Simulations of Experiments – SIMEX

As advanced light source technologies have developed, scientists have written software to simulate different components of these facilities. However, none had tackled the entirety of a complete experiment, from the generation of light to the results—at least until EUCALL scientists developed SIMEX. SIMEX is a unique simulation framework that uses some of most advanced simulation tools and integrates them to mimic an entire light source beamline. It is a flexible, modular system that can be tailored for use at potentially any advanced light source.

Chains of calculators and interfaces

SIMEX is akin to a train: each car has to fit into the next just right, and what happens with the first cars affects the ones that come thereafter. The framework’s developers worked from a prototype developed between several institutes under the leadership of European XFEL. In the early version of the software, scientists could inputthe different features of their X-ray FEL experiment and get a model diffraction pattern. SIMEX expanded upon
this framework by integrating simulations of different constituent parts of advanced light source facilities. This means SIMEX could now potentially model an experiment at any advanced light source. SIMEX works by placing different constituent simulations, or “calculators”, in a chronological order. The simulation for the focusing of the light onto the sample is followed by the simulation for interaction of the light with the sample, and so on. At the end, the simulation acts as a virtual version of the desired facility beamline. In between the calculators are interfaces, which take the output of one calculator and make it standardized so as to turn it into the input of the next. At benchmarking tests at the LCLS X-ray FEL in California—which hosts some of the beamlines first modeled in SIMEX—scientists found that SIMEX data outputs match well with actual experimental data on biological imaging experiments.

An essential tool

One promising use of SIMEX is the ability to test the feasibility of experiments before they are performed. This feature will allow scientists to demonstrate that an experiment can be carried out with reasonable results ahead of the experiment allocation. Also, scientists can ahead of time determine the best conditions to optimize data collection, saving experiment teams precious time at the facility that would otherwise be used for tinkering with settings. As SIMEX grows in its flexibility, it will be able to simulate more instruments at more facilities. Tools such as SIMEX allow scientists to use advanced light sources more efficiently. They also support a widening of scientific applications through the ability to test the feasibility and to optimize the control of experiments at the facilities.

A further use of SIMEX is that it can be employed by new users of advanced light sources to learn about how to design and run a photon science experiment. SIMEX is foreseen to be integrated into an e-learning platform to train the next generation of users about how to plan and perform experiments at advanced light sources.

Image: An example SIMEX simulation of a structural biology experiment on proteins surrounded by layers of water of varying thicknesses. The goal of this simulation was to mimic how the water could affect the results of the experiment. TOP AND CENTRE: 3D visualizations of the protein and the surrounding water layer of thickness (top image) 3 Å and (centre image) 20 Å. The results from SIMEX, showing variations in the diffraction patterns, are insets in the molecular model. The patterns show that the sample particle has gotten larger with the additional water. BOTTOM: The simulated data, when translated to a line graph projection, reveals that the water obscures the physical details of the protein by making the sample appear smoother than it really is.

SIMEX Deliverables

 

Deliverable 4.1 - Design report and advanced photon-matter simulation software – short pulses (Deliver a document describing the requirements for simulations of femtosecond laser-target interaction and transient plasma properties probed using ultrashort pulse x-ray sources, e.g. X-ray FELs) / Submitted 30 September 2016

 

Deliverable 4.2 - Design report and advanced photon-matter simulation software – long pulses (Deliver a document describing the requirements for simulations of nanosecond laser-target interaction and target design for dynamic compression experiments using ultrashort pulse x-ray sources, e.g. synchrotrons or X-ray FELs) / Submitted 30 September 2016

 

Deliverable 4.3 - Interoperability of simulation workflows (Report describing the interoperability of the simulation platform, including packages by various Partners) / Submitted 29 September 2017

 

Deliverable 4.4 - Generation of simulated coherent scattering data from plasma and non-plasma samples (Report about generation of and results for simulated coherent scattering data from plasma and non-plasma samples) / Submitted 29 September 2017

 

Deliverable 4.5 - Testing, validation, and example workflow (Report about the implementation and testing of high performance computing measures for the Simulation)