FRC Team 229 used a similar type of tank tread bought from an automation supply company in 2012. The belts were also steel lined and had a similar breaking strength. The team tore through plenty of them and at $220 a piece it was a big problem for our wallets. What we later determined after busting through at least 3-4 sets of belts that tensioning is a key part of belt drive. Due to the nature of FRC you are constantly altering directions and going from tension to slack on the drive belts. If a belt slips off your drive cog or becomes miss-aligned it has high potential to form a crack at the edge of the material. As soon as that happens there is a point of imbalanced stress concentration on the cross section. This game in particular has plenty of points (ie steel objects that could fracture a belt). My guess is that your issue initiated as a small hairline fracture in the material, then after time on the field alternating between tension and slack then applying a peak stall the belt reached its max.
Using these belts I would make sure to keep an eye out for any small sources of cracks in the flexible liner. Once this happens you may need to replace the belt to be safe. The other thing I can recommend from my 2012 experience is that you need to make sure you have a way to tension the belts (statically, or passively in both directions) and not over tension the belts i.e. applying too much torque to your turnbuckle. You may be surprised how much force you can put on those belts through tensioning.
I’ve got a drive base in our shop from 2010 with the original belts still on it. We use to practice having a defender harassing the current season’s robot. Brecoflex belts are indestructible ONLY if they’re very carefully designed around.
That drive base also has no adjustable tensioner.
EDIT: Here’s our guidelines for a tank track system:
- Sideloading is the quickest way to kill a belt system not built for sideloads. All pulleys will need their center groove accurately machined, and wherever the belt is not contacting the ground some sort of rigid plane should be present to stop the belt from tracking too far out of center.
- The belt loop length needs to be very accurately machined. The system we were using had 0.25" 6061 plates for each side of each track module. This didn’t allow for the possibility for either too loose or too tight a loop. Multiple components retaining your track hardware will allow for misalignment, and each time you split the system into multiple parts you add more possibility for either lengthening or tightening the belt loop.
- A tight belt is just as responsible for failure as a loose belt. Overtensioning a steel belted system will make it fatigue and tear. Once the plastic and tread around a seam has torn free, all the fatigue and stress moves to the steel belt for that section, which will fail in short time. On our 2010 system each belt has maybe .01" of vertical flex when the belts were new. Turnbuckle style belt tensioners lead to ‘crank-er-down’ practices which will overstress the belts and lead to premature failure.