EMShip – 8th cohort

Composite materials

Report – Tensile test

GROUP 5

MARCO TÚLIO DAL CÓL DOS SANTOS

TUMUL PURWAR

ARRSHAN S P

November / 2017

1
1 Objective

The purpose of this report is to present the aspects of designing and manufacturing a sample of a
composite material that is subjected to an unidirectional tensile load.

2 Given problem

The given problem requires to design a shell for containing petroleum products. This shell will have a
volume of 350 liters.

Figure 1: Shell geometry.

The table 1 show the data of the cylinder.

Table 1: Data.

Internal pressure
100
(N/cm2)
Length (m) 6
Diameter (cm) 90
min 1.5 with top
Thickness (mm)
coat
Factor of Safety 1.4

• Characteristics of the components

o Resin
Specific weight: 1.116 g/cm³;
Tensile strength: 520 kg/cm²;
2
 E modulus: 38,000 kg/cm²;
At 25°C and with 2 % hardener PMEC: hardening after 20 minutes.
o Gel coat
Specific weight: 1,056 g/cm³;
Thickness: 0.5 mm.
o Roving fabric
Surface weight: 350 g/m²;
E modulus: 75,000 MPa;
Tensile strength: 2,500 MPa;
Specific mass of glass: 2,560 kg/m³;
Thickness: 0.35 mm.
o Mat
Surface weight: 450 g/m²;
E Modulus: 75,000 MPa;
Tensile strength: 2,500 MPa;
Specific mass of glass: 2,560 kg/m³;
Thickness: 0.65 mm.

2.1 Calculation

 Estimation of Yield Strength and Elastic Modulus

𝛔𝐜 = 𝐤. 𝐤 𝟏 . 𝐤 𝟐 . 𝛔𝐅 . 𝐕𝐅 (%) + 𝛔𝐌 . 𝐕𝐌 (%)

𝐄𝐜 = 𝐤. 𝐤 𝟏 . 𝐤 𝟐 . 𝐄𝐅 . 𝐕𝐅 (%) + 𝐄𝐌 . 𝐕𝐌 (%)

Where:

Ec : E modulus of the composite (longitudinal)

EF : E modulus of the fiber

3
EM : E modulus of the resin
VF (%) : volume of fibers (%)
VM (%) : volume of resin (%)
k1: misalignment coefficient of the fibers
k2 : coefficient of contiguity of the fibers
k: % in volume of fibers with angle α versus load direction

We can consider K1 and K2 as 1. The table below presents the calculation that has been done using
the Excel spreadsheet.

Table 2: Volume of fiber and mat calculations.

Thickness (mm) Surface weight (g/m2) VF (%) VM (%) k

Roving Fabric 0.35 350 0.391 0.609 0.500
Roving Fabric 0.35 350 0.391 0.609 0.500
Roving Fabric 0.35 350 0.391 0.609 0.500
Mat 0.65 450 0.270 0.730 0.375
Mat 0.65 450 0.270 0.730 0.375
Gel Coat 0.5 528 0.000 1.000 0.000
Gel Coat 0.5 528 0.000 1.000 0.000
∑ 3.35 0.227 0.490

The results of E modulus and the Yield Strength is shown on the Table below.

Table 3: E modulus and Tensile yield strength for each material.

Composite Mat Fiber k1 = k2 K

σ (N/mm2) 197.26 52 2500
1 0.30
E modulus (N/mm2) 41869.08 75000.00 75000

Since the composite should withstand a pressure of 100 N/cm2, it was decided to use 2 layers of gel
coat, 3 layers of roving fabric and 2 layers of mat along with top and bottom layers of gel coat. Therefore,
the total thickness of the composite is 3.35mm. Hence, follow below the Hooper stress calculation.

4
 𝐏 × 𝐝 𝟏𝟎𝟎 ∗ 𝟗𝟎 𝐍
𝑺𝒕𝒓𝒆𝒔𝒔 = = = 𝟏𝟑𝟒𝟑𝟐. 𝟖𝟑 = 𝟏𝟑𝟒. 𝟑𝟐 𝐌𝐏𝐚
𝟐𝐭 𝟐 ∗ 𝟎. 𝟑𝟑𝟓 𝐜𝐦𝟐

Considering the factor of safety of 1.4, we get 188.04 MPa. Hence, the required yield strength should
be higher than that of Hooper stress, which is 188.04 MPa, so that it is sufficient to withstand the design
stress.

3 Manufacturing

The measurements of the specimen required are taken and accordingly the required amount of gel
coat is calculated using its specific weight. 2 % of the calculated weight of the gel coat is added with a
hardener and stirred well for uniform mixture.
The gel coat is applied thoroughly in the given space with a brush and left for 30 minutes for curing.
After that, 3 layers of roving fabric and 2 layers of mat are cut according to the taken measurements. The
amount of resin required for each roving fabric and mat is calculated accordingly for the size of the fabric
and mat by using the respective specific weight.
The roving fabric is kept on the workspace and the resin is applied thoroughly over it using a brush
and made sure it fills all the voids in the fabric. A special type roller is then used to roll over and remove all
the entrapped air bubbles. The specimen is left for some time for curing.
On the next step, a layer of mat is kept above the cured fabric and resin is applied on it thoroughly
filling all the voids in them mat. The roller is then used again to remove any entrapped air.
The above 2 steps are repeated 3 more times to complete the composite with totally 3 layers of
roving fabric and 2 layers of mat in an alternative manner.

As the last step, the top layer of Gel coat is applied thoroughly over the composite specimen and left
for curing.

 Calculation of the gel coat mass

Length = 50.4 cm
Width = 35.3 cm
Thickness of layer = 0.05 cm

5
20% has been added in order to compensate the loses during the manufacturing. Also, an additional 2% of
hardener is required into the mixture.

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐆𝐞𝐥 𝐜𝐨𝐚𝐭 = 𝟓𝟎. 𝟒 ∗ 𝟑𝟓. 𝟑 ∗ 𝟎. 𝟎𝟓 = 𝟖𝟖. 𝟗𝟓 𝐜𝐦𝟑

𝐌𝐚𝐬𝐬 𝐨𝐟 𝐆𝐞𝐥 𝐜𝐨𝐚𝐭 = 𝐕𝐨𝐥𝐮𝐦𝐞 ∗ 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐰𝐞𝐢𝐠𝐡𝐭 = 𝟖𝟖. 𝟗𝟓 ∗ 𝟏. 𝟎𝟓𝟔 = 𝟗𝟑. 𝟗𝟑 𝐠

𝐓𝐨𝐭𝐚𝐥 𝐦𝐚𝐬𝐬 𝐨𝐟 𝐆𝐞𝐥 𝐜𝐨𝐚𝐭 = 𝟗𝟑. 𝟗𝟑 + 𝟐𝟎%(𝐥𝐨𝐬𝐬) + 𝟐%(𝐟𝐨𝐫 𝐡𝐚𝐫𝐝𝐧𝐞𝐫) = 𝟏𝟏𝟐. 𝟕𝟏 𝐠

 Calculation of the percentage of void in fabric

Fabric

Weight of roving fabric per unit area if only fiber is present = 2560 kg/m3 * 0.35mm

Surface weight of roving fabric = 896 g/cm2

% of fiber in roving fabric = 350/896 = 39%

% of void in roving fabric = 1 – 0.270 = 60%

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐯𝐨𝐢𝐝 𝐢𝐧 𝐫𝐨𝐯𝐢𝐧𝐠 𝐟𝐚𝐛𝐫𝐢𝐜 = 𝟓𝟎. 𝟒 ∗ 𝟑𝟓. 𝟑 ∗ 𝟎. 𝟑𝟓 ∗ 𝟎. 𝟔 = 𝟑𝟕. 𝟗𝟐 𝐜𝐦𝟑

 Calculation of the mass of resin required

Length = 50.4 cm
Width = 35.3 cm
Thickness of layer = 0.035 cm

20% has been added in order to compensate the loses during the manufacturing. Also, an additional 2% of
hardener is required into the mixture.

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐭𝐡𝐞 𝐥𝐚𝐲𝐞𝐫 = 𝐋𝐞𝐧𝐠𝐭𝐡 ∗ 𝐖𝐢𝐝𝐭𝐡 ∗ 𝐭𝐡𝐢𝐜𝐤𝐧𝐞𝐬𝐬 = 𝟓𝟎. 𝟒 ∗ 𝟑𝟓. 𝟑 ∗ 𝟎. 𝟎𝟑𝟓 = 𝟔𝟐. 𝟐𝟕𝐜𝐦𝟑

𝐠
𝐌𝐚𝐬𝐬 𝐨𝐟 𝐫𝐨𝐯𝐢𝐧𝐠 𝐟𝐚𝐛𝐫𝐢𝐜 = 𝐒𝐮𝐫𝐟𝐚𝐜𝐞 𝐰𝐞𝐢𝐠𝐡𝐭 ∗ 𝐀𝐫𝐞𝐚 = 𝟎. 𝟎𝟑𝟓𝟎 ∗ 𝟓𝟎. 𝟒 𝐜𝐦 ∗ 𝟑𝟓. 𝟑 𝐜𝐦 = 𝟔𝟔. 𝟐𝟔 𝐠
𝐜𝐦𝟐
6
 𝐦𝐚𝐬𝐬 𝐨𝐟 𝐫𝐨𝐯𝐢𝐧𝐠 𝐟𝐚𝐛𝐫𝐢𝐜 𝟔𝟔. 𝟐𝟔 𝐠
𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐫𝐨𝐯𝐢𝐧𝐠 𝐟𝐚𝐛𝐫𝐢𝐜 = = = 𝟐𝟓. 𝟖𝟖 𝐜𝐦𝟑
𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐦𝐚𝐬𝐬 𝐨𝐟 𝐠𝐥𝐚𝐬𝐬 𝟐. 𝟓𝟔 𝐠/𝐜𝐦𝟑

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 = 𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐥𝐚𝐲𝐞𝐫 − 𝐕𝐨𝐥𝐮𝐦𝐞𝐨𝐟 𝐫𝐨𝐯𝐢𝐧𝐠 𝐟𝐚𝐛𝐫𝐢𝐜 = 𝟔𝟐. 𝟐𝟕 − 𝟐𝟓. 𝟖𝟖 = 𝟑𝟔. 𝟑𝟗 𝐜𝐦𝟑

𝐌𝐚𝐬𝐬 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 = 𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 ∗ 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐰𝐞𝐢𝐠𝐡𝐭 = 𝟑𝟔. 𝟑𝟗 ∗ 𝟏. 𝟏𝟏𝟔 = 𝟒𝟎. 𝟔𝟏 𝐠

𝐓𝐨𝐭𝐚𝐥 𝐦𝐚𝐬𝐬 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 = 𝐌𝐚𝐬𝐬 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 + 𝟐𝟎%(𝐥𝐨𝐬𝐬) + 𝟐%(𝐟𝐨𝐫 𝐡𝐚𝐫𝐝𝐧𝐞𝐫) = 𝟒𝟎. 𝟔𝟏 + 𝟖. 𝟏𝟐 + 𝟎. 𝟖𝟏

= 𝟒𝟗. 𝟓𝟒 𝐠

 Calculation of the percentage of void in mat

Weight of mat per unit area if only fiber is present = 2560 kg/m3 * 0.65 mm

= 1664 g/cm2

Surface weight of mat = 450 g/cm2

% of fiber in roving fabric = 450/1664 = 27%

% of void in roving fabric = 1 – 0.270 = 73%

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐯𝐨𝐢𝐝 𝐢𝐧 𝐦𝐚𝐭 = 𝟓𝟎. 𝟒 ∗ 𝟑𝟓. 𝟑 ∗ 𝟎. 𝟔𝟓 ∗ 𝟎. 𝟕𝟑 = 𝟖𝟒. 𝟒 𝐜𝐦𝟑

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐭𝐡𝐞 𝐥𝐚𝐲𝐞𝐫 = 𝐋𝐞𝐧𝐠𝐭𝐡 ∗ 𝐖𝐢𝐝𝐭𝐡 ∗ 𝐭𝐡𝐢𝐜𝐤𝐧𝐞𝐬𝐬 = 𝟓𝟎. 𝟒 ∗ 𝟑𝟓. 𝟑 ∗ 𝟎. 𝟎𝟔𝟓 = 𝟏𝟏𝟓. 𝟔𝟒 𝐜𝐦𝟑

𝐠
𝐌𝐚𝐬𝐬 𝐨𝐟 𝐫𝐨𝐯𝐢𝐧𝐠 𝐦𝐚𝐭 = 𝐒𝐮𝐫𝐟𝐚𝐜𝐞 𝐰𝐞𝐢𝐠𝐡𝐭 ∗ 𝐀𝐫𝐞𝐚 = 𝟎. 𝟎𝟒𝟓 ∗ 𝟓𝟎. 𝟒 𝐜𝐦 ∗ 𝟑𝟓. 𝟑 𝐜𝐦 = 𝟖𝟎. 𝟎𝟔 𝐠
𝐜𝐦𝟐

𝐦𝐚𝐬𝐬 𝐨𝐟 𝐦𝐚𝐭 𝟖𝟎. 𝟎𝟔 𝐠

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐦𝐚𝐭 = = = 𝟑𝟏. 𝟐𝟕 𝐜𝐦𝟑
𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐦𝐚𝐬𝐬 𝐨𝐟 𝐠𝐥𝐚𝐬𝐬 𝟐. 𝟓𝟔 𝐠/𝐜𝐦𝟑

𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 = 𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐥𝐚𝐲𝐞𝐫 − 𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐦𝐚𝐭 = 𝟏𝟏𝟓. 𝟔𝟒 − 𝟑𝟏. 𝟐𝟕 = 𝟖𝟒. 𝟑𝟕 𝐜𝐦𝟑

𝐌𝐚𝐬𝐬 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 = 𝐕𝐨𝐥𝐮𝐦𝐞 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 ∗ 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐰𝐞𝐢𝐠𝐡𝐭 = 𝟖𝟒. 𝟑𝟕 ∗ 𝟏. 𝟏𝟏𝟔 = 𝟗𝟒. 𝟏𝟓 𝐠

𝐓𝐨𝐭𝐚𝐥 𝐦𝐚𝐬𝐬 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 = 𝐌𝐚𝐬𝐬 𝐨𝐟 𝐫𝐞𝐬𝐢𝐧 + 𝟐𝟎%(𝐥𝐨𝐬𝐬) + 𝟐%(𝐟𝐨𝐫 𝐡𝐚𝐫𝐝𝐧𝐞𝐫) = 𝟗𝟒. 𝟏𝟓 + 𝟏𝟖. 𝟖𝟑 + 𝟐. 𝟐𝟔
= 𝟏𝟏𝟓. 𝟐𝟒 𝐠

7
4 Tensile test

From the composite manufactured, six specimens were taken, two from vertical cut, two from
horizontal cut and two for oblique (45 degrees) cut.
The thickness and width of the specimens are measured and the average for each one is taken, as
follow the Table below.

Table 4: Thickness and width of each specimens.

Thickness (mm) Width (mm)

5V(1) 2.8 20.26

5V(2) 2.93 20.2

5H1 2.78 20.19

5H2 2.95 20.25

5o1 3.03 20.1

5o2 2.83 20.23

The six specimens were submitted to a tensile force on its edges by a speed movement of 2mm/min.
The data was then plotted in a graphic as presented on the Figure bellow.

Table 5: Data of the tensile test.

8
 Figure 2: load x displacement of each specimen.

Table 6: Results of the UTM.

4.1 Discussion of results

The first thing that can be observed is that the standard deviation is quite high. Also, the differences
between specimens along the same direction (1V1 and 1V2, for example) are significantly high, meaning that
the variance is high too.

9
 Table 7: Details of each specimens calculated on Excel spreadsheet.

Component H1 H2 V1 V2 O1 O2 Avarage
Width (mm) 20.3 20.2 20.2 20.3 20.1 20.2 20.2
Thickness (mm) 2.8 2.9 2.8 3.0 3.0 2.8 2.9
Area (mm2) 56.7 59.2 56.1 59.7 60.9 57.3 58.3
Force (N) 9162.1 8506.7 8600.8 7813.8 6145.7 6420.9 7775.0
Stress at Failure
161.5 143.7 153.2 130.8 100.9 112.2
(MPa) 133.7
Strain 0.0284 0.0286 0.0295 0.0276 0.0255 0.0273
Young’s Modulus 5193.2 4739 5686.6 5024.4 3956.8 4109.8

Additionally, according to the theoretical values obtained, the values of Young modulus and tensile
strength for specimens 1H1, 1H2, 1V1 and 1V2 should all be the same, while the values for 1O1 and 1O2
should be lower. What was observed, however, was that the horizontal specimens (1H1 and 1H2) presented
higher tensile strength, followed by the vertical specimens (1V1 and 1V2), while the oblique specimens (1O1
and 1O2) were considerably weaker.

Table 8: Tensile strength comparation.

Tensile strength (MPa)

Specimen
From
From experiment
calculation
1 161.5 188.1
2 143.7 188.1
3 153.2 188.1
4 130.8 188.1
5 100.9 188.1
6 112.2 188.1

As can be seen, all of the results from the experiment are significantly different than the values
expected from the calculations. We cannot know for sure why the results are so different than expected, but
we can list the possibilities:

• We don’t know where the specimens were taken from in the plate; because the specimens are
of significant length, a part of them could have been taken from a region of the plate where the
fabric was not well placed.

10
 • The manufacture process was not very good; The layers of the fabric & mat are only placed in
vertical & horizontal direction and are not placed oblique direction. And some handling errors
like (excess or lack of resin or hardener) might have had a bad impact on the final result,
changing the k1 and k2 coefficients.
• The specifications of the materials used for making the composite were not checked
beforehand; although they were given initially in a sheet of paper, we should always check the
specifications in loco (for example, on the description on the side of the resin container).
• The design is made for a specimen in cylindrical form and the stress is expected to be radial and
axial. But the specimen is tested with only axial direction while the specimen is manufactured as
a flat plate.

We can calculate the actual k1*k2 values for each specimen by just dividing the result obtained by the
theoretical result, since we used k1*k2=1. The following values were obtained.

Table 9: K factor values.

Specimen k1*k2 for Tensile strength

1 0.858
2 0.763
3 0.814
4 0.695
5 0.536
6 0.596

5 Conclusion

The group had almost no previous experience with making composite materials, so it was expected
that the manufacturing process would present problems in the end. Indeed, what was seen was that the
specimens taken from the final product were of poor quality with regards to total strength. However, many
lessons were learned about how to make composites, and the group is certain that, if the material was to be
done again, the result would be better.

11
 From the point of view of education, the group has acquired experience and a good sense of how it
feels like to work with composite materials, and how the process is performed. Since many boats/ships are
made with composite materials, it was very important for the group to get to know this better.

6 Questions

 Which type of manufacturing process would be more suitable for this type of composite? Why?

The best manufacturing process would be the filament winding, because the model requested for
designing is a cylindrical shell, so filament winding is the best option since its composition can be uniformly
distributed. Also, winding in different angles increases the strength in different axis.

 How to increase mechanical strength without increasing thickness of composite?

It is possible to increase the strength by avoiding voids during the manufacturing process, once these
voids are avoided, the mechanical strength of the material increases.
Aligning the layers of fibers with Horizontal, Oblique & in Vertical directions would increase the
strength of the material is different directions.

 How to increase reinforcement content?

To increase the reinforcement content in the composite, the number of filaments for a specific area
during the production of the fabric must be increased. Because once it is done, the volume of voids
decreases at the same time that the resin volume decrease, while the volume of fiber increase.

 What would be the thickness of an Alu 6061-0 plate for the same tensile strength?

The average maximum stress supported by the specimens was of around 132.72 MPa,
with a force of 7775 N, and having a thickness of 3.35 mm and diameter 90 cm.
An Alu 6061-0 material has a yielding strength of 55MPa Hence to calculate the thickness required to
withstand the force of 7775 N at which it fractured is:

7775
𝑃= = 132.72 𝑁/𝑐𝑚2
(2.9 ∗ 20.2)

12
 𝑝∗𝑑 1.327 ∗ 900
σ= = 55.2 =
2∗𝑡 2∗𝑇

𝑇 = 10.18 𝑚𝑚

13