Page 1 of 14

41. Flow of gases and steam through nozzle

Author: Ji korpk, skorpik@fme.vutbr.cz : last update 2015-04
A nozzle is a channel with stepless variable of flow area nevertheless here
described formulas can be use for calculation flow through all types of channels for
example orifice plates, blade passages, gaps etc. In these cases is necessary to take into
account losses that are caused their shape. Flowing of gas through the nozzle is a
process with a decrease pressure and an increase kinetic energy according the First laws
of thermodynamics for open system and the Hugoniot condition.

Converging nozzle
A mass flow rate through a converging nozzle is function pressure at the inlet pi
and the exit pe (back-pressure) of the nozzle:

Equation 1/id101. The flow velocity at the exit of the nozzle in i-s diagram at adiabatic expansion.
(a) formulas for calculation of leaving velocity of gas as function of static state of gas at inlet of nozzle; (b)
formulas for calculation of leaving velocity of gas as function of stagnation state of gas at inlet of nozzle.
A [m2] area of flow; c [ms-1] velocity of gas; i [Jkg-1] specific enthalpy of gas; s [Jkg-1K-1] specific
entropy; [-] heat capacity ratio; r [Jkg-1K-1] individual gas constant; T [K] temperature of gas; p [Pa]
pressure of gas. The subscript i denotes the state at the inlet of nozzle, the subscript e denotes the state at the
exit of nozzle and the subscript c denotes the stagnation state of the gas. This equation is called Saint
Vnantova-Wantzel equation [2, p. 350]. This equation is can derived from First law of thermodynamics for
open system for flow of ideal gas without friction. The derivation of the equation for a calculation of the
velocity of gas at the exit of the nozzle shows the Appendix 101.

05.2015

Page 2 of 14

The velocity of gas ce is function the inlet temperature Ti and pressure pi according
last equation and a maximum velocity is reached at expansion in the vacuum pe=0:

Figure 1/id514. The leaving velocity of gas at the exit of a nozzle.

pat [Pa] atmospheric pressure. Gas properties: =1,4; r=287 Jkg-1K-1; ti=20C; pi=pat; ci=0.

Remark
The pressure ratio pep-1i is usually denoted by the letter or c (c=pep-1ic).
The leaving mass flow rate of gas at the exit is calculated from Equation 1 and the
continuity equation:

Equation 2/id334. The equation of mass flow rate through the nozzle.
m [kgs ] mass flow rate of gas through nozzle; v [m3kg-1] specific volume of gas; m [-] variable parts of
equation that is function of pressure ratio. The derivation of the equation for a calculation of the mass flow
rate of gas at the exit of the nozzle shows the Appendix 334.

-1

According this equation is true, if the pressure on the exit nozzle pe is decreasing
then the mass flow rate m is increased only to a pressure ratio where the mass flow
rate should be decreasing:

05.2015

Page 3 of 14

Equation 3/id515. The maximum of mass flow rate of gas through the nozzle.
The curve 1-a-0 corresponds to Equation 2*. The maximum mass flow rate m* is reached at a pressure ratio
*c. According Equation 2 should following a mass flow rate fall. In really the mass flow rate is a constant
and equals m* from point *c to an expansion to vacuum c=0. The pressure ratio for the maximum mass
flow rate of gas through the nozzle is called the critical pressure ratio (therefore index *). The derivation of
the equation for a calculation of the critical pressure ratio of gas at the exit of the nozzle is shown in the
Appendix 515.

*Bendemann ellipse
The curve 1-a-0 is similar with an ellipse, therefore this curve is usually substituted by
the ellipse for case routine calculations. This ellipse is called Bendemann ellipse:

Equation 4/id162. A formula for approximate calculation of mass flow rate through the nozzle according
Bendemann ellipse.
This formula can be use only for pep*. The derivation of the equation of Bendemann ellipse is shown in the
Appendix 162.

The isentropic index can be various for individual type of gas and therefore its
critical pressure ratio can be also various:
gas *c [-]
gas
*c [-]
-----------------------------------H
0,527
air (dry)
0,528
He
0,487
superheated steam 0,546
CO2 0,540
saturated steam
0,577
Table 1/id699. The critical pressure ratio of some gases.

Gas velocity reaches the speed of sound at the critical and a lower pressure ratio in
the nozzle throat (the narrowest area of the nozzle), this state of the flow is called a
critical state of flow. The equations of flow for narrowest area of the nozzle be can
derived by substituting equations for the critical pressure ratio (Equation 3.) to the
above equations for the gas velocity and mass flow rate:

05.2015

Page 4 of 14

Equation 5/id516. The equations of the critical flow in narrowest area of the nozzle.
These quantities are called critically (critical velocity, critical mass flow rate, critical pressure ratio...). max
this constant are listed in a table according type of gases, pressure ratio and for case ci=0; i* [Jkg-1] critical
enthalpy (at critical enthalpy reaches isentropic expansion the critical velocity).

3D plot of the equation for mass flow rate of gas as function the pressure ratio and
the back-pressure is called flow 3D characteristic of the nozzle:

Figure 2/id335. 3D characteristic of the nozzle.

The curve a lies on a surface which has the axis pic as its normal, it is a case when pic=constant. The curve
b lies on a surface which has the axis p as its normal, it is a case when p=constant. For case expansion to
the vacuum pe=0 the mass flow rate has linear dependence for tic=const., the curve c. The curve d lies on a
surface which has the axis m as its normal, it is a case when *c=constant and the pressure ratio is variable.
The air flows through a nozzle its velocity is 250 ms-1, its pressure is 1 MPa, its temperature is 350C at the
inlet of the nozzle. Surroundings pressure behind the nozzle is 0,25 MPa. The narrowest area of the nozzle
has 15 cm2. Find if the flow through the nozzle is critical flow. Calculate the leaving velocity and the mass
flow rate of air. The properties of air are: cp=1,01 kJkg-1K-1, r=287 Jkg-1K-1, =1,4. You do not solve a
situation behind the nozzle.
Problem 1/id102.
*
[-]
0,5283
tic [C]
380,9406
pic [MPa]
1,1848
c
[-]
0,2110
vic [m3kg-1] 0,1584
c*
[ms-1]
467,9865
max [-]
0,6847
m* [kgs-1] 2,8087
Problem 1: the results.

Ideal shape of converging nozzle

An ideal shape of the nozzle is smooth, parallel with streamlines (on the inlet even
the exit to avoid a rise of turbulence through sudden change of direction of flow
velocity at the wall), on the exit must be uniform velocity field (this condition is
05.2015

Page 5 of 14

confirmed by experiments [4, p. 319]). It means the exit velocity should be in axial
direction of the nozzle. This condition must also satisfy the streamlines at the wall of the
nozzle:

Figure 3/id475. Influence shape of the nozzle on the direction of the exit velocity.
(a) cone nozzle; (b) ideal shape of nozzle; (c), (d), (e) usually shapes of nozzles; (c) so called Vitoshinsk
nozzle or Vitoshinsk converging nozzle (this shape is use for reduction passage between two passages with
different diameter or as blower nozzle of wind tunels); (d) shape of nozzle by lemniscate ; (e) shape of
nozzle for outlet of bottles; R [m] radius of nozzle; l [m] length of nozzle. The cone nozzle has significantly
lower velocity coeficient than the nozzle with shape (b) (definition of the velocity coefficient of the nozzle
is shown in chapter "Flow in nozzle with loss").

The described condition for shape of nozzle is also true for non-circular passages
and blade passages.

State at exit of converging nozzle

The flow behind exit of the nozzle can have two states:
(1) The leaving velocity at the exit of the nozzle corresponds the sub-critical or
accurately the critical pressure ratio, pep*.
If there is not any channel behind the narrowest area of the nozzle that separates the gas
stream from the surrounding environment then the leaving gas stream is deformed and
gradually mixes with the surrounding gas. In this case the state of the leaving gas
(velocity, temperature, pressure) will be the same as the surrounding gas at some
distance from the exit of the nozzle-the leaving gas is in thermodynamic balance with
the surrounding gas:

05.2015

Page 6 of 14

Figure 4/id984. The outflow from a converging nozzle at the critical pressure ratio.
Photo from [3, p. 5].

(2) The pressure ratio is less than the critical ratio, pe<p*.
The leaving velocity is equal the sound speed but the leaving pressure is bigger than the
pressure of surroundings gas, therefore expansion of the leaving gas continues and the
velocity of gas increases to supersonic according the Equation 1. The gas stream area be
must increased according Hugoniot condition. The divergent gas stream forms oblique
shock waves on border between the stream and the surroundings gas. These shock
waves are reflected to the core of gas stream and they are decreased an efficiency of
expansion (they cause pressure drop). The expansion is ended when the pressure is
equal the surroundings pressure and a next process is similar the previous case it means
gradually leveling of the gas state with the surroundings.

De Laval Nozzle (converging-diverging nozzle)

For better efficiency of gas expansion behind the narrowest area of the converging
nozzle (it is the case p*>pe) is necessary made the appropriate conditions. It means a
divergent channel must be added to the converging nozzle behind narrowest flow area
of the nozzle (so called critical flow area, because the speed of sound is reached here):

05.2015

Page 7 of 14

Figure 5/id103. De Laval nozzle (CD nozzle)-direction of expansion.

(a) converging section of the nozzle; (b) divergent section of the nozzle. Ma [-] Mach number; l [m] lengh
of diverging section of nozzle; [deg] the flare angle of the diverging section. The velocity of gas is
subsonic Ma<1 in the converging section, and is sonic Ma=1 in the narrowest area (throat), is supersonic
Ma>1 in the diverging section.

Figure 6/id983. Supersonic exhaust of gas from de Laval nozzle.

The oblique lines inside flow are oblique shock waves, which arise at the exit edge of the nozzle and are
reflecting from the boundary of the stream. Photo from [3, p. 23].

05.2015

Page 8 of 14

Figure 7/id517. i-s diagram used in the description ideal expansion of gas through a CD nozzle.
Calculate a diverging section of the nozzle from the Problem 1. Calculate Mach number on the exit of the
diverging section. The flare angle of the diverging section is 10.
Problem 2/id104.
ce [ms-1] 686,6286
D [cm]
5,0067
l [cm]
3,6375
Ma [-]
1,6730
Problem 2: the results.
Water steam flows through a de Laval nozzle. Pressure and temperature of the water steam is 80 bar
respectively 500C at input to the nozzle. The leaving pressure is 10 bar. The mass flow rate of water steam
must be 0,3 kgs-1. Calculate base dimensions of the nozzle and state of water steam at the exit. The flare
angle of the diverging section is =10.
Problem 3/id336.
c [-]
0,1250
*c [-]
0,5460
p* [MPa]
4,3680
c* [ms-1] 615,4186
D* [m2]
6,4475E-3
ce [ms-1] 1054,9313
De [m2]
8,7083E-3
l
[m]
1,2920E-2
Problem 3: the results.

Flow in de Laval nozzle at non-design states

For case good design of de Laval nozzle is reached the pressure pn during
expansion, which is the same as the back-pressure, this pressure is called the
designated pressure of the nozzle. The non-design state are changed the parameters of
working gas at the inlet or exit of nozzle. These parameters are changed from various
causes (for example a regulation of the mass flow rate). If pe>pn (over-expanded
nozzle) then the length of the divergent section of the nozzle is longer than it is need. If
05.2015

Page 9 of 14

pe<pn (under-expanded nozzle) then the length of the divergent section of the nozzle is
shorter than it is need. At pressure which is the bigger than the designated pressure can
arise the normal shock wave inside de Laval nozzle:

Figure 8/id105. Pressure and velocity variation with location inside a de Laval nozzle for several backpressures.
Subscript 1 denotes a state in front the normal shock wave; index 2 denotes a state behind the normal shock
wave.

pe>pb
At this back-pressure the velocity at the throat does not reach speed of sound - it does
not reach critical state, the curve a. Inside the converging section of the nozzle is done
subsonic compression (increasing of pressure and decreasing of velocity-in this case the
converging section of the nozzle acts as a diffuser) to the pressure pa.
pe=pb
At this back-pressure the velocity at the throat reaches speed of sound-critical state, the
curve b. Inside the converging section of the nozzle is done subsonic compression to the
pressure pb.

05.2015

Page 10 of 14

pb>pe>pd
At this back-pressure the velocity at somewhere inside converging section of the nozzle
occurs non-continuity of the velocity and pressure (is developed the normal shock
wave). Behind the normal shock wave is subsonic flow and the gas is being compressed
to pressure pc.
pe=pd
At this back-pressure the normal shock waves arises exactly at the exit of the nozzle.
pd>pe>pn
At this back-pressure the normal shock waves arises behind the exit of the nozzle. This
normal shock wave is unsteady (free flow) and it alternately develops and disappears
(similar situation is in case an converging nozzle where back-pressure is smaller than is
the critical pressure pe<p*)
pn>pe
At this back-pressure continued the expand of working gas behind the exit of the nozzle.
Similar as previous case arise sonic effects.
*Remark
A develop of the normal shock wave inside divergent section of the nozzle be can
assumed from the Hugoniot condition. A smooth change the supersonic flow on the
subsonic flow is allowed only in throat of the nozzle. The position of normal shock
wave inside the diverging section of the de Laval nozzle is can calculate through the
equations for stable normal shock wave:
Find the approximate position of normal shock wave inside de Laval nozzle from Problem 2, if the backpressure is increased about 0,52 MPa. The specific losses of the normal shock wave is calculated
in Problem 1 in the article ffects at high velocity flow.
Problem 4/id862.
x
iic
Ma1
T1
p1
c1
Ma2
T2
p2

[mm]
9,3611
[kJkg-1] 384,75
[-]
1,3230
[K]
484,489
[MPa]
0,4144
[ms-1]
583,7204
[-]
0,7745
[K]
584,0188
[MPa]
0,7772
Problem 4: results.
x [mm] the position of the normal shock wave from the throat. This calculation does not include the flow
losses, only losses of the wave are included.

The normal shock wave inside the nozzle is not usually stable [4, p. 363] therefore
it can cause a vibration of the nozzle and a vibration neighboring machines. This
unstable of the shock wave increases noise of the nozzle.
05.2015

Page 11 of 14

A operation back-pressure has an influence on length of the nozzle of a rocket

engine. During flight of a rocket through atmosphere is changed surrounding pressure
with the altitude. Therefore the nozzle of the rocket engine for first stage are designed
on atmospheric pressure (pressure near ground) and the next stages are designed on
smaller pressure (according the altitude of the ignition of this engine). The engines of
the last stage of the rocket are designed on expansion to vacuum [1].

Flow through oblique cut nozzle

If there is supersonic flow inside a oblique cut nozzle then the stream of working
gas is deviated by an expansion fan from the nozzle axis. This expansion fan is
developed on shortly side of the nozzle. Similar situation arises for case a blade passage
at the end blade passage (see a chapter de Laval nozle as blade passage[42]):

Figure 9/id106. Flow through an oblique cut nozzle- for a case the critical state in the throat.
left a converging nozzle; right de Laval nozzle. [deg] Mach angle; [deg] deviation of the stream from
the nozzle axis.

A change of the back-pressure p2 at the end of a converging nozzle influences the

direct of the leaving stream:
p2>p*
In the throat of the nozzle is the pressure p2 because the flow is subsonic. The direct of
the leaving stream is the same as axis of the nozzle.
p2=p*
If the back-pressure p2 is the critical pressure then this pressure must be in the throat of
the nozzle. The velocity stream is equal the sound velocity at this location. The
expansion of working gas does not proceed, therefore the stream does not deviate from
the axis of the nozzle.
p2<p*
In this case the critical pressure is in the throat (cross section A-C) and the backpressure is at the cross section A-C'. Between the cross section A-C and A-C' there is
an expansion fan. The stream is deviated from the axis of nozzle with angle during
flows of working gas through the expansion fan.

05.2015

Page 12 of 14

The flow through oblique cut de Laval nozzle is the same as the supersonic flow
around an obtuse angle. The start of expansion is at pressure p1 on the cross section A-C
and the end of expansion is at back-pressure p2 on the cross section A-C'.

Flow in nozzle with loss

In the previous paragraphs is described adiabatic expansion of working gas in the
nozzle without losses. This expansion is called isentropic expansion. But the expansion
in the nozzle is influenced by a friction or also internal loss heat that arises through
inner friction of the gas and a friction gas on surface of the nozzle. This friction heat
decreases of the kinetic energy of the gas at the end of the nozzle. The friction heat is a
loss between the kinetic energy at the nozzle exit for case isentropic expansion and the
actual kinetic energy at the nozzle exit:

Figure 10/id108. Flow in a nozzle with the loss.

zp [kJkg-1] specific loss in the nozzle.
Index iz denotes the state for case the isentropic expansion.

At pressure p*iz can be the velocity in core of the stream equal sound velocity but
the velocity on the periphery of the flow area can be subsonic due boundary layer on the
surface of the nozzle. The mean velocity at the throat of the nozzle is smaller than sound
velocity respective the mean kinetic energy of the stream is smaller than kinetic energy
for case sound velocity. The mean velocity stream occurs of the sound velocity at the
pressure p* - it is position behind the throat. In this the flow area is p*<p*iz, therefore the
velocity can be supersonic in the flow core and subsonic on the periphery.
The nozzle loss is can calculated through energy parameters of the nozzle as a
velocity coefficient , a mass flow coefficient (discharge coefficient) and a nozzle
efficiency :

Equation 6/id569. The energy parameters of the nozzle.

05.2015

Page 13 of 14
[-] velocity coefficient; [-] mass flow coefficient; [-] nozzle efficiency; ciz [ms-1] the leaving velocity
at flow without loss (isentropic expansion); miz [kgs-1] the mass flow rate at flow without loss (isentropic
expansion). The values of the velocity coefficients are shown in [4, p. 328] (for convrgent nozzle
including the cone nozzles) a [4, p. 348] (for Laval nozzles).
Calculate the throat area and the exit area of the de Laval nozzle and its efficiency. Through the de Laval
nozzle flows the water steam saturation. The mass flow rate is 0,2 kgs-1. The stagnation pressure at the inlet
of the nozzle is 200 kPa and pressure at the exit is 20 kPa. The velocity coefficient of the nozzle is 0,95.
Problem 5/id109.
A*
ce
Ae

[m2]
6,9934
[ms-1] 803,0844
[m2]
17,2275
[-]
0,9025
Problem 5: results.

Questions for study

(1)

What is the leaving velocity of the converging nozzle? It is known

the enthalpy and other properties of working gas on the inlet
and the exit of the nozzle.

(2)

What is an equation for calculation of the mass flow rate for

the nozzle?

(3)

What is the critical pressure ratio for the converging nozzle?

(4)

Sketch the 2D characteristic of the nozzle for the mass flow rate
as function of the pressure ratio m-c (m-mass flow rate,
c-pressure ratio between the inlet of the exit of the nozzle).

(5)

Describe (sketch) the 3D characteristic of the nozzle and highlight

at least one case of the flow for p0=constant or p=constant or pe=0
or m=constant.

(6)

What is maximum leaving velocity of the converging nozzle and in

what case?

(7)

What can be impacts of the non-design state of the de Laval nozzle?

(8)

Sketch i-s diagram for the expansion of the de Laval nozzle (flow
with loss).

(9)

Define of the velocity coefficient, the mass flow coefficient

and the nozzle efficiency.

References
1. TOMEK, Petr. Kde jsou ty (skuten) kosmick lod?. VTM Science, 2009, leden.
Praha: Mlad fronta a.s., ISSN 1214-4754.
2. KALK, Josef, SKORA, Karel. Technick termomechanika, 1973. 1. vydn,
Praha: Academia.
05.2015

Page 14 of 14

3. SLAVK, Josef. Modifikace Pitotova pstroje a jeho uit pi proudn plynu

hubic, 1938. Praha: Elektrotechnick svaz eskoslovensk.
4. DEJ, Michail. Technick dynamika plyn, 1967. Vydn prvn. Praha: SNTL.

Citation this article

This document is English version of the original in Czech language: KORPK,
Ji. Proudn plyn a par dzou/tryskou, Transforman technologie, 2006-02, [last
updated 2015-04]. Brno: Ji korpk, [online] pokraujc zdroj, ISSN 1804-8293.
Dostupn z http://www.transformacni-technologie.cz/proudeni-plynu-a-par-dyzoutryskou.html. English version: Flow of gases and steam through nozzle. Web:
http://www.transformacni-technologie.cz/en_proudeni-plynu-a-par-dyzou-tryskou.html.
Ji korpk, LICENCE

www.transformacni-technologie.cz

05.2015