Accelerated lifetime testing of Accelerated lifetime testing of

vacuum insulation panelsvacuum insulation panelsDouglas Smith, Steve Wallace, Dan Martinez and

Bob Keller

NanoPore Insulation LLC

3721 Spirit SE, Albuquerque, NM 87106 USA

Quentin Shrimpton

NanoPore Insulation Ltd.

Brimfield, UK

What is NanoPore Insulation LLC?What is NanoPore Insulation LLC?

• NanoPore Insulation LLC is a joint venture of NanoPore Inc. and Sealed Air corporation (a 4 billion+ $ packaging company with extensive expertise in high-speed vacuum packaging and barrier films)

• NanoPore Insulation LLC has three US-based VIP production lines including a new automated facility and a UK plant producing VIPs to the same specifications.

• NanoPore is the only producer of nanoporous carbon-silica vacuum insulation panels (VIP’s) and has been producing these for >10 years.

• NanoPore has developed and continues to develop a range of barrier/core materials optimized for various applications

• In ’09, we will produce >1M VIPs and >3M in ‘10.

Lifetime of InsulationLifetime of Insulation

• The thermal performance of all thermal insulation degrades over long time periods because of:– Water adsorption/absorption

– Structural degradation from thermal and barometric cycling, water adsorption, etc.

– Outgassing

– Temporal decomposition for foams

• How to define lifetime for VIP’s? Difficult given the logarithmic change of thermal conductivity with pressure/time

• The lifetime of a VIP is more dependent on both its’ use conditions and dimensions than other insulation

• We normally define predicted lifetime as the time required to lose 10% of the thermal resistance under use conditions

Key issues in VIP lifetime Key issues in VIP lifetime

• Operating temperature and humidity

• VIP design

– Core material

– Barrier film

– Getters/desiccants (if any)

– Thickness (affects the surface area to volume ratio)

– Seal layer (material type, thickness, width)

• Barrier film specifications are only a starting point since they will degrade over time with:

– VIP production

– VIP handling and system integration

– Barometric pressure cycling

– Thermal cycling

• Lifetime performance in a “system” can be very different (better or worse) than that for the VIP only.

Effect of flexing Effect of flexing • We know that the WVTR and OTR (hence lifetime) of metalized

films changes dramatically (>100x worse) with flexing.

• Impossible to relate laboratory flex testing back to the real world of VIP use in production and application!

• EvOH-only does not degrade but not nearly as good to begin with!

• No data on either mEvOH or mPET with an EvOH seal layer

OTR (cc/m2d)

Virgin Flex/Lab 1 Flex/NP 2 cycles Flex/NP 10 cycles

Met #1 <0.01 0.75 1.5 4.3

Met #2 <0.0005 0.15 na na

EvOH <0.2 0.06 na na

WVTR (g/m2d)

Virgin Flex/Lab 1 Flex/NP 2 cycles Flex/NP 10 cycles

Met #1 0.02 0.20 0.12 0.25

Met #2 0.01 0.084 na na

EvOH 2.12 2.43 na na

Temperature dependence on Temperature dependence on

barrier film permeationbarrier film permeation

0.0001

0.001

0.01

0.1

1

20 40 60 80 100 120 140

Temperature (oC)

WV

TR

(g

/m2d

ay

)

B4-coated B1-virginB1-2 twists B1-10 twistsB2-virgin B3-virginF1-virgin

• There is a strong temperature dependence on gas/vapor

permeation rates through VIP barriers. Going from 20 to

80 oC represents a ~20x increase in aging rate

• Results for:

– current virgin films

– current flexed films

– next generation barriers

Water Vapor: Water Vapor: TheThe problem for problem for

nonnon--nanoporous cores!nanoporous cores!

• Typical ambient water

vapor pressure is 5-30 mbar

• In buildings, the VIP’s

coldest side sets the vapor

pressure (<ambient)

• For foams and fiberglass,

the useful capacity of the

desiccant is used up at <5%

relative humidity!

• Foams/fiberglass VIPs are

just not applicable for

building applications,

impossible to add enough

desiccant!

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.1 1 10 100 1000Pressure (mbar)

Co

nd

uc

tiv

ity

(W

/mK

)

silica-carbon

Foam

Fiberglass

ambient

humidity

Water Vapor in ConstructionWater Vapor in Construction

• During summertime, maximum water vapor pressure is the inner wall (22 oC/100% RH) or ~25 mbar independent of the atmospheric

temperature/humidity.

• For winter, even at 100% RH, the lower outer wall temp. means that water

vapor will actually be leaving the VIP, not entering!

• Because of the high adsorption capacity in the 70-80% RH range, water vapor

pressure in the VIP should never exceed 20 mbar in most geographies (except

tropical)

0

5

10

15

20

25

30

35

40

45

0 5 10 15 20 25 30Temperature (oC)

Vap

or

pre

ssu

re (

mb

ar)

0

4

8

12

16

20

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Relative Humidity RH %

We

igh

t o

f H

2O

(w

t %

)

Water vapor pressure at 100% RH Water adsorption on NanoPore core

Does the core really matter?Does the core really matter?

• Compare a silica-carbon VIP with no desiccant to a commercially available (for

warm applications) fiberglass VIP with a desiccant

• 25 mm thick panels and both have similar metalized barriers

• Store at 80 oC, cool to ambient and measure the panel pressure. Convert the

pressure to the room temperature conductivity

• Fiberglass VIP conductivity doubled in two days!

• Silica VIP reached a pressure plateau and has now been aged for over ~450

days at 80 oC with only 20% loss in thermal performance.

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

0 20 40 60 80 100

Days at 80 oC

Co

nd

uc

tiv

ity

(W

/mK

) a

t 2

0 o

C

silica-carbon

Fiberglass

Long time 80 Long time 80 ooC accelerated testingC accelerated testing

• Compare 25 mm silica-carbon VIP’s with no desiccant and

different barrier materials

• The two metalized barriers have the same 3xmPET structure!

• Store at 80 oC, cool to ambient and measure the panel pressure.

0

10

20

30

40

50

60

70

80

90

100

110

120

0 50 100 150 200 250 300 350 400 450 500

Time (days)

Pre

ss

ure

(m

ba

r)

EvOH EvOHSiO2 #1 SiO2 #1

metallized #1 metallized #1metallized #2 metallized #2

Thermal cyclingThermal cycling

• In a 50 year service life, a VIP will go through >18,000 thermal cycles

due to diurnal temperature variation!

• Silica CTE is different than plastic barriers and other building

materials.

• Thermal cycling induces repeated stress on the barrier material as the

plastic barrier expands and contracts with ambient temperature

variation.

• Stress from barometric cycling is much lower

Material CTE (ppm /oC)

Silica 0.5

PET 73-92

Aluminum 22

Mild steel 12

Wood (perpendicular to grain) 30-50

Wood (parallel to grain) 3-5

Polyurethane 40-160

Masonry brick 6-9Glass 6-9

Thermal cycling (cont)Thermal cycling (cont)

• In buildings, although the interior side of the VIP is essentially

isothermal, the outer VIP surface temperature can vary significantly

depending upon the climate zone.

• Not normally addressed in VIP lifetime analysis! Not an issue for

many VIP applications such as refrigeration (where temperatures are

essentially constant with time)

• Effect of thermal cycling is system design dependent.

• Depends on CTE and modulus of adjacent layers.

– Best case is a VIP only (just CTE mismatch between core &

barrier)

– Worst case is a VIP bonded directly to a metal with a rigid

adhesive

• Stress relief at the VIP interface is a major design consideration

Thermal Cycle TestingThermal Cycle Testing

• Multiple silica/carbon VIPs in an Aluminum/PU

foam/VIP/PU foam/plastic sandwich

Construction Diagram

Chemlite Liner

Aluminum Shin

(0.050" Thk)

Vacuum

Insulated Panels(½" Thk) (x3)

PU Foam Board

(½" Thk) (x2)

Gap filled with foam

1 ½

"

38"

All mating surfaces

glued with PU glue

Build-up

Thermal Cycle Panels

• Attached Aluminum and FRP skins during second gluing operation

• Mounted heat flux sensor to inside of skins at center of door

• Filled voids at edges with expandable PU foam.

• Built-up core (PU/VIP/PU) first

• Attached surface mounted, Type-T thermocouples to each side of VIP’s (8 total)

• Packing foam attached to side of VIP to prevent wear

• Applied PU glue to all surfaces

Thermal Cycle TestingTest Set-up

• Composite panel

mounted to chamber

• Interior cycled from

50 °C to –30°C (i.e.

much larger swing

than normal building

applications to

accelerate aging

• Exterior~20 °C

• Measurements

– 2x Heat flux sensors

with integrated TC

– 8x Surface TC

– 8x Embedded TC on

VIP

Heat Flux

Sensors (x2)

Thermocouples(x18)

Composite Door

Temperature Profile

-30

-20

-10

0

10

20

30

40

50

60

70

80

90

100

110

120

130

0 30 60 90 120 150 180 210 240 270

Elapsed Time (Min)

Te

mp

era

ture

(°F

) .

Environmental

Chamber

8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00

Time

-30

-15

0

15

30

45

60

75

90

105

120

135

Te

mp

era

ture

[F

]

21

Inner Skin Temperature

Outer Skin Temperature

Inner VIP Temperature

Outer VIP Temperature

Single cycle• Because of the low thermal diffusivity of VIP’s, a single cycle takes

~6 hours to stabilize at high and low temperatures

• In climates with large diurnal variation (such as New Mexico!), this

shows that it is not just thermal conductivity that matters but also

thermal diffusivity. With more thermal mass or thicker VIPs, this

effect becomes more pronounced!

Thermal resistance versus time• 647 cycles (>160 days), there was minimal thermal resistance change

• After 647 cycles, a temperature controller malfunction caused the

chamber to spike to >100 oC so the test was stopped and panels were

pressure tested

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

0 200 400 600 800

# of thermal cycles (-30 C to 50 oC)

No

rmali

zed

th

erm

al

resis

tan

ce

NV1DT

NV2DT

NV3DT

NV4DT

Post thermal excursion analysis

• Hot side PU showed significant thermal degradation as

compared to cold side

• All VIPs were intact! Pressure was measured and found to

be between 10 and 13 mbar despite the hot side excursion

Cold side PU

Hot side PU

VIPs

Summary• For long-life building applications, only nanoporous core

materials such as silica are appropriate because of the

difficulty of maintaining water vapor pressure below 1 mbar

for cores with larger pores (foam/fiberglass).

• Assuming that water vapor is not an issue (i.e, core yields

good thermal performance in the ~20 mbar range such as

for some silica panels), 80 oC accelerated testing indicates

lifetimes well over 30 years with minimal (<20% thermal

performance degradation depending on the barrier material).

• This study seems to be the first long term, accelerated

testing on thermal cycling of VIPs in an integrated system.

• Thermal cycling means that VIP system design is an

important factor in VIP lifetime.