Our conversation has consumed me to the point that I actually had a dream about our correspondences' (correspondii? Hmm....) last night. I'll probably get conflicting responses from the "arm-chair engineers" out there, but here I go anyway:I understand what you are putting yours thru - but a big drop followed by a hard turn is different than sliding under power through rough chopped out track. We can run the dunes/ woods / huge whoops / big jumps all weekend w/o any issue - but 12 laps on the track and they get loose. Its side loading simular to a torque spike that seems to cause it. 99% if the Mavs are not driven in this manner and will never have an issue. But those of us that race this terrain are having it. The hubs in my Rhino & Teryx had a tighter spline fit - so maybe its a clearance issue with the hubs.
You're right about the excessive clearance(s) between the male/female spline interface on the stock Maverick stubaxles, and the cast iron, splined drive flange that mates to it . Loosey-goosey to say the least, resulting in the condition suggested by 'AReed' regarding the "hub pivoting on the splines". In my dream, our desert Mav is competing against narrower-tracked vehicles in some fictitious and foggy arena. The visualization of that image instantly clarified the ROOT cause of your dilemma. Leverage. Specifically, the moment of inertia, and how weight transfer is inversely influenced as the distance between a force and the center of mass from which it rolls or rotates increases (inertia = mass times the distance of rotation, or the 'lever arm', squared). Although the g forces generated while cornering may be similar for both cars (the co-efficient of friction for rubber is pretty constant, so theoretically, your style of racing can impose no significant increase in lateral g-forces over ours), the amount of weight transferred to the out side tire due to stock geometry is far greater than that of a car with a modified/widened geometry. It's my assumption that, in order to be agile in the corners, the track width on your Maverick is far narrower than the 77" that I built into ours. Not to bore you with the math, but by widening our track width 6 1/2" (per side) over stock, the load (or weight transferred) on the outside wheel while cornering, is met with roughly 40% more resistance than that of a stock Maverick (not considering the increased distance between roll centers and the center of gravity of a narrower track verses ours as a result of the excessive operating angles of the un-loaded inside control arms while cornering, which further harms your cause).
Although the simple "side-loading"explanation you offered may be suffice for the lay-mans, it did not satisfy my overly analytical, obsessive and ANAL mind ! Actually, side lolading implies that the threads of the stub axle are in either 'tension' or 'compression'. You're obviosly not pulling off the inside wheel due to tension, nor shearing off the outside wheel due to compreassion. Rarely does a joint ever "push" apart. It is failing due to excessive lateral or bending loads applied as the body rolls and the weight transfers to the outside tire. This condition is aggravated by the axle torque, as well as poor manufacturing tolerances of the splined interface. While beefier , higher alloy stub axles may more effectively resist these forces, the proper solution lies in the geometry of the suspension, and its effect on the roll center locations at the front and rear of the vehicle (because you have unequal length a-arms instead of a rear trailing arm geometry of say a Polaris, your roll centers are figured differently, and conversely, move differently as the suspension travels). Springing the vehicle so the a-arms are as level to the ground as possible when entering a corner will move your roll center height closer to the ground level, rather than some dimension below it, as the OEM settings encourage. This will help reduce the weight transfer to the outside tire. The double edged sword here is softening up the effect of the spring rate during weight transfer during cornering, while leaving enough spring rate not to bottom hard off of the jumps. Even though our car weighs more than 2,000 lbs, we are able to run spring rates lower than that of cars hundreds of pounds lighter than ours. This is a direct result of the low motion ratio I designed into the mounting of our coil overs. Add to that the damping and valving diversity our Fox 2.5" internal bypass truck shocks offer, and BINGO, a well balanced race car that doesn't fail before its time! We are still fine tuning the compression valving in the last 25% of our wheel travel, and are pretty close to what I feel is the optimum setup for our car.
In conclusion, I appreciate your patient respones and input you guys have offered. It stimulated me to apply the fundamental principles I've learned over years of researching proper road racing suspension and chassis design, and how it applies to our world of long travel off-road racing geometry, to understand exactly what is happening here. It is my philosophy of design that tells me, "If I can not quantify a problem, I can gain no understanding on how to best solve it". It is why it took over 2,000 hours of time for me to design and build our current race Maverick, WITHOUT the use of "CAD" design software. But then again, we DNF'd 5 of the 6 races we entered last year, so how much clout can I really have? Ha Heee! S.Y.A.T.C. (home-made for See You At The Checkers!)