Seventh Gen rear hub bearing issue & upgrade

[Last American Indian]

I would offer this cautionary story if you plan to build a ground pounder as they are called! I can not speak about all FWD’s as I have not built the suspensions of all manufacturers, but of all GM & the like there is a serious issue with McPherson strut/knuckle suspension vehicles, especially when it comes to building an aggressive handling suspension. This is due to the illogical application of the wheel bearing design! 

In a traditional rear wheel drive the bearings are installed on a traditional spindle with two tapered roller bearings. This design can not fail, I.E. the wheel falls off! Other than to have a bearing go bad, there is little else to occur beyond stranded! As it would be virtually impossible, due to bearing failure, for the wheel to come off!

That is not the case with a FWD! The rear hub bearings are held in place by bolts that thread into the knuckle. Which pushes the wheel hub into the knuckle. The wheel obviously bolts to the hub. If the hub bolts that are threaded into the knuckle fail for any reason, the hub has nothing to keep it in the knuckle, or the wheel on the car!

This was such the case for me! Fortunately I caught my problem before it became a dilemma or catastrophe. When I built this suspension many years ago I built everything new from the ground up, including hub bearings. They were done more than correctly! The fit to the knuckle was a little tighter than factory & the torque was 5 pounds more. And yes it has taken nearly 20 years to raise its head, but that’s only about 15,000 miles. 

So it went like this. For the last half of last summer & most of this past summer every time I would make a hard, I.E. aggressive right turn I would get this rattling noise from the left rear that kind of sounded like decelerating exhaust while in the cornering sequence. Straight forward driving it seemed quiet. Over & over again I would check & recheck, exhaust, brakes, tire rub, suspension clearance, suspension component torque, etc., but nothing! It drove me nuts, because I knew something was wrong. I took out the back seat pulled the access plate to the fuel pump. Now it appeared more metallic then. Took it on the freeway & now I would hear a intermittent hard metallic crack. That was it, I knew I needed to put it up in the air & go through the whole back suspension piece by piece, oh great!

Well I didn’t have to go far. Removed left wheel, caliper, rotor & bingo! There it was! Looking through the access holes of the bearing hub tire mounting flange was a hub mounting bolt. Loose & bent. I replaced it & checked the torque on the other 3 bolts. All were loose! I check the other side. One needed re-torque slightly, the rest were good.

My take on this is this. The hub bearing mounting setup is poor at best, but the front is better than the back. The front use 3 larger bolts mounted from the back. This is better because in hard cornering, the loaded wheel, this places the bolts in compression & the bearing receives more of a angular vertical load. So while this is far less than optimal, it’s by far better than the back.

On the other hand the back is abysmal! In a hard corner the rear load wheel has a similar load as the front, but the bolts are a third of the diameter & they’re being stretched & the following wheel is also being stretched because they hold the bearing hub from the front. 

 

I have a new design for these that I am building this winter. I will share it with you folks so if you chose you’ll be able to incorporate it as well. This design will be more in the vain the a spindle type attachment that will have a means to be a fail safe design that would prevent this type of bolt failure. 

This post is to make anyone who wished to do any of the aggressive modifications I’ve done aware of the possibility. I wouldn’t think you would see this type of issue in a non modified vehicle. I have no doubt that the aggressive nature of the Indian’s suspension is the reason for the occurrence. When you’ve increased the track width by a foot, stiffened the suspension components so they transfer power effectively to the ground were it belongs & you actually use it in the manner you built it for; it’s not surprising! 

What is surprising is that any head engineer would let such a flawed design make it to production! Let alone prevail for decades & hundreds of thousands of vehicles. The simplest of design change would have kept the possibly of the catastrophic failure like the wheel coming off, should have been the mind set no matter how unlikely. 
 

I will follow up with this shortly.

Edited January 8 by Last American Indian

[Last American Indian]

One of the other issues with the factory rear design are the bolt heads & fit. The compression part of the bolt head, the diameter, is to small & the fit of the bolt shank is far to small. The diameter of the through hole in the brake backing plate is .395 to .400, but the through hole in the hub bearing is .460! With the bolt size being 10mm (.390.), the more important fit is the bolt size to the hub bearing! So the whole setup is garbage! The clearance fit of the bolt to the hub bearing should be no more than .015 total, .0075 on a side, with .010 total or .005 on a side being ideal.

So before I move on with the detailing of this modification I want to show just how poor of an engineered setup this is! As I said before the head engineer that ok’d this design is a disgrace to the craft! 

Look at the following pictures. First is the bearing hub bolt. The shoulder of the bolt head is .682 thousands by .100 thick. Yet the hole in the bearing hub is .470 thousands. Put in perspective that is only .106 thousands on a side bigger that the hole! On top of that the bolt shaft is .388 thousands in diameter. That is typical of a bolt thread as they are always a few thousands small diametrically, of which case means there is .294 total difference in diameter & .147 on a side! That is not how you engineer bolts & their respective fits. Considering the fact that these 4 bolts on each hub carry the entire transferred lateral load of the rear suspension; it isn’t surprising that under aggressive handling they fail.

Additionally, the last picture shows just how little contact, I.E. clamping force there is generated from this setup! The thin dark line that runs around the outer edge of the hub bolt is all that is making contact with the hub! A mere .106! That pitiful! 

[Last American Indian]

So here is the start of this improvement modification. My advice would be get a spare set of knuckles. So obviously I would recommend sandblasting or glass bead the knuckles if they’re not new.

Now drill the two 10mm threaded holes that are on the opposite side from the caliper mounting flange of the knuckle deeper, about 2.250” deep total. These two holes are blind. The setup to do this should be done on a mill as you will see in the attached pictures. 

Now you will flip the knuckle over to spot face the backside of the two short 10mm through holes. This is done to make a flat surface for the new bolt arrangement that will be used. This arrangement will change the two through holes from 10mm x 1.5mm to 7/16 14 thread. This is done by drilling those two holes out to the tap drill size of a 7/16 x 14 bolt. Then tap the hole with a 7/16 tap. You will notice that drilling out the 10mm taped holes does not remove all the threads. That’s ok! Metric threads are a very different animal when compared to US threads! American threads are far & away superior to metric, fact based from the principles of what threads are! I don’t just say that because I’m American. Metic threads are weaker, strip more often & in general defy the good principles of threading, which is considered an art form! And has been from almost the beginning. I can speak to this later if someone wants to know more, but for now you’ll never know that there was ever a 10mm thread there when finished!

Next you will need to open up the two holes on the brake backing plate that corresponds to the 7/16 bolts. Drill them out to 7/16. 

[Last American Indian]

Now on with the modification that needs done for the other side of the knuckle. While the other 10mm threaded holes are now at around 2.250” deep you’ll need longer bolts. I would seriously suggest using 304 stainless steel! Why? Steel & aluminum don’t get along with each other & since these are blind holes the likely hood of there being electrolysis between them is inevitable with corrosion that will freeze them up, being that they are in a blind hole. This makes that corrosion difficult to effect because of that. So the difficulty of getting them out later maybe significant. So basically you I’ll make two long studs & set screw them. Where they interface with the set screws the threads will need to be removed by undercutting. For reference I used 8x32 thread set screws. I also cut a screwdriver slot into the top of each stud to use for installation.

[Last American Indian]

At this point the hub bearing is secured soundly. The two 7/16 bolts now only have .016 thousands clearance on a side when in the bearing hub! That’s a decent fit & will eliminate the bearings hub from any rotational movement. We also are now going to have a much larger clamping force on each hole of the bearing hub, but we are not done yet. Now comes the part that makes the complete failure of the bearing hub, I.E. wheel coming completely off the car, highly unlikely. Not to mention a much stronger more solid & robust setup that will now, like many other components that have been modified on the Indian, to reduce or eliminate flex, & subtle vibrations that keeps the suspension from transferring all the energy you can to where it belongs, the ground!

 All 4 bolts are locked in place in the knuckle. The two 7/16 bolts have lock washers & are torqued to 60 ft lb. in the knuckle, more than what the factory specs call for. The two 10mm studs are locked in place by the 8-32 set screws. And now we are going to machine castle nuts for each bolt & drill each bolt for a cotter pin. This arrangement is as close as can be to a spindle arrangement from a safety & secure perspective. This is now a solid racing setup for hard aggressive cornering!

 

Edited January 11 by Last American Indian

[Schurkey]

Remove all rust from mating surfaces.  New bolts, clean and inspect the female threads.  Use a drop of Loctite 271 thread locker on each bolt unless the new bolts come with a thread-locker preapplied.  Torque to spec using a torque wrench that is properly calibrated. 

 

No machining needed.  No headaches, no excessive cost, no excessive labor, no excessive down time, no reinventing what works perfectly fine on millions upon millions of vehicles.

Edited Monday at 03:08 AM by Schurkey

[Last American Indian]

On 1/12/2025 at 10:08 PM, Schurkey said:

Remove all rust from mating surfaces.  New bolts, clean and inspect the female threads.  Use a drop of Loctite 271 thread locker on each bolt unless the new bolts come with a thread-locker preapplied.  Torque to spec using a torque wrench that is properly calibrated. 

 

No machining needed.  No headaches, no excessive cost, no excessive labor, no excessive down time, no reinventing what works perfectly fine on millions upon millions of vehicles.

Wow! Who knew! Thanks for that extremely rare knowledge!

 So, for the rest of you that actually read this post & understood that the Indian is nowhere close to a regular GP. I.E. pulls over 1.1 g’s in a 300’ skid pad test & over 4g’s in abrupt powered right hand turn among other pure performance handling maneuvers & have an interest in building such a ground pounder! As is detailed in Last American Indian (a ground pounder build) the members ride section. This post was for you! So you understand how much lateral g force there is exerted in this type of build & how some FWD pieces need extra attention! Enough said!

[pwmin]

I'm intrigued for sure.  To put in that extra percent or two takes a monumental effort.  Not worth it for most people, but I appreciate your hard work and dedication for sure.

[55trucker]

To be honest * & completely neutral on this post* I don't even Know what a 7th gen w body is?

[Last American Indian]

4 hours ago, pwmin said:

I'm intrigued for sure.  To put in that extra percent or two takes a monumental effort.  Not worth it for most people, but I appreciate your hard work and dedication for sure.

PWMIN, I get that & I say that at the very beginning of the post. That said I thought that if, & it may be a big if. That anyone who went down the path that I laid out in my members ride post they might be made aware of this issue. Also this may look extensive, but it’s not that much work & it does not equate to one or two percent. The best 300 foot skid pad test I have ever seen out of any FWD Grand Prix is .76 to maybe .79. To get to the 1.1 range you need more than 40%. All it is ok. No one needs to do it! But just as a side note this happens to enough regular FWD’s of every mfg as well. Always to the rear hub/aluminum knuckle combo! The loctite fails as does the torque load due to the disparity in metallurgy & thermal expansion.

1 hour ago, 55trucker said:

To be honest * & completely neutral on this post* I don't even Know what a 7th gen w body is?

7th Gen ran from 1997-2003. Many people call this incorrectly the sixth generation, but it is the 7th.

[55trucker]

6 minutes ago, Last American Indian said:

 

7th Gen ran from 1997-2003. Many people call this incorrectly the sixth generation, but it is the 7th.

Hmmmmmmmm........there are only 3 gens of the W platform

1st Gen 88 thru '96, 2nd Gen '97 thru '03, 3rd Gen '04 thru 08....the Impala did continue as a 3rd gen W platform vehicle until 2016

 

[rich_e777]

The forum OGs started the count over when the G-body Grand Prix's went FWD into the W-body back in the 80`s.

I get both sides of the argument but has there ever been an official sit down at one of the old meet ups to determine who is right over this?

[Last American Indian]

2 hours ago, 55trucker said:

Hmmmmmmmm........there are only 3 gens of the W platform

1st Gen 88 thru '96, 2nd Gen '97 thru '03, 3rd Gen '04 thru 08....the Impala did continue as a 3rd gen W platform vehicle until 2016

 

Ya, that’s not how a name plate from a manufacturer is done! Ever! A name plate, in this case the Grand Prix, is determined by a manufacturer body change! Not a platform change or a discontinuation & then a revival & not even a drivetrain change. Case in point the Camaro was discontinued in 2002 & then revived in 2010. 2002 was the 4th Gen. 8 years later in 2010 the 5th Gen was resurrected. The OGs in their mind didn’t want a FWD vehicle to  be associated with their RWDs. But irrelevant of what they want or how they want to spin it, the Grand Prix is one name plate line be it RWD or FWD! Fact!

2 hours ago, rich_e777 said:

The forum OGs started the count over when the G-body Grand Prix's went FWD into the W-body back in the 80`s.

I get both sides of the argument but has there ever been an official sit down at one of the old meet ups to determine who is right over this?

 

Edited Wednesday at 12:54 AM by Last American Indian