'88 wheel bearing and hub upgrades

[Series8217]

maybe I missed it, but is there any reason why you're sticking with a cartridge bearing?

I'm not sticking with it; I'll still make the hub-cartridge knuckles eventually. This is an interim solution. It's something I can build and use in a month, rather than the new knuckles which are going to take a month of design work alone, including designing the fixtures and sourcing parts, and then be tricky to fabricate because of how critical all of the dimensions are.

[fieroguru]

I think its a cool idea! Any comparison info on the diameters of the bearings and the seperation distance between the two?

[Series8217]

I think its a cool idea! Any comparison info on the diameters of the bearings and the seperation distance between the two?

Here is a photo of the disassembled Fiero hub next to the GTI parts:


I disassembled and measured the dimensions of the VW GTI MKIII and 1988 Fiero front bearings:

Code: Select all

Application:                    1988 Pontiac Fiero  VW GTI MK III
Bearing / Assembly:             7466939             191498625A
Ball Diameter:                  12.71               10.3
Inner Race Diameter:            33.7                46
Outer Race Diameter:            59.1                66.6
Inner Race ID (Hub Shaft OD):   28.29               40
Outer Race Wall Thickness:      5.34*               2.2**
Row Spacing (Center-to-center): 17.84               18.1

Note: All dimensions are in mm.
* - the outer race is unsupported by the knuckle (0.1mm clearance fit)
** - the outer race is press-fit into the knuckle for support

With respect to the separation distance, the effective bearing spread is dependent on the contact angle, which I haven't figured out how to measure. See this note from Timken's bearing engineering manual:

When a load is applied to a tapered roller or angular contact ball bearing, the internal forces at each rolling element-to-outer raceway contact act normal to the raceway. These forces have radial and axial components. With the exception of the special case of pure axial loads, the inner ring and the shaft will experience moments imposed by the asymmetrical axial components of the forces on the rolling elements. The effective center for tapered roller bearings is defined as the point at which the lines of force normal to the outer ring raceway intersect the bearing axis. As an approximation, it also applies to angular
contact ball bearings. The effective spread is then defined as the distance between the bearing effective centers for a two-bearing system. It can be demonstrated mathematically that, if the shaft is modeled as being supported at its effective bearing center rather than at its geometric bearing center, the bearing moment may be ignored when calculating radial loads on the bearing.

If we assume that the effective bearing spread is the same between the two bearings, the GTI MK III bearing is stronger by virtue of its 12.6% larger outer diameter.

There are other reasons the new setup should be stronger and longer lasting than the original Fiero setup:
* Less deflection due to a stiffer hub shaft (40mm vs 28mm diameter)
* Stiffer outer race assembly due to the bearing being supported by the press fit into the knuckle and bearing caps (over 7mm thickness total), instead of a 5.3mm thick housing.
* New parts are available for the GTI MK III. The '88 Fiero bearing assembly has been out of production for over a decade, and new OE units are not available. The aftermarket ones fail catastrophically or wear out in one weekend. The remaining OE bearings are almost 30 years old and most have 100,000+ miles on them; needless to say, they aren't in great shape. Low-mileage units should last a season on the track, but I can't find any more of them.

[fieroguru]

Looks good!
I think the increased number of ball bearings will reduce the point loads at the races as well and should help improve life.

[The Dark Side of Will]

If we assume that the effective bearing spread is the same between the two bearings, the GTI MK III bearing is stronger by virtue of its 12.6% larger outer diameter.

Looks good!
I think the increased number of ball bearings will reduce the point loads at the races as well and should help improve life.

Diameter, number of balls and the ability to get better/more consistent heat treatment in the races of the ACBB than in the housing of the hub cartridge.

[Series8217]

Parts and materials ordered.

I need to get a tool to measure the ~72 mm bore with 0.005mm accuracy. Should I use a dial bore gauge, or a T bore gauge and outside micrometer? I'm only going to use these tools for this project, so I want to minimize my investment while meeting the accuracy requirement.

[Emc209i]

Just wanted to say I think having the option to install German parts that had to pass Deutsches Institut für Normung standards on my Fiero would be tits. Carry on!

https://www.youtube.com/watch?v=AWYxde1BTx0

[FieroWanaBe1]

I need to get a tool to measure the ~72 mm bore with 0.005mm accuracy.

I dont think VW's are manufactured to that accuracy, do you really need it?
You are looking at least a 250-300 Outside micrometer and a T-bar. T-bars are not ease to use accuractly.
An inside micrometer, just as difficult to use, is even more, unless you can live with a low resolution and in inches.

[The Dark Side of Will]

A tripod style inside mic is going to be the only thing that can consistently give you +/-0.0002".
I have access to them. Ship me your knuckle and I'll get the measurement.

A bore into which a cartridge bearing like that is to be a press fit needs to be quite precise in order to adequately support the bearing, retain it via friction and not alter its shape/size significantly.

For example, 0.0005" difference in bore diameter for a spherical bearing/monoball can easily make 10 inlbs difference in the preload torque.

[Series8217]

A tripod style inside mic is going to be the only thing that can consistently give you +/-0.0002".
I have access to them. Ship me your knuckle and I'll get the measurement.

Being that I need the tools to verify the dimensions of the final parts, I think I'm going to need the tools anyway.

[FieroWanaBe1]

If you will be farming this out to a machine shop, they would be able to verify starting and finished dimensions for you. And not only that, you can verify a current calibration on their measurement hardware, and measurement process, if you choose to be that anal, and demand a print-process-review before rendering their service.

Housing fit P7 (heavy shock loading on a static housing) is between 51-9 microns under according to NTN, and 51-21 microns from SKF & Timken. Giving you .002" to .0008" under nominal on the interference fit.

[The Dark Side of Will]

I'll see what I can do about finding out a good place to get one... They're $600+ tools...

[Series8217]

If you will be farming this out to a machine shop, they would be able to verify starting and finished dimensions for you. And not only that, you can verify a current calibration on their measurement hardware, and measurement process, if you choose to be that anal, and demand a print-process-review before rendering their service.

I'm making the prototypes myself, and I probably won't put these parts into production. I will more likely wait for the C5/C6 hub knuckle design to start making parts, and that will be produced by WCF.

Housing fit P7 (heavy shock loading on a static housing) is between 51-9 microns under according to NTN, and 51-21 microns from SKF & Timken. Giving you .002" to .0008" under nominal on the interference fit.

What tables are you using for that information? Timken doesn't offer a housing fit table for angular contact ball bearings; only for shaft fits. If we use the radial ball bearings guidelines, the housing fit is M7 (Heavy shock loads, indeterminate load direction, outer ring not displaceable axially).

Automotive wheel bearing installations are unique in that both the housing and shaft have interference fits.. The tables from Timken, NTN, and SKF assume either an interference fit on the housing bore or the shaft, but not both. The bearing must have specific internal clearances to accommodate interference first on both, so the safest option here is to replicate the VW GTI knuckle bore to ensure the resulting clearances are as intended. I'm using VW hubs (which have the "shaft" that presses into the inner ring) so I don't have to worry about that fit.

[FieroWanaBe1]

NTN offers wheel hub specific engineering advice in their catalog:
http://www.ntnamericas.com/en/website/d ... s_4601.pdf

Unfortunatly they list R7 for the housing fit. I do not know of a R7 ISO fit standard, I assume they mean P7.

***correction:
here is a more complete iso tolerance table:
http://ace-m.com/pdf/Mating%20Tolerance.pdf
32-62 microns interference is R7
***
Automotive applications are always hard to specify, the environment is one of the harshest available, the service life is nothing like an industrial application and the volume's resulting precision of OEMs is laughable to industrial standards.

Try talking to a chain or belt supplier about motorcycle drive chains and they will tell you the speeds and power is impossible.

[Series8217]

NTN offers wheel hub specific engineering advice in their catalog:
http://www.ntnamericas.com/en/website/d ... s_4601.pdf

AWESOME link! Thanks!

[Series8217]

I identified another candidate for a bearing/hub combination:

2004 Honda CR-V Hub (42210-S0H-000, $109) with 1990 Civic Wheel Bearing (44300-SB2-961, $68). The Civic wheel bearing has a 72mm outer diameter, 40mm length, and 38mm ID for the hub shaft. The CR-V hub has a 5x114.3 bolt pattern. I don't have one on hand so I can't measure the flange offset of the hub, but based on photos it looks to be greater than the GTI flange by a significant amount.

The CR-V wheel bearing (44300-SAA-003, $68) can be used but requires a larger bore (73mm), and according to NTN's catalog for the equivalent bearing, has the same load capacity as the Civic bearing.

This is the CR-V hub flange:


Due to the lightweight design of the hub flange it would not be possible to drill it out for a 5x100 bolt pattern. However, it does start with the same wheel stud diameter and thread pitch as the Fiero, unlike the GTI hub which has 14mm thread holes. Since the CR-V hub is $109 compared to the $201 GTI MK III hub (1H0407613B) which requires $40 ARP studs, the CR-V hub reduces the cost per side by $132 and results in a stronger knuckle due to the smaller bearing bore. However, the CR-V hub is likely weaker due to its lightweight construction.

Honda CR-V + Civic Bearing Pros and Cons (compared to MK III GTI bearing / hub assembly):
Pros:
* Cheaper by $132 per side
* No hub modifications required
* 5x114.3 bolt pattern (very common pattern)
* Same thread pitch as Fiero bolts
* Factory stub axle available from FWD CR-Vs (GTI requires cutting down an outer CV joint end)
* Leaves more material in the knuckle for greater strength

Cons:
* Cannot redrill to other bolt patterns
* 64.1mm hub bore (Fiero and GTI are 57.1mm)
* Stiffer, probably weaker hub
* Increased hub flange offset
* Hub durability not proven in motorsports

[Series8217]

I decided to make these parts I would perform external machining and rough boring myself, but outsource the honing operation for the final internal dimensions. I need to make some changes to the design to accommodate alignment pins for boring and final assembly. I also need to identify a shop to perform the honing operation.

[Series8217]

Direction change again...

I took a look at a J-body front bearing again and the 73.5mm bearing carrier steps down to 72.5 mm at a distance of 10.5mm from the mounting surface of the flange. If I were to take a little off the bearing and a little off the Fiero knuckle (70mm) to meet the size mismatch in the middle, the amount I'm boring out the knuckle is pretty much the same as the 72mm GTI bearing... and ultimately a lot less work than making two bearings retainer caps and honing the modified knuckle / cap assembly. It's also $200 cheaper per corner for new parts...

[The Dark Side of Will]

Pics?
Those need to be held together by bolts, right?

Also, you won't need the same degree of precision machining the housing for the J-body bearing.

How does the offset compare?

[draven]

Direction change again...
If I were to take a little off the bearing and a little off the Fiero knuckle (70mm) to meet the size mismatch in the middle

70mm? 2.75" ?

Did you mean 7mm?