top of page

CLEMENTINE

It was 1992, Satellites cost at least $250 million dollars and took at least 5 years to develop. ‘Faster, Cheaper, Better’ was becoming a new buzzword, a heretical slap in the face to the old-space mantra of “Fast, Cheap, Good, pick any two.”  A small team was  challenged to build a spacecraft that would go from a blank sheet of paper to launch with only 20% of a traditional spacecraft budget and schedule: 22 months and $55 Million to map the moon. A sharp, 29 year old spacecraft engineer, Bill Purdy was selected to lead the mechanisms team. There were only two unbreakable rules – launch on time, make it work.  The engineering and production had to be done very quickly – yet be done right; it had to work.  

 

The project was a chaotic ride from start to launch to on-orbit operations.  Risks were taken, lots of risks.  Problems were encountered and wrestled to the ground, lots of problems.  Sleep was traded for making the launch schedule. The “way it has always done” was mercilessly scrutinized, and rules that did not add value were discarded.  Rigorous testing replaced redundancy, decisions were made by smart individuals rather than committees, and there was no requirement that designs be perfect.  Rules of good engineering, good testing and personal responsibility were honored.  

 

And ...  it worked.  

This was the first map of the full moon, including the dark side, and this map was done in 5 different wavelengths.  And, to everyone’s surprise, we discovered that ice is trapped in deep craters in the South Polar region of the moon, enough water to support a human colony on the moon’s south pole.  

Bill brings the lessons learned from the journey of “it has to be inexpensive, it has to be done fast, and it absolutely has to work” to the groundbreaking projects of new space.

Windsat

WINDSAT

The WindSat payload on the Coriolis mission was a low cost, single-string technological innovation, replacing the high costs of reliability based on redundant design with single-string reliability based on intelligent design and an emphasis on thorough testing.    

At the time, space-borne microwave radiometers could determine the magnitude of the wind over the ocean but could not determine the direction.  An instrument which had never been attempted was needed, with absolute state of the art sensitivity since the wind direction signal was so faint as to be nothing but noise to previous space-borne wind velocity measurements.  Key elements of the instrument included development of ultra sensitive receivers, an extremely precise, 1.8 meter 22 channel antenna and a spin mechanism containing over 100 slip rings that would spin at 30 rpm continuously for its life.  

As a result of Clementine’s and other successes, Bill was selected to head the spacecraft mechanical design. WindSat took advantage of Bill’s diverse skill set as he led the mechanical design, served as antenna systems engineer and led the optical alignment of the sensor.  Bill applied his mechanisms experience as technical advisor to the lead engineer for the spin mechanism.  Post launch, Bill’s role evolved into providing mechanical and systems insight for the on orbit calibration and validation effort.  Bill has worked this project from cradle to grave (Not Yet!)  starting on this project the day it was initiated in 1997 and is presently supports flight operations monitoring the system performance.

The Windsat instrument was designed to meet a 1 year life requirement, but remains fully functional at 14 years life.  The addition of a wind direction measurement has proven a valuable aid in the forecasting of hurricane track and intensity, and provides many other benefits to maritime forecasting. In fact, you can see today’s Wind direction map at https://www.nrl.navy.mil/WindSat/data/wind/AscendingWindMap.php

Bill’s experiences as systems engineer on WindSat are a rich source of lessons learned necessary to successfully implement the innovative, single string satellite designs crucial to realizing the potential of new space.

Looking Closer

LOOKING CLOSER

Early in Bill’s career he was responsible for the pyrotechnic  separation devices used for mission critical launch restraints for a major space system, and simply put, if just one of these mechanisms failed in flight, the mission would be a total loss, with the spacecraft failing to release from the launch vehicle.

A key part of the separation system supplier’s testing was to operate each mechanism with pressurized Nitrogen gas to verify performance margins, as a test with the actual pyro devices was not possible.  The mechanism release pressure was used to indicate performance; lower release pressure meant better performance.  When Bill reviewed the data for the flight mechanisms, it showed all units had passed by meeting the requirement of release at less than 800 psi.  On his own initiative and intuition, Bill decided to look more closely at the statistics of the data to identify patterns that might not be immediately obvious. His review showed an average release pressure of 450 psi with a standard deviation of 30 psi.  He also noticed that one separation device had released at only 235 psi, which suggested amazingly good performance but strangely, performance outside of the statistically expected range.  Bill asked that this unit get retested because it was outside the expected range, and almost too good to be true.  Reexamination confirmed it was too good to be true.  Surprisingly, the retest showed a release pressure of 1135, out of spec and indicating unacceptable performance.  

 

The closer look identified a potentially catastrophic coincidence: (i) the separation nut had been mis-assembled and consistently performed poorly, and (ii) during the first test of that unit the pressure had been measured erroneously;  a much lower test load than specified had been applied, leading to a too-good-to-be true release pressure.  The ensuing investigation found root cause for the failure, found the error in the test procedure, and the test was repeated, rejecting the flawed release device but exonerating the rest of the lot.  

Subsequently, all the release devices operated successfully in flight. If Bill hadn’t asked to have anomalous test repeated, the misassembled unit would have flown on the spacecraft.  Would it have failed in flight leading to a mission failure and multi-hundred million dollar loss?  Possibly.

Fortunately we will never know.

bottom of page