dren wrote: ↑16 Aug 2017, 02:20
The translated article insinuated it was dynamic balancing.
That can't be right. Is there really anyone who doesn't use dynamic balancing? Especially for a part that rotates at such a high speed dynamic balancing should be standard practice. I don't think there has ever been any rotating part that hasn't been dynamically balanced where I worked (mainly heavy engineering in the paper industry). Even every wheel on road cars gets dynamically balanced.
What I can imagine is that they maybe have been balancing the assembly at 2000rpm (or similar) and therefore wouldn't be accurate enough for operation at 125000rpm. And now they are balancing at 20000rpm+.
That also might explain that some MGU-H units seemed to work (maybe because they got the balancing good enough) and a lot have not.