Powering particle accelerators with lasers has the potential to turn what is at present a hugely expensive but vital scientific tool into something far more accessible. Dr Thomas Kluge has been leading a project investigating the crucial phase just before the emission of ultra-intensity laser pulses that are used to accelerate ions.
When we think of particle accelerators today, our minds tend to conjure up vast structures spanning multiple countries, such as the Large Hadron Collider at CERN. But what if we were able to take the power of these devices and shrink them down to the size of a room?
Thomas Kluge and his colleagues Michael Bussmann and Marco Garten have been running a PRACE-supported project that has been investigating the acceleration of ions using high-power lasers. Laser-powered particle accelerators have been the subject of much experimental work over the past 15-20 years. Ultra-intense lasers are needed to accelerate particles to the velocities required for practical applications, but so far, the ion beams that can be produced do not have the required energy for such tasks.
“Our project was carried out at the Piz Daint machine hosted by CSCS in Switzerland,” says Kluge. “Our simulations are linked very closelywith experimental work that is carried out by our collaborators, and through our work we aim to support their experimental findings and advance knowledge in this field of research.”
The allocation we received from PRACE was so important to our work and to the field of laser-powered ion accelerators in general. Our data will help experimentalists to understand what is actually happening to the target and allow them to optimise their lasers with a rational approach.
Thomas Kluge, Project Team Member
Using an in-house code developed specifically for simulating laserplasma interactions on massively parallel computing resources, the team has been investigating the interactions of ultra-intense lasers with extremely thin copper foils, as well as plastic and liquid crystal targets.
As Kluge explains, with targets as thin and fragile as this, it is extremely important to fully model the evolution of the laser pulse. “In real-world experiments, a so-called plasma mirror is used to suppress the small amount of light intensity that arrives before the main laser pulse,” he says. “However, this device cannot suppress it completely, so there is always some leakage of light before the main pulse. Although this might not seem important, even small amounts of energy that leak through can heat the target, causing it to expand and distort before the main pulse arises. Therefore, we have to model this pre-pulse phase in our simulations in order to get the full picture.”
The simulations carried out in this project were the first ever to include this all-important pre-pulse phase that occurs just a few hundred femtoseconds before the main event, as well as being the first to show the targets in 3D and at full density. The results have shown the impact that the leakage has on the targets and how this in turn impacts the acceleration of ions. “What we found is that there is an optimum amount of leakage for accelerating ions,” explains Kluge. “If there was no leakage, the ions would gain less energy. Our simulations can help experimentalists to optimise this process and thus optimise the ion beams that they produce.”
Previous simulations of these lasers using lesser computational resources had to make certain initial assumptions about the pre-pulse heating and expansion of the target. “Our work has shown that these assumptions are actually not sufficient and do not fully reflect the dynamics of what is happening,” says Garten. “By simulating the few 100 femtoseconds before the main laser pulse arrives, we can see that there are some crucial things happening at this stage that affect the outcome of the simulation and the ion beam that is produced.”
The figure shows the free electron density (i.e. those electrons that have been ionised from the atoms) in a Cu foil after the laser irradiation.
These simulations have shown up details that previously had not been accessible to researchers working in this field. These details will be crucial for further optimising the ion acceleration process, as Bussmann explains: “If you don’t include these details in the simulation you get very different results,” he says. “This is why the allocation we received from PRACE was so important to our work and to the field of laser-powered ion accelerators in general. Our data will help experimentalists to understand what is actually happening to the target and allow them to optimise their lasers with a rational approach.”
What the team now want to continue to explore is how variations in certain parameters change the results for specific lasers and setups. This will help improve understanding of the role that these parameters play and the connections between them. However, this work will require even larger amounts of computational resources. “We need to carry out lots of these simulations and at greater detail so that we can start to add error bars to our results,” says Bussmann. “Only then will we really know how well we are actually predicting the outcomes of these experiments. This will probably require exascale resources, which we hope to be able to access in the near future with PRACE.”
As with much cutting-edge computational work being done today, one of the main challenges the researchers face is dealing with the huge amounts of incredibly high-quality data they have produced. “We need to find ways to bring all of this data to the community so that everyone can benefit from it,” says Bussmann. “This is a real challenge for the world of high-performance computing at the moment, and it will only become harder with the dawn of exascale computing.”
Collaboration with researchers at CSCS throughout the project has helped the team develop and establish workflows that will benefit others working in this field in the future. “These workflows, which can only really be established with large allocations like the one we received, will trickle down and benefit smaller groups,” says Kluge. “We want to make everything that we have produced in this project available to others so that they can continue to improve our understanding of this fast-developing field of research.”
Project title: Radiation imprint of ultra-intense laser heating of solids
The resources awarded were:
- 109 000 000 core hours on Piz Daint hosted by CSCS, Switzerland
Research field: Fundamental Constituents of Matter
- For more information see www.hzdr.de/crp.