Finding the Way (Again): Building the Air Force’s New Century Series
We’ve got to kill the major defense acquisition program as it is today, and replace it with something that looks like the Century Series development of the Early Air Force.
Dr. William Roper, the Air Force’s senior acquisition official, has set a new goal for the development of combat aircraft: Create an Air Force acquisition process that can design a new fighter every four years, and maintain that high development tempo for the next generation of programs. His proposal explicitly hearkens back to the Century Series aircraft, built for the Air Force during a flurry of modernization in the 1950s. All told, six fighter/interceptor designs had their first flight between 1953 and 1956 (three more designs remained unflown), resulting in 5531 aircraft delivered to the Air Force alone.
Roper’s is a worthy goal, designed to break the Air Force out of a paradigm in which it takes entire careers to field a single fighter aircraft, often late, over budget, and unable to meet initial requirements. It’s been decades since an Air Force fighter reached (or exceeded) a planned production run — the last was the 2231 F-16 deliveries between 1978 and 2005. The subsequent programs, F-15E and F-22, delivered far fewer aircraft than anticipated, and the total F-35 buy has yet to be determined. But while the Century Series looks like a great model, the conditions under which those aircraft were designed and purchased was vastly different from the conditions facing Air Force acquisition today. To achieve Roper’s goal, the Air Force would have to make major changes to the way acquisitions work, from the requirements process to the decision-making that enables it — all areas where the service continues to face challenges. The New Century series will require a lot more than a better design phase.
Figure 1: An F-105 Hunter-killer team in Vietnam. In the foreground is the two-seat F-100F Wild Weasel variant of the Thunderchief, with a F-105D single-seater in the background. The Wild Weasel mission focused on the destruction of enemy radars and was often flown in mixed teams (U.S. Air Force photo)
The Century Series
The Century Series aircraft were not part of a unified program. Instead, they were products of a flurry of proposals — some unsolicited by the government — to deliver aircraft in a way that incorporated lessons learned during the Korean War. The acquisition environment of the 1950s was vastly different — services could buy aircraft as needed without a long, painful oversight process that demanded that every requirement be studied and “validated,” and every capability justified. The philosophy at the time remained little changed from that of World War II, and incentivized corporations to conduct independent design and prototyping without having to meet a stack of government-imposed specifications. The process rewarded rapid production, often with hundreds of changes following the initial design. Even the name came afterwards — the aircraft series was later dubbed the Century Series because the aircraft numbers followed a set sequence from F-100 to F-108.
There is an elephant in the room that must be addressed early. The Century Series aircraft weren’t very good. The North American F-100 Super Sabre saw the Air Force order over 270 aircraft based on a company mockup. Despite impressive aerodynamic performance, the F-100A was so dangerous that it was taken out of front-line service after seven years and 47 losses — only to be returned to service in the Berlin Crisis. The later F-100D was a better aircraft, although over 500 were lost in accidents, overshadowing 198 combat losses. Arguably, the F-100 was the best of the bunch and the only “real” fighter in the lot — the subsequent aircraft were straight-line, high-speed fighter-bombers and interceptors.
The McDonnel F-101 Voodoo was a bomber escort turned mediocre nuclear strikefighter that actually provided credible combat service as the RF-101C, a high-speed reconnaissance aircraft. The Convair F-102 Delta Dagger was an aerodynamic disaster that never met its performance requirements and was undistinguished in Vietnam. The Republic F-103 ran into engine and airframe design problems and was never built. The storied Lockheed F-104A Starfighter could do one thing well — zoom to altitude in (literally) record time, which was exactly what the Air Force requirement dictated and not what the Air Force actually wanted. In Vietnam, its air to air record was 0 to 1; Tactical Air Command cancelled more aircraft than it accepted. The F-104 later found wide use as a (heavily redesigned) export fighter.
The series never really got better. The Republic F-105 Thunderchief, referred to by the “Thud” it made hitting the ground, was another straight-line nuclear strikefighter that could haul a lot of bombs a long way but had a nasty habit of exploding after taking battle damage. Despite valiant efforts by its aircrew and maintainers, around half of the 833 F-105s produced were destroyed in accidents or in combat. The Convair F-106 Delta Dart was a redesigned F-102 that still holds the world airspeed record for a single engine combat aircraft, but was the last interceptor built for the Air Force. It never flew in combat and was not exported. The North American F-107 was declined without a competition, in favor of the F-105, and the F-108 Rapier never made it past the mockup stage. The F-109 designator was never assigned.
The Century Series may not have been the best jet aircraft the United States ever produced, but they weren’t the worst either. Despite real shortcomings, the initiative produced large numbers of aircraft quickly, and pushed the state of the art and sometimes beyond. But it did so in a freewheeling development, test, and acquisition environment that is vastly different from what the Air Force faces today. While risks were managed, the design and fielding of aircraft that were at the leading edge of technical knowledge entailed substantial risk — and substantial reward. Today, the service will face challenges related to policy, the law, and the industrial base if it attempts to re-create the pace of the Century Series. Perhaps most importantly, the Air Force also faces challenges with culture and leadership.
Figure 2: Only three North American XF-107 prototypes were built. This aircraft is in the Air Force Museum Collection in Dayton, Ohio. (U.S. Air Force photo)
Back in the Day
The Century Series was built at a time when the Air Force could order aircraft built on a mockup and the fighter/interceptor force was so large that failure or delay of a model was barely notable Indeed, some aircraft programs were kept in play precisely because others were failing. There was a degree of program redundancy that ensured that the failure of an aircraft program was a crisis only for that airplane’s manufacturer — any maybe not even then. The sales of the Convair F-106 and the McDonnell F-101B were a direct result of the delays and inadequacies of the Convair F-102. For every aircraft type ordered, there were three or four other designs submitted by other manufacturers.
In 1952, when the F-100 first flew, there were no fewer than 13 aircraft manufacturers building fighters, attack aircraft, or interceptors for the U.S. military. Between 1950 and 1960, Boeing, Convair, Douglas, Fairchild, General Dynamics, Grumman, Lockheed, Martin, McDonnell, North American, Northrop, Republic, and Vought all built fixed-wing combat aircraft. For Republic, North American and Convair, their Century aircraft were the last fighters they ever built.
The design process for the aircraft was fast — the prototype F-100 flew 27 months after the proposal was rolled out. The F-102 took 33 months from proposal to first flight, and the F-104 a mere 17 months. After being introduced, initial models rarely spent much time in front-line service before being superseded by improved models. They were short-life aircraft not by design, but because they weren’t worth keeping. The F-100A lasted seven years, the F-105B less than five, and the F-104A barely a year before being transferred to the Air Guard. The contracting authorities matched the design times — the Air Force was able to let contracts quickly because the acquisition process not only permitted it, but encouraged it — a marked contrast from today’s deliberate, cumbersome, and ponderous processes.
The services ordered aircraft in the hundreds without a long, convoluted process that relied on “validated” requirements. The Air Force put out requests for proposals like Christmas cards — with solicitations for fighters or interceptors issued in 1946 (Penetration Fighter), followed by the advanced supersonic interceptor (1949), 1954 interceptor (1950), bomber escort (1951), F-100 replacement (1953), and long-range interceptor (1955). The F-100, F-104, F-105, and F-107 were unsolicited, company-funded proposals untethered to any advance requirement from the Air Force.
Figure 3: Century Series Timelines
The Century Series were not the only jet fighter/attack aircraft introduced in the 50s. The Air Force also introduced the Republic F-84F Thunderstreak, although it was pretty much obsolescent when purchased, and missile-armed variant of the F-89H Scorpion. In the 1950s, the Navy fielded front-line jets just as quickly: Grumman’s F-9 Cougar and F-11 Tiger, the Douglas F-4D Skyray and A-4 Skyhawk, the McDonnell F-3H Demon, the Vought F-8 Crusader, the North American A-5 Vigilante, and three way-late-to-need variants of the F-86 Sabre, the North American FJ-2/FJ-3/ FJ-4 Fury series. Buried in the later years (1958) was the cream of the crop, the McDonnell F-4 Phantom II, arguably the best jet fighter built in the western world.
There is another key difference between the Century Series and today’s aircraft: software. While these aircraft all had electronic systems, they were not the software-driven systems in today’s aircraft. The first transistor radio just barely beat the F-100 into service, and modern, language-driven software hadn’t been invented yet. FORTRAN entered use in 1954, followed by FLOW-MATIC in 1958 and BASIC in 1964. The Apollo Guidance Computer was the first aviation-rated integrated circuit computer — introduced in 1966, after the last of the Century Series rolled off the line. Additionally, the Century aircraft did not require the years of testing and development that the Defense Departments requires today. This would have been impossible — the state of the art was advancing so rapidly that any specifications would have been obsolete before they were implemented. Instead, designers and developers were encouraged to use their best judgement — an approach that is explicitly rejected today.
For the Air Force to restore a multi-pronged aircraft development effort similar to that of the 1950s, the Air Force will require flexible acquisition authorities from Congress and buy-in from industry partners. The service itself will have to generate intelligent requirements, invest in development and test infrastructure, re-imagine the requirements and acquisition workforce, and shift its culture toward rapid-decision making rather than failure avoidance. The problems are nested — we need to treat the tangle as an interconnected series of problems and not just pull on individual threads.
Changing an Ineffective System
Today’s acquisition process is handicapped by design — the laws dictating what the services can and cannot buy are not designed to produce value for money, empower the services, or facilitate rapid acquisition. Grouped under the “Joint Capabilities Integration and Development System” (JCIDS), the status quo is a process-driven straitjacket, overly focused on creating an equal playing field and spreading around federal funds to as many congressional districts as possible. Industry is effectively prevented from producing unsolicited proposals, since companies know they would have to compete for a contract for any successful proposal — even if the government has to invent competition where none exists. Textron Aviation built the Model 530 Scorpion as a multirole reconnaissance/attack aircraft on its own dime, fielding a unique design without a modern counterpart. Air Force interest was limited to inviting the aircraft to participate in Phase I of the Light Attack Experiment — on Textron’s dime. Under such conditions, there is zero incentive for industry to develop suitable products without government money up front.
Figure 4: Defense Acquisition University’s Latest Acquisition Chart. A high-resolution copy can be downloaded here.
Today’s industrial base is a shadow of its former glory. Only Boeing and Lockheed Martin build fighter aircraft for the U.S. military, with Northrop Grumman relegated to building bombers. Textron and Sierra Nevada have attempted to enter the market with turboprop light attack aircraft, thus far unsuccessfully. Most of the aerospace companies of 1952 are gone, absorbed, or dissolved.
Nevertheless, there is still a capability to design fast and well. Boeing partnered with Saab for the Air Force’s advanced trainer program (T-X) — the pair won the competition with a new-design that rolled out in less than four years from project start to first flight — and flew before the final request for proposal was issued. The aforementioned Scorpion was constructed in secret over 17 months in a plant that hadn’t built a combat aircraft since 1975. There may be less industrial capacity than there was in 1952, but it is still there. If the Defense Department and Congress can eliminate the innovation-suppressing elements of JCIDS, it may become useful again.
Figure 5: The Textron Model 530 Scorpion in the ramp at Holloman AFB during the 2017 Light Attack Experiment (Author)
Open Mission Systems
The longest part of the process of developing an aircraft is not hashing out its planned lifespan (design life). Designing an aircraft that lasts 3000 hours (the F-100A) is little different from an aircraft intended to last more than 18,000 hours (T-6A). What drives the long fielding timelines are the systems integration challenges, particularly in software. The T-X rolled out of the production line ready to fly with no need for integration of sensors and weapons because it doesn’t have either. If the T-X were converted for a combat role (AT-X or FT-X), the systems integration phase could take a decade or more under the old methods. Software can have a long timeline. In 2014, the F-22 software was so exasperating that the program office carved up their bulky software development project into short, rapidly executable chunks. Called the Scaled Agile Framework (SAFe), the process was later adopted for the F-35. But the fundamental problem remains for most other aircraft.
To address this problem and shorten aircraft development timelines, Roper has proposed the use of open-mission systems. Open-mission systems would be derived from a common software baseline, separate from the flight controls and thus not requiring recertification every time a line of code is changed. A transition to an open system would also allow the avionics developed for one aircraft to be migrated to the next, not unlike the RADICAL proposal from 2017. Indeed, this approach was used for some of the Century Series aircraft, as part of the 1954 interceptor project. This project was split into two elements: Project MX-1554 for the aircraft, and MX-1179 for a common set of avionics and weapons. Hughes Aircraft won the avionics contract over 50 competitors in 1950, before any aircraft were selected. The Hughes systems were installed on the F-89H Scorpion, the F-101B Voodoo, the F-102A, and its successor the F-106. Derivatives of the system were installed on foreign fighters as well. The Air Force Research Lab already has an open mission system standard; its use should be mandated.
Requirements and Acquisition
A new Century Series will require a shorter, less cumbersome requirements process. The F-35’s requirements document was issued in 2000, and has had to undergo multiple revisions since then — a process now approaching its 20th year. With no fighter programs since then, there are few airmen who are capable of writing an achievable requirements document, as it requires a deep knowledge of the state of aviation technology. Over the intervening two decades, the acquisition workforce was slashed, with many capabilities outsourced to contractors. The long partnership between those who fly aircraft and those who build them has dissolved, and re-establishing it will be easier said than done.
Neither acquisition nor aviation is a pick-up game — it takes years to learn the job and become effective. In the past, aviators and acquirers would combine forces to develop and acquire aircraft. Aviators flew the aircraft as they were acquired, which assured that the people buying the aircraft had a deep understanding of its intended use. To recreate that, the Air Force would have to adjust its 3-year assignment cycle, where uniformed personnel pick up and move every three years. Three years into a multiyear project is no time to dissolve the team.
The selection process for competition winners will have to change as well. Today’s process is designed to be “fair” to all possible industry competitors, using a definition of fairness that ensures that nobody who knows anything about the competing aircraft or their vendors will have influence over the selection. That could easily result in the selection of an aircraft with a panel that has no aviators. Selection boards should always include individuals who have been involved in the process, from requirements writing to capabilities analysis and research to test and evaluation.
Experimentation and Support
If the Air Force is going to move out on a series of rapidly developed aircraft, tied together by an open missions system core, it will need an ongoing development effort that ties together avionics and systems development alongside development of the airframes. That will require demonstrator aircraft and a software integration lab, along with an experimental venue that enables a process whereby aircraft are continuously flown and adapted based on experimental discoveries. I conducted two quick studies for aircraft to fill such a role, including ex-Navy TA-4J pulled from the boneyard and an off-the-shelf new design. I found it is entirely possible to obtain suitable aircraft almost immediately. Either method would allow experimentation with Open Mission Systems, sensors, communications, and even weapons, independent of any single program but supporting many. So far, those proposals have floundered because there is no constituency for experimental aircraft today — in stark contrast to the thinking that helped make the Century Series a reality. But if we’re going to share mission systems, we need the aircraft, coders, and software integration capability in-house, for the long term.
Conclusion: Trying Again
No list of the changes needed to make a new Century Series happen would be complete without mentioning the necessary cultural change. The Air Force believes that it embraces innovation; my experience reminds me that it absolutely does not. Innovation requires departing from the pack and taking risk — behaviors that are not rewarded when promotion time comes around and not valued when it’s time to build a budget. The Century Series efforts combined a high tolerance for risk with a constant flow of money that paid for designs, mockups, prototypes, and straight-into-production aircraft. Companies submitted countless unsolicited proposals, because if the Air Force didn’t buy into one proposal, it would examine the next. The experience gained ensured that the next generation of aircraft would be even better, as exemplified by the F-4 Phantom II.
Century aircraft were fielded rapidly and imperfectly, but a healthy aircraft industrial base ensured that aircraft could and would be redesigned rapidly, sometimes even before the model being redesigned ever flew. Most importantly, the culture of the aircraft industry encouraged risk-taking and tolerated failure. The aircraft it produced were flawed by modern standards, but they were cutting-edge at the time, constantly pushing the boundaries of what was possible to achieve. That foundation can be rebuilt if it is important enough to today’s Air Force; indeed, it’s a good time to invest in the necessary changes. But make no mistake — changes will be necessary if the Air Force is to reach the goal of nothing less than a complete rebirth of its ability to rapidly field combat aircraft.
Figure 6: Postscript. The Air Force borrowed two Navy F-4Bs for evaluation, repainting them and giving them the short-lived designation of F-110A. Pictured is one of those aircraft, painted in Tactical Air Command livery but still sporting its Navy bureau number on the tail and the Navy FJ-4U designation under the right engine. The F-4 is still in service internationally more than 50 years after its maiden flight. (U.S. Air Force photo)
Col. Mike “Starbaby” Pietrucha was an instructor electronic warfare officer in the F-4G Wild Weasel and the F-15E Strike Eagle, amassing 156 combat missions over 10 combat deployments. He is the last American aviator to reach 1000 hours operationally in the mighty Phantom II. As an irregular warfare operations officer, Colonel Pietrucha has two additional combat deployments in the company of U.S. Army infantry, combat engineer, and military police units in Iraq and Afghanistan. He is currently assigned to Air Combat Command. The views expressed are those of the authors and do not necessarily reflect the official policy or position of the Department of the Air Force or the U.S. Government.
Top Image: A U.S. Air Force North American F-100D Super Sabre aircraft (tail # 56-3415) fires a salvo of 2.75-inch rockets against an enemy position in South Vietnam in 1967. This aircraft was lost with its pilot, 1st Lt. Clive Jeffs, after an engine failure near Nha Trang on Mar. 12, 1971. (U.S. Air Force photo)