30mm blower VS 30mm fans

On this topic I will try to evaluate if the transition from one small blower to two small fans is effectively better

Until now I'm relying on a 30x30x10mm blower (this one)
Relevent specs :
*CFM 1,4 or 0.039m3 /min
*static pressure 0,285 dans H2O (71,0 Pa)
*dB 29
*speed 7 500 tr/min

This very small blower was interesting for its compacity, but does not exist in 24V.
I think it could be better to use two 30x30x10mm fans instead, for a more even cooling (two sides instead of one), and to use more standard parts (makes repairs/replication of the project easier)
(e.g. this one).
Relevent specs :
*CFM 3,5 or 0,098 m³/min
*static pressure 0,100 dans H2O (24,9 Pa)
*dB 25
*speed 9 000 tr/min

I learned around here (and the specs confirm it) that blowers are known to have a better static pressure, which is interesting for small or complex fan ducting (too many fan duct on thingiverse are not adequate for axial fans, the air flow bounce back).
But axial fans generally have much more volumetric output, which is better in term of cooling power, I guess.

I tried to calculate if the input-output surface difference in my last fan duct design would be a problem (for the small fan), but as the duct is quite straight it seems that it should work.

That will be easier to verify empirically (= a bridging test with the blower and one with the fans).



Edited 2 time(s). Last edit at 11/26/2017 09:18AM by Emmanuel.

---

about // liberapay // flickr // wiki // thingiverse - github

Today I discovered SimScale.
I used it to perform a Computational Fluid Dynamics analysis on this shape.
Making the volume usable for the online tool was a bit tricky (STEP was better than IGES or STL) but here is the result



I guess it's ok.

---

about // liberapay // flickr // wiki // thingiverse - github

Looks good ^^


(too bad, they need to be mounted the other way)

For my test I have two "UDQFH2H08" that are almost as good as the reference in the first post.
Relevent specs :
*CFM 3,18 or 0,09 m³/min
*static pressure 33 Pa
*dB 19
*speed 8600 tr/min

By comparison the fan used in the e3D hotend (this kind) is 30dB / 10k RPM / 4,08 CFM (quite noisy), for future tests I will use something in between (this kind) 25dB / 12k RPM / 4,7 CFM

By the way, that made me think : why did they changed from two 30mm fans to two 30mm blowers between the Ultimaker 2->3 ? (The only reason I see is the slightest more compact form factor, for the dual head).
We have both in the local FabLab and I have the impression that the 3 have less cooling but it is normal when you look at the specs.

Edited 3 time(s). Last edit at 11/29/2017 09:37AM by Emmanuel.

---

about // liberapay // flickr // wiki // thingiverse - github

Ok, so as planned :

Blower (CFM 1,4 or 0.039m3 /min)


Two fans (CFM 3,18 or 0,09 m³/min each)


With 454% more CFM the result is not 454% better, but at least a little more
Especially the fourth step is more consistent.
I will make other tests, to see not only on bridging but also were the plastic would typically goes up (small surfaces + heat build-up).

There is also a probable way to improve on it by changing the shape of the duct to be closer to the nozzle (maybe).

---

about // liberapay // flickr // wiki // thingiverse - github

Since the modification, the experience seems better overall.
More specifically, I printed this, a geometry with small cross sections that was especially tricky before at small scale.
Or I would have to print several parts at the same time to let them cool and avoid the thermal build-up :

---

about // liberapay // flickr // wiki // thingiverse - github