Is TPU print initialization supposed to take forever?

I have a H2D and I’m printing TPU 90A. I just got one print done (successully).

The pre-print initialization took forever (probably close to 45 minutes) and I thought something was broken because the extruder just kept moving in a square: scanning, scrubbing the hotend, going around the front corners, then repeating the whole process with nothing happening that I could see.

Is this normal? I can’t find anyone else talking about it or any documentation on what’s going on here or if there’s any way to shorten this cycle.

I understand the H2c can take a while to do its pre print checks, up to 30 mins if i remember correctly .
BambuLab H2C Full In-Depth Review: Extensive Tests vs SnapMaker U1, Prusa XL & More
Bambu Lab H2D In-Depth Review: Dual Extruder, Multi-AMS, 10W/40W Laser, Blade & Pen All Tested
check these reviews

My H2D is unbelievable slow until it actually starts printing.
Constant hotend switches and heating up and cooling down hotends (why on earth multiple times???), cleaning nozzles a few times. Without any exaggeration I can state that for smaller prints my MK4S can finish the print including initialization and bed levelling where the H2D has not even started printing.

But: My H2D is not that slow that it would take 45 minutes for init as you described. If I remember correctly a mixed TPU/PLA print I have done had probably around 15 minutes init time.
Is it possible that it did a full bed levelling?

The different temperatures are there for a reason. Nozzle cleaning (scraping over the perforated steel section) happens at printing temperatures. The bed leveling (needs a clean nozzle) is done at 140C.
After that the nozzle offset needs to be done at printing temperature.

The initialization process (especially the nozzle swapping) could possibly have some optimization, but that won’t be huge time savings IMHO.

Are you sure? Haven’t clearly noted the temperatures on my H2D yet, only that it goes up/down more than once during init.
Imo that wouldn’t make sense as there’s always a bigger chance of oozing at print temperature. My other printers so far (not all of them had nozzle cleaning so only for the ones that have it) were cleaning the nozzle slightly over melting temperature (like PLA at 160 - 170 °C). This makes any residue on the tip soft enough that it can easily be wiped off but the filament isn’t so liquid that it might be oozing again.

Imo it would also help reducing time to load the filament from AMS after all other init steps have been done. That would reduce a heating/cooling cycle and it would also reduce the chance of oozing and therefore less time-consuming cleaning requirements.

Anyway, I can live with it. But I doubt that there is no significant room for improvement here.

Lets say it like this:
The nozzle cleaning starts at a reasonable high temperature (and I am sure it’s related to the extrusion temperature). It runs over the pad twice and finally waits on the pad until 140C before it starts with bed leveling.