For experiments performed at LLNL’s National Ignition Facility (NIF), the goal is not just to measure the interaction between the target and the laser beams. Researchers also need to understand what’s happening to the material in the target and how it changes over time. This capability to see inside the target as it interacts with NIF’s 192 laser beams is especially useful in high-energy-density experiments and those exploring material strength.

A decade ago, NIF experiments relied on a technique known as long-delay backlighter (LDBL), in which lengthened optical fibers allowed selected laser beams to arrive up to a microsecond after those that are directed at the target. The delayed laser beams irradiated an x-ray backlighter behind the target, using the radiation produced to see what transpired during the experiment; the data was then recorded by diagnostic equipment. The LDBL method provided extraordinarily high-speed imaging of material interactions. As an example of the strong collaboration between Computing and NIF staff, this seemingly simple change required specialized software to redefine timing values for the lengthened fibers as well as operational expertise to provide the requested delay without introducing error.

However, the first-generation solution was not perfect. Highly trained specialists still needed to install and verify the correct lengths of fiber before every shot, often doing so during off-hours. Dave Mathisen, the shot timing systems engineer and shot automation lead architect, recalls, “The LDBL system functioned properly, but it was incredibly time- and labor-intensive while only supporting a limited set of beams. The whole process, including the preparation of the fibers and manual inspection of hundreds of lines of data, took hours. We were motivated to improve upon the original process.”

As NIF’s requirements and technologies evolved over time, the LDBL technique provided a springboard for a new backlighter solution. Mathisen continues, “The backlighter use case presented a unique systems engineering challenge. Our goal in the next phase was to improve the efficiency of conducting these shots by providing a seamless process for proposing, reviewing, scheduling, and conducting them.” The fiber-delay backlighter (FDBL) project was born, with Mathisen at the helm.

The introduction of FDBL automation addressed all of the LDBL issues. According to Mathisen, the LDBL experience helped the team mature the operations related to providing backlighters as well as carefully consider the software and human relationships involved. “Easing the path for operations is all about smoothing the rough corners on these interactions. Improving throughput means maximizing strengths and minimizing weaknesses,” he explains. “I think of the whole system as an ergonomics problem, where we make an incredibly complex and sophisticated tool easy for a human to manage.”

The new FDBL system organizes and controls the full lifecycle of an FDBL shot from the initial planning stage to the experiment itself, requiring human input or action at only a few key junctures. For example, when a shot is requested by the Responsible Individual (RI), they input a packet of information that allows the system to automatically determine the corresponding fiber needs. With the RI’s approval, the system orders the fibers, places them in inventory, guides their installation, and automates the related timing changes during the shot. Once that shot is complete, the system guides the actions required to return NIF equipment and software to a non-delayed configuration. This system of improved procedures and toolsets places the expertise for executing FDBL shots inside the system, eliminating the need for timing experts to be on hand.

Since its implementation, the FDBL automation has reduced the execution time for these types of shots by four to six hours each—a significant increase in efficiency that quickly accumulates. In addition, the process allows optimization of shot scheduling. For instance, an RI can see other shots requested in a certain time frame and organize them to minimize the time spent changing fibers between shots, further reducing the cost of beam-delay experiments. Mathisen adds, “We’ve turned a very complex process into a far simpler one for all involved. It’s usable, versatile, reliable, and efficient.”

Moreover, according to Mathisen, “The success of the FDBL project has set the stage for other wide-scale, interdepartmental systems engineering projects aimed at increasing the performance, mitigating obsolescence, and optimizing shot efficiency, such as the High-Fidelity Pulse Shaping system [HiFiPs].” Over the past 20 years, he continues, “I’ve seen a theme of efficiency advancement from automating a shot to parallelizing tuning shots with target preparation, which helped us move from 130 shots per year to more than 400. Everything we do is aimed at maximizing NIF’s capabilities and shot rate.”

Mathisen notes that these types of improvements provide significantly richer capabilities without ballooning operations staffing. He says, “We’re not simply running an expensive machine as much as possible to get the most out of it. We’re giving our physicists, country, and partners a better view through a wider window.”

The multidisciplinary FDBL collaboration has brought together shot experimentalists, laser engineering, facility operations, and dozens of contributors from the NIF Computing Division. “Our greatest resource is our people, including their dedication and depth of institutional knowledge and the passion with which they apply it,” states Mathisen. “We thrive on advancement.”