heat exchange
Posted: Wed Nov 22, 2023 3:10 pm
A desirable property of a cantilever fan is that when an edge of the fan is placed close to a heated surface, the boundary layer of heated air is physically scraped off the surface and thus improves the heat transfer rate. I envisaged comb shaped fans interleaved through the gaps between the fins of a heat exchanger. The fan blades would have the dual purpose of being a fan and scraping the boundary layer off the heat exchanger fins.
I first experimented with two fans of dimensions 140mm long, 3mm tall and 0.5mm thick, running at 20hz and sandwiched between two heated flat plates. I found the heat transfer rate in the swept area of fan was 49.5 watts/m^2/k. This is the heat loss in respect to the surface area of the heated plate. These results were encouraging.
I then changed focus to seeing if two comb shaped fans could be interleaved between the fins of a heat exchanger. I 3D printed a dummy heat exchanger out of transparent resin in order to see how the comb “teeth” moved when surrounded by the heat exchanger sheets. I have attached a photograph of a fan comb and it interlaced with the dummy heat exchanger.
The minimum configuration that would move air through the heat exchanger is two sets of fan combs. This is consistent with the description in the published paper, where two fans are required to accelerate a cuboid of air forward (and right/left). What I found was that some of the comb “teeth” had a tendency to get out of phase and clash/hit each other. I speculated that this was a consequence of only being 3mm tall. Very small variances in the thickness or length of a fan tooth will very slightly chance it’s resonance frequency. If a fan is slightly out of resonance, it can significantly alter the phase of it’s sinusoidal movement. As a result, it can clash with an adjacent fan tooth.
It might be possible to eliminate this clashing. Perhaps by running all the fans slightly out of resonance, or tightening up the manufacturing tolerances.
I first experimented with two fans of dimensions 140mm long, 3mm tall and 0.5mm thick, running at 20hz and sandwiched between two heated flat plates. I found the heat transfer rate in the swept area of fan was 49.5 watts/m^2/k. This is the heat loss in respect to the surface area of the heated plate. These results were encouraging.
I then changed focus to seeing if two comb shaped fans could be interleaved between the fins of a heat exchanger. I 3D printed a dummy heat exchanger out of transparent resin in order to see how the comb “teeth” moved when surrounded by the heat exchanger sheets. I have attached a photograph of a fan comb and it interlaced with the dummy heat exchanger.
The minimum configuration that would move air through the heat exchanger is two sets of fan combs. This is consistent with the description in the published paper, where two fans are required to accelerate a cuboid of air forward (and right/left). What I found was that some of the comb “teeth” had a tendency to get out of phase and clash/hit each other. I speculated that this was a consequence of only being 3mm tall. Very small variances in the thickness or length of a fan tooth will very slightly chance it’s resonance frequency. If a fan is slightly out of resonance, it can significantly alter the phase of it’s sinusoidal movement. As a result, it can clash with an adjacent fan tooth.
It might be possible to eliminate this clashing. Perhaps by running all the fans slightly out of resonance, or tightening up the manufacturing tolerances.