Vous êtes sur la page 1sur 4

February 18, 2016

Aerospace Structures Lab

EXPERIMENTS ON THIN-WALLED PRESSURE VESSEL

Aswathi Krishna, Aucknoor Amogh, Avinash Chandra, Bukhya Rishwith,


Chandrakant Kaushal, Ch. Dileep Kumar, Fahd Bin Abdul Hasis
Indian Institute of Space Science and Technology

INTRODUCTION
A cylindrical or spherical object which is subjected to an
internal pressure is an important practical problem. Such a com-
ponent is called a pressure vessel. Pressure vessels are one of the
most widely used structural components. In aerospace industry
they have a special importance in designing rocket motor casing,
fuel storage tanks for cryogenic engine, air chambers for large
scale wind tunnels etc.

THEORY
In many applications involving pressure vessels, it is Figure 1. All the stresses on the element of the cylindrical vessel.
reasonable to assume that:
(i) The material is isotropic
(ii) The strains resulting from the pressures are small
(iii) The wall thickness, t of the pressure vessel is much smaller
than some characteristic radius i.e. t = r0 ri << r0 , ri . This is
called the thin cylinder assumption.
(iv) End effects can be neglected.

Because of (iii), it will be assumed that there is negligible


variation in the stress field across the thickness of the vessel.
The cylinder has three different principal stress values, the
circumferential stress(h ), the radial stress, and the longitudinal
stress l , which acts in the direction of the cylinder axis. The
radial stress is much smaller compared to the other two normal
stresses and hence can be neglected. r and rz (refer fig.1) are
zero throughout the thickness. Hence this is a case of plain stress.

Figure.2 shows the free body diagram of a cylinder with


Figure 2. Free body diagram of thin cylinder under loading
closed ends. Longitudinal and hoop stresses in a thin-walled
cylindrical pressure vessel can be derived easily by invoking ba-
sic force balance. It turns out that, for a cylinder with closed
ends, axial = Pr Pr
2t and hoop = t . For open ended cylinder, and the slope of axial strain v/s hoop strain should be numerically
hoop = h /E and axial = h /E. Hence the slope of hoop equal to the Poissons ratio of the material. For closed cylinder,
stress v/s hoop strain should be equal to the Youngs modulus hoop = (hoop axial )/E and axial = (axial hoop )/E.

1
PROCEDURE
The experiment has 2 configurations i.e open and closed end
configurations. When the ends are clamped, the load transfer oc-
curs directly to base structure and there is no longitudinal stress DISCUSSIONS
developed and it is called as open end configuration, while the * The strain transformation relation is given below, where, xx0
other configuration has its ends as free and thus develops the lon- represents the normal strain in x direction in the new orien-
gitudinal stress and is called as closed end configuration. tation and xx represents the same in initial configuration.

1. For open end configuration, the end of cylinder was clamped


with the structure. The air valve was tightened and zero was (1 + cos2) (1 cos2)
xx0 = xx + yy + xy sin2
set. 2 2
2. The chamber was pressurised using the hand lever in the
steps of 0.2 MPa till 3 MPa. In order to find out the orientation of the plane where xx
3. The stress values obtained from the six strain gauges at- goes to zero in the open ended configuration, replace the
tached to the cylinder were noted down. LHS with 0. Assuming that this orientation is found with
4. For closed end configuration, the clamping screw was loos- respect to the initial configuration (Strain Gauge 1), xy = 0,
ened. xx = h and yy = l = h .
5. The same procedure which was followed in case of the open
end cylinder was repeated. (1 + cos2) (1 cos2)
0 = h h +0
2 2
1
RESULTS = tan1

* The pressure vessel used for the experiment was made of an
Substituting = 0.33
aluminium alloy, Al 6063, which had a Poissons ratio of
0.33 and Youngs modulus of 69 GPa. The slope of longitu- = 60.1245o
dinal v/s hoop strain in open cylinder configuration provided
a good estimate of Poissons ratio. observed = 0.33497.
Thus for an open ended cylinder made of Aluminium, strain
* The slope of hoop stress v/s hoop strain in open cylinder
gauges inclined at 30o with the axis of the cylinder, will
configuration provided a good estimate of Youngs modulus
theoretically read zero strain. For an unknown material, this
of the material. Eobserved = 68.88 GPa.
knowledge provides a way of evaluating the Poissons ratio.
* In open configuration, strain gauge inclined at 30o (strain
* Strain gauges 3 and 5 are perpendicular to each other. Hence
gauge 3 in fig. 3) always gave a value close to 0 for normal
the sum of strains measured by strain gauges 3 and 5 should
strain.
be equal to the sum of strains in gauges 1 and 2, which are
also normal to each other. This is because of the trace of a
* The Mohr circle of strains, showed a negative value as
matrix is invariant under co-ordinate transformation. This
the minimum normal strain for the open configuration.
statement was verified with the observed data.
This is because in the longitudinal direction there are no
direct strains. Hence Poissons effect dominates. For
* The angle corresponding to the maximum or minimum
a tensile hoop stress, the longitudinal strain developed
shear strain is given by the formula:
by Poissons effect will be compressive. However, for a
closed configuration the normal strains are always lead-
ing to elongation. This can be seen from the following: 1 (yy xx )
= tan1
axial = (axial hoop )/E. Also axial = hoop /2. So 2 2xy
axial = (hoop (0.5 ))/E, and as for normal materials As xy = 0 f or the initial con f iguration
can not be more than 0.5, axial strain in case of closed
= 45o
cylinder can never be compressive.

* The readings of strain gauge 1 and 6 were very close, Thus strain gauges inclined at 45o with the initial configu-
indicating that sufficiently far from the ends, the strains do ration, will correspond to a maximum/minimum shear strain
not vary along the axis. value. Since this is the topmost/ bottommost point on the

2
Mohr circle. Thus the normal strain measured by these strain
gauges will be the average of maximum and minimum nor-
mal strains. This statement was verified with the observed
data. This statement was verified with the observed data.

References
1. Solid Mechanics Part-I by Kelly, section 7.3, Pg-186
2. http://ocw.mit.edu/courses/materials-science-and-
engineering/3-11-mechanics-of-materials-fall-
1999/modules/pv.pdf

Figure 5. ANSYS Workbench 15.0, results of open end configura-


tion, hoop strain. Theoretical value=5.41104 , Software predicted
value=5.22104 , experimental value=6.09104 .

Figure 3. Schematic representation of the orientation of the strain


gauges used.

Figure 6. ANSYS Workbench 15.0, results of closed end configuration,


hoop strain for applied pressure of 0.2 MPa. Again only half of the
cylinder has been modelled. Similar boundary conditions. Theoretical
value=3.01105 , Software predicted value=3.4105 , experimental
value=3.8105 .

Figure 4. ANSYS Workbench 15.0, results of open end configuration,


axial strain for applied pressure of 3 MPa. Since the pressure vessel is
symmetric, only one half has been modelled. Because of symmetry, the
mid plain is not allowed to move along the axis. Boundary conditions:
fixed support on one side and remote displacement freezing rotations
and axial movement at the other end. Theoretical value=1.78104 ,
Software predicted value=1.72104 , experimental value=2.02104 .
Figure 7. ANSYS Workbench 15.0, results of closed end configuration,
axial strain.

3
Figure 8. Strain v/s pressure plot for cylinder with closed ends. The
markers indicate the observed values and the lines indicate the least
square fit line.
Figure 10. Hoop stress v/s hoop strain for open ended cylinder. Mea-
sured slope of the linear fit line is 68.88 GPa, very close to the actual
value of Youngs Modulus for the material.

Figure 9. Strain v/s pressure plot for cylinder with open ends. The mark-
ers indicate the observed values and the lines indicate the least square fit
line.

Figure 11. Longitudinal strain v/s hoop strain for open ended cylinder.
Measured slope of the linear fit line is -0.33497, the magnitude of which
is very close to the actual value of Poissons ratio for the material.

Vous aimerez peut-être aussi