Modeling the Response of Woven Fabric

1
Modeling the Response of Woven Fabric

• J. W. Simons
• ARA, Sunnyvale, CA
• D. Erlich, and D. A. Shockey
• SRI International, Menlo Park, CA
• Presented
• At FAA Tech Center
• July 20, 1999

2
Outline

• Objectives
• Approach
• Yarn model
• Fabric model
• Design model
• Current effort

3
Objectives

• Help design and analyze ballistic impact
  experiments
• Understand response of woven fabric
• Effect of friction
• Boundary conditions
• Yarn properties
• Weave geometry
• Optimize fabric efficiency
• Best protection/weight
• Develop a tool to certify and validate designs

4
Approach

• Develop detailed model to gain understanding
• Yarn model
• Explicitly model yarn geometry
• Simulate several simple tests
• fabric model
• Explicitly model woven fabric
• Simulate impact scenarios
• Simulate different conditions
• Vary design parameters to optimize performance
• Incorporate understanding into a simpler model
  that can be used as a design tool

5
Computational Tools

• LS-DYNA3D
• Network of SGI Workstations
• Computationally intensive
• Multiprocessor machines

6
Modeling Geometry

• Measured geometry from micrographs
• Get yarn cross-section dimensions
• Measure offsets for fill and warp

7
Yarn Properties

• Properties strongly directional
• Strong and stiff along fiber direction
• After it straightens
• Peak stress of 2.8 GPa (28 kb)
• Peak strain of 2.5
• Not stiff in other directions
• Does not take compression

8
Constitutive Model

• Orthotropic Model for yarn
• E 164 GPa in fiber direction
• E 1 in other directions and in shear
• Effective Poissons ratio 0.
• Fiber direction defined by nodal alignment
• Stress-based failure (2.8 GPa)
• Allow failure to take place over 1 strain
• Failure takes a minimum of 10ms
• Element erosion at failure

9
Yarn Model

• Yarn model geometry
• Cross-section from micrographs
• Crimp from interference
• Build up yarn by repeating section

10
Crimped Yarn Pull (axial)

• Small segment of crimped yarn
• Pulled left end at 20 m/s
• Yarn straightens and breaks
• Initial stress is small, peak stress is just over
  30 kb

11
Transverse loading

• Left end displaced upward at 20 m/s
• Large rotations and deformations without
  developing significant stresses
• Stresses develop after yarn straightens out
• Peak effective stress just over 25 kb

12
Impact of Single Yarn

• Model interaction between a projectile and a
  Zylon yarn
• Titanium projectile at 200 m/s
• Two-inch length of yarn
• Calculated effective stress in the yarn
• Estimate displacement wave in the yarn to be only
  about 320 m/s
• much less than sound speed in taut yarn of about
  14,000 m/s

13
Effect of Density

• Increasing the density significantly increases
  the resisting force
• Much of the fabrics effectiveness is due to
  inertia

14
Effect of Crimp

• Crimp just delays the resisting force
• Does not significantly affect the overall
  resisting force

15
Understanding the Woven Fabric Response

• Laboratory Tests
• Yarn pull
• Fabric push
• Interyarn friction
• Get load-displacement curves
• Observe how yarns interact
• Finite Element Calculations
• Develop material model for yarns
• Model weave explicitly
• Generalize to fabric model

16
Detailed Model

• Explicitly model woven fabric
• Model individual yarns
• Geometry from photographs
• Contact and interaction between yarns

17
Weave Model

• Weave Properties
• 30-45 threads per inch
• Depending on mesh density, layout not symmetric

fill
warp
18
Fabric Mesh Model

• Warp and fill yarns interwoven
• Small clearance between yarns (2 yarn thickness)

19
Computational Considerations

• Numbers of elements
• 100 elements/crossing
• 10 x 10 yarns 10,000 elements
• 90 x 90 yarns 3-in square 810,000 elements
• Calculational Time
• 15 e-6 s/element time step
• Time step 2e-9
• Velocity 100 m/s
• Time 0.5 ms (50 mm)
• Calculational Time 1 month (680 hours)
• Hope is can learn fabric response from smaller
  sections

20
Effect of Boundary Conditions

• 25 x 25 patch of Zylon fabric (3/4-inch square)
• Square steel impactor over 2 threads _at_ 120 m/s

Case 3 Not held
Case 1 Held on 4 sides
Case 2 Held on 2 sides
21
4 sides 25 yarns
Underside view
Straight under view
Topside view
22
2 sides, 25 yarns
Topside view
Underside view
Straight under view
23
Fabric Not Held
Topside view
Underside view
24
Boundaries

• Holding on 4 sides generates higher force than on
  2 sides but fabric breaks

25
Boundary Condition Load Shedding

• 4 Sides Held
• Center yarns fail locally
• Some load shedding

• 2 Sides Held
• Center yarns fail remotely
• More load shedding

• Not Held
• Low Stresses
• Yarns dont fail

26
Effect of Size

• Three cases 15 yarns, 25 yarns, 35 yarns
• Held on 2 sides
• Steel impactor at 120 m/s

27
Size Effect on Resistance

• Timing is different, but overall resistance is
  similar
• Unless it breaks or pulls out

28
Design Model

• Use shell elements with material model developed
  for Zylon fabric
• Example Fragment simulator impact
• 6-in square fabric
• Held on 4 sides
• Velocity 80 m/s

29
Current Effort

• Running on multiprocessor machine
• 4 CPU SGI Origin
• Correlate results with experiments
• Single yarn quasi-static tests
• 6-in square impact test
• Parameter studies on woven fabric impact
• friction between yarns
• shape of fragment
• Further development of shell element design model

30
(No Transcript)
31
35 yarns
Topside view
Underside view
Straight under view
Stress in held yarns
32
15 Yarns, Two Side Response

• From below see local failure, remote failure and
  pullout

33
(No Transcript)
34
Numerical Sensitivity

• Reduce shear stiffness to 0.1
• Numerical instability

35
Yarn Tensile Tests

• Constitutive properties for yarns
• Modulus
• Strength
• Rate effects

36
Push Tests

• Examine response mechanisms
• Deformation
• Damage
• Failure
• Effects of designchanges

37
Two Yarn Interaction

• Two short yarns are crossed
• Yellow yarn is lifted at 80 m/s
• Straightens and breaks
• Green yarn goes along for the ride

38
(No Transcript)
39
Effect of Boundary Conditions

• 4-sided reaches higher peak but breaks sharply
• 2-sided keeps force on longer and applies more
  impulse

40
Gage Length Effect

• Model to explain observed gage length effect in
  yarns and fibers
• Short (5 mm), compliant (E25 GPa) end on gage
  length (E180 GPa)
• Specimen modulus increases with gage length
• Model agrees with data

Gage Length
Compliant End