2 40 

 220 

 2.00 

 1.80 

 1.60 

 1.40 

 1.20 



.00 



±r^ 



V 



Foa 



-IT- -- 10 



■INTERMITTENT JET 

 EJECTOR TESTS 

 (LOCKWOOD) 



STEADY-FLOW EJECTOR TESTS 

 (MORRISON) 



= 10 



I 2345678 

 AUGMENTER LENGTH TO DIAMETER RATIO 



Fig. 2 - Comparison of static perform- 

 ance of steady-flow and pulsating -flow 

 ejectors (from Ref . 2) 



where p, p, and T are the static pressure, density, and temperature, respec- 

 tively, s is the specific entropy, u the local velocity, 7 the force per unit vol- 

 ume due to surface viscous stresses, and t the time. These equations show 

 that a reversible transfer of mechanical energy within a flow system is possible 

 only in regions where the local derivatives 'bp/'bt are not zero, i.e., in regions 

 of nonsteady flow. 



From this it must also be concluded that a transfer of mechanical energy 

 from one flow to another can be nondissipative only if both flows are nonsteady. 

 This conclusion is supported by the observation that the only known steady-flow 

 mechanism of direct exchange of mechanical energy between two flows is that of 

 the conventional ejector, where the exchange is entirely effected through irre- 

 versible transport processes, whereas considerably higher efficiencies can be 

 achieved through nonsteady flow induction. 



Of special interest, among the methods of nonsteady flow induction, is a 

 method that does not require that the interacting flows be nonsteady in all 

 frames of reference. As demonstrated in Ref. 4, a nonsteady process that ad- 

 mits a frame of reference with respect to which it is steady over certain re- 

 gions in space and intervals of time will be said to be cryptosteady over these 

 space and time domains. 



A flow which is steady and isoenergetic in a frame of reference f is 

 neither steady nor isoenergetic in any other frame of reference F unless it is 



1354 



