After three lectures on Density Functional Theory (DFT), a group of students from the TSM CDT undertook the task of coding it up within 8 weeks.

With a general knowledge base of around three lectures on Density Functional Theory (DFT), a group of students from the TSM CDT undertook the task of coding it up within eight weeks as part of their Group Programming option. The criterion for the projects success was to get the equilibrium lattice parameter of bulk Silicon to agree with CASTEP (a commercial DFT-based electronic structure code) to within 0.01 Angstrom (~0.2%). A fairly formidable target.

Luckily for the group, Imperial's Thomas Young Centre has a number of academics with expertise in electronic structure code. Guidance from Peter Haynes, who has worked on developing both the CASTEP and ONETEP DFT codes, and James Spencer, a specialist in computational science, proved a valuable resource. Following their advice the group based much of the project on the work done by Mike Payne and his group in the development of CASTEP, and Peter Haynes is now able to associate another TEP to his name in the form of MISSTEP, the students'  Mostly-Inefficient Surrogate-Software Total Energy Package.

After a couple of premature celebrations and subsequent debugging sessions in the last weeks of term the code was finally completed for submission before the deadline at midnight this Sunday. Over the Easter break it'll be documented in a report, and the group will present the work in front of Mike Payne at 2pm on Monday 9th May. Results for the total electron density and energy-volume curves (from which the bulk modulus can be calculated) are shown in the figures, and as for the lattice parameter? Correct to 0.001 Angstrom.