P-MAN VIII Final Report


NEW: Finding, identifying and documenting one new American crash site (a partial TBM-1C Avenger crash site, unknown BuNo) in a mangrove swamp on the east coast of Peleliu (Target 6).

Some will say this was the easiest crash site we have ever found, since on the day we went to look for it, we found it in the jungle right where we thought it would be within an hour of our starting. Upon reflection however, we probably spent more time on finding this site, before we ever started hiking, than any other (with the exception of B24 '453 - see P-MAN VI and P-MAN VII reports).

While investigating the crash site of another Avenger on Peleliu (see ATTACHMENT 8 below), the team knew from prior archival searches that a second Avenger had gone down on Peleliu somewhere. In fact, it was only after we completed our search for the first Avenger's debris field that we knew the second Avenger was somewhere else on Peleliu and not superimposed with the first: during P-MAN VII we confirmed that there must be a second separate Avenger crash site.

Upon return to the US, we began intensive research of the archival records.

A. Background for Search

During the past fourteen years, the members of the BentProp Project have studied the records in Washington, DC at the US National Archives and Records Administration to determine how many American aircraft were lost throughout Palau and, in particular, in the Peleliu area. From our research, we have determined that four American aircraft, not yet discovered, crashed on Peleliu in combat during the period of March to September 1944, all of which are associated with MIAs - meaning that their remains were never recovered. Other aircraft associated with MIAs may also have been lost around Peleliu for which records as yet have not been located. Such aircraft, because of their associated MIAs are separate from the large mixed collections of Japanese and American aircraft debris, not associated with MIAs, which exist on Peleliu in at least six locations (see map in ATTACHMENTS of all aircraft sites identified by the BentProp Project) .

These four aircraft and their crews include:

  • TBM1C Avenger from VT-51, USS San Jacinto, 15 September 44
  • SB2C1 Helldiver From VB-2, USS Hornet, 30 March 44
  • F6F5 Hellcat from VF-13, USS Franklin, 10 September 44
  • F6F5 Hellcat from VF-13, 10 September 44

From our research, the two F6F Hellcats appear to have crashed near the Peleliu runway and may have been salvaged during or immediately after the battle, by either Japanese or American forces. Large boneyards still exist on Peleliu and we cannot help but wonder if one of these is the final resting place for one or more of these aviators. While finding these two Hellcats and their aviators is unlikely, our search continues.

Of the other two aircraft, only the Avenger, with a crew of three, was seen to crash onto Peleliu. The specific location of the Helldiver's loss was not reported. The squadron's After Action Report states two parachutes were seen to open and land in the water near the eastern shore of Peleliu - Peleliu elders convey stories of seeing two parachutes during that time, as well as these Americans being captured, although their fate remains unknown. This is still under our investigation and this crew remains on our MIA/POW list.

The Avenger was reported to crash in AREA 130 (Gunnery Target Map for Peleliu) which lies on the eastern side of the North-South Runway. Based on date, time and location, this same Avenger's crash was recorded on both still camera and motion picture film which still exist at the US National Archives and Records Administration. This photographic and film footage evidence has been located and reviewed by members of the BentProp Project. Analysis of the photographic/film evidence predicted a possible search area on eastern Peleliu, most likely in the mangroves. [NOTE: The BentProp Project has a crew list and Bureau Number for this aircraft but because this is under investigation as a possible MIA site, we have agreed with JPAC not to release the crew names or aircraft identifier. This crew was from the same squadron as then-Ensign, and future US President, George Herbert Walker Bush. However, having been shot down over Chichi Jima a few days before, ENS Bush did not participate in Operation STALEMATE II. In 2003, several members of the BentProp Project had the honor of privately meeting and discussing our work in Palau with former President Bush.]

Briefly, evidence of this crash and its approximate location on Peleliu includes:

  • After Action reports from the USS San Jacinto and Squadron VT-51 (US Navy Torpedo Bomber Squadron-51)
  • Still Photographs
  • Motion Picture Film Footage
  • Interviews with living VT-51 veterans on that mission.

Courtesy of R. Stinnett

Left: Ensign George H. W. Bush in VT51 Avenger, 1944. Courtesy of R. Stinnett           
Right: Former President George H. W. Bush with Pat Scannon, 2003 © T. Krasny, 2006

Peleliu Airfield, Post-invasion, June 1945 Under American Control
NOTE: The US Military extended the runway 2061 feet and built east road along runway


The members of the BentProp Project respectfully requested and received permission from the Governor of the State of Peleliu and the Palauan Historical Preservation Office to do the following:

Conduct a search mission east of the East Road for the TBM1C of … (redacted), lost in combat 15 September 44 along east side of … (Peleliu, redacted) near mangroves. This mission would include a) interviews of Peleliu elders to determine if the crash site has been seen by Palauans after WWII ended and b) searching east of … (NOTE: This predicted search zone might be adjusted a little as a result of input from our interviews of the Peleliu elders).

If the BentProp Project team locates an Avenger debris field, we then propose to review the area to determine if the Avenger wreckage is a single crash site (which would suggest this is … aircraft) or part of a larger debris field with other aircraft (which would suggest this area is another aircraft junk yard and less relevant). As part of this field review, we will conduct a search for a US Navy Bureau of Aeronautics aircraft number (on tail of aircraft or in cockpit areas), which would identify the aircraft and crew. We will also search for evidence of specific squadron markings on the wreckage for VT-51, … . As part of this review, the Team also, as appropriate, will a) collect relevant GPS coordinates, b) map the debris fields and c) photograph key findings for documentation purposes.

The BentProp Project team estimated between one and two days to complete the terms of this proposal.

The BentProp Project also proposed to write a report before leaving Palau (17 March 06) to share all of our findings with Governor Jackson (Ngiraingas), the Palauan Historical Preservation Office, U. S. Joint POW/MIA Accounting Command (JPAC) and the US Naval Historical Center (NHC).



B. Findings

The team traveled by boat from Neco Marine and landed on the north Dock. Tangie drove us then to the general area. The team split into three search patterns from the East Road:

  • Joe went 50-100 meters E of suspected area and headed S
  • Pat and Reid started N of suspected area, in jungle and headed S
  • Dan and Mike started S of suspected area into the jungle and headed N.

Both sides of the East Road were explored. Dan and Mike rapidly found an aircraft debris field off to the the East and the team then re-assembled near that site. The first question to answer was: was this a combat loss or a boneyard of mixed debris?

While most of the team began searching the immediate area around the debris field, Pat and Joe continued searching to the Southeast from the aircraft field and to the West across the East Road. This led to discovery of a collection of mixed aircraft parts 120-80 meters to the Southeast of the above aircraft field. No other aircraft fields were found inside of the 50 meter radius perimeter of Dan's and Mike's discovery. Thus the field within that crash site appears NOT to be contaminated by other adjacent aircraft parts.

Because the team is limited to superficial investigations only, out of respect for these archeological sites, our investigation must be considered preliminary until appropriate investigation by JPAC can be made. This situation is similar to the FM-2 Wildcat site we found during P-MAN VI, which remains closed pending JPAC evaluation. The Team's preliminary assessment of the superficial area of the aircraft field around is characterized by the following observations:

  • This field is made up of a substantial, but incomplete, portion of a single Avenger Torpedo Bomber, which held a three man crew.
  • A combination of the Team's ground search and scanning, using a Minelab1000 Excalibur metal detector, indicates that additional metal may be found underneath this immediate area, although no attempts were made to unearth them.
  • The field consists of three major adjacent sub-fields (see field drawing in ATTACHMENT): a) engine with main landing gear #1, b) main landing gear #2 and c) outer port wing - all within a 25 meter diameter; several additional smaller sub-fields also lie within this diameter. [NOTE: the two main landing gear assemblies, the outer leading edge slot on the wing and the engine assembly with Hamilton-Standard propeller are only consistent with an Avenger.]
  • Canopy aluminum framework, with fragments of plexiglass, was found between the engine and the second main landing gear.
  • Parts consistent with the rear .50 cal machine gun turret were found between the second main landing gear and the wing.
  • Although several rounds of expended and unexpended .50 cal and .30 cal ammunition were found within this field, none of the three .50 cal nor the .30 cal machine guns were found in the ground search.
  • The tail wheel assembly, a substantial collection of electrical equipment and other debris (including a portion of the engine) were found lying on top of the port wing
  • … (redacted)
  • This field (contrasting with Avenger 19656 recovery operations in 2005) lies within a small area (25 meter diameter) and is consistent with a vertical crash (known to have occurred for the avenger in question).
  • No empennage, which would have the painted USN BuNo, was located from the ground search (possibly the tail separated from the fuselage, which would be consistent with the vertical crash). Likewise, no cockpit BuNo identification plate was located in the ground search.
  • The field does not appear to be contaminated by parts from any other aircraft.

To gain a more detailed sense of this field, the three sub-fields are characterized by the following observations. In the discussion, note that when objects are referenced by a "letter" designation, e.g., "A," the letter refers to a hand-drawn map of the debris field that appears farther down on this page.

  1. Engine with Main Landing Gear #1: The R2600 engine, used in Avengers, is the most visually obvious aspect of the entire field. The cylinder stack lies approximately vertically, with several of the 14 cylinder heads (two rows of seven) missing [NOTE: some found lying within the 25 meter crash diameter]. The forward crankcase housing is shattered and bent forward, along with the propeller shaft, approximately 30º [NOTE: part of this crankcase housing with an attached aluminum plate was found near the wing (see below)]. The Hamilton-Standard propeller remains attached to the engine via its shaft, with two bent blades present on or near the ground, while the third blade (in the most superior position) is sheared off at its base and is missing. As one faces the front of the propeller, the left propeller blade's outer tip (approximately 2 feet) is snapped off and lies separately near the engine. Both blade tips (the one on the right blade remaining intact) retain remnants of yellow paint. The propeller spinner is fractured open but at least one spinner piece lies separately nearby. This spinner has a yellow painted stripe or band at its base, approximately 4 inches wide (NOTE: This may be a VT-51 squadron marking - see below). As one looks at the front of the propeller, along side and in front of the right propeller blade, lays a partially complete main landing gear assembly (designated #1). Gears, aluminum and other debris lie around the rear base of the engine.


  2. Main Landing Gear. #2: This assembly, characteristic for a TBM Avenger, lies 2 meters to the west of the engine. This assembly appears more complete than #1 and includes the wheel and brake assembly onto which the wheel was inserted. Tire debris (labeled "A") with a diamond tread pattern lies immediately along the southern edge of the landing gear on the jungle floor. Two meters to the north of this landing gear lays a pile of debris with gear and tooth assembly (Labeled "B"). We believe this is part of the gun turret used to rotate the turret. In this same immediate area, lies a panel of aluminum with piano wire (labeled "B1"), which may be part of the bomb bay door. Three meters east of this area lays a small debris pile with an aileron tab, labeled "D". Approximately 2 meters to the Southeast of "D" lays a bent aluminum framework, labeled "C", consistent with being part of the canopy; fragments of plexiglass remain in the grove of the frame.

    6 photos above © R. Joyce, 2006

  3. Outer Port Wing: This third sub-field is perhaps the most complex area, especially for a limited ground search. The metal detector suggests more metal may lie under the ground in this area. The Avenger has a folding outer wing and this wing section appears to be what remains of the port outer wing. It lies with its ventral surface up (painted white) and blue-painted aluminum can be seen on the dorsal side from its jagged edges on either end. Along the Southeastern edge of the wing can be seen the ventral aspect of the leading edge slot, labeled "G". A square (~5 inches in each edge) hole, labeled "I", can be seen in the middle of the wing, with some writing. This is the result of the radar antenna being ripped from the wing. The antenna shaft still lies in this area. On the northern end of this wing lies a large pile of mixed electrical and hydraulic debris, labeled "E". The hard rubber tire (used on aircraft carriers) on the tail wheel assembly (labeled "F"), also characteristic of a TBM Avenger, lies on top of the wing. The upside down wing tip, labeled "H" also lies on topmost part of this pile of debris. A piece of the forward crank case housing from the engine area was found immediately along the west side of the wing. This piece has an aluminum plate still intact and attached. The surface of the plate is blackened and pitted in the middle (? melted) but letters and numbers can still be seen - suggesting that with appropriate cleaning, some information concerning the engine might be determined. Also in this same area (western edge of wing), a spoon, stamped with "USN" on the front side of the handle and "STAINLESS STEEL" on the back of the handle, was found.


  4. Additional Parts (labeled "L" through "Q"): As noted on the drawing of the field (below), a number of other smaller parts lie near or outside the three sub-fields. However, the parts actually found are small (except for the wing folding hinge at "H") and none offer any additional identification that we observed.
  5. Key Parts NOT Found During Ground Search: Some key parts not found include 1) any component of the empennage (with the BuNo), 2) much of the fuselage and cockpit (with BuNo), 3) any armament, and 4) the starboard wing.

C. Summary and Working Conclusions:

At present, the available evidence supports this site as consistent with a TBM-1C Avenger lost from VT-51 on 15 September 44 with the crew of three MIAs. However, definitive evidence was not found.

The evidence for this specific Avenger includes:

  • The aircraft in this field is an Avenger, based on characteristic main landing gear, outer wing with leading edge slot and radar antenna and tail wheel assembly; as well as an engine (R2600) used in Avengers (although not exclusively).
  • No other aircraft parts/debris were observed within 100 meters of this field so that this appears to be the debris field of a single aircraft.
  • Based on archival footage and photographs showing the crash of an Avenger on 15 September 44 at a time and place consistent with the after action report summary and interviews of living squadron members, this Avenger field falls precisely within the predicted crash area of Avenger in question.
  • The Avenger in question suffered a rapid, vertical descent, which should have resulted in a compact debris field - this is what was observed with engine, canopy, wing, tail wheel assembly and turret components all within a 25 meter area.
  • No other Avengers (including from VT-51) are known missing in the immediate area of Peleliu.
  • The Avenger in this field had a non-standard painted band around the base of the propeller spinner. Avengers of VT-51 had a band around their spinners, as seen in the photograph below. One VT-51 veteran in a recent interview reports that all but one of the VT-51 spinners were painted white, the exception being TARE-9 (TARE was the squadron call sign, with "X" painted on the vertical stabilizer), which was painted yellow. So far, we have only been able to locate black-and-white photos, which confirm the presence of a spinner band.

Although the available evidence supports this Avenger field as being that of the Avenger in question with three MIA crewmen, no specific aircraft identification was made during this ground search by the BentProp Project team and no personal effects, other than one spoon, were located.

A variety of reasons could explain why no such identification and observations were made. Concerning the aircraft's specific identification, recent interviews with surviving members of VT-51 who flew on this mission confirm that the Avenger lost that day was hit and crashed consistent with the available photos and film footage: namely, this Avenger experienced a catastrophic event that resulting in almost instantaneous translation of forward motion into a rapid, vertical (flaming) descent. Very close review of the photograph above suggests other flaming parts fell separately from the main body. Loss of the empennage is one explanation which could account for the sudden vertical descent - and would explain the lack of an empennage at the crash site. As importantly, because the BentProp Project team's search was a superficial ground search, the possibility exists (and is suggested by the metal detector search) that additional parts may lie beneath the surface at or near this field. In addition, parts of this aircraft may have been covered during the crash by the observed parts (compressed "accordion-like") or may have been covered or obscured during the construction of the East Road, which started some time after 15 September 44 and which aerial photos show was completed by June 1945 [NOTE: compare 1944 and 1945 photos of Peleliu field above; also comparison of the East Road from June 1945 to 2006 indicates no apparent change to the road.] Finally, although this crash site lies in an uninhabited area of Peleliu, this area is known by the local crab-hunters. Over a period of almost 61 years, the possibility exists that some parts of this aircraft may have been salvaged.

Concerning the absence of personal effects and human remains, several explanations are possible. First and foremost as noted above, the BentProp Project team conducted a ground search only. Such a search would not include excavation in and under the large debris to look for such evidence. Secondly, the very rapid descent resulted in a focused and explosive impact, which could affect distribution of human remains and personal effects in an unpredictable manner. Thirdly, it is possible that remains were located as early as 1945, and in the absence of aircraft and personal identification, might have been recovered and sent elsewhere as unknowns [NOTE 1: no mention is made of this site in the US Army Graves Registration Unit 1947 report, suggesting they did not recover remains from this site.
NOTE 2: Post-mission, extensive efforts to determine/fix an association between the yellow propeller spinner have not added to the information contained herein. Our current assessment remains that this band was painted on after arrival to VT-51 (making this a unit identifier).
NIOTE 3: Numerous GPS readings and maps have necessarily been deleted from this report, pending investigation].

Taken together, the archival research predicts the location of a VT-51 Avenger lost along the East side of Peleliu and the field, at the predicted location, is that of a single aircraft, an Avenger, possibly from VT-51 (e.g., yellow propeller spinner). As well, the location and profile of the crash site is consistent with the recent recollections of living former VT-51 squadron members. The team members of the BentProp Project, who were present at the site, place a 90% probability that this Avenger in the Avenger in question and thus may be the site of up to three MIAs lost as a result of the crash.

Accordingly, the BentProp Project recommended in a summary document that consideration by the Governor of Peleliu, the Palauan Historical Preservation Office, JPAC and NHC be given to the high probability, with all the above noted caveats, that this area may contain the remains of up to three American airmen, Furthermore, the BentProp project respectfully requested that the area be appropriately protected until JPAC and other appropriate American authorities, in coordination with all affected Palauan authorities, review our findings and make a determination about recovery operations.

All the members of the BentProp Project, including those who conducted the ground search and those who provided technical support in the United States, wish to express our sincere appreciation to the Governor of Peleliu, the Palauan HPO, Mr. Tangie Hesus and other Palauan authorities for their support and assistance in permitting us to conduct and complete this search.

Redacted drawing of the Avenger site. © P. Scannon, 2006

Flag ceremony at the new TBM engine.
L-R: Pat, Mike, Reid, Joe, Tangie. © D. O'Brien, 2006

Special Note - In 2008, a JPAC recovery team conducted a recovery mission at this site. The JPAC team did not recover remains, but they did recover sufficient information, including personal effects, to confirm that the aircraft at this crash site is, as we had previously concluded, the TBM flown by Francis Waters, with crew members Hyman Atun and Paul Bensman, which was shot down on 15 September 1944, the first day of the invasion of Peleliu.

Go back to the top menu


Page last modified 11 October 2009