I accept your explanation about the diesel radiator. I still hold that its completely possible to build a radiator that will last for the TJ. It wouldn't look like the one that's in it - or in my diesel for that matter, obviously.
Need to get Joe in here and pick his brain on that one. I suspect the limitations on all the things that make a radiator work will also severely complicate the final design and ultimately the efficiency of the cooling abilities.
If you bump up the tube wall thickness to increase the pressure handling ability, it will take longer for the heat tranfer to get to the fins. If you add more fins to help overcome that, you restrict the air flow through the core and knock any gains back down to below what would be gained by adding more fins.
Since I don't see many issues with the actual cores on the TJ radiator, I suspect it may be very possible to just make a better tank for each end. I don't know that anyone would pay for it but a nicely stamped aluminum tank with welded water necks would be where I would start. The o-ring seal would have to be improved and someone a lot smarter than me will have to figure that one out.
What we need is a furnace brazed copper core and then some very robust brass tanks. I suspect we could deal with soldering the tank to the core if we could get rid of the solder in the tube to fin interface.
Pretty much all of what's said here checks out.
1. Modify fin density as last resort. It has a pretty significant impact to airflow, and it can get bad especially fast when you have an engine mounted fan that needs an inch of blade tip clearance inside the shroud.
2. Heat conduction has a lot of parallels with electrical conduction. Thicker material is more resistive just like a longer wire has more electrical resistance than a shorter one of the same gauge.
3. my "data" is based on what I see online, but the common failure modes seem to be split tanks and failures at the tank-to-core crimp seal. Aftermarket replacement radiators usually skimp on tube count and fin density which hurts capacity, and they're probably skimping on the thickness and maybe material for tanks leading to a short lifespan. It also appears that Mopar seemed to be the only one that applies the e-coat, which is an important guard against corrosion. Question: do we see a lot of OEM or aftermarket replacements fail by springing a tube leak?
4. On the surface it seems like aluminum tanks welded to an otherwise identical core would knock a lot of them out. The aftermarket all-aluminum radiators that have such a short lifespan are skimping somewhere...if it's at the weld, they have bad welding practices/QC, if it's pinholes in the tube then it's material thickness and lack of e-coat. I haven't looked into it but could very well be skimping on tube count and fin density as well if they don't perform. They're used by the thousands, if not millions in HVACR with 650psig pressure ratings and 15-20 year life expectancies, but it's not exactly the same as an automotive application since though it does get vibrations being attached to rotating equipment (fans and compressors), it doesn't get the shock of a suspension bottoming out or twisting up the front end going over a boulder. (Note: I'm mildly curious about keeping the core exactly as is, but fabricating aluminum tanks to crimp on instead of factory. Would at least cut out the tank splits)
5. Making the exact same (dimensionally) core out of all copper with soldered/brazed tanks would address all of the issues, but
a. I don't have any experience with copper, flattened-tube cores to know what issues there might be with duplicating the current aluminum design
b. copper is about twice as strong, 50% more dense, and 3x the cost, so there would be a compromise between cost and how much thickness can you take out without ending up back where you started.
c. round tube (like a heater core) copper needs more volume to pack the same heat transfer, so you need a thicker core and an electric fan to handle the added air pressure drop.
