The Restoration of a 1956 T-33 ROCAT Ejection Seat

The T-33 aircraft is one of the very first jet aircraft produced and has served its purpose well. The profile of the aircraft with its classic wingtip tanks are easily recognizable . The choice to restore a T-33 ejection seat has been a good one for several reasons. Although both the aircraft and the ejection seats have undergone many changes, they were produced in sufficient numbers that there are still a few flying throughout the world. These seats are relatively simple to restore and do not require sophisticated tools or knowledge to work on them. 

Photo: http://www.aviationsmilitaires.net/display/aircraft/91/t-33

The T-33 trainer or "T-Bird" as it came to be know, evolved from the F-80 fighter which also bore the name Shooting Star. By adding three feet to the fuselage to accommodate a second pilot the world's first jet trainer was born. When the cockpit grew to two places the six machine guns were removed. 

Design work for the P-80 began in 1943 with the first flight on Jan. 8, 1944. The F-80, then P-80, was the first jet fighter to enter squadron service in the Army Air Force and saw action in Korea. As more advanced jets entered service, the F-80 took on another role - training jet pilots.  It was inspired by the German ME-262 and shows many similar features.
  

The T-33 is historically significant to illustrate the evolution of both first generation jet aircraft and ejection seats.  The aircraft has a straight wing configuration which limited it to subsonic speeds even though its mach meter wishfully hoped it could achieve mach 1.5  Air compressibility becomes increasingly important as speeds exceeds 300 mph, and when they are excess of 400 mph it is the foremost design consideration which requires swept back wings.  The whole concept of compressibility was new and poorly understood in first generation jets. The T-33 was first developed in 1946 at a time when the attrition rate of jets was sixty-one aircraft per 100,000 flying hours, and was only reduced to forty in 1948. The ten year average between 1985 and 1995 is 1.5!

(Picture source unknown.)

The two-place T-33 jet was designed for training pilots already qualified to fly propeller-driven aircraft. 

Originally designed the TF-80C, the T-33 made its first flight in March 1948. Production continued until August 1959 with 5,691 T-33s being built. In addition to its use as a trainer, the T-33 has been used for such tasks as drone director and target towing, and in some countries even as a combat aircraft. The RT-33A version, reconnaissance aircraft produced primarily for use by foreign countries, had a camera installed in the nose and additional equipment in the rear cockpit.

With the conclusion of WW II, a new era began to unfold in Naval Aviation - the jet age - creating new requirements for training aircraft. Navy policy had been to utilize obsolete fleet aircraft in the advanced training stage but, with the development of jet fighters, there were no old or obsolete jet fighters available for training. This led the Navy to acquire three P-80A Shooting Stars from the Air Force to evaluate the suitability of jet aircraft for operations aboard carriers and for training.

Testing began with the delivery of a P-80A to NAS Patuxent River on June 29, 1945. On November 11, 1946, a P-80A was used for catapult launches, free takeoffs and arrested landings aboard USS Franklin D. Roosevelt. However, this was not the first shipboard testing of a jet. On July 21, 1946, an FD-1 (FH-1) Phantom had operated aboard Roosevelt.

By 1946, the FJ-1 Fury and FH-1 Phantom were in the fleet and the need for a jet training aircraft was evident. To fill this gap, the Navy procured 50 Lockheed P-80s from the Air Force to use as interim familiarization trainers for fighter pilots until enough Navy jets became available. The P-80 Shooting Stars were designated TO-1s and were not equipped with arresting hooks or catapult fittings. They were assigned to VF-6A and VMF-311. They acted as the training squadrons for pilots and maintenance personnel, providing basic operations and tactics in jet aircraft. On May 5, 1948, VF-17A, flying the FH-1 Phantom, became the first Navy jet squadron to become carrier qualified. 

Some T-33s kept two machine guns for gunnery training and some allies used T-33s in the 1960s, but T-33s continued to fly as currency trainers and test platforms right into the 1980s.

The T-33 served many missions during its time. It was used to transport pilots to other bases to pick up tactical aircraft, and afterward led them home on long over-water flights. The aircraft also provided realistic combat training. While deployed as the "enemy" during exercises, T-33 pilots flew far out to sea and returned to penetrate the radar defenses of Okinawa, testing the interception ability of jets defending the island. The aircraft also served as a high-speed courier service for the delivery of classified materials, as well as ferrying aircraft parts to various bases in the Pacific.

The T-33 is one of the world's best known aircraft, having served with the air forces of more than 20 different countries for almost 40 years. Many are still in use throughout the world. T-33s have also been built under license in Canada and Japan. A reconnaissance version, RT-33, was built for export. (reference Wikipedia)  Canada was by far the largest licensed producer of the T-33 with 656 variants designated as the CT-133 for the RCAF.
 

Ejection seats of all varieties are becoming more difficult to find all the time but my advice to anyone planning to undertake such a project is to find one as complete as possible to start with.  Locating a parts manual which shows the seat variation which you have is a big help as well.  Many of the larger aircraft parts dealers who may have some parts in stock prefer to deal with part numbers.  Parts are more difficult to source as time goes by but with diligence, determination and patience many can still be found.  I cannot stress the words diligence and patience enough.  The parts shown on this Blog came from Canada, the United States, Mexico, the U.K., Holland, Belgium and Greece and this project has been in progress for four years.  It is wise to check with your customs to make sure the parts you seek are not in some way restricted.  In most cases obtaining parts are the result of being in the right place at the right time and networking with friends who know friends who know friends.  With the assistance and advice of a very good friend, I was finally able to locate a seat which had potential.
   

A picture is 1000 words.  This is a good graphic depiction of what the restoration often felt like.

"The reason I'm doing this is because it feels so good when I quit." 

These seats went through many modifications in their service life but they can be divided into two basic types.  The first is the ballistic variety and the later variation the rocket propelled or ROCAT type.  The ballistic seat was basically a charge which propelled the crew out of the cockpit.  At higher speeds there was some concern about clearing the tail of the aircraft and the weight of the pilot was a major consideration.  The ROCAT or rocket powered seat on the other hand provided the necessary propulsion and it did not compress the spine and organs as severely when activated.  It provided a steady burn propulsion rather than a sudden jar.  This allowed for ground level ejections at 70 knots.

“Of necessity, to prevent incinerating the rear-seat occupant, the ejection was sequenced rear-seat/front-seat regardless of who initiated it. Ironically, after more than 20 years of operation, there was suddenly some concern that some pilots in the front seat might strike their knees on the windscreen rail on ejection.  The T-33 cockpit was small but the rear seat had slightly more room. 

An illustration showing the leg room for an average sized pilot wearing a backpack chute.
(picture source unknown.)

Early ejection seat noting minimal knee clearance in foreward cockpit.  As a result, all pilots flying the T-33 were measured for clearance - thigh length, rather than height being the deciding factor. A number of pilots who had been flying the T-33 for years found themselves restricted to "rear-seat only."[2]

At first glance, both the ballistic and the ROCAT seats look similar but the ballistic seat is quite a bit simpler and seems to be more common.  The ROCAT seat on the other hand has some differences which are not immediately evident.  The most distinguishing features are the ‘kicker’ strap with its man/seat separator under the bucket (upper right picture) and the shoulder strap inertia reel in the headrest.(upper right picture).  

This seat diagram from an original manual is exactly the same as the one which underwent restoration.  (picture source from T-33 manual.)

The total force impose on the body during ejection can be in the neighbor hood of  8 to 12G.  With the ballistic catapult, the rate of onset of G. is very high and after the seat has moved only a small distance, the force applied to the pilot is considerable higher than if you had been moving at the same time as the seat.  A T-33 cockpit is quite small and the seat cannot be adjusted fore and aft.  When the T-33 was flown from the front seat, the pilot often opted to wear a seat chute which would give more room for his knees to clear the windshield frame of the canopy in the event of an ejection.  A back pack chute was more commonly used in the rear seat with the survival seat pack.

The ballistic seat also has a roller bar for the shoulder straps below the headrest.  Many of the parts are the same for both seat types but in undertaking a restoration, it was most important to find a manual with the correct part numbers for the seat.  That is not an easy task in itself.  The manuals are copy written so part numbers often need to be scavenged from a variety of sources in bits and pieces.  Most commercial sources for parts want part numbers and life is much easier with those /N numbers.  For example, my seat was fitted with an MA-6 shoulder harness retractor.  These are used in many configurations but the differences are significant enough that only one will fit the seat correctly.  The most useful sources for both information and parts were networking with other collectors who had similar seats and also by talking to various members of the Flight Clothing Forum.

My personal preference with any static restoration is to leave things as they are in their original state.  As long as they do not have parts missing or broken, I try to just clean pieces so that they are in their best display condition and leave things alone.  Unfortunately in the case of my seat, its lengthy service and years in forgotten storage had taken a cosmetic toll.  Although the seat did not have physical damage, many parts had been striped and the paint had pealed and worn and in many places it had taken on an unpleasant mottled look so I decided to restore it as close as possible to its last colour.  The frame is welded stainless steel while the bucket and other parts are aluminum.  There was evidence that the seat had been painted several times in its lifetime.  The aluminum had traces of original primer typically used with fabricated stock aluminum.   The bare aluminum had been prepped with zinc chromate and the seat had been originally painted olive green.  The bottom and inside of the seat pan had been later painted florescent orange and the last coat was grey.  Each time it had been painted it appeared that the only parts which were removed were ones which would not operate properly if they were sprayed over again.  Most ballistic seats I have seen were olive green although there are some Canadian seats which were black.  Many of the later ROCAT seats were done in the medium grey.  In refinishing, I decided to leave the replacement parts in the original colours that were used when their data plates were applied.  My ROCAT seat ended its flying career in the U.S. Navy so it was finished in those colours.
   

Regardless of whether sandblasting or paint stripper is chosen to clean the seat, parts need to be removed to achieve a smooth finish and to prevent damage to parts in the process.  Either way it is not necessary to remove ALL the paint but it is important that the metal be smooth.  I masked any part that I wanted to be left in their original state for example the labels which were surprisingly in good condition.  I decided to remove the paint with Circa 1850 Paint Remover rather than to sand blast it.  In hind sight I feel that was a good decision.

 

Seat frame and bucket stripped and ready for zinc chromate priming.

The seat bucket and frame were not mass produced.  By using paint stripper, many unexpected secrets were revealed.  In stripping the paint, I found layout marks on the aluminum to indicate where holes needed to be cut and drilled.  The seat was not designed to be totally disassembled either.  In places, the aluminum seat was riveted to the stainless steel frame so that even if all the bolts were removed the seat itself could not be separated from the bucket frame.  Using paint stripper gave me the opportunity to very faintly discern the word “pilot” and the tail number of the aircraft “56-?593” on the seat.  A bit of research on the tail number and the process of elimination revealed that it could be 56-1593 which went to the Military Aircraft Storage and Disposal Center (MASCDC) as TC 893 on November 16, 1969.  In order to keep the provenance of the seat, I reapplied the stenciling as it was when I repainted it.

Removing parts was a learning experience itself. Many of the bolts have Philips heads and I learned that in order to remove them NOT TURN THE BOLT WITH A SCREWDRIVER! Instead HOLD the bolt with the correct size screwdriver and turn the nut. Other things I’ve learned? ALWAYS put the same bolts back in the pieces they were removed from. The bolts may look the same but they are often have slightly different lengths and belong in specific locations. When it comes to reassembly, the parts go back together in the reverse order to the way they were removed. Even though this may seem obvious it can save a lot of grief and be easily overlooked. Murphy’s Law: No matter how many pictures you take, you probably won’t see the information you need when it comes to reassembly. I wanted to save the original headrest and as a result it required a great deal of care to remove it without damage. I have reattached it with Velcro so that it can be interchanged with an original spare.

Once the seat had been taken down to bare metal I heeded the advice of someone who did an outstanding job of painting a motorcycle with spray cans. He had repainted his bike with very impressive results. His advice was to shake the can well, keep the can upright, spray in short bursts, start the spray slightly before contact with the part and don’t aim for a finished product the first time. Keep coats very light and spray over when the previous coat has set up. BE PATIENT! This process helped to prevent the paint from running. The odd time when it did, I simply let it dry thoroughly and lightly sanded it before I redid it. I used zinc chromate on the aluminum to give it a base coat. Because my seat is the later ROCAT type, I decided to paint it the gray color that was on it last. I went to an automotive paint store and bought automotive primer which turned out to be very close to the original grey of the seat. The colour was so close that I didn’t feel it was necessary to spend the extra money to have the paint on the seat digitally color matched. The flatness of the primer did not quite match the original sheen so I finished the seat with a coat of Krylon, Satin, Clear Transparent spray which was just enough to take away the total flatness of the primer.

Reattaching parts in the reverse order presented a new set of problems because many were already removed when I obtained the seat.  Many of the parts for this ROCAT seat are very difficult to locate.  Some of the missing parts in the head of the seat had to be attached before others would fit.  The “seat adjustment jackscrews were missing and were one of them.  Fortunately they are common to both the ballistic and ROCAT seats and one would think that they would be easier to locate.  Not so.  With a great deal of searching I was able to finally locate one which was suitable.  They too were stripped and repainted in their original colours.

T-33 seat adjustment jackscrew.  This piece is located in the headrest of the ejection seat and is frequently missing when a seat is located for restoration.

Once a project of this type is begun, it is amazing how even locating such simple things as correct cotter and clevis pins to install parts can become difficult to source.  Once again, friends and connections with like minded individuals proved more valuable than such conventional sources as E-bay.  Because the number of these aircraft that are still flying is becoming less and less each year, some aircraft owners have either stockpiled what they think they will need or modified their aircraft to accommodate more advanced technology.  As a result, parts are becoming more difficult to find from conventional aircraft suppliers and in many cases they are removing those that they have from their inventory and because of a small demand it is not worth it for them to stock them anymore. 

The “kicker” strap and it’s inertia reel are extremely difficult to locate for the ROCAT seat. After a great deal of time, patience and searching the correct one was finally located.

Inertia reel for kicker strap ready to be attached under the seat.

Inertia reel with kicker strap.

The MA-6 inertia reel which operates the shoulder harness for the T-33 is located in the headrest.  It too has many variations for use in a number of aircraft.  The basic body is the same but they have been adapted for a variety of seats.  The correct reel for the T-33 has the shoulder harness release mechanism on the left side whereas some of the other MA-6 reels have a longer shaft and release from the right side.  I had to make one out of two that were different.

Correct release when mounted on seat.

Variation of MA-6 with long shaft and release on  the opposite side.

Fortunately many parts for these variations can be interchanged  but care must be taken when making the conversion.  The first step was to install the correct shoulder harness.  This harness is quite different from the older  ballistic type.  The ROCAT shoulder harness has a longer strap which attaches to the reel in the headrest.

Correct ROCAT shoulder harness.

Correct shoulder harness for ballistic seat. Note metal attachment. for clevis at the top of the picture.

Attaching the shoulder harness to the MA-6 is not intuitive.  There is a strong spring in the MA-6 and the release lever on the reel must be used to allow the shoulder harness to be extended as far as it will go.  An Allan key is used to hold it in place while the harness is installed.  The harness is fed through the drum on the reel and a wedge is used to hold it in place.  When the shoulder harness is held and the Allan key removed the shoulder harness should retract back into the reel.

The Allan key holds the spring so that  the shoulder harness can be removed.

The ‘wedge’ holds the shoulder harness in place.

                                                                        

Shoulder harness and MA-6 reel ready to have release lever switched.

Care needs to be taken to make sure that the shoulder harness is oriented correctly when installed so that the release lever will operate from the correct side of the seat.  (The large side of the reel needs to be on the left side of the seat.) 

Two Philips head screws which hold the cover on the MA-6 reel need to be removed to make the changes.  The cover needs to be held in place when this is done because the cover is spring loaded.

Cover removed from MA-6.

Release lever removed from ‘donor’ MA-6

Release lever shaft on correct side and cover replaced ready to be installed on seat.   

    

Shoulder reel installed in headbox.

Shoulder reel connected to release mechanism.

I am a firm believer that it is the details on any project that are noticed the most and this is particularly true with the T-33 seat.  Details take the most time to research and to locate but they are well worth the effort and make for a rewarding project when it is completed. 

The T-33, ROCAT seat utilized three initiators to activate the ejection sequence.  Certifying an ejection seat is a costly undertaking due to the fact that both the chutes and the initiators become time expired and require regular upgrading.  In Canada, many of the T-33’s which are still airworthy do not have the seats armed but there are companies which can provide this service.  The initiators are explosive devices and very carefully controlled and regulated.  Even when they are time expired they are rebuilt so the possibility of acquiring an inert example for display is  quite remote.  The M32 initiator on this seat is an inert example with French labels.  The catapult and all the initiators on this seat are inert originals.  Excellent reproduction initiators are sometimes available.  It is unlikely that an inert example of the catapult will be found however the exact specifications are available to manufacture a reproduction out of plastic tubing.

One configuration of the M5 series catapults.  Specifications are only slightly different on other variations in the series.

Live initiators awaiting installation on aircraft. (source unknow.)

Inert M3 initiator left.  Reproduction M3  initiator on right.

Many of the actual initiators look the same as evident in the difference between the M3 above and the M32A1 diagram below.  In most cases the visual difference is in the length.  The above reproduction initiator was cast in resin using an original for a mold but tubing fairly close to the correct size can be obtained from a plastic shop in order to make very close representations.  

Diagram  of a M32A1 initiator showing dimensions to make a reproduction from plastic. 

The braided gas lines are the same as the no.4 fuel lines used by mechanics who build high performance engines and can be ordered from auto supply dealers either with or without fittings on the ends.  The ones purchased at an auto supply dealer are often blue whereas the aviation grade fittings are stainless.

It took four years of searching to find pictures, diagrams, part lists and the correct parts but with determination and perseverance the project came together.  The major components of the variations on the T-33 seat are the same.  The major difference is the various components which were mounted on the frame throughout the years.  Often a T-33 seat will have "additional" holes visible where older equipment was mounted.  An example is the square mounting bracket with the hole in it seen in the right side of the seat in the pictures below.  It is for the inertia reel on the earlier seats and the bracket was not removed in subsequent variations because it did not interfere with anything.   Below are the before and after pictures:

          

Before and after pictures of T-33 ROCAT seat.

Zero-delay lanyard configuration

SH-2 bail-out bottle as configured for an early T-33 ejection seat.  The bottle is positioned in the front of seat cushion of the SA-20 seat parachute and does not have the canvas pouch when installed.

Seat dressed out showing “Remove before flight” flag and H-2 bail out bottle.  The bail out bottle  has been removed for display purposes.  The SA-20 seat chute is complete with the original canopy.

The back of the seat showing the MA-5 catapult which does all the heavy lifting.  The M5A1 catapult  was fired from the bottom and supplied almost twice the “punch” as the ballistic catapult did.  Some of the earlier  T-33 rocket catapults were fired from the top.

A rare view of the bottom of  a T-33 ROCAT seat showing the rotary actuator (take-up real) which operates the seat/man separator ("kicker strap").  It is possible that the pilot may be unconscious upon ejection.  The operation of the kicker strap is timed so that it tightens to throw the pilot away form the seat and automatically pulls the ripcord on the parachute upon egress.  

There were only 58 years between the first manned controlled flight and the first time humans went into space.  The T-33 is a significant part of history as one of the very first jet aircraft produced and both first generation and second generation technology were used.  Its lengthy service was largely due to its durability.   One of the features of restoring a T-33 seat which makes it different from restoring many other ejection seats is that all the major components are quite visible whereas on newer seats some of the details which illustrate how the seat would have functioned are not visible and for static display purposes the seat can be restored without an untrained eye noticing that something is missing. The T-33 ROCAT seat is a good one to restore because historically it is a second generation seat and it is a good way to learn how an ejection seat functions.  Newer seats function in much the same way but with much more sophistication, advanced technology and efficiency.

Instructions for Proper Pilot (Canadian)Tie-Down After 1966.

Ensure sea is in the full up position.

Tighten seat pack Airlock fasteners and tuck excess strap under seat pack cushion.

Attach tie-down straps to D-rings (two) at rear of bucket and to the negative G-strap ring (one).  Tighten straps and attach loose end of negative G-strap to Velcro hooks.

Route the oxygen hose over seat armrest, through the loop, and stow on the oxygen stowage plug.

Lock the shoulder harness.

Place shoulder harness loops on the tongue of the left half of the HBU-12/P lap belt.

Place the parachute arming key on the tongue of the lap belt with the two nipples facing the right side of seat and fasten the lap belt.  Tighten lap belt and shoulder harness. 

Wrap the radio cord around the shoulder harness and tie with the lose ends of harness.

Tie loose ends of lap belt and wrap seat pack lanyard around lap belt and tie or attach to strap DOT fastener.

Carefully lower seat to snug up all straps.

At this point in time, an ejection seat of any type is difficult to obtain.  I have been told that more than one hundred T-33 seats were used to test modifications made to them in Canada alone. and when aircraft were taken out of service they were stripped of parts before they were decommissioned and sold at public auction.  Now none are sold to the public and most end up being shredded and some of the aluminium probably ends up being used to make pop cans.  A sad ending for a noble and historical bird.  If you want to display your seat as if it were just removed from an aircraft the catapult will have been the first to be removed anyway. Both the ROCAT and the ballistic catapults are attached to the seat by a single pin which is held in place by two keepers.   The pin for the ram is taken off the seat by removing the keepers on the pin and tapping the pin out with a drift from a hole in the head box on the left rail of the seat.  (See arrows on picture below.) Penetrating oil and the use of a “hot wrench” technique were required to remove the pin on this INERT! ram.  If you are fortunate enough to locate an initiator or catapult MAKE ABSOLUTELY CERTAIN THAT IT HAS BEEN CERTIFIED INERT BEFORE ATTEMPTING TO WORK ON IT!!

   

                       

                           

    

Pin removal from seat.  Note holes indicated in last picture to assist with removal of pin.

 

Top of ROCAT catapult.  When re-installing the catapult, the tops on both the ROCAT and the ballistic rockets are clearly marked that they are for the T-33A aircraft and how it should be positioned when the pin is inserted. 

Bottom of M5A1 catapult.  In order to prevent the occupant in the back seat from the blast of ejection, the ejection sequence is timed so that the back seat egresses slightly before the one in the front.  

  

SA-20 seat chute showing release mechanism (first picture).  Note zipper (last picture) which provides access to H-2 bail out bottle.  The SA-20 chute on this seat is complete with canopy but time expired and was removed from a T-33 in Dominica.  The seat cushion has a zippered pouch in the seat cushion for the bail out bottle which is typical on most SA-20 chutes.  (See detail photos. )  

The instructions for removing the seat were commonly displayed on a tag either on the back of the seat or on the headrest.  Most of the labels on this seat are original however reproductions can be made to the exact specifications.  The best source for this is a shop which does engraving for trophies.

Neither the chute nor the rocket were left attached to the seat when it was removed from the aircraft but for display purposes both are shown.  The rocket is one of the rare parts to locate and very few are displayed with the seat.

My intent was to make the seat as complete as possible.  However, as with any project, it will be a work in progress and I am always looking for parts to make improvements.  I intend to build a 1/48 scale T-33 model aircraft to display with the seat.  Future plans include restoring a 1953 T-33 ejection seat to display with this one to be able to provide visual comparisons of both first and second generation ejection seats.  The T-33 was one of the few aircraft in which such a transition has occurred.

The pre-flight external check on theT-33 was straightforward but there were noticeable differences from the earlier traing pilots underwent in propeller powered Harvards. Undercarriage locks and tip tank pins had to be removed and stowed. A red stripe around the rear fuselage indicated the location of the engine turbine. There were warning decals for the ejection seats and canopy jettison - all features unique to jet aircraft.

The cockpit was compact, exciting and intimidating with its unfamiliar equipment, instruments and ejection seat. There were red-guarded switches, red and amber warning lights, and warning decals generously distributed around the cockpit. The seats and canopy had their own safety pins to prevent inadvertent ejection or jettisoning. The small cockpit was quite comfortable for pilots "of average stature" but taller or heavier pilots found it cramped.

The strap-in procedure was almost a ritual and it had to be performed correctly and quickly; the instructors had little patience with students who dawdled, especially in the prairie winter.

Armament switches were on the left rear console although they were inactive. They were followed by the fuel switches, throttle, flap lever and temperature control rheostat. On the left wall were the unique landing lever, aileron booster on/off control, fuel high pressure cock, oxygen regulator and the inevitable row of circuit breakers. On the right console were the IFF (Identification Friend or Foe), 12-channel VHF radio, light controls and switches and the start-control panel. On the side were more circuit breakers, the seat adjustment switch, a manual canopy crank and a map case.