Failure of OriginalTF Inner Leg Assembly

C Neumeyer4/10/3

0

10000

20000

30000

40000

50000

60000

70000

0 0.5 1 1.5 2

Time (sec)

Itf (amp)

110201

110200

February 14, 2003: Following our morning “test shots”, the firstplasma attempt of the day resulted in a loud bang (heard on thecontrol room audio monitors) accompanied by a plume of smoke(visible on the control room video monitors).

Test Shot

Failed Shot

• Target level was 53.4kA which produces Bt=4.5kG

• Fault occurred just prior to flat top as the current passed 50kA

• Several protective devices tripped within milliseconds....

- TF power supply fault detector section overcurrent, - TF Analog Coil Protection (ACP) overcurrent,- TF Rochester Instrument System (RIS) overcurrent, - TF ground fault current relay.

Initial inspection revealed that one of the TF “flags” on the bottom end of the machine was displaced radially by about 1 inch

TF Flags

Subsequent inspection revealed the extent of the damage

Peak Damage Region Mating Flags

PEAK DAMAGE AREA

Peak DamageArea

Start & Finish Leads(V = 1kV)

Fault Scenario• An open circuit fault (the flag joint

opening up) led to multiple turn-to-turn and turn-to-ground faults (to the hub and umbrella assemblies) at or near the high-voltage terminals of the TF circuit.

• Spike of fault current from the power supply shunted the coil inner/outer leg assembly.

• Once the power supply tripped, the current spike decayed.

• This was followed by an L/R decay of the coil current as the coil released its stored energy. The L/R decay can be modeled by fault with V=125V and R=500.

• The energy dissipated in the arc was of order 1.4MJ.

Spike Current

Coil Current

Current Decay Waveshapes

Varc=125V Rfault=500

Normal vs. Fault L/R Decay Curve Fit to Fault Model

Other “glitches” in current decay are explainableby mutual coupling to OH power supply circuit along power cable run

MECHANICAL/ELECTRICAL FAILURE OF JOINT

TF Flag

Hub Assembly

Hub Assembly

TFBundle

“Keensert[tm]” Shim

Preload(bolts)

EMLoad

HISTORY OF TF OPERATIONS• Approx. 7200 Shots mainly at 3kG, 4.5kG• Limited number at 6kG

3kG

4.5kG 6.0kG

FAULT

TF Pulse Spectrum

Max I2T = 5.46x109 A2-sec, T = 65oCRated I2T = 6.5 x109 A2-sec, T = 80oC

3kG

4.5k

G

6.0k

G

PRECURSORSEvidence of a problem surfaced after ‘02 run period

Also, loose bolts and broken inserts were discovered

RESPONSE TO PRECURSORS• Corrected Various Defects

–Resurfaced non-planar flag faces and chamfered bolt holes (bottom)

–Improved bolt washers and bolt retention (top and bottom),

–Retorqued bolts (top and bottom) and replaced 4 “keenserts” with “tap-lok” inserts (bottom),

–Replaced G-10 flag shims with inflatable epoxy shim design (top and bottom)

• Initiated more detailed FEA• Returned to 4.5kG limit• Initiated more regular inspections

Too Little, Too LateShim

FACTORS LEADING TO FAILURE• Design Factors

– hub stiffness not adequate to react moment– communication of load from flag to hub uncertain with G10 shims – bolt thread and shoulder engagement too small– bolts necked down too far at threads, not enough on shaft– dual shear/preload function of bolts– lack of feature to facilitate joint resistance measurement w/o disassembly

• Quality Factors– frequent manual reworking of contact surfaces– non-planar flag surfaces– shoulder bolt concentricity

• Operational Factors– monitoring of joint integrity too infrequent, too imprecise

ALL OF THESE FACTORS HAVE BEENADDRESSED IN THE NEW DESIGN