The regression equation for the three-factor test of temperature, humidity, 

 and vibration becomes now 



P = 1.1117 + 0.0013594T - 0.0014441H + 0.063866V - 0.000045701 T"" 



+ 0.000015696H2 - 0.0025789^2 - 0.000020144TH - 0.00039512HV 



- 0.00047497VT 

 where 



P = output power in milliwatts 



T = temperature environmental factor 



H = humidity environmental factor 



V = vibration environmental factor 

 TH = temperature-humidity-interaction factor 

 HV = humidity-vibration-interaction factor 

 VT = vibration-temperature-interaction factor 



The coefficient of coiTelation that was computed for the regression equa- 

 tion had a value of 0.906. This high value of correlation is indicative of the 

 closeness of fit between the regression equation and the true-response equation. 

 This was substantiated by several estimations of the output power using the re- 

 gression equation. The estimations are shown in table 8. 



From the above regression equation, one observes that synergistic effects 

 do occur and have a detrimental effect on electronic equipment operating in com- 

 bined environments. It is these additional interacting effects acting in conjunc- 

 tion with the main effects that influence the failure rate and performance of ship- 

 board electronic equipment. 



The contribution of each environmental factor may be obtained from the 

 regression equation. If a unit change is assumed for each environmental factor, 

 then, the magnitudes of the coefficients of the various factors represent the de- 

 grees of contribution. After the scrutiny of the interacting terms, one notices 

 that the humidity-vibration and the vibration-temperature interactions contributed 

 about equally to the degradation in output power of the if/ am amplifier module. 

 However, the temperature-humidity interaction contributed approximately one- 

 tenth less to the degradation in output power than the other two interactions. 



During a particular combined-environment test, the significance or insig- 

 nificance of the different main and interaction terms will depend to a large degree 

 on the type of electronic equipment being tested. The susceptibility of electronic 

 equipment to the different environments and their interactions will definitely vary 



15 



