Effect of Boiling Water Near an Electronic
Laboratory Balance
Frederic N. Rounds
August 31, 2017
Abstract
A hundred-gram test weight is placed on a .1mg
electronic balance. A series of baseline weighings
are taken followed by series of weighings with a pot
of boiling water situated above the balance. A statistical
comparison is made between the no-boiling state and the
boiling state. A significant difference is found at greater-
than the .1mg level.
I. Introduction
The first question is how to produce a very high frequency mechanical—as opposed to electro-magnetic—oscillator. Candidates for such an oscillator would be substances in phase transition: for example, solids to liquids and liquids to gases. Water is a good example. The phase transition from liquid to gas is extremely fast. A idea for the order of magnitude of the transition speed can be obtained from the following formula from Thermodynamics:1
Where hfg - enthalpy of vaporization. Water is 2256 Kjoules.
Let Power (P) = Q/t. Hence, Pt = 
and, therefore, t = /P. So if:
and, therefore, t = /P. So if:
P = 1100 Watts, t = 2256/1.100 = 2050 seconds/Kgram of water, or about 2 seconds per gram. 36 grams equals 1 mole of water, or 6.023 X 1023 molecules. Hence, each molecule vaporizes at an order of magnitude of 10-22 seconds. Therefore, water boiling into an open environment produces a multi-particle system of very high frequency phase changes
The second question is that if high frequency mechanical oscillations can be generated, what effects would it have on the environment, such as weight (force on a mass in a gravitational field). Would such effects be measurable in a laboratory? The current analysis of scientific literature provides no specific experimental data regarding the interaction of boiling water with the surrounding physical conditions other than thermal effects, such as humidity and temperature. Time, mass, and standard physical constants do not appear impacted by boiling water. To think along such lines implies a considerable unreality factor. Fortunately, a test for boiling water effects can be done in a laboratory with an electronic balance with at least a .1mg sensitivity. Other necessary components can be fabricated with relative ease. Of course, the objective in creating the experimental environment is to eliminate problematic influences, such as acoustic noise, temperature and humidity variations, structural vibrations, electrical fluctuations, and possibly structural couplings that may exist between the water and weighing systems.
II. Experimental Design
 |
 |
Figure 1. Schematic Diagram of the Testing Apparatus. This configuration is the author’s design. The balance is sensitive to .1mg and it has an RS232 interface, USB, or other computer readable output. The dimensions are not critical. They box should be designed to accommodate the size of the electronic balance.
Figure (2) shows photographs of the test apparatus. The left is with the door open. Note the sound proofing on all surfaces except the bottom. In the right photo the door is closed and the stove and boiler are suspended from a shelf above the balance’s housing. During operation the door is closed and sealed using masking tape.
Figure (1) shows the schematic experimental configuration and equipment. The design objective was to create two isolated systems: The water boiler and the measuring equipment. The two systems should not touch each other and no channel should exist between these systems to transmit vibration. Boiling water does vibrate and make noise that can cause air currents.
To alleviate as much inter-system communications as possible without using elaborate vacuum equipment, the stove and associated boiler pot are suspended from the ceiling using thin steel wires. The wires are attached to the ceiling using a single point hanger, such as an eye bolt or hook.
The electronic balance is enclosed in a plywood box. I used 5/8-inch plywood to make this box, all seems are sealed with wood glue, and the entire interior is lined with sound proofing2, including the hinged door. The only access would be for power and data cabling. I drilled a one-inch hole in the back of the box, inserted the necessary cabling, then taped over the cabling to provide an airtight fitting. The only part of the box that is not sound-proofed is the bottom upon which the balance sits. A sensitive electronic balance needs a hard support surface. I used the 5/8-inch bottom sitting upon a ¾-inch, painted, plywood work bench which is secured to the wall. Of course, a heavy granite or ceramic table would be preferable to wood. The door to the box must close tightly and create a good seal. I used masking tape to close the door and seal around all seams.
The balance is described in the footnotes.3 It is sensitive to .1mg. It has an RS232 interface which connects through a null-modem cable to a lap top computer running Windows XP. The balance can be programmed to operate at 1200 bps at continuous data flow of approximately 10 samples per second.
In order to capture the balance data, the computer needs data acquisition software, such as BC-Wedge8. This software will gather the data and deliver it to files or to an Excel spreadsheet. I recommend data acquisition software that can remove non-numeric characters from the balance’s output before arriving at the spreadsheet.
The stove is simply an off-the-shelf, 1100 Watt, signal burner. I attached the hanger wire to the bottom of the stove, so that it can be suspended from the ceiling. The boiler is made of 20-gauge titanium. It is 14 cm in diameter and 7 cm in height. I obtained the pot from a camping mess kit. The heat coils completely cover the boiler.
The spacing and dimensions of all components are shown in Figure (1). Probably variations on these configurations are possible, but I have not tried a wide variety of configurations at time of this paper.
Figure (2) are photographs of my assembled test system.
Also, before beginning the actual experiment, understanding how the electronic balance behaves in successive measurement cycles is critical in trying to verify experimental significance. I checked this type of balance for linearity using the following protocol:
1. Calibrate the balance with 100g test mass.
2. Close and seal the box.
3. Collect 100 samples to spreadsheet
4. Stop the data logger for 5 seconds
5. Re-start the data logger
6. Collect another 100 samples to another column on spreadsheet
7. Stop experiment
8. Continue to next replication
The author replicated this protocol 9 times to create 9 pairs of readings. The difference between each member in the pair is used as a measure of balance’s linear stability. The results of the measures are shown below.
The testing protocol for each experiment is as follows:
1. Calibrate balance with a 100g test mass.
2. Leave test mass on balance pan.
3. Close and seal the box.
4. Collect 500 or more samples of weight onto spreadsheet.
5. Stop collection.
6. Add 100ml of water to pot
7. Turn stove to highest power.
8. Weight until full boil occurs
9. Start data collection again for another 500 or more samples. These samples need to go into another column on the spreadsheet.
10. Continue to next replication
This protocol should be replicated at least 10 times, but the author did 22 replications. Each replication should follow the same protocol, including re-calibration and re-sealing the box. Therefore, to the best possible means, each replication should be independent from any previous experiment run. Each replication should, therefore, have a baseline run with no-boiling, followed by a boiling run. These experimental pairs will be used to test the difference between the no-boiling and boiling states.
III. Results
The electronic balance’s stability was checked for linearity using a calibrated 100-gram weight. Nine (9) replications were performed. Each run consisted of 100 samples followed by turning the data logger off for approximately 5 seconds. Then the logger was turned on again to gather another 100 samples. Ultimately, I obtained 9 on-off-on pairs. I took the difference between the samples within each pair. Finally, I computed the mean of the differences followed by the mean of the means. The final mean was .0001. The Stability Test Chart below shows the distribution of the sample means for all replications. The linearity test appeared to fall within manufacturer specifications.3 The maximum variation between the samples in the paired set was .3mg. Therefore, I used .3mg as the significance boundary. To qualify as a significant test the means of the actual results needed to be greater than .3mg (.0003 grams).
Stability Test Chart. Nine replications were done to check the linearity of the balance. The average of the means for the sample runs was .1mg. The linearity falls into the advertised specification of .2mg [see footnote].
Results Table. Twenty-two (22) replications were done using a 100-gram calibrated weight. Each replication consisted of a baseline run followed by a run with water boiling above the electronic balance. The difference is taken within each baseline-boiling pair. The mean of the differences are computed for each replication pair.
The Results Table is shown above. I summarize the results as follows:
1. Seventeen (17) out of the twenty-two (22) runs met the .3 mg mean significance level.
2. Replications 11, 13, and 16 failed to meet significance because the means were below .3mg.
3. Replication 14 had a .3mg mean, but the z-test showed the null hypothesis to be true. The number of values less than .3mg are greater than the number greater than .3mg.
4. I marked replication 3 as borderline (BL) because even though the mean is .0002 the z-test disproved the null hypothesis. .0002 is below the cutoff of .3mg, but it is greater than or equal to the linearity of the balance.
5. Replications 1, 6, 10, 12, and 19 showed lighter weights under boiling as opposed to non-boiling conditions.
6. In replications 2, 4, 5, 7, 8, 9, 15, 17, 18, 20, 21, 22 appeared heavier during boiling than non-boiling.
7. Each replication is tested with a single tail z-test. The probability (p-value) must be 0 to four decimal places to be considered as disproving the null hypotheses. The alpha value is also considered 0 to four decimal places. Since the z-test is single tail, any negative values that appeared in the data were not tested, except in cases in which the data was heavily skewed in the negative in which case the negative values became the tail for testing.
IV. Conclusion
The results rule against the null hypothesis that no difference exists between non-boiling and boiling states. Further study is warranted, therefore, the test conditions need to be refined to eliminate sources of error. For example, the electronic balance should be replaced with a gravimeter9 with accuracy to the milligal level. The boiler and measurement system should be further isolated by enclosing the measuring equipment in a vacuum. The entire experiment should be enclosed in a noise-free environment that is not subject to vibration.
At this point I do not see much value in speculating as to the cause of these weight effects. In experiments of years past4,5,7 I was guilty of proposing a number of reasons for gravitational anomalies that occurred during laboratory experiments. However, as we often experience, these experiments could not be replicated.6 Also the correlation of separated events, such as boiling and weighing, does not imply causation.
Though not reported definitively in this paper, I have tried to rule out boiling water’s vibration and noise. I made an audio recording of the boiling water. I played this recording four (4) times. I did not experience variations in the data at the .3mg level. I located the speaker in the same location as the boiler. Similarly, I tried to eliminate external vibration and noise using insulation as described, but also in the environment. I performed measurements with and without room air conditioning. I could detect no variation due to these external noise effects.
Humidity can affect the behavior of and electronic balance. I didn’t take strict precautions against changes in humidity other than sealing the balance in the insulated box and the process of re-calibrating and re-sealing between each replication.
I also did two (2) replications using alcohol instead of water. The weight variations occurred in both cases and—strangely—when the alcohol completely evaporated the weight measures returned to baseline values (non-boiling weights). Further testing needs to be done with other materials and I hope to report this research later.
A great deal of work is underway now in high frequency laser pulses at the attosecond (10-18 seconds ) range.10 These pulses are able to visualize chemical interactions at the outer electronic layers. I imagine the goal would be to shorten these pulses further to allow access to nuclear and sub-nuclear levels.
Footnotes
1Thermodynamics DeMystified. Merle C. Potter. McGraw Hill. New York. P. 74. 2009
2MuteX Sound Proofing. Model No. MUS268051. ¼ inch thick.
3https://www.joyfay.com/media/import/lab/JFDBS00005.pdf. Electronic Balance User’s Manual.
MXBAOHENG is supplier to Amazon. https://www.amazon.com/0-0001g-Digital-Analytical-Balance-Precison/dp/B0156E01QW/ref=sr_1_27?ie=UTF8&qid=1505686976&sr=8-27&keywords=MXBAOHENG+balance
Characteristics :
1. Electromagnetic Force Compensation Technology
2. Aluminum Alloy die cast base and ss platter.
3. Super Bright LCD display with backlight.
4. RS232C Interface
5. Clear Glass Windshield as standard.
6. Over load protection, Bubble Level Adjustment, Under hook
7. Full capacity Substraction
8. Multi Weighing unit conversation g/mg/ct/oz.
9. Stabilization Time, 2-3 seconds.
10. Check Weighing, Parts counting, percentage weighing
11. Universal power adaptor supplied as standard
12. Certificated calibration weight supplied as standard.
13. Internal calibration or external calibration for your choice.
14. Ce approval.
Parameters:
Model JF1204
Capacity 0-120g
Readability 0.1mg
Repeatability ±0.2mg
Linearity ±0.2mg
Output Interface RS232C
Pan Size Ø80mm
Dimensions(L*W*H) 34cm×21.5cm×35cm
Packing Size(L*W*H) 48.5cm×33cm×48cm
Gross Weight 8600g
Power Supply 110V 60HZ/220V 50HZ
1. Electromagnetic Force Compensation Technology
2. Aluminum Alloy die cast base and ss platter.
3. Super Bright LCD display with backlight.
4. RS232C Interface
5. Clear Glass Windshield as standard.
6. Over load protection, Bubble Level Adjustment, Under hook
7. Full capacity Substraction
8. Multi Weighing unit conversation g/mg/ct/oz.
9. Stabilization Time, 2-3 seconds.
10. Check Weighing, Parts counting, percentage weighing
11. Universal power adaptor supplied as standard
12. Certificated calibration weight supplied as standard.
13. Internal calibration or external calibration for your choice.
14. Ce approval.
Parameters:
Model JF1204
Capacity 0-120g
Readability 0.1mg
Repeatability ±0.2mg
Linearity ±0.2mg
Output Interface RS232C
Pan Size Ø80mm
Dimensions(L*W*H) 34cm×21.5cm×35cm
Packing Size(L*W*H) 48.5cm×33cm×48cm
Gross Weight 8600g
Power Supply 110V 60HZ/220V 50HZ
4Rounds, Frederic N. (1998). "Anomalous Weight Behavior in YBa2Cu3O7 Compounds at Low Temperature" (PDF). Proc. NASA Breakthrough Propulsion Phys. Workshop. And https://arxiv.org/pdf/physics/9705043.pdf
5https://www.wired.com/1998/03/antigravity/
6N. Li; D. Noever; T. Robertson; R. Koczor; et al. (August 1997). "Static Test for a Gravitational Force Coupled to Type II YBCO Superconductors". Physica C. 281 (2-3): 260–267. Bibcode:1997PhyC..281..260L. doi:10.1016/S0921-4534(97)01462-7.
7Podkletnov, E.; Nieminen, R. (1992). "A possibility of gravitational force shielding by bulk YBa2Cu3O7−x superconductor". Physica C: Superconductivity. 203 (3–4): 441–444. Bibcode:1992PhyC..203..441P. doi:10.1016/0921-4534(92)90055-H
8BC-Wedge Software. Downloadable from www.taltech.com. Records data from RS-232 port on the balance. The data can be sent directly to an Excel spreadsheet.
9http://scintrexltd.com/wp-content/uploads/2017/02/Guide-High-Precision-Land-Gravimeter-Surveys.pdf
10https://phys.org/news/2010-05-attoseconds-world-shortest.html. Attosecond laser pulses can detect behavior of outer electrons.
