HD1963MMAPU

Stage I Stage II

Stage III

Introduction

The Minuteman Missile System I, II, and III was cutting edge technology. Funding that went into missile programs between 1950 and 1958 provided the ingredients that made it possible to put man on the Moon, and the technologies used for Minuteman were even more advanced. Much credit has been given to the Inertial Navigation System and the first application of Digital Computers using then new Integrated Circuits. Less has been said about the new solid propellant motors and thrust vector control with electronically control hydraulic servo systems. These controls were significant contributors to the Minuteman missile series great success.

For those skilled in controls systems, the Minuteman Nozzle Controls Units (NCU) were almost a work of art – from their Servo-Actuators to their hydraulic Auxiliary Power Units (APU) described here.

Minuteman I had three solid propellant booster stages each equipped with four tilting nozzles which were positioned by NCU’s to achieve the desired Pitch, Yaw and Roll thrust vector control. Each NCU had four hydraulic servo-actuators, one for each tilting nozzle, powered by a single APU with each positioning servo actuator controlled by it’s own dedicated semiconductor electronics. The NCU’s were in the form of an X with servo-actuators at the end of each arm and the APU plus electronics located at the center. The unit was mounted at the end of each arm to each nozzle base, and covered with an ablative rubber material to protect it from the over 2000 degree gasses expended from the nozzles. Hydraulic fluid and wiring passed through drilled holes in the NCU structure. The system was powered by a battery which was activated by the release of electrolyte – there were no battery power switches. The NCU was commanded by wires encased in an external raceway extending from the Guidance and Control section at the top to the NCU at the bottom. As one stage burned out, it was separated and the next stage took over. The missile preferred to fly back end first, as the center of pressure was forward of the center of gravity. This required very snappy attitude control to keep the nose pointed within two degrees of the desired heading. Side winds when lifting from a Silo launch and unbalances during stage I separation, which occurred while aerodynamic pressure on the body was still significant, placed high demands on the servo controls.

The APUs delivered 3000 psi on stage I and II and 1500 psi on stage III. The servo-actuators were powerful enough to shove a garage car hoist backward. The “small” electric motor on stage I could produce enough torque to deliver 7.5 HP of hydraulic power. The small compact size of these units were deceptive, observers often were not aware of their power and rapid response to commands.

The need for small envelope, limited weight, instantaneous response under severe environments were demanding and especially when topped of with exceptional reliability requirements. The systems would be required to rest in silos for many years then perform without malfunction within one Minute. Launch procedures could be validated but the vehicle itself could not be operated, like a fire cracker, it either worked or didn’t – it had to be reliable.

Prior to and during the early Minuteman I worked in our hydraulic test labs which were dedicated to evaluation and development of hydraulic flight control elements under extreme conditions. I worked closely with Art Greer and Lou Purpura who were developing the concept of how we would power the hydraulic servos. They coordinated the design concept for the NCUs, and came up with the packaging concept for the APU’s. They wrote the specifications for them and selected Vickers to be the sub contractor. At this time Ling-Temco-Vaught was selected to be the servo-actuator sub contractor for three kinds of servo-actuators also built to our specification.

Shortly after pre-prototype units arrive for initial testing, Art and Lou were made Project Engineers and I was move from the Test Lab to be the Lead Engineer for the development of the hydraulic control systems. Thus I became intimately involved in bringing to reality the design concept they evolved.

These units would prove to be enormously successful: The packaging and design was a work of art and the subsequent method of R&D testing, malfunction analysis and immediate design correction, resulted in an exceptionally reliable unit. Previous Auxiliary Power Units were a sub system of independently developed parts. Minuteman, faced with unique problems required radical innovation while using proven technologies.

To reduce envelope and save weight, functions were merged -- for example the pump was placed inside the fluid reservoir, eliminating the need for a pump case. The reservoir was a bellows because it was required to hold or release extra fluid when the “unbalanced” servo actuators were in the extend vs retract position. In the above view the left dome is the end of the reservoir and moved in or out as the bellows, inside the larger collar, expanded or contracted. The bellows also served as a spring to apply pump inlet pressure. Inadequate pump inlet pressure can cause “cavitation”, a phenomenon which can erode metal at the pump inlet.

Stage I, II, III pumps were the same size, the angle, motor and reservoir size were changed

The valve plate was much thicker than shown and auto controlled to meet demand

Conventional hydraulic input/output fittings were omitted and replaced with a dual passage port that plug-in to the main NCU structure, with it’s drilled passageways that served as hydraulic lines. The servo actuators had a similar plug in – thus no hydraulic lines or fittings were used.

The only exposed external seal was that on the hydraulic pump to electric motor shaft. This seal was to hold up very well, but for a while there was serious debate between engineers and with quality control as to what constituted a leak. The shaft would become “wet” with oil during operation, and after repeated tests, could develop a drip – that had not dropped. It was resolved that a “wetted” shaft was not a leak, and if a drip didn’t drop it was not a leak. Over and over I calculated to be sure there was a reserve in the reservoir to account for some drips. The units had to survive for many years in a silo – they could not be inspected unless removed and stripped down.

When Art and Lou went to the project office three young engineers: Clarence Asche, Ron Frazini and George Leonard, were assigned to help me -- they were an excellent choice. We had not been party to what had gone on, so one of the first things we did was gather all correspondence and calculations others had left behind, and placed them in order by date. We then went through them from start to present. We could tell that many persons had been involved so we decided to make our own design calculations for sizing the actuators, pump and battery. We discovered that actuator demands were based on pure Pitch or Yaw, they did not take into account a combination of both pitch and yaw when moving 45 degrees to a main axis which represents worst case demand. We also found that they did not take into account servo valve leakage. The servo valve is a “hydraulic amplifier”, it converts constant flow “leakage” to move a pilot valve in response to a 0 to 8 ma command. For this application, it was a significant omission. For example the pumps are turned on 1 minute before launch and during this first minute battery power is drained to provide leakage flow, independent commanded demands. Then while state I is controlling flight stage II and III are draining the battery through leakage flow. Then again stage III leaks while stage II is controlling. The problem is compounded because leakage flow is very expensive, being produced at very poor hydraulic efficiency -- a little bit of high pressure leakage flow is high costs demand on the battery.

Once we cross verified our findings I went to find Ray Curci, the head Project Engineer. He was busy in a meeting so I handed his secretary a note which said:

“There is a good chance the pending (first) launch will fail.”

“The actuators, the APU’s and Batteries are undersized.”

I asked her to hand the message to him, that I’d be at my desk.

A few minutes later Ray left his meeting and was at my desk with note in hand saying, “Just what the hell is this!!”

Ray, a very competent engineer, reviewed and understood our calculations. He said be ready to implement a redesign as soon as I can get word on holding up the first launch. He knew this required complete redesign of actuators, pump motors and the battery. It came at a time of much political pressure to get off a first launch. In 90 days we were able to have the first launch, the units had been completely redesigned, delivered, tested and install in the vehicle – a remarkable achievement.

Two days later I went with Vickers to their motor supplier to describe the change and the nature of the demand duty cycle. I learned that the armature commutator segments were held in place by wrappings of “piano wire” – the armature rotated at 30,000 rpm and became progressively hotter. Armature heating was a real concern, there was no way to carry heat away. I was quite impressed with the capability of this specialty motor manufacturer. Many highly skilled specialty businesses were located in southern CA. The position transducers for example were being built in a home converted to manufacturing by Gary Collins an ex North American Engineer – his transducers were being used on every missile and aircraft system in the free world including the Minuteman.

We at Autonetics did the initial R&D and qualification testing for Minuteman I. Vickers performed those tasks on later versions. We were causing units to fail under severe test conditions. The ability to define environments to be expected and to apply those conditions via a test lab had only recently become an option. Test equipment could now apply gausian random vibration inputs. These and other realistic environmental tests put the hardware through grueling conditions. Up to 14 g’s accelleration through atmospheric conditions and severe shocks when blasting near by structures apart were parts of the environment. When a unit failed it was sent back to Vickers for failure analysis. At our periodic meetings they would present their findings and submit their design modification fix for approval. I kept a personal log of these failures and fixes. Once revealed and fixed, that failure mode never occurred again. The same thing applies to the servo actuators – we knew with confidence we were fielding a quality product.

Though I was dealing with the components, people doing the system came to me for ideas on how to fill the system with hydraulic oil. When it came time to add fluid the lines were full of air and there was only one “fill” port. I knew the oil must be free of entrained air as well as removing air from the system. I had the fellows get a vacuum pump from the test lab I had operated and used it to suck air out of the system. We also set up a system to acoustically agitate the new fluid while evacuating it’s chamber, to remove air entraped in the fluid. We then switched from vacuum pump to new fluid reservoir and let atmospheric pressure push the new fluid into the NCU cavities. It worked very well.

It was about this time when I was home on vacation that I stopped by the Ford Garage, operated by my Uncle, and visited with his main mechanic Howard Saum with whom I’d worked pending return to school after service. Howard had just overhauled a hoist on a Ford Tractor and was having problems getting the hoist to work. Standing in back of the tractor I asked where is the pump? He said in the transmission case. I asked does it use transmission oil? He said yes. I asked do you have an air hose handy? He said right above your head. I reached up for the air hose, cupped my hand over the fill port to the tractor case and squirted air into the case to pressurize the pump inlet – I figured it had a bubble of air in it’s inlet. As if by magic the lift operated. I said nothing as I let loose of the air hose and enjoyed the look that came over Howard’s. Though he said nothing his expression was saying, now how in the hell did he know to do that? I’d learned much from Howard and it was a pleasure to return the favor. I had become very aware of their being air in fluid when Bob McCoy one of our engineers ran tests on a actuator made of plexiglass. You could see bubbles form and absorb as pressure was applied and released. It was a remarkable demonstration on what goes on out of sight.

Sec A-A

The square box on top of the motor is a “Radio Noise Filter”. DC motors with their brushes making and breaking contact as the armature commutator segments whiz by causes noise spikes on the power cable which can radiate as noise into other parts of the system – it is essential to reduce this noise to acceptable levels. We were to find that it was essential to ground equipment and maintain attention to signal grounds to avert problems due to electronic noise.

I was particularly fascinated by how Vickers designed their pumps. The pump of each stage used the same piston block, but cocked at a different angle to adjust for flow capacity. The pumps were set to be constant pressure at variable demand. The pump had a servo actuator with controlled the valve plate input output port – the pump powered itself to meet demand only, providing the amount of flow necessary to maintain a designated pressure.

The piston barrel was a bit like that of a Colt 45 pistol, with the pistons being about the size of a 22 rifle bullet. But what caught my eye were the fixed angle “Kingsbury” pads by which the rotary member “skied” on hydraulic fluid – the metal parts never came in contact yet carried very high thrust loads. Shortly after I hired in, my boss Paris Stafford and I were “told” by our boss George Keller to take a class being given at nights at UCLA on Lubrication and Bearings. Two of those lectures had been on hydrodynamic lubrication. One application being the automotive drive shaft bearings and another application being Kingsbury bearings. Kingsbury from England, came up with the idea of how to make a shaft end bearing. He cut a washer like plate into pie like segments and ran a pivot through the center of each segment, which extended radialy from the center. When immersed in oil the plates would rotate as if skies supporting a rotating shaft on a dynamic layer of fluid. This design was applied to support the end of large generators at Boulder Damn – even sustaining operations when during an accident the oil was flooded out by water. The bearing held up the heavy armature while it skied on a film of water.

On these pumps the Kingsbury bearing was in the form of fixed “tilted” pads, unfortunately there is no photo showing these.