Volume 112
您当前的位置:首页 > 期刊文章 > 过刊浏览 > Volumes 108-119 (2025) > Volume 112
Scale-up of CFD-DEM simulation of chemical looping gasification by GPU acceleration
Christoph Graf *, Yannik Lichtmannegger, Jochen Ströhle, Bernd Epple
Technical University Darmstadt, Department of Mechanical Engineering, Institute for Energy Systems & Technology, Otto-Berndt-Str. 2, 64287, Darmstadt, Germany
10.1016/j.partic.2026.02.020
Volume 112, May 2026, Pages 62-73
Received 28 November 2025, Revised 4 February 2026, Accepted 16 February 2026, Available online 12 March 2026, Version of Record 18 March 2026.
E-mail: christoph.graf@est.tu-darmstadt.de

Highlights

• GPU-CPU coupling accelerates fluidized bed CFD-DEM simulations by a factor of 80.

• Gas phase simulation time is dominant up to large number of particles.

• Reaction calculation accounts for the majority of simulation time.

• The approach allows industrial scale simulation on standard workstation computers.


Abstract

The chemical looping gasification process is an efficient technology for chemical recycling and bioenergy. While the process has been tested in lab and pilot scale, a scale-up to demonstration or industrial scale has not yet been attempted. A potential tool to assist the scale-up is coupled computational fluid dynamics (CFD) and discrete element method (DEM) simulation. However, with conventional simulation techniques the required simulation time is unfeasibly long. In this work the simulation was accelerated by a hybrid approach using both, graphics processing unit (GPU) and central processing unit (CPU) achieving a speed-up of 80 times. An analysis of the main influences on the simulation time was conducted to improve the simulation efficiency. With these results an upscaled simulation of a 200 MWth plant was performed achieving plausible results in a reasonable time. Thus, this work provides a proof of concept for CFD-DEM simulation of fluidized bed gasification at industrial scale.

Graphical abstract
Keywords
Discrete element method; Chemical looping gasification; GPU; Fluidized bed; Scale-up