1. NCSX VACUUM VESSEL HEATING/COOLING PL Goranson Preliminary Results February 17, 2006 MDL Testing of Coolant Tracing NCSX

2. Talk Overview P. Goranson GOALS TEST SUMMARIES TEST RESULTS K. Freudenberg ANALYSES ANALYSES/TEST COMPARISONS P. Goranson CONCLUSIONS/RECOMMENDATIONS/STATUS

3. NCSX GOALS Compare thermal performance of smooth tube vs. braided corrugated hose. Benchmark ANSYS analyses. Project the test results across to VV cooling performance. Determine baseline design for VV heating/cooling tracing configuration. Evaluate alternative conductive materials (thermal caulk). Measure pressure drop in braided hose. Test effectiveness of Copper ground strap (doily)

4. NCSX MDL TEST SETUP

5. NCSX TEST ASSY Utilized existing MDL resources –Oven –Thermocouple controllers –Braided hose –Building compressed air for cooling Setup geometry was not meant to be completely duplicate VV system. A helium source was not available nor was very high pressure air. - Curved SS plate represented VV NSTX center stack mockup with studs - Cu saddles were fabricated to correct geometry - Hose was prototypic Smooth tubing test used Cu tube with SS insert to represent tube wall - permitted using same saddles and at same time duplicating hose flow parameters (h, v) - large SS tube overloaded building capacity Test Assembly With Braided Hose ss Tube Assy Saddle Tube Gasket

6. NCSX AIR FLOW/PRESSURE/TEMP MONITORING

7. NCSX TEST COMPONENTS Saddles –Cu precision fit to hose & tubing insert Saddle Gaskets –1/16” Grafoil –0.005” Cu shim Tube and Hose gaskets –0.005” Cu shim –0.01” Grafoil Clamps 1/8” x 1” steel to permit large clamping force

8. NCSX TEST PROCEDURE Test assembly was covered in insulation and heated to test temperature by the oven. When assy temperature equilibrated, oven was turned off; door remained closed. Temperature of thermocouple at varying positions around saddles were recorded in 2 minute intervals, for 30 minutes total. Air flow (scfm), temperature, inlet pressure, and outlet pressure across test section were monitored. Tests were performed using different combinations of gasket materials under the saddles and under the tubing inserts.

9. Straight tube Approximation TC Locations NCSX

10. TEST RESULTS braided with both Grafoil smooth, Grafoil gasket braided, Cu gasket smooth with both Grafoil braided, Grafoil gasket smooth, Cu gasket Braided hose does cool.

11. NCSX Ground Strap (doily) Results of the test were disappointing. Test performed with 0.04” x 1” x 6” Cu strap wrapped and clamped to hose with Grafoil. Other end bolted down by stud, washer, and Grafoil gasket. Results - temperature drop after 15 minutes – 6.5 C - temperature drop after 30 minutes – 13 C

12. NCSX FACS Un-cooled plate temperature fell less than 2 degrees C in 30 minutes, none in 15 minutes. Cooled plate dropped by as much as 24 C in 15 minutes. Cu foil under the braided hose decreased performance. Grafoil under the braided hose increased performance. Grafoil under the smooth tube improved performance. Grafoil under the saddle improved braided both hose and smooth tube performance. Grafoil was required under both the saddle and the tube insert to attain the theoretical cool down time of the smooth tube. Pressure drop in braided hose is ~6 times the equivalent smooth tube with same I.D., At these flow regimes, twice the manufacturer’s estimate.

13. NCSX CONCLUSIONS Where we are headed: Braided hose does not perform as well as smooth tube but is good enough. We propose to go with it. Neither tubing nor hose is proving to be as effective as previously predicted. Braided hose is: more cost effective (more studs cost a lot less than segmented hose) easier to install quicker procurement time more suppliers probably more reliable The VV will require twice the number of studs and saddles, and a redistribution of some of the tracing, regardless of whether or not the tracing is tubes or hose. Presently, the only alternative to hose is segmented hose made up of short braided sections intervened with shot straight tube sections, essentially a lot of braided hoses with end fittings welded together. None of the thermal caulks tested has proven to be advantageous. Only one, Pyro-Putty 950, adhered and exhibited structural integrity through bakeout; unfortunately its conductivity was 7.4 time worse than saddles.

14. NCSX Design Changes Simplified saddle assy Saddle is two pieces rather than one Reduces machining Assures good alignment and clamping Two clamps rather than one Permits spacing tubes further apart in some locations One saddle cools almost as well as two so wider spacing can be more effective WAS NOW

15. STUD LOCATION INDICATED WITH 1INCH RADIUS BALLS STUD LOCATION THE RIGHT HAND SIDE REPRESENTS A 5 INCH SPACING BETWEEN STUD LOCATIONS THE LEFT HAND SIDE IS THE ORIGINAL SPACING. THE SPACING VARIES FROM A NOMINAL OF 8 INCHES +/- 2 INCHES. THE TOTAL NO. OF STUD LOCATIONS ON THE RIGHT HAND SIDE IS ~ 356 FOR A HALF PERIOD THE TOTAL NO. OF STUD LOCATIONS ON THE LEFT HAND SIDE IS ~ 226 FOR A HALF PERIOD THE STUD LOCATIONS HAVE NOT BEEN OPTIMIZED RELATIVE TO THE MAGNETIC LOOPS. TOP VIEW

16. STUD LOCATIONS ON THE FRONT AND BACK SIDE 5 INCH SPACING BETWEEN STUDS NOMINAL 8 INCH SPACING +/- 2 INCHES 5 INCH SPACING BETWEEN STUDS Outboard view Inboard view

17. STUD LOCATIONS ON THE FRONT AND BACK SIDE NOMINAL 8 INCH SPACING +/- 2 INCHES 5 INCH SPACING BETWEEN STUDS Inboard view

18. HEATING/COOLING CLAMP ASSY GRAFFOIL 2.5X1.25X.06 COPPER 1.0X1.25X.333 GRAFFOIL.01 THICK STUD 5/16ODX3/4LG WASHER NUT HEATING/COOLING TUBE.48OD CLAMP