Earhart Radio Part I: The 3105 Donut

Any plausible theory about where Amelia Earhart and Fred Noonan’s final flight ended has to account for what is recorded in the radio room logs of the U.S. Coast Guard Cutter Itasca. Itasca was stationed just offshore of Howland Island, the flight’s destination, in support of the flight. One of the tasks of Itasca’s radio room personnel was to listen for voice radio messages Amelia would be transmitting on the frequency of 3105 kilocycles. When Amelia was first definitively heard in the Itasca radio room at 2:45 AM ship’s time, few if any of her words could be made out over the sound of static [1]. As she flew her Lockheed Model 10 Electra closer to Howland Island, the reception quality of her radio messages improved. Radio operators of the day rated the radio signals they heard on a progressive signal strength scale ranging from an S-1:"Hardly perceptible" to S-5: "Very good, perfectly readable" [2]. An Itasca radio room log provides signal strength ratings for three of the last four times Amelia’s messages were received, and in all three cases these were S-5 receptions. The strength of Amelia’s final few messages suggests the Electra was close to Howland Island at this time. Itasca’s chief radio man, Leo Bellarts, recalled in an interview that Amelia was coming in so strongly he went out onto the Itasca’s deck expecting to hear the Electra’s engines and see the plane fly over [3]. Amelia and Fred also believed they were close to Howland at this time. An Itasca radio log entry for the first of Amelia’s three S-5 messages, received at 7:42 AM ship’s time [4], records Amelia’s (radio call sign KHAQQ) message as: 

 

KHAQQ CLNG ITASCA WE MUST BE ON YOU BUT CANNOT SEE U BUT GAS IS RUNNING LOW BEEN UNABLE TO REACH YOU BY RADIO WE ARE FLYING AT 1000 FEET 

 

While Itasca was hearing Amelia quite well, this log entry tells us Amelia hadn’t heard any of Itasca’s attempts to contact her on the same frequency. For reasons that aren’t clear, two-way voice communication between the Electra and Itasca on 3105 kilocycles was never established before Itasca stopped hearing from Amelia. The only radio signals from Itasca that Amelia acknowledged hearing were strings of Morse code ‘A’s that Itasca transmitted on 7500 kilocycles as a homing signal. Itasca radio log entry for 8:00 AM ship’s time reports Amelia saying she had heard Itasca’s ‘A’s but was unable to take a bearing on them with the radio direction finding (RDF) equipment on board the Electra [5]. Amelia had arranged prior to the flight for Itasca to transmit the homing signal on 7500 kilocycles, apparently unaware that her RDF equipment was incapable of taking bearings on signals sent on frequencies higher than 1500 kilocycles. I think this error more than any other sealed Amelia and Fred’s fate. Had Amelia been able to take bearings on the Itasca’s ‘A’s from 8:00 AM onward she would have had at least 40 minutes to home in on Howland Island, enough time to cover roughly 80 miles if the Electra’s groundspeed was about 120 mph, a reasonable estimate. Amelia’s 8:00 AM message was the next-to-last time she was heard in the Itasca radio room. The last time she was heard was at 8:43 AM ship’s time, another S-5 reception. The Itasca radio log entry for this reception reads:

 

KHAQQ TO ITASCA WE ARE ON THE LINE 157 337 WL RPT MSG WE WL RPT N ES S THIS ON 6210 KCS WAIT

 

Translated from radio room shorthand, Amelia is saying that she was flying north and south on a 157-337 degree (magnetic) course line, the final course line Fred Noonan’s celestial navigation would have provided, and that she would repeat her message on 6210 kilocycles, the only other frequency on which her radio equipment allowed her to transmit. It would seem that Amelia and Fred were flying back and forth on the 157-337 line trying to spot Howland Island. Amelia may have wanted to repeat her message on 6210 kilocycles hoping that switching transmission frequency would resolve the communication problem with Itasca. Itasca listened for Amelia on 6210 and 3105 kilocycles but never heard her again.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

The Itasca radio room logs present a problem for the Nikumaroro Hypothesis. If the Electra was flying close to Howland Island from about 7:42 AM to 8:43 AM and its fuel supply was running low by 7:42 AM, it’s hard to believe that Amelia and Fred Noonan ended up at Nikumaroro, located some 400 miles (~350 nautical miles) southeast of Howland Island. A short article titled ‘3105 Donut’ in a 2008 issue of TIGHAR’s in-house publication TIGHAR Tracks [6] attempts to address this problem, or at least make it less of a problem for the Nikumaroro Hypothesis.

Recent software advances have made it possible to computer model the propagation properties of the Electra’s transmitting antenna to an unprecedented degree of accuracy.

As a result, the long-held assumption that the closer the plane was to Howland Island the stronger the signal heard by the Coast Guard would be, has been shown to be incorrect. A peculiarity in the antenna’s transmission pattern meant that if the plane was closer than about 80 nautical miles there was less than a 10% chance that Itasca would hear Earhart on 3105 kilocycles at maximum strength as recorded in the cutter’s radio log. Chances are the Electra was at least 80 and perhaps as much as 210 nautical miles from the ship at the time of the last transmission.

At 08:43-55 local time Itasca heard Earhart say, “We are on the line 157 337. Will repeat message. We will repeat this on 6210 kcs. Wait. We are running on line north and south.” The message came in at maximum strength. Given a newly discovered anomaly in the propagation pattern of the aircraft’s transmitting antenna, to have even a 10% chance of being heard at maximum strength, the Electra had to be somewhere within the “donut” shown. If on the line southeast of Howland, the plane was much closer to Gardner Island (Nikumaroro) than previously assumed.

 

TIGHAR’s 3105 Donut result put the Electra as close as ~160 miles (~140 nautical miles) to Nikumaroro when Amelia was last heard by Itasca at 8:43 AM. Amelia and Fred would still have to have decided to give up searching their immediate vicinity for Howland Island, which they thought they were close to, and flown steadily to the southeast for an hour or more to get to Nikumaroro. That still seems pretty implausible, but if you want to believe in the Nikumaroro Hypothesis, I guess you’ll take whatever reduction in implausibility you can get. 

A reasonable question to ask about the 3105 Donut result is whether it is correct. TIGHAR never released a report explaining the nuts and bolts of the software modeling work underlying the result, but in 2011 a member of the group’s online forum started a thread to discuss the 3105 Donut result [7]. The TIGHAR member who did this work, a guy named Bob, responded to questions and comments posted there and in so doing provided insight into his work. A lot of what is discussed in this thread goes over my head, but I think I’ve managed to grasp a few key points. In response to one question Bob explains:

“The “donut” was computed by the ICEPAC propagation model, using the dorsal antenna gain pattern computed by 4NEC2.  The SNR at the Itasca was computed versus Electra distance in 20-mile increments along the LOP, from 20 miles to 340 miles.”

Further along Bob explains:

“It’s possible that there was direct path propagation at short distances, due to excitation of the airframe, but ICEPAC only calculates path loss for an ionospheric path. However, at 1,000 feet altitude (where Earhart said she was flying then), the horizon distance is about 38 miles.  So outside about 40 miles, there wouldn’t be any direct path, and skywave would govern.”

So there were two key parts to Bob’s modeling work: modelling the radiation pattern of the Electra’s transmission antenna using 4NEC2 software [8]; and calculating the signal-to-noise ratio (SNR) of Amelia’s radio messages as a function of distance using ICEPAC skywave propagation modelling software [9]. In the case of skywave propagation, radio signals from a transmitter are refracted back from the ionosphere towards the earth’s surface as they move away from their source; ‘skip zones’ can occur where the signals produced can’t be heard or are weakly heard, as illustrated in the figure below. According to Bob’s skywave propagation modelling work, the Electra was most likely between 80 miles and 210 nautical miles from Itasca when Amelia was heard at S-5 signal strength; at distances less than 80 miles his modelling work gave little chance of Itasca hearing her at S-5 signal strength.

Skywave and groundwave propagation [10]

There are two other ways Amelia’s messages could have reached Itasca: by line-of-sight transmission and by groundwave propagation. Bob’s TIGHAR forum posts explain why he ruled out these two possibilities. As the name suggests, line-of-sight transmission could only have occurred if the Electra was within visual range of Itasca, about 40 miles at the Electra’s last reported altitude of 1000 feet. Bob argues that the Electra was never in visual range of Howland Island because if it had been, Amelia and Fred would have seen Howland Island. I’m not convinced by this argument. According to the Itasca’s deck log [11], cumulus clouds covered 20 to 30 percent of the sky during the period when Amelia was heard at S-5 signal strength. What that cloud cover looked like can be seen in a photo of Itasca right offshore of Howland Island on the morning the Electra failed to arrive [12]. The Waitt Institute’s YouTube channel has a video shot from a helicopter flying at 1000 feet near Howland Island [13] that gives a sense of the difficulty of spotting the island under these conditions. If the Electra had gotten really close to Howland Island, maybe within five or six miles of it, the island would have been pretty hard to miss. But if they were twenty or more miles away? The Waitt Institute video tells me Howland would have been pretty difficult to spot at such distances. Whatever one might think about how easy it would have been for Amelia and Fred to spot Howland Island, it’s fair to say that Bob didn’t rule out the possibility of the Electra being within 40 miles of Howland based on radio wave propagation physics, he simply offered his own personal judgement about this possibility. 

Groundwave propagation, like skywave propagation, allows radio signals to travel beyond the horizon at the source location. A key difference compared to skywave propagation is that in the case of groundwave propagation signal strength tends to steadily decrease with distance from the transmitter. The ‘donut hole’ result that came out of Bob’s ICEPAC skywave modelling isn’t something you’d expect to see for groundwave propagation of Amelia’s signals to Itasca. What led Bob to use the ICEPAC skywave propagation model rather than a groundwave propagation model? This is where his 4NEC2 software modeling of the Electra’s antenna radiation pattern comes into the picture. In the same TIGHAR discussion forum thread in which Bob made the comments I’ve excerpted from above, there is a series of exchanges between Bob and a forum member named Chuck about whether Bob had correctly modeled the radiation pattern produced by the Electra’s transmission antenna. At one point Chuck quotes something Bob had written in 2001 about the Electra’s transmitting antenna that explains why Bob chose a skywave propagation model [14]:

“In order for there to be a vertical electric field component, and hence a ground wave, it is necessary that the radiating antenna be vertical.  If the antenna is horizontal, the electric field is parallel to the earth’s surface, and is immediately shorted out, causing the entire wave front to collapse, thus preventing ground wave propagation.  And there’s the rub.  The Electra’s antenna was horizontal, hence there was no groundwave.”

In 2001 Bob believed that the Electra’s transmitting antenna couldn’t radiate groundwaves because the net polarization of the radio waves it radiated at 3105 kilocycles was horizontal, thus Amelia was transmitting with what was effectively a horizontal antenna. The forum exchanges with Chuck indicate that this was still Bob’s opinion in 2018 when he obtained his ‘3105 Donut’ result. Chuck argued that the net polarization of the radiated electric field was vertical due to the contribution made by a vertical wire leading up to the ‘vee’ antenna on top of the Electra, so Amelia was transmitting with what was effectively a vertical antenna. Chuck had come to this conclusion based upon his own antenna modeling work, some of which is presented in the discussion thread. If Bob’s antenna modeling work was wrong, his 3105 Donut result wasn’t valid since there would then be no justification for using a skywave propagation model to determine how the signal strength of Amelia’s radio messages as heard by Itasca would have varied with distance.

 

I lack the expertise needed to adjudicate Bob and Chuck’s dispute, so this wasn’t something I thought I’d ever post about. But a while ago I came upon a YouTube video of a presentation that an expert in high frequency radio communications named Tom Vinson gave to the Brazos Valley Amateur Radio Club (BVARC) in 2020 [15]. This presentation rather convincingly shows that Bob’s 3105 Donut result is wrong.

The back story to Tom Vinson’s talk is that the ocean exploration company Nauticos [16], which has searched for the Electra on the sea floor near Howland Island, contacted experts at the Rockwell-Collins Corporation [17] asking whether it was possible to determine how far the Electra was from Howland Island based on the signal strength ratings of Amelia’s radio messages recorded in the Itasca radio logs. Tom was part of a team of Rockwell Collins personnel with extensive experience in radio communications that tackled this problem. The Rockwell Collins team took a comprehensive approach, investigating pretty much every factor that would have influenced the signal strength of Amelia’s messages as heard in the Itasca radio room. The figure below from Tom’s talk identifies the different aspects of the problem that were considered. In many cases, the research into some part of the Electra-Itasca radio link involved a combination of modeling and measurement work. Tom’s presentation describes the team’s work far better than I ever could, so I’ll simply refer readers to the video of his talk.

Courtesy of Tom Vinson

The end result of the Rockwell Collins team’s work were Electra-Itasca distance estimates for signal strengths S1 through S5. The figure below from Tom’s talk presents these results as ‘range donuts’, minimum and maximum Electra-Itasca distances consistent with the five signal strength ratings. Tom mentions in his talk, and the figure’s text states, that the signal strength ranges overlap. I suppose this is because of uncertainties associated with the many factors influencing the signal strength versus distance relationship. Tom’s talk makes clear that the Rockwell Collins team modeled Amelia’s signals reaching Itasca by groundwave rather than by skywave propagation. My guess is that the S-5 range in the figure would have been depicted as a disk rather than a donut had it been meant to display the results quantitatively; I’m pretty sure signal strength wouldn’t have dropped off to less than S-1 at close-in to Howland for groundwave propagation. Or perhaps the ‘hole’ for the S-5 range is meant to convey the idea that if they were close enough, Amelia and Fred would certainly have spotted Howland Island (note added after completing this post: Tom confirmed that the S-5 range was a disk, not a donut).

Courtesy of Tom Vinson

The Rockwell Collins team validated their groundwave propagation model by comparing the model’s prediction of the SNR-versus-distance relationship against measurements of the same for an airplane transmitting at a frequency very close to Amelia’s 3105 kilocycles while flying over the sea at an altitude of 1,000 feet. The agreement between model and measurement was excellent, see the figure below. The purpose of these measurements was to validate the Rockwell Collins team’s groundwave model; the plotted data isn’t meant to represent how signal strength varied with distance for Electra to Itasca radio transmissions. But I think this plot does illustrate that for the case of radio signals propagating as groundwaves there would have been nothing like the signal strength ‘donut hole’ result that came out of Bob’s modeling effort.

Courtesy of Tom Vinson

About 33 minutes into his talk, Tom Vinson says that at Amelia’s 3105 kilocycle transmitting frequency “the only thing radiating on that aircraft was the lead, that six-foot piece of wire. That vee on there is a capacitive top hat at that frequency…She was running 50 W power input to a six-foot piece of wire…They were hearing her overnight when it was skywave and then of course hearing her real loud when it became seawave in the morning in daylight”. Note that ‘seawave propagation’ is simply a term for groundwave propagation happening over the sea. What Tom says here sounds a lot like the point Chuck made in his TIGHAR forum exchanges with Bob, i.e., that Amelia was transmitting her messages on 3105 kilocycles with a vertical antenna.

The key result that came out of the Rockwell Collins team’s work was an estimate of how far the Electra could have been from Itasca when Amelia’s radio messages came in at S-5 signal strength. Due to a non-disclosure agreement with Nauticos, Tom wasn’t free to precisely state the result, but he did say the result put the Electra within 100 miles of Howland Island. I’m not sure if Tom meant 100 statute or nautical miles, but for the purposes of this discussion it doesn’t matter. What does matter is that the Rockwell-Collins result is pretty much the inverse of TIGHAR’s 3105 Donut result.

Given the comprehensive nature of the Rockwell-Collins team’s work and their expertise to carry it out, I have to believe the Rockwell Collins team got it right and TIGHAR’s guy Bob got it wrong. The flaw in the Bob’s 3105 Donut result apparently was a consequence of incorrectly modeling the radiation pattern of the Electra’s transmitting antenna, which led to the mistaken use of a skywave propagation model to determine how the signal to noise ratio of Amelia’s voice messages varied with distance.

Tom Vinson’s BVARC presentation covers the Rockwell Collins team’s research up to the end of 2019. He didn’t discuss work the Rockwell Collins team did in 2020, though he did play part of a Nauticos video that gives a brief overview of that work. What motivated the 2020 work was that in 2019 the Rockwell Collins team obtained a Western Electric Model 13C transmitter, the same transmitter model that Amelia had abord her Lockheed Electra. It is quite possibly the only surviving example of this piece of equipment, and it was the only piece of equipment whose performance characteristics the Rockwell Collins team hadn’t yet experimentally evaluated. The Rockwell Collins team refurbished the Western Electric 13C, which allowed them to experimentally replicate the entire Electra-to-Itasca communication link using the same radio equipment on board the Electra and Itasca when Amelia and Fred went missing. This was a more integrated evaluation of the signal strength-versus-distance relationship than the Rockwell Collins team’s previous work had permitted.

The Rockwell Collins team used a Beech 18, an airplane very similar to Lockheed Model 10 Electra, as a stand-in for Amelia’s Electra. The Beech 18 was equipped with the refurbished Western Electric model 13C transmitter and an antenna matching the configuration of the transmitting antenna on Amelia’s Electra. The stand-in for the Itasca was a boat called the Nellie Crockett. The Nellie Crockett’s mast was fitted with an antenna matching Itasca’s receiving antenna connected to a refurbished CGR-32 receiver, the same receiver model used in the Itasca’s radio room. The Beech 18 flew at an altitude of 1,000 feet over the Atlantic Ocean transmitting recordings of a woman reading Amelia’s words as recorded in the Itasca radio room logs. The simulated Amelia messages were transmitted from the Beech 18 at various distances from the Nellie Crockett. What was heard through the CGR-32 receiver onboard the Nellie Crockett was recorded, along with parameters such as the signal strength of the simulated Earhart radio transmissions and the distance of the Beech 18 relative to the Nellie Crockett. This is a very brief outline of the work done in 2020. In separate interview videos that can be found online [18,19], Tom Vinson and another key member of the Rockwell-Collins team, Rod Blocksome, discuss the 2020 work far better than I possibly can. In his interview video Rod Blocksome says that this work provided a tighter constraint on the Electra - Itasca distance range for the S-5 signal strength receptions. Neither Blocksome nor Vinson say what the refined distance estimate is, presumably because of non-disclosure agreements. So, we’re still left with the result Tom Vinson provided in his BVARC talk, that the Electra was 100 miles or less from Howland Island when she came in at signal strength 5 in the Itasca radio room. 

A Beech 18 on display at the National Air & Space Museum [20]

Amelia Earhart's Lockheed Model 10 Electra [21]

The Rockwell-Collins result doesn’t rule out the Nikumaroro Hypothesis, but it makes it seem rather unlikely. As I’ve already said, Nikumaroro is about 400 miles (350 nautical miles) southeast of Howland, about a two-hour flight from the closest point consistent with the Rockwell Collins result. I see no reason why. with fuel running low, Amelia and Fred would have chosen to fly steadily southeast for two straight hours, headed for parts unknown, when they thought they were close to their destination, the only place with a runway to land on they could possibly reach. In the circumstances they found themselves in, continuing to search nearby for Howland Island was clearly their best hope for survival. Amelia’s final message about flying back and forth on the 157-337 course line suggests that this is exactly what she and Fred were doing. In her final 8:43 AM message, Amelia said she would repeat her message on 6210 kilocycles, but Itasca never heard such a message. A good explanation for that is that the Electra ran out of fuel shortly after Amelia sent her 8:43 AM message. Flying at an altitude of 1000 feet, the Electra would not have stayed in the air more than a minute or two once its engines stopped. Tom Vinson speculates that the word ‘wait’ in Amelia’s last radio message -- KHAQQ TO ITASCA WE ARE ON THE LINE 157 337 WL RPT MSG WE WL RPT N ES S THIS ON 6210 KCS WAIT -- marks the moment the Electra’s engines began to sputter, and that she was too busy trying to fly the plane to an intact landing at sea to repeat her message.

 

There is more to say about the matter of the Electra’s fuel supply. Amelia expected that by closely following engine power and flight altitude recommendations provided by the Electra’s manufacturer she would have a few hours of flying time left when she got close to Howland Island. Tom Vinson said in the Q&A session that followed his BVARC presentation that Nauticos had commissioned Caltech researchers to estimate fuel consumption for the flight. The Caltech researchers concluded that the Electra ran out of fuel shortly after Amelia was last heard at 8:43 AM; altitude and course deviations she made, probably to avoid bad flying weather, as well as higher-than-expected headwinds, ate up the anticipated fuel reserve. According to Tom, an earlier-than-planned climb to high altitude when the Electra was heavily laden with fuel was particularly costly in terms of fuel consumption. As far as I know, TIGHAR hasn’t come up with its own fuel consumption estimates, it simply assumes that Amelia had enough fuel to fly for a few more hours when last heard by Itasca at 8:43 AM. 

 

TIGHAR hasn’t had much to say about the Rockwell-Collins group’s research results, which had been in the public sphere at least as early as 2020 when Tom Vinson’s BVARC talk appeared on YouTube. The only time I know of that TIGHAR discussed the Rockwell Collins team’s work was to comment on a 2025 Nauticos press release that reads:

CAPE PORPOISE, Maine, Aug. 19, 2025

Nauticos, a leader in deep-sea exploration and historical research, today announced plans to launch a fourth expedition aimed at locating Amelia Earhart's Lockheed Electra, driven by newly acquired proprietary data revealing the likely resting place of the aircraft. This cutting-edge research offers the most precise information yet about Earhart and navigator Fred Noonan's final position before their disappearance on July 2, 1937.

After extensive restoration and analysis of an identical radio system used by Earhart and Noonan, Nauticos has determined their approximate location at 8 a.m. on the day they vanished. This groundbreaking discovery significantly refines the search area near Howland Island, the intended destination of Earhart's ill-fated flight.

About a week after the Nauticos press release appeared, TIGHAR sent a mass email to its membership with the following commentary:

The press release is a bit misleading. The “newly acquired proprietary data” was acquired in September 2020 with a Twin Beech provided by Dynamic Aviation in Bridgewater, Virginia. N18G is now for sale. If you scroll to the bottom of the Dynamic Aviation sales ad there’s a video describing the test flight. 

There are flaws in the Nauticos reconstruction that invalidate the data. To get accurate readings you have to duplicate not only the transmitter and the power, but also the precise antenna length (including lead-ins) and its surrounding electrical ground, i.e., the aircraft itself. In other words, the small Twin Beech doesn’t work. In 2019, TIGHAR Senior Researcher Bob Brandenburg computer-modeled Earhart’s radio signal propagation and discovered an anomaly. The long-held assumption that the closer the plane was to Howland Island the stronger the signal heard by the Coast Guard would be, was incorrect. If the plane was closer than 80 nautical miles there was less than a 10% chance that Itasca would hear Earhart on 3105 kilocycles at maximum strength as recorded in the cutter’s radio log. Chances are the Electra was at least 80 and perhaps as much as 210 nautical miles from the ship at the time of the last transmission. We passed that information to Nauticos at the time, but they ignored it.

I’m guessing this was written by TIGHAR Director Ric Gillespie. Gillespie claims the work done in 2020 is invalid because the Rockwell Collins team failed to reproduce Amelia’s Electra and its transmitter antenna closely enough. It’s not clear what, if anything, Gillespie actually knows about “antenna lengths (including lead ins)” as configured in the 2020 Rockwell Collins work. How large an error might have been introduced by using a Beech 18 as a stand-in for Electra or by not closely reproducing the Electra’s transmitting antenna geometry, if that is even true, Gillespie doesn’t say. He offers nothing in the way of modeling results or experimental measurement data to support his critique, and I doubt he has anything of the kind to offer. The person who did the 3105 Donut work, Bob, passed away several years ago and I suspect TIGHAR currently has no one capable of providing a technically informed critique on the Rockwell Collins team’s work. Ric Gillespie is claiming that a team of engineers with decades of experience in radio communications failed to see flaws Gillespie seems to have readily identified. I find that very hard to believe. Rod Blocksome mentions in his YouTube interview that the Rockwell Collins work was reviewed by a University of Maryland electrical engineering professor. This reviewer apparently failed to spot any significant flaws in the work. When I began reviewing material to write this post, I contacted Tom Vinson with questions about the research he discussed in the BVARC presentation. Coincidentally, while I was exchanging email messages with Tom, a TIGHAR member forwarded me a copy of the TIGHAR email. I asked Tom about the TIGHAR critique, and he stood by his team’s work. He also told me the team had made measurements to ensure the 2020 flight experiments were an appropriate simulation of the Electra - Itasca radio link on the morning Amelia and Fred went missing.

After dismissing the Rockwell Collins team’s research, the TIGHAR email goes on to repeat the original 2008 TIGHAR Tracks 3105 Donut Hole article nearly word for word, closing with “We passed that information to Nauticos at the time, but they ignored it”. This is a classic bit of TIGHAR self-aggrandizement -- we’re the experts, we tried to set Nauticos straight, but they didn’t listen. I suspect Nauticos simply decided it wasn’t worth the bother to set TIGHAR straight. 

Unfortunately, for many TIGHAR followers pretty much everything they’ve been told about the Earhart disappearance has come from TIGHAR. But at TIGHAR ideas and evidence that run against the Nikumaroro Hypothesis are sometimes discounted if not outright ignored. That certainly was the case with the Nikumaroro sextant box and with ‘Artifact 2-2-V-1’ the piece of aluminum from a World War II era C-47 transport plane TIGHAR found on Nikumaroro, as discussed in my previous posts. The possibility that the castaway of Nikumaroro was a crewman from the Norwich City, the freighter that ran aground there in 1929, has been researched carefully by a former TIGHAR member [22]. That research effort was apparently conducted without any significant support or encouragement by TIGHAR. I don’t expect TIGHAR to willingly give the Rockwell Collins research its due, but I hope TIGHAR members reading this post will follow watch those Vinson and Blocksome videos at the links provided and consider the implications for the Nikumaroro Hypothesis. Perhaps some TIGHAR member who views those videos will post on the TIGHAR discussion forum there to make the Rockwell Collins research more widely known to TIGHAR membership.

 

Thank you to Tom Vinson, Kenton Spading, and Chuck for providing comments on a draft of this post. All errors are my responsibility, of course. 

 

Comments, corrections, additional relevant facts, differing viewpoints, etc., are welcome.  Send to gardnersghost@gmail.com

Footnotes

[1] Memorandum written by Richard Black, U.S. Department of the Interior, July 29, 1937. Black acted as liaison between Earhart and the various U.S. Government agencies supporting Amelia’s flight. He was at Howland Island on July 2, 1937. Black’s memo includes as an attachment a transcript of Itasca radio log entries obtained from Lieutenant Commander Baker, the Itasca’s Executive Officer. Earhart’s 2:45 AM message was largely unintelligible but two reporters listening in the Itasca radio room recognized Earhart’s voice. Itasca Chief Radio Man Leo Bellarts believed that he had heard Amelia say “cloudy and overcast”. https://documents3.theblackvault.com/documents/earhart/126_PI_154_1-O_Box14_9_12_21_Equatorial_Islands_Aviation_1936_1951.pdf

 

[2] The Radio Amateur’s Handbook. Newington, CT, page 364.  American Radio Relay League. 1936. Available at: http://www.tubebooks.org/Books/arrl_1936.pdf

 

[3 Leo Bellarts’ comments were made in an interview conducted in 1973 by Earhart researcher and author Elgen Long. This part of what Bellarts said was incorporated into a mock interview that podcaster Chris Williamson constructed from excerpts of audio recordings of Long’s interview with Bellarts. Chris Williamson’s mock Bellarts interview can be accessed at: https://podcasts.apple.com/us/podcast/khaqq-calling-itasca-a-conversation-with-leo/id1645810327?i=1000579875388

 

[4] Separate logs were kept for two positions in the Itasca radio room. For the purpose of this blog post I refer to the radio room position #2 log typed in real time, i.e., the ‘raw log’. Digitized scans of this log can be found at: https://www.archives.gov/college-park/highlights/earhart-log

 

[5] Note that Amelia heard Itasca’s Morse code ‘A’s using her RDF loop antenna on top of the plane. The loop antenna was separate from the vee antenna Amelia was relying on to hear Itasca’s 3105 kilocycle voice messages.

 

[6]  The 3105 Donut. In TIGHAR TRACKS, Volume 24, #4, October 2008, page 3. https://tighar.org/Publications/TTracks/2008Vol_24/1008.pdf  

 

[7] TIGHAR Forum, 3105 Donut thread: https://tighar.org/smf/index.php?topic=285.0

 

[8] NEC Based Antenna Modeler and Optimizer. https://www.qsl.net/4nec2/

 

[9] General Information on the VOACAP Propagation Prediction Model. https://www.voacap.com/2023/itshfbc-help/icepac-general.html

 

[10] Introduction to Radio Equipment. U.S. Navy Training Courses, Edition of 1946. Chapter 21, page 22. U.S. Government Printing Office, Washington, D.C. (1946). Chapter 21 can be accessed at: https://maritime.org/doc/radio/index.php

 

[11]  Itasca Deck Log: https://media.defense.gov/2022/Mar/14/2002955993/-1/-1/0/370702-G-CG969-1001.JPG

 

[12] The cloud cover photo appears about twenty seconds after the start of this video: https://vimeo.com/469779444 Notice that the Itasca was producing smoke as a potential visual aid to Earhart. Meteorological conditions were such that the smoke stayed near the sea surface, probably making it less useful as an aid to spotting Howland Island.

 

[13] POV: Howland Island. Waitt Institute YouTube channel: https://www.youtube.com/watch?v=m9c3yZ0xeHw

 

[14] https://tighar.org/smf/index.php?topic=285.msg5767#msg5767  Note that I’ve corrected an error in Bob’s word choice, as pointed out by Chuck in the discussion thread. Bob meant to write ‘electric field’, not ‘electrostatic field’. Bob acknowledges the error later in the thread.

 

[15] The Search for Amelia Earhart’s L-10E Electra: An HF Systems Engineering Approach. Tom Vinson, NY0V.  Brazos Valley Amateur Radio Club (BVARC) YouTube channel: https://www.youtube.com/watch?v=bdA6bm9tL8M

 

[16] Nauticos LLC Website: https://nauticos.com/

 

[17] Rockwell Collins Wikipedia page. https://en.wikipedia.org/wiki/Rockwell_Collins

 

[18] Meet the Engineer on a 26 Year Journey to Find Amelia Earhart’s Plane. Amelia Rose Earhart YouTube channel: https://www.youtube.com/watch?v=pwYo1kq89xc

 

[19] Episode 1: How to Find Amelia Earhart with Nauticos Radio Engineer Rod Blocksome. Amelia Rose Earhart YouTube channel: https://www.youtube.com/watch?v=yN4A-ENyChU&t=28s

 

[20] Photo of Beechcraft D18S Twin Beech at the National Air and Space Museum. https://airandspace.si.edu/collection-objects/beechcraft-d18s-twin-beech/nasm_A19761792000

 

[21] Photo of Amelia’s Lockheed Model 10E Electra in the San Diego Air & Space Museum collection. Catalog number 01-00091572. Viewable online at: https://www.thisdayinaviation.com/amelia-earharts-lockheed-electra-10e-special-nr16020/

 

[22] A Lost Sailor or Amelia Earhart? Lost Norwich City Crewman: Potential Sources of Human Remains Discovered in Gardner Island (Now Nikumaroro Island) in 1940. This paper can be downloaded at Academia.edu. https://kentonspading.academia.edu/research#papers

Is 2-2-V-1 a Piece of the Sydney Island Wreck? Update #5

In my last post I mentioned that I was unable find any measurement data on TIGHAR’s web site corroborating TIGHAR’s often-made claim that 2-2-V-1 is piece of .032 inch thick of 24ST aluminum sheeting. Ric Gillespie recently remarked on TIGHAR’s online forum that three sources had examined 2-2-V-1 and found it to be .032 inches thick [1].  When I checked reports these sources had submitted to TIGHAR [2,3,4], however, I couldn’t verify that what Gillespie said was actually true. Some of the reports do say that 2-2-V-1 is .032 inches thick, but they seem to be restating the thickness value reported by TIGHAR rather than reporting their own measurement results.

A long-time TIGHAR observer who read my post pointed me to a document on the TIGHAR web site that does provide an actual thickness measurement for 2-2-V-1. In 2015, Lehigh Testing Laboratories in New Castle, Delaware performed chemical and mechanical tests on several specimens of aluminum sheet submitted by TIGHAR, including a specimen cut from 2-2-V-1 [5]. A table on page five of the report listing the measured dimensions of each specimen reports that the 2-2-V-1 specimen was found to be .030 inches thick.

This measurement result raises the possibility that 2-2-V-1 is not a piece of .032 inch 24ST aluminum sheet as TIGHAR has long supposed it to be, but instead that it is a piece of .028 inch 24ST aluminum sheet. Note that I’m talking here about nominal thicknesses, not actual thicknesses, of aluminum sheeting. As discussed in my last post, at the time that artifact 2-2-V-1 was produced the thickness tolerances for Alclad 24ST sheet were .0025 inches for both nominal thicknesses, .028 and .032 inches. This means that nominal .028 inch 24ST aluminum sheet manufactured back then might actually be as much as .0305 inches thick, and nominal .032 inch 24ST aluminum sheet might be as little as .0295 inches thick. A thickness of .030 inches falls within the acceptable ranges for both nominal thicknesses and so it isn’t possible to know from this one measurement whether 2-2-V-1 was fabricated from a piece of .028 inch or .032 inch aluminum sheet.

The same person who pointed me to the Lehigh Laboratories thickness measurement pointed me to a thread on TIGHAR’s online discussion forum, excerpted below, in which Ric Gillespie reported his own thickness measurement results for 2-2-V-1 [6].  This part of the discussion was prompted by TIGHAR’s announcement that Lehigh Testing Laboratories would be analyzing specimens from 2-2-V-1.

++++++++++++
TIGHAR announcement quoted by forum member:  Lehigh Testing Laboratories, Inc. was recommended to TIGHAR by Prof. Eager at MIT as the best lab for trying to learn more about 2-2-V-1 (and other aluminum artifacts) through materials analysis.  By pure coincidence, LTL is located in Wilmington Delaware about 45 minutes from TIGHAR HQ. I contacted LTL and dropped Tom Eagar's name.  They responded with enthusiasm and today we set out a program of testing that should give us the answers to a number of important questions.  Those answers could confirm our fondest conclusions about 2-2-V-1 or they could blow them out of the water... I won't try to list the alphabet soup of technologies LTL will be using to do this work.   They're doing this pro bono because such is TIGHAR's reputation in the scientific community that they consider it an honor to be asked to help with our investigation. What they'll be doing represents thousands, perhaps tens of thousands, of dollars worth of work if hired at commercial rates.  The results we'll get will be scientifically sound, whether or not they're what we want to hear.

Forum Member:  Would it be possible to have LTL confirm the gauge of the aluminum?  I believe the NTSB is the only one that measured it at .032, and I think at least one of their other measurements was slightly off (the convergence of the rivet rows?).  It would be helpful to know for sure what the gauge is.  Just a suggestion.

Gillespie:  Your belief is in error. The thickness of the sheet has been checked many times by many people.  All it takes is a micrometer.

Forum Member:  Alas I have no micrometer at my disposal at the moment nor access to the piece and am therefore dependent upon the written reports of others in this regard.  Could you please direct me to a source(s) that describes and documents any of the many gauge measurements that were made (other than the NTSB report already mentioned), preferably one that includes a description of the tolerance levels involved.  Thank you.

Gillespie:  Okay.  Here's a written report especially for you. I have a micrometer and I have access to the piece and I have measured it numerous times. It's .032".  You can choose not to believe me and wait for the LTL report which will include much more detail.
++++++++++++

Gillespie’s answer may be flippant, but it nevertheless contains some useful information. His belief that 2-2-V-1 is a piece of .032 inch 24ST aluminum sheet is based at least in part on multiple thickness measurements that he himself made using a micrometer. But remember what Gillespie said in questioning Tom Palshaw's micrometer skin thickness measurements on the NEAM C-47 wing [7]:

“I know from experience that measuring skin thickness with a digital micrometer is difficult and frustrating.  Measure three times and get three different answers. With a little confirmation bias you can get any result you want.”

Tom’s follow-up measurements made using an ultrasonic thickness gauge show that Tom’s original micrometer measurements were correct. I think it’s fair to ask whether Gillespie’s own thickness measurements on 2-2-V-1 might have suffered from confirmation bias. When Gillespie made his measurements, what sheet thicknesses did he think 2-2-V-1 might have been manufactured from? Current aluminum product catalogs like the one reproduced below don't list a sheet thickness between .025 and .032 inches for 2024-T3 Clad aluminum, the modern designation for 24ST Alclad. As discussed in my last post, Alcoa product literature from the early 1940s list .028 inches as an available thickness for Alclad 24ST sheet. Did Gillespie know that when he made his measurements? I wonder if Gillespie arrived at his firmly stated .032 inch thickness for 2-2-V-1 through the following process:

1. Gillespie measures 2-2-V-1’s thickness several times with a micrometer; his measurement results cluster around some value between .032 inches and .025 inches, but closer to the former value;
2. Gillespie knows that Alclad 2024 aluminum sheet is currently produced in nominal thicknesses of .025 and .032 inches but is unaware that decades earlier it was produced in a nominal sheet thickness of .028 inches;
3. Since Gillespie’s micrometer measurement results for 2-2-V-1 are closer to .032 inches than to .025 inches, he concludes that 2-2-V-1 is a piece of 24ST Alclad sheet of .032 inch nominal thickness.

Thyssennkrupp Aluminum Stock Guide [8]

If any readers out there know of TIGHAR data that convincingly settles the matter of whether 2-2-V-1 is a piece of nominal .028 inch of .032 inch aluminum sheet, please let me know where to find it. If there is no such data — and I suspect that there isn’t — TIGHAR needs to arrange to have quality measurements made to try to settle this question.  Tom Palshaw noted in his most recent report update that factors such as corrosion, shape, installed fastener effects, and stress induced changes to original dimensions would need to be carefully considered in making thickness measurements on a beaten up, weathered piece of aluminum sheet like 2-2-V-1, and so this is not a job for an amateur wielding a micrometer, with all due respect to Ric Gillespie.

Why is it important to accurately characterize 2-2-V-1’s thickness? Tom Palshaw has shown that 2-2-V-1’s rivet pattern matches the rivet pattern on the upper wing of a C-47 at the New England Air Museum (NEAM). It is reasonable to think that the rivet pattern match would extend to many other C-47s, including the C-47 that crashed on Sydney Island. Tom has confirmed through ultrasonic thickness gauge measurements that NEAM C-47 wing skin is .032 inches thick at the location of interest, the same thickness that TIGHAR has long reported for 2-2-V-1. C-47 repair manuals indicate that the skin of a C-47 should be .028 inches thick at the location of the rivet pattern match, and so it is not a sure thing that the Sydney Island C-47 would also have a seemingly non-standard .032 inch skin thickness at the matching location. On this basis Ric Gillespie can say that the rivet pattern match to the NEAM C-47 wing is a “crazy coincidence”,  and TIGHAR’s claim that 2-2-V-1 is a piece of Amelia Earhart’s Electra can perhaps limp along, badly wounded but not dead. But the .030 inch thickness value measured by Lehigh Testing Labs suggests that if 2-2-V-1 might really a piece of nominal .028 inch 24ST Alclad.  If careful thickness measurements show this to be true, there would be no reason left to think that 2-2-V-1 wasn’t a piece of the wing of the Sydney C-47 wreck. The match in rivet patterns between 2-2-V-1 and the NEAM C-47 wing makes it reasonable to think that the Sydney C-47 wing matched 2-2-V-1’s rivet pattern at the location of interest. If 2-2-V-1 is nominal .028 inch aluminum sheet then it matches the thickness for a C-47 wing at the matching location given in repair manuals. There would be no physical attribute of 2-2-V-1 that can't be matched to a C-47 wing. The claim that 2-2-V-1 is a piece of Amelia Earhart’s Electra would truly be dead.

Comments, corrections, additional relevant facts, differing viewpoints, etc., are welcome.  Send to gardnersghost@gmail.com
+~+~+~
Footnotes/References
[1] https://tighar.org/smf/index.php/topic,2074.msg43572.html#msg43572
[2] (NTSB) https://tighar.org/Projects/Earhart/Archives/Documents/NTSB_Report/ntsbreport.html
[3] (MMR) https://tighar.org/Projects/Earhart/Archives/Research/ResearchPapers/Artifact2-2-V-1/MMR%20Project%20No.%20128927%20Report.pdf
[4] Letter from MIT Professor Thomas Eagar to Ric Gillespie https://tighar.org/wiki/2-2-V-1#cite_note-3
[5] The Lehigh Testing Laboratories Report can be downloaded at: https://tighar.org/Projects/Earhart/Archives/Research/Bulletins/75_Findings2-2-V-1/75_Report_of_Findings_2-2-V-1.html
[6] https://tighar.org/smf/index.php/topic,1490.msg35664.html#msg35664
[7] https://tighar.org/smf/index.php/topic,2074.msg43575.html#msg43575
[8] http://www.tkmna.com/wcm/idc/groups/internet/documents/web_content/mdaw/mdiy/~edisp/d_004646.pdf

Is 2-2-V-1 Piece of the Sydney Island C-47 Wreck? Update #4

Since my last post on the topic of Tom Palshaw’s report, there have been new developments that are worth discussing. In a recent post on TIGHAR’s online discussion forum [1], Ric Gillespie rejected the report’s conclusion that TIGHAR Artifact 2-2-V-1 is a piece of the wing of a C-47 transport plane that crashed on Sydney Island. Gillespie writes:

“TIGHAR Artifact 2-2-V-1 did not come from a C-47.  The Wing Plating diagram in the C-47 structural repair manual shows the skin thickness in the entire area where the rivet pattern allegedly matches 2-2-V-1 is .028".  The artifact's skin thickness is .032”

Tom Palshaw measured the skin thickness on the C-47B wing at the New England Air Museum as .032 using a micrometer at the edge of the skin.  .004" is an easy error to make. It's a small but important discrepancy.  The NTSB lab, Professor Eager at MIT, and the Massachusetts Materials Research metallurgical lab all measured 2-2-V-1 as .032”.

The C-47 manual dates from September 1942 and was updated in 1945.  There is no indication that the basic structural components of the wing were changed. With this new information, the rivet pattern on the C-47 wing, although remarkably similar to the artifact, becomes another of the crazy coincidences we sometimes encounter.”

In response to this critique, Tom Palshaw has updated his report with new skin thickness measurements on the NEAM C-47 wing at the location of interest. These measurements were made by a level 2 nondestructive testing technician from Bombardier Business Aircraft Service Center using an ultrasonic thickness gauge. I take it that this technique is not prone to sort of errors that Ric Gillespie suggests might have afflicted Tom’s original measurements.  Five ultrasonic thickness gauge measurements yielded skin thicknesses of 0.032, 0.031, 0.032, 0.031, and 0.032 inches. These new measurements confirm that the skin of the NEAM C-47 wing at the location of interest is the same thickness that TIGHAR has reported for 2-2-V-1.

 Wing location matching 2-2-V-1 is marked in yellow (Markup by Palshaw of a DC-3 repair manual; See Ref. 8)

As Ric Gillespie points out, C-47 repair manuals indicate that the skin of the NEAM C-47 wing should be .028 inches thick at the location of interest. It isn’t clear why the NEAM C-47 wing’s skin at the location of interest is thicker than repair manuals indicate it should be. Damaged aircraft skin is often replaced with sheeting in the next higher thickness produced by manufacturers, which in this case would be .032 inches thick. Tom tells me he’s sure that at the location of interest the C-47 wing skin is original factory-installed Alclad 24ST aluminum alloy sheeting, not a later repair. For what it’s worth, I’ll note that .028 inch 24ST aluminum alloy sheeting is not listed as a standard size normally carried in stock in 48 inch widths in the Alcoa's Aluminum in Aircraft published in 1941.  Would sheeting of that width be needed for fabricating the skin at the location of interest on the C-47 wing? If so, perhaps this is a clue that should be pursued in understanding why the NEAM C-47 wing skin is .032 inches thick at the location of interest. Something else that suggests .028 inches is not your typical aluminum alloy sheet thickness: the 1943 version of Aluminum in Aircraft has only one mention of .028 inch aluminum -- in a table of sheer strengths of spot welds for various metal thicknesses

Aircraft sheet metal thickness table, Aluminum in Aircraft, Published by Alcoa in 1941

Spot weld shear strength table, Aluminum in Aircraft, Alcoa 1943

As things now stand, Tom Palshaw has shown that a location on the wing of the NEAM C-47 closely matches 2-2-V-1 in terms of rivet line spacing, rivet pitch, irregularities in pitch of its -5 rivets, and that it has the same .032 inch thickness that TIGHAR has reported for 2-2-V-1. The NEAM C-47 wing and 2-2-V-1 match in every physical attribute determined so far.  From the standpoint of physical dimensions the match between 2-2-V-1 and the NEAM C-47 wing has been better established than the match between 2-2-V-1 and Amelia Earhart’s Electra. As discussed in my last post, past TIGHAR efforts have failed to demonstrate that 2-2-V-1 fits within the boundaries TIGHAR defined for the putative source location on Earhart’s airplane, the window patch installed at Miami in 1937.

The letters ‘AD’ that are faintly visible on the artifact’s surface are another feature that is consistent with the idea that 2-2-V-1 is a piece of the Sydney Island C-47. To help factory workers identify aluminum sheet stock, Alcoa applied material markings to the aluminum sheeting it produced. The Aluminum Markings web site documents a clear trend in how Alcoa’s material markings for Alclad 24ST aluminum sheeting that aircraft skins are made of changed over time.  Photos of aircraft manufactured before 1942, including Amelia Earhart’s Lockheed Electra, consistently show the marking to be  ‘ALC 24ST’.  Photos of aircraft manufactured from 1942 onward predominantly show the marking to be  ‘ALCLAD 24 ST’, though some photos show aluminum aircraft parts marked ‘ALC 24ST’.  So the photographic evidence —a lot of it— indicates that by early 1942,  had Alcoa started to mark 24ST Alclad sheeting it manufactured with a new material marking, ‘ALCLAD 24ST’. This line of evidence therefore tells us that the window patch attached to Earhart’s Electra in Miami in 1937 could not have the letters ‘AD’ seen on 2-2-V-1 because because Alcoa didn’t start using the ‘ALCLAD 24 ST’ material marking for several more years.

Remnant of 'AD' on Artifact 2-2-V-1. See reference [5] for source

B-25 factory photo, circa 1942. Note 'ALCLAD 24S-T' markings to left of female worker

Underside of Earhart's Electra, circa 1937. The material marking reads 'ALC 24ST'. Full photo at upper right

The similarities between the NEAM C-47 wing and 2-2-V-1 are just too great to be dismissed as a ‘crazy coincidence’.  The problem with the assertion that the wing of the Sydney Island C-47’s can’t be the source of 2-2-V-1 because repair manuals indicate that skin at the location of interest shouldn’t be .032 inch thick is that an actual C-47 wing, the one at NEAM, shows that C-47 wings were not always manufactured with the skin thicknesses indicated in repair manuals.

The right thing for TIGHAR to do, given the evidence Tom Palshaw has produced, is gather evidence that might shed further light on the Tom’s conclusion that 2-2-V-1 is a piece of wreckage scavenged from the Sydney C-47. To misrepresent a piece of the Sydney C-47 wreck site as a piece of Amelia Earhart’s Lockheed Electra would be a disservice to those who died in the Sydney C-47 crash and to the crew of the Electra. Surely TIGHAR would want no part of such an abomination.

TIGHAR should carefully document how well the rivet pattern on the NEAM C-47 wing matches 2-2-V-1. Laying 2-2-V-1 over the NEAM wing at the location of interest and documenting the match with photographs would take no more than ten or fifteen minutes to do. While at NEAM to conduct this comparison, TIGHAR can also demonstrate whether or not 2-2-V-1 fits within the boundaries defined for the dimensions of the Miami window patch in a previous TIGHAR study [2]. This too would be a simple exercise that would only take minutes to carry out. These comparisons should be done with the full cooperation of Tom Palshaw. I find it odd that Ric Gillespie visited NEAM only a few weeks ago to conduct measurements on the NEAM Electra but gave Tom Palshaw no advanced notice of his visit.

Another basic piece of information that needs to be better documented is 2-2-V-1's thickness. Although TIGHAR has repeatedly stated that 2-2-V-1 is .032 inches thick, I can find no clear information about how 2-2-V-1’s thickness was determined. A few posts further along the discussion thread in which Ric Gillespie dismisses Sydney Island C-47 as the source of 2-2-V-1, he comments that “The NTSB lab, Professor Eager at MIT, and the Massachusetts Materials Research metallurgical lab all measured 2-2-V-1 as .032” [3]. I’ve reviewed all three of these sources [4,5,6] and as best I can tell, 2-2-V-1’s thickness was not actually measured by NTSB, MIT or the metallurgical lab. Where the thickness of 2-2-V-1 is stated in these sources the authors appear to simply be stating the artifact thickness reported to them by TIGHAR. There certainly is no description in any of these sources of a measurement method or a presentation of actual measurement data.

Surely the 0.32 inch thickness TIGHAR has reported is based on an actual measurement, but how that measurement was made and what the actual measurement result was has never been reported, as far as I know. Did Ric Gillespie make these measurements using a micrometer? If so, then the claim that 2-2-V-1 is .032 inches thick is based on the same measurement technique that Ric Gillespie has said could easily be in error by .004 inches. Gillespie in a later post on the same TIGHAR discussion forum states: “I know from experience that measuring skin thickness with a digital micrometer is difficult and frustrating.  Measure three times and get three different answers. With a little confirmation bias you can get any result you want” [7]. Here it is worth quoting something Tom wrote in his latest update on the NEAM C-47 skin thickness measurements:

“Measuring the thickness of a new sheet of ALCLAD is as simple as using a standard 1" micrometer. The metal is smooth, flat, and clean. Once the metal has been installed, or exposed to the elements, several factors can affect the accuracy of the measurement. These include a paint coating, corrosion, shape, installed fastener effects and stress induced changes to its original dimensions.”

2-2-V-1 is not a fresh new sheet of aluminum alloy, it is a beaten up sheet of aluminum alloy that undoubtedly has been exposed to the elements for many years. The Massachusetts Materials Research Laboratory report on 2-2-V-1 describes it thusly:

“Overall, the artifact presented a grey, oxidized aluminum appearance with isolated regions of buff-colored calcareous deposits and thin, green discolorations. Both the buff and green deposits were reported to have been previously tested and found to be consistent with coral and algal growths”

Note that before making his new measurements on the C-47 wing, Tom gently removed accumulated surface coatings from his measurement locations using a Scotchbrite pad. Were surface deposits accumulated in the years of exposure of 2-2-V-1 to the elements removed before its thickness was measured? Tom points out that at the time artifact 2-2-V-1 was produced, the thickness tolerances for Alclad 24ST sheet was .0025 inches for manufactured sheeting in either the .028 or .032 inch thickness. For all we know, 2-2-V-1’s actual thickness falls within the range falls within the range for nominal .028 inch sheeting, but TIGHAR has erroneously reported it to be .032 inches thick. For example, if 2-2-V-1 was nominal .028 inch sheeting whose actual manufactured thickness was .029 (well within the manufacturing tolerance) with a .001 inch thick coating of weathering/coral/algal residue, and a .001 inch micrometer measurement error was made, the measured thickness would have been .031 inches (.029+.001+.001=.031). The individual making the measurement might have  concluded that 2-2-V-1 was .032 inch thick since this was the closest nominal Alclad 24ST sheet thickness to the measurement result.

Clearly then another thing TIGHAR must do is document what 2-2-V-1’s thickness was when manufactured. Given the .0025 inch manufacturing tolerance, it may not be possible to definitively exclude that 2-2-V-1 was either nominal product thickness, .028 or .032 inches. The determination of 2-2-V-1’s original manufactured thickness is yet another activity that TIGHAR could and should carry out in collaboration with Tom Palshaw. The effects of factors Tom mentions in his update such as shape, installed fastener effects, and stress induced changes to original dimensions would need to be carefully considered.

It would also be useful to accumulate more information about rivet patterns and skin thicknesses of surviving C-47s and DC-3s in museum and private collections. Tom’s update includes ultrasonic measurements of the thickness of the skin of the NEAM DC-3 at the location of interest [8], and there the measurements indicated a skin made of .028 inch nominal sheeting. Accurate skin thickness measurements made on other C-47s and DC-3s might reveal a pattern relevant to the question of whether the Sydney C-47 is the source of 2-2-V-1.

Comments, corrections, additional relevant facts, differing viewpoints, etc., are always welcome.  Send to gardnersghost@gmail.com
+++
Footnotes/References

[1] https://tighar.org/smf/index.php/topic,2074.msg43572.html#msg43572
[2] Earhart Project Research Bulletin #73, The Window, the Patch, and the Artifact.  https://tighar.org/Projects/Earhart/Archives/Research/Bulletins/73_StepbyStep/73_Step_by_Step.html
[3] See the TIGHAR discussion forum post at reference 1
[4] https://tighar.org/Projects/Earhart/Archives/Documents/NTSB_Report/ntsbreport.html
[5] https://tighar.org/Projects/Earhart/Archives/Research/ResearchPapers/Artifact2-2-V-1/MMR%20Project%20No.%20128927%20Report.pdf
[6] Links to the MIT emails can be found at https://tighar.org/wiki/2-2-V-1#cite_note-3
[7] [https://tighar.org/smf/index.php/topic,2074.msg43575.html#msg43575]
[8] DC-3s and C-47s are to a great extent the same airplane, the former manufactured for use as a civilian airliner while the latter was made to be a military transport. The wing diagram above comes from a 1940s-vintage DC-3 repair manual