Small anele Z&ray scattering measurements were used to detcrmfne the size and shale 01 asphxltene and preasphakene mIcelIes in soIution. The fraction of the coal-derived ltqufda forming mIcoUes in solution was as high as 49%. The fraction depends upon the concentration and molecular size of (he asphattenes and pceasphaltenes, the 1-and/or hydrogen bonding properties of ttae solvent and the agitationhfost systems studied were palydispersed with lhs majority of the micelles bcingspherical with a diameter of 22-38 A_ Mhelles in the 80--100 A range were also apparent but at 6--12% of the number of thti smaller size. Variattanr in the acatbrlng intensity over the scattering angles and the varialians of the absolute intensity oPscattered X-rays from saIutions irradiated Immediately after ultrasonic agitation and upon extenrkd sitting, suggested the presence of a low tevel Rot in most of the solutions. Coal-derived liq;lids are complex mixtures of liquid and semi-liquid canstituents. They can be broadly classified as oils, resins, asphalknes and preasphaknes. Oils and resins are nominally soluble in pentane, asphaltenes are insoluble irl pentane but soluble in benzene or toluene and preasphaftenes are insoluble in totuene but soluble in t4rahydrofur.m (THF) cw pyridine. The mobcular wd;lht, number of condensed aromatic rings, the number of nitrogen, oxy@n and sulfur hrterotltoms and the concentration of metals increases in going from the average molecule in the oil fraction to the average molecufe in the preasphaltene &a&ion [I] a Most of the oxygen in the asphaltenc and preasphaltene fractions is phenolic oxygen and most of the nitrogen is basic ring nitrogen. The physical state of coakkived liquids depends significantly on the reIative amounts of each fraction. Anomalies sre observed when coal liqciefaction slurries containing high levels of asphaltenes and preasphaknes are Ekered in the 2OU-320°C temperaturu range [2). The anoma.

No abstract

No abstract

1 0 k -Fig. 1tions can be obtained, however, by staying "near" the optimum procedure in particular cases.For example, consider the circuit of Fig. 1, whose voltage transfer ratio may be synthesized by a refinement of the method suggested by Bobrow and Hakimi.,where A ; = A 2 -1. Finite amplifier input conductance Gi and output resistance R, can be included as part of the appropriate Y blocks. If F(s) = N(s)/D(s)is an arbitrary real rational function, then the realization is made by chosing a polynomial Q(s) with n distinct negative real zeros where n is one less than the maximum of the orders of N and D . Then F(s) becomes and N / Q and D/Q can be expanded by partial fractions to form positive and negative admittances. These RC admittances can be identified with appropriate terms in (1). ( Y , and Y3 are realized first and subtracted from D(s) to realize Y , and Y,.) The Horowitz optimum decomposition can be used for N or D, but not both, since the same Q must be used for both. For many functions an optimum choice can be made for D to reduce pole sensitivity (this is possible only if N is of no greater order than D ) ; or in some functions where N and D are of the same order, a compromise choice for Q can be found. The optimum choice for D cannot be exactly realized here since D in (1) contains Y 3 and Y4 from N .To make the example more speciffc, consider a normalized three-pole low-pass Chebyshev transfer function with one decibel ripple. N(s)0.491307 _ -D(s) -(s + 0.494171)[(~ + 0.247085)' + (0.965999)'] (3) The optimum Q(s) for the conjugate pole pair is (s+0.997104). The real axis zero in D cannot be exactly realized in the optimum pattern;' but a good compromise choice is to pick the zero in Q(s) to nearly cancel the real zero of D(s), say (s+ 0.5). For this example, the optimum sensitivity of the denominator to changes in A;, IS?;l, equals 3.03 at s=j. The compromise choice permits realization of N(s) and D(s) and yields a sensitivity of the overall transfer function to A;, ISAil, of 6.67, with A , = A , = 10. The realized sensitivities SAj and S;; are equal. If, as Bobrow and Hakimi Q(s) is arbitrarily chosen and is of the same order as D(s), Q(s) = (S + 0.3)(s + 0.7)(s + 0.8), and A , = 5, A , = 10, the realized sensitivity is ISA;I= 75.6.We have constructed both the arbitrary Q circuit, and the nearoptimum Q circuit using simple transistor amplifiers. The improvement in sensitivity is evident in the noticeably easier practical alignment of the latter circuit.The effects of space charge on the operation of transverse-wave amplifiers are studied using a simple optical electron-beam model suggested by H e m a n n [I 1. Some of the interesting results are summarized in this letter.In the analysis, a quadrifilar helix pump structure is used for the dc pump and the phase constant B, of the pump structure is set to satisfy the requirements of the desired modes of coupling.The spacecharge effect manifests itself in the equations of motion as a radial field term which, in the presence of a dc quadrupole field, r...