glenntws wrote:Ok, so what did I do? I started of with two pipes. One clear pipe with a inner diameter of 40mm and one with the Special shaped internal part (inner diameter also 40mm). Both pipes are 520mm long, have a wall thickness of 2,5mm and both are enclosed in a air volume.
I then simulated both pipes with the same variables. The outer air volume had a in-temp of 100ยฐC and a velocity of 50m/s. The inner (exhaust) volume had a in-temp of 1000ยฐ , a pressure of 5bar absolute and a velocity of 100m/s.
While I won't show you screenshots of the normal exhaust simulation (useless, it's just a simple flow in one direction), I can show you some details about the vortex-style pipe.
http://fs5.directupload.net/images/170222/78hlcdrb.png
Here you can see how the heat get's trapped in the center. The vortices indeed are able to generate some kind of heat shield.
It's fascinating that these turbulent motions also make up for a barrier between the "pins" that look into the center and the exhaust gas.
http://fs5.directupload.net/images/170222/dtbf9wio.png
Here I also enabled some vector-arrows to show, how the flow looks like in these little chambers. You can see good, how there are building up little circles of flow, generating the "heat shield". Interesting is, that the vortices have a different shape for every chamber, maybe the flow that went out of one chamber directly get's into the next one and creates some different kind of motion. I don't know if I did the Simulation too short (only did 400 interations, the divergence still was pretty high) but I'm not able to do these big simulations with my current PC in a acceptable time window
The most important thing however is: Did we save heat energy in this process??? And, according to Autodesk CFD, it indeed does. The heat energy transfer on the normal pipe was 28767,5 watts. The energy Transfer on the vortex-shaped pipe was at only 18632,8 watts!!! That's a plus of 54,4%! And i think we all now what could be done with a extra of 10 Kilowatts > more MGU-H recovery potential.
The only negative effect is the flow resistance. In the vortex style pipe, it's 2,24 times higher... But nevertheless, I am more than sure that this is the right way to go.
The only thing that's important is the shape of these vortex inducing elements. And I'm pretty sure the ones I used weren't nearly as optimal as possible (only did one version, the Simulation is sooooo time consuming....).
Anyway, I think we really found something that MB and the others could use