of the F/P_c ratio may not be accurate, but between 5000 V and 6000 V the values must be quite right.



Plotting the current as a function of voltage on a logarithmic scale shows, for the conditions in this setup, that the current increases with U4.21 in the region above 6 kV. Just above 9.4 V breakdown occurred. No limiting effect just underneath.





Influence of dielectric plate on efficiency.

Another similar experiment comes from Willy Guns. In this setup, a bakelite plate having a dielectric constant of 3.5 to 5, size 20 cm x 20 cm, was used. The Cu wire (0.1 mm, l=20 cm) is glued on the 1.5 mm thick bakelite plate. The wire follows the diagonal of the plate. On the other side of the plate, a copper plate is fixed just under the wire (length 20 cm and 2.4 cm high).

Height of the plate = 2.4 cm. See the photos below.

For the generation of 0.83 g, the power required at 30 kV is only 0.9 W! This gives an efficiency of 0.83g/0.9W = 0.92 g/W.

In a similar experiment with Cu plate (height 3.4 cm, length of plate = 20 cm, distance of the wire = 30 cm):

U = 30 kV, I = 250 µA and F = 1.59 g
P = 30 kV x 250 µA = 7.5 W!

This gives an efficicency of 1.59g/7.5W = 0.21 g/W.

In comparision with the above 0.92 g/W, this is very low.

U in kV F in g I in mA P in W F/P %F %P/F
10 0.03 0.25 0.0025 12.00 0.036 13.012
15 0.1 0.5 0.0075 13.33 0.120 14.458
20 0.45 4.5 0.09 5.00 0.542 5.422
25 0.73 14 0.35 2.09 0.880 2.262
30 0.83 30 0.9 0.92 1.000 1.000












Willy Guns has exciting results : at 20 kV the force is still at 54.2% of the force at 30 kV. The efficiency however is 5.422 times higher, giving an efficiency of 5 g/W!, 5 times the efficiency for the best standard lifter. Note however that this setup is not yet able to counteract the extra weight of the dielectric introduced (the bakelite plate), but is the first experiment that successfuly shows how the introduction of dielectric can improve efficiency in a drastic manner.