Now for something completely different…

So the reason I’ve been messing around with rocket motor design lately is that an old idea I started dabbling in several years ago, that of a truly miniature ramjet, something that could fit in the palm of your hand. I have read a report on a tiny ramjet designed to operate at Mach 4, however I’m thinking of aiming for something around Mach 1.5.

There are many videos on YouTube of small subsonic ramjets being fed by a leaf blower that guzzle propane, making a lot of heat and no doubt some small amount of thrust all while looking spectacular, but nothing like a usable engine. The engine I had in mind would operate at supersonic speeds to operate as a sustainer stage on a small rocket. But how would I test the engine, especially the aerodynamics that are involved? There’s no leaf blower around that can deliver air at those speeds, while trying to problem solve and tune the engine while flying it on a rocket would be long, tedious and expensive so what I need is a wind tunnel! Fortunately in my other guise as a post-graduate researcher I have access to a small supersonic wind tunnel in my department, but the test section is small and being a recirculating tunnel trying to investigate fuel flow and combustion would be problematic, plus I think the lab manager might have a few things to say about that…

The other advantage of my research is that I specialise in the use of Computational Fluid Dynamics, otherwise known as CFD, well rather I specialise in the large Reynolds Number flows that occur around Naval ships, but those skills transfer well to my problem. An issue that I repeatedly find myself telling undergrad students when they are first introduced to CFD, is that it you will always get an answer! The data might be complete rubbish, but it can and still look good enough to present as ‘the correct result’ despite being wrong. Sometimes as somebody with a fair degree of experience in CFD I can see straight away what is wrong, even when that work is being presented at a conference by somebody getting many times my salary.

Other times I find myself trying to unpick what a particular CFD code is doing and has once resulted in contacting with the creators of an extremely popular CFD package to tell them that their program has a bug. The moral of the story is that CFD is a helpful tool, it can substantially reduce the costs of testing but it should be used with caution and it does need to be validated.

So what else can be used, well fortunately there’s a fairly cheap and easy method of using a shallow water channel to investigate supersonic flows, and by shallow I mean a few mm’s deep, so something that can be made fairly easily and cheaply. Being so shallow the flow is analogous to two-dimensional gas flow, with any vertical change in water height assumed to be negligible. It would also allow me to look at effects like intake buzz, something that would be prohibitive as a quick CFD project. The maths underlying it is fairly simple and I have some old aquarium pumps that can be rigged up to provide a source for the water flow.

So a plan has begun to form… Thesis or this, this or thesis… I think I can do both.

Sarah