Friday April 3, 2015

Lightning striking two Air Alaska planes flying into Sea-Tac airport (Seattle) on Wednesday evening this week (Apr. 1).

In this lecture we will look at lightning strikes to airborne vehicles such as airplanes, space vehicles, and even briefly at dirigibles.  We will also look at preliminary results from a difficult recent experiment that sought to determine the conditions necessary for successful rocket triggered lightning initiation.

As the table below shows, aircraft are "struck" by lightning about once every 3000 flight hours or about once per year for commercial airliners.  The majority of strikes occur when the plane is inside a cloud and in precipitation.
Aircraft type
(typical cruising altitude)
 Strikes
 Flight hours
 No. hours between strikes
Piston
(10,000 - 15,000 ft)
808
 2,000,000
 2,475
Turbo prop
 389
 1,291,000
 3,320
Jet
(~30,000 ft)
 521
 1,741,000
 3,340
Combined
 1,718
 5,032,000
 2,930
source: Uman and Rakov (2003)

Usually the metal body of the airplane is able to safely conduct the lightning current from one end of the plane to the other without catastrophic damage.  The metal body also acts as a Faraday cage and partially shields sensitive electronics and flight control systems inside the plane from direct or indirect damage.  There are enough openings in the body of the airplane however that damage to the interior does occur.  Sparking across poorly bonded surfaces is a particular concern because it may ignite fuel.

Some catastrophic accidents have occurred of course.  A few are summarized below.

December 8, 1963, PanAm Boeing 707
The aircraft was making its initial approach into Philadelphia and was in a holding pattern at 5000 feet altitude near Elkton Md when it was struck by lightning.  Sparking, near the fuel tank vent outlet it seems, ignited fuel vapors in the left wing reserve fuel tank (the outermost fuel tank).  That then caused the center and right fuel tanks in the left wing to explode.  73 passengers and 8 crew members were killed.  After further investigation and study, the thickness of the metal enclosing the fuels tanks in 707s was increased and bonding between metal in the airframe and fuel access plates and filler caps was improved.  This was apparently the last time lightning was responsible for commercial US airline casualties (ref).

March 9, 1976, Imperial Iranian Air Force Boeing 747
The plane was probably near 6000 feet altitude near Madrid, Spain, when radio contact was lost.  The US National Transportation Safety Board was given permission to participate in the accident investigation because the 747 was so widely used worldwide.  It appears that lightning sparking ignited fuel in the vicinity of a motorized valve that was inside the fuel tank.  It is believed that the explosion broke the left wing of the aircraft into 15 major pieces.

February 8, 1988, Fairchild Metro III
This 2 turboprop commuter airliner was approaching Dusseldorf at an altitude of about 3000 feet when it was struck by lightning.  The lightning apparently completely disabled the entire electrical system including battery backup systems and the cockpit voice recorder.  The plane was inside a cloud and the pilots may not have been to read their instruments, could not lower the landing gear, and had only limited control of the airplane's flaps.  A wing separated from the aircraft which went into a spiral dive and crashed.  19 passengers and 2 crew members were killed.

February 26, 1998, US Airways Fokker F28
The aircraft was traveling from Charlotte, NC, to Birmingham, AL, when it was struck by lightning.  There were no immediate problems but within a few minutes both of the aircraft's hydraulic systems failed.  Lightning apparently side flashed to the hydraulic lines and punctured them allowing the hydraulic fluid to leak out.  The landing gear separated from the aircraft on landing and the plane skidded about 1100 feet in grass along the side of the runway but there were no casualties.

November 14, 1969, Apollo 12
The space vehicle initiated a lightning strike to ground 36 seconds after launch when at an altitude of about 6400 feet and another intracloud discharge 52 seconds after launch at an altitude of about 14,400 feet.  Nine nonessential instrumentation sensors were permanently damaged.  Several critical systems were upset but recovered.  The mission was able to continue to the moon and return.

March 26, 1987, Atlas-Centaur 67
We mentioned this earlier in the semester when discussing the enhancement of electric fields by conductors.  The vehicle apparently triggered a 4-stroke cloud-to-ground discharge 48 seconds into the flight.  Electrical transient signals changed the contents of one memory location in the computer control system.  A sudden change in the flight direction caused the vehicle to break up.

May 6, 1938, the Hindenburg
The Hindenburg, the largest airship to ever fly, had apparently been struck by lightning several times without incident (the metal frame of the vehicle provided some protection).  After completing a trans-atlantic flight it caught fire while docking in Lockhurst, New Jersey.  The cause of the fire was apparently corona discharge from the landing structure that ignited hydrogen gas coming from a small leak on the vehicle.  The airship burned from end to end in 34 seconds (though there is some dispute about this figure, it may have been faster than that).  I didn't realize it but 62 of the 97 passengers and crew survived the disaster.

Saying that aircraft are struck by lightning may be misleading because in perhaps 90% of cases it is the airplane itself that initiates or triggers the lightning discharge.  Here are three examples caught on video.

In this first case a Boeing 747 is struck shortly after take off from Komatsu Airport in Japan in the winter.

Note the clear branching that points away from the two extremities of the airplane, evidence that positively and negatively charged leader processes began at and moved away from the plane ( here's a video of the discharge).

A Qantas L-1011 was struck on Sept. 5, 2004 (this video I think) during its descent into the Sydney airport.

On April 23, 2011 an Emirates Airlines Airbus A380 was struck during its final descent to London Heathrow (source of the still photograph below)

Here's a video (from the Washington Post article) where you can see the lightning channel sweep along the fuselage.  Here's a pretty good written description from the videographer himself.

In the unusual cases where the airplane intercept a natural lightning the plane is near the ground.  This video seems to be one of those situations.

Uman and Rakov (2003) and Uman (2008) discuss four research programs in which instrumented aircraft were purposely flown into thunderstorms in order to better determine the conditions that could lead to lightning trigger and to evaluate whether lightning test standards were adequate.  Some "hardening" of these aircraft was probably done for the research projects and there were no serious incidents caused by lightning.


1964-1966
US Air Force Cambridge Research Laboratories
"Rough Rider Project"
used an F-100F aircraft
 (a 2 seat trainer version of the F-100, I believe)
studied thunderstorms in Florida

Data obtained from 49 lightning discharges the recording system may not have had sufficiently fast time response
to accurately determine current rise time and peak dI/dt values


Four US Air Force North American F-100C Super Sabre jets 
flying in formation
(source of this image)




1980 - 1986
NASA Langley Research Center Storm Hazards Program
used an F-106B aircraft

About 1500 thunderstorm penetrations at altitudes ranging from 5000 to 40,000 ft.  Struck by lightning 714 times.  Almost 10 times as many strikes were recorded at high altitudes than at lower altitudes even though the numbers of penetrations at the different altitudes were about equal.
A US Air Force F-106B Delta Dart
This is a two-seat trainer version of the aircraft
operated by the New Jersey Air National Guard
source of this image





1984, 1985, and 1987
USAF/FAA Lightning Characterization Program
used a Convair - 580 twin engine turbo prop aircraft
A Pionair CV -580 aircraft (Pionair is a New Zealand based airline)
source of this image





1984 and 1988
no data apparently from the 1984 experiment
the 1988 experiment was conducted in the south of France

A particular interest were the processes that occur at the beginning of lightning discharges triggered by the aircraft.
A Transall C-160D transport aircraft
operated by the German Air Force.
The Transall is built by a consortium of French and German companies.
source of this image


The lightning strikes seen in the videos and many of the strikes that caused the accidents mentioned above occurred at a relatively low altitude.  The plot shown at the left of the figure below shows the relative frequency of lightning strikes plotted as a function of altitude.  These data come from studies conducted in the 1950s and mid 1970s and are adapted from Fisher and Anderson (1999).


The distribution of lightning strikes has been positioned next to a figure discussed in our Feb. 20 class.  The right figure shows the locations of charge centers involved in both cloud-to-ground and intracloud discharges determined using multi-station measurements of electric field changes at the Kennedy Space Center in Florida.  This aim is to give some appreciation for where lightning strikes to aircraft occur relative to the main charge centers in thunderstorms.  The data at right really just apply in Florida.  The lightning strike data at left come from locations around the world.  Some of the US data probably came from Florida.  The fact that the 5 histograms are similar leads me to believe the comparison above isn't too unrealistic. 

Most strikes appear to occur near or just below (colder than) the freezing level (0 C).  Older piston and turbo prop aircraft had a cruising altitude between 10,000 and 15,000 feet.  Modern commercial jets fly at and above 30,000 feet but are usually struck when passing near the freezing level during their ascent or descent.

Next we will look at some of the results from the instrumented flights into thunderstorms.  The description of the lightning discharge initiation process is remarkably consistent among the studies.  We will present information obtained from 3 of the aircraft mentioned above.


This first figure is from Mazur (1989) and shows the typical initiation of a discharge by the F-106B.  Generally inside the thunderstorm fields on the aircraft body will be due to charges induced on the plane by the external field and also due to charge present on the aircraft.  The discharge begins with a weak negative corona discharge from one extremity of the aircraft (Step 1 above).    The corona discharge causes a positive excursion of the field.  The negative corona slows or stops and is followed by development of the positive leader from another extremity of the plane (Step 2).  This causes a relatively slow (milliseconds) negative field change.  Lightning discharges initiated by the other aircraft often seem to begin with the positive leader (negative corona is not observed).  That was the case with the CV-580 as we will see.

The positive leader carries positive charge from the plane and leaves the plane negatively charged.  Negative corona begins again and produces a somewhat faster positive field change (Step 3).   Current pulses are observed on the aircraft during this stage.  The negative corona develops into a negative stepped leader.  The positive leader is accompanied by a continuous current Ic that grows to several hundred amperes.  Current pulses, Ip, in the stepped leader may have amplitudes of a few kiloamps.

An example of data recorded by the F-106B are shown below.

Once the discharge is underway it seems to resemble a natural intracloud discharge.

Here's an illustration of lightning initiation on the CV-580.  It starts with the positive leader step.

Note the bipolar discharge doesn't necessarily travel away from opposite ends of the fuselage.  The points at which the positive and negative discharges originate must depend on the specific shape of the aircraft, aircraft charging, and on the orientation of the ambient electric field.

Current and electric field recorded on the CV-580 at the beginning of a lightning strike. All of the studies seem to indicate an ambient field of about 50 kV/m is needed to trigger a lightning discharge (see table below).  Figures C and D show the 4 ms interval in the middle of A and B on an expanded time scale.  This is the interval in which the higher amplitude current pulses begin to appear on the record.
Aircraft
 Q (mC)
 E (kV/m)
Convair 580
 - 0.66
 51
Transall C160
 -1.2
 59

Average values of the charge on the aircraft and the ambient electric field just prior to a lightning strike to the aircraft (from Laroche et al., 2012)


Next, in the second half of this class, we will discuss an experiment that sought to determine the conditions above the ground needed to trigger lightning.  The experiment was conducted at the rocket triggered lightning site (the International Center for Lightning Research and Testing) operated by the University of Florida near Gainesville, Florida.  The experiment and results discussed below are from Willett et al. (1999).

The basic set up is illustrated below

The plan was to launch an electric field sounding rocket about 1 second before sending a launch command to one of the triggering rockets.  It would take about 2 seconds for the command to reach the triggering rocket (the firing signal was sent pneumatically) and another 2 seconds roughly before the triggering rocket reached triggering altitude.  The electric field as a function of altitude could then be determined just prior to a triggered lightning discharge.  Launches were attempted after the active periods in storms to try to avoid having a natural discharge occur during a triggering attempt.

The sounding rocket is very similar to one described and pictured in Marshall et al. (1995) which is reproduced below:


The rocket is just over 2 m tall.  A big part of the outer body of the rocket spins.  Six windows (three are shown above) alternately cover and expose 8 field mill stators mounted inside.  I've included the figure caption because the arrangement of the inner stators is not immediately clear to me.  The eight field measurements made by the rocket are enough to determine all 3 components of the ambient field.  Data from the rocket were transmitted to a telemetry station about 2 km from the triggering site (the receiving antenna would follow and track the sounding rocket).  Streak cameras and video cameras were also located at the telemetry site.  One major disappointment of the experiment was that no photographs of the upward leader with the streaking cameras.  No direct measurements of leader velocity were made.

The rocket was fired at an 80 degree elevation angle, would reach a peak altitude of just over 3500 m in about 24 seconds and fall to ground about 2 km away from the triggering site.  Total duration of the flight was just less than 1 minute. 


Useable data was obtained for 10 of 15 attempts and lightning was triggered in 9 of these 10 cases.

An example of field measured as a function of altitude.  The field is remarkably constant once above the space charge layer found near the ground.  The abrupt drop in field near the end of the record was caused by a triggered lightning discharge..

The wire from the triggering rocket was connected to a 207 mΩ shunt so that low amplitude currents flowing in the wire during the upward leader process that initiates a triggered lightning flash could be measured.  The downward dart leader and subsequent that is subsequently triggered attached to a ground lightning and were not measured.


This is an example of currents recorded in the triggering wire at the beginning of a successful triggered discharged.  The current waveform recorder was triggered by the 1st precursor.  This is the first sign of a discharge forming at the upper end of the wire.  The triggering rocket was 112 m high at the time of the 1st precursor.


The first precursor shown on a much faster time scale.  Most precursors consisted of just a single pulse like this.  The oscillations are caused by the current reflecting repeatedly off the top and bottom ends of the wire. 

At just over 600 ms into the discharge an unsuccessful leader was initiated.  The triggering rocket was 260 m high at this point.  This "failed leader is shown on a faster time scale below.


This failed leader lasted about 1.8 ms during which time it propagated about 35 m at a velocity of 1.9 x 104m/s.

Ultimately at about 900 ms into the discharge, when the triggering rocket was at 307 m a successful leader was initiated and ultimately led to a triggered discharge.  This leader is shown below.  Note the continuous current that develops once the leader is underway.



The table below summarizes the results from the experiment, conditions at the time of  initiation of a successful trigger.
Launch
 height (m)
of the triggering rocket
 electric field (kV/m)
Potential (MV)
2
 447
 12.4
 -5.0
6
 307
 14.5
 -4.7
7
 295
 19.4
 -4.1
8
 279
 15.0
 -3.6
9
 312
 13.4
 -3.7
12
 336
 15.9
 -4.6
13
 230
 18.5
 -3.6
14
 324
 17.3
 -4.2
15
 299
 16.4
 -4.1


Potential above is the voltage at the triggering height relative to the ground (the integral of electric field with altitude)

By comparison there was one case where lightning wasn't triggered.  The E field in that case ranged from 2.5 to 4.5 kV/m and the potential at 500 m was -1.4 MV.










References:

Marshall, T.J., W. Rison, W.D. Rust, M. Stolzenburg, J.C. Willett, and W.P. Winn, "Rocket and ballon observations of electric field in two thunderstorms," J. Geophys. Res., 100, 20815-20828, 1995.

Willett, J.C., D.A. Davis, and P. Laroche, "An experimental study of positive leaders initiating rocket-triggered lightning," Atmospheric Research, 51, 189-219, 1999.

Uman, M. A. and V. A. Rakov, "The interaction of lightning with airborne vehicles," Progress in Aerospace Sciences, 39, 61-81, 2003. (available at: http://www.pas.rochester.edu/~cline/FLSC/Lightning%20Report.pdf )

Laroche, P., P. Blanchet, A. Delannoy, F. Issac, "Experimental Studies of Lightning Strikes to Aircraft," Journal Aerospace Lab, Issue 5, December 2012 (available at
http://www.aerospacelab-journal.org/sites/www.aerospacelab-journal.org/files/AL05-06_0.pdf)

Washington Post article on lightning strikes with details about the Qantas strike and the Emirates Airlines A380 strike.
(available at: http://www.washingtonpost.com/blogs/capital-weather-gang/post/airline-safety-is-the-fear-of-a-lightning-strike-warranted/2011/05/17/AFD2vs5G_blog.html )

Mazur, V. "A Physical Model of Lightning Initiation on Aircraft in Thunderstorms," J. Geophys. Res., 94, 3326-3340, 1989. (available at: http://onlinelibrary.wiley.com/doi/10.1029/JD094iD03p03326/pdf