Yes, increasing the pressure of the air forced through the hollow screw would increase the volume of air flowing in. The question is “will that volume of air cool the rotor in a significant way?” Has anyone looked at the airflow pattern inside the motor?
As I stated earlier, the rotor is isolated/insulated from the case by the air gap between them.
My questions are based on the issues addressed in the twice yearly R&D updates from the various R&D Departments, including one that designs motors from scratch, that I sat through for 15 years at my last job.
The fact that only one motor on each side is getting hot is definitely indicative of some sort of mechanical or electrical issue, generally gearboxes that use multiple of the same motors should heat relatively evenly.
Buying all new CIMs is definitely a good call, I would also recommend taking a multi-meter to your speed controllers (with motors unplugged) and verify that they are, in fact, outputting the same voltage. Also, check to make sure that all the speed controllers are all in either coast or brake mode (but not a combination of both, also coast is probably better for a drivetrain).
If you want a non-permanent mount, those steel ties are good, but I would still recommend using either the thermal tape that MisterTim posted above, or using regular thermal compound. Either of these options will vastly improve thermal transfer over just mounting the heat sink straight onto the motor.
Why do you think the root cause of the motors being hot is that they had been used the previous year? Do you have data showing the temperature difference from year to year? Have you made changes that affect how the motor is used in your drivetrain? What is the airflow over each motor like and how has that changed over the last year?
This is an engineering competition. Engineering decisions should be based on critical thinking informed by solid objective data. It appears that you are basing decisions on comparisons of single subjective data points taken under different conditions.
Before we had a counter-spring on our robot arm, the motor required constant current and heated up.
Our first reaction was to put am-2865 fins on the mini-cim. We had them laying around the shop from First Choice a couple years ago.
When we added the spring, the heating-up went away, but we left that heat sink on it.
So then we go to Wisconsin. Senior Robot Inspector Big Al Skierkiewicz actually made a point to come up to me and say “you know, that heat sink doesn’t do anything. If you don’t make weight, you should remove it”. Also, “all the heat goes out the shaft.”
Initially, though I know that the heat in a CIM is in the armature, I wanted to argue that the heat from the armature obviously does make it to the magnets and the frame, at least some of it does. I know all of the heat gets out–some way or another. The heat sink can remove that frame heat more easily, over time. But yes, I understand that in a match, this won’t be much time.
Another mentor on my team wisely advised me to not argue with Big Al.
Also, I’ve wiped water on a hot CIM and watched it evaporate immediately, indicating absorption of heat. Heat of vaporization indicates really quick usage of calories. I’ve seen people in this string talk about not mixing water and robots, but if you just wipe the outside of a CIM with a damp towel, you’ll cool it really fast. And I don’t have seven bucks to waste on canned air every match.
I think the difficulty of unmounting and remounting CIM motors in most applications is a big factor on why people don’t do this. It’s also important to remember that using canned compressed air upside-down (so it’s cold) can actually crack the internal ceramic components of a CIM motor, causing permanent damage. Even when we use it on the outside of the motors, we try to use it sparingly.
Let me be a little more precise… The motors were used previously and were realistically in an unknown state when they were installed on this year’s robot. So, we suspect that something bad probably happened to them in a previous year. Further, two of the CIMS were known to be new, and two were known to be old.
It would be lovely if we had the ability to do the sort of in-depth analysis you suggest. But, the fact is that we noticed the problem toward the end of our last event. After that event, we stuck the robot in a bag and had to come up with potential solutions that we could implement in our 6-hour unbag window, along with other things that we wanted to do. Replacing the CIMs was one of the solutions.
A thing on the super backburner is compiling data from a few tests I did in the evening back when I worked at VEX, but here is the gist of it.
Two vented screws on a CIM, run on dyno at constant 40 amp load till a thermistor (stuck through vented screw, so measuring ambient temperature inside) read 130deg F, stop test, time how long it takes to get back to 80deg F using various methods.
Method 1, set motor on wooden bench and wait, ~1hr 20min
Method 2, set motor on wooden bench with 120mm KOP fan pointed at it, ~20min
Method 3, Blast 80psi compressed air through the vent holes, ~9 minutes
Method 4, Shoot freeze spray inside, ~15sec (hard to really measure this one accurately, but it was really fast and overshot 80deg)
These numbers are from memory, but definitely in the territory.
Someone else can definitely perform the same test and post with more rigor, I got what I needed to know from it, and as such haven’t revisited it in awhile.
The internal temp vs power map was also interesting, and good motivation to keep motors cool.
I’m not sure ambient internal temp is that much better a proxy for armature temp. Method 4 specifically seems unlikely to be that accurate since you’re more likely measuring all the newly evaporated gas. It’d be interesting to see the same series done but using the armature resistance as the proxy for armature temperature. It’d be even better because the armature resistance increasing with temperature is the whole reason for the power loss, so you’d be directly measuring the variable you really care about.
To do it accurately, you’d probably need to index the motor to a known location whenever you’re measuring the resistance to make sure the brushes are seated about the same. Then use a strain gauge/wheatstone bridge setup to measure the armature resistance since you’re talking about 0.393% change per degree C. So probably less than 40% change in an already small resistance. Your initial condition could still be the ambient internal temp, or run it till you think you hit steady state at some current, or cut it off when dyno RPMs drop below some threshold.
I’m curious - has anyone ever tried putting some sort of fan/heat sink on the CIM shaft? This wouldn’t work with most of the Vex gearboxes, but the end of the CIM shaft is exposed or extendible on the WCP shifting gearboxes, the AM Toughbox series, and the Vex single stage gearbox. It seems that putting an impeller or fan shape which is thermally and mechanically coupled to the shaft would do a better job of carrying heat away from the armature than a heat sink on the outer case.
While arguing might not be the right thing to do, asking how and why is great and you should do so the next time you see Al. (Maybe he’ll chime in here?) I think that your mentor should have encouraged you to ask.