Spindle Fan Failure Headaches

Spindle cooling fans can be a weak point on CNC equipment, causing expensive and time-consuming breakdowns. Here, pros commiserate and offer advice. January 27, 2008

I am having problems with my HSD spindle fan. We have broken 6 different cooling fans on the top of the spindle. They have just self-destructed. I think there are a number of issues with these $15.00 plastic parts.

1) The fans wobble when installed on the shaft.

2) The fans are constructed from numerous smaller parts. Therefore any vibration causes the fan to fall apart, specifically the joint between the center and the fins. Simply falls apart.

3) It appears that the thickness of the fans is ever so slightly too big. My guess is that when the spindle is assembled, the fan installer taps the snap ring with a small mallet, or uses a press to seat the snap ring. In the process the joint between pieces is compromised. All it takes to break it is a little vibration.

Every fan that has broken has done so when there sounds like some resonance or vibration caused by the cutter. I have been using a 1/2" diameter cutter, 2 flute, upshear with chip breakers. 17,000 RPM and 125IPM feed rate, 3/4" depth of cut in both fir and oak. Three fans that have self-destructed this time around have only lasted about 620 lineal inches of cut. Has anyone else had a similar problem? HSD has given us differing answers as to a possible cause and solution.

Forum Responses
(CNC Forum)
From contributor J:
As for a cooling fan, we once had to replace our metal one for several hundred dollars because the fan itself failed, not the fan motor.

From the original questioner:
My fan failure is in the fan itself. Mine does not have a fan motor. It is driven by the spindle itself. Slow the cutter down and the fan slows down. So the fan itself is falling apart. Over and over. And in this case it is plastic.

From contributor G:
Sorry I can't help you with the fan, but you are way out of whack on your feed speeds vs RPMs. Slow your RPMs down to 12K at that feed speed or increase your feed speed to 650IPM at that RPM. This may eliminate the problem.

From the original questioner:
Thanks for the info on feed speed and RPM. I originally had it programmed for 17000 rpm and 260 ipm. That combo breaks bits. Yeah, new bit, new collet, torqued down, and work piece clamped securely to table. The replacement bit did not burn, even after all parts have been cut. Chip size varied somewhat based on with grain or cross grain cuts. Mostly chips. So you think after breaking 2 - $50.00 bits I should increase my feed speed? What kind of a chip load are you figuring? Somewhere in the back of my thick head I think I should be able to run my speeds based on what works, not on whether the chosen speed is going to take out the fan. Another way to say that is "where in my chip load/RPM/feed speed calculations do I enter fan destruction?"

From contributor B:
I'd look at Grainger and McMaster-Carr for replacement metal fans. Seems like that would permanently solve the problem. My HSD 915 and 919 spindles both have fans with separate electric motors. That way the tool vibration doesn't transfer up to the fan blade in a direct path. A much better design in my opinion.

From contributor G:
I am not sure if the feed had anything to do with your fan problem. You are right, as tools work differently in different shops, but usually only by 10%. If you had two bits break, a number of things could have and did happen. Two main problems are tool length too long and/or the tool heated up because of the FS vs RPMs, and broke. Just remember to make chips, not dust.

From contributor B:
What are you using for hold down? We used to break a lot of bits holding with vac pods until we worked out all the kinks in the system.

From the original questioner:
Thanks for the responses. Contributor G, I do not think it is a heat issue. I have gotten a lot of very good information from you guys at Courmatt, both over the phone and at the AWFS show, so thanks for that. That info has focused my attention on FS and RPM, with a careful look at the chips being produced. I obviously still have much to learn. Bit length was a little long because we were making curved 3" thick white oak rafters. My cutting length was 2.125" so I took it in three passes with no pass greater than .75".

Contributor B, we are holding the pieces down mechanically. Block and cam is what we call it. We use a block screwed to the spoil board and then clamp the billet using cams. The billet was big and heavy to start with, 3" thick, 21.0625" wide and 61.125" long. Clamped with 6 cams. I think it was held down good. We have used that method on a variety of products, with good results.

I still say that the fan should not self destruct because of what is happening on the other end of the spindle. Of course "should not" is the operative phrase. Someone mentioned making my own fan, so we'll see. Use my CNC to make replacement parts for it.

From contributor O:
Hey, just remove the plastic fan blades and mount a small computer fan on top of the spindle. Works fine and the airflow is constant - better than the original and cheaper. You can then add a separate fan speed control if you like.

From contributor T:
We too had been having some issues with needing to replace the fan motor as well as the blade. The original (blade) was metal and the first replacements were plastic. The plastic ones were almost impossible to get seated correctly and would cause a vibration soon after installation. I switched back to the metal one and the problem completely went away.

From the original questioner:
I rigged up my dust collector to pull air through the motor to cool it and I have ordered an electric fan that will need to be attached and wired to replace the HSD fan. So I am fixing it myself.

From contributor J:
Keep in mind liquid cooled spindles allow you to get rid of the fan altogether. Those HSD cooling fans are a pain in the butt to maintain. I do agree that support from spindle manufacturers is lacking. Although if you were interested in selling more spindles, it would make sense to use a cheap fan that breaks and causes the spindle to overheat. Not to mention the downtime you can not afford, thus the need for a spare spindle. We have had good luck with the liquid cooled spindles on DMS machines.