Deixis Online

The unthinkable has become a reality. A nuclear strike has hit the U.S., and a retaliatory long-range ballistic missile loaded with a nuclear warhead streaks toward its target. As it begins reentry an enemy defensive missile explodes nearby creating extremely high temperatures and radiation aimed at disabling the incoming weapon.

Should this doomsday scenario ever occur, the U.S. must ensure that its nuclear weapons arsenal will be able to withstand such defensive missile blasts and successfully strike their targets.

But how to do so?

Until U.S. ratification of the Nuclear Non-Proliferation treaty in 1970 nuclear tests could be conducted to confirm the viability of nuclear weapons, but the treaty put a halt to such tests.

Prior to 2005, the job of creating a hostile neutron environment similar to what a nuclear weapon system might encounter was given to DOE’s Sandia Pulsed Reactor (SPR), housed at Sandia National Laboratory, Albuquerque, New Mexico. When researchers wanted to determine if the electrical components of a nuclear weapon system could withstand being bombarded by a defensive missile explosion they placed the components in the SPR, hit them with short-pulsed neutrons and gamma rays and then examined their condition.

But then came 9/11 and everything changed.

There was concern that if the nuclear material used in the reactor fell into terrorist hands it could be used to produce “dirty bombs,” or other kinds of nuclear weapons. To eliminate even the remotest possibility of that occurring, the pulse reactor was deactivated. That left those within DOE who are responsible for maintaining the viability of the nation’s nuclear weapons stockpile without an important testing tool.

Compounding the situation is a freeze on the building of new nuclear weapons that is also part of the Nuclear Non-Proliferation treaty. As a result, the U.S. must rely on its existing nuclear weapons stockpile, which is not immune to the effects of aging, and must, therefore, be periodically refurbished. And, “If you refurbish the old systems you have to requalify them,” said Robert Hoekstra, who heads the 13-person Electrical and Microsystems Modeling group at Sandia that has taken on the task of ensuring that the new electrical components used in the refurbishing process can survive extreme environments.

Their goal is to use DOE’s high-powered computing capabilities to model the effects of hostile environments on the electrical devices used in nuclear weapons as a substitute for testing in the SPR. In anticipation of the possible eventual need for such computational modeling, Hoekstra’s department began creating the necessary computer software some two years prior to 9/11. As part of this effort, Hoekstra’s team has been working in conjunction with other groups within Sandia that are providing data necessary to do the modeling, including codes representing material properties and radiation sources. Due to the complexity of the task, and the crucial need for proven accuracy, the development efforts are ongoing, with recent test results that have been promising.

“We had the first prototype demonstration earlier this year and that was very successful,” said Hoekstra. “It was a big milestone toward proving that we can use this modeling for new systems qualification.” But the actual use of this software to qualify new nuclear weapon components is still several years away.

The project designed to accomplish this monumental task is called Qualification Alternative to the Sandia Pulsed Reactor — the acronym being QASPR (pronounced Casper). QASPR consists of several computer codes, with two of the primary codes being developed by Hoekstra’s group. They are known as Xyce (pronounced Zeiss), and Charon, the more powerful of the two.

The River Styx

Because of its ability to solve a broad range of transport problems, including semiconductor physics, Hoekstra’s team named the program Charon, after the Greek mythological boatman who transported the dead across the river Styx.

The name Xyce is a variant of SPICE (Simulation Program with Integrated Circuit Emphasis), a widely used circuit simulation program first developed at the University of California-Berkeley, around 1970. Today there are numerous variants of SPICE that are used commercially in the electronics industry, but, says Hoekstra, none is as powerful as Sandia’s Xyce.

The prototype of Xyce was written some nine years ago by Eric Keiter at Sandia shortly before Hoekstra joined the department. He and Scott Hutchison then worked with Keiter to create the first production version. The team is now working with Xyce version 4.1.

A high-fidelity code in its own right, Xyce can take a computerized snapshot of the electronic forest by modeling the logic and timing of multiple circuits, while Charon can bore down into the individual trees by examining the workings of a single device to the point that it can model the movement of electrons inside a semiconductor material.