Some progress!
The maths doesn't work. Okay, it doesn't work well enough.
One of the biggest challenges with fluid mechanics is that it can be extremely sensitive to very small changes, particularly at the kind of scale we're working at when developing the core technology of dry powder inhalers. That perplexing "transitional" region of flow - not fully turbulent - yet definitely not laminar. You can't simply say that the square root of the differential pressure is proportional to the flowrate - and neither is the differential pressure - it's annoyingly somewhere in between.
Even tiny changes in draft (expansion) angles can have a pronounced effect on the way the air flows. Surface finish affects the friction at the wall and consequently the losses through the system. The real issue is that a change in the volumetric flowrate directly changes the velocity of the air; and the kinetic energy in that airflow is proportional to the velocity squared - so even relatively small changes in the flowrate can produce large and significant changes in the energy that's available to do work on the formulation. And these differences in energy lead to undesirable variation in device performance.
So it was back to the drawing board - the mathematical predictions of airflow resistance were not good enough to be useful - and we needed to be in a position of confidence before going forwards with the next stage of the αeolus technology development...
We are now the proud owners of 43 precision chokes - thanks to Complete Fabrication, and the slightly less proud borrowers of a DFM 2000 mass flowmeter - thanks to Intertek. We will be running 83 experimental configurations over the next week or so to fully map out this poorly understood design space. The regression coefficients we obtain will be used to drive a parametric CAD model of our new αeolus technology. At last we should be able to develop the principle based on sound empirical data, and have good confidence in the design. We will know that we are forcing as much air energy as is physically possible through the small dimensional constraints of the blister. Sometimes it's just best to measure things... :-)
Mechanical Engineer (Analysis, Test, Sustainability of Medical Devices)
4 年Great to hear about your work progress. Sounds like you are chipping away at the problem. Experimental data is v good for debunking a ‘good’ theory, but also for building the basis for the theory that follows it. Good luck with the next phase!
Drug Delivery Devices and Intelligent Healthcare Specialist - helping clients achieve patient and commercial benefits
4 年Spot on David, we did the same with Orbital, the CFD just couldn't be relied upon and we had to do loads of empirical testing.
Founder & Engineering Director at 3P innovation Ltd.
4 年Nice work.
Managing Director | Driving Innovation and Excellence in Rapid Manufacturing at FRP
4 年Let me know if you need any of the ones that work best CT scanning so you can confirm the *actual* nternal geometry ??