Literally thousands of different BB spindles exist out there, their sizes settled on by everything from trial and error to finite element stress analysis to rote imitation. But TSDZ2 has the smallest diameter I've ever seen-- even on a child's bike-- by a whole millimeter. And then they cut a groove right where the stress is highest. It's not a small mistake. It's a small component used where everybody else in the world, for generations, agrees you need a bigger one. And where for generations we've known it should not be grooved. That's textbook negligence. If the manufacturer were from a respectable country, they would already have been sued out of existence.
The snap ring grooved Viscount BB spindle of the 1970s was a chronic problem, and it was over 13% bigger in diameter than the one we're discussing.
Hi Chalo,
I thought I'd chime in regarding the shaft / snap-ring groove issue because I'm an engineer. Simply, your analysis either makes some assumptions or is incomplete, however you are PROBABLY correct. Here's what an engineer without full information would say:
The shaft diameter could well be adequate and even have an additional margin if, and this is an important if, it is made of an appropriate material that is processed in the right ways. It's
not ONLY metallurgy but also whether cast, forged, extruded, rolled, etc. In this instance, rolled could be an effective choice but the starting materials need to be fairly free of any debris ("impurities"). Forging would, by far, be the best choice here. And if you're silly enough to insist on a stupid snap ring groove, forge the damned thing in with plenty of radius! It could work fine with no breaks.
I am well informed about rotational power transmission for early Porsche and VWs. In short, the original requirement for the input shaft (set in 1947) that takes power from the engine and transmits it to the transmission was for 20 PS BHP (25 SAE) with a maximum RPM of about 4200 or thereabouts. Power kept being increased and the original design was inadequate so the materials and manufacturing processes were continually improved leading up to a full order of magnitude increase in power still being handled by the same fundamental diameter.
The early ones did not but later units did have a snap ring groove used to retain the inner race for a replaceable roller bearing, and the inside diameter of that snap ring is ~23 mm and easily handles around 150 bhp with plenty (but unknown to me) margin to spare.
All that said, it's a fairly safe bet Tongsheng isn't spending more than the minimum they can get away with, and, just eyeballing one, looks to me that there's no care taken to have a good radius on the snap ring groove bottom, but I could be mistaken about that. A key factor in their decision making might very well be complete or at least substantial ignorance about the relative mass differential between "Americans" and their native population. What's adequate but not robust for a typical Chinese citizen is surely inadequate for a typical citizen of the USA.
As for product liability, your remarks are spot on, however I think the complaint was based on the non-specific nature of the comment rather than the targeted criticism of your later explanation. As I got _burned_ into me at a young age, it's often not what you say but how you say it!
...On another topic, I'd love it if Tongsheng got its proverbial -blank- together; the design has a lot of potential and if that torque sensor is an issue then make a bigger one! Duh! But I haven't paid a lot of attention; maybe they're now producing bigger / better units now? However, the firmware issue(s) that I've seen some comments about give me pause.
RTIII
I hope mine lasts as long, but I am fine with it if it doesn't as I'm aware I'm not ( always ) using it for it's intended use case. 
