Solar air heater aerodynamics

[Vekke]

I will put this also here as its aerodynamics I am trying to optimise here and not all read the home section.

Thanks to www.eqmsol.com Gaurav Kamde and Anand Kulkarni making this simulation possible.

This is a opensource project so final version design details and drawings will be released to everyone so you can make it easily yourself and help to save our climate and save yourself little bit money on heating costs.

https://youtu.be/9LD-mVdRsKA



All comments and ideas are welcome what you think will happen and what you think is the correct way to solve the puzzle.

[freebeard]

I suspect that the depiction is MSPaint rather than CFD?

I thought I suggested before, move the inlet and outlet to opposing ends of the two outer fences. See if that doesn't better equalize the flow.

Also the corners could have a quarter-round or straight 45 degree baffle to smoothen the air flow.

[Vekke]

Originally Posted by freebeard

I suspect that the depiction is MSPaint rather than CFD?

I thought I suggested before, move the inlet and outlet to opposing ends of the two outer fences. See if that doesn't better equalize the flow.

Also the corners could have a quarter-round or straight 45 degree baffle to smoothen the air flow.

Yes the linked picture is MSpaint my guess what will happen in the 2nd CFD simulation, but on the linked youtube video there is pictures and simulation of that flow for the first version.

In first sketch versions I had baffle rooms at the top and bottom. Meaning those holes that are now in the baffle plate were on the top and bottom walls. Like in the first one was the 4 breathing holes at bottom. But it causes little problems if that is what you meant. Sure it would create nice and more even flow but how to control the flow how much goes on top vs down is not so easy. Also easiest would then CNC cut two of those baffle plates vs now only one. Also will the air be turbulent all the way that 2m distance is the entry is rigth at the beginning.

Also the area of the collector would get little bit bigger. That insulation foam does not costs much, but it adds also little bit more complexity to the construction.

If you mean corners 45 degree walls that cuts the collector surface area that is no go as you want to get maximum effective surface area for that PC sheet size as its one of the most expensive parts of the system. Of course it can be build using even cheaper clear materials to get the costs down even more.

If I just move the inlet to lower left corner and exit rigth upper corner I would bet it would not solve all those problems seen as the air will go easiest route. and there is more surface area to choose the easiest route. Meaning there would still be those hot spots where there is minimal air flow. If you look different versions of intercoolers or radiators both styles are used and also one that comes back to same side where the entry is. In this case that migth also work really well. if you put the air first go through the visible side and after that the back side which is hotter. The you would get double the distance but also about double the air speed with those dimensions. Visible side should stay coldest to avoid heat escaping through the PC sheet.

I am still pondering is it better to have slower air flow speed on the visible top side or on the down side. less volume on the visible side or little bit more and have thicker boundary layer near the polycarbonate. At the moment I am shooting for that more volume slower flow speed and thicker boundary layer.

However when the working basic construction is found, its easy to test what flow speeds and entry points will offer best performance. Also which side should have more air flow. I think I have PC sheets for 6 prototypes.

That cfd simulation made things easier as you get to see basic flaws before building any prototypes.

[freebeard]

The double-pass sounds interesting. If the plate is conductive, the corners would still contribute and the airflow will be compromised anyway. IMHO, of course. The ends of the fences might angle toward the fan to manage what goes where.

Testing lots of iterations should [get/keep] you on track.

[Vekke]

Second CFD simulation is ready. System went to rigth direction. Biggest issues came from fan position and it was not rotating in this simulation. Also the triangles together with the divider plate caused too much restriction so on the visible side was too little flow.

https://youtu.be/1cgwMMo786Y

[freebeard]

I think the vena contracta problem could be resolved by cambering the ends of the dividers to the angle of the air flowing off the fan.

Unless that's a fabrication problem.

[samwichse]

Silly question, but why not just use a standard solar water heating panel and a liquid radiator?

Way more flexibility in mounting and interior installation, buffering built into the system, and water is a more efficient heat exchange material than air. I guess I don't see much upside to this direct air design?

[Vekke]

This is much cheaper and easier to make by anyone than air water systems. 2m2 part costs can start from 30€, if you use cheapest front glass solutions. So paypack time is vey fast under one year.

That one 2m2 can produce about 1000-1500kwh of heat energy per year.

[freebeard]

Air leaks don't cause structural damage.

[Vekke]

Plus you can at the same time get your fresh air from outside need filled without losing heat as the air is already heated when it enters the house. So in final version before that fan there will be small fresh air hole where it sucks partly fresh air mixed the already hot air that comes from house for recirculation.