V.NATARAJAN

A recent article by Dr.P.K.Iyengar in the Times of India [1] has stressed the need for continued nuclear testing and cautioned the Government of India against any hasty decisions on signing the CTBT. Dr.Iyengar's long and close involvement with the atomic project including Pokhran-I, and as a former Chairman of the Atomic Energy Commission should naturally compel one to consider his advice carefully. It is thus unsurprising that others [2,3] have referred to his remarks in support of a broadly similar stand. While listing out his reasons against signing the CTBT, he gives a brief and possible description of the fusion secondary of the thermonuclear device (also referred to as Shakti-1 or S-1) detonated on Pokhran-II. This note attempts to analyze those portions of Dr.Iyengar's article and lay out the conclusions that can be drawn from it.

Technical Background

A thermonuclear device/weapon generally has two stages - a fission primary and a fusion secondary. Radiation from the primary is used to compress and heat the secondary, setting off the fusion process. In this section, we look into two aspects of a fusion secondary - the fuel Lithium Deuteride (LiD) and the efficiency of the fusion process.

(i) LiD : Existing as a solid at normal temperature and pressure makes it a convenient fuel for the secondary in almost all modern H-weapons. Lithium (Li) occurs naturally as two isotopes, Li-6 (~7.4%) and Li-7 (~92.6%). It participates in the fusion process by capturing a neutron and producing tritium (T), which fuses with deuterium (D) to produce a highly energetic neutron,

D + T à He + neutron (+ energy)

Pure Li-6D (comprising 100% of the lighter Li isotope) has an energy content of 64kT/kg, while pure Li-7D (comprising 100% of the heavier isotope) has an energy content of 38.5kT/kg [4]. This implies that if a kilogram of Li-6D were to be fully used up in the fusion process, it would yield 64kT. It also implies that LiD made of natural Li has an energy content equal to the weighted average between the two extremes i.e., 40.3kT/kg. In practice, LiD with enriched lithium (with Li-6 content anywhere from 40% to 95%) is used in the secondary [4]. It may be mentioned here that India was reported to have been working on enriching lithium back in the eighties [5].

(ii) Fusion Efficiency : A major concern in designing the secondary is to ensure that the burn within it consumes most of the LiD fuel, before it (the secondary) disassembles and explodes. If the fusion yield for a design with 1kg of Li-6D were found to be 32kT (either post-facto or through prior numerical simulations), the efficiency of the burn could be taken to be 50% i.e., half the fuel in the secondary burned before disassembly. A fusion efficiency of around 50% appears to be in the characteristic ballpark range of many thermonuclear weapons [6,7].

It must be mentioned that the above definition for efficiency deals only with the burn occurring within the secondary and the resulting yield. A broader definition should necessarily consider other factors – among others, the size of the primary and the physical distance between the two stages. A miniaturized design is arguably more ‘efficient’ than a larger design for the same yield and fraction of LiD burned.

Analysis

Parts of interest from Dr.Iyengar's article are quoted below, and an analysis is attempted with the help of the previous section.

"The thermonuclear device had two parts: a `boosted-fission' trigger, and the actual thermonuclear part. The boosted-fission trigger would have yielded at least 20 kilotons, which means the thermonuclear yield of this device could only be around 20 kilotons. About 400 grams or only around 500 ml of LiDT is needed to produce this much energy."

400g of LiD (with a volume around 500cc; actual specific gravity of LiD at normal temperature and pressure is ~0.9) with a yield of 20kT gives an energy content of 50kT/kg (which corresponds to ~45% enriched Li-6). It is evident here that Dr.Iyengar has used a fusion efficiency of 100% to calculate the amount of LiD needed for the fusion yield of 20kT. If the BARC designers had designed for an efficiency of 50%, the probable mass of LiD would be 800g (1000cc). With a 'more realistic' design efficiency of 20% for a first generation thermonuclear weapon, probable mass of LiD is ~2kg (2500cc). The important point to note is that a 20kT yield with 400g of LiD is for an ideal case of 100% burn - in the real design, there are bound to be allowances for lesser efficiencies by using more LiD to yield the same 20kT. The caveat here is that this real design would have to ensure that the larger amount of fuel is uniformly compressed and heated i.e., merely adding more LiD to an existing design doesn’t automatically lead to a higher yield.

"In my judgment, this is a very small volume for the core of the fusion secondary. It, therefore, seems that the secondary (fusion) device burnt only partially, perhaps less than 10 per cent. To get any burn at all in such a complex device is in itself a significant achievement, but clearly the next step would be to improve the device to get greater burn, leading to greater efficiency and smaller size."

Dr.Iyengar proposes here that the actual design yield for fusion was perhaps around 200kT (for 4kg or 5000cc of LiD with 100% efficiency), with the secondary achieving a 10% burn for an actual yield of 20kT. One is not aware if there are other reasons (undisclosed to the public) for assuming an efficiency of less than 10%. However if the reported low volume of LiD (~500cc) calculated for 100% efficiency is the only consideration, the higher mass/volume of LiD for the realistic efficiencies of 50% or 20% calculated previously would pre-empt any such objections. Thus there is no reason to suppose that the 20kT (or ~30kT as suggested by later BARC reports) fusion yield was not close to the actual design yield.

Conclusions

Dr.Iyengar's starting point in his article is the BARC-reported fusion yield of 20kT (or 30kT). He does not - as has been suggested by some - claim that the fusion yield was any less than this.Complete burn of the fusion fuel or 100% efficiency is an ideal scenario likely never approached in real designs. It is not known if the design efficiency for fusion in S-1 was 10% or 20% or 50%. However, if the actual device yield was roughly equal to the design yield (at whatever efficiency), it follows that the device worked as planned. It would be completely incorrect to provocatively assert that the "thermonuclear explosion…was a fizzle" [3].

As Dr.Iyengar goes on to write, if a higher design efficiency is desired – either purely in terms of the yield or in terms of the size of the device/weapon - further testing would be unavoidable. His article also details other technical reasons for not signing the CTBT anytime soon. Coming from such an eminent source, one trusts that they are reaching the eyes and ears of the decision-makers.

References

[1] P.K.Iyengar, "In Testing Times: Repercussions of Signing the CTBT", The Times of India, February 17, 2000 - http://www.timesofindia.com/170200/17edit4.htm.

[2] V.R. Raghavan, "Dangerous Nuclear Uncertainties", The Hindu, March 13, 2000. http://www.indiaserver.com/thehindu/2000/03/13/stories/05132523.htm. Gen. Raghavan refers to a lecture given by Iyengar based on his article in the Times of India.

[3] Bharat Karnad, interview at Rediff.com, June 10, 2000 - http://www.rediff.com/news/2000/jun/10inter.htm

[4] Carey Sublette, "Nuclear Weapons Frequently Asked Questions (NWFAQ)", Section 4 - http://www.fas.org/nuke/hew/Nwfaq/Nfaq0.html

[5] David Albright, "The Shots Heard 'Round the World", The Bulletin of the Atomic Scientists, Vol. 54, No. 4, July/August 1998 - http://www.bullatomsci.org/issues/1998/ja98/ja98albright.html

[6] Section 4.5.2 in [4]. Putative description of the 'Tsar Bomba' tested by the Soviet Union in 1961 - 1700kg of Li-6D yielding around 49Mt at 50% efficiency.