I would expect there to be a rating or a rule of thumb defined by experience. That’ll define what ambient can be once you find out how much it raises above it during loaded long running.
Oh, here’s an opportunity for me to think this through more thoroughly! 
The spindle didn’t come with a max temperature rating. For a rule of thumb, I’m pretty sure that a cheap spindle isn’t using the highest-precision spindle bearings possible, so there’s probably some slop. A lot of folks with these spindles apparently just run a submersible pump in a bucket of water with no radiator or fan at all, and I haven’t heard of seized bearings from overheating in that case. (Doesn’t mean it hasn’t happened.)
It’s a 1.5kW spindle, but most of that energy is going into the product being cut, so I don’t have to worry about dissipating 1.5kW of heat. The warmer the water gets over ambient, the more heat energy it will dissipate with a constant airflow at a constant ambient air temperature.
There is a certain amount of air resistance in the radiator, so the friction of moving the air through the radiator raises the steady-state temperature above ambient. (I’m sure I’ve seen that discussed here for why the CW-3000 style radiator is bad for lasers; it will run the water a bit above ambient even with the laser not firing.) Because heat transfer is a function of temperature difference, this will reach a steady state when heat transfer in from friction matches heat transfer out from convection. This is what I believe I was measuring when I left the system running for about an hour without the spindle running. I should have done that first, but as long as I waited enough, it’s all the same. It’s about 1°C above ambient, it looks like. Similar to what has been reported with the original radiator.
This is a 3-phase motor, so it is probably at least 85% efficient. That would mean 1.5kW max power * 0.15 max assumed heat loss = 225W max heat output.
With the original radiator, the unit was rated at 50W/℃. Even if the new radiator (with this fan and pump) achieves only 10W/°C, that would be 22.5°C max over unloaded steady state to dissipate 225W.
Obviously that would be running the spindle at the edge of stalling all the time, which I’m not going to get close to; even if I intended to, the large, relatively flexible setup I have (1M by 1.5M C-Beam) has enough compliance that it would flex and fail long before I hit that limit.
By this napkin math, I’m not going to worry about it.
3 Likes
I was told that this sealant [#5] contains teflon. https://amzn.to/3iUHw3g
Out of curiosity I checked and it does not:
… but this one (T+2) does: https://amzn.to/3oUIikL
I probably will switch to using the teflon …
3 Likes