Peer Review

NSTX TF Hub Assembly Friction Coating

C Neumeyer10/27/04

RequirementsTestingCalculation UpdateImplementation

Friction Coating for Hub/Box Interface

Hub/BoxAttachment Studs

Potting

Hub/Box Interface

Joint must not slip under applied EM loads

FlateraldR from effective

point of application to flag

inboard edge

Moment

(lbf) (in) (in-lbf)Outer Layer Flag

3174 8.9 28286

Outer Layer Link

1899 18.8 35767

Total perOuter Layer Flag/Link

5074 12.6 64053

Inner Layer Flag

1979 9.1 18061

Inner Layer Link

1330 19.0 25200

Total perInner Layer Flag/Link

3039 14.2 43261

Total from All Turns 158233

13.02056400

Outer Layer Loads Are Design Drivers

Flex Link loads updated out to r = 24”

FverticaldR from effective

point of application to flag

inboard edge

Moment

(lbf) (in) (in-lbf)Outer Layer Flag

7379 4.3 31719

Outer Layer Link

2575 15.1 38934

Total PerOuter Layer Flag/Link

9955 7.1 70653

Inner Layer Flag

1845 4.3 7930

Inner Layer Link

644 13.8 9734

Total PerInner Layer Flag/Link

2489 6.2 17663

Total fromAll Turns

268775

Friction Shear (frs) at Interface Must Exceed Loadw/adequate Safety Factor

M_em

F_ frs

F_ frs

In-Plane Reaction to Moment Out-of-Plane Reaction to Lateral Load

F_frs ~ f*M_em/r/2f = load fraction to frsApprox. 42% per FEA

r

F_frs ~ f*F_em/2Assume f = 100%

Max Single Surface Radial Shears at Hub-

to-box Interface(Total load along face in lbs)

For 6kG TF-loading with max moment (70000 in-lbs)

SOFT – 3479 lbs (Run 54ng)

EOFT – 4569 lbs (Run 65nb)

In both cases, the maximum hub-to-box shear surface is the one away from the plasma

Courtesy Irv Zatz

Design Load and Safety Factor

Net load per interface is resultant of in-plane and out-of-plane at 6kG….

€

Ffrs_ IP = 4569lbf

Ffrs_ OOP = 2537lbf

Ffrs_ NET = Ffrs_ IP2 + Ffrs_ OOP

2 = 5226lbf

Average Box Stud compression force with three bolts ….

€

Fnormal = 3*5500 =16500lbf

Desired Friction Coefficient for SF = 2 ….

€

COFSF= 2 =2 *5226

16500lbf= 0.633

Grit-Surfaced Sample from Braziler Technology, Inc

Test Fixture

• The test fixture consisted of a floating clamp with a load cell to measure the clamping load. Right view

• Bottom end view of grit sample between two sacrificial clamping blocks

Measured COF ~ 0.6

• Tests on coated SS ~ 0.6

• Tests on uncoated SS ~ 0.2

• Coating provides 3x improvement

Sample Grit 060/080 060/080 SS Strip

Run 1 lbf 140.00 141.00 45.50Run 2 lbf 120.00 127.00Run 3 lbf 120.00 103.00Run 4 lbf 100.00 141.00Average lbf 120.00 128.00 45.50Std Dev lbf 16.33 17.93

Static Coefficient of Friction Avg 0.61 0.65 0.23Test RunsNo. Box Bolts 3.00 3.00 3.00Force/Bolt lbf 5000.00 5000.00 5000.00Area sq in 15.40 15.40 15.40Pressure psi 974.03 974.03 974.03Sample Width in 0.225 0.225 0.337Clamp Height in 0.500 0.500 0.500Sample Surface Area sq in 0.113 0.113 0.169Clamp Force lbf 99.00 99.00 99.00Clamp Pressure psi 880.00 880.00 587.54Friction Coefficient 0.61 0.65 0.23No. Surfaces 2.00 2.00 2.00Shear Force to Slip lbf 120.00 128.00 45.50Handle Dia in 0.115 0.115 0.115Handle Area sq in 0.010 0.010 0.010Handle Stress psi 11558.88 12329.48 4382.74

Break Away Peak Force

Courtesy Tom Kozub

Calculation

COF for Coated Plates 0.600 0.200 0.200

At 6 kG 6 6 6 kGTotal In Plane EM Moment 70653 70653 70653 in-lbf% In Plane Load to Hub/Box Shear 42% 42% 42%Moment Arm to Interface 3.25 3.25 3.25 inIn Plane Shear Load per Interface 4569 4569 4569 lbfTotal Out Of Plane EM Load 5073 5073 5073 lbfOut Of Plane Shear Load per Interface 2537 2537 2537 lbf

Field Strength kG 6.00 4.50 3.56 kG

In Plane Load 4569 2570 1607 lbfOut Of Plane Load 2537 1902 1504 lbf

Resultant Load 5226 3198 2201 lbf

Min Bolt Tension 3900 3900 3900 lbfAvg. Bolt Tension 5500 5500 5500 lbfMax. Bolt Tension 6500 6500 6500 lbf

Min Resistance to Load 7020 2340 2340 lbfAvg Resistance to Load 9900 3300 3300 lbfMax Resistance to Load 11700 3900 3900 lbf

Min. FS Friction Resultant 1.34 0.73 1.06Avg. FS Friction Resultant 1.89 1.03 1.50Max. FS Friction Resultant 2.24 1.22 1.77

WithCoating

WithoutCoating

Courtesy Mike Kalish

Implementation

• Two prospective vendors identified, and samples received from each…

Braziler Technology, Inc, East Windsor CT

White Engineering Surfaces Corporation, Newtown PA

• President of Brazilier has been very cooperative…

-answered many questions-Appears very interested to develop product-Visited PPPL-Researched options and came up with preferred scheme using diamond grit abrasive with resin binder (non-conductive and non-magnetic)

• Included White on bidder’s list but do not know much about capability

Highlights of SOW for Coating• Surface coating shall consist of a diamond grit with waterproof heat resistant resin binder, or other binder material as may be proposed by subcontractor in writing and approved by PPPL in writing. Abrasive shall be grit size in the range 60-80 based on the Coated Abrasives Manufacturers Institute (CAMI) standards. Materials shall be non-conductive and non-magnetic.

• Two 1/2” SS sample plates supplied by PPPL shall be coated and returned to PPPL for test

• PPPL may elect to proceed with the production run at any time, independent of the status of the samples, based on schedule considerations.

Eight Surfaces to be Coated

Top Assemblies Bottom Assemblies

• Surface coating shall not exceed 0.015” thickness based on 8 measurements per disk located 90 degrees apart on ID and OD, before and after coating

Note: actual delta dimension hub to box will be <= 15 mils depending on extent to which diamond grit bites into SS. This will reduce the potting thickness on the middle disk side

Potential Issues

1) Samples tested used “electroless nickel” binder, not identical to resin-based binder

-final COF could be different

2) Substrate material of samples used unknown material, not 304SS

-final COF could be different

3) Present plans call for assembly of boxes, potting, detachment, testing, and re-assembly

- impact of multiple compressions on coating is unknown, although Braziler thinks it will not be a problem - final COF could be different - fit-up and seating during re-assembly could be impacted by presence of coating

•Testing of sample parts in advance of committing to coating of actualparts would reduce risk of “gotcha’s”

•Sandblasting would probably be required to remove coating if it is found undesirable