274 replies to this topic

[Johnny D]

Any heat treat shop can test hardness.

J~

[Bench Racer]

Any heat treat shop can test hardness.

J~

True

I searched hardness testers and there are hand held testers. ??? Accuracy/repeatability.

[Johnny D]

What I've seen is a unit the size of a drill press.

Has the table or bolted to it. Something real solid.

 

The machine hits the sample, bigger the mark, softer the material and whatever the machine reads.

J~

[ppridday]

There are many types of handheld hardness testers available, from the "smack it with a hammer" brinell type to a diamond tip microdur that we used to use on turbine blades.  There are plenty of places that can check them for cheap or free if you get in good with a car guy..

 

Paul

[SaulSpeedwell]

Here's my concern about trying to "failure analysis" our way out of this rabbithole:  If you test the hardness, you'll get a number, but you won't know:

1.  Was the owner's torque wrench calibrated?

2.  Were the studs torqued evenly?

3.  Did the owner have a wheel to wheel as he daringly divebombed for 11th place on the first turn of the first lap of the 2nd Annual Dutch MacGillicuddy God Save The Queen Mother's Day Memorial Regional?

4.  What wheel, what spacers, what offset?

 

This is, so far, potentially anecdotal B.S -  but -  those running correct offset wheels and stock studs are generally not reporting this issue.  But, they may also be autocorrelated to people not violating 1-3 above?

 

Crappy and/or Chinese hubs have been with us since 2002.  But what things have NOT been with us since 2002? 

[SaulSpeedwell]

The stud to hub system stresses follow the "thick-wall" model, as the thickness of the material is >> than 1/10th the radius of the hole.  In the thick wall case, the tangential (parallel to the hole) stresses are much greater than the radial (perpendicular to the hole) stresses.  Hoop stress is the tension stress measured at the inside wall of the hole, and would result in a failure at the hole itself.  If the forces induced by the press fit interference are at play, then the primary failure stress would be tangential stress, not radial or hoop.

 

Yes.  Thanks!  Put another way, hoop stresses are in play, but don't matter unless we are having cracks at the edge of the hole?  I'm imagining the hoop stresses as being the CAUSE of the compressive and tangential stresses radiating "around" the stud holes?

 

Absent my analytical and terminology fubar, I keep picturing the stress and strain induced by too much interference ... I'm picturing it as if we were FEA-ing it, and I'm seeing "red" around the studs and very possibly - likely!? - extending to the stress riser at the 90 degree of the hub flange?  If I'm crazy, please coach me.

[Glenn Davis]

Agree with all of this on the fatigue side.  I have no idea how fatigue might lessen the ultimate strength of the hub over time.  The tangential stresses of the press fit are actually compressive.  In a right hand turn, the inside of the left hub, at the bottom of the rotation, is in compression.  This would add to any tangential stresses created by the press fit.  I'm sure shear stresses, as well as other dynamic forces play a role.

Quoting myself to correct a thought blunder.  Radial stresses are compressive, tangential are tensile.  Same theory, but the additive force would be on the outside of the hub in a turn.  Apologies Keith.

[Glenn Davis]

Yes.  Thanks!  Put another way, hoop stresses are in play, but don't matter unless we are having cracks at the edge of the hole?  I'm imagining the hoop stresses as being the CAUSE of the compressive and tangential stresses radiating "around" the stud holes?

 

Absent my analytical and terminology fubar, I keep picturing the stress and strain induced by too much interference ... I'm picturing it as if we were FEA-ing it, and I'm seeing "red" around the studs and very possibly - likely!? - extending to the stress riser at the 90 degree of the hub flange?  If I'm crazy, please coach me.

 

Reading your post made me go back and review my posts and I saw I swapped the stress.  I wrote compressive for tangential, which is incorrect.  Radial stress is compressive, tangential is tensile. The hoop stress is the circumferential (tangential) stress at the wall of the hole.  It is tensile.  Hoop stress is typically a thin wall stress and causes radial failure at the wall.

 

I, too, see the colorful stress lines of a solidworks model in my head.  I'm picturing the "red" circumferential stress pattern of one stud/hole interface intersecting with the same "red" pattern of the adjacent stud/hole interface at a point equidistant from both.  Thus, the highest additive tensile stress would be at the points in between the holes as opposed to being at the holes themselves.

 

It would be nice to know if the studs from different manufacturers are nominally different in diameter.  This could be a solid clue.

 

Thanks for keeping me honest Saul!

[FTodaro]

Here's my concern about trying to "failure analysis" our way out of this rabbithole:  If you test the hardness, you'll get a number, but you won't know:

1.  Was the owner's torque wrench calibrated?

2.  Were the studs torqued evenly?

3.  Did the owner have a wheel to wheel as he daringly divebombed for 11th place on the first turn of the first lap of the 2nd Annual Dutch MacGillicuddy God Save The Queen Mother's Day Memorial Regional?

4.  What wheel, what spacers, what offset?

 

This is, so far, potentially anecdotal B.S -  but -  those running correct offset wheels and stock studs are generally not reporting this issue.  But, they may also be autocorrelated to people not violating 1-3 above?

 

Crappy and/or Chinese hubs have been with us since 2002.  But what things have NOT been with us since 2002? 

Mark I get your point, but we have had hundreds of racers race thousands of hrs on hubs and Most do not fail. however we have had some failures that are unexplained, and no known abuse, just use.

 

I reported last year the rear hub failure i had and had no know contacts of any significance. Agree that its far from science, but if we can pull it off, its worth the attempt to take a look. You never know we may learn something.

[Michael Cola]

Here's my concern about trying to "failure analysis" our way out of this rabbithole:  If you test the hardness, you'll get a number, but you won't know:

1.  Was the owner's torque wrench calibrated?

2.  Were the studs torqued evenly?

3.  Did the owner have a wheel to wheel as he daringly divebombed for 11th place on the first turn of the first lap of the 2nd Annual Dutch MacGillicuddy God Save The Queen Mother's Day Memorial Regional?

4.  What wheel, what spacers, what offset?

 

This is, so far, potentially anecdotal B.S -  but -  those running correct offset wheels and stock studs are generally not reporting this issue.  But, they may also be autocorrelated to people not violating 1-3 above?

 

I agree with Saul....there's a lot of critical information about how this part was used that is missing, which will prevent us from drawing a true conclusion about why the part failed.

Analysis of the fracture surface will give us an idea of the failure mode (I'm suspecting fatigue cracking that initiates at the stud hole...see my comments below).

Hardness and other materials testing are only helpful if we know what those numbers should be (design spec), or if we have other data to compare it to.

 

It would be nice to know if the studs from different manufacturers are nominally different in diameter.  This could be a solid clue.

 

Agreed, but you also need to know how different the hub's hole diameter for the studs are too.

 

How many people actually measure the side of the hole in their hubs and compare it to the diameter of their new ARP (or alternate brand/source) wheel studs before pressing them in?

I suspect (as others have mentioned) that some of the failures may be due to higher than desired interference fits, which caused excess stress in the hub around the bolt holes (hoop stress) which eventually led to cracking and fatigue.

 

I am in the process of installing ARP wheel studs on my car, but I've only started to do the rears (I will be getting to the fronts tomorrow).

 

I bought new Dorman 930-550 Rear hubs, pressed out the stock-style wheel studs in my vice, and then measured everything with a pair of calipers:

Hub hole diameters measured ~0.550"

Stock-style wheel studs had a outer knurl diameter of ~0.570".

     - "Stock" Interference fit = 0.570-0.550 = 0.020"

ARP Wheel Studs (PN 100-7720) have a knurl diameter of 0.579˝ (verified with calipers)

     - ARP Interference fit with original hole diameter = 0.579-0.550 = 0.029"

 

That means that the ARP studs installed in the same hole w/o modification would result in ~1.5x the interference fit of the original stock-style studs. I think a 0.029" interference is too much, so I'm thinking of drilling/reaming the holes to a larger size (9/16" Drill bit would make it ~0.562") in order to maintain a similar interference fit as the stock studs had.

 

The only reference I can find regarding to ARP's suggestion for press fits comes from an ARP vendor rather than ARP themselves:

"Per ARP recommendations, a .005 - .010" diametrical press fit is desired."

Assuming that this is really ARP's recommendation, this would mean that the "stock" stud press fit is twice what ARP recommends. I bet that a lot of people have to be running ARP's in these same hubs, and I'm sure they work, but if the hubs are stressed beyond design (via use of spacers, different offsets, etc) then perhaps it can become an issue and lead to failure.

 

I tried looking in the machinists handbook and other ANSI standards for specifications on interference/press fits, but all of the ones I saw refer only to smooth bore and smooth shaft press fits, rather than a smooth bore with a knurled shaft, which is likely different.
 

I'll post up some numbers on the fronts when I get a chance.

[Glenn Davis]

 

I bought new Dorman 930-550 Rear hubs, pressed out the stock-style wheel studs in my vice, and then measured everything with a pair of calipers:

Hub hole diameters measured ~0.550"

Stock-style wheel studs had a outer knurl diameter of ~0.570".

     - "Stock" Interference fit = 0.570-0.550 = 0.020"

ARP Wheel Studs (PN 100-7720) have a knurl diameter of 0.579˝ (verified with calipers)

     - ARP Interference fit with original hole diameter = 0.579-0.550 = 0.029"

 

That means that the ARP studs installed in the same hole w/o modification would result in ~1.5x the interference fit of the original stock-style studs. I think a 0.029" interference is too much, so I'm thinking of drilling/reaming the holes to a larger size (9/16" Drill bit would make it ~0.562") in order to maintain a similar interference fit as the stock studs had.

 

Thanks.  This was what I recommended somebody do way back in the thread.  Interference is measured relative to the radius, so your numbers need to be divided by 2.  Having said that, the ARP studs are still in excess of their recommendation by ~.005"

[Michael Cola]

Thanks.  This was what I recommended somebody do way back in the thread.  Interference is measured relative to the radius, so your numbers need to be divided by 2.  Having said that, the ARP studs are still in excess of their recommendation by ~.005"

Agreed that interference is typically given on a radius, however, that ARP recommendation I found says "0.005-0.010" diametrical press fit" which means the numbers I presented are correct.

Might be worth calling ARP to confirm that though...

[ChrisA]

I bought new Dorman 930-550 Rear hubs, pressed out the stock-style wheel studs in my vice, and then measured everything with a pair of calipers:

Hub hole diameters measured ~0.550"

Stock-style wheel studs had a outer knurl diameter of ~0.570".

     - "Stock" Interference fit = 0.570-0.550 = 0.020"

ARP Wheel Studs (PN 100-7720) have a knurl diameter of 0.579˝ (verified with calipers)

     - ARP Interference fit with original hole diameter = 0.579-0.550 = 0.029"

 

 

The original dimensions of the Dorman studs was with no doubt impacted when pressed into hub during assembly. I just measured two new unused '99 front studs. Both measured at the centerline of the knurling.

- OEM = 12.82mm/0.505"

- ARP = 12.85mm/0.506

[Michael Cola]

The original dimensions of the Dorman studs was with no doubt impacted when pressed into hub during assembly. I just measured two new unused '99 front studs. Both measured at the centerline of the knurling.

- OEM = 12.82mm/0.505"

- ARP = 12.85mm/0.506

 

The dimensions of the dorman stud knurling don't seem to have changed dramatically, since the knurling still has most of the black oxide coating on it. If it had been dramatically altered, I'd expect the coating to be gone in that area.

 

See attached pictures:

 RearHubHole.jpg   90.86KB   2 downloads  RearWheelStuds_DormanVsARP1.jpg   157.73KB   1 downloads

[Keith Novak]

I'm having a very hard time with the concept that the interference fit is the cause of the cracking.

 

I make my money in the airplane biz where entire planes are riveted together.  Interference fits and hole filling shear fasteners are the name of the game.  We experience cracking all the time but when it's due to something occurring at the holes, the cracking is radial to the holes.  In this case, the knurled fasteners will generate stress concentrations so I would expect to see radial cracking at the hole even more.  They're not failing that way though based on the photos.  I could see the interference fit contributing to another failure mode, but it being the primary cause of failure doesn't seem to compute in my head at least.

[Glenn Davis]

Agreed that interference is typically given on a radius, however, that ARP recommendation I found says "0.005-0.010" diametrical press fit" which means the numbers I presented are correct.

Might be worth calling ARP to confirm that though...

Good catch.  Given that, the recommended interference is 2-1/2 to 5 mils.

 

Your other post is a bit confusing to me.  New front hardware is that much different than rear?

[Glenn Davis]

I'm having a very hard time with the concept that the interference fit is the cause of the cracking.

 

I make my money in the airplane biz where entire planes are riveted together.  Interference fits and hole filling shear fasteners are the name of the game.  We experience cracking all the time but when it's due to something occurring at the holes, the cracking is radial to the holes.  In this case, the knurled fasteners will generate stress concentrations so I would expect to see radial cracking at the hole even more.  They're not failing that way though based on the photos.  I could see the interference fit contributing to another failure mode, but it being the primary cause of failure doesn't seem to compute in my head at least.

How thick is the material compared to the hole?  I don't think anybody is saying that the only stress involved is the stress caused by the press fit.  If that were the case, I would tend to agree with you.  However, when you have multiple stress sites at play, the stresses can be additive at other points on the hub.  Add that to the stresses induced when driving and possible material differences and I don't have any problem thinking that the press fit stress may be a player.

[Keith Novak]

How thick is the material compared to the hole?  I don't think anybody is saying that the only stress involved is the stress caused by the press fit.  If that were the case, I would tend to agree with you.  However, when you have multiple stress sites at play, the stresses can be additive at other points on the hub.  Add that to the stresses induced when driving and possible material differences and I don't have any problem thinking that the press fit stress may be a player.

I think it contributes but I don't think it's a major player.  I totally agree about the different stresses combining but generally one is dominating the equation.  The stress at the hole will be fairly localized I would think.  The relative thickness to hole size is fairly comparable to planes with thinner skins but smaller fasteners.  I also think about cracking in thicker material sections both with respect to airplane and car parts I think if it were a major player, there would be some accompanying signs of radial cracking at the hole.  I am however mostly familiar with failure in aluminum parts from a practical standpoint so I could be way off base.

[SaulSpeedwell]

I'm having a very hard time with the concept that the interference fit is the cause of the cracking.

 

I make my money in the airplane biz where entire planes are riveted together.  Interference fits and hole filling shear fasteners are the name of the game. 

 

I "get " what you are saying .... but where in the airplane biz do you cram an interference fit in so close to a 90 degree angle in a casting that is locally hardened on the other side of said 90 degree flange? 

 

I'm not trying to be a smarty-pants, I'm just thinking about how this application is being overstressed in every possible way .... and trying to mentally decide whether the stud interference could be the tipping point. 

[Mark McCallister]

Somewhat understand the intricacys of testing steel. The fractured cross section of the second picture posted had suspect looking material color/density. If the material is suspect many things being discussed become secondary. Blueskying is occuring and I am not saying the things being talked about are not relevant.

Saul has a video http://www.youtube.c...h?v=ls9rR1yCwrs of improved front hubs. He has considerable more knowledge of these hubs. He has hardness tested the outer bearing races.  Don't know if he hardness tested the hub/bolt flanges. 

 

Interesting thread.  What is this blue grease in the video?  Not sure I've ever seen blue...