I’m wondering about the accuracy of the density data for the TPU HF. The physical properties sheet lists the density as 1.22 g/cm3. I’m printing items that require very accurate weights. So far, the TPU HF is considerably lighter than advertised. For example, normal PLA has a density of 1.26 g/cm3. When I print my model in PLA it weighs 168g. I print the same model with TPU HF and it’s 140g. Has anybody else seen this? Also, while preparing the model in Bambu studio it gives the weight as I’d expect. Both TPU HF and PLA have similar weights (after slicing, no supports). TIA
Filament moisture can have a drastic effect on actual density.
And of course printing parameters such as nozzle diameter, line width/z-offset as they can lead to increased air inclusions for different filaments.
I agree, although I think this is a different problem. I designed the part in Fusion 360. When I use the density (TPU or PLA) in Fusion, the weight agrees with what Bambu Studio estimates. But, when I print the part, the PLA part comes out almost perfect and the TPU print is off by more than 22%.
Did you rule out moisture yet?
Even new spools can be fully saturated. And TPU takes up moisture quite quickly.
It may only be a few grams per spool, but, when evaporating, water occupies over 1500 times its liquid volume. That’ll cause quite a difference in density.
Edit: TPU has 3 timed faster water uptake than PLA according to Bambu and is listed as requiring drying prior printing. Optional for PLA.
Looking at other sources, they both seem to saturate at up to 1%. But TPU does so within around a day (depending on conditions).
The model I’m printing should weigh 180g. It came in at 140g. So, too big a difference for moisture. I mic’d all the measurements and everything was good. Their old TPU 95A worked fine for me. I’m used to seeing a couple grams difference from the calculated value to the printed value. I think their data for the new TPU HF is off (considerably).
Well, a variation in density should be pretty easy to verify by weighing and measuring samples with usual household means. It is certainly possible. At least, you should be able to claim a new roll from Bambu if you can verify that their QC failed on that.
Do not rule out moisture though without verification. With 1g of water taking up a volume of over 1.6l when evaporated for example in a hot nozzle, and hence more volume than a full roll in the first place, humidity can certainly explain any kind of density difference in low pressure processes such as FDM. While this does not explain differences between PLA and TPU that you are seeing, the much higher water uptake speed of TPU can be the cause.
Also, my BVOH appears to take up moisture faster than my AMS desiccant. So that roll keeps my desiccant dry rather than the other way round. 
Not sure about TPU though. I’d recommend to always consider rolls soaked fresh from the bag.
So you can certainly check the density of your filament prior to printing. Then you can use that density to weigh in your part. If you still have differences, look at moisture.
You could also try the same print again after drying your filament thoroughly just before printing. It does not sound like you have tried that yet.
I would calculate the volume of 1m of filament based on the average diameter of your filament, then use the advertised density to obtain the expected weight of 1m of filament. Then cut a piece 1m long and put it on the scale and compare.