Update on:

Improved Properties of Nanocomposites for Flywheel Applications

Sandia National LaboratoriesAdvanced Materials Laboratory

1001 University Boulevard SEAlbuquerque, New Mexico 87106

(505)272-7625(505)272-7336

tjboyle@Sandia.gov

This work was supported by the Office of Electricity of the Department of Energy under Contract DE-AC04-94AL85000. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy.

Timothy J. Boyle (1815), Mathias C. Celina (1821), Nelson S. Bell (1816), Benjamin J. Anderson (1833)

November 2, 2010

Funded by the Energy Storage Systems Program of the U.S. Department Of Energy through Sandia National Laboratories

The current energy “regulation profile” is complex with demanding, small, time energy perturbations, which will be magnified upon introduction of alternative energies.

Optimal AC grid usage requires accurate, rapid, and continuous adjustments (frequency regulation) of power pulses.

Wind Ram

pingP

erce

ntag

e

Per

cent

age

Sola

r Sha

ding

A 20 MW flywheel energy storage resource accurately following a signal

Flywheels provide “near instantaneous” response

• Zero direct carbon emissions

• 85% system efficiency at transformer

• Zero storage degradation

• 20-year design life > 150,000 100% DoD charge / discharge cycles

In order to maintain efficiencies, speed is necessary.

• Faster, more effective than conventional generation– Up to 100x faster than a traditional generator

• Available separately – without generation

• Low operating cost (no fuel)– Displaces higher cost deployments of traditional

generators

– Could reduce CO2 emissions by 80 % (KEMA study)

A coal-fired power plant poorly following a regulation command signal

What is a flywheel?

1898 illustration of a White and Middleton stationary engine; note the large twin flywheels.

Energy is stored in the rotor as kinetic energy, or more specifically, rotational energy:

ω = angular velocity, I = moment of inertia of the mass about the center of rotation

Ek = ½ • I • ω2

The amount of energy that can be stored is dependant on:

σt = tensile stress on the rim, ρ = density, r is the radius, ω is the angular velocity of the cylinder.

σt = ρ • r2 • ω2

Flywheels have become much more advanced, requiring complex composite rim materials to maximize efficiency

All flywheels have similar issues – the ‘need for speed’ - kills!

Composite Rim

Vacuum Chamber

Magnetic Bearing

Shaft

Motor

Hub

Commercial Flywheels are typically designed for

(max energy) / cost

not

(max energy) / weight

For commercial interests, energy increase

must result in a higher

kWh/$

Motor/Generator

Hub

Composite Rim

VacuumHousing

Magnetic Bearings

Motor/Generator

Hub

Composite Rim

VacuumHousing

Magnetic Bearings

Approach: incorporation of nanomaterials into rim composite to improve flywheel performance.

The main focus of this project is optimization the rim materials to allow for increased flywheel performance

Modification of matrices’ physical properties by nanofillers is controlled by 4 phenomena:

(I) Hydrodynamic effects,(II) Occluded polymer effects,(III) Bound polymer effects,(IV) Interaggregate attraction.

The filler properties can be manipulated by 3 primary components of the nanoparticle which relate to its performance in a matrix:

(I) Surface area (related to size and aggregation),(II) Filler surface energy,(III) Filler structure in the matrix.

100

ns –

40 d

ays o

n 51

2 pr

oces

sors

32 coated nano particles in decane (not shown)

Improve conductivityImpact strengthStorage modulus,

Wear resistanceOptical propertiesFracture toughness

• Characterize/evaluate existing high quality flywheel materials

(a) Existing operating material to establish baseline(b) Evaluate different resins(c) Evaluate nanoparticle effects on resins(d) Evaluate nanoparticle effects on fibers

• Measure and optimize resin processes(a) Cure kinetics determination(b) Monitoring of cure chemistry

• Nanoparticle syntheses/selection(a) Size(b) Shape(c) Phase(d) Functionalization

• Characterize/optimize nanoparticles/matrix interaction(a) Surface charge(b) Rheology(c) Viscosity

• Feedback loop

Approach based on defining ‘state-of-the-art’ system and elucidating nanoparticle fillers effects

Transverse failures are observed at higher spin speeds in current flywheel materials

• Inner RingHoop x-section Radial x-section

• Outer RingHoop x-section Radial x-section

Voids

SEM images reveal ‘void’ formation occurring in both the inner and outer rims

Microdroplet test of the epoxy/ E- glass fiber adhesion strength undertaken

Epoxy Microdroplet

on E-glass fiber

0

0.1

0.2

0.3

0.4

0 0.5 1 1.5

Load

[N]

Position [mm]

0

0.02

0.04

0.06

0 0.5 1

Load

[N]

Position [mm]

Adh = 50 MPaAdh = 10 MPa

100 oC max cure 150 oC max cure

Fiber w/ microdroplet

Microdroplet de-bonded by calipers

Ultimate strength of adhesion found to

depend on the cure schedule:

Droplets cured to 150 oC have 5x higher adhesion strength

Adh = ForceInterfacial Area

Isothermal cure kinetics of the anhydride epoxy reaction realized from IR spectroscopy measurements

• Cure chemistry as a function of time (t) and temperature (T)

• Correlation with cure viscosity• Use data as model input for autocatalytic behavior

Epon862:LSK81 = 1:1 system Time (Hrs)

0.001 0.01 0.1 1 10 100 1000

ATR

pkht

anh

ydrid

e 17

77 c

m-1

0.0

0.2

0.4

0.6

0.8

1.0

120C110C100C90C80C70C60C50C

Time (Hrs)

0.01 0.1 1 10 100 1000

ATR

pkht

1777

cm-1

0.0

0.2

0.4

0.6

0.8

1.0

120C110C100C90C80C70C60C50C

1/T [10-3K-1]

2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2

Shift

fact

ors

aT

0.1

1

10

100

1000

Eact= 73.5 KJ/mol

Cure viscosity monitoring using dielectric analysis and its correlation with cure chemistry

• Interdigitated electrodes detect cure state via dielectric loss spectroscopy

• Remote sensing of cure conversion as a function of time (t) and temperature (T)

• Correlation with cure chemistry and rheology• Input for 3D models linking evolution of chemistry and physical properties

161_+75%ECA100_1%imi_90C_hot plate

Time (mins)0 50 100 150 200

log[

ion

Visc

osity

]4

6

8

10

12

14

11010^210^310^45x10^410^5

Ceramic nanoparticles selected as filler based on multiple intrinsic characteristics/properties

solvothermal

Solution precipitation

100 nm

Aldrich(Rutile)

50 nm

Hybrid(Anatase)

50 nm50 nm

SOLVO(TiO2-B)

100 nm

Aldrich (Anatase)

50 nm

HI(Anatase)hybrid

(i) Sarwar et al. J. Sol-Gel Sci Tech. (2007) 44, 4. (iv) Sumfleth et al Polymer (2008) 49, 5105.(ii) Adler-Abramovich et al. Angewandte Chemie (2010) 49, 1-5. (v) Sangermano et al. Macromol. Mater. Eng. (2006) 291 517.(iii) Kane et al. J Appl. Cryst. (2009) 42, 925.

The surface chemistry of the TiO2 nanoceramics were studied by comparing ζ-potential measurements.

2 3 4 5 6 7 8 9 10 11-80

-60

-40

-20

0

20

40

60

80

Zeta

Pot

entia

l (m

V)

pH

Aldrich Anatase Wires Anatase Aldrich Rutile Rods Rutile Wires Rutile

Surface Modification in volatile solvent (acetone)

Dispersion of Solids in chosen phase

(resin, hardener)

Hi Shear Mixing with second phase

Degassing

Casting or Processing

Controlling the dispersion of the nanoparticles requires tailored surface chemistry

Variations noted for different TiO2 nanomaterials, which will allow for fine tuning interaction.

Surface Modification in volatile solvent (acetone)

Dispersion of Solids in chosen phase

(resin, hardener)

Hi Shear Mixing with second phase

Degassing

Casting or Processing

10 100 10000

5

10

15

20

25

30

Inte

nsity

% (%

)

MTHPA Average PSD (nm)

Rutile Rods Solvo Wires Hybr Wires Aldrich Rutile Aldrich Anatase

MTHPA Solvent - No additve

10 100 10000

5

10

15

20

25

30

Inte

nsity

% (%

)

MTHPA Average PSD (nm)

Rutile Rods Solvo Wires Hybr Wires Aldrich Rutile Aldrich Anatase

HHMPA Solvent - No additve

10 100 10000

5

10

15

20

25

30

Inte

nsity

% (%

)

MTHPA Average PSD (nm)

Rutile Rods Solvo Wires Hybr Wires Aldrich Rutile Aldrich Anatase

NMA Solvent - No additve

• NMA fluid gives smaller particle sizes and most stable dispersion. • MTHPA and HHMPA fluids lead to particle aggregation and sedimentation; however, molecular structures are not

greatly different between the three resin curatives. • Of the resin curatives properties, NMA has highest viscosity of the three systems tested.

The dispersion of the nanoparticles is highly dependent on the curative agent and nanoparticle

Optical Light Scattering

HHMPAHexahydro-4-methyl phthalic anhydride

NMANadic methyl anhydride

MTHPAMethyl 1,2,3,6-tetrahydro phthalic anhydride

Dot

s –A

Dot

s –R

Squa

res

Wire

s-H

Wire

s-S

Dot

s –A

Dot

s –R

Squa

res

Wire

s-H

Wire

s-S

Dot

s –A

Dot

s –R

Squa

res

Wire

s-H

Wire

s-S

Dissemination of technical information asPapers, Patents, and Presentations is forthcoming

Papers: (iii) Celina et al. “Cure reactions of advanced composite resins explored by high

temperature micro ATR-IR” 241st ACS National Meeting, Anaheim, CA. Program Area: POLY: Division of Polymer Chemistry Symposium (POLY002) Polymers for Energy Storage and Delivery

(ii) Bell and Boyle “Nanoparticle stabilization mechanisms in epoxy curative fluids: wetting interaction and Van Oss model parameters” (in prep for J. Materials Chemistry)

(i) Boyle and Ottley “Structural Characterization of a Novel Family of Cesium Aryloxide” (in prep for Inorganic Chemistry).

Patents/Technical Advances:None

Project initiated February 2010

Presentations:(a) Saad et al. “Synthesis of Barium/Cesium Alkoxide Precursors for the Application of Nitrogen Phosphorus

Detectors“ Rio Grande Regional Meeting (OCT 2010 – Albuquerque).(b) Ottley et al. : “A Novel Family of Cesium Alkoxides as Novel Resin Catalysts” 241st ACS National Meeting,

Anaheim, CA. Program Area: Materials Chemistry (upcoming)(c) Boyle and Bell: “Novel Precursors for production of complex well-characterized nanoceramic materials”

241st ACS National Meeting, Anaheim, CA. Program Area: Materials Chemistry (upcoming)(d) Celina: “Cure reactions of advanced composite resins explored by high temperature micro ATR-IR” 241st

ACS National Meeting, Anaheim, CA. Program Area: POLY: Division of Polymer Chemistry Symposium (upcoming)

Summary and Conclusion

• Voids observed in the rim materials offlywheel

• Ultimate strength of adhesion found todepend on the cure schedule

Time (Hrs)

0.001 0.01 0.1 1 10 100 1000

0.0

0.2

0.4

0.6

0.8

1.0

120C110C100C90C80C70C60C50C

microdroplet testing of epoxy-fiber strength

isothermal cure kinetics

nanoceramic filler materials

10 100 10000

5

10

15

20

25

30

Inte

nsity

% (%

)

MTHPA Average PSD (nm)

Rutile Rods Solvo Wires Hybr Wires Aldrich Rutile Aldrich Anatase

NMA Solvent - No additve

Optical scattering of NMA

• Nanoceramic materials developed with different morphologies.

• Variations noted for different TiO2nanomaterials surface chemistry and behavior in resin curatives, which will allow for fine tuning interaction.

• Isothermal cure kinetics of the anhydride epoxy reaction realized from IR spectroscopy measurements

SEM image of outer ring

Future Aims• Optimize interaction of nanoparticle and curative agent.• Introduction of morphologically varied NP into fully

characterized system for impact determination.•Complete characterization of epoxy/fiber (glass and

carbon) adhesion through microdroplet de-bonding measurements for different cure schedules

• Measure transverse yield stress of epoxy/fiber composites for different cure schedules

• Investigate 3-D nanoparticles• Explore measuring stress strain behavior in pure resin

cylindrical specimens using compression.• Develop the 1D FE models linking resin cure with position-

dependent thermal conditions.• Develop the measurement of thermal gradients and

maximum cure temperature variations in thick resin specimens as a function of cure time.

• Initiate nanoparticle on C-wire surface• Magnetic component for ‘hubless’ design study initiated.

Motor/Generator

Hub

Composite Rim

VacuumHousing

Magnetic Bearings

Motor/Generator

Hub

Composite Rim

VacuumHousing

Magnetic Bearings

Increasing the strength of the rim of the flywheel is necessary to store more energy.

Ek = ½ • I • ω2

σt = ρr2ω2

• Kinetic energy depends on speed of wheel which is limited by the tensile strength of the rim material

Approach is to initially determine the properties of the ‘state-of-the-art’ commercial system (Beacon Power) through (i) mechanical and (ii) chemical testing. With this baseline information, we can then introduce well-characterized nanoparticles - both in (iii) physical (i.e., size,shape) and (iv) chemical properties (i.e., solubility) to elucidate/optimize changes.

Time (Hrs)

0.001 0.01 0.1 1 10 100 1000

ATR

pkh

t anh

ydrid

e 17

77 c

m-1

0.0

0.2

0.4

0.6

0.8

1.0

120C110C100C90C80C70C60C50C

(i) Microdroplet testing of epoxy-fiber strength

(ii) Isothermal cure kinetics determined by IR spectroscopy

(iiii) nanoceramic filler materials

10 100 10000

5

10

15

20

25

30

Inte

nsity

% (%

)

MTHPA Average PSD (nm)

Rutile Rods Solvo Wires Hybr Wires Aldrich Rutile Aldrich Anatase

NMA Solvent - No additve

Dot

s –A D

ots

–R Squ

ares W

ires-

H Wire

s-S

50 nm

(iv) Optical scattering of different curative agents

Our approach is to introduce nanoceramic fillers into the

resin to increase the strength of the rim material. These

nanoparticles will not significantly increase the

weight of the rim but should allow for faster spin speeds

and thus more stored energy