Rickover and Rigged Research

Obviously, adverse data could have had serious consequences.

And, an effective way to avoid adverse data is to rig the research.  One set of rigged research is 
the Baker-Just equation which has its roots at Rickover's Bettis.  This is documented on pages 345 and 346 of:

Nuclear Power from Underseas to Outer Space

John Wistar Simpson

American Nuclear Society, 1995 - Technology & Engineering - 467 pages

John Simpson, former president of Westinghouse Power Systems Company and past president of the American Nuclear Society, provides a vibrant account of the events associated with the birth of the nuclear industry. Simpson's account of his career and the many turns it took is formidable. 

Following are excerpts from pages 345 and 346:

The question was raised whether ignition of zirconium, even if locally initiated, could spread in an uncontrollable manner to a major fraction of the zirconium contained in a core. The accident condition under which such a reaction was considered most likely to occur was a loss-of-coolant accident in which a coolant pipe would suffer a major break.

If it could be shown that Zircaloy-4 reacted with water in a predictable manner at all temperatures, up through its melting point, this should assuage fears of unpredictable and uncontrollable reactions.  These data were required in less than three weeks.  An experiment was devised by Bill Bostrom in which samples of zirconium tubing similar to those used in Shippingport were heated  inductively while immersed in water; the released volume of  hydrogen was collected and used to monitor the rate  of reaction. Barricades were set up around the water container and a system of mirrors was devised so engineers could control metal temperature during the reaction.  Another engineer 
 observed the change in hydrogen volume as it bubbled  up through the water column.  Temperatures were gradually 

increased in consecutive experiments until, as a grand finale, zirconium was melted under water and the amount of reaction monitored during melting. 

These experiments, completed during the required time  period, demonstrated that the occurrence of high temperature reaction rates could be extrapolated from those measured at lower temperatures, and that no new or unexpected phenomena intervened that would endanger reactor plant safety. Crude as these initial experiments were, the  kinetic data derived from them and the conclusions 
drawn have been supported by subsequent experiments
and analyses.  Obviously, adverse data could have had serious consequences.

http://pbadupws.nrc.gov/docs/ML1009/ML100950085.pdf

Please read page 1 of 8, ignore the rest.