Riou island's F-5B Lightning, Rhône's delta, France
Pilot : Commander Antoine de Saint-Exupéry
July 31, 1944, 8h45
Borgo-Porreta, Bastia, Corsica, France
take pictures of targets in the area of Annecy, Lyon, Châlons/Saône
Zone of crash
Sea, Riou island, Marseille, Rhône's delta, France
F-5B-1-LO, LAC 2734
Cliquer pour agrandir
Antoine de St-Exupéry

Simplified preliminary study report

Due to the importance of the pilot, this page will evolve in the coming days (history, pictures and technical sketches).



- Acknowledgements

- Introduction

- Facts and geographical location

- 1st Part : authorisation for investigations and recovery

- 2nd Part : study, analysis & identification of the remains : left gear assembly

- 3rd Part : study of the forward boom part holding the turbocharger

- 4th Part : aircraft engine

- 5th Part : search for an identification reference on the left forward boom

- 6th Part : conclusion



The Aéro-Re.L.I.C. association thanks all persons who have allowed, at all levels, to perform undersea investigations, often made difficult because of the wide area and the depth where the wreck of Lockheed Lightning F-5B serial number s/n 42-68223 was located.

Mrs : Habib Benamor, Jean-Claude Bianco, Pierre Boissery, Jean-Claude Cayol, Jack Curtiss, Brian Cyvoct, Daniel Ellena, Laurent Lhuillier, Arnaud Maquaire, John K. Mattison, Gilbert Pachoux, Jean-Louis Robin, Alexis Rosenfeld (Photocéan), Sylvain Scoccia, Luc Vanrell (Comptoir des Sports Sous-Marins), Christian Vigne, Nicolas Vincent.

Particular thanks are given to maritime authorities of the Mediterranean area, who accepted all our requests for investigations and recovery :

- Direction des Affaires Maritimes des Bouches-du-Rhône, quartier de Marseille

- Département des Recherches Archéologiques Subaquatiques et Sous-Marines, Fort Saint Jean, Marseille

- Préfecture Maritime, Toulon

Particular thanks are also given to two subaquatic engineering companies, and their presidents, who graciously lent important technical means, as well as highly qualified personnel, during the investigations on this aircraft :

- Comex, Marseille. President and CEO Henri-Germain Delauze

- Géocéan, Aubagne : President and CEO Pierre Becker



Facts and geographical location

On May 23rd, 2000 Luc Vanrell, director of company Le Comptoir des Sports Sous-Marins, located in Marseille, in the course of a dive east of Ile du Riou, southwest of the Calanques, identified scattered and heavily damaged debris, for what appeared to be a twin-boom aircraft.

L. Vanrell had known this site for several years, a true debris field due to its extension as observed by him, starting from the eastern side of Ile du Riou, and stretching eastwards for half a mile.

The goal of his reconnaissance dive was to try defining the type of aircraft involved, thanks to precise characteristics which could possibly still be seen on some existing elements. These scattered debris lied on a sandy bottom, littered with shellfish, between 50 and 62 m deep.

Away from the eastern tip of Ile du Riou, in an eastern direction, at a depth of 54 m, lied the main wreck, here called " main wreck ", because it was the most important.

Although this main wreck was partly covered by remains of trémail nets (of vertical-pulling type), its study allowed identifying the type of aircraft involved.

It thus was ascertained that it was a part of a left boom, of an approximate length of 2 m, still equipped with its turbocharger, and most probably belonging to a twin-engined twin-boom American aircraft, a single-seat Lockheed Lightning of the Second World War.

Around this main wreck, lied various other debris, which could not be easily identified, due to their small size. Only a big engine could be identified, nearly completely buried in the silt, with a row of rocker arms emerging from the sand.

A little deeper, 62 m below sea level, a complete main landing gear assembly lied, alone and devoid of any attachment. Following this dive, L. Vanrell officially declared these marine remains to DRASSM (Département des Recherches Archéologiques Subaquatiques et Sous-Marines) of Fort Saint Jean, in Marseille, on Friday May 24th, 2000.

At the same time, L. Vanrell contacted the Aéro-Re.L.I.C. association, specialised in serarching and identifying crash sites of WW II aircraft in Southern France, to have a detailed advice on the type of aircraft involved.


1st Part

Request for authorisation to investigate and recover remains found by L. Vanrell, in order to identify them

Two years later, in agreement with the inventor, a request for authorisation to investigate the site was transmitted by Aéro-Re.L.I.C. to the Préfet Maritime of the Mediterranean area, in Toulon, on July 3rd, 2003. Copies were transmitted to DRASSM and to Direction des Affaires Maritimes des Bouches-du-Rhône, responsible for the site involved.

The main goal was to recover the remains, in order to perform visual analyses.

On October 3rd, 2003 with letter Premar/AEM/NP Sitrac N° 1136, the Préfet allowed Aéro-Re.L.I.C. to proceed with the recovery of the remains found in Ile du Riou by L. Vanrell, with the usual restrictions associated to the fact that these remains were known to be of military nature.

The remains were then recovered and officially taken in charge by the Minibex and Janus II ships, belonging, with their crew, to Comex.

The remains were transported to the premises of the Géocéan subaquatic engineering company, in order to be cleaned, dismounted and analysed there, for identification, and later storage with the appropriate preservation conditions.


2nd Part

Study, analysis & identification of the remains :

The left main landing gear

The first investigations of the remains found by L. Vanrell in the Riou area were done by the bureau of Aéro-Re.L.I.C. : Philippe Castellano, Brian Cyvoct, Daniel Ellena, & Christian Vigne.

An important cleaning and dismantling work had previously been performed by Laurent Lhuillier, an experienced sheet metal worker as well as subaquatic yards responsible, employed by Géocéan. He was instrumental in this respect, thanks to his competencies.

It should be noted that the remains described in this report, in the light of their appearance, have certainly been dragged and trawled many times over the previous years.

As this is a warbird, it can be considered in a first step that the complete break up of the aircraft, besides the impact with the water surface, is due to fishing activities (mainly dragging and/or trawling ones), over a span of at least 50 years. During the following years, up to the official dive of L. Vanrell, i.e. about ten additional years, these remains were again displaced due to small fishing activities, as shown by numerous bits of trémail nets, old or recent, left all over the site.

The landing gear and its upper square-section fulcrum (photo by P. Castellano)

On the side of hardware, besides aluminium and stainless steel parts who do not seem, after a quick visual inspection, to have been too much damaged by their long stay in a marine environment, electrolytic action has on the contrary been very active elsewhere.

This factor has without doubt accelerated the degradation of the other remains, because of their specific composition (aluminium and magnesium alloys in simultaneous contact with forged steel, copper, bronze, brass, lockwires,…).

The first remain to be analysed has been the easiest one : the left main landing gear (Lockheed Part n° 196127L), with its wheel missing.

This part, characteristic of a Lockheed Lightning P-38 fighter, or its F-4/F-5 photographic adaptation, exhibits modifications typical of the later version of this aicraft.

The fact this is a left gear is confirmed by the position of its arm (Part n° 225297-3), mounted on the strut assembly in a forward position.

The wheel, as well as the hydraulic braking system, has completely disappeared. Only the two main conical roller-bearings, in stainless steel, made by Timken, remain in place.

The most interesting part of the gear is its upper fulcrum, with pins at its two extremities, the rearward rotation of wich allow storing the main landing gear in its well.

Aircraft of the US Air Force (called the US Army Air Forces between 1941 and 1947) are referenced by a letter corresponding to their missions, like at the time Pursuit for fighters or F for photographic reconnaissance, followed by a chronological number, like P-38, F-4 and F-5 for the Lightning. Subtypes have letters, like F-5A, F-5B,… A X prefix is given to prototype versions (i.e. XP-38, XP-38A, XP-38K). To distinguish minor modifications, numbers are added, in " locks " of 5 (later, even smaller modifications, then can be numbered 2,3,4,..) : e.g. P-38H-1-LO, P-38H-5-LO (LO is a factory code, in this case the Burbank premises of Lockheed in California).

Starting with P-38H with USAAF serial number s/n 42-66802, the 76th P-38H-5-LO produced by Lockheed, a significant modification was brought to the fulcrum : originally built from steel with a tubular hollow cross section (Part n° 225189), it then switched to a full square cross section made from cast magnesium (Part n° 147667).

This modification, together with, among others, the improved GE B-33 turbocharger, are the main reasons for a switch from Block 1 to 5 : P-38H-1-LO to P-38H-5-LO. All Lightnings after s/n 42-66802 were equipped with this new fulcrum, which precisely corresponds to the one found near Ile du Riou. Note the P-38H also introduced the improved Allison V-1710-89 and V-1710-91 engines boosting take-off power from 1,325 to 1,425 h.p.

This led Aéro-Re.L.I.C., as soon as L. Vanrell declared his remains, to limit the possibilities to the following Lightning versions : P-38J (this introduced the first significant physical change in the Lightning series, a chin radiator under the engine, for improved cooling), P-38L, two-seat P-38M, F-5B, F-5E, F-5F and F-5G. Also, after a long and detailed analysis, Aéro-Re.L.I.C. concluded that no P-38H-5-LO, P-38J, nor any of the subsequent models, had ever been officially recorded as having been lost in the Mediterranean area during and after the war.


3rd Part

Analysis of the forward boom and its turbocharger

In front of the left turbocharger of F-5B 42-68223, from left to right : L. Vanrell, P. Castellano, J.C. Cayol (photo by B. Cyvoct)

The boom part, of a length of 1.9 m, a width of .9 m and height of .63 m, weighs about 150 kg, turbocharger included : belonging to the Forward Boom, it is made of the Stainless Steel Structure Assembly (Part n°197401L), essentially the upper part of the Forward Boom, accomodating the turbocharger. It initially withstood a mechanical compression, alongside its main axis, probably because of a vertical impact with the water, followed by a significant break-up, most probably in the course of subsequent trawling and dragging.

The forward part of the Structure Assembly was ripped off at the level of the vertical access door to the landing gear fulcrum (at the level of the middle of the outer wing chord). The whole assembly is of rigid box-type construction, being in interface, at the front with the firewall (which has disappeared) behind the engine bay, and below with the landing gear well.

The rear part of the Structure Assembly was similarly ripped off at its rear frame, at Station 225, i.e. at the leading edge of the two main air scoops, used to cool the main engine radiators (one scoop and one radiator of elliptical shape on each side of the boom).

Upper part of the stainless steel structure assembly

The upper part of the forward boom accomodates three main distinctive elements :

- 1) the turbocharger

- 2) the transition assembly

- 3) the Y-shaped duct assembly


1) : The turbocharger The turbocharger is a General Electric GE B-33, of .6 m diameter, weighing approximately 90 kg. It could be neatly removed from its box by Laurent Lhuillier. A massive and compact part, it did not suffer from electrolytic effects, at least outside (it was not dismounted, although screws, nuts and bolts could be easily removed), except in the lower right area, where a quarter of the shell has been destroyed. The aluminium case did not exhibit any specific marking, nor any riveted identification plate.

The turbocharger, with its pipes. To the right, the landing gear (photo by P. Castellano)

2) : The transition assembly

Bolted to the upper forward part of the turbocharger, this tube (Part n°11-249) collects the exhaust gases from the Y-shaped duct. Of 20 cm diameter, it is made of stainless steel, with a thickness of 5 mm. Reference II-377B, electrically etched with a pencil, has been found. This piece has withstood loads alongside its main axis, which made it move backwards, and bend with an important deformation of its attachements at its two extremities. Underneath, there is a cover in stainless steel, used for thermal protection : it incorporates a welded plate with punched reference N° 12-1153-1 SOLAR. The latter name, which appears twice on the remains (on the boom and on the transition assembly), corresponds to a Lockheed subcontractor, specialised in the manufacturing of stainless steel parts. It was located, during the 40's, in San Diego in California.

3) : The Y-shaped duct assembly

This duct (Part n° 234379L) is attached to the forward part of the transition assembly with a clamp made in stainless steel. Its forward part is Y-shaped, corresponding to the two tubes collecting gases from the two rows of 6 cylinders of the V-12 Allison engine (note that the engines of the Riou Lightning have not been found). These two tubes have been severely deformed, due to the strong rearward loads applied :

The right pipe was pushed backwards, such that its angle is at nearly 90° with respect to its original position

The left pipe was pushed even more, the angle being greater than 90° with respect to the original position.

All the above remarks lead to think that the aircraft impacted the sea rather violently, with a combination of high velocity and steep angle. The shape of the turbocharger and its piping shows that the impact probably pushed the engine backwards, leading to an important overall deformation. It is highly probable that these dynamic deformations are linked to the impact, and thus have no direct link with the numerous fishing activities (trawling, dragging,..)

To obtain more detailed informations on the structural deformations under such accidental compression of the remains of the wreck of Ile du Riou, the authors suggest the use, in the near future, of the wide experience of BEA (Bureau d'Enquête Accidents) in Le Bourget.

Lateral sides of the stainless steel structure assembly

The left and right sides of the forward boom have also been analysed : no Karman, outer left wing or central section elements, be it upper or lower surfaces, have been found on the lateral sides of the Structure Assembly.

1) Left side

- Various parts still exist on the shell, mainly :

- An air scoop, oblong shaped (Part n° 195446L), to provide fresh air to the turbocharger. This part has been crushed and released from its housing by the small fishing activities

- Three rear fittings for the gear left door (insted of the five original ones). One of the three is completely detached, the other two still are in place

- various guides and brackets, well identified.

The most interesting part of the remains is the oblong air scoop. Like for the fulcrum of the gear leg, it was mounted on a later generation of the aircraft, thus more recent than the P-38H-1-LO.

This scoop (.19 m diameter, .74 m length, .34 m high), replaced the airscoop, bent at 90°, which equipped the former versions of the aircraft.

Right side

- No specific element remains. The fin has been completely torn away, carrying away with it the right side if the boom. As the aicraft has been identified as a military Lightning, thus being a war machine, it was not appropriate to direct our investigations on suspect traces such as impacts, possibly coming from shots, either ground ones (AA/Flak) or aerial ones (7.92 and 13 mm bullets or 20 mm shells, as is the case for a P-38 lying in the Bay of La Ciotat).

In fact, after careful analysis of the surfaces of these remains, no particular mark or hole caused by possible shots could be found. In the light of the low percentage of recovered parts, this does not however preclude a crash due to shots, either ground or aerial ones.


4th Part

Aircraft engine

As indicated in the introduction, an engine, apparently of liquid-cooling type and at least 6 in-line cylinders, lied not far away from the left boom. L. Vanrell, in spite of the significant depth of the site, 54 m , tried to remove sand from the remain. A second row of cylinders then appeared, confirming it was a liquid-cooled V-12 engine. This type of engine precisely equipped the Lightning. It thus became necessary to find the identification plate of this engine.

In his report of May 24th, 2000 concerning the remains of Ile du Riou, L. Vanrell mentioned the plate details as being : DB-601.305-001 EZ11, followed by other factory references. They thus clearly correspond to a German Daimler-Benz DB-601 motor, a liquid-cooled inverted V-12 (crankshaft in upper position, cylinder heads in lower position). This type of engine equipped during WWII the single-seat Messerschmitt Bf 109E fighter, a version nicknamed Emil, as well as the Messerschmitt Bf 110 and Me 210, the Dornier Do 215, the Heinkel He 100, the Italian Macchi C 202 and Reggiane Re 2001,…


5th Part

Search for an identification reference for the left forward boom

All references found on the left landing gear, the left forward boom and its turbocharger have been carefully recorded and identified by us. Those punched on the landing gear and the turbocharger and his pipes not giving away any information with respect to identification of the aircraft, we concentrated on the boom. Only two references exist on this boom, but they are located inside the Structure Assembly holding the turbocharger, which means that they could only be found by removing this turbocharger.

- Analysis of the turbocharger box's left sidewall :

On the upper part of the Inboard Web (Part n° 230085L) of Steel Structure Structure Assembly, in a forward position under the Steel Plate for protection of the pilot, in company of a Solar stamp, digits were punched during manufacturing: 197 371 2. This is a generic part number, as shown in the Lockheed technical documentation, refering to an assembly support built by the Solar subcontractor. It thus does not allow identifying the aircraft.

- Analysis of the turbocharger box's right sidewall :

Just above the the corner joining the Inboard Web (Part n° 230085L) and the Diagonal Shear Web (Part n° 230109) of the Stainless Steel Structure Assembly, in front of the supercharger inboard support, the following reference, manually punched , was found : 2734L.

The Lockheed constructor's number which allowed identification of F-5B 42-68223 (P. Castellano pictures)

This reference is essential, since it is not a generic part reference, but what historians call the manufacturer's constructor number, as given by all aircraft manufacturers on their assembly line.

The Lockheed Aircraft Corporation (LAC), now merged with Northrop under the name of Lockheed Martin, including Lockheed Martin Aeronautics Company of Fort Worth in Texas (with other plants in Palmdale and Marietta), has always given its aircraft 4-digits c/n's (starting with c/n 1001,..).

This number is described by the company as a LAC MSN (Manufacturer Serial Number).

As an example, our work on a P-38 wreck in Baie de La Ciotat, where no military serial number could be found more than 50 years after the crash, confirmed it as military s/n 43-2543, as soon as we could find its LAC MSN 3653.

Thus, LAC MSN 2734 identifies without any doubt the remains of the aircraft lying near Ile du Riou. The L letter following 2734, for Left, is of particular shape, as it corresponds to an inverted 7. Most probably, for practical matters, the worker simply used the " 7 " stamp he already was using, to quickly identify the left side of the main part.

Note that our experience with various Lightnings has shown that, according to the factory where a given aircraft was built, the location of these references can vary. Two factories were involved with the Lightning : Vultee in Nashville for the P-38L-5-VN, Burbank for all others.


6th Part

Preliminary report conclusion

The presence of LAC MSN 2734 number on the left forward boom of the aircraft remains lying near Ile du Riou has allowed its formal identification.

LAC MSN 2734 corresponds to a Lockheed Model 422-81-21, USAAF type F-5B-1-LO Lightning, as modified on the Burbank (California) line into a photographic version of the P-38J-10-LO (Lockheed Model 422-81-22). This aircraft had its reception flight on December 18th, 1943 and was handed over to the US Army Air Forces (USAAF), with serial number 42-68223, 42 corresponding to a buy during Fiscal Year 1942.

Military archives show that Lightning 42-68223 belonged to French squadron Groupe de Reconnaissance GR II/33, commanded by René Gavoille, within the 3rd Photo Group of MAPRW (Mediterranean Allied Photo Reconnaissance Wing) of the 12th Air Force.

This aicraft went missing on Monday July 31st, 1944 during an aerial reconnaisance mission in the area of Annecy, Lyon and Châlon-sur-Saône, with Commandant Antoine-Marie de Saint Exupéry as the pilot.

As this preliminary report is transmitted to the French maritime authorities of the Mediterranenan area, the official status of Commandant Antoine-Marie de Saint-Exupéry still is : " pilot missing, died in the service of France ".


Philippe Castellano, President

Brian Cyvoct, Secretary

Phillipe Jung, Translator

La Napoule, April 12, 2004.

© Copyright Aéro-Re.L.I.C. / B. Cyvoct 2004. All rights reserved.