Replaced Rear Wheel Bearings - Problem / Question

[Oiluj]

I replaced the rear wheel bearings but am not satisfied with the results. I'm guessing it might be operator error...

Before assembly I did a visual check on the bearing mount surfaces and verified that the axle bearing was well seated on the shaft prior to assembly.

I tightened axle nuts down to spec, (200 ft-lbs), but both axles feel a bit "stiff" and do not rotate that freely.

My manual gives a axle-nut torque spec of 181 to 239 ft-lbs, but also gives a "preload" spec of 3.9 in-lb. I'm assuming this 3.9 in-lb is the maximum rotational resistance?

Torque required to (slowly) rotate the hub is at the upper limit of about 4 in-lbs, but feels too stiff to me. I packed the bearings with "black" disc brake wheel bearing grease, (what I already had in the garage).

End play is almost non-existent, but if I loosen up the axle nut and lightly tap the axle end to remove a bit of bearing pre-load, end play increases to about 0.015", which is outside the 0.005" spec. At this end play, rotational resistance is about 2 in-lbs. Thing is, I can't seem to get a reasonable end-play less than 0.015" without it becoming "stiff". This seems to be it's "happy place".

Any thoughts?

[JimmyZ]

The internal clearances on the ball bearings is more critical than the torque on the bolt. The nut gets staked and serves a similar function to it's cotter pinned brothers on the front axles. (Tapered rollers I know) Both serve to maintain proper "preload". If you visit the Timken or SKF bearing sites you'll find some info regarding the effects of improper preload. Internal clearances in a ball bearing are very important. The fish scale method of determining preload is pretty good as long as your scale is good. Things such as the tightness/looseness of the axle or bores also change the int clearnaces. Bearing tolerances are also a factor. Preload on a single row bearing isn't what you want proper int clearance is.

You've tried rotating the axles in both directions a few times right? Even though this is something you would do for tapered rollers it can't hurt on the ball bearings. Perhaps there's a bit of sand in there giving your fish scale a hard time?

If I recall isn't there a spacer involved? It's possible that the spacer is under it's tolerances.

Just throwing some ideas out there.

Jim

Edited October 15, 2008 by JimmyZ

[Oiluj]

I appreciate the ideas, and a good point about axial pre-load like on a timken bearing. The rear bearings appear to be "offset" ball bearings. Also, I did check using calipers that the new bearings were the same thickness & OD as the old bearings.

I did rotate both directions quite a bit. The assembly was in a relatively clean environment, (my garage), so no sand or grit to worry about.

I'm also thinking it might be possible that the aftermarket bearings came from the factory with more "play" than the original bearings. Lastly, since I'm not yet satisfied, I haven't "peened" the nuts, so I can R&R the bearings if I have to...

[JimmyZ]

I'm also thinking it might be possible that the aftermarket bearings came from the factory with more "play" than the original bearings. Lastly, since I'm not yet satisfied, I haven't "peened" the nuts, so I can R&R the bearings if I have to...

I'm thinking that you should see if there are any specs for the spacer/crush sleeve. It's possible that it has worn or some other part ID/OD is out of tolerance. Was there any evidence of seizure with the old bearings? (Or impact damage/deformation from an accident?) Calipers are just a quick sub for a good set of mikes. You're looking for tolerances below .001". Sure, you can use a standard to get a set of calipers pretty close but to reliably read .0001" increments a mike is needed. I think a -0 to +.0005 fit ID/OD would be a good tolerance for a ball bearing in this app. (+.0002" would be ideal)

It might even be that you got a bearing which was made "too well". The internal tolerances might be so tight that fitting it to a shaft or bore might cause it to bind.

2c

Jim

[Carl Beck]

The fish scale method of determining preload is pretty good as long as your scale is good.

Was a fish scale used to take the measurement in the first place? Or was some other method used?

Interesting thread... but the use of aftermarket parts may be what is throwing the Nissan Spec.'s for Nissan supplied parts - out of wack.

Were Nissan supplied parts NLA?

FWIW,

Carl B.

[Oiluj]

Arne,

Don't know if the Nissan parts are still available, but get many of my auto parts from a NAPA store just 4 blocks away from my house.

At this point I'm thinking that 0.015" end play is probably better than high rotational resistance. Figure I'll try running them for a 100 miles or so to "break them in" and then see if I can re-adjust them to reduce the end play.

[John Coffey]

My manual gives a axle-nut torque spec of 181 to 239 ft-lbs, but also gives a "preload" spec of 3.9 in-lb. I'm assuming this 3.9 in-lb is the maximum rotational resistance?

Torque required to (slowly) rotate the hub is at the upper limit of about 4 in-lbs, but feels too stiff to me. I packed the bearings with "black" disc brake wheel bearing grease, (what I already had in the garage).

End play is almost non-existent, but if I loosen up the axle nut and lightly tap the axle end to remove a bit of bearing pre-load, end play increases to about 0.015", which is outside the 0.005" spec. At this end play, rotational resistance is about 2 in-lbs. Thing is, I can't seem to get a reasonable end-play less than 0.015" without it becoming "stiff". This seems to be it's "happy place".

Ignore the torque numbers (get rid of the torque wrench) and focus on end play and pre-load spec. End play is more important then pre-load spec so if you can get under .005 with a pre-load of 4 then I would go with that. You can also source a copper crush washer the same ID/OD as the inside end of the companion flange and use that to get to the specs.

EDIT: If you're going to tear the assembly down again, clean out that black grease and run Redline Synthetic EP grease.

Edited October 15, 2008 by John Coffey

[JimmyZ]

Was a fish scale used to take the measurement in the first place? Or was some other method used?

Carl B.

I remember somewhere seeing a scale being used to measure the force required to rotate the axle. You simply pulled one of the wheel studs with a fancy fish scale.

So I'm getting off my bum and cracking open an FSM and Haynes manual. According to the Haynes manual the "tug" should be 25-30 ounces and endfloat should be under .0057". (pg 124)

The '73 FSM from Carfiche doesn't have the "fish scale" pics which I could swear were in the '72 FSM. My '72 FSM was destroyed in a move. Perhaps I'm remembering a Clymer manual I once had. There seemed to be more detail in earlier FSM's IIRC.

The whole idea of using the fish scale is to provide a means of checking for interference in the bearing's internal clearances. (Kind of like double checking your math) Single row ball bearings should have NO preload. They should be located precisely to curb shock loads or misalignment.

What I think has happened here is that Julio's spacer/distance piece is out of tolerance. (Or they got swapped from one side to the other) According to the 73 FSM they used different spacers for different manufacturing tolerances. The castings had an A,B or C stamp on them which coincided with a different spacer.

If you wanted to be really nitpicky You could take the entire thing apart and measure the distance between the bearing bores and compare that to the spacer measurement. The spacer measurement would need to be 0"- to - .005 of that measurement assuming the bearings were flat across their faces. (Preferably 0")

Using a copper washer would be a poor choice for the long haul. It would compress and eventually give excessive slop. Turning a new spacer on a lathe would be best.

2c

FWIW etc

Jim

PS This is what happens when Machinery's Handbook is your bedside reader.

Edited October 16, 2008 by JimmyZ

[Carl Beck]

I tightened axle nuts down to spec, (200 ft-lbs), but both axles feel a bit "stiff" and do not rotate that freely.

My manual gives a axle-nut torque spec of 181 to 239 ft-lbs, but also gives a "preload" spec of 3.9 in-lb. I'm assuming this 3.9 in-lb is the maximum rotational resistance?

Torque required to (slowly) rotate the hub is at the upper limit of about 4 in-lbs, but feels too stiff to me.

AND

I remember somewhere seeing a scale being used to measure the force required to rotate the axle. You simply pulled one of the wheel studs with a fancy fish scale.

So I'm getting off my bum and cracking open an FSM and Haynes manual. According to the Haynes manual the "tug" should be 25-30 ounces and endfloat should be under .0057". (pg 124)

The '73 FSM from Carfiche doesn't have the "fish scale" pics which I could swear were in the '72 FSM. My '72 FSM was destroyed in a move. Perhaps I'm remembering a Clymer manual I once had. There seemed to be more detail in earlier FSM's IIRC.

The whole idea of using the fish scale is to provide a means of checking for interference in the bearing's internal clearances. (Kind of like double checking your math) Single row ball bearings should have NO preload. They should be located precisely to curb shock loads or misalignment.

PS This is what happens when Machinery's Handbook is your bedside reader.

Hi Jim:

Yes, I had that same picture of the FishScale in mind when I ask if Oiluj had used that method. From his words I couldn't tell if that was the case or if he was using a torque wrench (on a lug nut on a stud) - as he used only the term "torque" to slowly rotate the hub -in that sentence.

At any rate.. the FishScale is used in the FSM on the Front Wheel Bearing.. now that I check my 72 FSM... for both new bearings and re-using the present one...

PS - This is what happens when 85% of your memory is "graphic" rather than textual.

FWIW,

Carl B.

[Oiluj]

I didn't use a fish scale, but did the same thing with a known weight hanginging from a stud. That's how I got the 4-5 in-lbs torque, (which semed high to me).

I did figure the problem out. It's wasn't the bearings, but the seal on the inside of the hub. When I removed the inside rear hub, the rotational force was much lower.

Turns out the new grease seal had a rough edge with molding flash, and the flash was creating a lot of drag. I gently ran some 1200 grit sandpaper to remove the flash, followed by some polish compound to clean-up the surface finish of the seal edge. (Learned this trick from a heavy equipment field mechanic).

This appears to have solved the problem. Once that was done, I was able to adjust the axle nut to eliminate end-play and still have acceptable rotation force.

I knew it had to be something simple, (or operator error). Got one hub done last night and plan to do the other tonite. Whew!

[doradox]

Ignore the torque numbers (get rid of the torque wrench) and focus on end play and pre-load spec. End play is more important then pre-load spec so if you can get under .005 with a pre-load of 4 then I would go with that. You can also source a copper crush washer the same ID/OD as the inside end of the companion flange and use that to get to the specs.

EDIT: If you're going to tear the assembly down again, clean out that black grease and run Redline Synthetic EP grease.

That 200 lb*ft of torque is a key part of the strength and fatigue life of the rear hub assembly. Backing off the nut to get the correct end play is not really a good idea.

Steve

[beandip]

Oiluj . I recently did the exact bearing replacement and using the same grease. I used OEM bearings. Torqued to 200lbs the same as you. I found the same stiffness you experienced also. I have driven the Z now about 300+ miles and the other day I lifted the rear and checked the rolling resistance and all is vary good. I recommend that you set the torque just as the manual says and drive the car. Check the wheel for heat once in a while for the first 100 miles . I think you will find as I did , all is good .

My 2 cts Gary