Pivot Bearing Press Fit Tolerances

Question

I have a Privateer 161 mountain bike. Ever since I bought it, most suspension pivot bearings run roughly when pressed into the frame. After the first 3 months I had the bike, I removed some of the bearings to inspect and found that the bearings ran fine when not in the frame.

2 years later, I found that the main lower pivot bearings were completely seized and would not rotate. I removed some of the bearing seals carefully, while the bearings were still in the frame, and found that there was no grease in the bearing whatsoever. After some coaxing and spraying a lot of wd40 in them and re-greasing, I still couldn't get them to budge. So, I bought a new bearing kit and some bearing extractors and bearing presses with appropriate drifts.

These lower main pivot bearings were extremely difficult to remove and the new bearings were also difficult to press in, even after cooling in the freezer. Upon removal, I did some measuring of the housing on both drive and non-drive side and also of the old bearing and new bearing.

Old bearing:

New Bearing:

Housing Non-Drive Side:

Housing Drive Side:

So, I know that bearings are often interference fit and in this case, we have a few things:

1. Interference fit of old bearing - 32.02mm - 31.85mm = 0.17mm = 170 microns/micrometers
2. Interference fit of new bearing - 32.01mm - 31.85mm = 0.16mm = 160 microns/micrometers

My question is this: Is this fit within industry standard for tolerances, or is it too tight of an interference?

I just can't imagine that the bearings should run so roughly when in the frame or be so difficult to install/remove with appropriate presses/bearing extractors. I've contacted Privateer and they tell me this is within tolerance, but after some research I am getting conflicting information. Also, they never told me what the tolerances were, just that it is within tolerance.

Answer 1

Whether there's a real industry standard here is debatable. What's almost universally standard is that people who answer emails at bike companies for a living are mostly there to deflect when given the oppurtunity. That is even more true now that there's never enough supply and everything is always crazy. By bringing the question of numerical tolerances along a theoretical 2d plane into the conversation, you've given them more of that oppurtunity than necessary. The real problem is the bearings on your bike aren't working.

The .17mm you're coming up with as the highest measurable amount of interference tells part of the story, but not all of it. Bearing fit tolerance classes also spec a minimum amount of interference that will be part of whatever interference class they're trying to hit. The other thing is that there can be all sorts of alignment/machining problems that can cause the kind of issues you're having that won't make themselves apparent by measuring point to point with a caliper. That said, yes it seems like a very high number.

You have new Enduro MAX bearings, which are among the best available, plus appropriate tools. I think the best next step here is put one in and see how it feels. It should still feel smooth when pressed in, and not significantly different from how it feels in your hand. Clean and grease the bore to get as accurate a read as possible for how hard it's going in.

If it feels rough and notchy once it's in, yes there is a problem regardless of what the company says. In practice there is some gray area here because a lot of bikes are less than stellar in this regard but still work, but there are also ones where there are legitimate issues, and it sounds like you could be in that camp. There are some different things you could consider at that point. The bores could be re-machined, but not cheaply or simply by any means if you're also going to do it in a way where everything is referenced properly to everything else for alignment. You could go nuts and just use sandpaper to rough out a little more clearance by hand, then fit the bearings in with retaining compound and let it cure while everything is assembled, notion being that by doing it that way the misalignment you introduce probably gets magically averaged out by the retaining compound taking up the space. That could go wrong but it's also very possible for the results to be better than what you have.

Ultimately it's hard to get to real conclusions with these issues because it takes advanced tools and skills to measure the alignment of the bores to each other and to the frame center plane, which is a common and major potential contributor to issues if there are any. It's not really done in practice. Given that, there's no crime in turning to more subjective judgment, i.e. does your newly installed bearing feel like garbage or not, does the linkage move smoothly when cycled by hand, etc. If you present your case to the warranty dept by simply asking if X amount of measured interference is correct, it threatens to oversimplify the conversation in their favor.