I just returned from a vacation on our Hanse 400, which also has a 3JH4E with SD50 saildrive. Our first day out we noticed the engine sounded different and I went below to find the engine was hopping around and both mounts had failed. I limped into port and was lucky to find a mechanic to weld on some replacement studs to hold the engine until I got some replacements a few days later.
The failure was very similar to what was described by Giulietta in the original post, except both studs had failed immediately below the nut that supports the engine (not between the nut and the engine tab). I've tried unsuccessfully to upload some photos (800x600, <50kB). Our engine was also delivered in 2006 and has 900 hours on it.
I showed the photos to a mechanical engineer who had the same comments that "Bob" had earlier in this thread. He too was surprised that the stud itself was used to support the vibrating load instead of some kind of spacer. He explained that threaded rods are subject to failure fatigue if they are subjected to flexing as these ones would be. The threads weaken the rod because the strain isn't evenly distributed across the rod but is greater on the "inside" surface of the thread and less on the "outside" surface. (similar to the effect of a small nick in a wire that will cause it to fail by focussing the strain in one spot).
This explanation made very good sense to me and it was very interesting to read the corroborating explanation and same suggested solution from Bob. The diagram in the original post doesn't quite make as much sense to me because the rod wouldn't flex so much at the interface of the nut and the engine tab (could it be that the drawing is not quite right?)
If the engine was bolted directly to the mount or if a spacer was placed between the engine and mount then the threaded stud would only function as a fastener and not a standoff and would not be subject to this failure mode.
I'll be sending some photos to Yanmar. One would hope that they have seen enough of this failure to realize there is a flaw to be addressed - perhaps with a field upgrade to add a standoff between the mount and the engine tab. I improvised one with my replacement mounts because otherwise I have no reason to trust that this won't happen again. I found some oversized stainless nuts which I placed between the top of the mount and the lowest nut (the nut that holds the stud on the mount), with a washer in-between. This provides the same spacing as the original installation but essentially creates a "stack" between the engine and the mount, with the stud used only to hold it together. (again, I have a photo but can't upload it).
I was very happy to see Bob making the same suggestion so hopefully I've made an improvement. I think he is bang on with his comment about using threaded fasteners for structural members.
Thanks very much to SailNet, and to Giulietta, Bob and Quiksilver and others for sharing their information and insights.