- Joined
- 27 December 2005
- Messages
- 17,713
- Reaction score
- 26,192
[links removed as broken - see posts later in topic - Admin]
Last edited:
DEVELOPMENT OF THE REDHEAD/ROADRUNNER by Erwin Ulbrich, Jr.
During World War II, the introduction of radar and solid rockets allowed the development of the fi rst generation of antiaircraft missiles. Typical of these were the Nike missiles positioned around Los Angeles at the end of the War.
In the 1950s, a new generation of jet fighters and bombers were developed, and it was obvious that a better anti-aircraft missile was then required and a typical one in the U.S. was the MIM-23 Hawk missile. It was to engage, both subsonic to supersonic threats, from ground level to 65,000 feet altitude. Missile crews for the Hawk were trained at Ft. Bliss in El Paso, Texas, at the McGregor Range, and at nearby White Sands Missile Range (WSMR) in New Mexico which could be used for fl ight testing missiles. Collectively, the three bases provided a very large expansion for the live fire of rockets and missiles for the U.S. Army.
The Columbus Division of North American Aviation (NAA) won a contract in 1959 to build a target drone for the Hawk missile crews that had to be economical and reusable. When it was operating near the ground, it was to be capable of flight 300 feet above the ground at 0.9 Mach and it was then called the Roadrunner. When it operating in the stratosphere, it was capable of flight at Mach 2+ up to 65,000 feet altitude and it was then called the Redhead. Both versions were to have the same analog flight control system with a switch for setting various gains and constants. Requirement The missile was mounted on a Little John launcher with the booster rocket developed from the artillery rocket’s booster. It was capable of a “Zero” launch where it had to become operational directly from the launch.
There were only two aerodynamic surfaces; elevons controlling both pitch and roll and indirectly yaw. They had no authority at rest and one to two seconds after the booster was fired, the airspeed was sufficient for them to start controlling. That meant that for those two seconds, the booster thrust had to be closely aligned to the center of gravity (CG) of the missile/booster because the booster generated 5,000 pounds of thrust and the missile only weighed 880 pounds. If there was a misalignment, there would be a very large rotational torque in pitch or roll or both which could not be controlled.
After the two seconds, the flight control would start to balance the two elements. At about four seconds, the amount of air and fuel going through the ramjet would ignite, providing thrust. At about fi ve seconds, the booster would burn out and aerodynamic drag would strip it off the missile. The CG will then move and the elevons, using full authority, would rebalance the forces. Normally, a vertical transient of about 5 g was encountered.
Design
The missile can be seen to have a radome on both ends. Inside each is a Luneberg Lens which provides a strong radar return to the missile battery radar. In the forward radome is a pitot system which provides barometric static pressure and total pressure and also launch point static pressure. This system was designed by Rosemount and featured a compensated static port on the boom not the fuselage as in most aircraft. The aft radome would be blown off the missile, releasing a braking rocket and a parachute for recovery when desired or when 7 Fall 2013 the ramjet fuel was gone.
Behind the radome in the front was the avionics bay which held the autopilot and the command radio receiver, the vertical gyro and the rate gyros. A short whip antenna protruded out the top of the bay, and it picked up the guidance commands which came from a transmitter in the Command and Control Center where the missile pilot was located. In front of the rear radome was a bay containing the missile battery and the accumulator/pump/actuator package. It could be set for various gains and damping in the auto pilot.
The two bays were connected by a wiring tunnel. This is the black strip shown in the picture above which was taken about two seconds after launch. The black smoke is from the booster. The white smoke is from an igniter in the engine that has not yet started the engine fuel burning. The elevons are about to pitch the missile up. On the pitot boom is an angle of attack sensor which was only used for fl ght test telemetry.
When the engine ignited and the booster was gone, the missile flew more or less at a small pitch angle. It can be seen that the missile was statically unstable, that is, there was more mass (the engine) above than below. This meant that the flight control system had to function for it to fly or it would turn over.
During flight test, it was determined that the missile made outside turns unlike a conventional airplane. This did not affect its use as a target but it would be uncomfortable for a passenger. The vertical gyro design was borrowed from the V-1 missile although it was made by the Iron Fireman Company in the U.S. It was air-erected and it had stops on its motion. If it hit a stop during its operation, the gyro would tumble and it would crash the missile. This was used by Spitfire pilots during the war to crash the V-1s. It did not affect the missile’s use as a target.
The autopilot controlled the missile in pitch, roll, and altitude. The missile speed was set by the engine fuel controls which were in the engine pylon. The engine at rest looked like a titanium garbage can and it is hard to imagine it operating like a supersonic blowtorch. It did supply significant sonic vibration to the missile in flight. The ramjet was supplied by Marquardt in Pomona and was designated as MA-74.
At the Command and Control Center, the missile was tracked by a radar receiver which produced the ground track of the missile in flight. Hooked up to this was a 3D model of the terrain which could produce a differential altitude signal for the range above or below the launch point. This could be sent to the missile to profile its flight over the terrain. The autopilot could also be set for hold pitch or hold altitude or both remotely from the Command and Control Center. There was a very tight specification for this profiling of 300 feet altitude plus or minus 60 feet.
A particular piece of terrain known as Hummingbird Gap between two mountains was given as an early fl ight test objective. It would be impressive to see the missile come roaring out of this gap. All systems on the missile were analog in nature. The 28 volt battery voltage was regulated down to 24 volts DC using a constant voltage circuit. The four operational amplifi ers in the autopilot were driven by double-ended 24 volts AC from a 400 Hz inverter. Their functions included: the Starboard Elevon Valve Driver, the Port Elevon Valve Driver, the Pitch Summing Amplifier and the Roll Summing Amplifier. Shown is the breadboard for the autopilot. Individual transistors, which were invented about five years earlier, were used throughout. The actual autopilot was developed a few months later. This was entirely a Columbus Division product. Two principal criteria were low cost and sturdiness because of the engine vibration. These were met using a deep drawn aluminum box that cost $12. It was modified with beryllium coated spring clips to hold the four operational amplifiers. This box and contents were thoroughly tested over a wide frequency range in all axes. All internal electrical joints were hard-wired. Life testing was not extensive as the total life of the autopilot was thought to be a few hours. There were only two connectors on the autopilot; one for the flight control system and one for test which was also used in flight for connection to the radio for Command and Control. The system required a system tester and this was also a Columbus Division product where the first article was designed and produced in three days. A careful inspection will show that the toggle switches are upside down. I took a lot of ribbing over that.
The autopilot also contained a pitch programmer which provided a climb out signal for the missile when it was fired. This was a durable spring wound timer which was built by a vendor in Riverside, Illinois. When I went to inspect their factory, I was surprised to see that the employees worked in the dark over a bowling alley. The management told me that they were specially trained people who were hearing and vision impaired. The timers always worked fine. A hallmark of this effort was the use of an analog simulation of the missile flight including as much hardware as possible. One instance was the actuators. As soon as the actuators were done, they were integrated into a simulation rig. Another was the vertical gyro which fl ew in a large three axis Carco table. At the end of the design effort, every autopilot that was made was flown on the analog simulation to assure records of what flight test should be like and then it was put in a sealed box (by Quality Control) and taken from Columbus to El Paso in a shipping container which was painted day-glow red so that the courier could keep track of it in baggage handling.
Flight Test Fight test at WSMR started before any hardware was ready and consisted of using the booster which was manufactured by Rocketdyne in Texas to loft wooden poles which had similar inertial characteristics to what the missile would have. At the end of three firings, very little data was acquired and all of the poles crashed into the ground. It did show the important role that the flight control system would play during launch. Both engineering and flight test did the testing and since there was no place for civilians to stay at WSMR, the Columbus staff stayed in El Paso, and early each day drove 90 miles on a private road at 90 mph with the headlights on to WSMR. This was reversed in the evening. Several cars were worn out during this test program. When the first sets of hardware were ready, it was time for the actual missile flights. The first three of these were a repeat of the earlier tests and at the end of three flights there were only about 10 seconds of data.
The crashes were powerful and devastating. The Customer and NAA Management realized that something drastic had to be done. The obvious problem was booster misalignment and the solution was given to both Flight Test and to Engineering. I was given the job of “Keeper of the CG” and working with the Weights Group, we came up with our best estimate of the CG location. This work was hindered by WSMR safety regulations which required that the missile without fuel and the booster be measured separately and then the fuel weight had to be added analytically to the measured date to provide an estimate. The big question was how many air pockets were in the fuel tank behind anti-slosh baffles, etc. We at least could control the amount of fuel put in. These safety provisions were reasonable if we should ever drop the missile and it went off. The other group was the Flight Test engineers and they made the most important invention to help solve the problem. They built a conical gage that fit into the booster nozzle with a long rod aligned with the thrust vector. They then used land surveyor’s tools from a distance to adjust the interface between the missile and the booster to align the thrust vector with our calculated CG. As it came out, the missile symmetry put the CG more or less on a vertical plane in the center of the missile. The nozzle could be adjusted to be in this plane also. That still left two dimensions to be accounted for. When everything was done, there was a successful flight.
When a turn was initiated we discovered the missile’s tendencies to make outside turns. I did not witness this since there were only a few people allowed in the Command and Control Center during flights. The telemetry was clear however. This required a rework of the equations in the flight simulation so that they agreed and from then on there was a close correlation between the two. The simulation of the first 5 seconds of flight was accurate from the start which helped a lot. We also could use the simulation then to find the limits of CG location that we could live with. These limits were in the small fractions of an inch as I remember and different for lateral and longitudinal errors. Missile Flight Program When the engineering was done, I was given other assignments associated with lunar landing simulation which became important at Columbus since we were some years ahead in that area.
The missile, after about a year of testing, went to McGregor Air Force Base for actual training use. In June of 1963, it was designated as the MQM-42, Redhead/Roadrunner. It was used for about 15 years into the 1970s. The missile, on paper, had a 250 mile range which seemed to be a lot but at Mach 2, it was only about a 10 minute flight. At Mach 0.9, it is longer but the friction is greater at low altitude. This meant that you would get only one engagement per launch. In use, the actual altitude ceiling was given as 59,000 feet. There were other drone competitors but none could beat the low cost of the ramjet versus a jet engine or a rocket engine. Subsequently, many other missiles used ramjets. I never got any data on how much reuse the missile could provide. It was obviously a problem to parachute it into a rock pile with the wind blowing, but other than that, many of the interchangeable parts would be durable enough for rework and reuse. I never got to see the missile fly. In retrospect, this was one of the most fun jobs I ever had. Almost every day, you could go out and at least on the computer, you could shoot rockets.
About the Author: Erwin Ulbrich is a retired flight control engineer who worked for seven aircraft companies during a 45-year career, culminating in the design of the B-2A avionics system involving the integration of over 200 computers. In the 1990s, he taught at UCLA Extension on the integration of all of the modern avionic safety systems into the cockpit to help provide the great increase in aircraft safety that we now experience. He and his charming wife, Myrna, have lived in Whittier, California for 50 years where he changed the thrust vector to his employers but not his residence.