Originally Posted by mitchellcht
The only time I see a rad sandwich not being a" optimal idea" is using two radiators with a high fin per inch count. Lets say you have two 35mm rads sandwiched, wouldn't that equal 70 mm, in which you could use 3 fans sandwiched, single rad you could use 2 fans, so thus giving you less airflow. By taking away the rad and putting the pump on 3 got him better temps, so how did that not show flow problems?Plus to me in that picture it look like he split the flow between the 2 rads, so reducing the flow which made it seem as if the 240 rad was being restrictive, hence high temps and high pump speed, now no rad, and the pump on 3 meaning he has more flow than before giving him better temps.
I think I get where you're coming from with the 35/70mm thought experiment, but any sandwich is still going to perform worse than using the same equipment not sandwiched. I'm not sure if you clicked through to the link in my first post in this thread, that offers firm experimental evidence which supports what I'm saying about rad sandwiches sucking, but let's assume you didn't. Apply your thought experiment in reverse--take a 70mm rad and slice it in half lengthwise into two separate 35mm layers. You're going to get better total performance if each half has its own fresh supply of ambient-temp air flowing across it, than if one half used the exhaust heat coming off the other half as its intake. That would send the air/water delta for that half into the toilet, and that's what making two rads into a sandwich does. Even if we grant that two 35mm rads sandwiched with three layers of fans would perform better than one 70mm rad with two layers of fans (which I am skeptical about, and I'll only explain why if you force me to, since this is already getting quite tedious and long-winded
), the bottom line is that there's no real-world scenario where those are your only two choices. Two 35mm rads not in a sandwich will perform better than either of those two options.
In the case under discussion, the OP's improvement also came partially from putting eight fans on the bottom 480 instead of six. The 240 was doing almost no cooling work under load. In fact IMO it's quite plausible that it was actually adding heat back into the loop when the system was cranking, since the air being supplied into the 240's fins (vented out of the 480) was likely at a higher temp than the water.
Now that I've read this thread over again, I think you might in fact have been right about there being bad flow, but not because of the 240 rad.
Originally Posted by eliteone
... I also think I may have had an airlock issue on the top rad, as the fittings are pointing down, when I was playing with pump speed last night I heard some gurgling coming from the top rad.
At first I thought he had a new air bubble in his top 480 rad after reconfiguring, but now I think he was talking about the bubble being there before he reconfigured, and now he's cleared it out. So yeah, no sandwich, two more fans on the bottom rad, and no bubble in the top rad together easily explain the big improvement he's getting.
Regarding the idea that he had split the flow between the two bottom rads--his most recent photo
shows that the thing at the bottom is a rubber cap that doesn't lead anywhere, so it seems clear to me there wasn't any split loop going on. My best guess is that it's a QD which he uses for draining.