Where's Captain Eddie Rickenbacker? 


Captain Eddie Rickenbacker was America's top WWI fighter ace with 26 air victories. He successfully won control of Eastern Airlines from General Motors before WWII. Within a year after America's entry into WWII, Eddie Rickenbacker and a B-17D crew of five would face their greatest challenge for survival.

On 18 October 1942, Eddie Rickenbacker and a crew of five departed Honolulu, on their second leg shown in Figure 1, in a B-17D en route to Port Moresby, New Guinea with a stop at Canton Island a speck of land in the expansive Pacific Ocean. Rickenbacker was entrusted with a top secret oral message from Secretary of War Stimson for General Douglas MacArthur whose headquarters was at Port Moresby. Rickenbacker and the crew had earlier ground looped another B-17D that had blown a tire on take-off. The flight path shown in Figure 1 was a 10 hour-leg with an ETA (estimated time of arrival) at Canton Island of 0930 19 October 1942. The island of Canton was an atoll 8 miles long and 4 miles wide that failed to come into view as the ETA passed. At 1015 with fuel for four hours remaining, it was evident that the B-17D was lost, the tail wind underestimated. The navigator First Lieutenant John J. DeAngelis concluded that his octant was damaged during the earlier ground loop and provided wrong sightings that resulted in flawed navigation. The lost B-17D was unable to obtain radio direction finding assistance from Canton Island since the equipment had not been uncrated. A typical bearing pattern of a non-directional HFDF (high frequency direction finding) radio station is shown in Figure 2. The ill-fated B-17D finally had to crash land into the Pacific when its fuel was spent. A model flight log is shown in Table 1. Rickenbacker and the crew survived their perilous ordeal in a raft for 24 days and were rescued 13 November 1942 with only one death of a crew member.

In a "what if…." scenario, let us assume that another B-17D (with extended range fuel tanks) trailed the ill-fated B-17D and located it. Assume that both B-17D aircraft were in communication on HF (high frequency) radio and had prototype VHF (very high frequency radio line of sight) for communication (VHF in reality was not introduced until the 50's). The trailing B-17D directed the lost B-17D to point into the Sun. They compared their magnetic headings to the Sun. From this observation, the trailing B-17D could determine on which side of its course was the lost B-17D. The trailing B-17D requested that the navigator of the lost B-17D use his hand as a sextant holding it horizontal and using his fingers to measure the elevation of the Sun. The width of two fingers, when the arm is fully extended, represents about 5 degrees in elevation. The trailing B-17D performed the same procedure. Lt. DeAngelis' reading of the Sun was about 1 finger higher than the reading of the Sun in the trailing B-17D. Thus the lost B-17D was closer to the sub point of the Sun and separation between the two aircraft 2 1/2 degrees (which is 2.5 deg.X 60nmi/deg = 150 nmi ).

The lost B-17D was requested to hold a tight orbit as the trailing B-17D began to fly an aural box pattern as seen in Figure 3. The lost B-17D emitted a tone on the VHF frequency. The trailing B-17D marked its own location when the tone was first heard and maintained its course until the tone was lost when out of VHF range. The line of sight VHF radio range for the two B-17's ,each at 8,000 feet, is shown in Figure 4. The trailing B-17D then turned 90 degrees later followed by two additional 90 degree turns reestablishing VHF contact until it lost it again. The navigator of the trailing B-17D plotted his position at each point that he gained and lost contact of the tone and drew the chords connecting these points. He then was able to establish the location of the first B-17D by locating the intersection of the perpendicular bisectors to these chords (which establishes the center of a circle containing the chords) and the most likely location of the lost B-17D. Assume that the HFDF equipment at both Palmyra and Canton were operative and that coded radio bearings (lines of positions) were obtained from both Canton and Palmyra by the lost B-17D. Canton's message was y equals zero; Palmyra's message was 4y equals 3x (Y axis is east - west and X axis is north - south). Hint for decoding bearing messages: use basic equation y = mx (where m is the slope of the bearing) therefore m=y/x . Each HFDF station has its own unique coordinate system at its site. The slope is measured with respect to the X- axis. Since the flight leg is close to the equator, a degree of longitude is close in value to a degree of latitude (in nautical miles).

Figure 1.

Figure 1. Flight Chart of Lost B-17D (Y axis is east-west and X axis is north-south) Honolulu to Canton Is. 1660 nmi. at course 211°

Figure 2.

Figure 2. Typical bearings (in degrees) to/from an HFDF non-directional radio station (an aircraft obtains a bearing that is a line of position such as 045-225 not knowing at which side of the station it is located in the case of a non-directional radio station).

Original log
What if…log
B-17 leaves Honolulu @ 0030 Oct.18, 1942 ETA to Canton is 0930 Hawaiian time. Navigator does celestial with octant and shoots Sun during daylight. Believes tail wind was 10 miles per hour as forecast. same same
0830 Pilot begins letdown to 1,000 ft altitude looking for island 8 miles long by 4 miles wide same same
1015 No land in sight 4 hours gas remain same same
  Rickenbacker believes tail wind is stronger and requests assistance from Canton. But Canton has not uncrated their DF eqt.   Canton HFDF eqt. uncrated and placed on air.
  Palmyra Is. contacted for DF assistance.   same
  Navigator takes two more Sun shots and has pilot change course twice   same
  Palmyra Is. obtains bearings on B-17 as it orbits and provides steering course to Canton. Palmyra Is. Course not believed. Trailing B-17 requests Lost B-17 to orbit as it flys aural box pattern Both aircraft have prototype VHF (line of sight) radios. Trailing B-17 flys until VHF signal is heard holds course until tone lost; turns 90 degrees holds course then turns 90 degrees two more times. Regains signal and marks point when VHF signal is gained and lost . Plots course of box pattern by connecting points. The intersection of the perpendicular bisectors of the chords flown between contact and loss of contact of the tone is the location of the lost B-17.
  Navigator now realizes that octant was damaged in earlier ground loop. All celestial results are flawed. Lost B-17 is located Coded bearing messages are received by the lost B-17 from Canton and Palmyra HFDF radio. Navigator decodes the bearing messages and plots his position.
1436 Canton upon request fires AA shells for air burst at 7,000ft. Nothing seen by B-17. One hour's fuel remains. Too late to box the compass. Crash landing at sea. same Lost B-17's location is now established by both the trailing B-17 and lost B-17. The locations of the lost B-17 deduced by the aural box pattern procedure and the HFDF coded bearings are reasonably close. Lt. DeAngelis gives his pilot a heading alteration to intercept Canton Is. which reasonably agrees with the heading alteration deduced by the trailing B-17 navigator to enable the lost B-17 to reach Canton Is.
  Rickenbacker and crew (one death) rescued after 24 days at sea on 13 November 1942 near Funa Futi about 640 nmi SW of Canton Is.    

Table 1. Flight Log of Lost B-17D in red (with What if… scenario in blue)

Figure 3.

Figure 3. Using aural box pattern to locate lost aircraft general case (flight path directions arbitrary).

Figure 4.

Figure 4. Line of sight Range of VHF Radio (not to scale).

Where is Captain Rickenbacker's B-17 when located? Return to Figure 1 and note the coded bearing information (from end of previous paragraph) from Palmyra and Canton Islands' HFDF stations and plot them on the screen or on a printout of the page to determine the answer.

a. In the center of the orange circle
b. In the center of the green circle
c. In the center of the yellow circle
d. In the center of the blue circle

The answer is a

The slope of the bearing line of position from Canton Is. is zero (since Y=0) with respect to the X axis (north-south) and therefore the bearing line of position is perpendicular to the Y axis (passing through the approximate center of the orange circle). The slope of the bearing line of position from Palmyra Is. is (where y=3/4(x)) y/x=m=3/4. One can establish the line of position from Palmyra by stepping off 12 degrees of longitude westward from Palmyra and then stepping off 16 degrees of latitude southward. Connecting the line from Palmyra Is. to the extremity of southward line will establish the bearing line of position which passes through the approximate center of the orange circle. Another method to establish the bearing line of position from Palmyra : slope of bearing line of position is y/x which is 3/4 therefore solve for arctan (3/4) which is 36.9 degrees or its reciprocal 216.9 degrees. Using a protractor, pass a line from Palmyra Is. making an angle of 037 degrees (clockwise) from a vertical (north-south line) through Palmyra Is. and extend it beyond the approximate center of the orange circle.

Figure 5.

Figure 5. Captain Eddie Rickenbacker and crew in their lost B-17D located (by intersections of bearing lines of position and perpendicular bisectors).

True story of why Rickenbacker's navigator got lost: His flawed sextant sightings gave him erroneous celestial fix positions. According to one story circulating in navigation circles in WWII, DeAngelis requested that his sextant be collimated (realigned) after the fix ground loop incident. His request was turned down by Rickenbacker. The actual tailwind for this leg was about 31 miles per hour (21 miles per hour greater than forecast) as determined by a navigator whose B-17D departed Hickham Field an hour earlier for Canton Island. This would shorten the flight by about an hour over the forecast ETA. As a result the lost B-17D (which may have relied upon forecast wind information) started its descent too late and missed the tiny island of Canton. The direction finding radio station at Palmyra was non directional. It furnished bearings that were lines of position but not direction. A bearing from such a station could be interpreted as 045 degrees or its reciprocal 225 degrees (as an example) . A good DR (dead reckoning) position would resolve this ambiguity. The ambiguity of the non directional station is further resolved when one has information from two stations as the intersection of their lines of position represents the most likely position of the aircraft. However, the Canton Is. station was not in operation as the HFDF radio equipment remained uncrated. B-17 Flying Fortress specifications are shown in Table 2.

First flight July 28, 1935
Classification Bomber
Span 103 feet 9 inches (B-17G)
Length 74 feet 9 inches (B-17G)
Gross weight 65,000 pounds (B-17G)
Top speed 287 mph (B-17G)
Cruising speed 150 mph (B-17G)
Range 3,750 miles (B-17G)
Ceiling 35,600 feet (B-17G)
Power Four 1200-horsepower Wright R-1820-97 engines
Accommodation 2 pilots, navigator, bombardier, radio-operator, 5 gunners (B-17G)
Armament 11 to 13 machine guns, 20,000-pound bomb load (B-17G)

Table 2. B-17 Flying Fortress Specifications

A total of 6,981 B-17s were built by Boeing plants and another 5,745 B-17s were built by Douglas and Lockheed (Vega) plants.


Emmott, N. W. "A Plunge in the Pacific." Litton Avionics Newsletter. Woodland Hills, CA., Volume One, Number 4, August 1970, pp. 22-27.

Rickenbacker, Edward V. Rickenbacker. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1967