Friday, June 26, 2015

Aluminum Standard Flywheels by DrSnikkas Initial Test

You have probably heard of this German nerfer "DrSnikkas" and the billet parts he makes for flywheel guns; particularly the CNC milled aluminum cage with canted-axis concave flywheels and a feed guide that has been all the hype. However what I have here is something different, a set of billet flywheels for standard Nerf-design cages.

Thanks to Bobololo for providing these to be reviewed.

I am not sure that they are, or ever will be, a product. They were sent to Bobo along with one of those full cage kits as a "goodie" and it is my suspicion that they were a leftover prototype.

The quality of machining is quite good with only light chatter and a few burrs.

Compared to a Stryfe fly. The outside diameter is identical minus the taper found on the molded Delrin flywheel (draft angle for the injection molding), as far as I can tell without calipers anyway. The hub diameter is larger and you will need to mod your left cage half to clear which is no big deal.

Height is identical. Obviously these are for Stryfe, Rayven and RS cages, not Demolisher.

Rim thickness is about 1mm and very close to Stryfe or Rayven flywheels.

Now here is where things get a bit wonky. The bore size on these slips loosely over a standard 2mm motor shaft. There is no set screw or other means to lock it in place. I reported this to Bobololo who asked the manufacturer. The answer was to use superglue.

I am not sure why these are designed in this manner, as opposed to a press or shrink fit like most other hobby miniature motor attachments; or with a set screw or two.

For the purposes of the test, I did use CA. I can confirm that CA works excellently as a retaining compound - but it is very unforgiving. When installing a part onto a shaft with CA, one must slam it home in one quick motion, and here you need a ~1mm shim to place on the inner surface of the cage body to set installation depth of the flywheel. If you hesitate or the part binds, it can end up locked in the wrong position, and then you will have to remove it, clean everything, and start over. If you need to do that, apply solvent (I used Gumout carb cleaner) to the bore, and be patient. Heating with a lighter after drying off solvent can help loosen the CA.

That brings me to another worry point of this design. They have lightening holes in the web, which leaves very little material for structure. Prying or using a puller on the rim as with standard Delrins is not recommended. You could bend that web easily, so be very careful with removal and installation force.

for future and permanent installation (as in entire service life of the motor) one could use Loctite 609 or other anaerobic retaining compound designed specifically for this task, which would be a much easier install than superglue but harder to remove.


I used my RapidPistol equipped with Mabuchi FK180SH-3240 motors, which has recently been a test mule for some other stuff and is waiting to be cleaned up, bodyworked and painted. After getting them installed perfectly, the first thing I did was to record audio of a motor startup. Previously, I had recorded the same clip with the RS flywheels and the same battery, which was not discharged any more between tests.

This covers 3 observations that need to be made:
  • Effects due to inertia of the heavier aluminum flywheel.
  • Noise level.
  • Motor top speed as an indication of objectionable windage load.
About that latter point, I was quite concerned about these because of those lightening holes. I do know that serrated flywheels can have excessive air drag and impose an undesired load on motors, causing a reduction in available power for transfer to darts, reducing top speed, increasing response time, increasing current draw and motor heating, all of which are bad. A severe case (i.e. a small 130 motor driving a large serrated Buzzbee flywheel) could be overloaded even at static idle and result in a motor burnout.

This however proved not to be an issue. The motor temperatures stayed down and top speed did not decrease as indicated by audio, which suggests the load from web speed holes catching the air is negligible.

Noise and vibration, in person, were greatly reduced by the far superior balance of these versus the molded Delrin stock flywheels.

However, response was severely compromised. The greater moment of inertia produced by switching this design to a denser material completely killed the reaction capability. The chrono shoot sheet will cover that, but in the end I would consider these not worthy of HvZ service when used with FK180SH-3240 motors. As these are effectively the pinnacle of 20.4mm flat can motors as torque goes, this problem will only be substantially worse for 130 players.

The chrono shoot was uneventful. 40 darts were fired, with the second run (yellow trace) using reused darts. The blue trace is the stock RS flywheels with decent buildup.

Do note the sorry snap numbers, however. If you get surprised, the dart might go about 20 feet. The errors were due to insufficient velocity to trigger the F1 chronograph.

These things need either a lighter material or a bigger motor - badly.

This raises an interesting question. Which should it be? The benefits of metals (high rigidity and easy machining to a high degree of precision, without the costs associated with exotic materials) shouldn't be ignored. Of course carbon fiber composite, Delrin, nylon, impact-resistant polycarbonate, and other materials could reduce the mass and inertia of a flywheel, but what is the point of 20.4mm flat can motors anyway?

To fit in a Nerf style cage.

If we are talking about hobby grade hardware that uses machined flywheels, we are already seeing as a concept custom cages to drop into the receiver of the Nerf or other brand gun; and clean-sheet hobby grade designs aren't constrained by that either. It is just as easy to hook up any other motor we want. People are designing with the 20.4 flat cans because they are common, tested, and well known. But why not use a 280, or a 20mm round brushless inrunner, or nearly anything else?

Me and an old friend were here, doing that, in 2010. Hacking Nerf cages up to run 300 series motors. I did another, a 280 Stryfe, in 2012 because the FK180SH-3240 had not yet been found. Lest we forget the old school Banshee conversions, they may be poised for a comeback.

A final observation surrounding chrono testing. I fired 400 breakin rounds before chronoing - but no buildup developed. See that shine? These flywheels just don't seem to support its development.

Why is not clear. The surface finish on these is turned. They are not polished, blasted, or anything else. Perhaps that is a factor. Perhaps flywheels need to be polished. Another idea is the thermal conductivity of aluminum, which may be so effective at keeping the contact patch cool that foam melting is reduced. The flywheels DID get warm during the shoots, but this may have been conduction through the motor shaft from the armature windings.

In any case, I think metal flywheels might need some design reconsideration in order to account for this change in behavior, such as a gap decrease. This begs the question of why this would not be also done to a plastic equivalent - and whether a plastic flywheel may be able to inherently reach a higher velocity or greater accuracy due to the self-adjusting buildup process.

I did not go about measuring accuracy super rigorously because I do not have an optimal point of comparison (without an inner barrel) and I am broke, though if people complain enough I might create one. I did a 35' full auto mag dump with Yutoys June '15 koosh and both this RapidPistol and the Tacmod 3.1 which is shooting nearly identical velocities at the muzzle. The T3.1's group was about 6 inches tighter which is an expected result of the inner barrel and full breakin. You are looking at an ~18" radius spread with a central 18" diameter concentration for the T3.1. In the end, I saw nothing interesting about these, they shot exactly like other unbarreled flywheels. All my stock flywheels are selected and balanced sets and these are stock-geometry flywheels, so this is all expected.

Update: A retest with the same Stampede inner barrel used in the Tacmod inserted into the cage of the RapidPistol yielded no improvements.

1 comment:

  1. Thus has been the crux of my argument about "hobby grade Flywheel" designs, why are people bothering with such small sized obsolete motors if they are designing the cage, blaster and she'll from scratch?