Design, analysis and development of special purpose tools for composite manufacturing efficiency improvement

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 04 Issue: 07 | July-2015, Available @ http://www.ijret.org 115
DESIGN, ANALYSIS AND DEVELOPMENT OF SPECIAL PURPOSE
TOOLS FOR COMPOSITE MANUFACTURING EFFICIENCY
IMPROVEMENT
Naveen Bharadishettar1
, Yadavalli Basavaraj2
1
PG student, Dept of Mechanical Engg, Ballari Institute of Technology and Management Bellary, Karnataka, India
2
Professor and Head, Dept of Mechanical Engg, Ballari Institute of Technology and Management Bellary, Karnataka,
India
Abstract
Advanced materials are widely used in high performance structures that profit from their high strength; high stiffness and low
weight .Composite products are realized using different manufacturing processes to meet the functional, production rate, size and
shape of the part. Now days most of the industries are mainly focusing on improvements in manufacturing techniques to reduce
efforts as well as material wastage. Wet layup, compression molding, resin transfer molding, filament winding, pultrusion are the
some of the well established process in composites. The present work mainly focused on some of the improvements in layup
process and filament winding process. In layup process improvements, design and development of special purpose tool for noodle
preparation (i.e. Radius fillers) for composite T, I, H sections is carried out. Comparison made on the existing and proposed
design for the feasibility. In filament winding process improvements, design and development of special purpose mandrel for
complex shaped filament wound composite parts is carried out based on the industrialization concept. Later part of work extended
to the study of impact of winding angle on the filament wound composite part using NASTRAN software.
Keywords: Layup, Noodle preparation (i.e. Radius fillers), Filament winding, Mandrel, Winding angle, NASTRAN.
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
Continuous advances in the composite manufacturing
technique have lead to significant market acceptance. But
current challenge is to reduce human efforts as well as
material wastage. A general composite manufacturing
process is labor incentive, requires high skilled operators
and uses expensive non degradable material. The effort to
produce economically attractive composite components
desire several innovative manufacturing techniques.
Presently, industries mainly focusing on improvements in
the composite manufacturing techniques which can cause
cost effective as well reduction in human efforts. Especially
for composites, that the improvement in manufacturing
technology alone is not enough to overcome the cost hurdle.
It is essential that there be an integrated effort in design,
material, process, tooling, quality assurance and
manufacturing processes.
Most of the aircraft and marine structures are manufactured
by layup process using prepreg. During manufacturing of
structural members like T, I, H sections forms cavity at the
interface between web and flanges due to flexibility
disadvantages of prepreg. These cavities cause induced
cracks or premature cracks when member subjected to
loading. Composite radius filler “a noodle” of present work
better meets the challenges faced with such structural
members. To meet these requirements, it is necessa9ry to
think on performing unique operation for multiple parts,
proposal for new tool to optimizing the process to reduce the
operational cost and improving the quality. Lot of works
carried out in last few years on the Radius fillers from the
most of the aircraft industries. Stanely W. Stawski [1]
investigated radius fillers are manufactured using laminated
board made up of plurality of the composite materials
having substantially unidirectional fibers. Panagiotis E.
George, Kirik B.Kajita, Barry P.Van west [2, 3] proposed a
plurality of woven fabrics used to form composite radius
fillers. Stanely W. Stawski [4] introduced another new idea
about radius fillers i.e. inflatable radius fillers.
Filament winding is another composite manufacturing
technique especially for cylindrical and spherical parts.
Most of the uniform thickness parts are manufactured using
conventional mandrel. But some special cases like complex
shaped inner part profile and an asymmetric parts leads to
machining which causes a material wastage, cost incentive
and lower production rate using conventional mandrel. To
overcome these problems, it‟s necessary to think on the
special purpose mandrel for such parts.
As literature says, there are different types of mandrels are
used based on the geometry of the part to be produced. For
special cases like inner step profile, asymmetric parts, Kevin
Davis, James A. Yorgason introduced about inflatable
mandrel. Which have disadvantages over highly complex
parts [5]. Jhon C. Brooks investigated inflatable, collapsible
mandrel for constructing a filament wound pressure vessels
[6]. But typically plain mandrel, segmented mandrel and
molded mandrel are widely used by the industries now days.

_______________________________________________________________________________________
In this paper suitable type of mandrel is chosen among plain,
segmented and moldable mandrel based on the
industrialization concept.
Last part of this paper extended to the impact of winding
angle the strength of the filament wound composite tube.
P.Satheeshkumar reddy, Ch. Nagaraju, T.Hari Krishna
worked on design and analysis of filament wound composite
tube under pure and combined loading using ANSYS tool
[7]. Balya Bora investigated characteristics of filament
wound composite tube under combined loading condition.
Winding angle, level of orthotropy and various ratios of
loading condition are main concern of that study [8-9]. The
present work gives impact of winding angle on stresses and
strains of the part and also concentrated on the winding
angle optimization. This analysis is performed using
NASTRAN.
2. LAYUP PROCESS IMPROVEMENT
Layup process is usually preferred for thin walls. Most of
the structural members like T, I, H stiffened composite
panels are manufactured using prepreg layup process.
Fig- 1: Typical Composite T stiffener showing “Radius
filler”
Composite T, I. H sections are made by binding two „C‟ or
„U‟ channel tools together to form a web with flanges. These
channels generally comprises of prepreg plies of different
orientation. Due to flexibility disadvantages of forming
sharp corners by its own from the prepreg, tool are designed
in predetermined radius form at the sharp edge as shown in
Fig-2. Laid prepreg plies are bent according to the
predetermined radius form.
When the two „C‟ or „U‟ tools are joined at the webs after
layup, a „dimple‟ occurs along the flange because of these
radii. When such structures undergo various types of loading
condition, distortion occurs due to gap at the interface
between web and flange. Usually structures should have stiff
enough to resist the loads. So, voids or dimples formed
during the manufacturing composite stiffeners are filled by
using „a NOODLE‟ (i.e. Radius filler).
Hence, noodles are necessary for composite spars and ribs to
fill the gap formed at the interface between the flanges and
web. These radius fillers or “noodle” must have structural
adequate, easy processing and less expensive to produce.
Such designs of noodles often force to undergo post
manufacturing using special purpose tool. Manufacturing of
these fillers slows the production flows, increases the cycle
time and increases the cost.
Fig-2: Composite T- Stiffener processing.
2.1 Noodle Preparation Technique
Most of composite manufacturing industries have their own
way of processing noodles. Using hydraulic, pneumatic,
automated, other mechanisms are used to form these radius
fillers. Usually spars and ribs are larger in length, such
structural members requires discontinuous fillers are
required in order to avoid discontinuity in members. Present
noodles pressing tools have less capacity in length
parameter. So, present work concentrated on disadvantages
over conventional systems and also focused on other
parameters like operational time, production rate and
improvements in quality.
2.1.1 Design of Noodle Pressing Tool
Concept of Pneumatic compaction system:
New concept of pneumatic compaction system is proposed
for noodle pressing. The output from the proposed design is
identified the following need:
 Increase in the length of the noodle.
 Development of special purpose tool replaces
conventional tools.
 Optimization of process parameters like operational
cost, production time and improvements in the
quality.
In this system (as shown in Fig-3), pressure is applied on the
mold tool pneumatically. Air is supplied to the inflatable
tube which is located inside the rectangular frame. The tube
get expands and pressure is applied by means of flat plate on
the rolled prepreg. Based on the noodle shape and size, mold
tool is designed. Once plate come in contact with the mold
tool, rolled prepreg is turned into a required shape. Max
pressure on the mold tool is assumed to be 2 bars. For air
intake and out take air values are fixed at the top of the
frame. External air connections are connected to the air
values. Mold tools are designed in such way that, mold is
movable using roller bearings for material loading and
unloading.

_______________________________________________________________________________________
2.1.2 CATIA Model.
Fig-3 CATIA Model for pneumatic compaction system
Specifications:
Max capacity: 2 bar pressure.
Length of the system: 2000mm
Width of the system: 300mm
Table – 1: Parts specifications
Parts Dimension
(mm)
Material Feasibility
Rectangular
Frame
250x150x12 Mild steel Good in
strength
Die pattern
130x100
Aluminum
Flexibility
in changing
the pattern
Inflatable
tube
130x100x10 Rubber Ease for
Expansion
Press plate 130x7 Composite
Light
weight, Less
hanging
effect
Air values STD STD
Less cost
and
Available as
Std part.
Roller
Bearings
D=58, d=28,
B=16
STD
Max
capacity and
easy for
movement.
Handles STD STD Easy to
hold.
2.2 Comparison between Existing and Proposed
Design
At present, hydraulic hot press is used for the noodle
pressing. As discussed earlier, length of the noodle stick
manufactured using existing system is only 0.5 meters. But
most of the structural members have a length more than 2
meters. In order to meet 2 meter lengthened T section, 4 no
of 0.5meter noodles are required. Hence, it takes more
production time as well effects on cost and quality. Using
pneumatic compaction system structural properties like
structural integrity, distortion and structural strength are
improved. Proposed design focused on increase in the length
of the noodle stick. So can adversely reduces the production
time and cost.
Comparison made between existing and proposed design by
considering time factor to manufacture 2 meter length
noodle stick.
As per hydraulic hot press, (for 0.5m)
Tool preparation time = 15min.
Prepreg cutting time = 2min.
Rolling of prepreg = 2min.
Heating up of tool = 15min at 60̊ C.
Other processing time = 5min.
Total time taken for manufacturing 0.5m noodle stick =
40min. But tool preparation time and heating time is not
included for the further processing. Therefore total time
taken to prepare a 2meter noodle stick = 40+9+9+9 = 67
minutes. For pneumatic compaction system, it only takes
40minutes to prepare 2 meter lengthened stick.
3. FILAMENT WINDING PROCESS
IMPROVEMENT
Filament winding is special type of composite
manufacturing technique especially for cylindrical and
spherical parts. Simple process of wrapping a mandrel with
fibers becomes complicated when the composite part is
comprised of a compound-curved shape, inner step profile in
the part and asymmetric parts. Such geometries of filament
wound composite parts problematic for mandrel extraction.
Internal mandrels for complex shaped composites require
extensive design considerations to ensure the mandrel can
be removed after the composite is cured. This makes
mandrels very expensive to develop, difficult to fabricate,
and time-consuming to remove from a cured, complex
composite part.
When composite parts like stepped cylindrical profile and
asymmetric parts are fabricated on a conventional mandrel,
additional machining to be done on the part to get desired
part profile. Typically it is not advisable for machining in
order to get desired profile in the part because as it leads to
cost incentive, decrease in the strength due to fiber
discontinuities also lowers the production rate. It is more
desirable to remove the mandrel without affecting the part.
Hence winding the part to net shape by introducing a
stepped mandrel and other accessories will reduce the cost.
This paper focused on the mandrel solutions for such parts
considering different conceptual design related to plain
mandrel, segmented mandrel and molded mandrel. Mandrel
solution is given on basis of industrialization concept
considering mandrel deflection, material consumption and
winding time calculations related to filament winding
process. Industrialization is another best tool for design
feasibility.

_______________________________________________________________________________________
3.1 Types of Mandrel
Any changes in the pattern or part in filament winding
process needs an alternative solution for the mandrel.
Solution for such parts gives rises to distinguished mandrel
design. There are three different types of mandrels are
studied for the feasibility.
3.1.1 Plain Mandrel
Plain mandrels are nothing but conventional mandrel having
uniform diameter. These mandrels are limited to shapes with
a uniform inner diameter and symmetric parts. But some
complex parts like inner step profile, non uniform and non
symmetric parts suggests to additional Machining
operations. Manufacturing of such parts involves resin
impregnated fibers wound on the rotating mandrel until
maximum part thickness is achieved. Later proper
machining is accomplished in order to get a required part
profile. Salient features of plain mandrels are easy in
extracting the mandrel from the part as mandrel having
uniform diameter.
3.1.2 Segmented Mandrel
Segmented metal mandrels are accounted in order to achieve
an inner step profile of the part. These types of mandrel are
consisting of core mandrel and split mandrel. Core mandrel
and split mandrel together assembled to get exact part
profile and then assigned for winding. Grooves provided on
the core mandrel helps to fit together with split mandrels.
Split mandrels are of converging and diverging shapes.
These converging and diverging split mandrels are arranged
in succession. This arrangement benefits for easy removable
of mandrel from part. Initially core mandrel is disassembled
from the split mandrel. However, these are difficulty to
remove if the end openings are small.
3.1.3 Molded Mandrel
Moldable mandrels are the specially employed for unshaped
wound parts comprises of moldable material along with
plain mandrel. Molded mandrels have mold patterns on the
plain mandrel which is exact replica of the inner part profile.
Exact part profile can be achieved using this type of
mandrel. Moldable mandrels may be constructed in any
conventional molding procedure which is pre processing
technique comprises of sand particles and polyvinyl acetate
in 100:9 ratio. Extraction of mandrel from the part is done
using water. Hence it‟s also called as water soluble mandrel.
3.2 Part Description
As shown in the Fig 3-1 is considered for the selection of
proper class of mandrel. Industrialization is another best tool
for design feasibility. Fig 3-1 represents complex shaped
part have inner step profile as well outer curved shape
manufactured from filament winding process. Mandrel
deflection, material consumption and winding time are
considered.
Fig-4: Complex shaped filament wound composite part
3.3 Selection of Mandrel
There are three different types of mandrel are used for
filament wound composite tubes. Only plain mandrel and
molded mandrel are considered for the industrialization.
3.3.1 Mandrel Deflection
In this section, mandrel deflection is calculated for plain and
molded types by considering various lengths. Based on the
results obtained from above calculation feasible mandrel is
chosen.
Length of the mandrel is varied from 1 to 6m.
Below table shows mandrel deflection results for plain and
molded type.
Table – 2: Mandrel deflection results
Length of the
mandrel (m)
Deflection (mm)
Plain mandrel Molded mandrel
1 0.35 0.29
2 0.97 0.57
3 4.94 3.52
4 15.96 11.93
5 38.09 30.58
6 78.87 65.52
Fig-5: Deflection v/s Length of the mandrel

_______________________________________________________________________________________
3.3.2 Material Consumption
Material required to manufacture a part shown in the Fig – 4
using plain mandrel and molded mandrel is calculated.
Material consumption is another parameter for selection of
mandrel. Length of mandrel is varied and corresponding
material consumption is intended.
Material weight = volume of wounded material x Density of
material used. Below table represents a comparison
between the plain mandrel and molded mandrel considering
material consumption parameter.
Table – 3: Material Consumption results
Length of the
mandrel (m)
Material consumption (Kg)
Plain Mandrel Molded
mandrel
1 33.33 7.81
2 66.61 29.88
3 99.92 59.78
4 133.22 89.68
5 166.53 119.58
6 199.84 149.84
Fig -6 : Material consumption v/s Length of the mandrel
3.3.3 Winding Time
Mandrel selection using winding time calculation is another
new concept in this paper. The time taken to complete
winding for given part (Fig -4) effects on the production and
processing. Time taken to complete the winding is different
for different types of mandrel. Here for given part, Hoop
and helical winding patterns are considered.
For Plain mandrel:
Mandrel diameter = 36mm
Length of the mandrel = 6m.
From geometry:
Sin θ = Tan θ
Sin 45 = Y = 0.036 m.
Hypo = X = 50.91mm
For one revolution = 72mm of length is covered.
Total winding length in 6m mandrel is =4800mm
No of revolution required for full length of winding =
(4800/72) = 66.66 revolutions.
Speed of the mandrel =100 rpm.
Time taken for one pass
= 0.666min. = 39.96 seconds.
Length of the strand covered for 72mm length of the
mandrel = 159.93 mm.
X =
X =
X = 159.93mm
No of passes
=
= 5.091 passes. (Approximately = 6)
Time taken to complete one layer of 45̊ winding = 6 x 39.96
= 240 sec.
= 4minutes.
Similarly, for hoop winding
i.e. ±88̊ ;

_______________________________________________________________________________________
From geometry,
Sin θ = Tan θ =
Sin 88 = 1.257 =
Hypo = X = 36.02mm. Y =1.257 mm
Width of the strand = 10mm
I.e. Y = 10+1.257 = 11.257mm
For one revolution, 11.257mm of winding length is covered.
Total winding length = 4800mm.
No of revolution required full length of winding =
=426.40 revolutions.
Speed of the mandrel = 100rpm. =
= 4.26 minutes.
Total time taken to wind one complete layer of 88̊ winding =
4.26 minutes.
Each layer thickness calculations:
Bandwidth of the fiber strand = 10mm
No of strands = 12
No of fiber filaments in each strand = 50
Fiber diameter range from = 3 to 20 micrometers.
Each layer thickness =
Each layer thickness =
= 1.2mm
Winding patterns for given part is alternate 3 layers of
helical winding and 2 layers of hoop winding.
Total part thickness is = 57 mm
No of layers required to achieve part thickness is = 69
(Total 24 Hoop and 36 helical windings are compulsory)
Time for Helical winding = 36 x 4 = 144 minutes.
Time for Hoop winding = 24 x 4.28 =102.72 minutes.
Total time taken to complete 4800m winding is 246.72
minutes. I.e. 4.106 hours.
Based on the no of layers and winding patterns total winding
time is formulated and below table represents the winding
time for plain mandrel and molded mandrel for different
lengths of mandrel.
Table- 4: Winding time results
Length of
the
mandrel
(m)
Winding time (in hours)
Plain Mandrel Molded mandrel
1 0.411 0.096
2 0.821 0.368
3 1.642 0.983
4 2.463 1.657
5 3.284 2.358
6 4.106 3.078
Fig -7 Winding time v/s mandrel length
4. ANALYSIS OF FILAMENT WOUND TUBE
Analysis performed on the filament wound composite tube
using HYPERMESH and NASTARN software is the one of
the leading parts of these studies. In the most of these
studies the effects of winding angle on the strength of
filament wound composite tubes are investigated. These
investigation results are compared with ANSYS results and
briefly explained about FEA simulation of filament wound
composite tube using NASTRAN.
4.1 Modeling of Composite Tubes
Structural analysis is performed in order to investigate the
behavior of layered orthotropic tubes with different
materials. Level of orthotropy is also performed considering
same winding angle with different materials. Internal
pressure loading conditions are considered. The model is
prepared with MAT8 and PCOMP card property in
NASTRAN. As shown in Fig – 6 this model is constraint at
one node at the end of the tube in all degrees of freedom in
order to prevent instability in finite element analysis and
internal pressure is applied at the inner surface of the tube.
Dimension of the tube used in this study are given in Table-
3.
Table -5: Dimension of the tube
Length of the tube 400mm
Average radius 60.565mm
Tube thickness 1.13mm
No of layers 6
Each layer thickness 0.1883mm

_______________________________________________________________________________________
4.2 Results and Discussion
In this analysis, Carbon/Epoxy, E-Glass/Epoxy tubes are
subjected to internal pressure of 12.6MPa. All deformation
and stresses in corresponding directions are collected. Result
validation is done, by performing similar analysis to the
experimental and ANSYS results performed in literature
[10]. NASTRAN and ANSYS results are very close to the
experimental results, although some scattering involved in
the experimental results. It‟s important to note that FEA
results observed for composite analysis is stiffer than the
experimental. Experimental results and material properties
are taken from the [10].
Fig-8: FEA Model of filament wound composite tube using
HYPERMESH.
Table -6: Validation results
Test
Type
Material Experim
ental
results[
FEA
results
using
ANSYS
FEA results
using
NASTRAN
Internal
pressure
of 12.6
MPa
Carbon/
Epoxy
Axial
Strain =
2.290E-3
Axial
Strain=
2.023E-3
Axial
Strain=
2.430E-3
Hoop
Strain=
7.665E-3
Hoop
Strain=
6.251E-3
Hoop
Strain=
6.6081E-3
(a)
(b)
Fig – 9(a) (b) NASTRAN Results
In the first analysis, 6 layers, ±60̊ wound, C/Ep tube
subjected to pure internal pressure of 12.6MPa. Stresses and
strains are obtained for each layer. Then in order to
investigate the effect of winding angle on stresses and
strains in each layer, the same analysis is performed for ±75̊
wound, 6 layer, C/Ep tube. Finally in order to investigate the
effect of level of orthotropy , the same procedure is
performed for 6 layer, ±60̊ wound, Eg/Ep tube.
Stress and strain levels are represented as shown in Fig -10
to Fig- 14

_______________________________________________________________________________________
Fig -10 :Axial strain vls Layer No
Fig -11 : Hoop strain v/s Layer No
Fig – 12 : Shear stress v/s Layer No
Fig – 13 :Axial stress v/s Layer No
Fig – 14 : Radial stress v/s Layer No.
In second type of analysis, composite tube comprises of
multiple of layers wound at different angles subjected to
internal pressure loading is considered. Three different
winding angles are considered. In previous analysis only
effects of repeating winding angles and level of orthotropy
is studied. Three different winding patterns for three
different tubes are chosen for the analyses. An internal
pressure loading condition is considered.
Table 3.8 Different winding patterns for Tube 1, 2 and 3.
Layer No Tube 1 Tube 2 Tube 3
1 25 25 90
2 -25 -25 -90
3 45 55 45
4 -45 -55 -45
5 90 90 25
6 -90 -90 -25
Following results obtained and represented from Fig 15 to
Fig - 18

_______________________________________________________________________________________
Fig – 15: Radial stress v/s Layer No
Fig – 16: Hoop stress v/s Layer No.
Fig -17 : Axial stress v/s Layer No
Fig -18 : Shear stress v/s Layer No.
5. CONCLUSION
Most of the composite manufacturing industries have
continuously seeking for the advanced or/and improved
manufacturing techniques in order to meet global market
acceptance. In this scope of the work, layup process
improvement related to noodle pressing tool has carried out.
Proposed tool has improved manufacturing capabilities like
length of the noodle stick which adversely effects in the
production rate and most importantly focused on structural
integrity and distortion factors. Also some of the factors also
discussed related to quality. Another development related to
filament winding has proposed. pain mandrel, segmented
mandrel, molded mandrel are considered for the complex
shaped parts and selection of mandrel is carried out based on
the industrialization concept. Mandrel deflection, material
usage and winding time are performed for the feasibility.
Based on result as shown in (Fig – 5, 6, 7) molded mandrel
is chosen for given part. Third part of this paper is worked
on effect of winding angle on the filament wound composite
tube using NASTRAN software. Results as shown in Fig 9
to Fig 18.
REFERENCES
[1]. Stanely W. Stawski “Method of forming composite
radius fillers” United States patent, Dec 6, 1988; US4,
789,594.
[2]. Panagiotis E, George. Kirk B.Kajita, Barry P.Van West
“Laminated composite radius fillers” United States patent,
May13, 2013; US 6,562,436.
[3]. Panagiotis E, George. Kirk B.Kajita, Barry P.Van West
“Laminated composite radius fillers” United States patent,
May13, 2013; US 6,709538.
[4]. Stanely W. Stawski “Method of forming composite
radius fillers” United States patent, Dec 20, 2014;
US2014/0341641.
[5]. Kevin Davis; James A. Yorgason. “Method constructing
a inflatable mandrel” United states patent, Nov 9, 1993; US
5,259,901.

_______________________________________________________________________________________
[6]. Jhon C. Brooks “Method of Fabricating a collapsible
mandrel structure to be used in manufacturing reinforced
hollow tubes” United states patent, Aug 4, 1987; US
4,684,423.
[7]. P. Sateesh Kumar Reddy, Ch.Nagaraju, T. Hari Krishna
“Optimum design and analysis of filament wound composite
tubes in pure and combined loading”. International journal
of Engineering Research and Technology” Vol 1 issue 8,
October 2012.
[8]. Bora Balya, “Design and analysis of filament wound
composite tubes” Master of science thesis, Middle east
university, 2004.

Design, analysis and development of special purpose tools for composite manufacturing efficiency improvement

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Design, analysis and development of special purpose tools for composite manufacturing efficiency improvement (20)

More from eSAT Journals (20)

Recently uploaded (20)

Design, analysis and development of special purpose tools for composite manufacturing efficiency improvement