Monday, December 18, 2017

The Radome Road to a 32-inch Radome

HISTORICAL BACKGROUND:

During the design phase, just as the mission of the F4H-1 changed several times, so did the radar that was proposed to help the Phantom see its prey.  Early on, when the proposed aircraft mission was ambiguous, many proposals emerged. When it looked like the aircraft would be primarily an attack aircraft the choices were:
  • Westinghouse AN/APQ-56 (modified) with a Teledyne AN/APN-79 Doppler set for navigation
  • North American Aviation Autonetics Division’s NASARR (North American Search and Ranging Radar) which was being developed at the same time. This radar was optimized for the attack role and thus fit in nicely with the anticipated role of the AH-1 and the early F3H/F4H designs. The early F4H-1 models showed this pedigree with their 24-inch radome which had been designed around the NASARR radar requirements.
But in 1956, when BuAer changed the mission and dropped the cannons in favor of an all missile armament, the North American Design was dropped.  So, the scramble was on for a replacement radar that would fulfill the new mission requirements but still fit in the space carved out in the design for the NASARR. There were two immediate contenders. A modified Westinghouse AN/APQ-50 X-band fighter interceptor radar (similar to that flown in the McDonnell F2H-3 Big Banjo and Douglas F4D Skyray), and the Hughes AN/APG-51B (similar to that flown in the McDonnell F3H-2N Demon and Douglas F3D Skyknight).

Both were proven, although by no means state of the art designs, that used discrete components that were connected by long wiring harnesses. In an aircraft where real estate was at a premium, this was not ideal. Also, their design posed other problems. Long wiring runs connecting the components weakened the radar signals passing through them as well as made the introduction of noise into the system more probable.  And each time the signal strength was boosted, the amount of noise introduced into the system increased as well.

Westinghouse was the first to come up with a solution. The team at Westinghouse decided to combine the AN/APQ-50 modules into a cylinder shape that would fit into the aircraft nose right behind the radar antenna so that the signals did not have to travel through long wiring harnesses. McDonnell need only supply cooling air and electricity to a central inlet and Westinghouse distributed it. The entire radar was supported on an I-beam on which it could be extended out for easy maintenance. (This design feature was one of the improvements that helped the F4H-1 win the fly-off with the Vought F8U-3.)

But, during testing Westinghouse could not satisfy the most basic requirement of the radar and that was the range at which the APQ-50 detected a target aircraft. Westinghouse blamed the radar antenna. McDonnell's engineers had designed the F4H nose to hold the 24-inch diameter antenna of the NASARR radar they had intended to use. Westinghouse claimed that 24 inches was simply too small, as their calculations showed that the APQ-50 needed at least a 32-inch antenna to meet the required range. In testing they had confirmed the size at an outdoor radar range where they took a prototype of the 32-inch antenna and shot it at aircraft landing at Baltimore airport. The Navy approved the bigger antenna renaming the system the APQ-72, and gave Westinghouse a production contract.

By this time the prototype F4H aircraft (first 18 airframes) were in various stages of completion on the assembly line, all designed for a 24-inch antenna. The 32-inch antenna design changes wouldn't be implemented until the next series of aircraft (Block 3).

The 32-inch antenna would pose several engineering challenges. First came the fuselage redesign to widen the nose cylinder to allow for a 32-inch antenna. From FS 77.0 forward the fuselage was redesigned to deepen and widen the nose. (All future nose changes also took place from FS 77.0 forward like the RF-4, and F-4E)
Fuselage structure modifications forward of FS 77.0

Another technical challenge that they faced was building a radome for a high-speed aircraft that was 32-inches wide.  This had never been done before. Rain hitting a radome that large at Mach 2 could have some very serious consequences. The radome had to be structurally strong while being light-weight and aerodynamically correct. But most of all it needed to be transparent to the radar waves both outgoing and incoming. Radar manufacturers like Westinghouse worked with an exposed antenna in the lab and in the field trials, and then expected the radome manufacturers to provide a transparent dome that did not absorb or reflect the radar waves. Any imperfections in the radome, either in thickness or density, could distort the radar signal and give false information.

After extensive research into the state of radome technology, McDonnell chose the Brunswick Company in Virginia (maker of bowling balls and fiberglass boats) to construct the new radome. The Brunswick engineers came up with a way to wind fiberglass filaments around a conical form provided by McDonnell in the exact shape of the radome. The filaments are wound in alternate layers which are 90 degrees from each other. One layer of filaments, called circs, are wound around the form circumference and the other layer of filaments, called longos, are wound around the radome lengthwise. The layers of filaments are then saturated with resin, to seal it from absorbing moisture, and baked in an oven until it is cured into a strong yet flexible cone. The engineers at Brunswick then constructed a grinder connected to a special meter that measured the phase shift of an electromagnetic wave as it passed through a section of the radome. The grinder would grind the thick or dense areas of the radome until the whole radome registered identical readings. The bonded fiberglass shape is then covered in a layer of neoprene to act as a rain erosion barrier and a small metal nose cap is affixed to the radome to prevent rain and airflow from peeling back the neoprene shell. (I have seen a pin hole on the neoprene peel half the radome like a banana when it came back from a flight.)

(Note: the lone exception to this process was the RF-4s which had a painted radome (epoxy enamel finish) with a fabric/neoprene boot only covering the first 12-inches of the radome.)

F4H-1 RADOMES

Please understand that this isn't a definitive list. I am having a hard time finding pictures to validate each aircraft. I have included BuNos. of the ones that I have been able to verify and will keep it updated as I find more information. 

VARIATION 1 : All Metal 24-inch Nose with Metal Rib Structure (opens left)
Verified on BuNos: 142259a, 142260a, 143388a, 

This was an all metal nose with internal metal rib structure not intended to be compatible with radar, so these aircraft were not equipped with radar while equipped this radome. Many of these aircraft had an instrumentation pallet in the nose where the radar would have been.


Internal Structure of the 24-inch metal nose.



BuNo. 142259a with metal nose. Notice TAT probe installed later in test program.
BuNo. 143388a Showing to good effect the screw pattern on the metal nose.
BuNo. 143388a Showing screw pattern on metal nose.
BuNo. 143388a With radome open. Notice instrumentation pallet swings out to the right (just beyond the open radome).
VARIATION 2: Glass Fiber 24" Nose with Metal Rib Structure (opens left)
Verified on BuNos: 145307b,  



Again, aircraft fit with this radome would not have had radar installed because of the metal rib structure that underlies it.
BuNo. 145307b - The screw pattern can easily be seen showing the underlying structure.
VARIATION 3: Glass Fiber 24" Nose (opens upward)

Verified on BuNos: 145308b, 145315b, 145317b

For some reason three aircraft had a modified radome which opened upward instead of to the left like the others.  The radome was a new 24 inch glass fiber type, without internal structure other than the mounting ring so it would be compatible with a radar set. Externally I am not sure you would see any differences, just that internally it was hinged differently.



Structural Repair manual shows a glass fiber radome which opens upward.
VARIATION 4: Glass Fiber 24" Nose (opens left)
Verified on BuNos:  143389a, 143390a,

Aircraft equipped with this radome could have had AN/APQ-50 radars installed because of the lack of internal metal structure. 


BuNo. 143389a with 24-inch fiberglass nose.
VARIATION 5: Hybrid 32-inch Long Nose with Metal Rib Structure
Verified on BuNos: 143392a, 145311b 

These aircraft had a much longer radome which actually was built over a 24-inch radome structure. The nose appears to be much straighter on top as well as being more symmetrical in shape than the eventual production noses. This nose was used to test out the aerodynamic qualities of the proposed 32-inch radome. Since it had two layers of metal ribs (one on each radome) and a test instrument boom installed, these aircraft did not have a radar installed.
Illustration showing the outer structure built around the normal inner 24-inch structure. 



BuNo. 143392a showing its 32-inch nose

BuNo. 145311b showing its 32" nose
VARIATION 6: 32" Glass Fiber Nose
Verified on BuNos: 145313b, 146817c and subsequent aircraft.

This was to be the production standard for all future F-4(B,C,D,K,M,N,S) Phantoms starting with Block 3. The airframe forward of FS 77.0 was modified to accommodate the antenna of the APQ-72 radar and the larger radome it required. (FS 77.0 was the point where all future nose modifications started - RF-4s, F-4E,F, etc.)
The new 32" radome was laminated glass filaments bonded together and then covered with a neoprene exterior which was to keep rain from eroding the glass fibers. It had no internal metal structure other than the mounting ring and a small metal nose cap is affixed to the radome to prevent rain and airflow from peeling back the neoprene shell. The new radome opened to the right unlike its predecessors. This made a lot more sense as it wouldn't impede access to the cockpit when completely open.

32-inch production radome 






Revision History:


  •  18 Dec 2017 - Original Post


Sources:

  • Glenn E. Bugos, Engineering the F-4 Phantom II - Parts into Systems 
  • NAVWEPS 01-245FDA-3-1 - F-4A, F-4B & RF-4B-Structural Repair Manual 15 MAY 1965
  • Photos found on the Internet




Friday, November 10, 2017

Conformal Weapons Carriage and the F-4

In the 1960s and 70s the Douglas Aircraft Company, a part of McDonnell Douglas Corporation, maintained a program of airborne weapons research and development known as the Advanced Armament Technology Program. This programs focus was addressing problems with the high-speed delivery, separation, and impact of free-fall conventional weapons.  During the 60’s both the Navy and Air Force were conducting studies with their contractors into the use of bluff bombs.  A bluff bomb is merely a conventional warhead, less its tail section, turned backwards and then it is fit with a very simple star shaped tail casting (right on the warhead) and a blunt end plate. Not very streamlined, but it does pose some advantages as the studies bore out. First, the size of the munition was decreased allowing a much denser packing of the weapons. Second, the bluff weapons have a lot of airstream drag due to their shape, allowing the aircraft to escape bomb fragments even during low altitude delivery. Third, bluff bombs exhibited excellent separation qualities dropping through the turbulent boundary air surrounding the aircraft quickly and with little influence on the trajectory. Carried internally they presented a great opportunity. On an aircraft pylon their effect was akin to putting out the speed brakes.
Comparison of Conventional and Bluff Bomb
In November of 1967 Douglas performed a study for the US Navy in which they first established the aerodynamic gains associated with conformal carriage of weapons. Conformal carriage represented a new method of carrying stores close to the fuselage of an aircraft. Nine to twelve individual bomb racks are housed in an aerodynamically smooth fairing beneath the aircraft fuselage, thus considerably reducing the drag of the installation with or without stores when compared to conventional MER/TER carriage. This seemed like the logical way to carry bluff weapons externally.
Theory was put to extensive wind tunnel testing and these seemed to bear out the math that supersonic flight with stores (carried conformally vs. on pylons) was enhanced, that stores separation was improved because the relatively flat surface of the pallet provided smoother air flow over the weapons and the rigid structure permitted higher ejection forces. Because the separation was cleaner and more precise the ground impact should also be more predictable and accurate.  +Wind tunnel testing had shown that the best external weapons carriage arrangement was by grouping the weapons as closely as possible, with a minimal frontal area, and in a single layer close to the fuselage. The closeness of the weapons is only limited by the need to avoid contact with the weapons beside it. Once a minimal frontal area is established, the length of the pallet doesn’t add any significant drag, thus limiting the load only by weight and area. The advantages seemed to be considerable, so now the time had come to put away the slide rules and thinking caps and see if the theory carried into the real world.
The idea of a conformal pallet posed some engineering challenges, especially when adapting it to an aircraft that wasn’t designed for it from the start. Accessibility for maintenance and servicing would need to be maintained or alternative means would need to be provided since the conformal pallet would be a rather permanent (well, it wouldn’t come off easy) fixture.  By now the Navy had been joined by the Air Force in this program and they brought Boeing on board to design and build the conformal pallet. The F-4 Phantom II was selected as the guinea pig as it had a wide, relatively flat lower surface to mount the pallet, and a good load carrying capability. Both the Air Force and Navy felt that the F-4 also provided the best candidate for further development of the demonstration package into a retrofit for existing operational aircraft. The Navy made F-4B (Bu. No. 148371), the 56th F-4 built and highest-hour airframe in their inventory, available for the program.
The flight program started with data collecting “baseline” flights of the F-4B both with and without conventional weapons on multiple conventional pylon configurations. The data collected here was used to compare with future data collected with the conformal pallet installed. 148371 was then flown to Seattle so Boeing could begin modifying the aircraft and installing their pallet.
Boeing had been busy designing and building their prototype pallet. The Navy provided Boeing with a F-4 airframe hulk to help in the design of the pallet. Boeing took this airframe and made a female mold from the lower surface giving them the ability to then make a precise cast replica made of steel reinforced plastic. This replica was then used as a base for development of full-sized plaster master molds of the conformal pallet external components. Plastic molds were then made for either casting low-shrink concrete stretch form blocks or for use as fit gauges when hand forming bulkheads and fairing skins. This process allowed more work to proceed simultaneously and avoided the traditional lengthy process of lofting to establish mold lines and interface details.  All parts were handmade, some being “made to fit” during actual aircraft modification.
Modifications to 148371 itself were kept to a minimum for this program. Some alterations were necessary, but where additional changes were necessary for future maintenance and servicing, the problems were resolved on paper and left for any future implementation of the pallet. Changes that were made for this program included:
1.       Engine oil servicing points were moved to the MLG well.
2.       The Liquid Oxygen fill/vent valve was moved to the left forward missile well (converter stayed in place) – launcher was removed for these tests.
3.       Canopy air pressure gauges were moved to the right forward missile well – launcher was removed for these tests.
4.       The engine air start duct was moved to a point aft and outboard.
5.       The engine auxiliary air doors were replaced with a set of louvers on the exterior of the pallet operated by the door actuator.
6.       The centerline pylon electrical circuit was revised to carry signals from a new weapons management system devised for the conformal pallet.
There were other modifications that would have to be made if the conformal pallet became an operational reality, but these were enough to get things off the ground.
The dimensions of the finished pallet were listed as:
  • Height: 6 inches
  • Width: 96 inches
  • Length: 326 inches
There were 49 ejector mounting positions available in a seven row by seven column matrix. In this prototype version, three rows of up to four columns could be used at one time, limited by the rudimentary weapons management system circuit. The ejector mounting positions were repositioned by relocating the crossbeams in the pallet and then inserting the exterior filler panels as needed.  For bluff munitions there were a series of bolt-on fairings of varying heights to streamline the munitions. Conventional weapons which by design were more aerodynamic would not require any fairing. As installed on 148371, the pallet had what could best be called outrigger sections which extended below the engines and intakes, but there is nothing in the documentation I have, nor in the documentation for the proposed operational version which mentions these. It does not appear that these sections had any load carrying capability. 


Conformal Pallet loaded with NSRDC Bluff Bombs

Once the pallet was installed 148371 resumed flight testing, starting off with basic handling and performance testing with the conformal pallet installed. Over 200 weapons were released during the tests at speeds between Mach 0.6 and 1.6, and altitudes between 5,000ft to 30,000 ft. msl. Ripple and 30 degree dive releases were also investigated using Mk. 82, Rockeye II, NSRDC bluff, and M-117M6 bluff bombs. The overall results indicated that the F-4B with the pallet in place flew as well or better than the clean F-4B (no pylons installed). But of greatest significance was that with weapons installed the drag was reduced so much that “it is possible to fly the F-4 supersonically with external weapons nearly to the full extent of the flight envelope of the clean F-4.”

Loading aircraft with conventional Mk. 82 GP Bombs
(note absence of bolt on fairing used only for bluff bombs)
Overall the program was a success. The conformal weapons all showed superior separation and predictability when compared to their pylon mounted conventional cousins.  Even the Mk. 82, which was notorious for their poor separation behavior when ejected at subsonic speeds using conventional methods, successfully separated in over 100 tests, both at level flight and dives, and at velocities escalating into supersonic speeds. Weapons mounted on wing pylons decrease the aircraft’s longitudinal stability, but these tests showed that fuselage mounted stores do not affect longitudinal stability at all. Overall performance with fuselage mounted stores resulted in much the same handling and stall characteristics as a clean aircraft. In fact, in the report the conclusion was “Based on the foregoing qualitative analysis it can be at least anticipated that a conformal carriage system installed on an F-4 aircraft can in fact enhance that aircraft’s handling qualities to the extent of greatly improving the combat capability in the attack role. Range and speed performance have certainly been improved for an attack configuration both to and from the target. But also important is the agility of the aircraft particularly in situations requiring evasive maneuver and offensive action.”
Weapons tested during the flight phase of this program.
The program was a demonstration and test of what conformal stores could accomplish and as such it merely scratched the surface of the potential. Much needed to be done to make this a viable modification to the existing aircraft. Some of the questions that weren’t addressed in this program, but had to be considered before it could move forward were:
1.       Engineering, development and testing with weapon mixes.
2.       Engineering, development and testing with guided weapons.
3.       Engineering, development and testing with air-to-air weapons.
4.       Solutions to minor buffeting that was experienced by some weapons at high speeds.
5.       An improved ejector rack was needed for this configuration.
6.       Engineering, development and testing for improving weapon and ejector rack access during loading, arming wire hookup, ejector cartridge installation and removal, mechanical weapon release and emergency jettison of installed weapons.
7.       Engineering, development and testing of improved ground handling and loading methods.
The limitation of this program was that it applied to one single aircraft type. It could be argued that they had the ideal situation with the F-4 with its wide relatively flat lower fuselage.  Conformal carriage when applied to the F-4 aircraft resulted in an exceptionally effective weapon system for the combined air-to-ground / air-to-air role. And there is evidence that development of this concept for the F-4 didn’t stop here. In a Naval Weapons Center report titled “F-4B/J Aircraft Conformal Carriage Preliminary Design Study Report” dated January 1975, the concept is further developed to address the list of questions above and to include 30mm gun pods and larger weapons with 30” center mounting lugs. It also addresses most of the identified servicing / maintenance modifications that were needed to bring the conformal pallet into operational status.  The flow chart in this report had all design, testing, and development to be done, kits being delivered to the fleet, modifications performed, and the first units becoming operational by July of 1977.
More evidence of the viability and further development of this concept shows up in Boeing’s proposed “Super Phantom” which uses a conformal fairing to house additional fuel, avionics, and contains hard points for weapons.
One of my old drawings of the proposed Boeing "Enhanced" or "Super" Phantom
I don’t know why this concept wasn’t implemented. The data seems to indicate that it would have extended the F-4s capabilities and usefulness. We see the concept popping up here and again without much traction. The F-16XL was one such proposal. Probably the most used variation was the conformal fuel tanks on the F-15 Strike Eagle, but that really wasn’t a weapons carriage system. But with the move to stealthy aircraft carrying their weapons internally, conformal carriage wouldn’t have been around very long all the same. My guess is that with the war in Vietnam winding down, the perceived need was reduced and as a result the funds became less available. Also by the end of the 70s new aircraft were already slated to replace the aging fleet of Phantoms. The expense of converting the aircraft with a  limited life expectancy was deemed uneconomical.



Revision History:

  •  10 Nov 2017 - Original Post
  • 11 Nov 2017 - Added one of my old pictures of the Boeing Enhanced Phantom
  • 13 Nov 2017 - Added links to full drawings and information about pallet dimensions.

Sources:

  • Artwork by Kim Simmelink
  • Suspension Equipment Considerations by Robert L. Kyle, Douglas Aircraft Co.,
  • The Conformal Carriage Joint Service Development Program by James E. Nichols Jr., Naval Ship Research and Development Center, 
  • Conformal Carriage Flight Test Program by R.E. Smith, Weapons Development Dept., June 1973
  • F-4B/J Aircraft Conformal Carriage Preliminary Design Study Report by Edwin J. Zapel, The Boeing Aerospace Company, January 1975
  • Characteristics and Applications of Bluff Bombs, USAF Aircraft Compatibility Branch, Munitions Division, June 1975