Spark Ignition of Combustible Vapor in a Plastic Bottle as a Demonstration of Rocket
Propulsion: Details Allotted to an On-line Appendix
J.R. Mattox, Department of Chemistry & Physics, Fayetteville State University, Fayetteville, NC,
[email protected]
This document provides supplemental information and video recordings for the article entitled
“Spark Ignition of Combustible Vapor in a Plastic Bottle as a Demonstration of Rocket
Propulsion”.1
The author’s institutional website for this work is here.
Video Recordings
The video recordings specified below are available on-line from the American Institute of
Physics (http://dx.doi.org/ 10.1119/1.4972317).
The numbers given below (30_1-v…) correspond to the file names for the video recordings or photograph in
the AIP archive. For each video recording, a link is provided to a page on the author’s
institutional website which will perhaps furnish more information, and perhaps eventually better
video recordings.
Video Recordings of Successful Demonstrations at FSU
A whoosh demonstration with a 3.8-L milk bottle fueled with isopropanol: 30_1-v1. Link for
more information.
A whoosh demonstration with a 19-L bottle fueled with isopropanol at 20.5°C in the FSU
Planetarium;1 splint ignition was successfully used after spark ignition failed due to the
temperature: 30_1-v2. Link for more information.
A whoosh demonstration with a 19-L bottle fueled with methanol: 30_1-v3. Link for more
information.
A 3.8-L milk bottle launched as a rocket fueled with isopropanol: 30_1-v4. Link for more
information.
A 3.8-L bottle launched as a rocket fueled with methanol: 30_1-v5. Link for more information.
A 19-L bottle launched as a rocket fueled with methanol: 30_1-v6. Link for more information.
Dependence of Fuel Choice on Temperature
Isopropanol (also called isopropyl alcohol) is preferred for these demonstrations.1 However, an
ambient temperature of 21° C or higher is required for the partial pressure of isopropanol to be at
least stoichiometric in air (at 0.045 bar). Methanol has a higher partial pressure at any specific
temperature, and may be used for this demonstration (with more vigilance) down to 19° C (where
it has a stoichiometric partial pressure of 0.123 bar). Ethanol can only be used down to 22° C
(where it has a stoichiometric partial pressure of 0.065 bar). I note that these temperatures
substantially exceed the respective flash points for these liquids because the flash point test is
done with a flame which provides more energy for ignition than the spark used for this
demonstration.
I have found that to use spark ignition at room temperature, these alcohols must not contain any
substantial amount of water. Isopropanol is sold for household use with up to 50% water – far
too much for this.
Lower temperature demonstrations are possible with a variety of hydrocarbon gases. I have
successfully launched a 3.8-L HDPE bottle with 12% of the air displaced by methane (9.5% is
stoichiometric in air); and also a 3.8-L bottle filled with 6% propane (4.0% is stoichiometric).
The gas was added by displacing the specific volume of water with the bottle inverted in a water
filled basin. The bottle was capped underwater as soon as the water had been displaced.
Remaining water was quickly drained after allowing time for the gases to mix. It is expected that
these mixtures will ignite outside in any climate. I have not tried the whoosh bottle
demonstration with these mixtures, but expect it will also work if the gas mixture is contained
prior to ignition as shown in figure 3 of the printed article.1
Diethyl Ether
I have found that diethyl ether is a very suitable fuel for “whoosh demonstrations”. It is
temperature versatile (it can be used down to -36° C, where it has a stoichiometric partial
pressure of 0.034 bar). Because of its volatility, it presents more hazard for the combustion of
accidentally pooled dense vapor than the other 3 alcohols described in this work, but it is
expected that it can be safely managed with the procedures specified1. Also, its toxicity is less
than that of both methanol and isopropanol1.
At room temperature, the fueling procedure specified for the other 3 alcohols1 would result in a
vapor mixture far too rich to ignite. Instead, the specific amount of diethyl ether required in
solution with air should be added to the demonstration bottle as a liquid, and the bottle rapidly
capped. In this case, there is no need to spread the liquid on the interior surface of the bottle, it
evaporates completely within a few seconds. As with other alcohols, air will leave the bottle
around the cap to accommodate the evaporation of the liquid fuel. For 3.8-L bottles, 0.7±0.1 ml
of diethyl ether is recommended (0.56 ml is the stoichiometric amount); for 19-L bottles, 3.0±0.2
ml is recommended.
The evaporation of diethyl ether is rapid at room temperature, but the density will not initially be
uniform. Approximately 30 minutes are required at room temperature until diffusion (with a
diffusion constant in air of 9x10-6 m2/s) results in a uniform mixture in a 3.8-L bottle. However,
mixing can be accomplished rapidly by inverting the bottle repeatedly every ~5 seconds over the
course of ~1 minute.
I have placed, ~2 ml of diethyl ether in a 3.8-L bottle, resulting in a mixture that is too rich for
spark ignition. After demonstrating this, I taped this bottle to another 3.8-liter bottle containing
initially only air. I then equalize the mixture between the bottles as described above, and then
demonstrated spark ignition of both bottles sequentially.
Moisture Problems
Even small amounts of water can suppress spark ignition using the three specified alcohols1,
either in solution in the liquid alcohol or present in the demonstration bottle before adding dry
alcohol. This results from the consequent reduction alcohol vapor concentration according to
Raoult’s law.
I find that, small amounts of water do not inhibit the ignition of diethyl ether because of its high
vapor pressure. Even though reduced by water, it is still enough for complete evaporation. Thus,
combustion products from this demonstration can be immediately purged by filling a used
demonstration bottle with water, draining thoroughly, and immediately refueling with diethyl
ether. This is not practical for the other three alcohols.1 The combustion of methane and propane
described above should also not be inhibited by a small amount of water.
Occasional Bottle Rupture
I have occasionally witnessed both 3.8-L and 19-L bottles rupture during attempted whoosh
rocket demonstrations. The rupture of these thin plastic bottles has been observed to be loud, but
is not expected to be dangerous. A video of the rupture of a 3.8-L bottle fueled with methanol
(heated to 28° C – see below) is available: 30_1-v7. Link for more information.
A video of an attempted launch of a 3.8-L bottle fueled with diethyl ether that ruptured is
available: 30_1-v8. Link for more information. In this video, an 8 cm diameter disk that broke
away from the bottle along a mold seam is apparent traveling toward the camera. A photograph
that shows, from left to right, a pristine bottle of the same type, the ruptured bottle, and the
ejected disk is available: 30_1-v14.jpg. It is expected that high time resolution video will show
that these ruptures occur while these bottles are on the launch base.
These ruptures are thought to be caused by pressure resulting from deflagration. Detonation
would be very inappropriate for these demonstrations, with a supersonic flame front resulting in
a shock that induces ignition through compressional heating. This can produce pressure in excess
of 10 atm in a fuel air mixture for a fully developed detonation. Fortunately, detonation will not
occur in these relatively small volumes using the fuels discussed above with air, especially when
ignited with a low power spark2.
The loud sound is thought to result from combustion gases flowing through a breach in the bottle
at supersonic velocities, creating a “booming” sound (similar to a popping balloon).
I found that rupture with compressed air of a 3.8-L HDPE bottle produced this same sound. This
rupture occurred at a pressure of less than 0.3 atm gauge. Thus, the observed sound is not
resulting from a detonation of the vapor mixture, rather, it is from a deflagration that results in
sufficient pressure to cause the bottle to rupture. I have also produced the same sound by gluing
the lid on a 3.8-L HDPE bottle with cyan acrylic cement, placing it on the floor, and jumping on
it to induce rupture. Substantial impulse is required to thus rupture these bottle - I suggest that
one jumps in such a way as to contact the bottle with both feet, prepared to subsequently land
with both feet on the floor following rupture, or to cope with an impulse in a random direction if
the bottle doesn’t rupture.
The rupture of these bottles when ignited from a distance by spark1 is not expected to be
dangerous because the bottles either remain intact, or if a piece breaks away it does not acquire a
substantial velocity. The rupturing 3.8-L HDPE bottles usually open along a mold seam, as the
diethyl ether induced rupture described above.
I have observed that 3.8-L bottles rupture predominantly when the bottle fits tightly into the
launch base and is thus mildly retained upon ignition. I have also observed that if rupture does
not occur, a retentive fit in the base results in higher flights. So far, I have not seen isopropanol
nor ethanol fueled bottles rupture at room temperature, even using a retentive base. I note that the
vapors of methanol and diethyl ether in stoichiometric mixture with air have higher laminar
flame velocities (0.56 m/s, and 0.47 m/s respectively at room temperature)3 than ethanol and
isopropanol (both 0.41 m/s at room temperature)3. However, I have observed combustion of
isopropanol to cause rupture of a 3.8-L bottle with the bottle opening constricted by ~80% (by
clamping the bottle opening over a male garden hose fitting with a 13 cm length of vinyl hose
attached with a 1.5 cm internal diameter – allowing insertion of the ignition cable shown in
figure 31 through the hose into the bottle).
On two occasions, I have observed 19-L bottles to rupture upon attempt to launch as a rocket
when fueled by methanol using the base shown in figure 1 which does not provide any retention
(out of my ~100 launches of 19-L bottles – most fueled by methanol). I have not yet witnessed
any bottle rupture during whoosh bottle demonstrations (with unobstructed bottles with the
opening pointed upward). However, numerous “explosions” have been informally reported
(references not currently available) while attempting to perform the whoosh demonstration with
alcohol in 19-L bottles. I expect that all of these have resulted from bottle rupture during
deflagration.
Investigation of the use of hydrogen and acetylene for fuel, and the substitution of oxygen
for air
Spark ignition (from a substantial distance) enables safe exploration beyond limits previously
specified for the whoosh demonstrations. I find that ignition of a stoichiometric mixture of
hydrogen and air, although not excessively loud, always ruptures 3.8-L HDPE milk bottles, with
multiple fragments resulting (that are not hazardous beyond ~1 m because of the small thickness
of these bottles – 0.5 mm). I have also found this to be the case for methanol vapor in oxygen in
3.8-L HDPE milk bottles.
I have also ignited by spark a stoichiometric mixture of hydrogen and oxygen in a 3.8-liter bottle
(video is available: 30_1-v9; link for more information); and a stoichiometric mixture of
acetylene and oxygen in a balloon with a diameter of 15 cm (video is available: 30_1-v10; link
for more information). This balloon was simultaneously popped and its contents ignited by a
spark initiated flame on a small piece of cotton gauze saturated with liquid methanol. (I find that
a low-energy spark on the outside of a balloon will not initiate ignition nor pop the balloon.) For
both of these stoichiometric gas mixtures, the combustion was very energetic. Accompanying
very loud booms may indicate that detonations occurred. These mixtures with oxygen are clearly
inappropriate for class room demonstration! Hydrogen and acetylene in air are also expected to
be inappropriate for “whoosh demonstrations”. In air, at small volumes, their combustion is
expected to be a deflagration, albeit vigorous; but their laminar flame speeds are too fast (3.1 m/s
and 1.7 m/s respectively at room temperature)3.
I have also demonstrated the ignition, by the method described above, of a balloon filled only
with acetylene (video is available: 30_1-v11; link for more information). A substantial amount of
soot results.
Ignition at elevated temperature
I have also experimented with heating bottles in an enclosure containing warm water prior to
ignition. This heating was done after adding the alcohol to the bottle, and before spreading it on
the bottle’s interior surface. I thus was able to spark ignite a 19-liter whoosh bottle using
consumer 91% isopropanol (with 9% water) after being in an enclosure with water at 28° C
(video is available: 30_1-v12; link for more information). It would not spark ignite at the
temperature of the room (22° C) prior to warming. After the whoosh, a flickering, multi-colored
flame persisted for an extended interval (~3 seconds).
Applications of alcohol vapor combustion
At FSU on 12/1/16, I demonstrated the deflagration of a 48 cm diameter balloon filled with pure
hydrogen. Video is available: 30_1-v13, here is the link for more information. Ignition was
obtained by spark initiated combustion of the vapor of diethyl ether (0.5 mL of liquid) in a 2.3 L
chamber (at the bottom of the white PVC “potato canon” strapped to the ladder in the video).
Rather than a potato (which would be potentially hazardous as it fell), a wad of cloth (~200 cm2)
was placed in the barrel above this chamber. A hardwood twig, ~1 cm in diameter and ~10 cm
long and pointed at the top, was placed above the wad to pop the balloon just before burning
gases reached it. This twig is apparent in this video falling to the ground ~2 seconds after
ignition.
Video recordings with higher time and spatial resolution
Video recording of some of these demonstrations with higher time and spatial resolution is
planned. Edited segments are expected to be posted at the author’s institutional website for this
work which is here. Additional combustion demonstrations that may be developed in the future,
and may also be posted at this site.
Collaboration in video analysis
The author is willing to provide video files of these demonstrations for analysis by interested
potential collaborators.
Provision of launch bases for whoosh rocket demonstrations
The author is investigating the possible manufacture and marketing of whoosh rocket launch
bases for 3.8-L milk bottles, similar to that shown in figure 2 of the printed paper1. Inquiries by
e-mail are welcome.
References
1. Mattox, J.R., “Spark Ignition of Combustible Vapor in a Plastic Bottle as a Demonstration
of Rocket Propulsion,” The Physics Teacher, pp 33-36, Vol. 55, January 2017
2. Turns, S.R., “Introduction to Combustion: Concepts & Applications,” 3rd Edition, McGraw
Hill Education, India (2012)
3. “SFPE Handbook of Fire Prevention Engineering,” 5th Edition, Springer (2016)