>. P 0 0

cu r SIXTH RAREFIED GAS DYNAMICS 0 G= 2 % .u 2

SEPARATION IN FREE J@PS

NITROGEN AMD HELlTTN MMTURES

OF

by

D. I. Sebacher, R. W. Guy, and L. P. Lee

NASA Langley Research Center Langley Station, Hampton, Va.

Presented at the Sixth International Symposium on Rarefied Gas Dynamics

P li

(ACCESSION YUMBER) (THRU)

Cambridge, Mass. July 22-26, 1968

https://ntrs.nasa.gov/search.jsp?R=19680024825 2020-07-30T20:10:09+00:00Z

SIXTH RAREFIED GAS DYNAMICS

DlFRTSIVE SEPARATION I N FREE JmS OF NITROGEN AND €ELW MMTURES

D. I. Sebacher, R. W. Guy, and L. P. Lee NASA Langley Research Center, Haimpton, Va.

ABSTRACT

Measurements of re la t ive species concentration, j e t structure, and N2 rotational temperature using an electron beamarepresented f o r f ree jets of nitrogen-helium mixtures. The results a re compared with the predictions of the method of' characteristics for the temperature distribution and with Sherman's nearly continuum theory f o r diffusive separation.

INTRODUCTION

The effects of diffusive separation in free expansions of binary gas mixtures from a sonic or i f ice into a low- pressure region has received considerable attention i n recent years. ration using sampling probes and skimmers,l-5 and Rothe6 has studied a free j e t expansion of ar on and helium using

theoretical continuum model for diffusive separation of gas mixtures i n f ree j e t s and a comparison is offered betwekn th is theory and the data presented i n th i s paper. tions of the je t structure and temperature distribution using the method of characteristicsg'fl are a l so used as an approximate model with an effective specific heat r a t io ( y ) assumed constant throughout the f l o w f ie ld . The gases selected were nitrogen and helium because of their excellent emission characteristics when excited by an electron beam, the i r large difference i n molecular weight, the i r different specific heat ratios, and because rotational temperature of the heavier particles could be measured. The techniques using the electron beam i n mixtures of N2 and He have been described i n a previous paper1* where the exhaust from a supersonic nozzle was investigated. The uncertanties reported i n that a r t i c l e led t o the present study.

A number of investigators have detected sepa-

an electron beam. Zigan and ShermanTt 8 have developed a

Predic-

L-6062

FIED GAS DYNAMICS

RESULTS AND DISWSSION

The method used i n t h i s experiment consisted of survey- ing the flow f i e l d of the free jets of a gaseous mixture of nEe/nN2 = 2 (n = number density) w i t h an electron beam t o measure the species concentration variations and, therefore, the extent of diffusive separation. perature distributions down the J e t center l i n e were also measured along with the j e t physical dimensions using a flow visualization technique. and a convergent sonic nozzle were used t o vary the Reynolds number and or i f ice diameter. Figure 1 shows 8 sketch of the main features of the apparatus and of a f r ee j e t flow f i e l d issuing from a sonic or i f ice . experiment had an 11' convergent portion and a 3.17-mm- diameter throat t ha t was 10 mm i n length. or i f ices consisted of circular holes dr i l led i n a 1.53-mm- thick plate.

The N2 rotationaltem-

A ser ies of square-edged or i f ices

The nozzle used i n th i s

The square-edged

Species concentrations e - Measured rat ios of helium atoms t o nitrogen molecules along t he j e t center l i n e as a function of x/d f o r various or i f ices and the convergent sonic nozzle are shown i n figure 2. thus systematically varied by both changes i n o r i f i ce diame- te r and stagnation pressure. These measurements were made by comparing the electron beam excited 7016 a He l i n e with the P branch of the 4278 a band of as described i n reference 12.

Reynolds number w a s

The maximum values of nitrogen enrichment observed fo r the convergent nozzle and each of the square-edged or i f ices a re shown i n figure 3 as a function of inverse Reynolds number d o n g w i t h the maximum separation predicted by Sherman's theory. min imum values of n&/nN2 shown i n figure 2 and the theoretical curve i s based on an x/d of i n f in i ty since th i s gives the maximum predicted separation. the theory, the w a s assumed t o be a constant (1.545) throughout, and the parameter of E time C from reference 8 was evaluated at 2.24.

The experimental points were taken a t the

In applying y

The j e t produced by the 3.17-m-diameter convergent nozzle was studied i n some detail so tha t l ines of equal concentration r a t io could be contoured and the results are shmn i n figure 4. The first conclusion t o be drawn from

2

SIXTH RAREFIED GAS DYNAMICS

these data i s tha t the separation of gas species i n a free j e t of an N2 - H e mixture does not completely agree w i t h Sherman's theory nor with the experimental resul ts of Rothe for a mixture of monatomic gases. This is readily seen from the results shown i n figures 2, 3 , and 4. The degree of separation appears t o be a function of nozzle geometry as w e l l as of Reynolds number. In all the jets measured, the heavier gas i s enriched f irst at a short distance from the ex i t and alocg the center l ine, then t h i s trend i s reversed and the helium concentration i s found t o be greater i n a l l other parts of the flow field. The lack of agreement between the data and the near continuum theory is not only i l lus t ra ted by the fac t that the separation is generally greater than predicted (see fig. 3 ) but that it first reaches a maximum and then decreases w i t h a decrease i n Reynolds number.

The contours shown i n figure 4 would seem t o violate the principle of species conservation but one must consider two factors which could account f o r this distribution. F i r s t , the density gradients i n a free j e t are such tha t a large number of the gas par t ic les are concentrated i n a small volume near t h e jet exit where the N2 enrichment is a maximum, and the much larger volume where He enrichment is found i s a region of re la t ively small par t ic le concentra- tion. The helium-enriched internal shock of the j e t has a relat ively greater density than the region both inside and outside the j e t a t values of that normal diffusion would tend t o increase the helium concentration both inside and outside the j e t boundary, as seen i n figure 4.

x/d greater than one-half, so

Second, some separation process other than that pro- posed by Sherman m y influence free je ts of N2 - He mixtures. Since the j e t passes rapidly from a collision-dominated region near the or i f ice t o a v i r tua l ly collision-free region farther downstream, it i s possible tha t the difference in the l ight and heavy pa r t i c l e velocit ies along the stream- l ines i n t h i s vicini ty could give rise t o a relative enrich- ment of nitrogen i n the t ransi t ion region.

Rotational temperatures.- For insight into the probable species parameters along the center l i n e of the je t , N2 rotational temperatures were measured using both pure N2 and

3

SIXTH RAREFIED GAS DYNAMICS

the mixture of N2 .. He. figure 5 and tend t o verify the assumption that i n the mix- tures the different par t ic les have W f e r e n t temperatures and therefore different velocities as they pass into the' nearly collision-free region of the j e t . This mnclusion i s drawn from the f ac t that the nitrogen rotational tem- peratures (Q) for the mixture a re conslstently above the expected distribution f o r continuum flow a t x/d9s greater -than one-half (x is distance downstream from the or i f ice and the d is geometric orifice or throat diameter). The predicted temperature distribution was calculated by the method of characteristics f o r x/dls greater than 1 f o r the 7 ' s of the pure gases and f o r the mixture (7 = 1.545). t h e lower values of x/d the temperature distributions (dashed l ines) were based on the measured values of TR the pure N2 j e t , and the sonic conditions computed at x/d = 0 f o r the Y ' s shown i n figure 5 . This procedure was followed due t o the uncertainty of the method of charac- t e r i s t i c s at l o w values of x/a ( ref . u). These measure- ments are somewhat limited as a method of analysis, since the ro ta t iona l energy begins freezing at an x/d above 2 ( for t h i s ured values from the predicted distribution. Despite th i s limitation and the d i f f icu l t ies i n determining TR at low temperatur~s,13,14,15,16 figure 5 shows tha t at low values of x/d the TR measurements effectively indicate the M2 translational temperature.

Some typical results are shown i n

A t

i n

Pod) as indicated by the veering-off of the meas-

Jet boundaries.- The results of the electron-beam flow visualization experiment are shown i n figure 6 where the measured ra t io of the maximum radius of the j e t internal shock t o the exi t radius i s plotted against pressure r a t i o f o r pure N2, pure He, and f o r the mixture. Results fo r the mixture a re shown fo r two different Reynolds numbers (based on nozzle throat conditions). These measurements are com- pared with the predicted distributions using the method of characteristics f o r various y ' s based on the measured con- centrations. the method of characteristics solutions cannot be used t o predict accurately the f ree j e t boundaries for a gaseous m i x t u r e whose constituents vary i n y when diffusive sepa- ration takes place. t o be i n agreement with the predictions based on the the maximum measured N2 concentration ( 7 = 1.528) but the higher Reynolds number data show poor agreement w i t h t h i s

The conclusion &awn from th i s figure is that

The lower Reynolds number data appear y fo r

4

SIXTH RAREFIED GAS DYNAMICS

7, t h a t of t he i n i t i a l composition (7 = 1.545) and that of t he maximum measured helium concentration (7 = 1.588). In m y case, there is l i t t l e Jus t i f ica t ion f o r select ing a constant Y if gradients i n 7 exist both axially and rad ia l ly i n t h e j e t due t o variations i n t h e concentration.

1.

2.

3.

4,

5.

6. 7. 8. 9.

10.

11.

12. 13 * 14 15 16.

E. W. Becker and K. B ie r , Z. Naturforsch, ga, 975

P. C. Waterman and S. A. Stern, J. Chem. Pkyys. 31,

V. H. R e i s a n d J . B. Fenn, J. Chern. Phys. 39, 32b.O

R. R. Chow, U. of C a l i f . Research Rept. HE-170-175

N. Abuaf, J. B. Anderson, R. P. Andres, J. B. Fenn, and D. R. Miller, Rarefied Gas Dynamics, V. 2, p. 1317, Academic Press (1967).

D. E. Rothe, Phys. Fluids 9, 1643 (1966). F. Zigan, Z. Naturforsch, l7a, 9 (1962). F. S. Sherman, Phys. Fluids, 8, 773 (1965). E. S. Love, C. E. Grigsby, L. P. Lee, and

M. J. Woodling, NASA TR R-6 (1959). E. H. Andrews, A. R. Vick, and C. B. Craidon, NASA TN

D-2650 (1965). H. Ashkenas and F. S. Sherman, i n Rarefied Gas Dynamics,

V. 2, 84, Academic Pmsm (1966). D. I. Sebacher, AIAA Journal, V. 6, 51 (1968). P. V. Marrone, Phys. Fluids, V. 10, 521 (1967). F. Robben and IJ. Talbot, Phys. Fluids, V. 9, 644 (1966). H. Ashkenas, Phys. Fluids, V. 10, 2509 (1967). I?. Tirumalesa, AIAA Journal, V. 6, 767 (1968).

(1954) *

405 (1959).

(1963)

(1959)

5

v) 7 W J o - z € 0 CY I:

dl w M s a I 0 c3 7 X -

s a W M

7 0 CY I- o W J W

a d G

a d al

m II E E 00 o\

II II II II ao no cllo ao

W

id N 0 z E

E E

-.

m -a .-

h

m

W

L

Di 0 E

E E

-.

0 -

m -a .-

h

m

'm

m Iz E E

m I E E

0, II E E

tn I E E tn

II

I I 0 a

4

X L? II cv

d d

D

0 f s 0 c, 0 0 b

I

0 ffi@

h

(u

il

A

7

CY Lrl I v)

2 W

0 v)

> 7

00

0

0

0 0

0

c

I -4 0 5 Ma)

cu I 0 rl X m cu II

k 0

* -

E E c?

E E 'p i-4 00

M

**

I1 r-4

0 0

P

NASA-Langley, 1968