I really enter now in the TPU testing, really...
Some test with TPU 95A : infill pattern impact (there is a file were someone did exactly this on Makerworld, I decided to use it). At the end I can tell than grid, cubic or anyother relying on parallel lines are not what we want for TPU. Honeycomb infill, on the other side seem to be the perfect candidate for bumpers as it is strong in all direction, while still allowing for some flexibility. Gyroid maybe ok for tires in 95A TPU from what I could see from this test.
Based on this I proceeded in a second test which aims to approach what we would see witha tire printed in TPU 95A. I used 2 top and bottom layers, and gyroid infill 3% - 24% by 3% increment. The result is that it maybe fine without side walls, but with the sine walls, the side walls are too strong so it doesn't behave at all like a tire... So this confirms more or less what I had in mind with all my previous TPU and Foaming TPU tests.
Next in line, I have a roll of Recreus Filaflex 60A. it's not TPU but TPE. The filament is highly flexible and it is difficult to load due to the friction in the PTFE tube (no way to go thru the tube - this confirms the upper mount is absolutely required for this type of filament). At the end I did a cold pull with PETG to remove anything from the hot end before loading it, and then I loaded it from the upper mount (the shorter tube is allowing it to load, with a little bit of persuasion - I suspect that it would not have worked if I did not cold pull before.
Recreus has made a profile for the filament and it is available on makerworld. I downloaded it, and use it with my test model.
One thing I can already say while a long print is in progress is that when there is accumulation of material, this doesn't behave like the TPU : the material bonds together and it seems that it remains flexible. For tires this might be exactly what we are looking for... Anyway, first test print is in progress...