Lab 7 - Estimation of Internal Pressure of A Soda Can

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Lab 7: Estimation of the Internal Pressure of a

Soda Can

A Lab Report

Submitted By

Neel Nadpara

ME 3264 Applied Measurements Laboratory


Section 004L

Department of Mechanical Engineering


University of Connecticut

Lab Preformed On:

March 23, 2015

1|Page
Summary:
Beverage cans have been around since the early 1930s; however today over a half trillion cans
are produced each year. Today’s cans are an engineering feat with a thickness of only 70
thousandths of an inch; today’s cans are capable of holding internal pressures of 90 psi while
shaped to provide good pouring, an ergonomic grip, and are able to be stacked. The cans are
designed to be cylindrical to reduce to forces acting along any corners (where the can would be
weakest), they include a convex half sphere at the bottom to allow the can withstand greater
internal pressure, and a wide-mouth is included to allow for easy pouring and sipping. Moreover,
there are many similarly shaped products that are used for engineering applications such as
pneumatic cylinders and other gas containment vessels. These vessels contain a stress on both
the axis of the cylinder and along the circumference of the container. These stresses are known
as the axial stresses and hoop stress respectively. The primary objective of this lab was to
develop a working knowledge of strain gage and associated instrumentation, and to learn how to
apply it to determining pressure inside an unopened container of carbonated drink. The objective
was also to learn to use a caliper and micrometer for precision measurement. In this lab both
axial and hoop stress were calculated for a soda. This was done by employing the use of strain
gages that measured the change in pressure caused by a soda being opened. The pressure
obtained in the can was calculated using then the hoop strain and then the axial strain. An
internal pressure of 538 kPa ±2.7 kPa was determined; though it appears to be higher than the
typical 300 to 350 kPa which is likely attributed to numerous errors.

Experimental Procedure:

This experiment requires a number of equipment and materials. On the materials side the
following supplies were utilized: an unopened can of diet coke, a strain gauge, 320 grit silicon
carbide sandpaper, etching solution, neutralizing solution, adhesive, resin-cored solder, and
scissors. Some equipment and instrumentations were also required: a P-3500 strain gauge
conditioner, a digital multi-meter, an iGAUGING origin digital caliper, and an iGAUGING
SpeedMic micrometer (note these measuring instruments were not the ones that were provided
and were utilized as they were more precise and would therefore reduce error propagation). First,
the strain gauge needed to be affixed to the can. However, in order to reduce the chances of error
the surface of the can needed to be prepped. In order to complete this, any labeling needed to be
removed, for this a piece of 320 grit silicon carbide sandpaper was employed. The sandpaper was
used to gently sand the surface of the can in order to remove any paint or labeling in the region
where the strain gauge was to be applied and care was taken not to sand to deeply as to
significantly reduce the thickness of the can and generate a weak spot. Next, the same region was
cleaned with a degreaser. Afterwards, an etching solution was employed to etch the newly

2|Page
exposed surface of the can. Following this, a neutralizer was applied to neutralize the effects of
the etching solution. Then, the strain gauge was affixed to a piece of tape which was centered
half way up the can with the strain gauge in a horizontal position (to measure hoop strain). Next
the tape was peeled up to expose the backside of the strain gauge and catalyst was applied to the
region under the strain gauge that way the glue would set quickly. Subsequently, glue was
applied to the back of the strain gauge and the strain gauge was attached to the can. Then the
glue was allowed four minutes to dry and then the take was peeled back slowly to prevent the
strain gauge from being removed. Then, the soldering iron was turned on and allowed to heat up
and the tip was “tinned” with the solder. Afterwards, the wires from the strain conditioner were
soldered to the two terminals on the strain gauge. The wires were secured to the side of the can
via tape to prevent any pulling from causing the solder/wires from disconnecting due to strain on
the joints. Following this, the connections were tested using a multi meter for a resistance value
of about 120 Ohms. Once the gages were tested and working as expected, they were connected to
strain gage box in a quarter bridge circuit. Strain on the box was balanced to set the stain to zero
to allow us to determine the strain change when the soda can was opened. Afterwards, the
diameter of the can was measured with the iGAUGING origin calipers in millimeters. Care was
taken to ensure that the calipers were level when the measurements were taken. Five
measurements were taken in different places to generate an average diameter. Next, the can was
opened and the new output strain was recorded. Following this, the can was cut open and a large
rectangular section was cut off. Successively, the iGAUGING SpeedMic micrometer was
employed to make five measurements for the thickness of the can. In order to ensure accurate
measurements both the soda can and micrometer’s measurement faces were cleaned with
degreaser. Like the diameter five measurements were taken for the thickness and averaged to
account for variability in the thickness and error in the measurements. After obtaining the strain
and measuring the thickness and diameter calculations could then be performed to determine the
axial and hoop stresses and the internal pressure of the soda can be determined.

Equations Utilized for Calculations:

Hoop Stress Derivation


[Force generated by hoop stress acting thought vessel’s wall] – [Force
generated by pressure acting on vertical face of the sectioned fluid] =
0

Force Balance:

𝟐[𝛔𝟏 (𝐭𝐝𝐲)] − 𝐩(𝟐𝐫𝐝𝐲) = 𝟎


𝐩𝐫 𝐭𝛔𝟏
𝛔𝟏 (𝐇𝐨𝐨𝐩 𝐬𝐭𝐫𝐞𝐬𝐬) = , 𝐩 (𝐏𝐫𝐞𝐬𝐬𝐮𝐫𝐞 (𝐇𝐨𝐨𝐩)) =
𝐭 𝐫

Pressure (Hoop)
𝐭𝛔𝟏 𝟐𝐄𝐦 𝐭 𝛆𝐡
𝐩 (𝐇𝐨𝐨𝐩) = =( )( )
𝐫 𝐫 𝟐−𝐯

3|Page
Axial Stress Derivation
[Force generated by axial stress acting throughout vessel’s wall] –
[Force generated by pressure acting on the sectioned fluid] = 0

Force Balance:

𝛔𝟐 (𝟐𝛑𝐫𝐭) − 𝐩(𝛑𝐫 𝟐 ) = 𝟎
𝐩𝐫 𝟐𝐭𝛔𝟐
𝛔𝟐 (𝐀𝐱𝐢𝐚𝐥 𝐬𝐭𝐫𝐞𝐬𝐬) = , 𝐩 (𝐩𝐫𝐞𝐬𝐬𝐮𝐫𝐞 (𝐀𝐱𝐢𝐚𝐥)) = ( )
𝟐𝐭 𝐫

Pressure (Axial)
𝟐𝐭𝛔𝟐 𝟐𝐄𝐦 𝐭 𝛆𝐚
𝐩 (𝐀𝐱𝐢𝐚𝐥) = =( )( )
𝐫 𝐫 𝟏 − 𝟐𝐯

Internal Pressure of
Soda Can 𝟐𝐄𝐦 𝐭 𝛆𝐡
𝐩=( )( )
𝐫 𝟐−𝐯

Figure 1: Picture of the lab setup with major measuring equipment.

4|Page
Results and Discussion:

Measurement Data:

Unopened Diameters
Measurements 1 2 3 4 5 Average
Diameter (m) 0.065820 0.065940 0.066160 0.066110 0.065910 0.065988
Table 1: Table showing diameter measurements and calculations.

Material Thicknesses
Measurements 1 2 3 4 5 Average
Diameter (m) 0.0001210 0.0001210 0.0001140 0.0001230 0.0001300 0.0001218
Table 2: Table showing thickness measurements and calculations.

Strain and Voltage Measurements


Strain Voltage (V)
0.001742 N/A
Table 3: Table showing strain measurements.

Calculated Data:

Pressure and Strain


Axial Hoop
Pressure (kPa) 2958.137116 537.843112
Strain 4.01434E-08 0.001742

Pressure in Soda Can (kPa) 537.843


Table 4: Table showing the calculated values for axial and hoop pressure and strain along
with the internal pressure of the soda can.

5|Page
Comparision of Calculated Pressure to Ideal
Pressure

600 538

500

400 350
Pressure (kPa)

300

200

100

0
Actual Pressure Ideal Pressure

Figure 2: Graph showing the comparison of the pressure obtained and what should have been
obtained.

Discussion:

The results in Table 3 indicate that strain was measured to be .001742. The calculated hoop
pressure is 537.84 kPa while the axial was calculated to be 2958 kPa. We were informed that the
“ideal” pressure inside the soda can at room temperature is within 300 to 350 kPa. Therefore, as
seen in Figure 2 the experimental results are significantly greater than the ideal results. This is
likely due to an error with our measurements or our strain reading. The error can be attributed to
taking measurements as the calipers/micrometer may have been at a slight angle. We tried to
curtail the fact that the can would have varying diameters and thicknesses by taking the average
of multiple measurements, despite this there are still likely to be flaws. The most likely source of
error was the placement of the strain gauge. The strain gauge was supposed to be placed in the
center of the can perfectly horizontally. However, if the strain gauge was crooked, even by a
small amount it is likely that some axial strain was computed as well as hoop stress. This would
make the measured strain be much larger which in turn would make our pressure calculation be
much greater than it should have been. Also, since the sandpaper is an abrasive and removes
material this could impact our results. If the thickness in that region decreased significantly then
the strain measured would be greater than it should have been. There could have been
instrumentation error as well since the error of the strain conditioner was unknown and it is
unlikely that the known voltage used in the Wheatstone bridge was without some degree of

6|Page
uncertainty. Despite this, the experiment did manage to meet its objective of demonstrating the
use of these various instruments and showing the existence of hoop and axial stresses.

Data and Error Analysis:

In this experiment three measurements were taken and two material properties were obtained
from credible sources. Nevertheless, there is some uncertainty to each of these values which in
turn leads to a certain degree of uncertainty in our calculated solutions for axial and hoop
pressures along with our calculation for the internal pressure of the soda can. In order to
determine these uncertainties, the uncertainties were determined as seen in the table below and
the partials of each of the respective variables were taken and calculated in order to perform a
root mean square (RMS) calculation. Since the modulus of elasticity and Poisson’s ratio were
obtained and not measured we assumed the uncertainty to be half that of the last significant
figure. The steps to solve for error can be seen below:

𝟐𝑬𝒎 𝒕 𝜺𝒉 𝟐𝑬𝒎 𝒕 𝜺𝒂
𝑷 (𝑯𝒐𝒐𝒑) = ( ) (𝟐−𝒗 ) ; 𝑷 (𝑨𝒙𝒊𝒂𝒍) = ( ) (𝟏−𝟐𝒗 ) Equation 1-2
𝒓 𝒓

𝜹𝑷 𝝏𝑷 𝟐 𝝏𝑷 𝟐 𝜹𝑷 𝟐
𝝏𝑷 𝟐 𝟐
𝜹𝑷 (𝑹𝑴𝑺) = √(𝜹𝑬 𝜹𝑬𝒎 ) + ( 𝝏𝒕 𝜹𝒕) + ( 𝝏𝒓 𝜹𝒓) + ( 𝜹𝜺 𝜹𝜺) + ( 𝝏𝒗 𝜹𝒗) Equation 3
𝒎

Where the Partials were calculated as follows:

Partial Axial Hoop/Internal Pressure


𝜹𝑷 −2𝜀 𝑎 𝑡 −2𝜀ℎ 𝑡
𝜹𝑬𝒎 𝑟(2𝑣 − 1) 𝑟(𝑣 − 2)
𝝏𝑷 −2𝜀𝑎 𝐸𝑚 −2𝜀ℎ 𝐸𝑚
𝝏𝒕 𝑟(2𝑣 − 1) 𝑟(𝑣 − 2)
𝝏𝑷 2𝜀𝑎 𝐸𝑚 𝑡 2𝜀ℎ 𝐸𝑚 𝑡
𝝏𝒓 𝑟 2 (2𝑣 − 1) 𝑟 2 (𝑣 − 2)
𝜹𝑷 −2𝐸𝑚 𝑡 −2𝐸𝑚 𝑡
𝜹𝜺 𝑟(2𝑣 − 1) 𝑟(𝑣 − 2)
𝝏𝑷 4𝜀𝑎 𝐸𝑚 𝑡 2𝜀ℎ 𝐸𝑚 𝑡
𝑟(2𝑣 − 1) 2 𝑟(𝑣 − 2)2
𝝏𝒗
Table 5: Table showing partials for each of the variables uncertainty was to be calculated
for.

7|Page
Uncertainty and Error Propagation
Diameter Poisson's
Young's Modulus Thickness Strain Error
Uncertainty Ratio
Uncertainty (Pa) Uncertainty (m) (m/m)
(m) Uncertainty
0.5 ±0.0000005 ±0.000005 ±0.0000005 ±0.005
Young's Poisson's
Thickness Radius Error
Modulus Strain Partial Ratio
Partial Partial (kPa)
Partial Partial
Axial Error 9.732E-10 5559629 -2035.38 1673000000 447.703 ±1.8725
Hoop Error 0.000007679 4415000000 -16060000 304100000 321148 ±2.7345
Internal Pressure
0.000007679 4415000000 -16060000 304100000 321149 ±2.7345
of Soda Can Error

Table 6: Table showing the error propagations for the major calculated values of the lab.

Error Partial Comparison


1000
836.5
800
Error Amount (Pa)

600

400

200
4.866E-10 2.7798145 0.0101769 2.238515
0
Young's Thickness Radius Strain Poisson's
Modulus Ratio
Figure 3: Comparison of the error amounts for each respective partial.

From looking at the Figure 3 above it can be seen that the greatest error propagation is from the
strain measurement by far. This is can be expected since the uncertainty for each of the other
variables is relatively low while the strain partials are very large despite the uncertainty being
very low. Moreover, from the table above it can be seen that the overall error propagation in
units of kPa is ±1.8725 for the axial and ±2.7345 for the hoop and the internal pressure estimate

8|Page
of the soda can. However, this shows that the error propagation is very small compared to the
values that we actually calculated for the internal pressure of the soda can. Since we calculated a
value of 538 kPa for the internal pressure of the soda can with an uncertainty of ±2.7345 kPa the
error due to the uncertainty of our measurements is only .5% of our solution. Despite this, there
is still a fair amount of error that needs to be accounted for since our measured value is 188 kPa
off of the ideal value that we should have gotten (this is a percent error of 53.7%). From Table 1,
we see that the diameter is changing a small amount, as is the thickness. Despite taking the
average of five measurements there can still be a fair amount of error in these measurements as
the diameter and thickness of the can is not uniform which in turn can affect the accuracy of our
results. Additionally, there error with setting up the experiment as the strain gauge may have not
been perfectly placed perfectly centered and horizontally on the can which may have led to some
of the error. The strain gauge was attached to the can via glue and if the glue had not properly
cured, then the strain readings could be off. Another error can be attributed to the uncertainty of
the strain gage box itself, which was not known and the wires affixed to the strain gauge do add
some resistance while small it could lead to error. Moreover, the temperature could have affected
the results and lead to the error that we experienced. The can was given to us at room
temperature however, room temperature could vary day to day and the internal pressure of the
can should increase as the temperature of the can increases as the CO2 in the can will want to
come out of solution as the temperature increases. Furthermore, there are errors that can be
attributed to the material properties that we used. Aluminum beverage cans are typically 3XXX
or 5XXX series aluminum which are wrought aluminum alloys with the primary alloying
element being manganese and magnesium and silicon respectively. For our material properties
Em (modulus of elasticity) and v (Poisson's ratio), we had used material properties for 6061-T6
aluminum. This is not the same alloy that is used to create the soda can as 6061 aluminum is a
age hardening alloy and is not a wrought alloy. This means that there is typically little to no cold
work in this alloy, however beverage cans are formed by cold working a small aluminum
cylinder into its final can shape. Thus, the material is much harder and the values for Em
(modulus of elasticity) and v (Poisson's ratio) will be much larger since the slope of the elastic
region is much steeper due to the material being harder due to cold working. Additionally,
materials that are cold worked have residual stresses, this these stresses could lead to improper
strain measurements. Also, materials that are cold worked are also anisotropic which means that
the grain direction in the microstructure of the material all fallows a single direction. Another
example of this is wood where there is an orientation that is “with the grain” and one where the
orientation is “against the grain.” The material’s properties vary widely when measured with or
against the growth grain. This would account for why our measured hoop and axial stresses are
so varied. Moreover, this would mean that the Em (modulus of elasticity) and v (Poisson's ratio)
would be different values for both axial and hoop stress. Another error is the fact that the soda
can may not be uniform in its materials. The outside labeling and the internal coating that
prevents the beverage from obtaining a metallic taste could have impacted the mechanical
properties for the can. Both of these materials are polymers which have much different properties

9|Page
compared to metals (they tend to have much short hard to define elastic regions and very long
inelastic regions in the stress vs. strain curve). This could reduce the Em (modulus of elasticity)
and v (Poisson's ratio) thereby impacting the results. This experiment can also be improved in
several ways. The micrometer and calipers could be setup on a fixture to insure that the
measurements made are not crooked which could lead to error. There should be two strain
gauges used in opposite directions to measure the strain for both hoop and axial pressures so that
the axial measurements could be made more valid. Also, the multiple cans can be tested to
account for variations as a simple dent could lead to different results. Also, the material
properties could be calculated by using a tensile tester and taking the slope of the elastic region
stress strain curve and the ratio of the proportional decrease in a lateral measurement to the
proportional increase in length.

Conclusion:

The primary objective of this lab which was to develop a working knowledge of strain gage and
associated instrumentation, and to learn how to apply it to determining pressure inside an
unopened container of carbonated drink was fulfilled. The objective to learn to use a caliper and
micrometer for precision measurements was also achieved. The calculated hoop pressure is
537.84 kPa while the axial was calculated to be 2958 kPa. This variation may be caused due to
many reasons. One of the possible reasons is the inconsistency in the diameter during different
trials. It can be seen form the above table that the error due to thickness of the can is also
significant. The orientation of the gages may also affect the error and there are numerous other
errors that could have impacted our results. However, the primary objectives of the experiment
were fulfilled.

10 | P a g e

You might also like