Research Projects

COMFORT

COmputational Morphogenesis FOR Time-dependent problems

Topology optimisation of time-dependent problems — such as transient heat diffusion and time-dependent fluid flow — requires solving the governing equations forward and backward in time at every optimisation iteration. A single simulation can take 12+ hours, making design optimisation impractically slow (up to 200 days).

COMFORT develops specialised space-time methods that treat time as an additional spatial dimension and solve the full space-time problem in parallel, targeting at least a 10-fold reduction in time-to-solution. The approach has already demonstrated a 52× speed-up over sequential time-stepping, solving problems with 4.2 billion degrees of freedom in just 17 minutes on 64,000 CPU cores.

Key applications include the optimisation of fluidic oscillators (relevant for wind energy, cooling, and lab-on-a-chip mixing) and heat spreaders with phase-change materials.

PI: Joe Alexandersen  |  Ph.D. student: Magnus Højmose Appel (Sep. 2024 – Aug. 2027)  |  Postdoc: Sarah Nataj (Apr. 2025 – Mar. 2027)

HiHeat

Topology optimisation for High HEAT flux cooling

Components in fusion energy reactors (such as the W7-X stellarator's divertor) and in high-performance electronics must absorb and remove extreme heat fluxes under turbulent flow conditions. Standard topology optimisation methods assume laminar flow and low heat loads — neither holds here.

HiHeat extends topology optimisation to conjugate heat transfer problems with turbulent flow, using RANS (Reynolds-Averaged Navier–Stokes) models with wall functions. The density-based approach allows the wall definition to be driven implicitly by the design field, improving mesh efficiency. Current results already show a 29% improvement in thermal extraction with an 11% smaller pressure drop compared to existing designs for the W7-X divertor.

PI: Joe Alexandersen  |  Ph.D. student: Amirhossein Bayat (Feb. 2024 – Jan. 2027)

Vestas Aircoil: High-temperature heat exchangers

Advanced design for high temperature-difference heat exchangers using computational morphogenesis

Heat exchangers with very large temperature differences — used in exhaust gas recirculation, Power-to-X systems, and thermal energy storage — develop significant thermomechanical stresses and deformations. Conventional design methods cannot reliably deliver solutions that handle these loads while maintaining performance.

This Industrial Ph.D. project, in collaboration with Vestas Aircoil, applies topology optimisation and computational morphogenesis to simultaneously account for thermomechanical deformations, vibration, and pressure loads — producing designs that are both thermally efficient and structurally robust.

Industrial Ph.D. student: Morten Bjerre Jonathansen (Mar. 2025 – Feb. 2028)  |  Academic supervisor: Joe Alexandersen  |  Industry partner: Vestas Aircoil, Denmark

Fusion energy: electromagnetic–structure interaction

Contributing to the NNF Challenge project on continuously operating fusion power plants

First-wall components of fusion reactors — specifically, the gaps between tungsten tiles in stellarators — interact with the high-frequency millimetre waves used for plasma heating. Understanding and minimising these interactions requires advanced electromagnetic–structure simulations.

The MSOE group contributes a one-year postdoc position to the NNF Challenge project "Enabling Continuously Operating Nuclear Fusion Power Plants" led by Prof. Stefan Kragh Nielsen (DTU). The work focuses on identifying and implementing the best numerical methods for simulating wave–tile interactions, with experiments planned for 2027.

Lead PI: Stefan Kragh Nielsen (DTU)  |  MSOE contribution: 1-year postdoc (starting 2026)

CT-IPC: Conformal cooling channels for injection moulding

Design of conformal cooling channels using topology optimisation

In injection moulding, cooling channels must be positioned close to the mould surface and follow its geometry ("conformal" channels) to cool parts uniformly and reduce cycle time. This project developed topology optimisation formulations that control both cooling efficiency and temperature uniformity across the mould surface.

Postdoc: Yupeng Sun (2024–2025)  |  Supervisor: Joe Alexandersen  |  Industry partner: CT-IPC (Centre Technique Industriel de la Plasturgie et des Composites), France

ADeHEx

Homogenisation-based topology optimisation of heat exchangers

Microchannel heat exchangers offer exceptional heat transfer performance but are challenging to design because the channel features are far smaller than the overall device. This project developed homogenisation-based topology optimisation methods that treat the micro-scale channels as an effective medium, enabling the simultaneous optimisation of both the channel microstructure and the overall flow layout at practical computational cost. The project included a secondment at Brown University (CRUNCH group, Prof. George Karniadakis).

Postdoc: Hao Li (Oct. 2023 – Sep. 2025)  |  Supervisor: Joe Alexandersen  |  Secondment: Brown University, USA