hydrogen storage for automotive tanks using hydrostatic
TRANSCRIPT
Hydrogen Storage for Automotive Hydrogen Storage for Automotive Tanks Using Hydrostatic Tanks Using Hydrostatic Pressure Pressure RetainmentRetainment (HPR) (HPR) MicrostructureMicrostructureBhavin Mehta, Greg Banyay, Housila TiwariOhio University
David Hill, Saurin MehtaInergy Automotive
Mark Biernacki, Steve BarnhartDaimlerChrysler
Ohio University - Mechanical Engineering 2
Hydrogen StorageHydrogen Storage
y
x
1. Pressurized Gas Cylinderstraditional, but potential dangerous and requires expensive materials
2. Liquefied Cryogenic Storagebetter volumetric energy density, but difficult to insulate and high cost of liquefaction
3. Hydrides (Metal / Chemical)potentially safe, inexpensive, but very heavy and/or unstable
4. Innovative Techniques (ie: nanotechnology, HPR…)promising techniques that require more research/validation
Current Methods
Nonetheless, the problem is still not solved…
Ohio University - Mechanical Engineering 3
y
x
Ideal FoamIdeal Foam
This is a novel concept for gas storage. Gas is stored in small bubbles of a foam matrix, thereby forming a series of small spherical pressure vessels. The resulting stress in the material between the bubbles is in a hydrostatic state of tri-axial tension.
Structural Efficiency ={(100 + 100 + 100)/3=100%}
HPR Description
Ohio University - Mechanical Engineering 4
y
x
Ideal FoamIdeal FoamHPR Description
Advantages:1) Conformability
Automotive Frames are designed before gas tanks, so the tank must be designed around the frame – not vice-versa
2) SafetyIn the case of an accident, only the gas contained in the adjacent cells tothe fracture location would dispel at once
3) Weight SavingsIn theory, because the matrix material is in a state of hydrostatic tension,the material is being utilized 100% in all 3 cartesian directions – thus requiring less material
Ohio University - Mechanical Engineering 5
y
x
Ideal FoamIdeal Foam
SC Unit Cube52% packing efficiency
BCC Unit Cube68% packing efficiency
FCC Unit Cube74% packing efficiency
Expanded BCC Lattice
optimized cell size is @ 95% of the touching radius
Materials Examined:• Stainless Steel• Aluminum• HDPE
Idealized Model Summary
Ohio University - Mechanical Engineering 6
ObjectiveObjective
y
x
While ideal HPR pressure vessels have been proven a feasible concept, the material behavior of actual cellular materials must be examined.
This research has the objective of both developing an efficient methodology for examination of actual cellular polymers and applying these methods to actual foam.
Ohio University - Mechanical Engineering 7
Work Scope and ApproachWork Scope and Approach
1. Data Acquisition: foam samples, SEM, micro-CT
2. Image Construction: image filtration/segmentation in Amira
3. Mesh Construction: triangular surface generated, cleaned, and constructed into tetrahedral solid mesh
4. FE Model Validation: simulated compression test and comparison of foamyield strength to bulk material yield strength
5. FEA HPR Analysis and Results:using Altair Hypermesh/Optistruct
6. Results Interpretation: MS Excel spreadsheet/Matlab programs developed tofind meaningful tank parameters
Ohio University - Mechanical Engineering 8
y
x
TheoryTheory
1−=foam
p
p
void
VV
ρρ
FCC βcos4.. ⋅⋅= trLE
θβ coscos4.. ⋅⋅⋅= trLE
trLE ⋅= 2.. 223 23_ −⋅+⋅−= nnnN SCbubbles
2232 23_ −⋅+⋅−⋅= nnnN BCCbubbles
1364 23_ −⋅+⋅−⋅= nnnN FCCbubbles
SC
BCC
Governing Equations of HPR Spreadsheet
for ideal models…
for actual models… for all models…
( ) [ ] 22
1VABV
VTRp −+⎥⎦
⎤⎢⎣⎡ −⋅⋅
=ε
to determine various tank parameters given criteria such as pressure, material,temperature, and cell arrangement, certain equations were developed…
Ohio University - Mechanical Engineering 9
TheoryTheory
y
xfixed H2mass
Ohio University - Mechanical Engineering 10
fixed tank volume
Ohio University - Mechanical Engineering 11
y
x
Actual FoamActual Foam
DIAB Divinycell H130: Cross-Linked PVC Foam
Mearthane Durethane DF-630A: Elastomeric Polyurethane Foam
SEM Images
Ohio University - Mechanical Engineering 12
y
x
Actual FoamActual Foam
DIAB Divinycell H130: Cross-Linked PVC Foam
Mearthane Durethane DF-630A: Elastomeric Polyurethane Foam
image slice fromBrookhaven National
Laboratory(reconstructed .bmp)
reconstructed imagedataset from OU
micro-CT scanner(GE Microview Software)
mesh renderingin Amira
mesh renderingin Amira
X-ray Computed Micro-Tomography
Ohio University - Mechanical Engineering 13
y
x
Actual FoamActual Foam
Useable FE Model Generation1) Image Filtration: median – dilation – erosion
2) Selection of Region of Interest
• smaller regions were selecteddue to limited computing capacity
• the single large box represents the entirefoam dataset while the multiple smallerboxes represent the chosen models
Ohio University - Mechanical Engineering 14
y
x
Actual FoamActual Foam
Useable FE Model Generation3) Surface Smoothing
although a smoothed surface was more aesthetically pleasing,it contained skew elements from which valid FE models could notbe constructed
Ohio University - Mechanical Engineering 15
y
x
Actual FoamActual Foam
Useable FE Model Generation
for most models, the mesh did not needto be coarsened
in other words, the maximum possibleMesh density was desirable
however, to generate “large” modelsencompassing more foam volume, themesh needed to be coarsened
this was done using a re-samplingalgorithm that restricted elements fromintersecting
Ohio University - Mechanical Engineering 16
y
x
Actual FoamActual Foam
PVC
E (psi) 175000
σ, yield (psi) 6500 +/ 500
ρ (lbs/in^3) 0.047
ν 0.32
Material Properties
H130
E, tensile (psi) 20300
E, compressive (psi) 25375
Tensile Strength (psi) 609
Compressive Strength (psi)
363
Density (lbs/in^3) 0.00469
DF-630A
E (psi) 140Ultimate Tensile Strength (psi) 4700Compressive Yield Strength @
30% Deflection (psi)125
Density (lbs/in^3) 0.0173
Polyurethane
E (psi) 200
σ, uts (psi) 5500 +/- 500
ρ (lbs/in^3) 0.043
ν 0.32
BULK MATERIAL (FEA)FOAM PROPERTIES
DF-
630A
H13
0
OTHER PROPERTIES
Cell Diameter = 350 μm
Vvoid/Vstructure = 9
Cell Diameter = ~100 μm
Vvoid/Vstructure = 1.5
Ohio University - Mechanical Engineering 17
Compression TestingCompression Testing
Verification of DIAB’spublished values
DIAB Divinycell H45, H60, H130
Ohio University - Mechanical Engineering 18
y
x
Finite Element AnalysisFinite Element Analysis
Foam Model Validation
y = 17.237x + 0.0768y = 16.325x + 0.2062
R2 = 1
0
1000
2000
3000
4000
5000
6000
7000
0 50 100 150 200 250 300 350 400
Foam Stress: Compressive Load (psi)
PVC
Str
ess:
max
imum
Von
M
ises
(psi
)
Ideal Behavior H130_8 Linear (Ideal Behavior) Linear (H130_8)
Model Validation
Ohio University - Mechanical Engineering 19
y
x
Finite Element AnalysisFinite Element Analysis
Preprocessing
Fixed Center Node• localized stress concentration• large displacements
Face Single Node• localized stress concentrations• deformation restricted
Ohio University - Mechanical Engineering 20
Finite Element AnalysisFinite Element Analysis
y
xH130 Result DF-630A ResultH130 Result DF-630A Result
maximum stresses
probable governing stresses
Preprocessing6-Face Boundary Condition: • Deformation Imperfect
• Max stress loc. unpredictable
Ohio University - Mechanical Engineering 21
y
x
Finite Element AnalysisFinite Element Analysis
Preprocessing
Single-Face Boundary Condition• for purposes of examining foam
about the periphery of the tank which is attached to a rigid outershell
Ohio University - Mechanical Engineering 22
Finite Element AnalysisFinite Element Analysis
y
x
Model ID Ntetra4 Papplied σmax_V.M. PMAX σavg_V.M. Pprobable
H130_1_small 925915 50 3710.00 87.60 9.00E+02 3.61E+02
H130_2_small 335526 50 1210.00 268.60 3.00E+02 1.08E+03
H130_3_small 1338114 50 1440.00 225.69 4.10E+02 7.93E+02
H130_4_small 780144 20 342.00 380.12 9.71E+01 1.34E+03
H130_5_small 623316 20 1890.00 68.78 5.40E+02 2.41E+02
H130_6_small 834414 20 234.00 555.56 6.60E+01 1.97E+03
H130_5 341171 20 492.00 264.23 1.40E+02 9.29E+02
H130_8 739028 20 827.00 157.19 2.00E+02 6.50E+02
mean = 250.97 920.62
standard deviation = 160.02315 556.376882
all units in PSI
H130 Results: 6-face BC
Ohio University - Mechanical Engineering 23
Tank ParametersTank Parameters
y
x
Pressure vs. Hydrogen Mass for 60 gallon tank
DF-630A, Pressure vs. H2 Mass
y = 137.35x2 + 961.49x - 17.304R2 = 0.9999
0.00
1000.00
2000.00
3000.00
4000.00
5000.00
6000.00
0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 4.000
M ass of Hydrogen (kg)
Pres
sure
(psi
)
497003722
260,130 −⋅+⋅= HHgalH mmP 17961137
22
260,630 −⋅+⋅=− HHgalADF mmP
Ohio University - Mechanical Engineering 24
y
x
Tank ParametersTank Parameters
H130, V vs. P
y = 171.84x-0.7617
R2 = 0.9878
0
2
4
6
8
10
12
14
16
18
0.00 500.00 1000.00 1500.00 2000.00 2500.00
Pressure (ps i)
Tank
Vol
ume
@ 5
kg H
2 (m
^3)
DF-630A, V vs. P
y = 333.2x-0.8123
R2 = 0.9947
0
5
10
15
20
25
0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00
Pressure (ps i)
Tank
Vol
ume
@ 5
kg H
2 (m
^3)
7617.05,130 2
8.171 −⋅= HkgH PV 8123.05,630 2
2.333 −− ⋅= HkgADF PV
Tank Volume vs. Pressure for 5 kg Hydrogen
Ohio University - Mechanical Engineering 25
Tank ParametersTank Parameters
y
xH130, Eff vs. P
y = -3E-09x2 + 5E-05x + 0.0023R2 = 1
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%
7.00%
8.00%
9.00%
0.00 500.00 1000.00 1500.00 2000.00 2500.00
Pressure (ps i)
Mas
s Ef
ficie
ncy
(kg/
kg)
DF-630A, Eff vs. P
y = -4E-10x2 + 8E-06x + 0.0005R2 = 1
0.00%
0.50%
1.00%
1.50%
2.00%
2.50%
3.00%
3.50%
0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00
Pressure (psi)
Mas
s Ef
ficie
ncy
(kg/
kg)
0023.010510322
529130 +⋅×+⋅×−= −−
HHH PPη 0005.010810422
6210630 +⋅×+⋅×−= −−
− HHADF PPη
Mass Efficiency vs. Pressure
Ohio University - Mechanical Engineering 26
y
x
Summary of ResultsSummary of Results
H130 DF-630A 2007 FreedomCAR Goals
Withstand-able Pressure 900 psi 3000 psi n/a
H2 Mass for 60 gal tank 1.2 kg 2.5 kg 8.18 kg
Tank Volume for 5kg H2 343 gallons 132 gallons 36.68 gallons
Mass Efficiency 3.5% 2.3% 4.5%
Volumetric Capacity 0.0056 kg H2 / L 0.0104 kg H2 / L 0.036 kg H2 / L
Optimistically Scaled
Ohio University - Mechanical Engineering 27
Ideal HPR matrix Results for BCC StructureIdeal HPR matrix Results for BCC Structure
• Material: DIAB H130
• Table 1: Effect of varying bubble sizes on pressure and tank size
Unit Cube Edge (a) = 100 0.60 0.70 0.95 Bubble Radius 26.0 30.3 41.1
Von Misses @P=500 psi 485 869 2350Max Pressure( psi) 3763 2100 778
Mass of H2 (kg) 4.60 4.47 4.44No. Of Bubbles 3925 3925 3925
Tank Size 2.20 2.20 2.20Void Fraction % 13.15 20.82 51.95mH2/mStructure 3.97% 4.24% 6.93%
rTrTrT
Ohio University - Mechanical Engineering 28
5 Gal Tank Provided by Inergy5 Gal Tank Provided by Inergy
Ohio University - Mechanical Engineering 29
Simulation results for 5 Gal Tank with DIAB H130Simulation results for 5 Gal Tank with DIAB H130
This tank was simulated to withstand highest pressure before yielding.
Results obtained are as follows:
Maximum Pressure without yielding = 1148 psi for YS= 3650 psiMass of H2 in tank = 0.04904278 kgMassH2 / mass STRUCTURE = 3.01%
For Storing 5 kg of Hydrogen the tank size needs to be 1.86 m3 which is almost 3.95 (~4) times the size of actual tank.
So the actual tank using DIAB H130 would be of the following dimensions:
2710 mm x 1500 mm x 670 mm
Ohio University - Mechanical Engineering 30
Summary SheetSummary Sheet
Material Steel Aluminum PolycarbonatePolyurethane(Composite)
DIAB H130(Foam)
Radius 41.1 41.1 41.1 41.1 41.1
Max Von Misses Stress 2100 2130 2100 2100 2350
Max Pressure( psi) 9524 13451 2310 10452 778
Tank size (m^3) 0.34 0.28 0.93 0.32 2.2
Tank size (gallons) 77 63 211 73 500
Tank dimensions (mm) 1200 x 840 x 370 1150 x 800 x 350 1650 x 1140 x 510 1200 x 840 x 370 2710 x 1500 x 670
Weight (Kg) 1415 405 586 271 70
mH2/mStructure 0.35% 1.23% 0.85% 1.85% 7.07%
Ohio University - Mechanical Engineering 31
y
x
ConclusionsConclusions
• The ideal models illustrate that HPR is a feasible solution to the hydrogen storage problem for the following 3 reasons:1. Conformability2. Weight Savings3. Minimized Safety Risk
• Neither H130 nor DF-630A meet the goals set by the DOE FreedomCAR project
• An efficient method for examination of polymeric foam for HPR application has been developed
• Composite foam structures can be pursued, which could greatly increase tank performance
• Back to metals (Metal foams, hollow spheres etc)
Ohio University - Mechanical Engineering 32
y
x
Recommendations for Future WorkRecommendations for Future Work
If it is verified that a non-composite foam is not suitable…Pursue Composite Foam
attain samples and examine using the methods developedin this thesis
Develop relationships between foam yield strength/modulus andHPR behavior
Expand the scope of the work past static structural analysis, toconsider thermal loads, potential chemical reactions, permeability andsorption rates, tank refuel-ability, quantifiable co$t, etc.
If/When a practical foam is found, perform experimental tests of an actualhydrogen-pressurized tank
Ohio University - Mechanical Engineering 33
y
x
Questions?Questions?