First results from experimental setup

For the thesis I'm focusing my attention on the heat recovery performance when operating at laminar and turbulent flow conditions. The setup basically consists of the coaxial heat exchanger "ventilating" a "room" that is represented by two plenum boxes. The two boxes are connected with a duct, through which I'm moving air with a couple of small PID controlled desktop computer fans. The fan speed is varied so that a pressure differential between the two boxes, equal to what would naturally occur due to stack effect in a full size setup is maintain at all times.

Below are a couple of pictures of the actual setup:

I have already gathered a bunch of data, and I have been really surprised by how reliably the cheap little Arduino micro controllers have been performing. I must however admit that it was a real a challenge to design a system capable of accurately emulating the small pressure differentials of a natural ventilation system, and I don't think it would have been possible to pull off on the budget I had if it hadn't been for the great little pressure sensors from Sensirion that I used (I'm not being sponsored by them or anything, their sensors are just pretty great!).
Anyway, here's a histogram showing just how good of a job the pressure control system did at maintaining the pressure differential set point between the plenum boxes:

Very impressive indeed!

While I'm not done evaluating all the data for flow rates and heat recovery efficiencies, here are some preliminary results fresh out of MatLab showing a new framework I'm making to predict pressure losses in the transitional flow domain: 


So far it's looking very promising. I'll try to keep you posted, and if nothing else I'll put up the whole thesis report once it's done in a month or so.