I did a deep dive on fans in that form factor here:
I think the biggest thing to consider is likely going to be mass. This fan gets thrown back and forth, as it’s on the head assembly. Any fan that operates at the same voltage, draws similar current, and has similar mass should be fine, I’d think. Question is whether or not you can find one that throws more air with those specs.
You’re limited by the scoop geometry, and at some point trying to shove more air through it will have diminishing returns, but there’s probably some room for improvement.
being a mechanical designer, and having worked on a project over the last year or so, designing a unit that had to take airflow into consideration to keep from overheating with severely limited space, my brain automatically wonders about airflow improvements.
But the air assist fan is not in there to help with cooling. It is there to clear smoke from the area so it doesn’t impede the laser beam. Since it can affect the material I would think you would want a fan with the least amount of flow to just get the job done.
and you’re probably right. i’m still learning, definitely. the “mechanical designer” bit wasn’t to try to throw my weight around, more of an explanation of why i’m wondering these things. i’ll shut up now.
And more airflow would probably help with that. @jbmanning5 has talked about this before re: his other machines.
The issues that are really hard to get around are physical… mass and dimensions and power supply. These largely scale with fan capacity, and we appear to be pushing some limits already. The higher volume fans are largely combination fans, with multiple turbines inline to generate more airflow, which would probably break our physical constraints by a large margin. Think of how precise clean corners has to be. If we increase (or decrease) the mass of the head we’ll probably first see clean corners break down. Increase the mass enough, and we would eventually overwhelm the steppers and/or belts, which is a whole other level of bad.
So we’re already up against some pretty exacting physical constraints. Not sure how to deal with that.
If I were more inclined to improve the air assist, I would first look at creating a more concentrated, direct flow as opposed to just a larger flood of air. However, I’m not an engineer. I don’t know precisely what happens when you force feed fan generated air into a very small outlet to concentrate the flow. I think that you’re going to come up against static pressure problems. Having compressed air comes with its own host of problems, beyond just the all-in-one form factor GF worked to develop. One thing to note would be that generally air assist works (on other systems) at a pre-determined distance from the material. The table moves up and down, so the lens and air outlet maintain the same distance from the material regardless of how thick the material is. That won’t happen with the Glowforge since the lens moves and the table is static.
You’re definitely right that EVERYTHING is connected. The whole is greater than the parts.
i would definitely be intrigued to see how it could be improved… i wonder if a direct nozzle, similar to how coolant is sprayed directly on a cutting tool in a mill, would affect the cut. would definitely kill the need for the rear-mounted air assist to clean…
Most systems use an air compressor and the air either leaves via the nozzle, below the lens, and directly down and into the cut (the air is blowing through same place the laser beam goes through), or a separate fixture that is outside the head but directed right at the cutting point. How much air (SCFM) gets ran varies and is somewhat difficult to measure because pressure is adjusted in a larger area of tubing but it’s generally only from a few PSI up to 30-40. But it’s generally not going to cause flare-up — it’s pinpointed enough that it’s entire goal is to reduce/eliminate it.