Design Example: The SMICBM

When I returned to Aerospace in 1963, I was assigned to the Ballistic Missile Division located in San Bernardino, an entity separated from the main operations in El Segundo by about 80 miles. I was given the job of heading a department composed of several disciplines; I remember that two of them were reliability and reentry vehicles. I was on this assignment about three months, perhaps enough time to become familiar with the technologies associated with intercontinental ballistic missile systems, when I was asked to design a survivable intercontinental ballistic missile system in accordance with the MCD criteria.

I begged off from accepting the new assignment, explaining that the development of the MCD criteria had caused me no end of grief, and problems with top management if not with the entire aerospace industry. Several days later I was again asked to take on the assignment. This time the importance of this assignment was explained with the argument that "the Navy is walking away with Polaris." In addition, I was assured that I will be treated differently, that I would get a team of seasoned designers to work with me, and that I could confer with specialists throughout the company on an as need basis. It seemed that I had no choice but to accept the assignment. Besides, I sensed some inner voice telling me to go for the opportunity to apply the criteria, and I wondered what an MCD-designed system would look like since my analysis told me it would look differently.

I was given about three or four months to complete the assignment. Initially I examined R&D and operations (cradle-to-grave) costs of existing ICBM systems. I was much surprised to learn that the bulk of the system costs was in operations. This, coupled with the belief that a "bare-bones" solid rocket booster, unconstrained by weight and volume limitations, could be made very cheaply, led me to envision a "shell game" basing concept as most promising.

In this scenario, multiple ICBM housings, or "silos," would be built, and only one in many would contain a complete, higher cost missile. The high-cost missile component and other system elements that would comprise a functioning missile would be surreptitiously moved at intervals from one silo to another. Under perfect conditions, all silos would represent legitimate targets.

The following description of the system studied, the SMICBM, does not disclose details that do not contribute very much to the understanding of the application of the MCD criteria to the system design. Sensitive aspects such as why the system is survivable, the vital system command & control and the security measures proposed are not discussed.

Among the first tasks undertaken in formulating its design was an analysis of the viability of the basing concept. The analysis contained multiple parameters representing all the system elements. Assigning likely ranges of values to these parameters showed that the system was very viable. After the second or third progress report briefing was presented, an analytically inclined Air Force colonel confronted me with a recent, unpublished RAND Corp. report that showed that shell-game basing was much less cost-effective than other means of achieving survival, such as hardened silos, for instance. After some frantic effort, I showed that both analyses were identical, and that the RAND report reached its conclusions by only applying typical, minimum weight/maximum performance costs to the system elements, while I used much lower costs as a consequence of designing to the MCD criteria. This incident illustrated that expectations based on experience may be misleading when the MCD criteria are applied in designing a new system.

The ultimate design criterion was considered the maximization of the ratio of the destruction cost to the deployment cost of the system. Hence, the design objective was to minimize the denominator while accounting for the composition of the threat. This led to the conceptual design of the following system elements. As already indicated, all elements were optimized essentially in unison and subjected to tradeoffs, particularly design and cost interactions.

Silos, hardened to a low overpressure, had the configuration of partially buried Quonset huts. They were spaced about a mile apart so that the majority of single enemy warheads could destroy only one silo. Each silo contained a veritable "dumb" booster that incorporated only the ignition system for the solid rocket motors, and an air-conditioning system. The boosters had no size or weight constraints. Their only requirement was to propel a "payload" weight a given distance - with a comfortable margin - at minimum cost; hence, the rocket casings, propellant, nozzle design, and other components were at the discretion of the contractor. The booster might have had a simple suspension system to absorb ground motions, if one were found necessary. The top of the booster was approximately flush with the floor of the hut. The optimum spacing of the silos and their degree of hardening was determined by the system optimization analysis.

An upper stage, called the "integrated payload package" or IPP, composed essentially of a reentry package, vehicle guidance and thrust vector control systems, was installed only on a small fraction of boosters. A portable auxiliary ground equipment (AGE) unit that monitored the IPP and hut operating systems, and provided the command and control link, accompanied the IPP. The fraction of complete vehicles was determined by a cost analysis that found that about only one missile in ten should be "live". The missile got part of its name (Semi-Mobile) from the fact that only the IPP and AGE were moved from silo to silo on a random basis, probably not too often, and after they were sometimes processed through a depot-type facility. Only commercial-type, small-size trucks were used. Few depot facilities were necessary. The computer-based operation of the system and the efficiency of the depot-type facilities promoted low manpower requirements and low-cost operations.

The system was very well received by Air Force and Aerospace management, and study contracts were issued to industry. Hughes Aircraft was awarded the system study contract. Boeing studied the design of the IPP. Several propulsion contractors studied the booster design, and these studies included pressure-fed, liquid stages. This was the first time industry was exposed to the MCD criteria. I did not get a chance to witness their reactions.

These contracts were hardly underway when I was asked to return to the El Segundo facility and work on the preliminary design of the Titan III B. My job was turned over to a very hard-working individual who was assured that I would be available for consultation and visitations with him as necessary. In addition to telephone conversations, I did meet with him about every ten days. After about two months, my supervisor informed me that I would be "wasting my time" continuing to work on the SMICBM. I recognized that the project was doomed and would be canceled.

Many months later I received a urgent call from my replacement. He visited me and told me with great distress that the general officer in charge of the Ballistic Systems Division called him into his office and told him that he is canceling the project because it would not employ very many Air Force personnel that will be available to him. I confessed that I knew this would happen and assured him that it wasn't any fault of his.

About the time the SMICBM was first proposed to the Air Force, and for several years thereafter, there was a rash of contractor proposals on various shell-game systems and other survivable concepts. Every concept used a complete ICBM, of course designed to the minimum weight/maximum performance criteria. They were transported about on huge, special-purpose, multi-wheeled vehicles, sometimes on a continuous basis, sometimes between special garages where they were parked for a random period of time, or run around specially designed racetracks. Or the missiles were encapsulated and moved submerged within a gridwork of canals. The system that received the most study had missiles moving about on the existing railroad system. Some time along the line it became known as the MX Missile project. The final basing mode placed the missiles in hardened silos. It seems that after many years and many billions of dollars spent, the RAND Corp. survivable basing analysis prevailed. Thus deception-basing concepts were relegated to the junk heap.

Occasionally and for many years later, I would get together with several members of my SMICBM design team to discuss the various survivable basing concepts as money was lavished on industry to study them. We concluded that all concepts would have cost many times more than the SMICBM and would have taken much longer to R&D and install. Those concepts that used the shell game had to assume the ratio of live-to-dummy missiles since no back-up analysis for determining this ratio was ever made. We concurred that the most labor-intensive concept was railroad basing, and facetiously wondered why this was not accepted as the final system.
 

Are any Minimum Cost Design principles in use on today's ballistic missiles? [no discussions were submitted for this question]

Next Column: Design examples continued

Return to Published Columns