Changing a Rear Axle Shaft Bearing on a Jeep CJ AMC 20 Rear Axle

[waxman]

I feel like I am completely missing something here. I am attempting to replace the bearings on my 1985 Jeep CJ7 with AMC 20. As you'll see in the photo, I'm down to removing the inner race("B" in image). But I am looking at the axle shaft and tit has this flare("A" in image) sticking out which as far as I can see, there is no way I could possible get the new bearing over this flare out.

All of the forums and tutorials say I should be able to slide the component down the shaft for pressing, but I just don't see how this would be possible. It seems to me I have to remove the axle shaft from the hub ("C" in image)

Image---https://drive.google.com/file/d/1MyfZGH_B5jCT4d-_tzVfsTVR7ulv6LWu/view?usp=sharing

Can someone please set me straight?

Thanks!

[Jon]

Hi Waxman, the bearing comes off on the splined side after you remove the hub. A puller would be helpful. I hope this helps.

[waxman]

Thanks Jon.

 

That makes sense. After further research and forum reads, I think I may switch to the one piece Morris set. Seems better all the way around.

 

Thanks again!

[Moses Ludel]

Thanks for jumping in, Jon!...Yes, this is a "two-piece" axle shaft:  the hub separates from the axle shaft before removing the bearing.  Jon's suggestion is correct.  I use an OTC puller;  Harbor Freight has an alternative/facsimile.  Here's a link to general information about the use of a puller.  When force is high, I add the OTC 6574-1 accessory plate to prevent bending wheel studs.  Wear protective eye gear, this force can be high, and hammering on a puller wing increases risk of metal shards.  Watch both videos;  the second video covers how to protect the axle shaft threads.   I make a threaded protector sleeve to prevent flaring out the axle shaft end.  Avoid damaging threads or mushrooming the cotter pin hole:

https://www.4wdmechanix.com/HD-Video-Tool-How-to-Using-the-OTC-7394-Hub-Puller?r=1

Importantly, if you do reuse the axle shaft and hub, mark the hub/axle shaft alignment before removing the hub.  (Do not mix hubs side-to-side.)  AMC called for replacing the hub with a new hub if the hub is removed from the axle shaft.  If parts are okay and this is a first time hub removal since installed at the factory, many of us elect to reinstall the hub.  Check the inner hub for signs of double spline cuts or shifting on the splines.  Replace the hub if it cannot be secured properly.

Align the hub as originally installed.  These hubs come "bare" without serrations or "splines".  This means that the axle shaft actually cuts splines into a new hub during the installation.  The axle shaft has "teeth" that serve as splines when the hub is tightened properly.

Equally important is the tightening torque for the axle shaft nut.  My reference is from the 1985 FSM for your CJ Jeep.  This is a minimum of 250 lb-ft!  If the cotter pin slots do not align at this point, tighten past this point to the first slots that align.  Do not back the nut off to insert a cotter pin.  Install a new cotter pin.

If you fail to reach this torque, the hub will loosen.  In this design, the woodruff key is simply for alignment and will not keep the hub from spinning.

The AMC 20 two-piece axle shafts have gotten a bum rap.  Many Jeep 4x4s, especially with oversized tires, have spun the hub on the axle shaft.  In most instances, however, the hub has been removed and reinstalled without adequate torque on the nut—or force over time has loosened the axle shaft/hub assembly.  When the nut is tight enough, this is a secure arrangement.  I use 33" as the limit for this hub design, 35" or larger tires increase the risk of hub spinning.  With oversized tires or off-road rock crawling, I verify the nut torque periodically.

To emphasize how to tighten the nut, be aware that most air guns will have difficulty reaching the required torque setting.  The shaft's taper force and hard steel can create a false reading.  Even if your air gun seems adequate, always verify the torque on the axle shaft nut with a torque wrench that will reach this limit. 

Just to bring torque into spec or align the cotter pin slots, I have tightened these axle shaft nuts with a 3/4-inch square breaker bar and impact grade socket, using a floor jack handle to extend leverage.  In most cases, I find that the typical 3/4-inch drive torque wrench has enough length to get the job done or at least verify the torque setting.  When you don't want to break the bank on a 3/4-inch drive torque wrench, this one works, and I use it:  https://www.harborfreight.com/34-in-Drive-Click-Type-Torque-Wrench-63883.html.

I address the AMC 20 axle and axle shafts in my Jeep CJ Rebuilder's Manual:  1972-86 (Bentley Publishers and available through Advance Adapters, Amazon and elsewhere).  The aftermarket single-piece axle shafts are certainly an option, which I also discuss in my book.

Moses

 

 

[waxman]

Thanks Moses for the in-depth information. Very much appreciated.

I actually already DO own your book. It's awesome! I am using it for this project, but admittedly, I must have misread or just missed this topic. I should have slowed down and thoroughly researched before posting here. But doing so lead me to a great forum here, so I'm happy for that discovery.

 

Thanks again!

Gary

[Moses Ludel]

You're very welcome, Gary...Glad we could air this topic, it's valuable to others.  Pleased to have you participating at the forums.  Looking forward to your posts!

Stay healthy and keep your family safe...

Moses

P.S.:  For those who have the book, I cover AMC 20 axle shaft work in depth from Page 298-308.  There is a whole section (Pages 274-308) of the book devoted to the AMC 20 axle rebuild, including axle shaft removal, axle disassembly, differential/ring-and-pinion rebuilding and installing an ARB Air Locker upgrade.

[waxman]

 @Moses Ludel

I opted to go with  the "TenFactory" one-piece axle kit. According to the instructions, the bearing race should protrude from the axle housing with a .020 - .060  tolerance. If it protrudes further than  .060, then the spacer should be filed down until the tolerance is within the limit.  

I've attached a photo that shows that my set-up, where the race only reaches to the inner-bevel of the housing, which seems wrong to me. Maybe they sent the wrong sized spacer? The fact that the race doesn't extrude beyond the housing at all, makes me think something is not right.

Another question...What is considered "binding"? In attempt to troubleshoot this original issue, I went ahead and assembled the axle/hub on to the housing and bolted it down to spec. The instructions say there should be no "binding". When I rotate the hub assembly, there is a decent amount of resistance, and I do here a little noise as I turn. I just don't know if that is normal, or if resistance and noise is considered "binding", which means too much pressure on the bearing? Looking through tutorials and such, it is suggested that there should be "some resistance". This being my first attempt at this, I'm not sure what is considered proper resistance.

 

Thanks again for all of your help!!

[Moses Ludel]

Gary, your concern is justified.  What you strive for here is correct axle shaft end play.  This cannot be determined without all parts in place at each side of the axle.  Do not be concerned about the kit's spacer thickness unless you cannot set the correct axle shaft end play.  If the bearings remain too loose (too much end play) with no shims, you have an issue.  Look at the bearings on the axle shafts, and you can see that they protrude;  this means the bearings will extend into the bearing races.  If the bearing cages are not rubbing on the axle housing edge or chamfer, and if you can achieve the correct end play, you have it.

The end play is across the axle:  between one axle shaft, the spacer block at the center of the differential and the other axle shaft.  Think of this as clearance between the axle shaft inner ends and the spacer block.  If centered properly, the spacer block slides laterally.  End play gets read at the shimmed side of the axle housing.  

Read the section in my book relating to the axle shaft end play adjustment.  The end play is checked at one side.  The correct side is set up without the thin clearance shims.  You measure the clearance/end play at the opposite side.  My book outlines which side has no shims during setup and which side to shim and measure the end play.  Setting up for testing the end play requires installation of the seal plates, axle shafts, thin adjuster shims at the correct side and the brake backing plate(s).  Fasteners are tightened.  All of this must be in place and secure before checking the end play.

In the final setup, the axle shafts will run with a bearing clearance (a specific end play in thousandths of an inch).  Since the axle shafts float laterally, measuring end play at the one side is actually reading the clearance for both bearings.  These axle shaft bearings must not "bind" or run with "preload".  You will see my use of a dial indicator to check the axle shaft end play. 

Note:  Before any adjustments, make sure the bearings are seated on each axle shaft.  Most kits come with the bearings already pressed into place.

Set up both sides according to my book's steps (or a factory service manual/FSM for those who do not have my book).  Follow the guideline for which side to stack the thin adjuster/clearance shims.  Regardless of the instructions that came with the kit, you want correct bearing end play.  Too tight will cause bearing "binding" or hot running, failed bearings.  Too loose means excessive axle shaft run-out, bearing damage and brake drum-to-shoe drag. 

The only way to test this is with both axle shafts in place, hardware secured and measuring the axle shaft end play at the correct side.  Pushing the axle shaft inward will push the opposite axle shaft outward.  You then pull the test axle shaft outward to get your dial indicator reading.  You can improvise with a feeler gauge and straight edge, but a dial indicator is recommended and precise.  

Use the thin adjuster shims at the correct side to set bearing end play and running clearance.  You can then disassemble the axle shaft(s) and use sealant on the thin shims before final assembly and hardware tightening.  Gasgacinch (same as Edelbrock's Gasgacinch) works well between thin shims, serving as an oil or moisture barrier.  If you torque while the sealant is pliant, Gasgacinch will not distort the torque setting or shimming.  Make a final check of axle shaft end play after the hardware has been torqued properly to specification.

Read the manual, if you still are uncertain or have further questions, just ask...

Moses

[waxman]

Thanks @Moses Ludel,

I have read the section, in your awesome manual, multiple times but am a bit confused now.

When I was speaking to the axle kit manufacturer tech guy, he said that shims are no longer needed with the one-piece. He then said that there should be no play at all. He continued to say that if there is play, then the bearings are not fully seated in the race, and that can actually cause problems.

I assembled everything and there currently is no play.

I trust your vast knowledge on this and feel like I am beating a dead horse now, especially after you have taken so much time to help me.

I am just wondering why the instructions/advice I am getting on these new one piece axles, do not coincide with the instructions your book suggests. End play or no end play?

hmmmm.

[Moses Ludel]

Gary...I'm still here...All of my discussion has related to the stock OEM AMC bearing and axle shaft relationship.  This aftermarket axle design apparently flips the bearing around and uses a heavy retainer plate between the bearing and the axle flange.  Is that correct?  If so, each axle shaft bearing adjustment is set up independently.  There is no cross-axle adjustment of the bearing end play as I describe with the stock AMC two-piece axle shafts and their bearing orientations. 

AMC wants the bearings to have end play, which is noted in the FSM and my manual.  The range is 0.004"-0.008" with preferred 0.006", measured from one side.  (See Fig. 5-237 in the book.) This clearance/end play is split between the two axle shafts and bearings.  That's the OEM stock design.

Apparently, the way this one-piece axle shaft kit works, the outer retainer holds the bearing in a fixed location.  There is no cross-axle movement of the axle shafts, each axle bearing adjusts separately.  This is clear in the way the kit's bearing cup and cone face.  The shaft can only move inward to the point that the bearing fits against the fixed bearing cup. 

If no shims are used, the spacer ring provided will predetermine either the bearing clearance (end play) or a preload.  From what you shared, the only reference to bearing end play or preload is "no bind", which means the bearing can rotate without resistance.  Do the instructions make any reference to end play or preload?

Note:  The spacers provided with the kit set what the manufacturer considers the correct bearing end play or preload, whichever the manufacturer wants to achieve.  Reference to sanding off spacer ring material if there is binding indicates that they want the bearing cup to put less pressure against the bearing cone if there is binding.

You axle shaft supplier suggests that with its placement of the bearing on the shaft, and no shims in place, the axle shaft will seat at a position that provides the "right" bearing clearance (end play) or preload.  Preload is zero clearance (end play) plus some tightening beyond that point.  Did the tech share whether they want axle shaft end play, zero play or a preload?  That would be the question.  The only reference is to avoid binding, which is excessive preload.

Back to your original concern about the instructions calling for a bearing cup protrusion of 0.020"-0.060".  From your photo, you obviously do not have this cup protrusion.  Did you try to fit up the axle shafts without any shims to see what the shaft end play would be?  According to the instructions about cup protrusion, you should have lots of end play or clearance.  If you do not have that end play, is it possible that the 0.020"-0.060" figure reflects how much the cup fits into the bore and not how much it should stick out?  You have that kind of fit with the recessed cups.  With the spacers removed, the axle housing bores should have no wear on the shoulders where the spacers seat in the housing bore.

Post some photos here of your new axle shafts.  I need to see a side view of the bearing and outer bearing retainer plate.  We need to determine whether the axle shaft "floats" inward and outward—or does the bearing set at a specific, fixed point once the retainer plate gets tightened in place (without the shims).  I also want to see the housing bores where the kit's spacer sleeves fit.  If the bore seat(s) are worn away due to the original bearing cups rotating, the spacer sleeve will seat too far inward.  The result would be too much bearing clearance or end play.

Let's go from there.  The wild card is whether the axle shaft bearing is supposed to have end play like the OEM/AMC design, zero play or some degree of preload.  Using the OEM AMC guideline, each bearing (adjusted separately if that is the kit's design) would have approximately 0.003" end play.  The kit manufacturer would need to indicate what bearing clearance (end play), zero play or preload they expect to see. 

Moses 

[waxman]

Hi Moses,

You now see where my confusion was happening.

The tech guy specifically said that there should be no end-play. The instructions, like you noted, make no reference to end-play or pre-load. Just that it shouldn't bind.

As far as the bearing cup protrusion, this video most closely resembles my process, and the instructions for installation. https://www.youtube.com/watch?v=rmKC7FRf7-E

The bearing and retainer  get pressed down to a flanged ridge on the axle and it remains completely stationary. In the axle housing, the seal, spacer, and race are installed in the housing. Then you just insert the axle onto the shaft as far as it will go, which positions the bearing within the race. Tighten everything down to spec, and check for binding.

When you watch the video, you will see the part where he is measuring the spacer, as it protrudes beyond the axle housing face. This is where he measures, and files down the spacer, to obtain the correct tolerance. If you look at my image, the race does not protrude at all. This is why I am concerned.

I then talked to the tech at "TenFactory" and he said that I should be measuring the "bearing" protrusion and not the "race". (This differs from the attached video)So because I already had pressed the bearing onto the shaft, I couldn't use it. I grabbed a spare bearing and placed it in the race to measure. It measured to be within the acceptable tolerance range. So I went ahead and tightened everything down.

Now it is all assembled. It has no end-play". I can turn the hub with my hand, but it does take some effort.

Right now, I have two main questions I am trying to figure out:

Is it okay that there is no end-play?

What is considered "binding"?

On the binding question....Although the hub is a little tough to spin, it doesn't sound or feel like it is catching on anything. I just am not sure how hard it should be to turn the hub, and  if it is somewhat difficult to turn, does that indicate binding, in which case, I would need to go back in and shave down the spacer a bit.

This is such a critical point in the process, I want to be sure it's right before I reinstall brakes, bleed them, etc... I also don't want to fry the bearing and have to do this all over again.:-)

Because it's all hooked up right now, I'm reserved to unbolt it to take pictures at this time. I have a couple of calls out to some people that will hopefully help me determine if things are correct. If it is determined that I do need to go back in and shave the spacer, I will take photos of everything and send them your way. I want to be helpful to this forum and I know images really help, but if I can move forward on this project without dissembling,  the photos may not happen.

To your point about the housing bores, the seal, spacer and race all would only go in so far because the would stop at their designated bevel. With that said, it did NOT occur to me that they could wear down and possible cause the spacer to recess further. But like you said, that would result in too much end-play. My scenario has no end-play so I think we're probably good there. But really good point you make!

 

 

[Moses Ludel]

Gary...Now everyone seems to be on the same page.  You confirmed that the manufacturer expects a correct bearing fit without the shims.  The missing piece was, "I then talked to the tech at 'TenFactory' and he said that I should be measuring the 'bearing' protrusion and not the 'race'. (This differs from the attached video.)  So because I already had pressed the bearing onto the shaft, I couldn't use it. I grabbed a spare bearing and placed it in the race to measure. It measured to be within the acceptable tolerance range. So I went ahead and tightened everything down."  Conclusion:  It's not the bearing race height that they measure but rather the bearing height with the spacer and race/cup installed.

To answer your direct question about binding and what that feels like, tapered roller bearing bind is simply too much resistance.  A ball bearing would have a "notchy" feel similar to a bicycle head bearing set that gets adjusted too tightly.  Tapered roller bearings just get tight.  Realize that you're rotating the axle shaft mass plus the differential side gears and pinions when you try to turn the axle shaft flange.  A good way to determine what that mass represents is to loosen the backing plate attaching bolts a bit to increase the space.  This will relieve a bearing that is too tight—in the same way that sanding the spacer would do.  Sense what that rotation feels like, then re-tighten the bolts to specification.  If there is a distinct difference in resistance, that tension would be bearing preload.  If there is no change, the bearing has slight or no preload, and so forth...If you find excessive resistance, and all parts are in place and seated correctly, consider sanding the end of the spacer as per the manufacturer's recommendation.

So, at this point, you're comfortable with the position of the parts and the accuracy of the fit.  I do have some concerns, though.  You did not post a copy of the kit instruction sheet or a parts schematic for the axle shaft kit, so I can only go by your one photo.  There appears to be a seal inboard of the kit's spacer.  Is this seal provided with the kit?  It looks like the OEM type housing bore seal.  Do you also install an OEM style axle shaft seal/plate outside the brake backing plate?  Is that what retains the axle shaft, keeping the shaft and bearing from sliding outward from the axle?  That would be somewhat like the OEM AMC/Jeep method.  However, the bearing cup faces the opposite direction from OEM.  I'd like to see the relationship of these parts.

Please explain the parts orientation or post more photos of the parts sequencing and order.  I'm unclear what keeps the axle shaft from moving outward and how the bearing receives lubricant.  With an axle seal inboard of the kit's spacer and the retainer seal plate outboard of the brake backing plate, the axle shaft bearing does not get lube from the axle's differential.  The bearing is in a cavity between the two seals.  This would be like the OEM AMC arrangement.

If there is a seal inboard of the axle shaft bearing, the axle shaft bearings are isolated from gear lube.  You would need to pack each bearing as you would a tapered roller front wheel bearing.  Wheel bearing grease needs to be pressed up between the rollers from the wide end of the bearing cage to the small end of the cage—around the entire bearing.  A substantial amount of grease also needs to surround the bearing.  See Chapter 7 of my book if you're uncertain how to hand pack a tapered roller wheel bearing.

If you currently have the parts oriented the way the kit manufacturer wants, the axle shaft bearing grease should be repacked periodically.  Bearing grease will last longer if  1) properly packed with the correct lubricant and surrounded by additional grease in the cavity,  2) if the fit at the brake backing plate and outer retainer prevents grease from seeping out, and 3) if you apply a film of sealant on the flange end of the axle housing.  We normally do this to prevent water from contaminating the axle housing and bearings during stream crossings.

Once you have packed the bearing, install the axle shaft in the right parts sequence and tighten hardware to specification.  Again try rotating the axle shaft.  You should note a difference with the bearings lubed.  If the axle shaft turns freely and smoothly without any detectable end play, you have either zero end play or a slight amount of preload.

The bearing style (a larger tapered roller cone and cup) can be run with zero end play, or even a preload in some cases, without damage—if the lubrication is correct.  On a semi-floating axle shaft bearing, however, grease or oil will run hotter if set with a preload.  I can think of many examples where a similar bearing does run under preload.  Differential side bearings would be the best example, and they are substantially preloaded.  Another situation that comes to mind is the slight preload on the input shaft bearing of an NV4500 transmission.  Worth mentioning, both of these bearing examples use extreme pressure gear lubricant and not grease.

As for the kit bearing adjustment, the manufacturer has tried to establish a universal setting with these factors:  1)  The bearings seat on the axle shaft;  2)  the shaft is machined to precisely locate the position of the bearing;  and 3) the spacer/cup fits into an axle housing bore of a certain depth.  The manufacturer expects a degree of consistency for each of these measurements.  Seldom are axle housings that accurate, and that's why AMC uses an adjustable shim and dial indicator method.  

Your axle shaft kit manufacturer does not want "bind".  That would be excessive preload, which would ruin the bearing.  From your discussions with them, they apparently don't want end play, or they would share that measurement....They don't provide an end play figure.  The kit manufacturer apparently leans toward the tighter side rather than loose, the reason for suggesting that you "sand" material off the end of the spacer if there is bind (excessive preload).  Sanding the spacer would provide zero end play or some end play clearance.

While AMC wanted 0.003" end play per side (summed up as 0.006" total for both axle shafts), you need to be clear whether the kit manufacturer expects these bearings to run under a slight preload with no end play and no bind.  With the bearing tapers facing inward, this kit may be designed for a bearing preload and no end play.

To comment further on the bearing retainer method, sealing, lubrication or axle shaft end play, I need to understand the kit's design and the manufacturer's installation steps.  Some photos would help...In the photo you posted, it looks like the kit spacer is right against the seal face.  Is the spacer pressing against the seal face? 

Moses

[waxman]

@Moses Ludel

Thanks for your reply. I've attached the instructions and a screen shot of the parts included in the kit.

The info. from the tech regarding measuring the bearing, as opposed to the race, was obtained well after you and I began our conversation. But you are correct, that was the important missing piece.

Although the bearing came pre-greased, I did heavily pack it more. and I added extra grease in the housing bore where the race and spacer sit. You are correct, the differential fluid does not reach the bearing, because of the seal. So again, the bearing needs to have plenty of it's own grease.

To your mention of " you need to be clear whether the kit manufacturer expects these bearings to run under a slight preload with no end play and no bind", THAT is the perfect question I need to ask them. Knowing that would give the final piece to the puzzle. Should it have pre-load, or no?

Thank you for providing a way to test whether or not it is binding. Based on what you said, I think I may have to go in and sand down the spacer a bit. When the back plate bolts were looser, the hub rotated more freely. As I tightened the bolts down, it became increasingly harder to rotate the hub. I'll have to sand until I find that "sweet spot". Before I do anything though, I will ask them about the pre-load. Maybe they do want it tighter.

This has been such a learning experience for me and having this resource is invaluable!! I can't thank you enough!

amc20.pdf

[Moses Ludel]

Gary...The parts layout is helpful.  I understand that the conversation with TenFactory took place after our early exchanges.

Looks like a stock type outer seal/retainer plate.  The bearing cup and bearing direction are reversed from the OEM arrangement.  The new bearing cup is recessed in the bore, so the cup does not rest against the seal/retainer plate. 

My concern was how the axle shaft and bearing stay fixed in the axle housing.  It looks like the bearing's inner race stops against the seal/retainer plate.  Correct?  The part to the left of the bearings is apparently the lock ring for the bearing;  the lock ring helps keep the bearing from loosening on the shaft and prevents the axle shaft from sliding out of the axle housing.

Does the bearing inner race ride against the seal/retainer plate?  If so, this setup uses the outer seal/retainer plate as pressure against the bearing inner race collar to hold the axle shaft in place.  Correct?  If so, that would be TenFactory's motive for preloading the bearing and explains why they give no end play measurement beyond "no bind".  They apparently want the plate to apply some degree of pressure, short of binding the bearing with too much preload.  

If this is the design and intent, the rotating axle shaft bearing's inner race collar rubs against the plate.  True?   Moser does this, too.  I ran into this same setup when drafting the CJ book.  It's unorthodox by AMC standard, which calls for end play.  However, AMC ran the bearing in the opposition direction with the back side of the bearing cup (a larger circumference) resting against the retainer plate.  It's easier to brace the cup securely at that larger circumference and use shims to get an accurate bearing adjustment.  TenFactory, like Moser, apparently wants to tension the seal/retainer plate some.  They do not build end play into the equation. 

Either bind or your comfort level would be the only measurement short of rotating the axle shaft with a torque meter or spring scale in two modes:  1) with some bearing clearance/end play and 2) rotating when hardware is secure and the bearing is preloaded.  Again, TenFactory locates the bearing on the axle shaft and cuts the spacer ring to a specific height that they believe will be close to the right bearing preload.  They caution that if it's too tight, sand the spacer ring. 

Moses 

[waxman]

@Moses Ludel

Your reference to the Morris configuration is accurate. TenFactory has the same layout.

I have everything reassembled now and it runs great. Actually a much smoother ride. (My old bearings were on their last legs). Now it's a matter of hoping everything is how it should be. I'll find out soon enough if things are fine, or the bearings fry and back to step 1. Knock on wood!!!

I will post my results here.

Thanks again for taking the time and effort to help me. It's been so very helpful!!!

Gary

[Moses Ludel]

Gary, I have a concern:  In your one photo, you have the bearing outer race (cup) facing the opposite direction from factory/OEM.  If you installed the bearing cone to match this fit, you have the taper of the bearing facing toward the center of the axle.  I puzzled over why Ten Factory would use the seal retainer plate as a thrust surface, allowing the bearing's inner race collar to spin against the seal/retainer plate. 

I went to the Ten Factory website just now and opened the Ten Factory instructions for your kit.  If you installed the bearing cups/outer races as shown in your photo, the bearings are installed backward.  The illustration in the instructions is poor;  however, following the Ten Factory directions would indicate that your bearings are flipped over:

"2. Install bearing spacer into axle housing end. (Reference Fig. B) Be sure spacer fully seats in housing. WITHOUT the axle shaft installed, test fit bearing into housing end. Be sure bearing is installed as in Figure A with the inner race facing towards the center of differential. Make sure bearing is against bearing spacer.  Bearing should protrude .020” to .060” beyond housing end. (Reference Fig. B) If more than .060 extends beyond housing, remove bearing spacer, then evenly grind bearing spacer OR use original shims to “lengthen” housing end until these measurements are achieved."

The "mystery" around the bearing protruding 0.020"-0.060" outward from the axle housing flange may refer to the bearing cone protruding this amount when the cup and cone have been installed in the correct direction.  For testing, you would hold the spacer and cup in place while pulling the bearing cone squarely outward to get the protrusion measurement.

If you flip the cone and cup over, that would be the factory method for holding the bearing cup in place.  The seal retainer plate applies pressure at the outer (thicker) edge of the bearing cup.  The cup does not rotate, and there is nothing rubbing against the seal retainer plate...If I'm correct, your current bearing installation will cause the bearing's inner race/collar to wear against the face of the seal retainer plate. 

Below is the AMC/OEM relationship of the bearing cup, cone and retainer plate .  Note the direction that the cone and cup face.  Your new bearings should face this way, too:

To provide clarity for anyone installing the Ten Factory one-piece axle shaft kit, here is the Ten Factory installation instruction PDF.  You can zoom into the illustrations for more detail:

Ten Factory AMC 20 One-Piece Axle Installation Instructions.pdf

Follow up on my concern.  Your worst case scenario is the need to remove the axle shafts and carefully remove the lock rings and bearings.  (I would install new bearings and new lock rings.)  When removing the lock rings, relieve the lock ring before pulling the bearings, using the AMC FSM or my book's guidelines. 

Use a bearing spreader and tall press if available to remove the bearings without damaging the axle shafts:  https://www.harborfreight.com/bearing-separator-63662.html will do for occasional use at $30.  Check the seal retainer plates for damage like bowing, wear or seal injury.  These parts and the inner seals are likely still okay at this point but confirm. 

If you did not modify the spacer rings and the inner seals and seal retainer plates are still okay, you're only out the bearings and lock rings.  (There is an expensive tool that would enable removing the bearing cones without damage, though new bearings are common and not expensive.)  Lock rings and bearings should be available from Ten Factory.  Common bearings like these are also available from any quality bearing source by part number.

Moses

[waxman]

@Moses Ludel

You are correct in that the bearing is flipped with the bearing's smaller tapered end facing the differential. I read through your last post multiple times to be sure I understood your concern. If I understand you correctly, the concern is that because the bearing is flipped opposite the OEM settings, the  "bearing's inner race/collar to wear against the face of the seal retainer plate."

From what I can tell, there is only one direction in which the bearing can be installed, due to the fact that the bearing race is installed on the housing and not the shaft. Since the tapered end of the bearing can only fit into it's race one way, and the race is already installed on the housing, the small end of the bearing HAS to face the differential. If I were to flip the race 180 degrees and place it in the housing, the beveled side of the race(which receives the bearing) would be facing the differential. then there would be no way the bearing could enter it's race.

The only way I see that the bearing and race COULD be flipped, would be to make sure the bearing's outer race was on the bearing as it was pressed onto the axle. Then I would have to try to get the race to fit in the axle housing while it was on the axle. That would be a very tough thing to do because of the precise fit of the race into the housing. It was hard enough to get the outer race to fit in the housing by itself, so I could measure. Having it attached to the assembled axle would not only be really tough,but would not be congruent with the instructions. hmmm.

I don't know Tenfactory's intentions regarding what component is putting pressure on the bearing. Is it intentional to have the bearing's inner race/collar to wear against the face of the seal retainer plate? Or maybe the inner race/collar is meant to remain stationary, while the bearing rotates? Not sure.

I attached Morris' instructions for their one-piece axle kit. To me, It seems the same as Tenfactory's. But I'm the first to admit that the newness of this is quite a steep learning curve. So I could be wrong.

I appreciate your willingness to discuss this. As we progress through this conversation, I'm kind of in the mindset of "let's see what happens".

As for now, it is running great, and much smoother than before. If it does end up failing due to the things we're discussing, I will disassemble and re-evaluate. At that point, I will take video and pictures of EVERYTHING as to help this forum.

 

Thanks,

Gary

Morris.pdf

[Moses Ludel]

Gary...After posting my concern, I spent time viewing a half-dozen videos at YouTube before finding one that made sense:  The axle shaft bearings have a unique cone.  Since I had no other photos to weigh, I gathered that you had a typical bearing cone.  If you have the cup, bearing cone and spacer design shown in the video, the video shows why you do have the bearing oriented correctly:

https://www.youtube.com/watch?v=rmKC7FRf7-E

In the YouTube video, the bearing trial fit is done as a bearing assembly off the axle shaft.  (See 3:30-minute and elsewhere for the measurement method.)  The cone outer shoulder is part of the stack height measurement.  This explains why the outer race/cup recesses into the housing bore.  Once you determine the 0.020"-0.060" bearing cone protrusion, you're on the right track.  They use a 0.030" baseline goal.

You are correct, the outer race/cup is a slight interference fit into the axle housing to help resist cup rotation once in place.  Your kit's design works due to the special bearing cone design.  If your bearing cones look and function like the bearings shown in the video, you should be fine.  That special bearing race and spacer ring, if installed in the correct direction as shown in the video, should not create an issue.  My concern about the bearing cone rotating against the seal retainer plate has been allayed.

Make a point of reviewing the video.  Compare the comments and examples to your work and parts.  The YouTube video explains how the right parts will interface and function properly.  If your kit matches up with the parts depicted in the video, and if that was your method of installation, you should be fine.

Moses

[waxman]

@Moses Ludel

Thanks for the reply. I must have watched that video literally 30 times, to wrap my head around everything.  I'm relieved to hear that you concur with the way I have it set up. I'm still knocking on wood, as that is my general paranoid nature. ha.

Thank you again for the vast amount of time you spent on this with me! Seriously! It's very much appreciated.

Gary

[Moses Ludel]

Hi, Gary...The bearing depicted at the Platinum Off-Road video shows a unique cone design.  At approximately 2:30/2:31-minute and 5:40-minute onward of the video, you can see the flanged lip style race of the bearing cone.  This flange allows the bearing rollers to "walk" between the inner and outer bearing races while the axle shaft rotates.  This style bearing cone/assembly differs from the common bearing that AMC uses as OEM with the two-piece axle shaft. 

The only concerns you should have are whether the Ten Factory kit uses the bearing style depicted in the video and whether you installed the spacer rings with the ledged end facing in the correct direction, which it appears you did according to your photo.  (I can see the spacer ledge facing outward in your photo.)  The Ten Factory parts photo shows the correct bearing cones/assemblies if that's what came with the kit.

Here is the difference between the two bearing styles (photos courtesy of Crown Parts):

Above is the conventional AMC OEM axle bearing cone and race.  The bearing cone race does not have an outer flange lip;  there is only the inner race/collar that presses onto the axle shaft and rotates.  If installed "backward", the inner race of the bearing cone would be rotating with the axle shaft and rubbing/wearing against the seal retainer plate—this was my concern.

Above is the bearing that comes with a one-piece axle shaft kit (Crown Parts).  The cone has a flange lip as shown in the Platinum Off-Road video at 2:30-2:31 minutes and elsewhere.  The races differ from the conventional bearing cone.  With this special bearing cone design, the bearing cone's inner collar presses onto the axle shaft (locked in place by the lock ring) and rotates with the axle shaft.  However, the flanged lip stays fixed against the brake support plate (according to the video's CJ example at 6:42-minutes and onward)—the rollers "walk" around the axle shaft and inner race of the cone.  The cone flange lip does not rotate.

Platinum Off-Road has provided a real service to installers of these one-piece axle shaft kits.  The video fills the instruction voids in both the Rugged Ridge and Ten Factory's kits.  It is interesting to see that Platinum Off-Road applied the same yardstick we did in our exchanges about "no bind" and the correct bearing preload setting.  Yes, to your original question, this design points to a bearing preload.  As Platinum Off-Road and I both suggest, there's a sweet spot between too tight and not tight enough. 

Kudos to Platinum Off-Road for detailing this work.  YouTube is rife with poor instructional videos and shoot-from-the-hip, anecdotal experiences.  On that note, for those working on AMC (stock) two-piece axle shafts without the benefit of my book or an FSM, please remove the tapered flange wheel hub before attempting to remove the bearing.  You will save both time and risk of damaging the parts.  Here are some puller guidelines.  I always use the triangular plate with the puller to preserve the wheel studs:

https://www.4wdmechanix.com/HD-Video-Tool-How-to-Using-the-OTC-7394-Hub-Puller?r=1

On the two-piece axle shafts, reinstall the removed hub in exactly the same position as the original location.  If in doubt about a good fit, buy new flange hubs.  AMC wanted techs to install new hub flanges any time the hub got removed!  For brake service, the brake drums come off without removing the flange hubs.  Hub flanges seldom require removal.

During a bearing change-out, the two-piece axle shaft cuts "spline teeth" into the blank (new) cast hub.  Removing/re-installing the old flanges is risky.  Worse yet, most installers never follow the nut tightening procedure, relying instead on an air gun to cinch the nut—an air gun will seldom reach the seated torque setting due to the taper of the hard steel axle shaft.  Torque the hub nut as I describe in my book and also in the April 12th exchange here at Gary's topic. 

This should sum up the AMC 20 one-piece aftermarket axle shaft installation details.  The kit manufacturers need to provide detailed photos, better step-by-step instructions and video how-to similar to Platinum Off-Road's contribution...We could have saved considerable time.  

Moses