Collaboration
The centrifugal chain works.
The next one is axial.
AeroSol today designs centrifugal compressors end to end — sizing, mean-line, blade, throughflow, blade-to-blade, map, volute, stress, CAD. The architecture underneath it is not specific to radial machines: one geometry, a chain of solvers that all read it, and agents running the loop. That same architecture is now being extended to multi-stage axial compressors and to axial fans.
The work is on the compression side and stays there — compressors, fans and blowers. It is early, and it is open: if you have a rig, a measured map, or a design case you are willing to run alongside your own tools, there is a straightforward trade to be made here.
In development
What is actually being built.
Stated plainly, because a roadmap that reads like a brochure is worth nothing to an engineer deciding whether to spend time on it.
Multi-stage axial compressor
Repeating-stage mean-line across a multi-stage build, stage stacking with a real spanwise work distribution, and the throughflow solver extended over several blade rows rather than one. The hard part is not any single row — it is matching them, and holding the match off design.
Axial fans and blowers
A different regime with the same machinery underneath: the duty is volume flow and pressure rise rather than pressure ratio, the blade wants its own camber and thickness rules, and the correlations that matter are the ones that predict stall margin and noise rather than choke.
What is not in it yet
No cooling, no disc or attachment stress, and the structural side stays beam theory for screening. Turbines are not on the list either — the work is deliberately on the compression side. When any of that changes it will be said out loud, the way the current limits are.
The honest status
- Centrifugal is the product. It is working, it is checked against machines other people measured, and it is what a pilot study puts in front of you today.
- The axial compressor and fan work is development work, not a shipping feature and not a date. Nothing on this page should be read as a delivery commitment.
- Validation is the bottleneck, not code. A solver is only worth what it has been checked against — which is exactly why this page exists.
The trade
What would be useful,
and what you get back.
This is a small operation. The offer is direct access rather than a partner programme with a logo wall.
Reach out
Tell us what you have,
and what you are trying to design.
A rig, a data set, a stubborn design case, or a course that needs a tool students can actually see inside. One message is enough to start.