Background Image
Previous Page  15 / 21 Next Page
Information
Show Menu
Previous Page 15 / 21 Next Page
Page Background

Fig. 2. The layout chart of pressure gauge locations in one-diaphragm shock tube (at

the Institute for Problem Mechanics, RAS)

Fig. 3. Experimental time dependencies for the second (

a

) and third (

b

) pressure

gauges (Experiment 4)

the graphic curve of pressure gauge 2 (

z

= 5

.

78

m), it follows that the

observed maximum of the amplitude corresponds not to the initial shock

wave arrival, i.e., not to the time

t

9

ms, but to a later instant

t

30

ms.

This phenomenon might be associated with the fact that at the moment

of time

t

30

ms, at the point where gauge 2 is installed (approximately

near the center of the computed area), the shock wave amplification after

its reflection from the end surfaces of the shock tube is significantly higher

that its attenuation during its interaction with rarefaction waves. These

graphic curves also indicate that the further (relative to time) gas wave

motion in a shock tube results in shock wave attenuation and the transition

of the flow into the acoustic wave movement.

Thus, it may be noted that the unsteady gas flowing in the aerodynamic

shock tube has a number of features which require a more detailed

numerical study. The analysis of the characteristics of the flow in the shock

tube of the Institute for Applied Mechanics, RAS, at the initial stage of

study is preferably to be carried out using an approximate one-dimensional

model, the elements of which are presented in this paper.

ISSN 0236-3941. HERALD of the BMSTU. Series “Mechanical Engineering”. 2014. No. 1 17