New Polyimide Windows for MagLIF Experiments · 2017. 4. 14. · Our data at higher burst pressures deviates substantially from the extrapolated values Extrapolated values for 1.8

Polyimide Windows for MagLIF

Experiments

J. Sin - presenter

T. Bernat, C. Castro, J. Hund, A. Pastrnak, O. Stein, N. Petta

Target Fabrication Meeting

Las Vegas, Nevada

March 2017 Work supported under DOE

contract:

DE-NA0001430

Outline

• There is a long history of measurements on PI windows

- Burst pressures

- Deflection measurements

• We are testing higher pressure windows for MagLIF experiments

• Our higher burst pressures don’t follow a simple extrapolation from

earlier lower pressure data

• Analysis of deflection data following Bhandarkar et al. effectively

predicts window deflection with pressure

In 1997 Luxel published burst pressures for

thin PI windows below ~ 30 psi

This data fits a single interpolation formula over the entire range

Interpolation graphs

Bhandarkar et al.

LLNL-Luxel

TFM 2015

A couple more recent points can be added to the

1997 Luxel room temperature data

The new points fall very close to the interpolation

line for a 500 nm thick window

Bu

rst

pre

ssu

re -

psi

diameter - mm

0.964 microns

0.5 microns Interpolation

0.3435 microns

0.1024 microns

Film deflection and creep has been explored

by LLNL-Luxel (2015)

• 3.94mm ≤ window ID ≤ 5.5mm

• 0.208um ≤ thicknessfilm ≤ 2.94um

• 0 PSI ≤ pressure ≤ 90 PSI

Parameter Range

S. Bhandarkar et al., “Constitutive models for viscoelastic behavior of polyimide membranes”,

Fusion. Sci. Tech. 70, p332, Aug./Sept. 2016

Schafer has recently fabricated and tested

windows at higher pressures

• Magnetized Liner Inertial Fusion (MagLIF) experiments are being

conducted at SNL and LLE

• Windows are required to hold up to 14 atm (~205 PSI)

• Two issues

- Windows should be as thin as possible

- Knowing the window deflection may be important

• LEH windows use a 1.8um thick film

• Washer ID = 3mm

• Tested at pressures ≥ 60 psi;

- Window adhesive is in

“shear” mode

Sandia’s MagLIF targets operate at “medium”

pressures

SNL MagLIF LEH window PI Film

Washer

3 mm

bulge height

• Washer with 1.8um thick PI film caps parylene tube at one end

• Washer ID = 0.4mm

• Pressurized with 11 to 14 atm (165 - 205 psi);

- Window adhesive is in “peel” mode

LLE’s Small MagLIF Targets require higher

pressures

Small MagLIF assembly 400 um

Laser

PI Film

Washer

We tested various window parameters

Shear Orientation: Film

side facing down

Peel Orientation:

Film side facing up

• Size of ID

• Orientation

• Film Formulation

- Schafer PI “α”

- Schafer PI “β”

Ø3mm, 0.4mm

Computer

controlled

Regulator/

transducer

Pressure

source

Controller

Labview interface Window mount

m-e optical

displacement

transducer

Ballast volume

Data was taken with a computer controlled system

Our test procedure ramped pressure over ten

minutes and held it for 50 minutes.

Burst tests were carried out during a subsequent pressurization

• Typical bulge test ramp

Burst pressures of both films at 3mm IDs are

clustered closely together

A PI “β” window at 0.4 mm diameter did not burst up to 400 psi.

13 samples

5 samples

Our data at higher burst pressures deviates

substantially from the extrapolated values

Extrapolated values for

1.8 mm thick windows

Lower bound:

did not burst at 400 psi

Earlier data range

However, since the LLE 0.4 mm windows are required to go to only 200

psi, we could reduce the thickness with the PI “β” formulation

Bu

rst

pre

ssu

re -

psi

Diameter - mm

Bulge heights are also clustered closely together

3 mm window

0.4 mm window

6 samples

13 samples 6 samples

5 samples

• We used the method of Bhandarkar et al. with a modified

expression for high strains:

• Average deflection from 3mm windows is used to fit the data for

modulus curves

- The curves were fit to a

6th order polynomial rather

than a piece-wise fit

• The moduli of the two

formulations are similar

in shape but quite different

in magnitude

Data from 3mm ID windows was used to evaluate

strain and effective modulus

Instead of

The moduli shapes differ from the LLNL-Luxel data

particularly at low strains

• With the washer curved to prevent a stress riser, the effective

window ID changes with deflection

• This is especially pronounced at smaller deflections

A possible explanation for moduli differences

is the 3 mm window bonding

Effective window ID is

changing as film “wraps”

around the curve

• Used moduli from 3mm ID to predict bulge height of 0.4mm ID

windows pressurized at 200 psi

Bulge predictions using the effective modulus fit

quite well to the 0.4 mm window deflection data

Summary: Schafer is fabricating PI windows for

higher pressure MagLIF experiments

• Burst and deflection of PI films were tested at smaller ID’s and

higher pressures

• We fabricated and tested two different formulations of PI film

• Burst pressures are substantially different from earlier data in the

high-pressure regime

• Using the method of Bhandarkar et al., we measured and fit

effective moduli for two polyimides

• The moduli accurately predict window deflections with pressure