Look closely at the tire and...
Look closely at the tire and rim and you'll see large gaps at seven and nine o'clock where the rim guard on this Krawler has moved away from the rim. Getting the tire this tweaked was accomplished with a winch, but similar side-loading takes place on the trail. Despite what it looks like, and perhaps because it was freshly mounted, this tire held air down to 2 psi, complete with a rim guard, but using another wheel could've made all the difference.
Our thinking was that the outside bead is the most likely to give out, since it's hard to put much force on an inside bead during normal 'wheeling. Yet we had experience with, and reports of, inside beads going out on bead-locked wheels.
Jason Bunch of Tri-County Gear had noticed vehicles struggling with anything less than 7 psi in rockcrawling competitions. Soni Honegger had lost downhill-side inner beads while traversing a snowy slope, both coming and going. And we'd lost one just going downhill. Now the common ingredients were bead-locked wheels and rim guards.
To try to figure out what was going on, we rounded up a few tires and wheels, then winched a test vehicle sideways on aggregate concrete to see how each tire-and-wheel combination would act in a severe side-load situation. We kept the pull on the winch right at the threshold of dragging the tire sideways (which takes considerable force on aggregate concrete), while airing down until the tire would let air escape-or until it was below 0.5 psi, usually. This loaded the outside bead big-time, while folding the tire onto the inside bead, showing any weaknesses with the rim protector and/or rim design. We got some surprising results.
While we only had a select few tires and wheels to work with, some things became obvious, even with our limited supply. For example, not a single tire lost the outside bead. If the tire eventually did lose air, it was always the inside bead that let it out. Not exactly what we'd expected. Less surprising, a tire that'd been on a steel rim for 20 years just wouldn't let go of the rim.

On the outside of the same...

On the outside of the same BFG tire and wheel (this side had the rim guard removed) is an impressive sight. At 2 psi, the tire's still hanging on, but the gap is substantial-again, likely made possible by still-wet bead lube. Obviously, dirt could get in a gap like this on the trail, but without a rim protector it's also more likely to fall back out before the gap closes.

Again, the same Krawler and...

Again, the same Krawler and aluminum wheel, but now with the rim guard facing the outside. That's a pretty darn large gap between the rim and rim guard, and it goes for about 200 degrees. Any dirt trapped on that shelf doesn't necessarily fall out, which is probably why we hear that BFGs and other tires with rim guards can need frequent cleaning on the trail.

This Pit Bull Growler with...

This Pit Bull Growler with removed rim protectors didn't begin to leak until just below 1 psi on a basic aluminum wheel, once the inner sidewall was perpendicular to the wheel, which is almost exactly what happened when using an aluminum Allied bead-lock wheel. With the Allied wheel and the rim guard intact, the tire let air out at a significantly higher 3.5 psi.
The areas where the sidewalls folded, which seems to differ between radials and bias-ply tires, is also where the gap(s) occurred that allowed air to escape. See the "Test Results" sidebar for the specifics on the tire-and-wheel combinations we tried and what happened.
Rim protectors that are situated...
Rim protectors that are situated away from the edge of the rim, such as on this Cooper STT, can still help prevent rubbing the rim against street or trail objects, but really can't be responsible for trapping any dirt between the tire and wheel since the dirt should fall back out just as easily as with a regular, non-rim guard tire. Also, this placement doesn't affect the ability of the tire to hold air or prevent use of clip-on balancing weights.
It's obvious that rim guards placed close to the rim play a role, but the inside shape of the rim itself appears to be at least equally important. Consequently, bead-locked wheels can be just as susceptible to losing inner beads as are regular rims.
We assumed that the outside bead would let go first, for two reasons; it's hard to put much outward side load on the inner bead during regular four-wheeling, and it's usually the outer bead that ends up separated from the rim when a tire does lose air. Now we suspect that it's generally the inner bead that lets the air out, which then causes the outer bead to come loose if you don't stop immediately. Every tire and wheel we tested (that let any air out) acted that way. Also, it would explain why our old Centerlines, having virtually nothing for an outer safety bead, but with a nice, sharp machined groove on the inner bead, retains the tires extremely well. This goes against everything we believed, but based on our observations, it may mean that we should run those bead locks on the inside instead, where the cops wouldn't see them.