23
July 2003 PM Telecon
Summary
The list of critical technical issues was discussed in yesterday’s telecon with ORNL (1:30-5:50pm) and in a telecon with Brad Nelson and myself this morning. This was really the first opportunity to review the critical issues and updated plans with Nelson and Williamson since the July 9 telecon due to vacation schedules. The following is a summary of those discussions. I took the liberty of including as part of this summary, some developments that were not discussed in those telecons. Please read through this summary carefully so that action items can be picked up and we can all be singing from the same sheet of music. Please advise of any changes (corrections and omissions) that need to be made.
Winding form geometry. The design of the modular coil winding forms has been and still is work in progress. It is a problem that will not be solved until it is solved. The current forecast is as follows:
01 August ORNL
model of winding forms complete (Williamson)
04 August Suppliers provide their solutions to
NCSX project (Suppliers)
08 August ORNL updates their model based on
supplier inputs (Williamson)
15 August Drawings generated of prototype winding
form (Williamson)
19 August Product spec for prototype winding form
updated (Heitzenroeder)
22 August FDR for prototype winding form
(Heitzenroeder)
The forecast date for ORNL to complete modeling of the winding forms is a week away. It is recognized by all parties that completion of this work is critical to a successful PDR. We have been encouraged by the rapid progress made by the suppliers using modeling tools different from the ones we have employed, but we are still not there yet.
Clamp design modifications in trough region. The winding packs come quite close together (a minimum separation of 0.1”) at their base, but fan quite far apart at the top of the tee. The plan is to provide a “spring-loaded” clamp that presses the cooling plates against the winding in this region and to rely on EM forces to hold the winding against the tee. Review of the forces on the conductor at the time of the CDR give us a measure of optimism that this will work. HM Fan will provide a detailed picture of forces in this region to ascertain whether there are any structural requirements for clamps in this region. It is expected that these results will be available by 08 August. Delivery of the winding geometry is required in order for Fan to start on this task (Williamson).
Material properties
The
performance of the modular coils will be limited by thermal stresses. In order to assess the magnitude of thermal
stresses, an understanding of material properties is required. We need to know the stiffness, strength, and
coefficient of thermal expansion for the winding pack. In addition, we need to know if there are
thermal strains resulting from the cure cycle.
To gather this information, tests are being conducted at PPPL and
CTD. Tests conducted at UT for the QPS
program might also provide relevant data.
At
PPPL, we are measuring the change in length of potted cable conductor from
before curing and after curing. These
test results were provided after the meeting by Raftopoulos and have been
included in the meeting records. This
information will tell us the thermal strain we are starting out with.
At CTD, tension, compression, and flexure tests long with tests to measure the coefficient of thermal expansion between 76K and RT are being conducted. We have received initial compression and flexure data. The compression results are bewildering, particularly at RT, showing substantial scatter. The modulus data and drops markedly from RT (37 Msi) to 76K (5.3Msi). This is the opposite of what was expected. Its stiffness in compression is about twice copper at RT and 1/3 of copper at 76K.
The compressive strength increases over the same range. The compressive strength and modulus at 76K are 333MPa (47 ksi) and 36GPa (5.1Msi). The ultimate (yield) strengths of hard C107 copper is 42 (41) ksi at the same temperature. It is a bit perplexing why this material appears to be stronger in compression than either of its constituents.
Flexural data was only provided at RT so far. The data indicates a flexural strength of 191 MPa (27ksi) and a modulus of 27GPa (3.9Msi). Flexural data had much less scatter than the compression data previously reported.
More work is required to complete the prescribed test program and to understand the bewildering results compression results (CTD, Zatz). Zatz has contacted CTD and has elevated the priority for completing the CTE tests. Testing and developing of material properties and design criteria will be tracked against the following schedule:
08 August
– Testing completed and documented (CTD)
15 August
– Recommended material properties documented (Zatz)
22 August
– Design criteria updated to reflect CTD/PPPL test data (Zatz)
Stress Analysis
The global structural model is a linear model and is being developed by HM Fan. The model is for a full field period; includes the windings, tees, and shell segments; and has more detail than the model presented at the CDR. The winding packs are tied to the tee and do not accurately represent the local deformation of the conductor due to differential thermal strain between the winding pack and structure. An analyst from ORNL, Kevin Freudenberg, is developing a non-linear model of just the tee and winding (and clamps) that should provide a better understanding of the magnitude and response to differential thermal strains. Results from both models are expected by the end of August (Freudenberg, Fan). In addition, Steve Chay from ORNL is developing is simpler global model whose results will be used to guide the design development in the next several weeks.
For assembly, our strategy is to place each component in its ideal location, again using a shim-as-we-go approach. The metrology working group is working to identify a suite of measurement tools that will allow us to position the modular coils during field period assembly and machine assembly to within the prescribed tolerance (Raftopoulos et al). We are basically following the approach used on ATF. The coil winding forms on ATF (a 20’ diameter machine) were assembled round and flat to a tolerance of 0.010” using theodolites and a shim-as-we-go approach. Thus, there is an existence proof that this approach has worked before. Documentation of the successful ATF experience should be included as part of the PDR documentation (Nelson).
Tolerance requirements. Brooks has already conducted studies that the nominal 1.5mm winding tolerance can be relaxed for coil locations away from the plasma (even for the outer legs of the modular coils). Brooks will document the results of the analyses completed already and complete key missing analyses prior to the PDR. This work will be provided as part of the PDR documentation.
Other sources of field error. Brooks has calculated the field error from the bending magnets in the neutral beams and they do not seem to be a problem. We are concerned however about field errors from structural steel in the test cell floor and walls. The field errors from the structural steel will NOT be quantified prior to the PDR, but will be addressed in early FY04 (Brooks).
UT Coil. The UT coil is presently being tested at ORNL. Faculty and students from UT are modeling the coil. Test results from the coil will be used to validate the models, which in principle would then be applied to the actual modular coil designs. ORNL is requested to provide timely feedback on the results from the UT coil tests and modeling (Nelson).
Racetrack coil. The racetrack coil is currently being fabricated at PPPL. ORNL has requested that Teflon plugs be installed (that can be removed after fabrication) in locations that would provide visibility of the separation between the winding packs and the tee sections (Chrzanowski). Once it is fabricated, it will be tested at a facility currently being designed/built at PPPL. In order to finalize the design of the facility and to execute the tests, a test plan for the racetrack coil needs to be provided by ORNL (Nelson). At PPPL, we need to review the plans, resource requirements, and personnel assignments required to set up the facility and test the coil. A meeting has been tentatively scheduled for Monday 28 July at 3pm with Gettelfinger, Chrzanowski, Strykowsky, and myself to review these points. The design of the test facility should incorporate prototypical features of the modular coils where practical (Gettelfinger). For instance, if we are planning to use “kick-less cable” to feed the modular coils, we might use this opportunity to gain some experience with it. Also, we might incorporate a prototype of our design for the current feeds through the cryostat (avoiding ice balls and heat leaks to the coil).
Twisted racetrack coil. Completing the twisted racetrack coil is no longer a priority for the PDR. Basically, we recognized that this would not be possible without risking completion of other, higher priority tasks for the PDR. However, we would like to get the casting ordered as soon as possible. ORNL needs to work with Chrzanowski to determine how this can best be accomplished (Nelson). One of the things we had hoped to get out of the twisted racetrack coil was a demonstration that we could indeed easily apply the chill plates to a warped surface. During the telecon, Chrzanowski indicated that some tests had already been done on our twisted tee section. Chrzanowski should document those tests and determine if additional tests should be done to build our confidence in the feasibility of this approach and the reasonableness of our cost and schedule estimates prior to the PDR.
The vacuum vessel shell geometry and spool piece geometry have been defined and satisfy field period and final assembly requirements. The geometry of the shell has been passed to ORNL. ORNL is in the process of updating the vacuum vessel model, which includes all the ports, to be consistent with the revised shell and spool geometry. This work is expected to be completed by 30 July (Cole). The remaining critical issue is the design of the closure joint. The first decision is whether it should be bolted or welded. Engineering considerations would favor a bolted joint. The only issue with a bolted joint is whether the envelope required is too big to fit between the shell and the plasma. An assessment of the feasibility of the bolted joint concept will be provided on 30 July (Goranson). At that point, a decision will be made whether to go forward with the reference bolted joint concept or switch to a welded joint, such as that conceptualized by Tom Brown.
A revised support concept, that is consistent with our assembly scheme, was provided at this week’s telecon. It appears that the design as proposed, provides only three degrees of freedom (vertical, roll, and pitch) for positioning a field period of the vacuum vessel. Additional degrees of freedom for final positioning of the vacuum vessel should be provided (Goranson).
The revised support concept
provides vertical support for gravity loads only. Our reference plasma configurations should be vertically stable,
but fusion experiments have a history of operating with plasmas that are
different from those envisioned when the device was designed. Goranson indicated that there are
stellarator symmetric locations below the device that could be used for
reacting upward EM loads. Brooks will generate a vertical disruption case based on
artificially translating a high current, ohmic plasma off the midplane and
quickly ramping the current to zero in order to crudely simulate a vertical
disruption event. Dahlgren will
calculate the stresses in the vacuum vessel and the net loads on the support
points.
Goranson
indicated that with the current support scheme, the thermal growth will not be
symmetric about the midplane. That
seems okay, as long as the vacuum vessel is aligned with the midplane (defined
by the modular coils) at the nominal pre-shot operating temperature of
40C. A
systematic study should be provided as part of the PDR documentation to assure
that we have adequate clearance between the vacuum vessel assembly (including
the ports) and the surrounding coils and structures for all possible tolerance
and temperature combinations (Goranson).