First ideas for the design and construction of  UST_2  fusion experiment    



Abstract :  Present ideas about the design of a new fusion experiment. The size will be limited to the economical resources. A stellarator is chosen among other devices (FRC, RFP, Spherical tokamak...). At the moment only an estimation of the dimensions, parameters and features is established in order to  work out the feasibility, the approximate cost and to show the likely feasibility of a full 3D stellarator at low cost.



    Ideas that are analysed : 

Summary of different devices (Dipole, FRC, FRP, Spherical torus).  

Choice of the style of fusion device.


Size and style of stellarator.

Bo , TF power and ECRH frequency

Energy confinement time and Te max.

Plasma Heating

Frame for the coils

Vacuum vessel





   One and a half years ago a similar document was written to clarify the alternatives to design a small toroidal device for fusion research in engineering. The result was UST_1 stellarator which was finally designed and built as a partially optimised modular stellarator with external resistive coils on a circular toroid, plaster frame for grooves and copper vacuum vessel. Major R=125mm and minor radius = 21mm




   Choice of the style of fusion device


   Dipole, FRC, RFP, Spheromak, Spherical Tokamak and  IEC were analysed in  [1] [2] [3] . An insight but no solid reasons are obtained from the information.


   General attractiveness differs from feasibility in a small low cost device although it has some relation.


   Dipole is almost impossible for DT fuel because of the heating of the levitating ring (or huge size of the shield-ring) and sputtering of first wall of the ring. The issue of the cooling of the central ring is still critical even if D-3He is possible. The rest is relatively simple, high Beta and confinement. Pulsed operation possible.  Slightly large device. It is attractive but not simple for a low cost device. Not chosen here. 


    FRC is very attractive for D-3He and it has advantages even for D-T. Presently impurities are an issue and the temperatures achieved are very modest (~Tt=50eV). Easy maintenance, direct conversion possible, open divertor. In D-T fuel, one side can be opened for maintenance of the blanket. Low magnetic fields. The possibility of D-3He should be further proved. More funds necessary. The critical issue is reliable and low cost sustainment of flux without degrading confinement. Not chosen here because of the difficulties in the formation of the plasma, impurities and sustainment of flux or current drive.


   FRP is attractive if highly effective, efficient and low cost current drive is possible (no significant bootstrap, OFCD might be possible, NBI is impossible because about 1000MWe are necessary for TITAN reactor at 33% efficiency). D-3He is not promising. High loads but even less space than in ST. Fast control of instabilities necessary for steady-state.


    Spherical torus would be attractive if the central post needed lower electrical power (superconductors seem impossible and D-3He is not promising in ST due to relatively low Tau E, 250MW needed for a Cu-post reactor). Loads on first walls, divertor and central post are enormous but feasible because of the easy and low cost (small surfaces) maintenance through the central post opening. Li-walls. Bootstrap current is enough (80% or more achieved). No chosen here because : extreme power and difficult CS coils-central-post if inductive start-up and current drive, non-inductive start-up and current drive is not possible at low cost.



  Current drive and flux sustainment:  In all the devices except for Spherical Torus and Stellarators current/flux drive is the main issue (450MWe need to be supplied to drive 50% of the Plasma current in ITER in steady state [4],  pg 2529. Obtained supposing 33% of efficiency, value obtained in JT-60U NBI-N.).





In general, not specifically for a small low-cost device, insight gives the next order of attractiveness :


1) FRC is the best (D-3He, open divertor, direct conversion)

2) Compact Stellarator and Spherical Tokamak are equal.

3) FRP

4) Dipole

[no insight for Spheromak]


Anyone can discuss this.







A  Compact modular stellarator is chosen for the new small low cost fusion device. The reasons are : device and controls are simpler, plasma can exist at any temperature, long pulses so easier diagnostics, past experience in UST_1 will help. For reactors : 2nd stability regime might be possible (up to Beta 8%), they are also promising as reactors, stable, low recirculating power, other.





    Many systems in the present installation (UST_1) are usable for a stellarator 50% , two and even three times larger. The vacuum system might need an improvement of ~100 to increase the conductance from the diffusion pump to the stellarator but nothing more. The present control and data acquisition system is enough because plasma experiments are not the focus in this research. Heating system at 2.45GHz is very cheap but higher frequency is convenient.

   The power supplies  and/or superconductors are the main issues. Large copper coils or another very creative idea is necessary. Superconductors for UST_1 cost 3000. For UST2_0.2m_8L_0.1T HTS will cost ~12000 if a creative idea is not found. So it is compulsory to find an innovative solution.


   In principle  UST_2 is not intended to be constructed without some funding and/or interest from foundations, university, company, sponsor, etc. The construction of UST_2 will depend on many factors so it is not assured.


   Around 3000 might be obtained  from different institutions or from personal savings to build UST_2.




Size and style


   A style similar to MHH2, HSR-3, NCSX, QPS, W7-X (not compact) is adequate. Perhaps the definition of coils for one or more devices is received. At the moment only HSX coils are received but HSX style is not considered here because Ap seems excessive.


   A volume of 3.4 litres is obtained for a device ~50% larger than UST_1, and 10L for a 2 times larger, for the same Aspect ratio = 6 (for Aspect ratio =4 then plasma volume =~ 8L for 50% larger, twice then 25L ). Notation :  UST2_0.2m_8L... means major radius=0.2 and plasma volume=8 litres.


   The option of twice the size of UST_1 with Ap=4 results in even more powerful supplies for TF coils, so this alternative seems impossible at low cost.




Bo , TF power and ECRH frequency


   High Bo is decisive because Tau E is almost proportional to B, ECRH allows higher densities without additional complexities (EBW-overdense heating...). 


   In all the next cases the stellarator is  UST2_0.2m_8L...

  Conductor is 16mm2 but it is not checked if there is enough room for the winding pack in the inboard side.


At 0.1T, cut-off and densities similar to UST_1. Rough power supplies : 12 coils, 12 turns, ~0.8kA in conductor, 10kA-turn per coil, total power for 12 coils =~95kW, ~110V).


At 0.3T, then ~8GHz for 1st and 16GHz for 2nd harmonic. and ~1e18m-3 cut-off density. Rough power supplies : 12 coils, 18 turns, ~2kA in conductor, 30kA-turn per coil, total power for 12 coils =~650kW, ~340V). Code UST2_0.2m_8L_0.3T


Around 1e19m-3 for O-mode and 2e19m-3 for high density cut-off for X-mode are achieved at 28GHz and  1T for 1st harmonic. This values might be impossible due to Tau E and line radiation but it is really impossible due to cost and magnitude of the power supplies (12 coils, 24 turns, ~4kA in conductor, 100kA-turn per coil, total power for 12 coils =5MW, 580V).


  Any magnetron/klystron from 5GHz to 24GHz, with heating at 1st 2nd or 3th harmonic could be useful and must be sought.


   An enormous effort should be done to obtain low cost HT superconductors, otherwise the power supplies are mammoth size. However the alternative of superconductors was rejected for UST_1 because it was very expensive and difficult.




Energy confinement time and Te max.


Case UST2_0.2m_8L_0.3T

Iota supposed=0.333 ; Ap = 4 ;  Enchancement factor = 0.2 (similar to TJ-II and TU-heliac and twice than supposed for UST_1). ISS04v1 is used.


n=1e18m-3 , Te max = 10eV , P ECRH absorbed necessary = 2kW.  


n=1e18m-3 , Te max = 20eV , P ECRH absorbed necessary = 9kW.  


n=1e18m-3 , Te max = 50eV , P ECRH absorbed necessary = 120kW, impossible.


Thus Te max. will be about 20eV.


Energy confinement time Tau E=23ms for  P=2kW and Tau E=9ms for P=9Kw.




HTS superconductors or resistive coils


   Resistive coils have two main disadvantages : 1) The required power is huge and so power supplies

2) If banks of condensers are used then the pulse is very short and diagnostics are expensive or impossible.


   HT Superconductors have several advantages : 1) They do not need much power. A drawback is the low internal resistance of the power supply and the  low voltage. However creative methods may be found to obtain low cost, low power, low voltage, low resistance, high current power supplies.

2) It is profitable to dedicate an intensive intellectual  and experimental effort to obtain low cost HTS or to find creative methods to obtain HTS from companies (2nd hand, out-off-specifications, short pieces ...). The first e-mail asking for this possibility has not been answered yet. HTS for UST_1 costs 6000 from EAS.

3) HTS may withstand the stress of 0.1 to 0.3T. (Need rough calculation) Bi-2223 in overcooled N2 is almost able for 0.3T but Ic needs calculation. YBCO is ideal but not commercially available in long pieces and the price will be extremely high. Bulk YBCO was studied for UST_1 but some doubts remain.


Superconductor coils in a boiling pool could be a simple and low cost option.


Other possibilities (N2 Cooled Cu, etc) ?



Plasma Heating


   The same reasons as in UST_1. Only ECRH is relatively simple and feasible at low cost. From 2 kW for Te=10eV to 9Kw for Te=20eV is necessary. Line power about 4kW for the first stage (10eV). O-mode, X-mode and no EBW mode conversion is considered.



Frame for the coils


   The device to mechanise coils may be possible for Ap=4, but surely not for Ap=2.7.

    Some new ideas are emerging. At the moment no one is as economical as the "Toroidal Mechaniser".



Vacuum vessel


Methods:  explosion (style HSX).  Welding pieces (expensive and diffiuclt). Glass mould (perhaps, excessive viscous?), plastic (outgassing?, moulds?, forming-glue?), reinforced soldered copper (similar to UST_1)?,  cast aluminium?, other creative ideas?







    Necessary calculations and further research


   Calculations and estimations to decide about superconductors or resistive coils. Check if Stellarators with aspect ratio = 4 can be built with the "Toroidal Mechaniser". How to build the frame for the coils at low cost.  Expand the calculations carried out for UST_1 in relation to HTS.






[1] "Comparison of innovative concepts to help in the decision of the UST_2 fusion experiment. Dipole and FRC (Part 1)."  Vicente M. Queral. See "List of all R&D"

[2] "Comparison of innovative.... RFP and Spheromak. (Part 2)". Vicente M. Queral. See "List of all R&D"

[3] Comparison of innovative....  Spherical Tokamak and IEC. (Part 3)", Vicente M. Queral. See "List of all R&D"


[4] "Chapter 6 Plasma auxiliary heating and current drive"  ITER Physics Expert Group on Energetic PArticles, heating and current drive  et al.





Date of publication 19-03-2007