1. Octagonal Double StepOctagonal Double Step
Lap Joint RepairLap Joint Repair
Group 3:Group 3:
Zhi ChenZhi Chen
Vincent DuongVincent Duong
Qunjie (Jay) GuanQunjie (Jay) Guan
Rory MillerRory Miller
Spring Quarter 2008Spring Quarter 2008
SE 171 Aerospace Structures Repair – UC San DiegoSE 171 Aerospace Structures Repair – UC San Diego
Dr. KimDr. Kim
La Jolla, CA 92092La Jolla, CA 92092
2. Design RequirementsDesign Requirements
The repaired panel has to have at least 15%The repaired panel has to have at least 15%
more strength than the unrepaired panel.more strength than the unrepaired panel.
Maximize Performance IndexMaximize Performance Index
3. Design ConsiderationsDesign Considerations
Priority given to PI maximizationPriority given to PI maximization
ManufacturabilityManufacturability
Maximizing peak load while keeping theMaximizing peak load while keeping the
repair light (low mass)repair light (low mass)
Minimize the adhesive shear stress andMinimize the adhesive shear stress and
NN0k0k
4. Modeling and AnalysisModeling and Analysis
Preliminary computer modeling andPreliminary computer modeling and
analysis were conducted to estimate theanalysis were conducted to estimate the
stress distribution while the panel isstress distribution while the panel is
undergoing tensile loadingundergoing tensile loading
The preliminary modeling of theThe preliminary modeling of the
unrepaired panel gave clues as to whatunrepaired panel gave clues as to what
repairs are suitablerepairs are suitable
6. Modeling and AnalysisModeling and Analysis
From the preliminary model, it wasFrom the preliminary model, it was
observed that the stress concentrationsobserved that the stress concentrations
are greatest at the ends of the crackare greatest at the ends of the crack
Therefore, proceed to decrease theseTherefore, proceed to decrease these
stress concentrations by placing two holesstress concentrations by placing two holes
at the ends of the crackat the ends of the crack
Resulting peak stress concentrationsResulting peak stress concentrations
decreased by about 50%decreased by about 50%
8. Modeling and AnalysisModeling and Analysis
Step lap joint reducesStep lap joint reduces ττaa(x) and N(x) and N0k0k(x),(x),
therefore, the repair will be in the step laptherefore, the repair will be in the step lap
joint configurationjoint configuration
The number of plies is determined byThe number of plies is determined by
comparing the failure strengths of thecomparing the failure strengths of the
inner and outer adherendsinner and outer adherends
Dimensions of the plies should be as closeDimensions of the plies should be as close
to the recommended 1:50 ply drop ratio asto the recommended 1:50 ply drop ratio as
possiblepossible
9. Modeling and AnalysisModeling and Analysis
_
_
52000
42000
0.0760 .
0.012 .
ult i
ult o
i
ply
o ply
psi
psi
t in
t in
t n t
σ
σ
=
=
=
=
= ×
_ _
52000 0.0760 42000 0.012
7.8
8
ult o o ult i it t
n
n
n plies
σ σ=
× = × ×
=
= 4 plies on each side
10. Modeling and AnalysisModeling and Analysis
The shape of the patch was narrowedThe shape of the patch was narrowed
down to either a hexagon or ellipsedown to either a hexagon or ellipse
Hexagon is easier to manufacture thanHexagon is easier to manufacture than
ellipse; however, priority will be placed onellipse; however, priority will be placed on
the robustness of the peak loadthe robustness of the peak load
The modeling of the repair patch designsThe modeling of the repair patch designs
were then conductedwere then conducted
13. Modeling and AnalysisModeling and Analysis
From the previous two models, notice thatFrom the previous two models, notice that
the hexagonal patch is more robust thanthe hexagonal patch is more robust than
the ellipse/ovular patchthe ellipse/ovular patch
Hexagonal patch is easier to manufactureHexagonal patch is easier to manufacture
However, there are isolated locations thatHowever, there are isolated locations that
have spikes in the stress concentrationshave spikes in the stress concentrations
(i.e. corners of the hexagon)(i.e. corners of the hexagon)
Remedy stress concentration problem byRemedy stress concentration problem by
creating an octagonal repair patchcreating an octagonal repair patch
14. Final DesignFinal Design
Octagonal patchesOctagonal patches
8 plies total, 4 plies on each side8 plies total, 4 plies on each side
Ply Drop Ratios from top to bottom arePly Drop Ratios from top to bottom are
1:40, 1:30, 1:20, and 1:201:40, 1:30, 1:20, and 1:20
16. Modeling and AnalysisModeling and Analysis
The failure load is predicted to beThe failure load is predicted to be
2518 lb2518 lbff/in (2N/in (2Nxx), or 440.987 N/mm), or 440.987 N/mm
The failure load of the unrepaired panelThe failure load of the unrepaired panel
was 74 N/mmwas 74 N/mm
17. Modeling and AnalysisModeling and Analysis
After the failure load is predicted, theAfter the failure load is predicted, the ττaa(x)(x)
and Nand N00(x) is then calculated for the(x) is then calculated for the
proposed repairproposed repair
PeakPeak ττaa(x) is 6365 psi when N(x) is 6365 psi when Nxx is 1259is 1259
lblbff/in/in
The predicted failure mode was adhesiveThe predicted failure mode was adhesive
shearshear
The detailed dimensions of the repairThe detailed dimensions of the repair
patch are determined bypatch are determined by ττaa(x) and N(x) and N00(x)(x)
plotsplots
18. Modeling and AnalysisModeling and Analysis
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0
1000
2000
3000
4000
5000
6000
7000
x [in.]
AdhesiveShearStress[psi]
Figure 1. Adhesive Shear Stress at Failure (Nx = 1259 lb/in.)
X: 1.32
Y: 6365
19. Modeling and AnalysisModeling and Analysis
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0
200
400
600
800
1000
1200
1400
x [in.]
AppliedLoad[lbf/in.]
Figure 2. Outer Adherend Load at Failure (Nx = 1259 lb/in.)
X: 1.32
Y: 1259
20. Manufacturing ProcessManufacturing Process
The crack needs to be sealed with micro-balloons, turning the crackThe crack needs to be sealed with micro-balloons, turning the crack
into a butt jointinto a butt joint
21. Manufacturing ProcessManufacturing Process
The repair patch was produced via theThe repair patch was produced via the
wet lay up method with fiberglass/epoxywet lay up method with fiberglass/epoxy
22. The surfaces surrounding the crack areThe surfaces surrounding the crack are
sanded down to provide better mechanicalsanded down to provide better mechanical
interlocking faces for adhesiveinterlocking faces for adhesive
Manufacturing ProcessManufacturing Process
27. Manufacturing ProcessManufacturing Process
The end tabs were originally composed ofThe end tabs were originally composed of
Aluminum bonded by JB weld.Aluminum bonded by JB weld.
However, The end tabs became unreliableHowever, The end tabs became unreliable
after holes were drilled in.after holes were drilled in.
32. ResultsResults
The predicted failure load is 2518 lbThe predicted failure load is 2518 lbff/in (2N/in (2Nxx),),
or 440.987 N/mm.or 440.987 N/mm.
Max.Max. ττaa is 6365 psi.is 6365 psi.
The actual failure load is 3124 lbThe actual failure load is 3124 lbff/in, or 547/in, or 547
N/mm.N/mm.
The unrepaired panel’s failure load is 74The unrepaired panel’s failure load is 74
N/mm.N/mm.
The mass of the repair is 16.6g.The mass of the repair is 16.6g.
The PI is approximately 29.The PI is approximately 29.
33. Comparison of 2Comparison of 2ndnd
Ply 0 and 90Ply 0 and 90
Degree LayupsDegree Layups
0 Degree
90 Degree
35. ResultsResults
The possible failure modes includeThe possible failure modes include
adhesive failure, inner and outer adherendadhesive failure, inner and outer adherend
failure, and lastly, interfacial failurefailure, and lastly, interfacial failure
The predicted failure mode was adhesiveThe predicted failure mode was adhesive
shearshear
The specimen exhibited adhesive failureThe specimen exhibited adhesive failure
and outer adherend failureand outer adherend failure
39. Conclusion and RecommendationConclusion and Recommendation
The actual failure load achieved was 3124The actual failure load achieved was 3124
lbf/in, or 547.10 N/mm.lbf/in, or 547.10 N/mm.
Some discrepancies included theSome discrepancies included the
manufacturing process and human error.manufacturing process and human error.
Ply orientation was forgotten about, so one or two plies might be 90Ply orientation was forgotten about, so one or two plies might be 90ºº
degreesdegrees
The patches could not be cut to exact dimensions due to the given toolsThe patches could not be cut to exact dimensions due to the given tools
and the glass-fiber cloth’s flexibility and ease of warpand the glass-fiber cloth’s flexibility and ease of warp
The analysis was done for the original panel dimensions instead of theThe analysis was done for the original panel dimensions instead of the
modified tested panel specimen.modified tested panel specimen.
The weight of the adhesive butt-joint in the crack was not accounted.The weight of the adhesive butt-joint in the crack was not accounted.
40. Conclusion and RecommendationConclusion and Recommendation
Recommended Approaches:Recommended Approaches:
Do not use Aluminum and JB Weld end tabsDo not use Aluminum and JB Weld end tabs
Map the 0Map the 0º degrees and 90º degreesº degrees and 90º degrees
coordinate system onto the plies before anycoordinate system onto the plies before any
manufacturingmanufacturing
Design a mold for each step-lap ply onDesign a mold for each step-lap ply on
AutoCAD with exact dimensions, and then cutAutoCAD with exact dimensions, and then cut
it out on acrylic, which would be placed overit out on acrylic, which would be placed over
the glass-fiber cloth to make cutting easierthe glass-fiber cloth to make cutting easier
Editor's Notes
Tentative, add/edit slides as necessary, presentation time should be around 20 – 30 minutes.
Modeling and Analysis = Technical development on outline
Manufacturing process belongs to the technical development section of the outline.