I found a video showing how to change out the bearing insert in the flange housings that Egbert suggested. Now I see how it all goes together and can transfers axial loads. It all makes sense now.
https://www.youtube.com/watch?v=1jG52W5U_Cs
Here's a short video with a fellow describing a 4-bolt version of these bearings.
https://www.youtube.com/watch?v=9rLP-p4R3pg
Of note is the detail of the self aligning nature of the bearings (at about the 40 second mark). This makes these bearings great for an installation where the mounting is not properly square to the shaft - within 3 or 5 degrees I think.
On a setup with a toothed belt reduction gear like I am planning (Which is equivalent to Thunderstruck in configuration) I think this may not work. If the shaft goes off axis as the bearing allows it to do it will make the driven toothed belt pulley tilt out of alignment with the motor pulley. That will be a problem for the belt. I think rigid alignment between the pulleys of the toothed belt reduction drive is important. Otherwise the belt may walk off the pulley or if it is a pulley with a lip at the rim it will start to chew up the belt at the edge. It seems like any configuration that lets the shaft with the toothed belt pulley go off axis is a problem. And having that freedom to go off axis means getting the reduction gear assembly aligned to the prop shaft while at the same time keeping the toothed belt gear pulleys in alignment will be more difficult. Seems to me the toothed belt pulleys need to be rigidly mounted which means the thrust bearing cannot have this spherical freedom or movement even if it is slight.
Having another point where the shaft is supported radially would solve that problem? Maybe the cutlass bearing provides that support. But that's a bit flexible. But that doesn't address the alignment difficulty to the prop shaft created by the freedom of the bearing to rotate spherically in the housing. Seems to me if you want a system that is tolerant of axial misalignment that has to happen at the coupling between the prop shaft and the drive output.
This is also interesting (at 1.5x speed) with regard to choosing bearings for reliability in axial and radial load:
https://www.youtube.com/watch?v=cU0wQHcYSSU
I am not an engineer or an engineering student but I learned a couple things from it.
Thanks,
Dan Pfeiffer
On 2021-03-31 6:55 pm, Dan Pfeiffer wrote:
For these very nice bearings that Egbert found, how are they retained in the housing? It looks like just a press fit with no shoulder on either side and I don't see any set screws on the housing (just on the bearing to hold to the shaft). So what transfers the axial load to the housing? The bearing may be rated for the load but the load has to be transferred to the housing somehow? I also see these housings in a 3-bolt of 4-bolt flange which might be a more robust choice. Also is stainless as pointless as that may be.
https://www.kugellager-express.de/all-flange-bearing-housing-units
Thanks,
Dan Pfeiffer
On Thursday, 1 April 2021, 01:13:16 GMT+8, Egbert Van popta <motorstunter@gmail.com> wrote:I looked at the specs of the ucfl bearing at the skf site and did the math. The axial load range is more than sufficient for my 12kw motor. Radial load 15kn of which 20% can be used as axial load force. It's not a huge margin, but as it is a very cheap and easy to change bearing I'm happy with that. On this site they weirdly enough don't mention the 20% axial load , but they give tips to increase the axial load capability.On the end of this written piece they link to plummer block housings with spherical bearings. Isn't this what you are looking for?Egbert_._,_._,_