OK let's work through these Q's.
........However, I have noticed that on almost any mod the vendor states that stacking batteries will void the warranty. Simply, is stacking batts safe?
Those vendors are most likely talking about their single 18650 model or similar - designed for 3.7 volt use with one cell. They would be correct in this, because it may be using the device outside of the design parameters.
Series connection (stacking) of two cells in electronic APVs that are designed for this usage, as they have a regulator that cuts the voltage from the max potential of 8 volts (2 x 4.2v = 8.4v) down to the required voltage, is not dangerous according to the history we have - no explosive events (or serious meltdowns with flame) have occurred (or at least have been reported). This is probably because these devices, which commonly use 2 x 18350 or 2 x 16340, have built-in protection against over-current, the fault condition that causes battery meltdown. The electronics basically stops a serious fault progressing.
The incidents have all occurred in mechanical and basic electrical metal tube APVs, not electronic APVs or boxmods. Stacking batteries in these basic metal tube devices is intrinsically dangerous as the fusing or protections are proven to fail occasionally. There is a fairly long list of devices that have exploded or gone into a rocket-mode fail where the endcap blew off and the device jetted away; it's not a design fault in one or two models, it is what might be described as a feature of non-electronic metal tube APVs. These devices should either be used with a single battery, maybe using a Kick if a higher voltage is needed; or if two cells are stacked, then a comprehensive set of safety features are needed. The most important of these are very large gas vents, which in the final analysis are the only things that will prevent an explosion if a series of faults chains up to produce a two-cell battery failure that progresses to a major outgas. Fuses and so on have failed to stop several of these events.
As these events are rare, we can assume that a fault chain is needed. For example: damaged or counterfeit batteries; a faulty charger that overcharges or damages the batteries; no user check of the batteries with a voltmeter; a torn battery cover that allows a different circuit path to exist; a fuse not working for some reason; a shorted-out 510 adapter on the connector that causes full current to be drawn (not uncommon), etc. It may not happen often, but when it does, in a virtually sealed metal tube, then the outlook is not good.
It's true that these incidents are rare, only about 12 or 15 are known to have taken place. Around 8 have been reported in detail on ECF, with photos etc. There were many minor injuries but only two cases of serious injury. In most events, there was a period of several seconds where a first-stage outgas took place, with heat, and the user dropped the device, which then exploded on the floor. In the two cases of serious injury, either there was no warning, or no small gas vents that provided a vent for the 1st stage outgas and warned the user, and the devices exploded in the user's mouth (the top end was blown off into the mouth and face).
In my opinion you would be unwise to use a two-cell mechanical or basic electrical device without a good fuse and large gas vents, especially at the top of the tube where the gas pressure is demonstrated to rise sufficiently to blow the top end off. These events, though, normally involve cheap unprotected Li-ion cells, or even primary cells (non-rechargeables) charged by mistake - so users of IMR / Li-Mn / Li-FePO4 cells are in theory safe; if you believe what the battery labels say, because counterfeiting of batteries is widespread.
So in conclusion:
- Electronic devices with 2 cells stacked have no history of explosive failures; therefore, statistically speaking, they are safe.
- Single cell devices, ditto.
- 2-cell basic metal tube APVs are where the issue is found. Unless you are absolutely certain (and willing to bet your eyesight on it) that you have genuine AW IMR cells (for example), then you should consider fitting large gas vent slots or lots of holes, especially at the top end of the battery tube. The batteries can swell and block the tube, if they overheat. This is likely to make the fail event more serious. The two cases of serious injury involved the top of the device blowing off. If there were gas vents at the lower end, they didn't work.
- Gas vents at the bottom end may be good for warning you something bad is about to happen. That is all.
- Blowout endcaps (that function correctly when they are needed, which is not guaranteed) will probably stop an explosion. They also cause the device to jet off and set fire to the room (or even two rooms, where the device went through a door in one case). These rocket-mode fails have occurred.
- When a blowout cap works, the reaction vector drives the APV into the user's face - something else to consider. Also, the force is enough to chip an oak door if the endcap hits it, so you don't want your child in the way.
- You can't beat large gas vent slots on 2-cell basic metal tube devices and that's all there is to it.
- If you look at the Akston Hughes SL then you will see that proper design and engineering can make the slots virtually invisible, and also waterproof to IP65 (it's an electronic device but with multiple safety layers).
Also, most Vamo reviews I've seen say that AW IMR 18650 batts don't fit tightly and will "wiggle" in the tube, sometimes breaking contact and turning the device off. Is safe to use the magnet battery spacers? Additionally, if using the magnet, is it only to be used on the positive terminal?
A loose battery is not a bad thing, for various reasons. It is less likely the cover will tear; if the battery fails then the gas could escape easier; and most useful of all, you might have room to fit a mylar sheath around the battery to protect it (a thin tough strip of protective material that stops the cover tearing or wearing through). It might be better to sheath the battery in some way instead of using spacers - those need a too-short battery to work, anyway?
As to which end to fit a spacer, then in some circumstances it won't matter, in others it must be at the negative end. If you just have an uninsulated metal disc then it must go at the negative end. This is a bad idea anyway as it would bypass a hot-spring fuse (it will short out to the APV case at the disc edge). A good spacer would have a metal centre and an insulated outer. I'd probably still put it at the bottom end. It might not connect well though unless used at the top end (pos) due to a flat end on the spacer not connecting with the batt neg end. Not really considered this before though...
Finally, I am looking at an Intellicharge i4 charger. However, it (and just about every other charger I've seen) do not list 18350 as being compatible. Most of the chargers I'm looking at do list compatibility with 17335 and 16340 batts. My understanding is that the first 2 digits are the diameter of the batt and the last 3 digits are the length of the batt. So an 18350 would be 18mm in diameter and 35.0mm in length and a 17335 would be 17mm in diameter and 33.5mm long making it shorter than an 18350 meaning an 18350 should fit?
I don't know this charger, but if it takes any of the 18000 series cells, and also a 16340 series cell, then it will almost certainly take the 18350 (because the large 18mm fits, and the short 16340 for min length).
As you say, 18 = 18mm. An 18650 is 18mm dia x 65 mm long x cylinder format (0), at least in theory.
Anyway, according to the data on this page, it pretty much charges any battery ever made
http://www.nitecore.co.uk/chargers/intellicharge-i4.html
[except Li-FePo4 3v cells - note!]
