(SANITIZED)UNCLASSIFIED SOVIET AND EASTERN EUROPEAN RESEARCH PAPERS ON REFRIGERATION(SANITIZED)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80T00246A007500340002-0
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 320
• Document Creation Date: December 22, 2016
• Document Release Date: April 30, 2009
• Sequence Number: 2
• Content Type: REPORT

Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1 - a'- 6
One Flow Cascade Cycle
{in Schemes of Natural Gas Liquefaction and Separation)
A.P. KLEEMENKO, C.T.S.
Institute for Gas Research of the Academy of Sciences,
Kiev, U.S.S.R.
The use of a multi-component mixture in the capacity of a cooling
agent in cycles of deep and moderate cooling cycles, as is shown
in this paper, may present a considerable thermodynamic effect.
As is generally known, energy consumption in a refrigerating cycle
is determined by the degree of irreversibility of separate processes
constituting the cycle. In all the throttle cycles, which comprise
all the steam compression cycles of moderate cold and the combined
multi-flow cycle of,deep cooling Picte cycle, the losses caused by
the irreversibility of heat-exchange and throttling processes are
major factors which determine the effectivity of,deep cooling cycles.
The irreversibility of a heat-exchange process is determined by the
temperature difference in that process.
An ideal heat-exchange process, which is characterized by a zero tem-
perature difference, is theoretically possible only at a counter
current, when equal regularities of alteration of heat exchange of
cold bearer and consumer in the function of temperature occur:
G1 Cpl l ( t ) = G2 Cp2
where: G1 and G2
- amounts of substance in straight and
reverse flows
Cpl (t) and Cp2 i2 (t) - heat capacities of straight and reverse
flows in the function of temperature
As a result of unequality of heat capacities of the substances involved
in the heat exchange and the nonconformity of their relations to tempe-
ratures, it appears possible to create, in most cases, a temperature
difference only at one end of the heat exchanger.. The temperature
difference at the other end appears to be rigidly fixed by physical
properties of the substances involved in the heat exchange. The
values of these temperature differences are sometimes very great in
deep cooling. As a result of this phenomenon, the coefficient of
thermodynamic reversibility of the heat exchanging apparatus decreases.
In order to reduce the losses during the heat exchange processes in
cooling and liquation of gases, it appears necessary that the source
of cold - the cooling agent, by means of which the cooling and liqua-
tion of as is brought about, should have about the same relation
t = `'(i) as the gas that is being liquefied.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1 - a - 6
The use of one component cooling agents boiling at constant tempera-
ture predetermines high values of temperature differences at one
end of the heat exchanger (or in the middle of it) and causes high
values of losses on heat exchange and consequently higher energy
consumption.
Multi-component mixtures have variable boiling temperatures. When
using multi-component systems, it becomes possible to achieve the
validity of the relation determined by the equation (1) for the gas
which is being liquefied and the cooling agent of multi-component
composition. In such a case the temperature differences along the
entire "length" of the heat exchanger will be the smallest, whereas
the coefficient of thermodynamic reversibility of the heat exchanging
apparatus increases Energy consumption in the refrigerating process
will be lowered correspondingly.
According to our computations and data available in literature, the
utilization of the propane-butane binary mixture, instead of ammonia,
in a0refrigeEating cycle of moderate cold for achieving temperatures
- 25 - - 30 C causes a decrease of energy consumption by 15 - 25 /.
The throttling of any substance without exception in the liquid state
is always more effective than that of any substance in the gaseous
state. The coefficient of thermodynamic reversibility of the
throttling process of a fluid reaches very high values, amounting to
a figure of the order of 0.8 - 0.9 in the zone of temperature beyond
the critical.
The transition of a gas into the liquid state may take place not
only as a result of the condensing of the gas, but because of its
being solved in another fluid as well.
The presence of a solvent causes the transition of a gas into the
liquid state at temperatures far above those of condensation and even
at temperatures above the critical. This effect may be utilized in
refrigerating cycles. When cooling a compressed binary or a multi-
component gas mixture consisting of gases with various boiling points
the first component to liquefy is that having the highest boiling
point. The liquid thus formed dissolves the other components of the
gas mixture; and as a result the liquid is transformed into low
boiling components, which at the given temperature and in the absence
of high boiling components cannot condensate. A low boiling component
will be desorbed from such a multi-component mixture at the time of
its being throttled (similarly as steam is evolved in the process of
pure substance throttling).
The decrease of the temperature level of gas liquefaction processes
as a result of absorption raises the thermodynamic reversibility of
low temperature refrigerating cycles. The combining of both the
compression and absorption cycles permits increasing the thermo-
dynamic effectivity and simplifying the constructional design of the
installation.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1 - a - 6
-3-
The one flow cascade cycle of deep cooling developed by the author
and investigated by the author and his collaborators is based on
the theoretical propositions listed in this paper: a multi-component
mixture of hydrocarbons is being utilized as a cooling agent (other
cooling agents forming ideal solutions, as freons, may be used).
The combining of the composition of hydrocarbon mixture and pressure
permi s obtaining cold at any temperature level in the range up to
- 160 C and below that figure when operating under vacuum. The
mixture composition and pressure are selected so as to satisfy the
heat exchange at minimum temperature differences.
On the other hand, the mounting and the construction of the one flow
cascade cycle are such that only the throttling of the liquid phase
is being carried out in it. The latter phenomenon, as has been
mentioned previously, predetermines the high value of the coefficient
of the thermodynamic reversibility of the process.
The suggested one flow cascade cycle which has been investigated on
a pilot scale, may be used in the following cases:
a) In the processes of methane liquefaction.
b) In cycles of air liquefaction and separation - in the capa-
city of an element of deep preliminary cooling up to the
above mentioned temperatures.
c) In cycles of gas separation in cracking and pyrolysis.
d) In cold producing installations with parameters -80 - 100?C.
In addition to the aforementioned thermodynamic advantages of cycles
with multi-component cooling agents, the one flow cascade cycle has
the following peculiarity: heat emission at the low pressure side
takes place in conditions of constant evaporation of the mixture com-
ponents complicated by purely gaseous heat exchange. The heat
emission coefficients are hereby increased just at the low pressure
side.
Description of the One Flow Cascade Cycle Pilot Plant
The pilot plant for investigating the one flow cascade cycle
consists of the following elements': a) a compressor, EC, with:::-
intermediate cooling arrangements and an end cooler, b) a first
stage liquid separator, SL-I, and a second stage one, SL-II,
c) a heat exchanger consisting of two sections, HE-I and HE-II,
d) a receiver R and e) a system of throttling valves, BY-I, EY-2
and EY-3.
Investigations of two operating regimes were made at that plant:
a) a half-closed cooling cycle for natural gas liquefaction and
b) a closed cooling cycle for air liquefaction.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1 a - 6
- 4 -
Using the regime for natural gas liquefaction, the installation
operates as follows: natural gas, consisting chiefly of methane and
containing 3.5% of ethane, propane, and butanes, enters the arrange-
ment for gas drying and purification. The drying is Sarried out by
means of alum silica gel up to the dew point -40 - 45 C. Purifica-
tion from CO is carried out by means of solid potassium hydroxide
KOH. The dried and purified gas enters the three-stage compressor C.
When experimenting with gases containing propane and butanesonly,
the condensate, as a rule, was precipitated after the end condenser
E in the liquid separator SL-I. This condensate was throttled in
the widener (Throttling) valve EY-I and discharged into the low
pressure branch of the first section of the heat exchanger HE-I.
The steam phase was directed under high pressure from the liquid
separator SL-I into the high pressure branch of the same section.
There at the expense of the reverse flow cold and that of the evg.po-
ration of the condensate it was cooled up to the temperature -40 C.
In this way practically all the butanes and up to 80 - 90 % of the
propane of the mixture are condensated and a considerable amount of
propane and ethane was dissolved. 'The liquid obtained from the
process described is separated from the steam phase in the separator
SL-II and after having been throttled in the widening valve EY-2 is
directed to the low pressure branch of the second section of the
heat exchanger HE-II. The cold obtained on the, evaporation of the
liquid is transferred to the high pressure flow and as a result it
is fully condensated and partly overcooled.
After having been throttled, up to the pressure 1.2 atm. and separated
from the steam phase, the liquefied natural gas is transferred into
the storehouse along the pipeline LNG.
The operating conditions of the installation in; respect of pressure,
concentrations of hydrocarbons C , C , Ca were so chosen as to bring
their absolute amounts in the natura~i gas (in the pipeline LNG) to
less than one half of their contents in the initial gas in the
pipeline NG.
The installation works in a closed circuit underthe operating con-
ditions for air liquefaction (or for liquation of natural gas with
a very low percentage of heavy hydrocarbons), with both pipelines NG
and LNG detached. The cold was withdrawn from the system through
heat exchanger A-LA and was spent for liquefaction of air (or natural
gas containing 98.5% of methane and 1.2% of nitrogen.
Object and Procedure of Experiment
The object of the experiment was to determine the optimum para-
meters of the scheme: composition of cooling agent, pressure,
temperature and position of condensate sampling. Temperature was
measured by means of thermocouples located at sixteen points (see
numeration in fig.l). Testing of the cooling agent was carried out
by means of a gas-analyzer CIATIM-51. Experiments were performed
on a binary mixture methane-propane and a triple one methane-ethane-
propane. Results of the experiment with the binary mixture methane-
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1 - a - 6
-5-
propane in closed circuit are shown in fig. 2. 'The relation -
cooling agent composition = f,(p) (as characterized by methane
content) - working pressure is shown in curve "a". The relation:
temperature of liquid after throttling from liquid separators SL-I
and. SL-II - in function of working pressure - is given in curves
"b" and "c".
Energy consumption for obtaining 1000 kcal of cold at temperature
level -156 C is shown in fig. 3. During experiments with that in-
stallation on a triple mixture methane-propane-butane somewhat lower
energy consumption was obtained, the optimum being shifted to low
pressure area. For that optimum: Working pressure: 60 - 65 ata.,
Mixture composition: C1 - 65% mol, C3 - 20%m, 04 - 15%.
Temperature of liquid sampling from SL-II - from -50? C up to -30?C.
For obtaining cold at a temperature level -40 - 80 0 C. experiments
were carried out in a conventional one-stage steam compressioned
cooling cycle with a regenerative heat exchanger (i.e. with detached
'Liquid separators SL-I and SL-II, fig. 1).
Results of the experiment at a temperature of cooling water + 13 -
+ 15?C are shown in fig. 4. As can be seen from the graph, energy
consumption, when obtaining cold at the level -80 -1 -70 , constitutes
from 1.7 to 1.4 kwh per 1000 kcal. This energy consumption is lower
than that in two flow cascade installations (comparable in respect
of cold production) and considerably lower than the energy consump-
tion in an ammonia cycle with two. stage compression and operating
under vacuum.
Acknowledgement.
The author expresses his gratitude to engineer Visotsky, who carried
out the experiments in the investigation of schemes developed by the
author.
COPYRIGHT RESERVED
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80TOO246AO07500340002-0
xegeno':
GNU -Cas Z7ehy0'?c7 Iin Unit.
C-C`omoeesso2.
EC' teed C00& 2. EV-2
Si-i -SS,oo'eac'o2 'o/ & /io's Z.
Hf -%- Heal exchan/e2,1- sect ion.
Sd? !i Seooealo~ of Ciylaio's-/7
11E-1;-11eaC excha,7e2, ii- seclton.
R -Receil'e2.
.EV-% EV-2 EV-3 -E. pans.on Vgeves.
- Va e ve
HH9h ,o?essaze Meow
- - - Xow,ozessoze peow
%c a&eo' Cono'ensat
feow o,tSe,o'e Consume2 co&o'.
Fig. 1. One flow cascade cycle.
x.70
? Y
N*
50
40
t13'C
-50
-45
-40
-35
-30
50 35 60 65 70 75 60 e5
Fig. 2. Methane contents of cooling agent vs working pressure (temperature level-1560 C.)
and temperature regime of the liquid separators.
Approved For Release 2009/05/01 : CIA-RDP80TOO246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
60
Pessaze oea
20 1 2 3
4 S 6 7 6 9 1 11 12 C 11
H4 14 mo/ 3 16
%
Fig. 4. Operating conditions of steam compressioned cycle with binary mixture methane-
propane at temperature level -70 - -800 C.
-40
-4z 44
-46
+
aC
_gg
-52
56
60
60
-62
-
_7Z
1
_6
-60
-70
-72
74
-76
7
-78
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
U
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
pG(r:RMIko riod
Or WORK 9)P coPwp. .SStosJ w)'CM REAL
UFti.l":17.,i_i.L70!] (''J tT}:,vii 1 (e CO.aJressiou avec les 1'ri.orig4nes ou
H.S. WEINiINJG
The Moscow Bauman Higher Technical School, Moscow, U.S.S.R.
Soh?t1!A:IInE. Pour d4terminer le travail consomme, dans un corapresseur
compression aoiabatique ou isotherme, it faut des diagramnies ther:niriues, des
tables de vapour s,.ircbaiff4e out les 4c!uations correspond antes. On ne dispose
pas de ces renseiRneaents pour les fluides frivorig~nes nouveaux ou peu
connus, it est conc necessair?e cie trouver d'autres methodes, telle que celle
indinuee ci-dessous:
ou le coefficient A::iagrat
Pour les calculs, on utilise le coefficient de co!apressibilitl" ~- Pv
P V
0 0
FtT
Ces coefficients peuve;)t titre ca.icules avec une approximation suffisante
par similitu(:e avec lea 1(:. iit,iiis4s et corriges AL l'a,iae du t:?etit nombre
de donnees expgri:-ientales (Ion.. riispose.
Le cY i c t ) ; (_ : s i' aw)liC:+.'i: ou d : d' etat Pv = FtT f-~ P.
eF;'t ici iite va)':: ')'(C'= (''?t?(:) ari4'~r (%'aprES..s les hraptliques generaux optic
le Ct:?S O') 1'411; +'1Sr?OSt' (t,:; ,. ... i, :i Ue 1(i? )res 1C)!l et la temperature Critique
ou analytici o, - :(t.. s.i + a v;i i.eu. ti e,at, connue ainsi nue les parain~tres critiques.
Le volw:e (;u r,-az ~ i evtes est ;,lus grand que celui du ga.z
ideal, c'est- -:ire ). U. U.:r in co;!! ne e saturation ou pour les vapeurs
p)eu surchauff4es sori voiun)e est inferiear is cc!...:i (du gaz ideal ( 0).
L:a m&thode or?oposee pour le ctlcui du travail de compression consiste &
Co!n1!;.(.r?(.,) (if?S C li ^'r'u:ime5 in(ica.teurs (Tune machine ideale compri!nant le gaz
(ior(r,$ el, !U CU:?'r.)rease?ur? co(npriaant un gaz ideal ayant la mere valeur de It.
ou diagram:!ie iu(licateur est facilement cal.culee d'apres les for!aules
Connues.
La ii t lode eat valfible pour la compression isotherme comme pour la
(.o:-..i ir?ession adiabdtinue.
On c1a,:ne une inethode de calcul ap )roche assez precis pour le Bernier
e1F-:: )t? tie 1'4euation.
Le travail ubsorhe est egal &:
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
on donne des forriules de calcul cowprenant les coefficients ~ ou r' au
lieu de valables pour n' importe nuelles com"itions de 11-v-T.
Des calculs types pour diverses substances indiquent que les calculs
suivant la mLthode proposge aonnent sans peine la pr4cision n4cessaire.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0 2 _
Thermal dia ,rams, superheated vapour tables or respec-
tive equations are required for calculation of the work
consumption of an ideal compressor. In the case of little
studied substances there may be no such data available.
An approximate; analytical method requiring only .a
minimum number of P-u-T data foz~the determination of work
consumption in a compressor is discussed below.
The following dimensionless group numbers are used by
the author in the calculation formulae: the coefficient ~ _
and the Amagat coefficient J" P'Y
rl recommended in the. super-
he coefficient f~- - PIr is *
243,16-R
heated vapour region instead of the Amagat coefficient. The
coefficient9 related to the critical temperature TZ
'Z R be used in some cases.
Calculations of work consumption in a compressor are
based on the utilization of the equation of e'Lat e:
P-= RTtf3 P
J3 - is here a variable estimating the difference between
the volume of real gas and that of ideal gas having the 84me
value of the gas constant value R
21= 2T +J3 = P +
In the high pre:, zur?e region V4 'UL4j and ~'7 0 and in
the superheated vapour region IR15L ( and J/ d~.. .dl T TL) 06
p
d( 7' __ T r25)
T. =1
TJ
Integrating the right part . of equation (25)# we obtain
fi 7' ~ '1's
(26)
Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
2 2 -23
Cons-,equent ly f or equal e ~
7.7
~S 1_ Tsw
_L-- rj'~' (27)
whore oor.ecponds to the critical. prer-,oure of the
unJoiocln compound.
for
if the equation (26) to applioabl 1; Z~ / , the
it rust be true also at AY ; 4_ d . Than P= .P and
1 r 77
7 9 and the expression (27) fully coinotdca with tine
oritcria JV9 which is proof of the correctness of the
initially oade asouriptions. This is also oonfirned by the
fact that for well studied oonpouands with close a, id le,
deviation in the values of 7' obtained according; to
equation (27) lies within the knits of cxpet'tmental
aoouracy /3/.
I eornarison of oompoundc iitb oonewhet varying
and ' (Freons XI, 12p 1,39 21, 22) I= revealed that in the
most well I own range v 0, the ;::ximun deviation of the
values from the mean curve does not e:coeed I
"or the val uo3 ,fit' t o,S used in refrigerating en freer ng
(the experirrnc_ita1 data are not quite reliable at higher
_>-,es1;ure3) L mid f oz w~. iouo isothc nc up to the critical, the
m:;btnum deviation w s below 2,, S.
CoordinzAion of viorl: di:?cctci tow . ds the cstablishi cnt
o.' unity of witcriw on an 1r.1%;c.Lnt..:.tional scaleaa roll as the
specification and cIz ,ci2ic..4ion o the piIonc~?tic of : tandards
is of considerable iupo~?ta ce to thco y and ,-%a.~ctice.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Reforonoeo
I. Badylkes I. i-iorking aubstanoos of rofrircrat ing
machines. uoeoovl, Pichcho-promiedat, 1952.
2. i iani i. Report at the m eetin of the SCionti,2io
Council of VNIKhI, Soptombor, 1958.
-Properties of Commonly Ucod Refigerants, i RX , 1ioohington,
1357.
Copyright reserved
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
4W .
- 1 -
Un appareil mecano-electrique pour la mesure de la puissance
frigorifique
2-52
A Mechanic-electric Apparatus for Measuring Refrigerating
Capacity
J. CHYTRALEK1 et J. PETRMICHL2
Prague, Tchecoslovaquie.
SUMMARY. The refrigeration industry requires an apparatus for
measuring refrigerating capacity, capable of recording the amount of
cold produced per unit of time. The apparatus makes possible economical
operation for given electric power consumption. A mechanic-electric
apparatus was developed which may be connected with an ammonia
refrigerating gstem of 25,000-250,000 kcal/hr for evaporation tem-
peratures of -5 , -35' C. and for, temperatures of 00, +25? C. before the
regulation valve. The apparatus is made of two parts. The first one
includes sensing elements connected to potentiometers for measuring
the main characteristic values and for computing final results.
Manometers, thermometers and a flowmeter regulate the respective
potentiometers.
The other part includes a stabilized low voltage source with
specific voltage, a stabilized source of high voltage to supply the
electronic coupling, between computor and indicator, giving the
results of measurements. The apparatus solves the well-known equation
Qo = G.(il - i2), the result of which is given by a Deprez apparatus.
The temperature of the liquid refrigerant before the regulating
valve is measured by means of a thermocouple and the pressure of
saturated vapours after the evaporator is measured by means of a spring
manometer. The respective values of pressure and temperature proportional
to enthalpy values are transformed into electric voltage by potentio-
meters. The values of voltage being substracted from one another, a
voltage proportional, to the difference of enthalpy values i1 - i2 is
obtained. In the following section of the apparatus, the difference of
enthalpies is multiplied by the weight of circulating refrigerant.
This operation is also performed in the potentiometric transmittor,
controlled by an orifice plate located in the suction line. In both
last sections a correction is made for the charge of the specific
weight of superheated vapour, flowing through the orifice. The last
section is coupled with a Deprez indicator or with a recording apparatus.
1lnstitut de Recherches des Machines Thermiques.
2lnstitut de Recherches des Machines Frigorifiques et Alimentaires.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
2 - 2-52
Bien que l'idee du mesurement de la puissance frigorifique soit
aussi ancienne que le refroidissement lui-meme, it n'existe pas meme
a present aucun appareil de service eprouve qui pourrait indiquer la
quantite du froid,produite par unite de. temps, comme it y en a dans le
cas de l'energie electrique etc. Les appareils mis au point jusqu'a,
present ne donnent que des valeurs approximatives ou ils ne fonctionnent
avec securiteque sous conditions specifiques que l'on ne peut maintenir
que tres difficilement dans l'exploitation pratiqueo Nous trouvons donc
justifie l'effort de dormer a 1'industrie du froid un tel appareil
qu'indiquera sur l'echelle d',un indicateur avec des moyens simples la
puissance frigorifique du systeme de maniere qu'on puisse la controler
par unite de temps et, la consommation de l'energie electrique etant
connue, apprecier ainsi l'economie de liexploitation.
Les methodes principales de mesure de la puissance frigorifique,
utilisees jusqu'a present, sont les suivanteso
La mesure de la puissance frigorifique utile dans l'a ent de
transmission 'e at iiquide ou gazeux p.ex. a saumure, eau, fair.
On mesure le de it de 'agen de transmission et l'ecart des temperatures;
la mesure de la puissance frigorifique globale directement dans
la section de r'agen par mesurement du debit de 1'agent circulant dans
a section iqui e ou dans la section vapeur du circuit. Le debit du
liquide est mesure generalement avant le detendeur soit a, l'aide de bac
jauge, soit par un debitmetre. Un tel equipement demande?un regime
permanent, c'est a dire dans 1'evaporateur it ne dolt entrer que la
quantite du liquide qui vraiment sera evaporee. Le debit de la vapeur
est mesure par debitmetres soit dans la conduite d'aspiration, soit
dans la conduite de refoulement du compresseur. La mesure dans la
conduite d'aspiration est plus facile et convenable. La mauvaise homo-
geneite de la vapeur ~assante qui contient des traces de lair et de
Thuile peut etre remediee plus facilement et on la trouve aussi plus
basse que dans la conduite de refoulement. Cependant, en raison de la
compressibilite des vapeurs et de la possible aspiration des gouttes
de l'agent liquide de l'evaporateur, si la construction du dernier est
inconvenable ou si la conduite est courte, les donnees de la mesure
peuvent etre deformees.
Il y a encore plusieurs methodes indirectes et des methodes calori-
metriques se basant p.ex, sur le chauffage de vapeur dune certaine
valeur en determinant ainsi le debit de l'agent frigorigene.
A la base des experiences des mesurements accomplis dans quelques
entrepots importants et aussi en considerant la simplicite, it nous
semblait la plus convenable methode du mesurement de la puissance
frigorifique executee par le mesurement du debit de l'agent frigorifique
dans la section vapeur du circuit dans la conduite d'aspiration
derriere 1'evaporateur et les valeurs d'enthalpie specifique etaient
derivees de la temperature de sousrefroidissement avant le detendeur
et de la pression dans la sortie de l'evaporateur (Fig. 1 et 2).
L'appareil etait construit pour-un systeme (circuit) frigorifique
a ammoniac avec une puissance frigorifi ue entre 25000 et 250000 frig/h,
la temperature d'evaporation etant de -P?C a -35?C et la temperature
avant le detendeur de 0?C a +25?C.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
- 3 - 2-52
L'appareil solve.la relation bien connue Qo = G.(il - '3) et
fonetionne a la base du principe d'un calculateur analogique.
Les valeurs particuliere de la relation principale sont exprimees
par analogie mecano-electrique, dont on accomplie les calculs donnes
par la relation. La valeur resultante Q0 est indiquee sur l'echelle de
l'appareil indicateur a l'.aiguille ou par 11appareil registrant.
Le debit G (ou difference des pressions au debitmetre, installe
dans,la conduite d'aspiration) est indique par un debitmetre dont les
donnees sont transformees a l'aide d'un potentiometre en une valeur
electrique. Les valeurs de l'agent frigorigene sont donnees par les
pressions ou tem eratures respectives. La valeur d'?enthalpie de
l'agent frigorigene liquide avant le detendeur dans la gamme de 1'appa-
reil donnee est une fonction lineaire de la temperature liquide (Fig. 3).
i3 = A +,a. t3
(1)
peut~etre exprimeeaisement par un potentiometre lineaire et trans-
formee en tension e'lectrique, la relation (1) etant transcrite ainsi
en forme suivante
i3 = k2-(E2 + E2)
(2)
Pour simplifier le calcul nous considerons la valeur de l'enthalpie
a la sortie de 1'evaporateur comme pour la vapeur saturee. Nous y
commettons une certaine faute, la vapeur etant en realite surchauffee
(ou meme humide), mais cette faute nest as de nature decisive.
L',enthalpie de la vapeur saturee it peut etre donnee soit par la
pression, soit par la temperature sans 1'evaporateur
i1 = f( p,x) i1 = f( t,x).
Nous choisissons la fonction i = f ( p,x) parce qu''il est plus/facile
de mesurer'la pression clue la temperature de la vapeur saturee (Fig. 4).
La continuite de cette fonction peut etre remplacee par une parabole
exprimee par la relation
i1 = C + ( - a a 2 + b. + c) - (Fig. 5) . (3)
Un potentiometre non-lineaire exprimant la continuite de cette relation
nous donne une tension correspondant a il.
Il en resulte la relation
i1 = Kl. (El + Ei)
(4)
Nous obtenons la tension ,EI + El aux resistances Rl +,Ri arrangees en
serie. La tension E2 + E2 est formee a la serie des resistances R2 +
R2 et,R~, alimentees par la meme source comme les premieres resistances.
Les resistances fixes RI et R2 correspondent aux valeurs fixes de
l'enthalpie, pour pouvoir se servir de la gamine entibre des potentiometres
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
- 4 2-52
Ri et R2 pour les valeurs variables de~l'enthalpie.Afin que le rapport
entre les tensions(E1 + El) et (E2 + E2) corresponde au rapport entre
les enthalpies it et i , on doit arranger la resistance R~ en serie
avec les resistances (H2 + R2) (Fig. 65.
A l'aide des deux potentiometres representante"s les valeurs it et
i3 nous accomplirons la soustraction
i = k12 . E12 + (Ef - E2) t - i3 (5)
La constante k12 exprime la transformation des tensions en valeurs
d'enthalpie.
A l'aide d'un autre potentiometre R' on multip,lie la difference
entre les valeurs d'enthalpie par la racTne carree de la difference
entre les'pressions au debitmetre qui correspondra au debit G. La
tension resultante (conside're'e pour le maximum de la valeur du poids
specifique du fluide circulant) doit encore etre multi liee par le co-
efficient du correction de la variation de ce poids specifique due ~.
la surchauffage. Les vapeurs circulantes a travers du debitmetre sont
surchauffe'es et'il est ne'cessaire de determiner le poids specifique a
base de pression et de temperature. Dans la gamme choisie de l'appareil
et pour une temperature maximum de la vapeur surchauffee de +20?C
(Fig. 7),le poids specifique peut etre exprime par une relation
approximee
209 . P (kg/m3) (6)
T
Dans ce cas on ne peut pas considerer la vapeur comme saturee parce
que l'erreur en resultante serait grave. Une correction correspondante
eat introduite par lea deux autres potentiometres R` et R. Ces
resistances expriment la partie variable, lea resistances k4 et R5 la
partie fixe de la gamme du poids specifique.
Le patron du potentiometre R'.a e'te ajuste pour la variation du
coefficient de detente 9 en fonction de pression avant le debitmetre
(pour une certaine gamme moyenne du debitmetre).
La puissance frigorifique est donc donnee par la relation
Q0 = Sc Km mn33 . m4 n4 . n m5 5 . k12 . k3 . kA ( E12 + ( E1 - E2) ) (7)
ou K = 0, 01252 . cc 0 d2 . ~,"~''I est la constante du, debitmetre donnee
par ies dimensions geometriques de 1'element de mesure (diaphragme)
et de la conduite.
Km = max~ . max . 1max eat une constante pour lea valeurs
maximum des parametres variables qui derterminent la,quantite du debit
du fluide.
m3 9 m 4 ' m5
sont les rapports entre la resistance du curseur et
n3 n5 n5
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
- 5 2_52
M3
celle du potentiometre entier, dont 3 correspond a 1'indication du
debitmetre n3
m3 p
n3
n4 au manometre pour la correction du voids specifique.et du coefficient
de detente en fonction de pression et n5 au thermometre pour la correction
du poids s~ecifique en fonction de la temperature de la vapeur dans le
point du debitmetre.
1 r o C- ~.
- a ,, o A V = ~.. A( + Al (po +
n4 n5
2(fi
0 +7)
Les coefficients k12 et k3 sont les constantes qui expriment les
r&lations des differentes tensions et valeurs thermodynamiques. kA est
un coefficient exprimant le gain des amplificateurs cathodiques qui
separent les etages particuliers de calcul.
Les expressions determinantes les regimes sur les.,'potentiometres
particuliers et l'expression determinante la tension peuvent etre
ex rimees par une deflection de l'aiguille sur 1'echelle de l'appareil
N (divisions). k12 . k3 . kA = K, qui est la constante de l'appareil.
La relation (7) peut etre puis formee
Qo = K . Kc Km - r (8)
Parce que la constants KK est donnee par les valeurs maximum et
l'echelle de l'appareil st divisee e1 "n" divisions, it est necessaire
d'introduire N/n.
L'appareil propre se compose de deux potentiometres pour soustraire
et de trois potentiometres pour multiplier; ceux sont commandos par des
manometres, thermometres et par un debitmetre. La tension de mesure pour
lea potentiometres eat delivree par une source d'alimentation stabilisee,
dont la valeur precise est reglee par un rheostat et c'bntrolee par un
voltmetre. Les ages particuliers des operations de calcul sont
separes 1'un de l'autre par les amplificateurs cathodiques. Le resultat
est indique par un milliampbrembtre (avec 1'echelle en,frig/h) ou par
un appareil-registreur. Les tubes des amplificateurs sont alimentees
par deux sources stabilisees de tension anodique. La tension de
chauffa~e des tubes est aussi stabilisee. (Fig. 8). Levoltmetre,
amperemetre, le reglage de la tension de mesure et du zero sur milli-
amperemetre, is commutateur du systems et de la tension de mesure et
aussi les couplages pour les potentiometres et pour le registreur se
trouvent sur la plaque en face de l'appareil. (Fig. 9 et 10). Les
potentiometres sont commandos directement par les pick-ups places au
conduites du circuit frigorifique et connectes par les?cables avec
1'appareil. (Fig. 11).
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
-6-
2-52
Les pick-ups pour la mesure de pression sont les manometres a
ressort ou a membrane. Les temperatures sont mesurees a l'aide des
thermometres bimetalliques au a pression et le debit est mesure p.ex.
par un diaphragme et une balance pendulaire. Tous les pick-ups sont
munis des potentiometres qui transforment les deflections mecanique en
valeurs electriques;
On a fait des essais de fonctionnement de l'appareil par verifi-
cation de 1'influence des quantites diverses ( A P, Po9 t A, pc, tc)
pour les differences valeurs de la 4amme de l'appareil. 0 a compare
les resultats et les valeurs calculees correspondantes aux valeurs
ajustees. La limite superieure de l'erreur etait voisine de + 2%.
Les essais pratiques etaient effectues dans un entrepot frigori-
fique en regime continu et avec un equipement d'essai dans une usine
de construction des machines frigorifiques. La puissance frigorifique
a ete controlee simultanement a l'aide des methodes de mesure classique
(manometr-e differentiel, pressions et temperatures). On a trouve les
differences + 1% en regime stable et maximum t 5% en regime instable.
La limite superi-eure de l'erreur totale pendant la determination
de la puissance frigorifique ne depasse pas ? 51, l'erreur du a
1'instabilite de l'appareil ? 1%. L'instabilite de tension peut etre
controlee pendant la mesure et corrigee d'apres un voltmetre et par la
remission a zero de 1'appareil. L'erreur generale du aux.imprecisions
des patrons de potentiometres se monte a l%. La precision absolue de
1'indicateur a aiguille est de 1% et peut-etre encore plus elevee.
` Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
kcal/kg
130
120
i3
110
r5
1 2 3 4 5 6 7 8 9 10
x
Fig. 5..
Fig.
i.1(p,x)
x-1
3901-
0
PO
Fig. 4.
Fig. 6.
alp
Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
5
4
3,619
3
410 420 430 440
/ kcal/kg
Fig. 7.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Simulation Tests of Refrigeration Turbo-Compressors
Essais par simulation de turbo-compresseursfrigorifi u~es
ING0 FRANTISEK WERGNER
CKD, Praha, Czechoslovakia
SOMMAIRE Mdthodes d'essai pour un prototype de turbo-compresseur
dans un circuit de remplacement Mdthodes d'essais par simulation
d'un prototype de turbo-compresseur a l'aide d'un fluide de rem-
placement; principe des essais par simulation; analogie du phdno-
m2ne physique, thdorie et regles pour le choix des fluides frigori-
genes essay6s; effet des nombres d'analogie individuels; exemp.le de
mdthode d'essai par simulation; description du materidl frigori-
fique et du-circuit de remplacement approprid; technique de mesure
et appreciation des resultats
?
TEST CIRCUITS OF REFRIGERATION TURBO-COMPRESSORS
Testing refrigeration turbo-compressors for the purpose of es-
tablishing their characteristic curves,.i e determining the depen-
dence of compression on the intake quantity?of refrigerant and the
dependence of compression efficiency on that quantity, is a'com-
plicated task. As a rule there are no facilities at the manufac-
turer's premises for installing complete refrigeration equipment
for testing the turbo-compressor, due to the considerable cost and
space requirements involved A simplified, so-called substitute
cycle, is therefore used A?diagram of such a cycle for testing a
compressor of a two-stage installation with three-stage throttling
of liquid is shown in Fig 1 Cooling-water quantities in the cooler
and refrigerant pressures*forr the,tests should be adjusted so as
to prevent vapour condensation in the cooler The conditions in the
tirbo-compressor suction and additional suction should be within the
range of superheated vapour, so that they may be successfully con-
trolled by measuring the pertaining pressures and temperatures
Fig? 2 shows test cycle in diagram i-log P.
TURBO-COMPRESSOR TESTS WITH SUBSTITUTE MEDIUM
Refrigerant used in the chemical industry are usually of the
explosive type, such as hydrocarbon propane, C3H8 Ammonia NH3 is
also used which again exerts unfavourable effects on the-surroundings
and'plant workers In no case can these refrigerants be used in
plant tests, because it is hardly possible to provide such protec-
tion and safety measures to safeguard the plant and the workers on
the test bench and its surroundings against explosion, fire, poison-
ing, etc Turbo-compressors using similar refrigerants are, there-
fore tested by so-called simulation tests
In these tests substitute refrigerants or gases are used The most
generally available gas is air Most suitable refrigerant replacing
refrigerants of higher molecular weight, e g propane, can be one
of the freons, which is nonimflammable, nonexplosive and harmless
to human health. In our works propane is substituted by refrigerant
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
F12 (CF2GL2),be:.ng easily available Air can. be used to replace am-
monia.
The principle of 52.ir'.zl .t _on tests is based on the theory of analogy
of physical phenomena, these phenomena being ch r ,cter-..zed by flow
analogy numbers such as li.o:jn lde number and Mach number It is under-
stood that when changi.np; t>.e wc;r':.:inG ;r:edlum one must sj..r.!ultan.eous!y
change not only the Forking pa.?a'. ~t?er?s, such as pressures and tem--
peratur?es of the working cycle, but also the machine speed When de-
termining this change use is made of the fundamental principles of
theoretical problem of an alouyr Use, is also made of the fact that in
a certain working sphere it isnposssible to simulate satisfactory
precision the physical phenomenon of circulation in a machine even
when neglecting the -h
From equation (4) we find for the adiabatic cubic compressibility
the v:=1ue :41.1 2
l0 cm /d--ne
?~4 ... Z JC, 'A o S O
It also shows the result: of this work, We.found for the
adiabatic cubic composs:ihility of solid carbon dioxide at the
of a 78,;;? C the value
k? 's :7U S L01 cm2/dyine
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1-h-3
DISCUSSION
We have compiled in Table 1 some data measured previously
by means of ultrasonic methods for the elastic constants of alkali
halides (l J and solid argon (2J. Inspecting the values of the
Elastic constants appearing in Table 1, we find a considerable
decrease of c' and c44 in moving from typical ionic crystals
towards van di; Waals crystals. The weaker the cohesive forces,
the smaller are the elastic constants.
TABLE 1 - VALUES OF THE ELESTIC CONSTANTS
Crystal
Measuring temperature
cll . 10-11
dyne/cm2
C44 . 10 11
dyne/cm2
LiF
Room temperature
,
5,54
NaCI
It
85
4
1,26
KC1
3,98
0,625
KBr
3,45
0,508
KI
it
2,69
0,362
CO
19497?K This work
0,603
0,148
2
A
80?K
0,27
0,082
This work has been reported in more details in paper ['3,].
R 'FERENCES
1,, Bergmann L. Der Ultraschall, S.Hirzel Verlag, Stuttgart,
6th ed., p.609, 1954.
2, Barker, J.R. and Dobbs, E.R. Phil. Mag. 46, 1069, 1955.
3, Hovi, V. and Mantysalo, E., Ann. Acad. Sci. Fenn. A VI 24, 1959?
Copyright reserved
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1-g-11
Electronic Specific Heat of & and /."-Brasses at Low Temperatures
La chaleur op6cifique electrenique de laitons cet A aux
basses temperatures
VA LNC HOVI and KAUKO MANSIKK t
V ihuri Physical Laboratory, Univeers ty. of Turku, 'Turku,
Finland
S0MMA.IP2, Les ohaleu.rs sp&cifiques 6J.ectroniques de laitons
c&. et oast ete etudiees en forlction de la composition en
appliquant 'la forriul.e de Sommerfeld et en faisaiit tine hypoth6se
coneernant les electrons lihres., 13. eat demcntre cue les valeurs
theoriques sent en bon accord aver lea donnees experi:mentales de
Rayn.e Bans le cas du laiton ?' ? :. Pourtant pour le laiton une
omPara.:Ison si?r i?.aire ne peut e
~; tr:?e ~aa:te a cause du fait que des
don4eea experimentales con. venables semblent toujours manquer.
INTRODUCTION
According to Sommerfeld' s theory 1 the .electronic specific
heat is given by the equation
OO?136 V`; :.nl"3 T T mi.J_li.joules/mole,degr.ee
m a .
where V is the molar volume i n
in
cm3, a_ the number of free electrons
per atom, T the absol?.ite temperature, and. is a constant,. For the
valid .t.y of equation (1), one has to assume that the electrons have
a rest mass in , and that they move in a constant potential field.
r
In the case o1 menoval.ent simple metals formula (i.) gives a quite
good approx ration of theoretical electronic specific heat.
Although this model can be regarded as an oversimplified
one, we may 'Lake into account., to some extent, the. fact that the
al ectro_ls are moving in a periodic potential field due to the
lattice by giving; them an "; ffective14 mass m, . Thus, instead of
equation (~?) we have
e
(2)
z ;re compare e_uat'..on (2) ~Rith the experimental Iinear term of the
specri.t .: .cat; we ob:te:Yu
rrj ; Mn r3 y; % heer: ,
T (3)
By means of equat._i..on.s ( J . ) , (2) and (3) i is immediately possible
to carry out nwne ical ealcul,ations for the value of the electronic
specific heats of di fferen-t, ma:ls
It :i.s clear that. one can improve the theory by including the
effect of the periodic potential V(r) due to the lattice in the
wave equation
According to Bloch s wel.]_Lknown theorem the solutions of this
equation are of the for-.,n.
(4)
Approved For Release 2009/05/01 CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1-g-11
Uk(r) eik.r
where Uk(r) has the periodicity corresponding to V(r). We see that
the special case, Uk(r) = constant, represents Sommerfeld's model.
Furthermore, in the exact mathematical treatment of the
electronic energy of metals we should take into account the
perturbation caused by electron-electron interactions. One might
also expect that the degree of order in alloys would have some
experimentally observable influence on the electronic specific
heat at very low temperature. In fact, the recent measurements
by Rayne [2J show that the electronic specific heats of an
AuCu -alloy corresponding to the ordered and disordered states
diff4r slightly from each other.
In this work, we have determined, by means of Sommerfeeld's
theory, values for the electronic specific heat of CJC and
brasses as functions of compostion. In the case of 04,-brass we
have compared our theoretical results with the experimental data
of Rayne '31. However, for /3-brasses we have not found from
the available literature experimental values of heat capacity
measured at very low temperatures.
(5)
RESULTS AND DISCUSSION
By assuming that each copper atom will give one, and each
zinc atom two, free electrons to the system of free electrons,
the average number of free electrons per atom in a copper-zinc
alloy can be represented by the formula
nav = 1 + 10`?2 p , (6)
where p stand for the atomic percentage of zinc. If our assumption
is true, equation (6) is valid only for zinc concentrations of
OK5O atomic percentages. This is because, as we know from
experimental investigations of the electronic specific heat of zinc,
the.number of free electrons per atom in pure zinc is not equal to
two (cp L4__7 ). At zinc concentrations larger than 50 %, we may
suppose that only those of the zinc atoms corresponding to the same
number of copper atoms will behave according to equation (6). The
excess of zinc atoms would behave like the atoms in pure zinc as
regards the number of free electrons per atom. On the basis of
equations (1) and (6), it is highly probable that the values of
will have a maximum at about equal concentration of components.
In order to calculate values for V in equation (1), we
should know the average atomic weight ofmthe alloy. This can be
obtained from the formula
Mav ' pZZn + gZCu (7)
100
Here p and q are the atomic percentage of zinc and copper,
respectively, and ZZn and ZCu are the corresponding atomic weights.
Then
V =
m
Mav
------------ ,
alloy
(8)
01
where alloy denotes the density in g/cm3. For our calculations
we used the measured values of density appearing in paper (5-7.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
1-g-11
The theoretical data calculated-for the quantity 6 of o< and
a -brasses are given in Table 1. Table 2 shows the "effective"
mass ratio of Ck:-brass.as function of composition. The theoretical
and experimental values of ''are represented graphically in Fig. 1.
Table 1
Theoretical values of
.compositions
P V
(%)
av
Cr.00
1000
1,49
1?015
3.15
1,031
5098
i o 60
7,80
1.
0
78
8.i8
,
/
1.082
9,76
1.098
1.5-07
1,151
10,,-.,97
1,200
-2. L'!( 0 50
1r 245
29._19
lu292
3298
1 330
L0,00
1 4oo
k2 ou
10420
4r 00
1?44.0
LF6.00.
1,460
48,,oo
1,.46o
50,00
11500
obtained for, d.! and f3 at different
(cm5)
7,065
71.67
7,198
70243
7271
7,277
7 304
7=378
1.446
7.51 7
'17573
7618
7.707
77,:732
7 7`59
70784
7,,8o8
7832
(millijoules/mole-degree2)
0:,501;
.0,5o8
0.513.
00519
0.523
0.524,
0,529
0.540
0,551
C.,56l
0.571
r
0-5790 A 593.
0 ,,598
0.602
0.606
0,610
O A14
Table 2
The ~`effecti_vep? mass ratio of 0(,,-brass as a function of
compo;,_t:.on
(~)
O,GO
1,ax, the difference between
the condensing temperature Ti{ ax involved and the temperature Ta of surroundings
through which heat is removed from the condenser, being Lk max.
With the heat load of the space refrigerated decreased from Qo ,,,ax to value Q0,
the system starts operating in cycles composed of the period during which the com-
pressor is running and the period during which the compressor is at a standstill.
During the working period of the cycle, the evaporator temperature drops,
and rises during its off period. The system works with varying evaporator tempe-
rature and the mean difference between the temperature of the space refrigerated
Tr and the mean evaporator temperature Torn is decreased to Ao:
At the same time the difference between the mean condensing temperature Tk,,,
and the temperature T. of the surroundings cooling down the condenser, will be
decreased to Ak. ?
These conditions are shown in Fig. 1. After starting the compressor, at the mo-
ment T1, the evaporator temperature drops from value To, until at the moment T2
it- reaches value Toe at which the compressor is cut off by the thermostat or presso-
stat. The evaporator temperature will again -start rising until at the moment T3, at
temperature Tol, the compressor will be started.
The mean evaporator temperature during the compressor running period is not
identical with that during its off period. Under the simplified assumption that a
single-capacity system is involved, the mean evaporator temperature during the
working period of the compressor will be:
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
*
Tk2
r
T
T* ,
Tkm2
T i
Tm2
o
m2
U2
Ton,,
02 T mf
of
1 'r2-Zt
Tome= Tomin+(rot-Toin) e- dT
T2 - TI TC
Ta
Where To-evaporator time constant.
The mean evaporator temperature during the compressor off period is
T3
T,-(T,-To2)e-
Ta-T2 dT
Tc
The mean evaporator temperature during the whole cycle is
Toni Tom,. (T2-Ti) + Toma (T3-T2)
=
(1)
(2)
(3)
For evaluating the economy of the refrigerating cycle, only value Tom, enters the
picture.
Similar equations can lie written also for the mean condensing temperature during
the compressor running period Tkm,, mean condensing temperature during the off
period Tkm2 and mean condensing temperature during the full cycle Tkm.
The influence of the length of the whole on-off cycle T., T, on the size of value
Tom,. .comes to prominence by substituting extreme values for TZ Ti and T, _T2
into equations (1) and (2).
If, T3 - Ti -, 0, then ?
lim To, = l,m Tom, = Tom, lim TO2' = lim Toe = Tom
If, on the other hand, T3 Tt - ~c then lim Tot = lim Tome = Tr, lim Toe =
lim Tom,. = To min? -
By changing the length of cycle T,Tt in an on-off control system it is therefore
possible to change the mean evaporator. temperature during the working period of
cycle Tom, within the range of (To min, Tom).
Similar conditions prevail on the condenser side.
If, T, - Tl - 0, then
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Effect of On-Off Control on Working Economy of Refrigeration Equip-
ment
Influence du,reglage par tout ou rien sur 1'economie du fonctionnement du
materiel frigorifique
V. POLAK
Research Institute for Refrigeration and Food Engineering,
Prague, Czechoslovakia.
SOMMAIRE. La consommation de courant electrique du materiel frigorifique
avec reglage par tout ou rien de la production frigorifique varie suivant la longueur
du cycle de fonctionnement, en raison de 1'influence des etats transitoires. On
prouve que l'economie du cycle frigorifique augmente en, fonction inverse de la
longueur du cycle de fonctionnement et l'on montre 1'ef fet des constantes de temps
de l'evaporation et du condenseur. On indique on certain nombre de facteurs de la
conception du compresseur et daps son exploitation qui ont one influence defavorable
sur 1'economie de l'ensemble du systeme frigorifique en regime variable. -Pour
obtenir l'economie maximum du materiel frigorifique avec reglage par toute ou rien
de la production frigorifique, le compresseur et son moteur doivent titre confus aussi
du point de vue de l'economie en regime variable de fafon qu'on obtienne le maxi-
mum d'economie dans le fonctionnement avec des cycles courts.
The input of refrigerating systems with on-off control of refrigeration output
varies with the length of the on-off cycle. This is obviously the result of the influ-
ence of transient states. I intend to show factors to be taken into account, because
the design and adjustment of this type of refrigerating system carried out with
regard to these viewpoints, can contribute to a more economical operation.
Let us assume a single-evaporator refrigerating system with on-off control. The
refrigerating circuit is assumed to be dimensioned so that with the compressor
running continuously, the evaporator removes heat Qo max from the space refrige-
rated, the difference between the space refrigerated T,. and evaporator temperature
To min being Lo max,
At the same time the condenser is removing heat Qk max, the difference between
the condensing temperature Tk ,,,ax involved and the temperature Ta of surroundings
through which heat is removed from the condenser, being Lk max.
With the heat load of the space refrigerated decreased from Qo max to value Q0,
the system starts operating in cycles composed of the period during which the com-
pressor is running and the period during which the compressor is at a standstill.
During the working period of the cycle, the evaporator temperature drops,
and rises during its off period. The system works with varying evaporator tempe-
rature and the mean difference between the temperature of the space refrigerated
Tr and the mean evaporator temperature Tot, is decreased to Ao.
At the same time the difference between the mean condensing temperature Tkm
and the temperature Ta of the surroundings cooling down the condenser, will be
decreased to Ak.
These conditions are shown in Fig. 1. After starting the compressor, at the mo-
ment T1, the evaporator temperature drops from value Toy until at the moment T2
it reaches value Toe at which the compressor is cut off by the thermostat or presso-
stat. The evaporator temperature will again start rising until at the moment T3, at
temperature Tol, the compressor will be started.
The mean evaporator temperature during the compressor running period is not
identical with that during its off period. Under the simplified assumption that a
single-capacity system is involved, the mean evaporator temperature during the
working period of the compressor will be:
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
TT
T* Tmf
Tkm2
Ti
Tm2
o
mt
01
T642
Tani
O2 QM'
T
r-101
om
omn p2
roml = Tomin -- (Tot -To ruin) e- -' -1 dr
72-T1 .rc
t2
Where TC evaporator time constant.
The mean evaporator temperature during the compressor off period is
?Z3
Tr=(TT-Tone- '3 "2 tla (2)
r.~
The mean evaporator temperature during the whole cycle is
T To m1 (T2 -.Z1) + T11.2 (ZS -t2) (3)
Tons =
zs- T,.
For evaluating the economy of the refrigerating cycle, only value Tom,. enters the
picture.
Similar equations can be written also for the mean condensing temperature during
the compressor running period Tkml, mean condensing temperature during the off
period Tkm2 and mean condensing temperature during the full cycle Tkm.
The influence of the length of the whole on-off cycle T ,-T, on the size of value
Toml comes to prominence by substituting extreme values for T,T1 and T, ,-T2
into equations (1) and (2).
If, T3 - Ti -, 0, then
lim To1 = lun Tom,. = Tom, lim Toe = lim Toe = Tom
If, on the other hand, T,, T1 30 then lim To1 = lim Tom, = Tr, lim T02 =
lim Tom,. = To min-
By changing the length of cycle T3 T1 in an on-off control system it is therefore
possible to change the mean evaporator temperature during the working period of
cycle Tom,, within the range of (To min, Tom).
Similar conditions prevail on the condenser side.
If, Tz - T1 0, then
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
?~1
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
lim Tkl = lim Tkml = T,,,,, lim Tkz = lim Tkmz =Tkm
If, on the other hand, T3 - Tl - oo then
lim Tkl = lim Tkmz = Ta, lim T12 = lim Ticmi = Tk max,
By changing the length of cycle T3 --T1 in an on-off control system it is, there-
fore possible to change the mean condensing temperature during the working period
of cycle Tkmi within the limits of (Tkm, Tk max)
Since the change in the evaporator temperature is taking place at the same time
as the change in the condensing temperature, the value of the coefficient of per-
formance of Carnot refrigerating cycle may reach values within the range of
(Emin, Es), where
Tomin
Tkm - Tom
Values T?m and Tkm are changing with the relative heat load of the space refri-
gerated Q/Qmax. So under maximum heat load and with the compressor working
continuously, then Tom = To min and Tkm = TI, max. As heat load of the space re-
frigerated decreases, there is a drop of the mean difference Tr - Tom and of mean
difference Tkm = Ta. If Q - 0 then lim Tom = Tr and lim Tkm = Ta.
Tomin -F (1 - QQ do max
max
ES (6)
= Q
Tk max -Tomin (1 - (dk max -h d o max)
Q,nax
What values can be reached by relation Emi^/Es are shown in diagram Fig. 2,
compiled for relations Ak max = An max = 15? C.; mean temperature of sur-
roundings cooling down the condenser Ta = 20? C. This diagram shows the de-
pendence of relation Emin/Ea on the relative heat load of the space refrigerated
Q/Q1,ax at temperatures of the space refrigerated from -20? C. to -1-100 C. The
influence of suction pressure on the compressor refrigeration output has not been
taken into account.
/4
7 -
o? .c ~o ti
Fig. 2. Rise of the coefficient of performance of Carnot refrigeration cycle with decreasing
heat load.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
For example with relative heat load QIQmax = 0,5, a region very common in
refrigerating equipment, the relation c,nin/'ms's is about 0,7. Under these conditions it
should be possible in the extreme case to reduce power consumption by as much as
30%. In reality in all cases Tom. > T. ,n;n.and Tltm, < Tk inax and also To,n, <
Tom and Tkm, > Tknr
It is obvious that the coefficient of performance of Carnot refrigeration cycle will
grow with increasing mean evaporating temperatures and decreasing condensing
tenperatures.
From the foregoing, the extent to which the thermostat or pressostat differential
will influence the length of the on-off cycle and the values of mean temperatures
in the compressor running period Tom, and Tkm. appears. The magnitude of the
mean condensing temperature can moreover be influenced by the magnitude of the
condenser and evaporator time constants. As shown in Fig. 3, large condensed time
T2
Tk2 Tk max
-T m
T m2 Tksc.,a
km2__,-.,.
T T2 T3
Fig. 3. Effect of condenser time' constant on condensing temperature curve in transient
states.
constant will reduce the mean temperature in the period of working cycle Tkm,?
Small evaporator time constant will, at equal thermostat differential, shorten time
T2 - T1, thus indirectly reducing mean condensing temperature Tk,n.. Decisive for
the size of the time constants are heat capacities of these exchangers and the vapor
volumes of the low-pressure and high-pressure sides of the refrigeration circuit.
Heat capacity of water condenser is substantially higher than that of an air
cooled one. Heat capacity can be increased by increasing the weight of material or
the charge of refrigerant or water. The circuit volume can be enlarged by additional
tanks.
Both these alternatives of increasing the time constant will naturally increase
the investment costs, the dimensions and the weight of the equipment. It is obvious
that the most effective means of reducing difference Tom - Tum, and Tkm, - Tkm
is by decreasing the differential of the thermostat or pressostat. This is easily ac-
complished with the evaporator thermostat and the pressostat, their differentials
being extensive. ,
The conditions of maximum economy of the whole refrigerating system are
different from those of the refrigeration cycle, for the following reason: the economy
of a refrigerating cycle as shown above, increases with the decrease of the diffe-
rential, but on the other hand, there is an increase in the number of startings and
stoppings of the compressor resulting in power losses in the compressor proper, its
transmission and electric drive.
Starting and stopping will create unfavourable phenomena such as losses by
piston leakage, which are very important at low speeds, and losses by friction. On
the other hand, wiredrawing is smaller. Another unfavourable phenomena is high
inertia of the compressor mechanism, extending the period during which these losses
occur.
Exceptionally high are those losses 'that arise during transient states in a circuit
provided with capillary tube as expansion element. After stopping the compressor
there are heat losses due to noncondensed vapours entering the evaporator from the
condenser.
Losses of this nature occur also after compressor starting until the moment when
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-4
lim Tkl = lim Tkml = Tkm, lim Tk2 = lim Tkm2 = Thin
If, on the other hand, T3 - Tl - oo then
lim Tkl = lim Tktt2 = Ta, lim T1,2 = lim Tknnl = Tic max.
By changing the length of cycle T3 - Tl in an on-off control system it is, there-
fore possible to change the mean condensing temperature during the working period
of cycle Tkml within the limits of (Tkm, Tk max).
Since the change in the evaporator temperature is taking place at the same time
as the change in the condensing temperature, the value of the coefficient of per-
formance of Carnot refrigerating cycle may reach values within the range of
(Emin, s, ), where
Tomin
min =
Tk max - To min
ES _
Tkm - Tom
Values To? and Tk,,, are changing with the relative heat load of the space refri-
gerated Q/Qmax. So under maximum heat load and with the compressor working
continuously, then T?m = To min and Tkm = Tk max. As heat load of the space re-
frigerated decreases, there is a drop of the mean difference T,. - Tom and of mean
difference Tkm = Ta. If Q -, 0 then lim Tom = Tr and lim Tkm = Ta.
+ 1 -
T
Q
) d
Es
omin
Qmax
o max
(6)
-
1 -
T
T
Q
)
A
P
o min
k max -
I
Q
max
k max +
. max
)
What values can be reached by relation Emin/Es are shown in diagram Fig. 2,
compiled for relations Al max = A? max = 15? C.; mean temperature of sur-
roundings cooling down the condenser T. = 20? C. This diagram shows the de-
pendence of relation' Emjn/Ea on the relative heat load of the space refrigerated
Q/Qmax at temperatures of the space refrigerated from -20? C. to + 10? C. The
influence of suction pressure on the compressor refrigeration output has not been
taken into account.
o? .c 10 0
x,
0.2 0.4 0.6 08 E?= i1
s
Fig. 2. Rise of the coefficient of performance of Carnot'refrigeration cycle with decreasing
heat load.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
04
2
8
b
-
L
0
O
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-4
For example with relative heat load Q/Qmax = 0,5, a region very common in
refrigerating equipment, the relation s,,,in/a5 is about 0,7. Under these conditions it
should be possible in the extreme case to reduce power consumption by as much as
30%. In reality in all cases Ton,1, > To min and Tkmt < T,, max and also Ton,l <
Ton, and Tkml > T.
It is obvious that the coefficient of performance of Carnot refrigeration cycle will
grow with increasing mean evaporating temperatures and decreasing condensing
tenperatures.
From the foregoing, the extent to which the thermostat or pressostat differential
will influence the length of the on-off cycle and the values of mean temperatures
in the compressor running period Tnm1 and Tkml appears. The magnitude of the
mean condensing temperature can moreover be influenced by the magnitude of the
condenser and evaporator time constants. As shown in Fig. 3, large condensed time
Tkrn1T
T m
2
T
rs
T~
Tm2
T2 Tj
Fig. 3. Effect of condenser time constant on condensing temperature curve in transient
states.
constant will reduce the mean temperature in the period of working cycle Tk,,,l.
Small evaporator time constant will,, at equal thermostat differential, shorten time
T2 - T,., thus indirectly reducing mean condensing temperature T1,nl. Decisive for
the size of the time constants are heat capacities of these exchangers and the vapor
volumes of the low-pressure and high-pressure sides of the refrigeration circuit.
Heat capacity of water condenser is substantially higher than that of an air
cooled one. Heat capacity can be increased by increasing the weight of material or
the charge of refrigerant or water. The circuit volume can be enlarged by additional
tanks.
Both these alternatives of increasing the time constant will naturally increase
the investment costs, the dimensions and the weight of the equipment. It is obvious
that the most effective means of reducing difference To,,, - T0,n, and Tkn,i - Tkm
is by decreasing the differential of the thermostat or pressostat. This is easily ac-
complished with the evaporator thermostat and the pressostat, their differentials
being extensive. ,
The conditions of maximum economy of the whole refrigerating system are
different from those of the refrigeration cycle, for the following reason: the economy
of a refrigerating cycle as shown above, increases with the decrease of the diffe-
rential, but on the other hand, there is an increase in the number of startings and
stoppings of the compressor resulting in power losses in the compressor proper, its
transmission and electric drive.
Starting and stopping will create unfavourable phenomena such as losses by
piston leakage, which are very important at low speeds, and losses by friction. On
the other hand, wiredrawing is smaller. Another unfavourable phenomena is high
inertia of the compressor mechanism, extending the period during which these losses
occur.
Exceptionally high are those losses that arise during transient states in a circuit
provided with capillary- tube as expansion element. After stopping the compressor
there are heat losses due to noncondensed vapours entering the evaporator from the
condenser.
Losses of this nature occur also after compressor starting until the moment when
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Tkz T max
Tkm
Tkt Ta
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-4
the pressure in the condenser has risen to'such a level that a sufficient quantity of
refrigerant is condensed. Up to this point there exist also losses of energy connected
with rise of pressure above that value.
At the same time heat losses are brought about by leakage of liquid refrigerant
from the evaporator to the suction piping, similarly as in circuits with expansion
valves.
Efficiency of the electric motor during starting is low as well, due to increased
losses in the winding as well as hysteresis losses in the rotor.
It would be possible to reduce these losses at the cost of decreasing the starting
torque. In most cases this will be possible at relived compressor starting only. Here
as well the high inertia of the motor rotor has unfavourable effects in extending the
uneconomical operation during transient state. The reduced starting torque will
naturally extend starting time.
We can see, therefore, that for determining optimal conditions during starting,
several contradictory requirements have to be taken into account.
What has been said in the foregoing can be recapitulated as follows: The con-
troller differential with temperature on-off control exerts a considerable effect on
the economy of operation. The economy of a refrigeration cycle rises with decrease
of differential.
There is, however, a number of factors to be found under transient states, which
move the economical optimum towards higher differentials, thus deteriorating the
economy of the whole refrigerating system.
For safeguarding highest economy of operation it is, therefore, necessary to give
due attention to not only adjusting the controller differential for optimum economy,
but also designing the refrigeration circuits and compressor drives so as to achieve
their highest efficiency even under transient states.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-4
the pressure in the condenser has risen to such a level that a sufficient quantity of
refrigerant is condensed. Up to this point there exist also losses of energy connected
with rise of pressure above that value.
At the same time heat losses are brought about by leakage of liquid refrigerant
from the evaporator to the suction piping, similarly as in circuits with expansion
valves.
Efficiency of the electric motor during starting is low as well, due to increased
losses in the winding as well as hysteresis losses in the rotor.
It would be possible to reduce these losses at the cost of decreasing the starting
torque. In most cases this will be possible at relived compressor starting only. Here
as well the high inertia of the motor rotor has unfavourable effects in extending the
uneconomical operation during transient state. The reduced starting torque will
naturally extend starting time.
We can see, therefore, that for determining optimal conditions during starting,
several contradictory requirements have to be taken into account.
What has been said in the foregoing can be recapitulated as follows: The con-
troller differential with temperature on-off control exerts a considerable effect on
the economy of operation. The economy of a refrigeration cycle rises with decrease
of differential.
There is, however, a number of factors to be found under transient states, which
move the economical optimum towards higher differentials, thus deteriorating the
economy of the whole refrigerating system.
For safeguarding highest economy of operation it is, therefore, necessary to give
due attention to not only adjusting the controller differential for optimum economy,
but also designing the refrigeration circuits and compressor drives so as to achieve
their highest efficiency even under transient states.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Thermodynamic Investigations of the Working Cycle of the Philips
Machine
Recherches sur les transformations thermodynamiques dans le cycle de
fonctionnement des machines frigorifiques a gaz Philips
I. I. KARAVANSKY and L. Z. MELTSER
The Odessa Technological Institute of the Food and Refrigerating Industry,
Odessa, U.S.S.R.
SOMMAIRE. Le rapport etudie let conditions de fonctionnement de la machine
frigorifique a gaz Philips'a mouvement alternatif des pistons. On suppose que
touter les transformations de la machine sont reversibles grace a deux sources de
chaleur a temperature constante. A 1'inverse de Codegone, les AA. basent leurs
calculs sur un espace mort considerable (dans les conduits du regenerateur, du
refroidisseur d'eau et du congelateur) en proportion du volume du cylindre de la
machine. Une nouvelle methode, fondee sur ['application de ['equation energe-
tique generale pour une quantize de gaz variable est utilisee pour 1'etude du cycle.
Le but de la recherche thermodynamique est de determiner les quantites de chaleur
reellement echangees par seconde aux echangeurs de chaleur de la #nachine.
Le probleme est resolu a ['aide de 1'equation caracteristique et de ['equation
energetique de la quantite de gaz variable.
L'etude- fait apparaltre un ecart considerable des quantites de chaleur par
seconde echangees a chaque appareil par rapport aux quantites calculees sur la
base de l'equilibre thermique du cycle.
La deduction de Codegone suivant laquelle la chaleur provenant du regenerateur
est egale a CpiT ne s'est revelee exacte que lorsque les quantites de gaz a ['entree
et a In sortie du generateur sont egales a tout moment.
On donne une analyse de ['influence des principaux parametres du cycle stir
son rendement en se basant sur les relations derivees a savoir:
1. la relation de la puissance frigorifique par unite de volume avec ['angle direc-
teur a. Il est possible de determiner in valeur optimum de a pour les conditions
donnees.
2. L'influence du parametre de construction egal au rapport de la valeur maximum
des volumes utiles de in machine stir in puissance frigorifique par unite de
volume et le rapport des pressions limiter du cycle.
L'utilisation de la methode proposee permet une meilleure description des carac-
teres reels des processus non stationnaires qui se produisent daps la machine et
orientent la methode du choix des principaux parametres de la machine.
The analysis and calculation of the cycle of the Philips machine even simplified
and schematic have not as yet been completely elaborated. Kohler and Jonkers [1)
give. no solution of the problem of a theoretical determination of the per-second
heat exchange in the working spaces of the machine and the regenerator. Investi-
gations carried out by Codegone [2] dealt with the cycle of a machine having a
null clearance volume.
The investigation to be discussed in this paper has been based on the thermo-
dynamics of a variable amount of gas. The fundamentals of this thermodynamics
have been set forth by M. A. Mamontov [3). The application of the method
allows solving the problem in a general case and giving a more complete illustra-
tion of the processes occurring in the machine.
Fig. 1 illustrates our diagram with the clearance and the volumes of the heat
exchangers allowed for. The sinusoidal motion of the pistons, close to actual
conditions, was accepted for approximate calculation in all the investigations.
All the mechanical and thermal processes of the cycle of the machine are
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
supposed to be reversible. Due to the fact that only two external heat sources
participate in the process with the temperatures TC and TE, it follows that the
theoretical coefficients of performance is to 'be equal to the Carnot coefficient, i. e.
30
DO -41 il0 360a
Fig. 1. Diagram of the Philips machine and piston motion. 1-water heat exchanger; 2-rege-
nerator; 3-refrigerator.
The following nomenclature is used in the analysis:
VC -operating volume of the warm space;
VE - operating volume of the cold space;
W - ratio of the maximum operating volume of the warm space VC max
to that of the cold space VE max;
p - the lead angle of the maximum volume VC max in relation to the
volume VEmax; -
VmC - the total volume of the warm space clearance and of the water
heat exchanger;
VmE -the total volume of the cold space clearance volume and of the
refrigerator;
VR - the volume of the' refrigerant in the regenerator;
Go; GE; Gft - the weight of the refrigerant in the warm and cold spaces and in
the regenerator respectively;
TR - the mean temperature of the refrigerant in the regenerator;
TC - the mean temperature of the refrigerant in the warm space;
TE - the mean temperature of the refrigerant in the cold space.
The following is derived on working out the equation of state for each space and
regenerator separately and having summed them up by terms:
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-29
PVzed = GRTE (1)
where Vaed = VE + VCTC + Vm - is the volume of the machine reduced to the
temperature TE,
V. = V,?, + VR TE + Vmc 1-E - the clearance volume of the machine reduced
G = G, + GE + GR - the total weight of the refrigerant.
In case of the refrigerant being distributed along the length of the refrigerator
in compliance with a linear law the mean temperature of the refrigerant T11 is
determined according to the following formula:
TQ - TE (2)
TC ()
TR TC
to the temperature TE,
The equation (1) allows defining the pressure of the refrigerant by the set
motion of the pistons.
The initial figures in the performed calculations have been chosen close to those
obtained in the Philips machine and namely:
TC = 300? K, TE = 75? K, VC max + VE max = 300 cm3,
V111L = 70 cm3, V,,,E = 35 cum', VR = 105 cm3,
the speed of the crankshaft n = 1500 rpm.
These data were used in all subsequent approximate calculations. The values of
q and W were assumed to be W = 1; (p = 90?.
Let us consider now the heat balance of the machine. We shall use equation
of the first law of thermodynamics for a variable amount of gas (3).
dQ + rdG = du + Ada + rK dGK (3)
where dQ - is the heat exchange of the gas with -the walls;
a - the energy of 1 kg of entering gas;
7rK - the energy of 1 kg of leaving gas;
dG - the weight rate of entering gas;
dGK - the weight rate of leaving gas;
dU - the change of the total intrinsic energy of the gas;
Ada - the external work of expansion.
The energy of 1 kg of the escaping gas a, is equal to the enthalpy of the
escaping gas, while the energy of 1 kg of entering gas n is equal to the enthalpy
of the entering gas only in the case of the heat transfer in the gas supply duct
amounting to nought.
Let us make up an equation of the first law for the refrigerant of the warm
space during the period of one cycle:
QQ +.pidG =q,dU+q,PdVV +pi,dGk,
where: Qc - is the amount of heat removed from the refrigerant in the warm
space during one cycle.
It is evident that at the stable working conditions of the machine the amount
of refrigerant entering the warm space and that leaving it in the course of one
cycle are equal
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
yodG = g9dGK
and the intrinsic energy of the refrigerant in the warm space is not changed after
a single cycle:
pdU=0.
The equation will thus acquire a more simple shape:
Q = ArpPdVV.
Having made up the equation of the first law for the refrigerant in the cold
space in the course of one cycle, we receive:
QE = APPdVE,
where: QE-'S the amount of heat supplied to the refrigerant in the cold space
. during one cycle.
The coefficient of performance of the cycle being investigated (no matter what
the character of the piston motion) is to be equal to the coefficient of the reverse
Carnot cycle.
let us analyse the expression QE + TE T QC.
C
The following will be received by using formulae (4) and (5):
QE + TC Qc = dpP /dVE + TC dVc/.
And if we consider that dVE + 7T,E dVE = dVxed, and,P may be determined from (1),
C
then finally:
Qc - QE TC - TE ' .
In the case, when the pistons have a sinusoidal motion (.ree Fig. 1). the equation
(5) acquires, after integration, the following shape:
QE = AGRTE 2"TW sin q a
c2 1 h'/a2j /'
where: z = TE , a = 1 + rW + 2V,,, , c2 = 1 + 2rW cos 4p + r2 W2.
TC VmE
For the determination of Q0 it is best to use the relation (6).
Due to the fact that the heat flows in the spaces are non-stationary, the values
QE and QC do not provide a complete characteristics of the heat exchange taking
place in the apparatus of the machine. An analysis of the heat exchange intensity
in each space separately may help to fill up this blank.
With the equation of the first law for the cold space made up, the following
expression may be derived after a number of transformations:
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80TOO246AO07500340002-0
(t - the time in seconds).
Fig. 2 illustrates the 'raphs of dependence of the space heat exchange intensity
for two cycles with various values of W. It follows from the graphs that there
is a change in the direction of the heat flows in .he spaces in the course of one
cycle. It is worth noting that with W = 1 the refrigerating capacity of the cycle
Qr, is only 26% of `the total amount of heat supplied to the refrigerant in the
cold space (area abc in Fig. 2). The greater part of this heat is removed from
the refrigerant during the second "period of the cycle. For greater clearness the
values of QE and QC at W = 1 are represented in Fig. 2 by shaded rectangles.
W-1
O
-2
-3
-5
-6
-7
-B
O 90 foO 270? O
Fig. 2. Intensity of heat exchange during one cycle.
\ W4
Qv
The per second rates of flow dd~ and ddtL may be determined by differen-
tiating the equations of state for, the respective space.
dA ` - / V, -f V. / u dted 1 . (10)
Approved For Release 2009/05/01 : CIA-RDP80TOO246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-29
dGE G
dt Vzed
E dVed
Lz ed lit - / VE 'I' VrnE / d t
The equation of the first law is to be worked out for the refrigerant in the
regenerator in order to determine the per-second heat load of this apparatus.
With a number of transformations made the following expression is obtained:
Pmax.
Qm%n.
(12)
Some comparative calculations have been done to determine the influence of
separate parameters of the machine upon its operating characteristics.
The influence of the lead angle qp was investigated first.
dQR = /ic- Up / d GC -}- / iE- U0 / d1E
A'KAA .
z KrCek.
60
0l 0
P ax.
Pm %n.
o kfAA.
{EVroTet.
1000
600
400
30 60 90 12,0 fS0 f80 So ?
Fig. 3. Refrigerating capacity depending upon lead angle S.
Fig. 3 illustrates the graph of dependence of the refrigerating capacity of 1 kg
of refrigerant qE upon the angle p. Besides, the value of the refrigerating capacity
qEv, related to the total of the maximum operating volumes of both warm and
cold spaces at equal values of the clearance and of the maximum pressure of the
cycle were defined.
VCmax+VEmax
The graph of the dependence of the pressure ratio p'nax upon the lead angle cp
pmin
is also given in the figure. It follows from the graph that with an increase of q>
the pressure ratio pmax drops, and consequently, the compression and expansion
pmin
losses are lessened likewise. In order to obtain the maximum value of qEV with
minimum compression and expansion losses, the values of p ranging from 90?
to 110? are to be recommended.
For the determination of the influence of W the values of qE; qEV and pmax
pmin
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
were found at various values of W. A change of W was regarded as a redistri-
bution of the same total of operating volumes VCmax + VEmax between the
warm and cold spaces, the other parameters. of the machine remaining constant.
Graphs were drawn of the intensity of heat exchange between the spaces for
various values of W. An examination of these graphs indicates that with an
increase of W the exchange in the warm space is subject to insignificant changes,
and that of the cold space is reduced significantly due to a decrease of regene-
ration. The latter must have a favourable effect upon the losses of the machine.
Graphs for W = I .and W = 4 are given in Fig.2.
REFERENCES
1. KOHLER, J. w. L;, and JONKERS, c. o. Fundamentals' of the gas refrigerating machine
Philips. Technical Review, 16, (3), 1954.
2. CODEGONE, K. The Philips Refrigerating cycle. Proceedings of IX International Con-
gress of Refrigeration, 1955.
3. MAMONTOV, M. A. Some cases of gas flow. Oborongiz, 1.951.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
On the Energy Efficiency of Thermoelectrical Refrigeration
Caracteristiques de 1'energie du refroidissement electro-thermique
V. S. MARTYNOVSKY and'V. A. NAER
The Odessa Technological Institute of the Food and Refrigerating Industry,
Odessa, U.S.S.R.
SOMMAIRE. La relation entre le coefficient de rendement des machines frigori-
fiques a compression et a absorption, leur puissance et la temperature des sources de
chaleur est etudie d'apres les donnees des experiences et des calculs. On propose une
methode precise pour le calcul des systemes electro-thermique permettant d'evaluer
le coefficient de rendement du systeme. Cette methode tient compte de la relation
entre la temperature des jonctions des thermoelements et du courrant qui les tra-
versee (c'est-a-dire la relation entre les temperatures des jonctions et l'importance
des fleux de chaleur),
Des essais of fectues avec une installation electro-thermique d'evaporation a semi-
conducteur confirme les formules etablies et permet des comparaisons avec les
systemes exitants a compression, a absorption et a ejection. Les experiences ont ete
of fectuees sur une installation de pompe a chaleur d'evaporation a semi-conducteur
it etait possible de reduire la consommation de courant electrique de 4 a 5 fois
pour AT = 10? C et avec une temperature d'evaporation de 100? C par rapport an
chau f f age electrique direct.
Le rapport presente aussi les domaines rationnels d'application de 1'ef fet electric-
therrnique pour les installations f eigori f iques et les installations de pompes a chaleur.
Dans le cas de materiaux semi-conducteurs in tel domaine d'application rationnel
(du point de vue de la consommation de courant) pour les generateurs frigorifiques
s'etendrait entre quelques dizaines de kcallh seulement 'pour une difference de tem-
perature ne depassant pas 30 a 40? C.
Les installations electric-thermiques de pompes a chaleur sont commodes aussi
pour les puissances elevees. En utilisant le chauffage electro-thermique an lieu du
chau f f age electrique direct on peat realiser une economic de courant electrique im-
portante. Cependant 1'application de l'ef fet electro-thermique n'est guere justifie
avec une dif ferense de temperature depassant 40 a 50? C.
GENERAL PRINCIPLES, COMPARISON WITH THE USUAL
REFRIGERATING METHODS
Introduction of thermoelectrical refrigeration into practice requires preliminary
study of its advantages and disadvantages as compared with ordinary refrigeration.
Thermoelectrical installations are absolutely noiseless. They are devoid of moving
parts, have no need for special refrigerating agents, are easy to control and may,
with extraordinary ease, be transformed from refrigerating installations into heat
pumps. The ease of conversion to a heat pump opens up a promising field for the
use of thermoelectrical refrigeration in various types of installations requiring
maintenance of temperature on a given level (above or below the ambient).
However, although the power efficiency of thermoelectrical cooling has greatly
increased with the transition from metal's to semi-conductors, it is still far behind
mechanical and absorption units (the latter in the case of heat but not electrical
energy consumption).
At this time energy consumption is the predominant factor in the choice of a
refrigeration system. For very small capacities of the order of a few score kcal/hr
the energy consumption no longer plays a major part and other factors as, for
instance, initial cost, portability, reliable operation, convenience in control, etc.
characteristic of semi-conductor arrangements come to the fore.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
10 20 30 40 50 60 70 Q "
hr
Fig. 1. Efficiency of various refrigerating Fig. 2. The effect of heat exchange on the
systems for domestic cabinets. 1 - mechani- energy characteristics of semi-conductor ther-
cal, 2 - absorption, 3 - thermoelectrical. mal batteries.
We made comparative calculations based on test data for a large number of
domestic refrigerators with mechanical and absorption systems. Fig. 1 shows
the relation between the real coefficient of prformance and the refrigerating
capacity Q0 (of refrigerator without load), the real coefficient of performance
for the semi-conductor refrigerator being obtained by calculation, using the semi-
conductor materials we obtained from the Institute of Semiconductors of the
Academy of Sciences of the-U.S.S.R.
The value of Z for a semi-conductor thermocouple is related to the thermo-
electric forces (e), specific thermal conductivities (d) and specific resistances (Q)
of the arms as follows: '
Z = (el + e2)2
(1)
(VQ1 Al + 6;A2 The analysis led to the following conclusions. The present semi-conductor
refrigerator will consume several times as much energy as the mechanical type.
With regard to its energy characteristics it is approximately on a level with an
absorption refrigerator with electrical heating. The latter is about three times
lower in efficiency than absorption systems utilizing heat energy.
Obviously for large capacity refrigerators (and still more for industrial plants),
semi-conductor units are as yet unable to compete with mechanical and absorption
types. The energy characteristics of semi-conductor systems may be improved on
employing them as heat pumps, particularly for evaporating plants. This is due
to two factors: the relatively small temperature difference and the higher tem-
perature of the low temperature heat source in heat pumps than in refrigerating
plants. Owing to these two factors, thermoelectric losses diminish and the energy
efficiency rises.
Naturally, the search for new semi-conductor materials with high Z values
assumes paramount significance.
It should be noted that the existing method of calculating semi-conductor
thermal batteries for refrigerating plants and heat pumps requires further develop-
ment taking into account the influence of the heat exchange of the thermocouple -
junctions with the surrounding medium. To determine the heat and energy cha-
racteristics of a semi-conductor evaporating plant and to check the rating for-
mulae, a test was made of a small model of up to 150 kcal/hr capacity.
CORRECTED METHOD FOR CALCULATING SEMI-CONDUCTOR
BATTERIES
- The classical theory of thermoelectrical cooling and heating gives formulae for
calculation based upon the temperatures of the junctions T and T0. However,
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
these temperatures do not remain constant on change of the heat flow in the
thermocouple. In designing refrigerating and heat pump plants the temperatures
of the media supplying and receiving heat (To' and T') and the heat transfer
coefficients (ao and a) are usually known. Therefore, of practical interest are
rating formulae depending upon these quantities.
The densities of the heat flows at the cold and hot thermocouple junctions are
determined by the following formulae
1 A
2qo=eToi'2i 0l-1 (T-To)
(2)
R
2 q = eTi + #i2 01-! (T _ To)
(3)
The temperatures of the thermocouple junctions may be determined from the
general equation for heat transmission:
qo -o (To' TO),
q =a (T- Ti)
(4)
(5)
Solving the systems of equations (2, 3, 4, 5) and taking into account that the
coefficient of performance equals
(6)
q-qo
(el+1l I1+
+
2-a
2a/
(7)
el \1 2a/ 21a
In these formulae l is the length of the thermocouple and i the density of the
current passing through it.
The quantity e' is determined from expression (6), T' and T'0 being substituted
for T and To in the formulae for qo and q.
It should be noted that for ao N and a N the formula for a coincides
with the corresponding formula in the theory of thermoelectrical cooling and
heating. In that case a -, e'.
Calculations show that in order to obtain a values sufficiently approaching a'
the heat transfer coefficients should be high. For instance, for a semi-conductor
thermocouple 1 cm in length operating under the temperature conditions of a
domestic refrigerator, the value of a and ao should be measured in thousands of
kcal/m2hr?C. Such values of heat transfer coefficients under conditions of natural
convection are made possible only by extensive finning of the thermocouple junc-
tions. In experimental semi-conductor refrigerators the finning coefficient attained
the enormous values of 400 to 600!
In Fig. 2 the relation is shown between the performance coefficient of the
thermobattery and the heat exchange conditions for the case when a = a0,
.1 = 1 cm, To' = 273? K
1
T1=303 OK, Z=1,85.10-3 oK
Characteristic is the fact that the optimum current density is independent of
a and ao.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
The heat flow densities diminish with increase in thermocouple length, im-
proving the conditions for heat exchange. For constant heat transfer coefficients
Fig. 3. Schematic diagram of heat pump
evaporating plants. A - mechanical, B -
ejector, C - semi-conductor. 1 - liquid in-
let, 2 - condensate outlet, 3 - excess liquid
outlet, 4 - vapour outlet, 5 - compressed
vapour, 6 - evaporator, 7 - compressor, 8 -
ejector, 9 vapour outlet tube, 10 - thermo-
battery, 11 - condenser, 12 - starting elec-
trical heater.
Fig. 4. Electrical measurement circuit of ex-
perimental plant. 1 - boiler, 2 - thermostat,
3, 5 - thermometers, 4 - electrical filter, 6 -
thermobattery, 7 - rectifier, 8 - thermostat
heater, 9 - transformer, 10 - temperature
relay, 11 - Dewar flask, 12 -potentiometer,
13 - exterior galvanometer, 14 - switch.
the increase in length of a thermocouple, operating in a refrigerating unit, increa-
ses the value of the coefficient of performance.
It should be borne in mind in the calculations that under conditions of natural
convection the effect of the heat flows upon the values of a and ao may be
neglected while on boiling and condensation these relations may be easily deter-
mined with the aid of the usual heat transfer formulae.
For small current densities and intensive heat exchange the quantities _ e:
2a
and 2a are negligible. Formula (7) in that case is considerably simplified. Then
u
the optimum current density corresponding to emax is determined by the formula:
e (TI - T'1)
(Vi+ T1 2 T?1 Z -1) Pi
(8)
In the particular case of calculations for an evaporator plant, when T' = T'?
extreme values for e are absent altogether. The basic principles of this method of
calculation were checked experimentally in a study of the thermoelectrical eva-
porator plant.
EXPERIMENTAL THERMOELECTRICAL EVAPORATOR PLANT
In heat pump evaporator plants the secondary vapour is compressed in the
compressor or ejector (Fig. 3). The condensation temperature of the secondary
vapour then increases and the heat of condensation may be used as heat source
for vaporizing the liquid in the evaporator.
Mechanical evaporator plants have high conversion coefficients ((p), but they
are complicated as regards both equipment and operation. Ejector plants possessing
somewhat inferior energy characteristics require vapour of comparatively high
parameters.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-34
The substitution of a compressor or ejector by a semi-conductor battery makes
it possible to construct a plant outstanding in the simplicity of its design, noise-
free operation, compactness and flexibility of control. Moreover, the semi-con-
ductor thermocompressor allows change in temperature of the secondary vapour
without change in pressure. In this case as heat source, use is made of the heat
of condensation of the evaporating liquid, permitting considerable elevation in
the temperature of the cold junctions of the thermocouples and maintenance of
this temperature on a level close to that of the condensing vapour at the corre-
sponding pressure.
The schematic diagram of the experimental thermoelectrical evaporator plant
is given in Fig. 3 and the electrical measurement circuits in Fig. 4.
The evaporating liquid, heated to the boiling temperature in the regenerator
enters the evaporator where it boils, taking up the heat from the hot junctions of
the thermobattery. The vapour formed passes through a vapour outlet to the
condenser where it condenses on the cold junctions. In the exchanger heat ex-
change takes place between the outflowing condensate and the entering liquid.
The effect of scale formation was studied by coating the hot junctions of the
thermobattery with layers of different thickness of a cement and determining the
thermal resistance
In Fig. 5 are produced the energy characteristics representing the dependence
of the conversion coefficient qp on the current density i (heat load) and the thermal
resistance
The curves show that the conversion coefficient of a thermoelectrical evapor-
ating plant increases continuously with fall in load exhibiting no extremes. The rise
in q is due to decrease in temperature difference between the junctions of the
thermobattery on fall in current density.
Scale formation markedly impairs the energy characteristics of the plant. For
example, at i = 16 a/cm2 the formation of scale with a thermal resistance
c
wt
Fig. 5. Energy characteristics of thermoelectrical evaporating plant.
= 0.75 - 10-3 m2hr?C./kcal leads to a fall in p from 4.5 at ~ = 0 to 3.5 i.e.,
by 22%. The relative influence of the scale increases with decrease in load. At
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
7 a/cm2 an increase in from 0 to 0.75 ? 10-3m2hr?C./kcal lowers rp by
30%.
The energy characteristics of a thermoelectrical evaporating plant are given
by the approximate equation derived from formula (7) on neglecting the terms
ei ei
2aand--
A (1 1\
1+21 a
0
eT 1
ml L91i + 2
(9)
(10)
Both calculation and experiment show the absence in evaporating plants of
extremes for rp with respect to the current density i.
The circumstance bars the choice of optimum current density and hence of
optimum heat capacity of the plant on the basis of only the energy characteristics.
To solve this question one must also possess data concerning the dependence of
heat capacity on the current. Moreover, the technological and economic aspects
of the plant have to be taken into account.
Computation showed that, e. g., for X = 0.75 ? 10-3m2hr?C./kcal the opti-
mum value of (p is 4%nt = 5, iet,t being 10 a/cm2 and AT?pt = 100 C.
Fig. 6. Coefficient of transformation of various types of evaporating plants. 1 - mechanical,
2 - ejector, 3 - thermoelectrical, 4 - with electrical heater.
Fig. 6 shows the comparative energy characteristics of various types of evapor-
ating apparatuses. An examination of the curves shows that in this aspect the
semiconductor evaporating plant is still below compressor or ejector installations.
For all values of AT the conversion coefficient p of mechanical units is ca.
40% higher than for semiconductor installations: For ejector units the value of p
exceeds the semiconductor type only by 11% for AT = 5? C. and by 20% for
AT = 15? C.
As compared with direct electrical heating the semiconductor evaporating unit
gives considerable power economy. For example, at AT = 10? C. the energy
consumption decreases five-fold and at AT = 7? C. seven-fold.
Behrndt bogtryk, Denmark Copyright reserved
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-28
The Volumetric Efficiencies of Medium Capacity Refrigerating Com-
pressors
Le rendement volumetrique de compresseurs frigorifiques de diverses puis-
sances
V. F. CHAIKOVSKY, A. A. SHMIGLYA, K. I. SAVKOV,
The Odessa Technological Institute of the Food and Refrigerating Industry,
Odessa, U.S.S.R.
SOMMAIRE. Le rendement volumetrique d'un compresseur frigorifique depend
de sa conception, du fluide frigorigene. utilise, du cycle de fonctionnement et_ de la.
puissance de la machine. Des recherches poussees of fectuees sur divers compresseurs
ont permis de determiner certaines constantes daps la variation de Ia valeur du
rendement' volumetrique et ses elements.
Des experiences realisees par V.V. Lavrova (Institut de Recherches Scientifiques de
l'lndustrie Frigorifique d'U.R.S.S.) ont prduve que le volume de l'espace nuisible,
qui est represents par le facteur ,\, est de la plus haute importance daps les com-
presseurs a ammoniac a grande vitesse modernes, avec une vitesse de deplacement
de Ia sonpape ne depassant pas 40 m/sec.
Des rcherches experimentales sur des compresseurs a Freon de puissance moy-
enne (ale sage 100 mm et course 80 mm) ont montre qu'a 970 t mn les pertes volu-
metriques. incluses daps le facteur Al et provoquees par etranglement, chauffage et
fuites, prennent Ia plus grande importance. Les echanges de chaleur ne doivent
pas titre negliges, meme a de grande vitesses, oh la periode diminue, le fluide frigo-
rigene est en contact avec les parois du cylindre, ce qui n'est pas le cat pour let
petits cylindres oh le poids du fluide frigorigene par unite de surface de la paroi est
trey faible. Dans let compresseurs a ammoniac refroidis par air de_.100X 100 mm
et de 960 t/mn, Ia temperature an debut de Ia compression est voisine de la tempe-
rature de condensation et pour des valeurs elevees de Pc, P. est encore plus eleve.
Avec un accroissement de Ia vitesse de rotation du compresseur, les pertes volu-
metriques dues a I'etranglement de la vapeur daps let soupapes sont elevies, en
particulier daps let compresseurs a Freon. En raison du perfectionnement de Ia
conception des sonpapes de compresseurs frigorifiques et de Ia reduction du role.
de l'espace nuisible, l'influence de Al devient de plus importante.
Par exemple, daps un compresseur a Freon de 100 X 80 mm, l'espace nuisible avec
les nouvelles soupapes utilisees est de C = 1,94%. Avec la temperature d'evapo-
ration to = -15? C., la temperature de condensation.tc = 30? C. et la temperature
des vapeurs surchauffees t8 = 20? C., on atteint les valeurs suivantes A = 0.790,
Ac = 0.930 et Al = 0.849.
D'apres les resultats des recherches, 1'augmentation du rendement volumetrique
A = Ac ' ,t daps les compresseurs a Freon de puissance moyenne dolt titre obtenue
surtout par une augmentation de At.
The volumetric losses of reciprocating compressors are evaluated by the volu-
metric efficiency ,k which may be expressed as
A=Ac'A~
where ~? - is the clearance volumetric efficiency;
Ai - the factor considering the losses in the compressor due to vapour thrott-
ling in the valves, to heat exchange between the vapour and cylinder
walls and to leaks.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-28
Results of experiments on the volumetric efficiencies of two medium capacity
reciprocating Freon-12 compressors are given below.
The following compressors were tested: FV-12 (100 x 80 mm, 980 rpm.) - a
vertical, monoblock, two-cylinder, uniflow machine; FU-8 (67.5 X t 50 mm,
1300 rpm.) - a V-type, four-cylinder, non-uniflow machine with two removable
blocks, the angle between the cylinder centre lines being 90?.
A diagram of the experimental installation is given in Fig. 1.
a40
Fig. 1. Schematic diagram of the test installation. 1. Compressor; 2. Condenser; 3. Heat-
exchanger; 4. Liquid strainer; 5. Peep hole; 6.' Expansion valve; 7. Electric calorimeter.
The refrigerating capacity was determined by means of an electric calorimeter
with a secondary medium. The tests were carried out in accordance with existing
rules. Measurements were made with very precise instruments.
The FV-12 compressor was tested at a condensing temperature of t? = +30? C.,
evaporating temperature of t? = -15? C., a superheat of the vapour on the low side
Ate = 20? C. and at various values of the clearance volume (from 5.8 to 1.94%).
Variations of the clearance volume were obtained by using compressor valves of
various design.
Besides, tests were carried out on a FV-12 compressor, having a 1.94% clearance
volume at t? = +35? C. and-various evaporating temperatures.
2400
2300
2200
2100
2000
1x000
17000
16000
15000
14000
13000
. ? `/. 1 2 3 Al S C7
Fig. 2. Results of testing the FV-12 compressor at various values of the clearance volume
C% and at to = -15?'C., t? = +30? C. and Ats = 20? C.
a) Values of the volumetric coefficients.
b) The refrigerating capacity Qo and specific refrigerating effect KS.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
_
The FU-8 compressor was tested. at various evaporating temperatures and to
+500 C. The clearance volume of the compressor was equal to 4.66%.
The clearance volumetric efficiency A0 was obtained by calculation. The exponent
of the polytropic reexpansion curve m was assumed approximately equal to unity.
X200
20000
16000 -
12000
!N 1000 -
Fig. 3. Results of testing the FV-12 compressor'at to = 35? C. and C = 1.94%.
a) Values of volumetric coefficients.
b) The refrigerating capacity Qo and specific refrigerating effect K8.
As is known, at a low (up to 20? C.) superheat of the vapour sucked by the
compressor the volumetric efficiency is considerably affected by the evaporation
of the freon from the oil returned into the compressor, as well as by the cyclic
solubility of freon in the oil deposited upon the internal surfaces of the cylinder,
and also by the possible condensation of freon on the walls having a temperature
below the dew point.
The superheat of the vapour on suction in the tests was not less than 20? C.
The influence of the evaporation of freon from the oil on the volumetric efficiency
could not be essential under these conditions. The other factors are combined in 2i
due to the difficulties in their quantitative account.
4V
!3
fa
if
/0
0 ! 2 6 C
Fig. 4. The value Y' for the FV-12 compressor at to = -15? C., to = + 30? C. and
Qts = 20? C.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
The results of testing the FV-12 compressor are given in Figs. 2 and 3. Fig. 2
shows that with an increase of the clearance volume of the compressor, the clearance
volumetric efficiency 2, as well as ).I are reduced. The capacity of the compressor
depends largely upon the value of 21.
Figs. 2 and 3 illustrate that tthe values of 2c are higher than those of 1i. The
value of the volumetric losses 21 especially increases with the utilization of modern
valves having small clearance volumes. To determine the influence of the clearance
volume on the volumetric efficiency of the compressor, the change of the value W
(Fig. 4) has been investigated.
1 dA 100. 0
A do /0
The graph (Fig. 2) demonstrates that the specific (per one per cent of clearance
volume) change of the volumetric efficiency, depends upon the value of the
clearance volume.
For example, an increase of the clearance volume from 2 to 3 % causes a decrease
of the volumetric efficiency by approximately 0.04 of its value, while with the
clearance volume increased from 5 to 6 % the value of A - is lessened by nearly
0.11 of its value.
Fig. 2 shows that the values of the specific refrigerating effect K" kcallkw-hr
reduce at an increase of the clearance volume. This is explained by the fact that at a
decrease of the indicated work with a larger clearance volume, the relative value
of the indicator losses of the compressor and of the friction losses is raised. An.
increase in the absolute value of the indicator losses is caused in this case by the
irreversibility of the processes of compression and expansion of the refrigerant in
the clearance volume.
During the tests the compressor was operated with group disc-plate valves with
1.94 % clearance volume the linear clearance included. The area of these valves was
equal to that of 4.5 % clearance valves. The suction valve had flanges entering the
apertures in the seat of the discharge valve when in the upper end point.
The linear clearance of the FV-12 compressor is 0.5 - 0.8 mm
When assuming that the FV-12 compressor has only a 0.8 rpm linear clearance,
the minimum value of the clearance volume would amount to 1%.
An attempt to reduce the clearance volume of the compressor from 2 to 1 %
would cause an insignificant increase of the volumetric efficiency while the design-
ing of valves with quite a small clearance volume entails great difficulties.
It should also be taken into consideration that the presence of the clearance
volume has a positive influence upon the dynamic characteristics of the compres-
sor which is especially important for high-speed machines.
All this makes it possible to conclude that no significant advantages can be achieved
in medium capacity compressors by a decrease of the clearance volume below 2%.
Fig. 2 illustrates that at a small clearance volume the curves of the volumetric
coefficients of the compressor as a function of the relation P,/Po run more steeply
than in the case of large clearance volumes.
Valves with small clearance volumes are especially advantageous at large values
of the pressure relation.
Results of testing the FU-8 compressor are illustrated in Fig. 5. As it was in the
case of testing the FV-12 compressor, the values of ke are lower than of k?.
The present paper deals with the investigation of the clearance volumetric effi-
ciencies of freon compressors. It is though interesting to compare the results obtained
with those of testing ammonia compressors of medium capacity.
Tests of the ammonia monoblock, two-cylinder, uniflow, air-cooled compressor
(80 X 80 mm, 720 and 980 rpm.) have also shown that the value of 11 is lower
than that of X?.
For example,' at P?JP? = 3.44 and n = 980 rpm., A = 0.718, A? = 0.867, and
21 = 0.828.
At to = -15? C, t? = + 25? C. and n = 980 rpm. the volumetric efficiency
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
3-28
A = 0.672, and Aw = 0.847 - heating up coefficient, The low values of Al for this
compressor are explained first of all by the active heat exchange inside the cylinder.
too 3200
/
Fig. 5. Results of testing the FU-8 compressor at to = 50? C. C = 4.66%.
a) Values of the volumetric coefficients.
b) The refrigerating capacity Qo and specific refrigerating effect Ks.
The availability of intensive heat exchange is proved also by the values of the
cylinder wall temperatures.
Table 1 presents data on the wall temperatures measured at three points on the
cylinder (n = 720 rpm.).
TABLE 1. WALL TEMPERATURES MEASURED AT 3 POINTS ON THE CYLINDER
(n = 720 rpm.)
-10.0
-20.0
Temperature in the compressor suction pipe, ?C.
2.0
-7.8
Temperature in the compressor discharge pipe, ?C.
86.9
103.9
Mean temperature of cylinder wall:
in upper point of cylinder, ?C.
74.8
96.3
at point corresponding to half-stroke of piston, ?C.
58.3
74.9
in lower point of cylinder, ?C.
42.3
55.8
Parameters Evaporating temperature ?C.
The temperature of the cylinder walls greatly depends; upon the friction. At idle
running of the compressor the temperature of the cylinder walls was by 30? C.
higher than that of the surroundings.
At the same time experimental data obtained by a number of authors show that
in water-cooled medium capacity ammonia compressors the volue of Al is higher
than X.
The following conclusions may be drawn from the experimental data:
1. In medium capacity air-cooled Freon and ammonia compressors the values of
2,; are higher than those of 21.
2. The decrease of the volumetric efficiency of medium capacity Freon compres-
sors at an increase of the clearance volume is non-uniform. The larger the absolute
value of the clearance volume, the greater its relative influence upon the volumetric
efficiency.
3. In the case of medium capacity Freon machines a decrease of the clearance
volume below 2 per cent is inexpedient. Such a decrease may be useful only in the
case of low evaporating temperatures obtained at single stage compression.
Behrndt bogtryk, Denmark Copyright reserved
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
Appareil pour la detection et la telesignalisation des fuites d'ammoniac
dans les chambres frigorifiques
V. BERCESCU, A. CIOBANU et C. MIHAILOPOL,
Institut de Recherches Alimentaires, Bucarest, Roumanie
SUMMARY. The greater use of direct expansion refrigeration requires more
efficient measures to protect the products stored against damage from ammonia
leaks. A device based on colour change in presence of traces of ammonia in a filter-
paper strip soaked in an indicator was constructed for detecting ammonia leaks.
The colour change acts on a photoelectric cell, then an electronic apparatus am-
plifies the variation of the photoelectric current and transmits, the impulse to a
visual or sound signalling device in a machine room. In automatic plants, the de-
tector may block the automatic valve supplying ammonia from refrigerating units.
Les aliments entreposes dans les chambres frigorifiques peuvent etre deprecies
en raison des fuites eventuelles d'ammoniac provenant des elements refrigerants, ce
qui necessite leur detection immediate.
La. litterature technique indique des appareils electroniques sensibles aux trans-
formations de la resistance electrique de l'air'qui se produit en presence des com-
poses hallogenes (freons, chlorure de methyle).
L'appareil que noun presentons dans cette note, a comme principe le changement
de couleur d'une surface imbibe_e par une solution indicatrice tres sensible'aux modi-
fications du pH du milieu.
Le changement de couleur agit sur une cellule photoelectrique, dont le courant
est ensuite amplifie par un appareil electronique qui transmet l'impulsion a un
dispositif de signalisation optique ou acoustique, place dans la salle des machines.
Fig. 1. Le principe du fonctionnement de l'appareil. 1. Bande de ;papier-filtre avec indi-
cateur, impregnee de chlorure de calcium CaC12. 2. Surface etalon. 3-4. Cellules photo-
electriques. 5. Source de lumiere. 6. Ecran reflecteur. 7. Vcran separateur qui empeche
l'influence des surfaces comparees sur les cellules voisines. 8. Dispositif de reglage pour
la mise de 1'appareil au point .
Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
DESCRIPTION ET FONCTIONNEMENT DE L'APPAREIL
La bande sensible 1 est en papier-filtre et est imbibee d'une solution indicatrice de
bromophenol bleu ou de -rouge de phenol. Ensuite, elle est impregnee de chlorure
de calcium, qui, grace a sa propriete hygroscopique, absorbe l'humidite atmosphe-
rique et favorise la dissolution des vapeurs d'ammoniac.
Approved For Release 2009/05/01 : CIA-RDP80T00246A007500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
La lumiere fournie par la source 5 est reflechie par la bande sensible 1 vers la
cellule photoelectrique 3, et par la surface etalon 2, vers la cellule photoelectrique 4.
Dans i'absence des traces d'ammoniac, les courants photoelectriques resultants
sont egaux et de sens contraire, c'est-a-dire que leur somme est nulle. Pour ?mettre
l'appareil au point ozerou, on emploie le bouton de reglage 8 qui, agissant sur
1'ecran reflecteur 6 change l'illumination des surfaces comparees, de maniere que
les courants resultants dans les deux cellules photoelectriques soient egaux et de
lens contraire.
Lorsqu'il existe dans l'atmosphere des vapeurs d'ammoniac, la bande 1 change
de couleur et determine la modification du courant dans la cellule 3. Par suite, la
resultante des courants des deux cellules photoelectriques, aura cette fois-ci une
valeur determinee, differente de zero, par rapport a la concentration de 1'ammoniac
situe dans l'atmosphere.
En amplifiant le courant qui resulte, on peut actionner un relais qui transmet a
distance (dans la salle des machines) les signalisations d'alarme optiques et acousti-
ques; dans les installations automatiques, le courant peut actionner directement les
vannes automatiques qui fermeront faeces de dammoniac a la chambre frigorifique.
Pour prolonger la duree de l'utilisation des cellules photoelectriques, 1'appareil est
pourvu d'un moteur synchrone generateur d'impulsions, qui determine le fonctionne-
ment periodique et a des intervalles egaux du transformateur d'alimentation de la
source lumineuse 5 et des anodes.
A la reception de chaque nouvelle impulsion, la transmission d'un signal optique
dans la Salle des machines indique le fonctionnement de Yappareil. Le signal cesse
peu de temps apres s'il n'y a pas d'ammoniac dans l'atmosphere, mais it reste
accouple dans le cas contraire.
Le schema electrique du Principe de l'appareil est montre a la Fig. 2.
Les experiences ont demontre que 1'appareil a une sensibilite de l'ordre de
0,0000066 vol. NH3%.
En resume, on peut conclure que cet appareil presente les avantages suivants:
- it empeche la depreciation des aliments entreposes dans les chambres frigorifi-
ques;
- it reduit les pertes d'amimoniac dans les installations frigorifiques;
- it assure un controle automatique permanent de 1'atmosphere sous l'aspect de
la securite du travail daps les entrep6ts frigorifiques et aussi dans les autres indu-
stries qui produisent ou utilisent de 1'ammoniac.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
The Reason for the Invalidity of the Lewis Ratio in the Case of Air
Washers
Raison de l'inapplicabilite du rapport de Lewis dans le cas des laveurs d'air
A. A. GOGOLIN
The Scientific Research Institute of the Refrigerating Industry of the U.S.S.R.
Moscow, U.S.S.R.
SOMMAIRE. Le rapport du coefficient d'echange de chaleur seche par convection
au coefficient d'echange d'humidite, connu sous le nom de rapport de Lewis, est
tras important du point de vue de la theorie de l'echange de chaleur entre l'eau et
1'air sur laquelle reposent les principes du conditionnement d'air. De nombreuses
recherches ont montre que daps let conditionneurs a ecoulement et a pulverisation
le rapport de Lewis est tres superieur a la valeur theorique. On a exprime a plusieurs
reprises l'idee que le rapport n'est generalement pas valable pour les laveurs d'air.
Des experiences of fectuees par l'A. sur divers types de laveurs d'air (a ecoulement
ou a pulverisation) montrent que la valeur la plus elevee du rapport de Lewis est
due a une valeur superieure a la valeur theorique de la teneur en humidite de fair
s'ecoulant a l'exterieur. Cette difference est causee par l'humidification supplemen-
taire des endroits arrows seulement de facon sporadique et de la surface de petites
gouttes qui se sont rechauf fees i la temperature bulbe humide de l'air.
La preuve en est donnee daps la dependance etablie entre le rapport de Lewis et
la difference de temperature psychrometrique a la fin de la variation theorique de
l'etat de fair daps les conditionnei rs a contre-courant.
L'augmentation de la dispersion des gouttes entrat"ne une elevation du rapport de
Lewis. Cela prouve que le role le plus important est joue ici par l'humidite supple-
mentaire plutot que par la difference entre la temperature superficielle des goutte-
lettes et la temperature moyenne de l'eau qui est plus importante pour une vapori-
sation de plus grosses gouttelettes.
Le rapport de Lewis est specialement eleve pour de faibles coefficients de conden-
sation d'humidite Q totait (! _ pour lesquels le coefficient d'echange d'humidite
Q seche )
total diminue en raison de 1'existance de deux phenomenes inverses: la deshydrata-
tion et l'humidification de fair. Pour les coefficients de condensation eleves
1,5 - 2) le rapport de Lewis Arend la valeur theorique.
The ratio of the coefficient of convective heat exchange a kcal/m2hr?C. to the
moisture exchange coefficient b kg/m2hr, known as the Lewis ratio, is highly
important for the theory of joint heat and moisture exchange which is the basis
of the entire system of air conditioning calculations.
Lewis (1) obtained in 1922 the following, equation based upon theoretical
premises:
where: C'p - is the specific heat of moist air at constant pressure.
Merkel [2] developed a theory of calculating joint heat and moisture exchange
on the basis of the Lewis equation. and was the first to check experimentally this
equation in a small stream cooling tower. However, the ratio of b obtained in
these experiments amounted to 0.3 - 1 which considerably exceeded the ordinary
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
value of C'p = 0.24 0.25 kcal/kg?C. This divergence is explained by the fact
that the wetted surface in the apparatus used by Merkel in his experiments con-
sisted of metal Raschig rings. Due to the high thermal conductivity of the metal
convective, heat exchange took place also on the unwetted part of the rings, i. e.
on an area exceeding the area of moisture exchange. This caused higher sensible
heat exchange as compared with moisture exchange, and enlarged values of the
However, this ratio b considerably exceeded its theoretical value also in the
subsequent experiments [3, 61 carried out with both spray and stream air coolers
with a packing of a low thermal conductivity.
Thus an opinion arose that the Lewis equation was not valid in air washers on
account of the absence of analogy between the heat and mass-exchange due to the
additional heat flow at the diffusion of water vapour, thermal diffusion etc.
The influence of all these factors makes the Lewis equation approximate. Its
validity must be checked by experiments carried out on real size apparatus with
the hydraulic and climatic operating conditions of the 'experimental unit varied
in a wide range. Such experiments were run some time ago by the author at the
Scientific Research Institute of the Refrigerating Industry of the U.S.S.R. Tests
[3] were carried out on various stream and spray type vertical air flow counter
current air coolers packed with porcelain Rashig rings. A conventional spray type
horizontal air flow air washer with several rows of spray nozzles was also inve-
stigated [4]. All the tested air coolers were quite large with an air flow area from
1.7 to 2.4 m2.
The counter flow air washers were tested on air cooling and dehumidityfying.
In the conventional type spray air washer experiments were carried out under
the operation conditions characteristic of cooling, dehumidification, humidification
or heating of air.
1 2
it O - B 1 3 '!
4
e, =177t
Fig 1. The change of the state of the air in a counter current air washer with Rashig ring
packing. twi, tw2 and tm - water temperature initial, final and mean values, ?C; ti and
t2 - initial and final temperatures of air, ?C; iiand i2 - initial and final enthalpy of air,
kcal/kg; t,,2 - final wet-bulb temperature of air, ?C.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
On analysing the results of the tests it was noticed that an increased value of
the ratio b was observed in air cooling and dehumidifying tests in the case of the
state of the escaping air considerably deviating'towards a higher specific humidity.
An illustration of such a deviation is given in Fig. 1. Fig. 1 also shows the stepped
plotting of the theoretical process of the air state variation which may be made
for a counter current air-cooler considering the Lewis equation true.
With such a process the point corresponding to the state of the leaving air in
counter current air coolers is approximately on the straight line connecting on
the enthalpy - specific humidity diagram point A of the initial state of the air
with point F on the saturation line. In this case the temperature at point F is equal
to the mean temperature of the water t111.
The deviation of the process from the line AF observed in the experiments was
explained by two reasons. It was supposed [51 that this deviation resulted from
the difference between the temperature at the surface of the droplets and the mean
temperature of the water after leaving the air washer. There certainly is such a
temperature difference but it cannot be the sole cause of all phenomena observed
during the experiments, which is discussed below.
According to another explanation given by the author, the reason for the
deviation observed is in the fact that there ordinarily are two different moisture
exchange processes in air washers: dehumidification of the air on the main surface
of the water and its humidification on the part of surface heated by the air flow
to the wet-bulb temperature.
In stream air washers provided with Rashig rings humidification occurs on a
part of the surface of the rings only sporadically wetted by water splashes. The
thin water film wetting this surface is quickly heated up to the wet-bulb tem-
perature and, while evaporating, it humidifies the air. The droplets precipitating
on the surface of the moisture eliminator act in an analogous manner.
In spray air washers the-air is humidified on the surface of fine droplets heated
to the wet-bulb temperature, and on the wetted surface of the eliminators.
The humidifying surface is usually located at the end of the air passage (the
upper surface of the wetted Rashig rings, minor droplets carried by the air flow
to the end of the apparatus, eliminators).
Therefore, the process of treating air in an air washer may be conventionally
divided into two: air. cooling and dehumidification (line A-B, Fig. 1) and air
humidification running approximately along the isoenthalpic line (B-C). Certainly,
this division is somewhat schematic and on a certain part of the surface both
processes go on practically simultaneously. If such an assumption is true the degree
of deviation of the state of the escaping air towards a higher specific humidity
and the corresponding increase in the value of the ratio b should depend upon
the psychrometric temperature difference at the end of the first process (point B
in Fig. 1) which determines the intensity of the subsequent humidification process.
Such a dependence was obtained on the basis of the results of the experiments
with counter current spray type air coolers (Fig. 2). At minor psychrometric
a
temperature difference the ratio b approaches the theoretical value (i 0.245)
while at greater temperature differences it increases up to 0.35-0.4.
An analogous dependence was obtained also for air coolers with the Rashig
rings.
The curves in Fig. 2 show that the ratiob increases with a rise of the pres-
sure, i. e. at a finer dispersion of the water.
This is also a proof of the fact that the deviation of the state of the air towards
an increase in the specific humidity is resulted by its humidification which is higher
at a finer dispersion of the water. The influence of the difference between the
surface temperature of the droplets and their mean temperature is, apparently,
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Fig. 2. The dependence of the Lewis ratio upon the psychrometric temperature difference at
the end of the theoretical process for counter current spray type air washers. Water pressure
before nozzles: 1-1.2-1.8 atm; 2-0.3-0.6 atm.
1 21
a / 2 d 4
5 t.,to 'C
insignificant. Otherwise, the finer the spray, the lower would be the value of
a
bwhich is not the case. Besides, the ratio should in this case depend upon the cha
of the water temperature, but attempts to detect such a dependence were a failure.
It is interesting that the coefficient of moisture exchange, calculated on base of
the mean logarithmic difference of enthalpy, does not depend upon the climatic
conditions. What concerns the heat exchange coefficient is that it increases with
the growth of the ratioo a as there is in this case an additional convective heat
exchange occuring without any change of the enthalpy of the air (line B-C, Fig.
1). The relation of Sand the moisture condensation coefficient ~ (total-sensible
heat ratio) was also determined in the experiments.
The ratio b is at its largest value at small moisture condensation coefficients
when the humidification process is more intensive owing, to higher temperatures
of the water.
Vice versa, with moisture condensation coefficients amounting to ca. 2, when the
process of the change in the state of air approaches the line cp = 1 at a very small
angle, the ratio is close to its theoretical value of 0.24-0.25. A corresponding
dependence for a conventional spray type horizontal air flow air washer is given
in Fig. 3 which illustrates the value of -b not only for cooling and humidification
(i; > 1) but also for cooling and humidification of air (0 < ~ < 1), as well as
equilibrium humidification along the constant wet-bulb temperature line (~ ~' 0)
and results of several experiments at heating and humidifying air.
Given in Fig. 3 are also results of experiments carried out at the Scientific
Research Institute of Water Supply and Heating Installations (NIIST) of the
Academy of Architecture by E. E. Karpis and M. L. Zusmanowich at = 2.2 - 3.
It is seen from Fig. 3 that 0 and a
g? 2 - 3 the ratio b coincides almost ex-
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
A
I
~
as
2
Q4
it
3
~I
i
c;4
rs5
.~i
n~
Fig. 3. The dependence of the Lewis ratio upon themoisture condensation coefficient in con-
ventional spray type air washers. I. Cooling and dehumidification of air. 11. Cooling and
humidification of air. III. Increase of enthalpy of air and its humidification. IV. Increase
of air temperature and its humidification. 1- experiments by the author on cooling air; 2 -
experiments by the author on humidifying air; 3 - experiments by the author on heating
and humidifying air; 4 - experiments by Karpis and Zusmanovich (NIIST).
7 9 9 A7 0 A! -45 O 45 10 15 40 45 J0 5
actly with the theoretical value cited by Lewis. Close to the theoretical value is b
also in the process of heating and humidification of the air.
Characteristic of all these ranges is the absence of duality of the moisture
exchange processes. At equilibrium humidification of air ( 0) the temperature
of the water temperature cannot change the incline of the line of the air process
air flow. The process represented in the enthalpy-specific humidity diagram by the
line of constant wet bulb temperature goes on in all points of the water surface.
In the case of cooling and dehumidification the air at high coefficients of
moisture condensation ( = 2 -- 3) as well as at heating and humidification,
the -temperature of the water is certainly not constant.
However, as the line of the change in state of the air in the enthalpy-specific
humidity diagram approaches the saturation line at a very small angle, a variation
of the wter temperature can not change the incline of the line of the air process
in the apparatus. Therefore, the surface of the water may conditionally be assumed
as 'isothermal' in the sense of its temperature not exerting any influence upon
the direction of the air state variation. Thus we may consider the validity of the
Lewis equation in air washers proved in cases when a certain moisture exchange
process takes place throughout the entire water surface.
The process with high coefficients of moisture condensation is typical in the air
conditioning of mines.
Hewever, most of the air washers operate with joint dehumidification and
humidification of the air in which case the Lewis equation is not valid. Hence
these air washers should be calculated on the basis of empirical coefficients
directly giving the final state of the air and the direction of its change.
The above analysis of the validity of the Lewis equation in the, case of air
washers is highly important for the theory of their calculation.
With high -moisture condensation coefficients the flow of condensed moisture
coincides with the direction of the heat flow and has the highest value practically
met with in air coolers. At ~ = 0 the direction of the flow of evaporating moisture
is opposite to the direction of the heat flow.
The validity of the Lewis equation for both above processes proves the presence
of analogy between heat and moisture exchange at various directions of the heat
and moisture flows for comparatively small temperature drops and specific
humidity variations usual in air coolers. It is obvious that the influence of various
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
supplementary heat and moisture exchange processes distorting the Lewis equation
in air coolers is comparatively low. This conclusion coincides with the. theoretical
analysis of this problem made by Prof. Berman [7'].
All above considerations make it possible to conclude that the Lewis equation is
valid for cooling and dehumidifying air by means of finned surfaces in which case
there are no possibilities for additional humidification.
REFERENCES _
1. LEWIS, w. x..The Evaporation of a Liquid into a Gas. Mech. Eng., 44, 445, 1922.
2. MERKEL, F. Verdunstungskiihlung. Vorsthungsheft VDI No. 275,' Berlin, 1925.
3. GOGOLIN, A. A. and RUDOMETKIN, F. I. Heat Transfer in Air Washers for Air Condi-
tioning. -Cellection of Works of the Mechanical Section of VNIKhI, Pishchepromizdat,
1940.
4. GOGOLIN, A. A. The Calculation of Spray Type Air Conditioners. Kholodilnaya Tekh-
nika, (4), 1957.
5. KURILEV, E. S. Some Specific Features in the Process of Heat Exchange Between Drop-
lets and Air in Spray Type Air Conditioners. Works of the Leningrad Technological In-
stitute of the Refrigerating Industry, Vol. IV, Pishchepromizdat, 1953.
6. KARPIS, E. E. and ZUSMANOVICH, L. M. The Investigation of Air Cooling and De-
humidification in Spray Type Air Conditioners. Report of NIIST, Academy of Architec-
ture, Moscow, 1957.
7. BERMAN, L. D. On the Analogy Between Heat- and Mass-exchange. Teploenergetica,
(8), 1955.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Natural Convection in Air with Temperature Differences
Convection naturelle de fair avec faibles differences de temperature.
K. BRODOWICZ, Dr. Eng.
Technical University in Warsaw
Nowowiejska 25, Warsaw, Poland
SOMMAIRE. La repartition de la temperature etait of fectuee autour de tuyau
horizontal et a proximite de la surface de la plaque verticale. La difference de tempe-
ratures entre les surfaces du tuyau et de la plaque et celle'de fair etait connue et
le coefficient de transmission de chaleur etait determine.
L'experience a ete realisee pour une difference de temperatures de 0,5? C. Le coef-
ficient de transmission de chaleur par convection naturelle diminue en fonction de
la difference de temperatures. Il existe quelques equations empiriques pour le cal-
cul de la transmission de chaleur pour une tres faible valeur de Gr Pr (telle que
10-2 - 10-4), mail on ne pent pas les etablir pour la convection de Gr Pr (telle que
et elles n'ont si pas la meme signification qu'une faible difference de temperatures.
L'experience a ete of fectuee par l'A. pour confronter les resultats du calcul de la
transmission de chaleur et les f ormules etablies par d'autres auteurs. Les courbes
ont ete etablies d'apres les calculs e f f ectuees a partir des f ormules. La conclusion est
qu'elles sont voisines l'une de l'autre. L'A. a adopte l'interferometre Mach-Zhen-
ders pour mesurer la repartition de temperatures a proximite des surfaces.
NOMENCLATURE
The following nomenclature is used in this paper:
q
- heat stream, kcal/m2h
h - distance from origin plate, mm
a
- heat transfer coefficient,
T - angle from vertical
kcal/m2 ?C h
A - thermal conductivity kcal/m ?C. h
At
- difference of temperature, ?C.
Nu - Nusselt number
At x = tx-tf, At = tw-tf
Gr - Grashof number
tf
- temperature of surface, ?C.
C,n - constant
tw
- temperature of ambient, ?C.
Pr - Prandtl number
x
- distance from surface, mm
This paper considers a free convection in air for the horizontal pipe and vertical
plate with the little difference in temperature.
Determination of heat transfer coefficient "a" with a little difference in tem-
perature is very important when the instable heat transfer is considered.
The heat transfer gets smaller when the difference of temperature between the
ambient space and the surfaces gets smaller too. It is clear that heat stream
changes according to the formula:
q = adt
in which a changes with change of At. The coefficient diminishes when the dif-
ference of the density of the ambient air gets smaller until At reaches zero, and
in that case there is no convection at all. To determine the value of a for a little
At ?C. we can use Pohlhausen's theoretical formula
4 T
a = c VGr
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
Fig..1. a) Illustration of parallel fringes produced by difraction by the free convection in
air on the vertical plate. b) Isotherm for the free convection in air on the vertical plate.
Fig. 2. a) Illustration of parallel fringes produced by difraction by the free convection
in air on the horizontal pipe, b) Isotherm for the free, convection in air on the horizon-
tal pipe.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 CIA-RDP80T00246AO07500340002-0
1-17
Fig. 3. Average temperature difference as a function of the distance from the surface
of the plate, a. and pipe b.
at [?c]
,at cc1
1 2 3 4 5 6 7 8 9 10 X[mm] 1 2 3 4 5 6 7 8 9 10
Nu c(Gr Pr)n
C and n are determined by experiments for large At ?C. For a little At ?C. c and
n should be checked by experiment, this is the main subject of this paper.
Experimental method.
The visual method by using Mach-Zehnder's interferometer was adopted and
a L
at
40
was determined, and a was calculated from theformula:
~St
adt=A ax) X =o
nt
OH
?C
AUS
N
at O.
?C
o
f8
C
n
t9.6
G
SCN
at.
DT
52
c5
_
NN
180!
120?
"I N
fo x 1---7 1 2 3 4 5 6 7 8 9 fo x (mm]
p f.~
Fig. 4. Average relation At as a function of the distance from surface of plate, a. and
angle from vertical of pipe b.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
lnt,3*44801-1
\ 3001 BAUE R
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
t was obtained by measuring. The method also allows determination of the tem-
perature for each point in the boundary layer, which is very important, because we
can compare the temperature curves with other experiments.
d mK
CH
n
DT,
.52
ECK
?
AN
ct
96?
At
L3?
0
.08
4 5 6 7 8 9 to ::rKcat -J
Lm2 G
Fig. 5. Heat transfer coefficient as a function of the distance from origin of plate, a. and
angle from vertical of pipe b.
h [mm]
100
90
80
70
60
50
40
30
20
t0
/I
120 150 180?
41P
Fig. 6. Relation a am as a function of the distance from the origin of plate, a. and angle
from the vertical of pipe b.
Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
 Approved For Release 2009/05/01 : CIA-RDP80T00246AO07500340002-0
The experiments described in this paper were made with a copper plate which was
partly insulated, only one vertical face 110 mm X 200 mm was uncoverd. The
next experiments were made with copperpipe 0 = 29 mm and Cb = 5 mm
300 mm long. The pipe was placed in a horizontal position for the experiments,
plate and pipe were electrically heated, and for stable thermic conditions, pho-
tographs were made. To obtain conditions zlt