Beginnings

Ryan had proposed an experimental installation of a Westinghouse Model 19B of 1,400 pounds thrust in place of the standard General Electric I-16 of similar power in the production FR-1 aircraft and two airframes were allocated for evaluation. The installation was planned but never occurred. The follow on schemed improved models, the XFR-2 and -3 would have used the improved Westinghouse Model 19XB-2B of 1,600 pounds thrust but never flew. The later XFR-4 did fly as an engine test bed for the Westinghouse Model 24C (J34) installation. None of these installations included any form of tail pipe burning or reheat (hereafter afterburning).

Preliminary afterburner (AB) work at Ryan was initiated under BuAer contract NOa(s) 8605 and this proceeded apace using war-time German research on athodyd combustion by Dr. Otto Pabst. The initial February 1946 Ryan contract test report noted that the planned afterburner combustion work was very similar to their earlier study except that the incoming air temperature would be about 1,200°F. Simple tests were run using both gasoline and kerosene. It was reported that the incoming limiting air velocities would be higher and the combustion use less space than the smaller test setup, but that no problems in maintaining combustion were anticipated in the planned afterburner for a model 24C-4 unit.

In early 1946, Ryan planned the larger F2R-2 (Ryan Model 30) which would use a General Electric TG-100 turboprop in the nose and a Westinghouse 24C-4 engine in the tail being fed by two side flush intakes ahead of the leading edge of the wing. In relation to the development of this aircraft, Ryan proposed and received authority to charge experimental work on afterburning to contract NOa(s)-1322. BuAer approved coordinating the AB work with Westinghouse (WEC) on the understanding that WEC had indicated their approval as well.

Slipping engine deliveries of both the engine models and engine configuration mismatches to the drawings plagued the work as time progressed. One of the FR-1’s was diverted to be modified to F2R-2 configuration. Initial airframe design and specification reports did not include any information or data relative to the now planned afterburner installation.

In February, a General Electric I-16 engine and three tailpipe assemblies were allocated for work on the Model 30 project by BuAer. Since the General Electric TG-100 engine burned kerosene, Ryan asked that the I-16 engine be a kerosene burning engine, thus allowing Ryan to only provide for one type of fuel on the XF2R-2 aircraft.

In May, the first AB development report stated that no testing was underway but included design elements planned for in the unit. The burner assembly would provide a central main fuel inlet duct with radial fuel distribution outwardly through the struts to the burners. It would be a slip-fit inside the AB combustion chamber and attached only by a central tension bolt and mating joint in the central fuel duct (See Figure 1). Five pictures of the original test unit were attachments on the report but they were not retained in the file and have not been located elsewhere.

A new combustion/vaporizer section was to be designed and built, this having a double outer wall similar to the initial design. The fuel connections would be at the forward end with flow from front to rear between baffles, these having a much closer helical pitch. Six additional flat section vaporizer ducts would extend diagonally across the combustion chamber from the aft end of the annular vaporizer forward to the central burner connection. A reference nozzle would be used on the I-16 when supplementary combustion was not present. The inside diameter of the nozzle was 12.4". An eyelid type nozzle on the exhaust would be used.

An adapter diffuser section for the requested test J33 (as the I-16 was now designated) engine was to be designed and fabricated to allow the attachment of the AB combustion chamber.

The BuAer stated requirements were that the AB was to be designed to be attached to the exhaust extension flange of the then present X24C-4 engine without alteration. The control system had to be arranged to be operated by the same pilot control lever used for basic engine operation. BuAer stated they could not promise allocation of a 24C engine to Ryan for the development work. The cost estimate for the work was to be resubmitted to BuAer if altered from the beginning by the stated requirements.

Contract NOa(s)-1322 was terminated via Amendment 25 September 1946. Testing charges continued partially under contract NOa(s) 8216 for the XF2R-2 until the airframe items in the contract were cancelled on 6 March 1947. The airframe cancellation reason given was that General Electric’s TG-100 engine had ceased development and the TG-110 engine was cancelled. These being the only two suitable engines for the nose of the XF2R-2, it was decided to cancel the development contract. Expenses related to AB testing appeared to have continued to be allowed under contract NOa(s) 8205 which had been specifically created on 8 November 1946 to provide financial cover for the continuing AB work.

Model 1 and 2 Development Details

These were broadly similar with the differences noted where the Model 1 was varied:

General Description

The tail pipe adapter of the standard 24C engine was replaced with a diverging diffuser section. (The diffuser would be modified again in the Model 3 to improve internal flow characteristics. A circular plate was to be inserted under the ball and socket joint. This plate would also provide thermal shielding of the ball and socket joint and help with sealing against leakage.) A ball joint was provided to allow for vertical and lateral misalignment of the AB support fittings. In the Model 1, a bellows joint had been used at that location, but testing had shown that the high thrust being produced was being transmitted through the bellows and had resulted in serious deformation of the joint. The support fittings were mounted on opposite sides of the combustion chamber and were arranged to allow for expansion in the axial direction. The three sections of the AB were joined by means of flanges and clamp bands. The exit nozzle was provided with two eyelid type shutters to reduce the effective nozzle area to that corresponding to the “dry” non-AB operating condition. The Model 2 was not tested with the controllable eyelid nozzle. Two jack screws connected by flexible shafting to an electric drive motor were provided for controlling the eyelids.

The Model 1 vaporizer consisted of a jacket surrounding the combustion chamber which was of double wall construction. This formed an annular passage for fuel flow and provided a heat exchange area of 16.5 ft2. Six helical baffles, each making a one-half convolution within the length of the combustion chamber, were located within the length of the combustion chamber. Their purpose was to increase the flow path and velocity of fuel through the section. The baffles were seam welded to the inner wall to provide structural rigidity. The gasoline type fuel entered at the aft end and, after passing over the combustion chamber, was introduced into the burner grid. Testing using the lower mass flows of the I-16 test engine showed the fuel distribution to the nozzles was poor. This indicated that poor vaporization was being achieved even at a measured 160oF fuel temperature.

In the Model 2, the vaporizer was increased to 24.5 ft2 but added plate type heat exchangers located in the combustion chamber. These appeared as six fins in the radial planes. Two helical baffles making two complete convolutions were also added. Fuel now flowed by entering the forward end of the jacket, then aft along the combustion chamber and onward into each of the six internal fins. It flowed forward and then was distributed into the burner nozzles. Early tests showed a 29% improvement in fuel vaporization. Further improvement was needed as the vaporizer buckled prematurely during proof testing. The Model 3 would be redesigned to handle higher fuel pressures and would have even greater heat exchanger area. The fuel filled jacket was considered to provide cooling advantages to the metal and eliminate the need for additional air cooling around the AB.

Fig. 2. Combustion Chamber Exit. Paddle vaporizer fin design is visible.

Fig.3. Combustion Chamber Exit. Right side wall warping was due to uneven burning temperatures caused by incomplete vaporization.

A vapor type burner was utilized to obtain the shortest possible flame length and maximize flow velocity. This resulted in the minimum size combustion chamber. Forty-two (42) burner vapor outlets were provided in a symmetrical pattern. The burner outlet consisted of a disk with provision for admitting vapor to the upstream face. The vapor flowed outward and mixed with the highly turbulent air spilling over the periphery of the disk.

The planned fuel control to the AB was to be a function of the turbine discharge temperature. A constant speed variable delivery pump would furnish the required pressure and delivery capacity. A motor driven valve in the high-pressure fuel line would be used as a throttling device to control fuel flow and pressure at the vaporizer inlet. The impulse delivered from a sensing thermocouple located in the AB would be introduced to a control device which controlled the energization of the motor driven throttle valve. The fuel would enter at the forward end of the annular vaporizer. It was directed by baffles to flow in a helical pattern to the aft end of the combustion chamber, there entering six flat section vaporizer struts. It then flowed forward to the central manifold hub. The fuel would then be distributed radially by the burner manifold to the forty-two vapor outlets.

Fig. 7. Model 3 Vaporizer Nozzle Pattern Seen from Combustion Chamber Aft End. Horizontal bar across the bottom was a water-cooled rake.

A thermocouple element was to be arranged in the turbine exhaust to control the AB fuel flow as a function of the turbine outlet temperature. As the AB temperature increased, the turbine exhaust pressure and temperature would rise. This rise being sensed, the turbine exhaust temperature of the engine could be automatically made to operate at its maximum rated output when the afterburner was in operation.

A vapor type burner was utilized in both the Model 1 and 2 designs in order to maintain combustion in the shortest possible flame length and attain the maximum gas flow velocity. Forty-two burner vapor outlets were provided in a symmetrical pattern with each outlet consisting of a disk with provision for admitting vapor to the upstream face. This burner design was of the same proportions of a Pabst type. The Model 3 would be equipped with smaller fuel nozzles (0.125” reduced to 0.063”.) Also, provisions were made for an orifice plate to permit attainment of higher fuel pressures in the vaporizer.

The Model 1 had a 16.5" diameter nozzle. The Model 2 was tested with both a 14.0" and 16.5" nozzle. Testing indicated a nozzle area increase of approximately 50% was required to achieve the 3,000°F for afterburning. The thrust difference between a purely circular nozzle and the eyelid type was negligible relative to the probable accuracy of measurement, but the eyelid nozzle was a few (4.3) percentage points lower.

By the time of the XF2F-2 airframe cancellation, Ryan had progressed in their AB work on their Model 3. They had experienced difficulty in obtaining thermocouples that could reliably read in the 3,000-4,000°F range and had suitable durability. They were now being forced to use a General Electric I-40 engine for testing due to continuing unavailability of a 24C of any type.

Model 3 Testing

This was delayed until a model 24C engine became available, since the AB diffuser required a 24C output to produce accurate results. In the interim, minor testing using the Model 2 using the I-40 engine continued. This work was done to determine the internal component air flow (drag) losses. Completing the instrumentation in the test cell control room was planned. Performance of the AB on the I-40 engine would be predicted.

A specification was needed and was finally submitted in preliminary form by Ryan on 2 January 1947. This was for the Ryan Afterburner Model 39 for the Westinghouse 24-C Turbo Jet Engine, Ryan Report 3927.1, on Contract NOa(s) 8205. This document appears to be both a requirement statement and then a current design approach for approval consideration. Each section of this specification is summarized below:

General. The design was of a thrust augmentation (afterburner) device for use during takeoff and combat. Temperature increase was limited by the material properties. Cycle efficiency reduction was also a practical limitation. A two-position nozzle was being used to preserve the exhaust area when operating in the dry condition.

Operation and Control. The cockpit control was a two-position lever separate from the engine throttle lever. A possible change might be to combine the AB lever with the throttle lever as an over-ride position. Further testing and operation experience was needed before a formal recommendation could be made. Pushing the current control lever forward would open the main shut-off cock by means of a push-pull linkage, thus energizing the electric actuator to open the two-position nozzle and turning on the ignition system. The AB temperature control senses the temperature drop and opens the fuel control valve. The AB temperature restores the back pressure and raises the turbine temperature back to normal. The AB control would automatically maintain the maximum allowable turbine operating temperature regardless of aircraft speed or altitude. The first system to be considered would use an amplifier similar to that in a Brown flight test recorder. Four thermocouples wired in parallel would provide a fairly good average temperature indication. The second design under consideration would be that using four resister type bulbs in conjunction with a polarized relay as an alternative.

Preliminary Performance Data. Augmentation was to be guaranteed to be a minimum of 37% for static sea level conditions. Thrust loss in the dry condition at maximum engine RPM to be no more than 160 lb. The guaranteed maximum increase in fuel consumption in the dry condition would be 5 percent.

Installation. The AB was to be designed to attach to the 24C flange connection located 37.34" aft of the main engine supporting trunnion after removal of the normal tailpipe. Alterations to the engine would not be necessary. A quick disconnect band clamp was planned to be used, but initially a bolted connection would be used in the interim as changes to the engine flange design might prove to be necessary.

Weight. Estimated to be 242.4 pounds for the AB. Shut-off and throttle valves – 5.2 pounds. Electrical system – 10.9 pounds. Total: 258.5 pounds. Required to be provided by the aircraft manufacturer was engine driven pump type Pesco G13 plus adapter, these to be mounted on the emergency pump pad on the engine gearbox – 6.5 pounds.

Overall AB Installation Weight. 265 pounds.

Delivery. One (1) production type afterburner suitable for attachment to the 24C engine and usable in the Chance Vought XF7U-1 airplane.

Fig. 8. Model 3 as of 31 December 1946 Showing Measurement Pickups.

When the progress report for December 1946 was submitted, Ryan included a warning that the test schedule was predicated on the basis of using the I-16 engine. The contract delivery schedules were dependent of the availability of the 24C engine throughout the development program. The lack of availability of the 24C would seriously compromise the quantitative scope of testing. Expedited delivery of a 24C was requested.

BuAer reacted on 13 January 1947 to the schedule risk warnings due to the lack of an X24C engine at Ryan by transferring X24C WE002009 (an early X24C-2) from Allison to Ryan. Testing in January 1947 included the Model 1 burner evaluation. This suffered a vaporizer jacket failure similar to that experienced with the Model 2 and at the same fuel pressure of approximately 20 psi. In this case no external evidence of a pending failure was evident to the operator. The Model 3 produced a measured 23% augmentation running on the I-16 engine. A misaligned burner assembly resulted in bad fuel leaks that caused erratic combustion. The receipt of the X24C-2 engine interrupted testing while it was being installed. The Model 3 diffuser flange had to be modified to accommodate the hole pattern in the rear engine flange. Instrumentation installation in the test house had continued.

Other developments were the final work on a fuel vapor type valve adaptable to the Model 2. A coil type vaporizer that did not incorporate internal fins was designed. It would be employed to test upstream fuel injection, development of burner nozzles and flame holders. Burners and flame holders were being constructed for use on Model 3 and 4 vaporizers. Regenerative cooling of the nozzle and how it could be achieved was under study. The 2,000°F temperatures being experienced with the current on the Model 3 using the I-16 indicated higher temperatures were likely when the 24C was used in testing. Modifications to meter fuel instead of vapor appeared to offer simplified fuel control and was also under study. Changing to injection of fuel from the current four holes on the upstream side of the disk flame holder to a single hole in the center with the burner manifolds situated downstream from the disk was to be tried.

At the current point in the testing, little difference between trimmed eyelid type nozzles and parallel throat nozzles was found. The final design of the outlet nozzle was to be deferred until the final configuration of the AB was determined. An operating configuration of the eyelid type nozzle was being fabricated. Other configurations such as a central plug and a ring plug were being explored. These components would have to be cooled, complicating such designs. Given the eyelid design allowed the components to swing clear if the AB was in operation, it did not require a cooling solution. Ryan was fully aware of various investigations of nozzle designs via the reports from General Electric and NACA.

After correcting the fuel leaks on the AB in test with the X24C-2 (WE002009), running at 12,000 RPM, excellent stable combustion was observed with a steady blue flame extending aft of the burners only a few inches. The X24C-2 delivered its gas flow to the burner section of the Model 3 at approximately 100 ft/sec higher velocity and 22 percent lower static pressure than expected from the X24C-4. Both of these deficiencies increased the difficulty on initiating and sustaining combustion.

Either course would require alteration for 24C-4 operation and subsequent testing using a 24C-4 engine. Indications showed that with the possibility of operating with slightly oversize disks, stable combustion could be maintained at Military RPM. No tests of fuel rate or the drag penalty of such a configuration had been made, but would be in the future.

The theoretical performance of the Model 3 AB with various engines was included in the January 1947 report.

The first control system was to be an electronic amplifier type utilizing thermocouple sensing elements. An alternate design based on resistance bulbs in combination with a polarized relay would be available. Both designs would be pulsing types to provide maximum protection against hunting instability. Another control approach would be based on a follow-up type of control. It might make possible a more rapid control response in the event that the time constraints inherent in the engine and AB operational characteristics were relatively small.

In February 1947, BuAer reviewed the sample specification and asked that all future AB performance data be shown as a supplement to 24C engine performance. All performance data and plots from it should be in the form of thrust in pounds, fuel consumption in pounds per hour versus true airplane speed in MPH, not in the form of thrust horsepower and fuel consumption in pounds per second versus speed in MPH.

In mid-February, Ryan had to ask that the language of the contract be specifically modified to cover all expenditures related to providing accessory equipment and instrumentation required in AB development. Considerable costs had been disallowed and Ryan asked for a prompt, retro-active amendment to be processed so that reimbursement could be made for the disallowed costs. This was processed as Amendment 2 of the contract dated 5 March 1947.

BuAer had requested that Ryan inform Chance Vought and BuAer as to the structural adequacy of the XF7U-1 AB. Ryan asked Westinghouse on 24 February 1947 to state the 24C flange loads for XF7U-1 non-augmented flight conditions be provided based on the planned model of 24C to be used in the XF7U-1. Corrected installation duct losses (but not for inertia loads) was also desired. Chance Vought had earlier provided the maneuvering loads of the XF7U-1 to both Ryan and Solar.

Turbine information was required as well to allow for proper estimation of material strength reduction during AB operation. It was requested to be in the form of Mollier-type diagrams showing total pressure, total temperature, fuel and air flow and engine rotational speed as ratios of compressor inlet conditions.

The test results from February were reported in Report 3927-3 for the Model 39 dated 1 March 1947, under the Ryan master model number for the AB under development. The other model numbers referred to were actually sub-design model numbers. All testing covered in this report was done using the X24C-2 WE020009 and the Model 3 AB with minor modifications for each test. The testing results and conclusions were to be applied to a “Model 6” AB that was being designed. The most important conclusion was that, with increased fuel input and refined internal design, the 24C-4 augmentation in excess of the guaranteed performance could be achieved. The combustion chamber temperatures which had been maintained in tests at reduced engine RPM without incurring adverse mechanical effects was used as support for the conclusion. Only minor deformation of the heat exchanger fins was cited as a mechanical deficiency. Once again, it was emphasized that the optimum 24C-4 burner design could only be developed using an actual 24C-4 in the testing.

On 17 March 1947 Ryan was instructed by BuAer to return the I-40 Unit No. 44 and all testing parts to Patuxent River for disposition. This engine had been used for the XF2F-4 AB testing. The I-16 would remain at Ryan for use in order to lower the running time on the 24C-2 engine as the entire 24C engine situation at that time was critical. The delivery delays were being caused primarily by a long strike then underway at Westinghouse. On 20 March 1947, Ryan confirmed that their design would support the loads on the AB laterally and that the flange could carry the engine connection loads, both according to the earlier information from both BuAer and Chance Vought. On 28 March 1947, BuAer asked for costs and a delivery schedule for six standard tailpipes and six afterburners for the XFU1 program. The tailpipes and afterburners were to be interchangeable without rework to facilitate airframe testing. The Chance Vought’s required configuration of the tailpipe was ordered to be sent to Ryan. The request was responded to on 16 May 1947. Six AB units would be $49,996.02 and six straight tailpipes would be $6,091.74. The total cost was to be $56,087.75. Delivery would be complete within six months after receipt of a contract, with deliveries beginning 3-4 months after from contract start, depending on availability and procurement of accessory control items. The above memo has written notes on it related to the X24C-2 and the scheduling of its return to Westinghouse for overhaul. It also refers to the use of the pending X24C-4A (WE002024) being shipped for AB work.

The progress report of 10 April 1947 covered the activities and results during the month of March.

Fig. 16. Model 3 with Tubes Replacing the Vaporizer Fins.

Model 3

The Model 3 AB, at 12,000 RPM, turbine out temperature of 640°C and a nozzle exit diameter of 20" attained a gross thrust of 3,200 pounds during steady afterburner operation. This represented an increase of 44 percent over the Military static sea level thrust of the X24C-2 with the AB not operating. Higher increases were thwarted by apparent compressor stalling, a common problem of the X24C-2 when operated at 12,000 RPM or at rated turbine out temperature. Testing at this speed and temperature proved to be very difficult. Only very small further increases of the Model 3 appeared to be possible using the X24C-2. Calculations showed that using the current Model 3 on the X24C-4 engine when delivering 29.1 psi turbine outlet pressure would produce a 38% in augmented thrust over a straight tailpipe. Even increasing the maximum operating pressure of the vaporizer from 20 to 40 psi did not raise the fuel rate high enough to obtain maximum afterburner performance. The weakness of the vaporizer limited the fuel pressure, so cold liquid fuel had been injected into the diffuser just aft of the turbine. Satisfactory AB operation and controllability accompanied with a combustion efficiency of about 75 percent was possible at military power operation of the engine. Each vaporizer fin was first reduced in size by 50%, then replaced by a 1" outside diameter stainless steel tube. It resulted in improved fuel flow stability and some reduction in internal drag loss. No distortion or deterioration was observed in the tubes to date.

It was reported on 2 June that X24C-4 engine WE002024 had been damaged in testing when a second stage nozzle inner shroud seal strip had broken off. This resulted in numerous dents and fractures in several second stage stationary blades along with rubbing on the forward edge of the second stage of the turbine disk. The engine was replaced with WE002031 and parts for WE002024 were requested for local repair of the engine.

The burner operation was best when using the 1.75" diameter vapor burners used in prior tests. A 0.5" wide strip was used to join adjacent burner disks; these added additional flame holder surface. It was believed the arrangement could be simplified when the use of the X24C-4 testing began.

The electronic control unit was used in some tests and demonstrated the ability to hold the turbine out temperature to the rated value after the AB combustion was initiated. Engine RPM could be varied from 8-12,000 RPM without the turbine out temperature varying or fluctuating during continuous operation. A turbine out temperature rise of about 30°C would be experienced during the starting cycle. Variations in engine performance did not affect the control’s ability to control the turbine out temperature by varying the AB temperature. The control unit was being modified to permit automatic coordination of ignition, eyelid operation and fuel control from the start of AB operation to shutdown. This would require only an on/off switch to operate the AB on the test stand.

Starting the AB at 12,000 RPM was not difficult. The pilot cone previously used to obtain good starting and stable burning was found to be unnecessary. The vapor control valve, though now installed, was found not to be necessary as the external control system operated entirely satisfactorily.

Model 4

This used a coil type vaporizer and a gutter type flame holder installed in combination with an upstream injection ring. Inadequate fuel pump pressure made it impossible to obtain operation at high engine speeds. The tests would be repeated with the new high capacity pump that had become available. The goal was to determine the advantages, if any, of maintaining the liquid or partial liquid phase in the preheater. The results would then being compared to the Model 3 results.

The construction of the vaporizer was of four 0.5" outside diameter stainless steel tubes wound as parallel helices on the inside of a stainless steel shell. Fuel entered the aft ends of the tubes and transmitted from the forward ends to an upstream injector. Superheating of the fuel had lowered pressures beyond the capacity of the test pump. An AB start was found possible during engine start because of the flames coming from the engine nozzle during the start sequence. Operation lasted 10 seconds in the diffuser upstream of the gutter type flame holder as the speed increased, then the AB experienced a blow out.

Model 6

Fabrications drawings were completed and released with the exception of the burner grid and associated details. This was considered a production design to be used with the X24C-4.

Report No. 3927-21, Model 39, 29 April 1947. Description of the Afterburner Control System (Key elements included here)

The AB as designed was to produce an increase in thrust over the basic thrust of the 24C-4 power plant. It was intended to be available for short periods of operation and require the pilot to only position a single on/off control to operate.

It was stated that while Ryan was fully aware of the possible benefits of development of a fully variable nozzle opening size in overall engine power and economy, in the interests of expediency, the current AB development program was predicated upon the use of a simple two position nozzle. Further development using both flight testing and in the test cell was needed to move forward to a fully variable nozzle.

A non-numerical chart was included showing the performance improvements available by comparing a conventional nozzle, a large fixed nozzle, a two-position nozzle and a variable nozzle. (This chart can be interpreted as a sales pitch to encourage BuAer to continue development funding beyond the current design which was being schedule constrained for production availability for the first flights of the Chance Vought XF7U-1.)

End Control System Report

Fig. 17. Model 6B Vaporizer Section Showing the Ignition Device.

On 6 June 1947, BuAer notified Ryan that the NACA Cleveland wind tunnel was available for AB testing if an afterburner could be sent and arrive by 1 July. After that time the wind tunnel was not going to be available. The latest specification and drawings on the AB configuration were to be sent to the NACA Flight Propulsion Research Laboratory in Cleveland, OH ASAP and the Bureau notified of the action taken. The cover memo (dated 12 June 1947) on the May 1947 progress report (dated 1 June 1947) added that additional testing had been done on the Model 6B afterburner. Some improvement in operation was observed by minor modification, but the burner operation was still not stable throughout the whole range. Starting at Military was satisfactory with the eyelids closed. Burner operation improvement was now the main focus.

The May progress report itself revealed that the wide convolutions of the Model 6 resulted in excessive warpage and cracking of the shell. The Model 6B attempted to correct this by detailed redesign of some of the welding joints and also by use of heavier gauge materials. In view of the fact that the Model 3, with its narrow convolutions, had withstood considerable testing without cracking, a Model 7 design similar to the Model 3 with the addition of regenerative cooling over the nozzle surface, an alternate burner using a cascade of three concentric flame holders progressively staggered in the upstream direction from the outer ring was being constructed. The flame holders were to be in the form of flattened tubes which would serve to inject preheated fuel upstream from the openings in the front face. Fuel would also be fed from the vaporizer jacket to the rings through six struts which also acted as structural support for the flame holder rings. As the struts were slanted upstream, it was expected they would provide additional fuel preheating to improve combustion.

The Model 3 AB had been tested with a water spray ring installed within the nozzle and achieved 4,300 pounds thrust at 11,900 corrected RPM and a turbine out temperature of 1,315°F. This was a gain of 48% over the XFR-4 straight tailpipe. The fuel rate was 9,400 pounds of fuel an hour at a combustion efficiency of 76.5%. The Model 7 design would attempt to reduce the high running drag and produce stable burner operation. The Model 6B spray ring had achieved a 100% increase in fuel rate without an increase in the necessary fuel pressure.

The X24C-4 engine (WE0020031) replacing the damaged X24C-4 (WE002024) was tested with the FR-4 tailpipe and demonstrated a 105°F higher turbine out temperature and slightly lower turbine out pressure. From an AB standpoint, the engine was deemed inferior to the prior one. The original cover memo noted they had experienced considerable high amplitude low frequency vibration during some testing in some configurations. These vibrations may have contributed to the engine damage previously reported. Ryan’s opinion was that, although considerable progress had been made, an extension of the completion date of the contract would be necessary to achieve a unit suitable for service testing. A separate letter would be sent requesting such an extension.

The components of the automatic control system had been installed in the test house and initial tests of the thermal switch on the eyelids had been satisfactory. Packaging of the electronic fuel control and the fuel control valve had been completed. All efforts were turned to completing one AB and its associated components suitable for delivery.

On 2 July, in a telegram, BuAer told Ryan to suspend AB testing until the cause of failure of the engines being used in the testing could be ascertained. Ryan responded on 3 July with a long memo covering the three engines and detailing their problems. In a second memo on the same date, photos of the damaged parts were forwarded. A summary of the information is above.

No conclusive cause of the faults had been determined. Only the “occasional” rough combustion was put forward as possibly aggravating the service life of the part. The requested repair parts were mentioned and were asked to be expedited.

In the BuAer representative (BARR) response to Ryan’s memo on 8 July 1947 regarding the engine faults, he indicated that the failures were considered to be more attributable to vibration and buffeting imposed on the engine by unstable combustion in afterburning than the contractor indicated. Such periods of rough running were considered to more than “occasional” as Ryan stated. The Westinghouse field rep concurred with the BARR’s exceptions. The tech rep noted that the scored rotors could continue to be used in testing. The requested replacement rotors could not be balanced at Ryan due to lack of such facilities and that similar failures could be expected in 10-15 hours of continued testing. The parts requested should be forwarded and all such resumed testing be done under the watch of a special representative from Westinghouse or the Bureau of Aeronautics. This would allow work to continue and the cause of the failures of the AB and engines investigated simultaneously.

The July Progress report which covered testing in June, was delivered on 18 July 1947. It including a testing activity chart as the first page, reproduced here for readability.

Paragraph IIC related to the modifications to the Model 3B. The first removed the disks from the upstream ends of the fuel injector tubes and the adding of 0.625"-wide strips behind the radial struts. The latter was supplemented with the interconnecting strips forming a hexagonal pattern. The second modification was similar to the first modification except that the strip pattern was solely of two concentric strips 0.625" wide. (This was the AB configuration forwarded to NACA Cleveland for wind tunnel altitude testing. The test unit would be replaced with a production version when available.)

The Model 7 AB had been tested with the burning arrangement illustrated in Figures 6 and 7. After three hours of vaporizing operating time there was no serious evidence of structural bulging of the vaporizer convolutions or any visible leaks. The eyelid mechanism showed some backflow gas leakage at intermediate eyelid settings. Eyelid fitting was critical to sealing and free operation. Radiation shields were being added between the eyelid operating jacks and the vaporizer in an effort to correct overheating of those units.

Blow-out was experienced and a 2" flame holder channel was added to the forward ring. High fuel pressures were relieved by drilling more fuel orifices in the sides of the support tubes. High drag resulted in the intermediate fuel distribution ring being removed. Regardless of the changes, the Model 7 AB performance was inferior to the Model 6B in blow-out, stability and internal losses. Another fuel system with three concentric flattened rings (Fig. 8) was constructed, this allowing the mounting distance to be moved forward of the flame holder. This distance appeared critical to combustion efficiency and elimination of roughness. Testing was on hold awaiting the engine parts and repairs.

The control system had been tested in simulated operating conditions in the lab and had been satisfactory. It would be used with the Model 7 when testing resumed.

General improvements underway were: Double wall construction similar to the current construction but having formed convolutions on the inner shell to prevent buckling and a vaporizer assembly fabricated from steel tubing emanating from a heavy walled header section with the exit nozzle. A revised exhaust cone assembly was in fabrication, consisting of wider supporting struts for the inner cone and allowance for free expansion of the cone to more adequately resist the imposed operating stresses. Provision was made for installation of a pilot burner with ignition plug. The latter would be used to test an upstream injection system. Alternate eyelid arrangements were being studied. These would contact a spherical nozzle outer surface to give more area for sealing. Other designs to produce a more circular orifice in the closed conditions were being studied.

The July progress report (3927-8) was sent to BuAer on 18 August 1947. The cover memo stated the contract was about 80% complete with the design done and one production unit for the “24-D” engine suitable for the XF7U-1 90% complete. Testing had been held up since 1 July due to the failure of certain 24C parts. Replacement parts were in shipment but not received at the time of the report posting. The contract completion date was 8 July 1947 and a three month’s extension had been applied for.

The failures had all been reported in the June progress report. The July report now issued the modified statement regarding the causes, stating that the AB testing had been at or near the maximum engine operating limits and this was accompanied by varying degrees of pulsation which was associated with the combustion phenomenon. These may had been contributing factors in accelerating the failures.

Fig. 25. Model 7 in August 1947, Modified Burner Section.

Planned work moving ahead was testing the modified burner in the Model 7 followed by testing of the automatic control system. Following that, investigation of the characteristics of the pilot burner cone in conjunction with the No. 4 vaporizer would be conducted.

The testing at the NACA Cleveland wind tunnel was delayed by the preceding program in the tunnel. A Ryan representative was at the center to expedite installation and operation of the Model 3 AB.

Amendment 3 to the contract was issued on 13 August 1947. This extended the contract and reset the completion date to be eleven months instead of the original eight months. No additional funds were included. The new completion date would be 8 October 1947. The cover letter acknowledged that due to failure of engine parts to arrive during July and early August, a further extension would likely be necessary. BuAer notified Ryan on 7 September 1947 that since the contract was “Confidential”, all release of information regarding the AB development program being conducted under that contract had to be through the BARR for coordination with BuAer.

The cover letter announcing the change in endurance requirements included a statement that the BuAer agreed with the request to lower the classification of the contract to “Restricted” and so ordered the downgrade. The classification change became Amendment 5, issued 17 October 1947.

Data was forwarded to Wright Aeronautical in September regarding the performance of the AB design for the XF7U-1 (not named, referenced as a “specific airplane”) operating on a 24C engine. It noted the value of a variable nozzle for optimum attainment in aircraft performance over that of a fixed nozzle. Which engine Wright was contemplating development of an AB for was not stated.

Some contract changes would also affect the Solar Company’s contract which was developing an AB for the 24C in parallel with the Ryan efforts. The BARR on both of these contracts suggested to BuAer on 25 September that the Solar contract be modified to the same performance requirements as the Ryan contract and that the Ryan contract be further modified to replace the 150% augmentation over the standard sea level static Military performance of an I-16 engine to that of a 24C engine.

On 1 October,1947, the BARR asked Ryan to reconsider the price for the testing of the AB since the requirements had been considerably reduced from 150 total hours to 75 total hours by Amendment 4 in process. Originally, Ryan had bid $42,130 plus a fee of $2,106.50 for the 150 test hours. Ryan was to advise the Bureau at their earliest convenience as to a possible reduction in cost that could be acceptable.

The BuAer Power Plant Division manager asked on 6 October as to the status of the Ryan program versus the Solar program. The brief Cleveland wind tunnel testing (which was incomplete due to engine damage) had indicated the early (Model 3) Ryan AB would not operate successfully above 25,000 feet due to unsatisfactory combustion and susceptibility to blow out. The data from the Cleveland testing was found in NACA Research Memorandum RM No. E8J25e, published 13 December 1948. This report, “Altitude-Wind-Tunnel Investigation of Tail-Pipe Burning With a Westinghouse X24C-4B Axial-Flow Turbojet Engine”, combined the results from four (4) different manufacturers without identifying them. However, the results for “Configuration D” can be seen to be for the Ryan Model 3 design, the identification being based on a picture of the Configuration D afterburner at the rear of the report. The Ryan afterburner having those distinctive helical passages surrounding the combustion section, these allowing fuel flowing through them to be preheated. Some specifics on the AB were included:

The results of the NACA tests in 1947 must have been relayed to Ryan verbally in a timely manner, as Ryan was stating in memos in October that further altitude and forward speed testing would be necessary to determine the current operational characteristics. Attempts were being made to schedule such tests at Cleveland, but the facilities at either Fontana or AEL might have to be modified to accommodate the tests.

The memo went on to mention that four (4) Solar production ABs were due for delivery to Chance Vought on 15 November 1947, they would be required to be subjected to the same altitude and forward speed testing program. The final results would allow the determination of which of the two AB designs would be most applicable to the XF7U-1 airplane.

Chance Vought had recently stated that the XF6U-1 could be modified to accommodate an AB designed to work with the 24C engine. Modifications required to the XF6U-1 to accept such an AB would take approximately five months.

A survey of the AB situation by the Air Forces across the services and manufacturers noted that while augmentation could be expected from 30% to over 100% above Mach 1 using various liquid injection or bleed and burn designs, using the airframe’s basic fuel supply in a liquid injection system was the simplest. Navy projects at that time were for the XF7U-1, XF6U-1, XF10F-1 and XF4D-1. The Air Force projects were for the XP-90 and XP-91. Engine manufacturers were developing ABs for the J33, J47 and J37 engines. In summary, the 10 October memo stated the Air Forces recommended that Westinghouse be asked to manage the development of AB’s for the 24C engine, thus avoiding a very difficult situation of divided responsibility between the Power Plant (Division of BuAer) and AB contractors. Progress to date in that area was considered satisfactory under the existing conditions, with the Navy leading the effort.

Contract Amendment 6 was issued 10 October 1947. This extended the contract completion date to fourteen (14) months from the contract start date of 8 November 1946. This moved the expected end date to 8 January 1948. The completed single AB example was to be delivered to the BARR at Ryan by that date. A wire was sent on the same date as the extension was granted, this stating that it was urgent that the AB be completely developed by that date. On 14 October, BuAer informed Ryan that their assumption that Chance Vought would supply the AB fuel pump was in error, since the contract required that the delivered production type AB include a fuel pump and all other accessory components to make the AB a self-contained unit.

Ryan asked to be allowed to return WE002031 for overhaul on 17 October 1947. The first stage turbine blades were experiencing rubs and had fallen below the stated minimum clearance of 0.060". The service representative stated the blade growth was the cause of the unsatisfactory blade clearance. Recent attempts to start the engine with a straight tailpipe had experienced torching and failure to start. A third attempt without the tailpipe experienced the same results. Apparently, the blade rubs were preventing the engine from accelerating properly. It was requested that both turbine wheel sets be renewed as well as the second stage nozzle diaphragm. Engine WE002024 (21 hours since last buildup) would be used to continue the AB testing. The BARR stated the replacement for WE002031 would be used as a standby spare for the Ryan and Solar contracts during the forthcoming endurance afterburner tests.

The BARR continued to investigate the causes of the WE002031 blade growth and rubs and found that hot starts due to low voltage at the starter motor was the likely root cause in the slow starts. He stated in a wire on 20 October that the voltage at the batteries had read 24 volts with no load, 18.5 volts at engine start beginning and read 11 volts at the starter. This was insufficient to properly accelerate the engine and resulted in hot starts. In addition, the engine tachometer did not indicate until 3,000 RPM had been reached and was sluggish at higher RPM values. He formally asked Ryan to develop a proper source of starting power before any more starts were made. Ryan was making attempts to obtain a new tachometer and generator.

The lack of a clear, overall management guidance structure was highlighted by the BARR’s request on 22 October 1947 regarding AB production proposals from both Ryan and Solar. As background, he referenced the ongoing engineering studies in connection with the production sales orders for Grumman, Chance Vought and McDonnell that Ryan was then conducting. He noted that Ryan had also exchanged classified information in connection with AB design with Douglas, Santa Monica. Solar had recently been in contact with Chance Vought regarding production AB models. The situation raised the question as to whether aircraft companies interested in procuring ABs must deal through the Bureau or could negotiate directly with Ryan and Solar without including the Bureau. The BARR stated he felt that production of ABs should be under BuAer’s control until they were declassified.

At that time, problems with both AB designs were felt to be resolvable by the January or February 1948 time frame. He recommended that the acceptance tests be run using the development ABs, as the tests would involve vital specifications (not specified) that had not yet been written into the contracts.

Also on 22 October, the BARR notified BuAer that Solar, unlike Ryan and due to contractual requirements differences, did not intend to deliver many of the necessary component parts with their ABs, but would make recommendations as to the type to be used and sources of procurement in the near future.

In assisting the Ryan test efforts, on 28 October BuAer authorized the operating time on 24C WE002024 be extended to 100 hours (from 50) as long as the inspection procedures were observed. Efforts were underway to obtain another engine to replace WE002031. An order received the following day authorized the BARR to return WE002031 to Westinghouse in Essington for overhaul and modification into a X24-4B engine.

The progress report (3927-9) that covered the testing work done in August and September arrived on 28 October 1947. The testing (summarized) covered was:

Afterburner Test Program

Testing was resumed during the period. A long series of tests involving eight different burner arrangements using the Model No. 4 and Model No. 7 vaporizers. The burners were designed to provide decreased aerodynamic drag losses both with and without afterburner operation.

The planned next test series would be made with the addition of superheater struts in the combustion zone and extension tube and spouts to the fuel injecting grid. The test AB would duplicate the Model 3 AB in principle with the added feature of a regeneratively cooled exhaust nozzle. A No. 8 vaporizer was under construction, consisting of axial tubes distributed on the inner circumference of the combustion chamber shell. The downstream header would form a cooled exhaust nozzle and the floating upstream header would support the burner grid and allow for thermal expansion. The larger heat exchange surface would provide a high degree of fuel preheating.

The September Progress Report, Appendix A, Shows Results of Two Position versus Fully Variable Nozzles on Thrust

With primary operable engines available, all the contract extensions requested having been granted and with just over two months left in the contract until the single production AB had to be delivered, the Bureau on 3 November sent a memo clarifying the acceptance criteria for the unit. (The memo also addressed various items on the Solar contract of the same general nature.)

The unit should be capable of meeting all the specified requirements of Item 3 of Section C of the contract and have successfully completed the endurance requirements as recently amended before acceptance by the BARR. Any statements by any party stating that the endurance testing program was eliminated by the Bureau was entirely in error. The Bureau deemed such a testing program was most essential if the AB was to be completely evaluated before airframe installation.

The Bureau stated it had no objections to continued development until the allocated funds had been expended. At the conclusion of the expenditure, Ryan would have to deliver to the BARR one production type AB meeting all the specifications and requirements of the contract. “There is no significant bearing on the contractual items to be delivered to this bureau as to whether such funds are sufficient enough to run through until 8 January 1948.” (In other words, Ryan would have to pay to finish the work if not complete when the Navy funds ran out.)

The BARR issued a memo to the Chief of BuAer on 5 November regarding the earlier Cleveland tests of the Ryan AB. The reason for this memo is not clear. The BARR stated that the AB sent to Cleveland had been extensively tested and revised at Ryan but was the only AB available for the test given the time constraint of tunnel availability at Cleveland. When the engine in Cleveland experienced problems that ended testing, Ryan had stated they would send a replacement at a later date.

The Bureau notified the BARR on 7 November that the Aeronautical Engine Laboratory (AEL) was preparing a 24C-4A for shipment to the BARR, estimated shipping date before 14 November. They also stated the 2nd stage turbine seal failures were being found on other 24C engines unrelated to AB testing and Westinghouse was working on a solution.

Ryan notified the BARR on November 7 that a fuel pump and electric motor to drive it would be sent to Chance Vought. The procurement of the electric motor was in progress but no specific part number was available yet. The pump would be similar to the Pesco No. S-1557 pump used in AB testing at Ryan.

The October Progress Report was sent on 11 November 1947 and covered the month of October’s testing. The BARR’s cover memo best summarized the activities.

1. The contractor was attempting to reproduce configuration and performance of his No. 3 afterburner reported in Ryan Report 3927-4, which appeared to be the best configuration to date.
2. The nearest approach to a satisfactory configuration at the present was the No. 7c afterburner wherein smooth combustion was reported; however, the thrust augmentation was not considered within a satisfactory range.
3. The No. 8 afterburner had not shown satisfactory results to date.
4. The bench test of the control system was very good; however, it would require (an) “actual” operation test to prove its value.

Further, the BARR supported Ryan’s claim that testing time lost was primarily due to the inability to obtain engine repair parts during the prior periods.

A series of photos and drawings was supplied.

Fig. 28. October Progress Report Figure 1. Model 8 Test Configuration	Fig. 29. October Progress Report Figure 2. Model 7D Vaporizer	Fig. 30. October Progress Report Figure 3. Model 7D Vaporizer View from Diffuser

 

Attached to the October 1947 Progress Report were three pages of outlines of the various ABs tested or in development. These were there to help the report reader follow the details of the report. Not to scale. Some are a bit warped from the records' storage condition.

On 13 November 1947, BuAer gave direction to Ryan and Solar regarding any negotiations with aircraft companies, stating they were at liberty to conduct such negotiations provided the aircraft companies were Army or Navy contractors, and providing that Solar, Ryan and the aircraft companies adhered to the security regulations imposed by the Restricted classification of both the Ryan and Solar developments. To keep the BARR and BuAer completely aware of any negotiations, both companies were requested to provide the BARR with a copy of any correspondence involving aircraft companies desiring information for a Navy development.

The BARR updated BuAer on 14 November stating that there was no likelihood of Ryan completing the contract by 8 January 1948. In his opinion, he felt it would take until “December 15, 1948” to complete the contract. (The date might be an error, as in a later paragraph in his memo, he now “estimates February 15” as the date when both companies should complete tested afterburners.)

On 28 November 1947, BuAer authorized using 24C engine WE002025 for 100 hours of operating time in testing if the inspection procedures were observed. A 24C spare parts authorization was sent out on 1 December for 8 bolts for the 2nd stage turbine wheel and 24 special washers.

Fig. 34. Model 9 Prior to Testing.

Official Performance and Endurance Testing

It was reported on 29 December that testing of the (presumed) Model 9 AB had begun on 24 December and was estimated to be concluded on or about 25 January 1948. After five hours one of the two flame holder rings broke off. It was being repaired and not considered a major problem. The automatic control system had not functioned. (It is not clearly stated whether it was meant that it failed to function or that it was not part of the testing yet.) Normal afterburner operation had been observed. UNIVIS 54 lubricating oil was in short supply and this might delay further testing. (This was the standard 24C-4A lubricating oil at that time.)

On 30 December 1947, Ryan wrote a memo to the BARR expressing concerns over the suitability of the current 24C normal engine controls during AB operation. Referencing their considerable operating experience with the normal controls, they stated a more limited knowledge had been gained regarding the exact degree of coordination between fuel flow variation and exhaust nozzle area change required during the initiation and stopping of afterburning. Tests regarding integrating the various components of the AB controls with the engine controls were expected to provide a more quantitative understanding of the information requested previously. (The referred to memo was not in the files.)

Even as the Ryan contract was ending, BuAer assigned on 21 January 1948 engine WE002028 as a spare engine to the BARR for Ryan or Solar use. The engine pool for AB testing was now WE002028, WE002017 and WE002025. No other engines could be made available.

The BARR officially accepted the Ryan AB Model 9 from the endurance test on 30 January 1948 and shipped it (minus some components) to NACA Cleveland on the same day. Missing was the electronic control unit on which some components needed further testing and the Pesco fuel pump, which had not yet been received from the vendor. On 5 February 1948 Ryan requested the contract be extended to 31 March 1948 to allow completion of these outstanding items.

On 3 February,1948, BuAer authorized both the Ryan and Solar test facilities to be used for production ABs being built for both Army and Navy airplane programs. BuAer would control the test schedule and had to authorize all testing prior to proceeding. All operating expenses for the test facilities and the repair and replacement of Government Furnished Equipment due to contractor’s negligence would remain the contractor’s responsibility. With engines in very short supply, both contractors were urged to use special caution during testing. In the meantime, Ryan had continued testing the automatic control unit and the BARR reported on 17 February that the changes they were making to the temperature control fuel regulator would produce satisfactory results. They intended to test the modified control on one of the production afterburners being produce (by Solar) under Contract NOa(s) 5312 in the next two or three weeks. The BARR also noted that the variable nozzle control was not automatic but functioned from the ignition switch with a time delay. The system functioned with highly reliable ignition and burning but did not provide for false starts and blow outs. His opinion was that the control of the eyelids should be automatic and activated based on ignition temperature or pressure variation.

Contract Amendment 7 was processed on 25 February, extending the contract end to 17 months from contract start, the end date now to be 31 March 1948. No additional funds were allocated. Upon receipt, on 25 March, Ryan asked for another extension to 30 April to enable the outstanding items to be delivered from their vendors.

Engine WE002028 was shipped to Ryan on 8 March 1948. On 29 March 1948, the BARR wrote a memo to BuAer giving his view of the data in Ryan’s report 3927-31 on the Performance Specification (no copy found). He noted that AB performance was greatly influenced by engine performance. This was visible in the observed data from the last progress report on the endurance testing. The BARR felt the values Ryan submitted reflected the maximum values that could be expected of AB performance. In order to aid the Bureau in valuation of the report, he submitted the average values from Ryan’s “Summary of Endurance Test Data”, page 1 of that report.

Since the values did not represent the final configuration as improvements were incorporated subsequent to some of the test values, the BARR felt the contractual requirements were complied with in the final configuration with an expectation that a small overweight condition would likely be corrected on a production contract. He went on to say that NACA did not consider it important for their tests that the automatic control was not delivered with the afterburner. The control itself had undergone considerable changes since the 1,000 cycle test at Ryan. No satisfactory engine demonstration could be made until recently when the fuel flow controller was satisfactorily demonstrated. The time delay system of ignition and eyelids (mistyped twice as “high loads”) was not considered satisfactory by the BARR. A pressure sensitized control system for the eyelids and replacement of the electric screw jacks with air or hydraulic actuators was under consideration. The intent was to speed up the entire ignition cycle.

Regarding the payment of the 5% profit fee on the completed contract, the Chief of BuAer ordered the BARR on 30 April 1948 to authorize payment. Inasmuch as the afterburner was ordered sent to NACA after only a little over 30 hours of total testing without incident, the failure to complete the objective seventy (70) hour endurance test was the fault of the Bureau.

Ryan reported on 10 May 1948 that the electric driven fuel pump and automatic fuel control system to complete the afterburner assembly had been shipped (to the NACA). They submitted report 3927-36; Ryan No. 9B Afterburner Fuel Control System directly to BuAer on the same day. (This report was not found.) The above items closed out the deliverables on the contract and BuAer shortly thereafter settled financially with Ryan.

One More Chance

Fig. 35. Model 9B Installed at Ryan Ground Test Facility May-June 1948

Unaccountably, the Ryan Model 9B got tested again, not at NACA, but at the Chance Vought facility in Stratford, Connecticut. How it got re-routed there is unknown. Under Contract NOa(s) 9602, which covered Chance Vought ground testing of afterburners prior to installation in an XF7U-1, it was installed in the test rig on a 24C-4B serial WE002025. Although the fully automatic control system was in Chance Vought’s hands, it was not used in the following tests.

Beginning 19 May, 1948, test runs were made to adjust the eyelid closed position and obtain engine operating limits. The Solar exhaust collector was installed instead of the Ryan collector. The runs were at 10,000, 11,000, 11,500 and 12,000 RPM with the afterburner off. On May 21, the engine governor characteristics were noted with the AB off and eyelids both closed and open. The manual controls only were used in these and all subsequent tests. The Lear screw jacks and flexible drive shafts took approximately seven (7) seconds to open or close. With the AB fuel flow set at 4,000 lb/hr at 10,000 RPM, another run was made on 21 May at 10,000 RPM. Performance was very rough and unsteady, with thrust variations of plus and minus 500 pounds. It was considered this was due to a lean AB fuel-air ratio.

On 26 May, runs were made at 11,000 RPM and on 1 June one run was made at 12,000 RPM. No reliable data was taken due to the rough performance and uneven thrust. All starting was done with the eyelids closed and then these were opened manually as soon as AB burning was initiated. The measured elapsed time for an AB start from initiation was six (6) seconds at 11,000 RPM, followed by a seven (7) second lag for the eyelids to open. The delay was caused by the fuel system having to fill the fuel annular chamber of the exhaust nozzle and the fuel struts joining the annular chamber with the burner ring, the required fuel volume being approximately 2.33 gallons. At 12,000 RPM, the eyelid opening time was measured as 4.12 seconds, taken during measurements for developing a blowout detection switch. On all runs, performance was very rough and uneven. The Ryan AB was removed from the engine on 7 June 1948. Chance Vought found the AB unsatisfactory. Thrust loss dry of the basic engine had been measured as from 5-7%. The AB shroud diameter was too large to fit the smaller F6U-1 and a replacement smaller Solar shroud would not have allowed for adequate cooling of the AB.

The disposition of this production level Model 9B AB is unknown.

Post-Script

Ryan later accepted contracts to build afterburners to other company’s designs, such as the Westinghouse J46, but their further research on afterburners appears to have stopped with the Model 9B  failure.