Above-Ground Contained Plutonium Experiments
Seymour Sack, Laboratory Associate
Lawrence Livermore National Laboratory
Hydrotesting, using simulants for fissile materials, has been and will continue to be
an essential part of the primary design and evaluation process, both for new designs and existmg designs; e.g. WR stockpile primaries.
Although speed and economy dictate that most design experiments are numerical experiments, using ever more sophisticated computer codes, hydrotest data (and underground test data) are essential links between the numerical experiments and reality. Continuing improvements in accuracy and detail of the hydro and nuclear data mirror and drive corresponding improvements in the numerical experiments.
Pin techniques have been basic hydrotest tools for over 45 years and will continue to be basic for designing and understanding implosion systems.
Optics techniques have been used to study specialized details of implosion system behavior. The use of more sophisticated techniques, e.g. VISAR and Fabry-Perot interferometry, are providing even more detailed understanding, especially with the promise of multiple views.
Dynamic radiography has provided important knowledge of primary functional considerations and interactions of a primary with its surroundings, and will continue to be an essential design tool for the foreseeable future.
Deep penetration radiographic facilities, e.g. PHERMEX (LANL) and FXR (LLNL), provide the possibility of "core-punching", carrying the evaluation of an implosion system substantially deeper than can pin techniques - essentially as far as is attainable for a non-nuclear experiment. As such they provide even more stringent tests of implosion system calculations. FXR has somewhat more capability (image quality-resolution and penetrating power) than PHERMEX and will be substantially improved over the next 1.5-2 years by upgrades already funded and in progress. The projected beam parameters for the DARHT accelerator should provide additional resolution and penetrating power for thicker systems. Provided that intermediate milestones are met, there is a reasonably good chance that the DARHT design parameters will be achieved.
It is important to understand that core-punching, either with present capabilities or with the improvements represented by the FXR upgrade or the DAURT accelerator, is not a revolutionary new design tool. It is simply a more detailed implosion evaluation tool, making more stringent demands on the numerical tools which are ultimately the basis of the primary design and evaluation process. Its importance, like the importance of the other hydrotest techniques, is independent of the continuation of underground testing or the likelihood of more restrictive test limits or a CTBT. For most of the past 20 years and for the foreseeable future, the number of underground tests has been small enough that a responsible approach to nuclear testing of new designs should require previous use of all relevant hydrotest techniques (and thorough numerical use of those hydrotest results).
Even in the absence of underground testing and the absence of requirements for new primary designs, continuing hydrotesting with thc best available techniques and use of test results in upgrading numerical models will lead to a continuing and upgraded understanding of WR stockpile primaries and continuing maintenance of the ability to understand and respond to changes in stockpile systems - as well as the ability to respond with new designs to flew requirements.
For a balanced judgment of the benefits of DARHT, it is first necessary to dismiss most of the material presented in support of DARHT. The presentation approach seems to have started with the desire for a major experimental facility, followed by a somewhat forced and awkward creation of design necessities for the potential DARHT capabilities. The result ignores the major role of calculations in modern design practice and is somewhat inconsistent with historical practice at LANL The considerations outlined below are based on the writer's experience and understanding of primary design and the interrelationship of numerical experiments, hydrotest data and underground test data, as well as a reasonable degree of familiarity with past and present LANL practice and designs.
Plutonium is a complex material, and many features of its behavior in implosion systems are poorly understood. However these features arc best evaluated in detailed small-scale experiments, with facilities such as the diamond cell or gas guns using equation of state evaluation techniques or sophisticated optical techniques, and are not particularly amenable to radiographic diagnosis of full scale implosion systems. The expense and difficulties of the Appaloosa program, the limited test frequency forced by limited availability of Cider material and its lengthy and expensive reprocessing requirements, force a critical look at the utility of Appaloosa tests in developing understanding of a primary design. In fact, except for a special circumstance, the Appaloosa program has been dormant for the past 10 years, with several major and enduring WR stockpile primaries totally ignored. There has not been a pin shot for the B61, although the special alloy used in that system might have justified one. The W76/Mk4/Trident 1 warhead has not been evaluated in a core-punch test. The same applies to the W78/Mk12A/MMIII warhead, which has a very limited hydro and nuclear test data base. There appear to be no plans to remedy this situation.
The possibility of directly measuring "mix" has been presented. However modem understanding of primaries involves much smaller quantities than older models, and only a fraction of this smaller quantity is present in a non-nuclear phase. Measurement of even this portion requires very optimistic assumptions as to DARHT performance and unrealistically idealized assumptions about device symmetry and spatial distribution of that portion of "mix".
The possibility of determining the degree of supereriticality, for both normal implosions and safety-related non-ideal implosions, by mapping density or more precisely density x radius has been offered... While this is an interesting curiosity, if valid, it has long been superseded by now standard computer calculations. Safety situations with axial symmetry have been well in hand for more than 25 years, and an extensive nuclear test data base exists, ensuring that ever more refined calculations will remain in touch with reality. Over the past 5 years, equivalent techniques and a very extensive nuclear test data base have been developed for purely three-dimensional safety related situations. These techniques are continually improving in speed and sophistication - - a series of Appaloosa shots, starting 5-6 years from now and necessarily spread over several years for one particular geometry are hardly competitive with even current calculational methods.
The second axis of DARHT is intended to allow more accurate tornographic reconstruction and for a reduced number of hydrotests, especially for systems without axial symmetry, apart from possible increased background/noise problems for each image, and somewhat decreased flexibility in device location, the proposal seems to once more assume the need for a purely experimental determination of detailed implosion geometry - ignoring or at least slighting the computational developments of the last three decades. There is no benefit justifying the cost of a second axis - - considerations should be limited to a single-axis machine, SAHRT.
The one valid justification for DARHT (actually SAHRT) is that PHERMEX is old, difficult to maintain, with the expectation of great difficulty in obtaining replacement parts in the out years. Normal upgrade/revitalization practice is to plan for a replacement, which of course would incroporate modem developments and capabilities. Given business as usual, and an adequate line item budget, SAHRT certainly has as much merit as most past and present line items, and would deserve to be quickly approved. There are, nevertheless, some reservations. Following current schedules, it will be 5-6 years before DARHT/SAHRT is able to be used for real cxperiments. If we examine the very small number of major hydrotests conducted at LANL over the past three years (and the somewhat larger, but still modest number at LLNL), a period which included active Phase 2 and Phase 3 developments, we must be skeptical about the utility/utilization of an expensive new facility 5-6 years from now, with no weapon development programmed for the foreseeable future. More speculative reservations assume that the cost/duration of the DARHT development make it likely that it would be the last new development in penetrating radiography. Pulsed power approaches such as SABRE can be simpler and less costly, although any commitment as alternate to DARHT is clearly premature. Soviet developments in pulsed power and applications to dynamic radiography may become available to the West. There is the possibility that, as with all major projects of long duration, DARHT may be technologically laggard at completion.
If line item funding is sufficiently limited that DARHT/SARHT is not approved, we will not see a catastrophic loss of capability or flexibility. PHERMEX and FXR, with intelligent use of reduced opacity simulants, optimal collimation and device sealing if appropriate, can supply the late time implosion data for a stringent test of implosion calculations. The facilities also serve as mutual backups. The near term FXR upgrades, funded and in progress, should supply nearly the resolution and image quality claimed for DARHT/SAHRT. We should assume joint use of both facilities as appropriate, with visiting laboratories assured of adequate scheduling priority and charges limited to true incremental/avoidable costs. Aggressive use of both facilities will allow LANL to fill out its hydrotest data base on important stockpile systems, a program already well underway at LLNL. Such use, along with willingness to fund device hardware costs (a past limitation) should be a prerequisite to any action towards new and improved facilities. With an aggressive program of supporting calculations, this will help to ensure the existence of primary design and evaluation capabilities (often naively taken for granted) and the maintenance of those capabilities.