I can't find time to sit down today so just a quick response from me here, but those links are awesome. Good find and thanks for linking here.
It looks like they have E10 at around 155F boiling and pure gasoline tracking for ~185F. I say this realizing there's a range assumed there, but I'll talk single numbers for ease of conversation. Lower than I expected, but I think still higher than anyone has measured their tank. We've seen fuel return lines at 160F, I think. Tank should be cooler (I suspect a lot), but I guess we don't know by how much.
I agree if you're over the boil point/range in the tank then there's no stopping the problem. I assume though, that not everyone with this problem is into the boil range in the tank. Might be a bad assumption, but I just can't see the tank staying that hot in all these cases. I think instead, we're just seeing a very slow and gradual build up of pressure over what the purge system can vent.
I got my pressure sensors and gauge today. Might have to add a couple temp probes to my data log as well.
Thanks for the response and links. Good stuff.
Quick addition so you know I'm not ignoring it: I think fuel rail temps are relatively meaningless for the tank vent problem. The temps and pressures of fuel going back into tank are different enough that I don't think the rail conditions mean much. Very meaningful for vapor lock, obviously, but not so much the fuel geyser issue.
I don't know how liquid fuel temp in the rail is meaningless. It should be the single highest temp of the fuel delivery system.
I think if it as one event, from too much temp that effects vaporization and not enough volume or thermal transfer in the tank to end the process. The gradual build up of heat over time at the fuel rail and traveling back to the tank overcomes it's ability to cool it.
After all, if your fuel pump wasn't pushing a few air bubbles the car would start. Never mind the tank venting or gas leaking out the cap or liquid itself. If any fuel comes back to the tank as a vapor, it is pressurizing the tank.
Added heat from other sources matters to. Pump itself, fuel lines close to other heat sources etc.
You're essentially going from 3.5 atmospheres absolute to 1 plus or minus, plus head pressure of the volume in the tank because your return fuel line exits at the bottom of the pump, essentially at the same level of the pick up or sock filter. Plus excess bleed from the regulator and latent pressure from the injector side all the way back to the pump check valve.
Just say for arguments sake that rail pressure leaks down to 21 psi gauge after shutdown within five minutes. (from the FSM)
Various states of tune will dictate how fast it equalizes to zero gauge at the fuel rails.
Worsening the effect with the pressure loss.
During that time temp is going up on the rails from the block. That's easy to measure.
So that combined with the pressure loss on either side of the regulator only increases the likelihood of the pump attempting to push gas bubbles resulting in a hot no start.
There are multiple compression and decompression events happening in the whole travel path. The ones we're worried about are the pressure and temp in the tank and the pressure and temp in the return line. It's just pressure, temp, and volume. One can't change without affecting the other.
All that governs how and why, after it vaporizes/boils.
The process that gets it to that point is someone else's profession. But heat is what makes it happen.
The fuel tank absolutely should be close to ambient. You can see that on a gauge on the supply to the fuel rails. I measure 25-40 degrees F difference at all times above ambient.
For example, driving around all day start and stops, gas station, bank, beer store, UPS etc all day.
Every time I get back in and see liquid rail temp it gets higher and higher at each stop, depending on how long each stop took.
At each start up a 2 or 3 degree drop happens about every 2 seconds or less until it is equal to tank temp. (assumed)
Prolonged driving at low speeds has the same effect as multiple start and stops.
For a year or more this has been a constant. 25-40 above ambient. Driving through a hail storm or a big rain where the temp drops 20 degrees outside you can see this affect the rail temp also. It drops like a rock.
If you take j1000's info also you can surmise that adding an external cooler will keep the delta between rail and tank at 20 degrees F or less.
I wish someone would install a pressure sender on the top fuel tank cover where the connectors are, and a temp sender at the bottom of the tank.
My bet is there is an equalization point and multiple crossovers with temp and pressure. The gas volume expansion is undersold in the tank, greatly. I think this should be obvious to all. I realize the tank is not a fixed container and that some gas escapes, but in this instance where the volume of the gas overcomes the containers ability to vent you might as well consider it to be inflexible.
The shear volume of the tank under pressure and clicking the cap open a few clicks lessoning the ambient pressure inside the tank even just a few PSI will cause a lot more boiling than people think. Desaturation leading to boiling. This is what is happening when liquid and gas vent out your fuel tank.
Personally I haven't had the problem for years. I put in a new pump, walbro at first and then denso stock.
I did add a heat shield, didn't measure anything but it was solved before that with the pump. For me at least.
I don't run the engine cover or the black air train.
I also don't have cats, but that only results in a 100 degree drop at the flange closest to the Y pipe.
I have a pressure sender at one rail and temp sender at the other rail.
I thought this was the most important to keep track of. I have no intention of adding other monitors.
My coolant runs at194-197 F constant. Everything related to that system is new.