I’ve been comparing several PPA-CF filaments and noticed something interesting — BambuLab’s PPA-CF lists a much higher HDT (at 1.8 MPa) than other brands:
Siraya Tech PPA-CF — 82.5 °C
Phaetus NexPA-CF25 (PPA-based) — 83.0 °C
QIDI UltraPA (PPA-based) — 72.5 °C
BambuLab PPA-CF — 196 °C
All of these values are for unannealed, 3D-printed specimens (not injection-molded), tested under ISO 75.
QIDI doesn’t mention annealing, but since Phaetus manufactures their filament, it’s probably unannealed as well.
I also checked MyTechFun’s independent comparison table. His tensile and flexural results don’t show a huge difference between these filaments — stiffness and strength are fairly close, though BambuLab’s PPA-CF does have noticeably higher impact strength (≈ 8.1 kJ/m²).
Still, Bambu’s published tensile strength is about 40 % higher than other brands under the same ISO 527 test, which made me wonder:
Are Bambu’s TDS values possibly inflated?
Or could they actually be using a more advanced copolymer, like a high-crystallinity PA6T/66 blend with a higher 6T content, while others use less crystalline variants?
The latter could explain both the higher HDT and better impact resistance. But it’s hard to say for sure — the numbers look a bit optimistic, yet they might still be technically real if Bambu uses a premium formulation.
What do you think? Has anyone else tested this material side by side?
With PPA, the aromatic content plays a huge role in regards to HDT. PPA higher in aromatics will have a higher HDT right after printing. Low-aromatic PPA will have a lower HDT. That’s without annealing.
Annealing will greatly increase the HDT for low-aromatic PPA, but the higher the aromatic content, the less improvement there will be. Simply put, more aromatics → more crystalline structure (after printing) → higher HDT.
There are other consequences of the aromatic content in ppa, but I’ll leave it at that for now.
Siraya Tech might have made a mistake with how low their stated HDT is or there’s a huge difference between the way the specimens were printed. MyTechFun’s results for unannealed Siraya PPA-CF looked similar to that of Bambu PPA-CF.
I don’t think BL’s TDS has inflated values. Their TDS contains measurement errors, printing and testing methodologies, and there’s still room for improvement (imo). For these reasons, I am keen to trust their data much more than many other vendors. Any measurement without the measurement error is not measurement, it’s observation. However, getting any statistically significant result requires testing more than three samples and that takes time. For this reason alone, I treat all YouTube reviews as rough estimates of what to expect from a filament. ISO metods strictly require at least 5 samples and I think for ASTM the minimum is 6.
There are so many factors that play a role in mechanical strength of 3d printed objects that simply printing and breaking 1-2 samples is far from enough to validate or invalidate data from the TDS
Edit:
I forgot to add that while siraya makes good budget engineering filaments their TDSs are toilet paper. No info on precision → not an ISO conformant test
Based on tds themselves yes. BUT, as I mentioned, in Igor’s tests siraya did just fine. The result of hdt measurement may differ based on print orientation.
Is it for a critical part that you’ll only print once? If so I’d buy once cry once but the differences might be almost negligible, though the BL PPA hss a higher melting temperature so it should have a higher HDT without annealing
I saw a YouTube video that was showing that the Z direction tensile strength (layer adhesion) was not anywhere close to Bambu’s specs of the PPA-CF.
But later he managed to get close to the specified values only by printing with fan completely off (after talking to Bambu).
Unfortunately, the TDS doesn’t say the specimens were printed without fan, and also the default profile for the PPA-CF in Bambu Studio will have at least the fans turned on when printing walls (which can be a significant amount of the print area on a small test specimen but shouldn’t be too relevant in real-life prints).