Rules and methods for dimensioning embarked materials for...
-
Upload
nguyenhanh -
Category
Documents
-
view
218 -
download
1
Transcript of Rules and methods for dimensioning embarked materials for...
Rules and methods for dimensioning surface ship embarked materials subjected to underwater explosions.
Prof. Hervé Le Sourne & Mauricio García N. 2016 1
Department of acoustics and vibrations STX France
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 2
Main parameters:
𝑆𝐹 =𝑊
𝐷 Shock factor:
Bubble pulsation effect:
noncontact underwater explosion
W : Weight of the charge
D : Stand off distance
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 3
1. Objectives
1. - Key points identification for surface ships submitted to an UNDEX.
2. - Shock response rules for embarked materials.
3. - Simulation of a simplified structure submitted to an UNDEX.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 4
2. Rules review
1. DDAM is the most referenced procedure for embarked materials.
2. BV/043 German rules.
3. French rules.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 5
3. Taylor plate theory
𝑃𝑖𝜌𝑐
= 𝑃𝑟𝜌𝑐
+ 𝑣
Incident velocity
Plate velocity
Reflected velocity
𝑃𝑟 = 𝑃𝑖 − 𝜌𝑐𝑣
𝑃𝑡 = 2𝑃𝑖 − 𝜌𝑐𝑣 𝑣 =2𝑃𝑜𝜌𝑐
1
𝑍 − 1𝑒−
𝑡𝑍𝜃 − 𝑒
−𝑡𝜃
Considering that the surface is fixed
𝑣𝑖 = 𝑣𝑟 Incident velocity Reflected velocity
Supposing
First term
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 6
4. Impulse velocity approximation:
Pressure balance :
𝛽𝑖 =
𝜌𝑐𝜃
𝑚 sin 𝛼𝑖=
𝛽
sin 𝛼𝑖
𝑚𝑑𝑣𝑖𝑑𝑡
= 2𝑝𝐼𝑖 𝑡 −𝜌𝑐𝑣𝑖(𝑡)
sin 𝛼𝑖 2𝑝𝐼𝑖 𝑡 = 2𝑝0 sin 𝛼𝑖 𝑒
−(𝑡−𝑡𝑜)/𝜃
𝑣𝑖𝑚 =2 sin2 𝛼𝑖 𝑝𝑚
𝜌𝑐 𝛽𝑖
1 ( 1−sin 𝛼𝑖)
Spherical wave approximation (SWA).
(Barras, 2007).
where:
where:
Impulse velocity:
𝑥
𝑦 𝑧 LS-DYNA/USA
Initial impulse LS-DYNA
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 7
4. Impulse velocity approximation:
Along the y axis: 𝑢𝑥 = 0 𝑟𝑧 = 𝑟𝑦 = 0
Boundary conditions
Along the x axis: 𝑢𝑦 = 0 𝑟𝑧 = 𝑟𝑥 = 0
Full clamped conditions at the border
Finite element model
Materials: - High strength steel - Mild steel Objective: verify initial speed approach
(Ramajeyathilagam, K.; Vendhan, C.P.; Bhujanga Rao, V., 2000)
Simple plate analysis
0
0,02
0,04
0,06
0,08
0,1
0,12
0 0,0005 0,001
Dis
pla
cem
en
t (m
)
Time (s)
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 8
4. Impulse velocity approximation:
Initial speed results Data
Cowper Symonds material model:
Pressure based (no strain rate)
Initial speed (no strain rate)
Pressure based (strain rate)
Initial speed (strain rate)
Experimental Displacement (m)
Pressure based Displacement (m)
Shock factor
0,072 0,062 0,794
Strain rate must be considered
Strain rate effect
The initial velocity approximation underestimates the results.
𝑊
𝑅
𝛼𝑖
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 9
5. Pressure based approximation: Time history pressure for a single shell element.
𝑃𝑒𝑙𝑒𝑚𝑒𝑛𝑡 = 2𝑃0 sin 𝛼𝑖 𝑒
−(𝑡)/𝜃 −𝜌𝑐
2 sin 𝛼𝑖 𝑃𝑜𝑚
𝜃1 − 𝛽𝑖
𝑒−𝛽𝑖𝑡/𝜃 − 𝑒−𝑡/𝜃
sin 𝛼𝑖
Fluid structure interaction First term
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 10
Shock Factor 0,794 High strength steel
5. Pressure based approximation:
0,00
0,02
0,04
0,06
0,08
0,10
0,12
0,0000 0,0005 0,0010
Dis
pla
cme
nt
(m)
Time (s)
Pressure input first term
0
0,02
0,04
0,06
0,08
0,1
0,12
0,0000 0,0005 0,0010
Dis
pla
cem
en
t (m
)
Time (s)
Shock Factor 0,849 Mild Steel
Experimental
Pressure input (Reference)
Initial speed
Pressure input (Reference)
Experimental
Initial speed
pressure SWA (First term)
pressure SWA (First term)
Final deformation (slightly) found above the experimental results.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 11
5. Pressure based approximation: SWA full equation.
0
0,02
0,04
0,06
0,08
0,1
0,0000 0,0005 0,0010
Dis
pla
cem
en
t (m
)
Time (s)
Shock Factor 0,849 Mild Steel
0
0,02
0,04
0,06
0,08
0,1
0,0000 0,0005 0,0010
Dis
pla
cem
en
t (m
)
Time (s)
Experimental
Pressure input (Reference)
Initial speed
Pressure SWA
Experimental
Initial speed
Pressure input (Reference) Pressure SWA
Results similar to the reference.
Shock Factor 0,794 High strength steel
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 12
Taken from: https://www.geogebra.org/m/26707
• ANSYS does not have explicit solution option (LS-DYNA only)
Full transient simulation.
Yiel
d s
tres
s
Strain rate
Cowper Symonds Perzyna model
5. Pressure based approximation:
Same equation.
1/p = m and 𝛾 = D
Implicit solution Explicit solution
ANSYS & LS-DYNA material model.
D = 40 & m = 5 for steel.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 13
5. Pressure based approximation: Results validation using ANSYS.
Shock Factor 0,794 High strength steel Shock Factor 0,849 Mild steel
0
0,02
0,04
0,06
0,08
0,1
0,12
0,0000 0,0005 0,0010
Dis
pla
cem
en
t (m
)
Time (s)
0
0,02
0,04
0,06
0,08
0,1
0,12
0,0000 0,0005 0,0010Dis
pla
cem
en
t (m
)
Time (s)
Experimental
Experimental
Pressure input (Reference)
Pressure input (Reference)
Initial speed Initial speed
Pressure (ANSYS)
Pressure (ANSYS)
Results above the ones obtained using LS-DYNA.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 14
4 loading approaches tested
• Initial impulse
• Initial impulse + added mass
• Pressure only
• Pressure + FSI
Slightly overestimates but conservative!
Oscillates near experimental results.
Underestimates the damage
Largely underestimates the damage
4 Different loading approaches 60 calculations using LS-DYNA or ANSYS
5. Pressure based approximation: Results summary
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 15
6. Stiffened plate : ANSYS compared to LS-DYNA
Finite element model
Boundary Conditions
Along the y axis: 𝑢𝑥 = 0 𝑟𝑧 = 𝑟𝑦 = 0
Along the x axis: 𝑢𝑦 = 0 𝑟𝑧 = 𝑟𝑥 = 0
Full clamped conditions at the border
𝑥
𝑦
𝑧
Plate thickness: 12 mm
Quench
Steel Mild steel
Young Modulus (MPa)
400 250
Tangent Modulus (MPa)
631 350
Poisson ratio 0,3 0,3
Material Properties
¼ of the model
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 16
0,00
0,10
0,20
0,30
0,40
0,50
0,000 0,005Dis
pla
cem
en
t (m
)
Time (s)
0,00
0,10
0,20
0,30
0,40
0,50
0,000 0,005
Dis
pla
cem
en
t (m
)
Time (s)
𝐸𝑓 = 0.056 + 0.54 𝑡
𝑙𝑒 Shock factor increased until rupture
6. Stiffened plate : ANSYS compared to LS-DYNA
Mild steel SF: 0,6 Quench steel SF: 0,8
Results obtained by the two software are similar.
ANSYS
LS-DYNA LS-DYNA
ANSYS
Erosive law.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 17
LS-DYNA results ANSYS results SF LS-DYNA ANSYS ERROR %
0,44 0,25 0,25 2,89
0,55 0,33 0,34 5,35
0,6 0,36 0,38 6,16
Mild Steel
Quench Steel
SF LS-DYNA ANSYS ERROR %
0,66 2,94 3,04 3,40
0,77 0,35 0,36 4,82
0,833 0,36 0,40 9,86
6. Stiffened plate : ANSYS compared to LS-DYNA
Similar pattern on the distributions of plastic strains
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 18
7. Ship full section: ANSYS compared to LS-DYNA.
Same procedure applied to the stiffened plate.
Boundary Conditions
Restricted displacement at the
edges. Rotation is allowed.
Plate thickness: 10 mm
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 19
7. Ship full section: ANSYS compared to LS-DYNA.
Scantling dimensions First load step
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 20
7. Ship full section: ANSYS compared to LS-DYNA.
Points being measured
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 21
7. Ship full section: ANSYS compared to LS-DYNA.
Exact modeling: number of elements – stiffeners – load profile.
0,000,050,100,150,200,250,300,35
0,000 0,020
Dis
pla
cem
en
t (m
)
Time (s)
MS-Corner scantling SF-0,378
0,000,050,100,150,200,250,300,35
0,000 0,020
Dis
pla
cem
en
t (m
)
Time (s)
MS-Small scantling SF-0,378
0,000,050,100,150,200,250,300,35
0,000 0,010 0,020
Dis
pla
cem
en
t (m
)
Time (s)
MS-Large scantling SF-0,378
LS-DYNA
ANSYS
LS-DYNA
ANSYS
LS-DYNA
ANSYS
Slightly overshoot possibly due to the element formulation.
Maximum: Mild Steel SF: 0,378. Quench Steel SF: 0,489
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 22
7. Ship full section: ANSYS compared to LS-DYNA.
Plastic strain comparison using Mild Steel S.F.:0,33.
ANSYS results LS-DYNA results
Plastic strain distribution present the same pattern.
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 23
• The initial speed approach underestimates the experimental results.
• The results obtained by the pressure, neglecting the second term, overestimate the level of deformation.
• LS-DYNA and ANSYS end up having approximately similar results. Considering the rupture strain of the plate.
• Discrepancies occur between LS-DYNA and ANSYS. Those discrepancies are probably due to the solvers themselves and to the formulation of the shell elements used.
8. Conclusions
Rules and methods for dimensioning embarked materials for surface ships when subjected to UNDEX.
Prof. Hervé Le Sourne & Mauricio García N. 2016 24
MANY THANKS: Special thanks to: • Supervisor: Hervé Le Sourne. • Reviewer: Phillippe Rigo. • Reviewer: Lionel Gentaz.
• Sylvain Branchereau. • Marc Yu. • Clement Lucas.
But also to: • Department of Acoustics and
Vibrations @ STXFrance. • All of the professors and students
@ ICAM.
• z