Halfshaft Bolt Answers (all threads merged)

[analogdesigner]

FEA by Analogdesigner

Here's a layout showing the possibility of adding two more M8 bolts.

Gimbal, et al:

Here is a finite element analysis of the situation using a modified ARP 12 point fastener. This assumes that the retaining washer is allowed to flex inward while being supported at the outer edge:

The washer is too compliant! Jay

 

[Pete S.]

Hi Pete,
A quick calculation shows at about 200 mph the rear wheel will be spinning at about 3000 rpm. With the safety wire so close to the center of rotation and the wire being very low in mass, it would be difficult to detect an imbalance.
A tiny rock in the grooves of the tire would create a much larger imbalance and that might not be detectable either.

Thanks. I agree.

Pete S.

 

[Indy GT]

FEA Results

Jay, nice job on the bolt FEA analysis.

150Ksi at the blot head fillet might be a little high even for a good ARP bolt. Was wondering-
What thread coefficient of friction did you assume to arrive at the bolt load?
Or what bolt preload was used?
The model indicates almost a point reaction at the washer OD to react the bolt load. Is the circumfrential reaction lip at the washer OD really that small? I would think a bit more reaction shelf would allow less washer flexing and allow a more uniform load into the bolt head (ie more equal loading left-to-right of the head) as shown in the results figure. However the design is what it is if that is the extent of the reaction lip.

I also agree with the safety wire imbalance calculation, it will be no-factor. The low radius essentially on the centerline (or very close thereto) of the axis of rotation makes the wire mass inconsequential in its contribution to vibration.

 

[Gimbal]

Here is a drawing showing a much thicker washer design. One with existing M8 bolts and one using Flat heads. If the new hardened washer is not stiff enough this might be an easy solution. I'm not sure how much stiffer the cap screw layout would be nor do I know if the Flat head bolts would be weaker. Any chance analogdesigner could do an FEA on this?
Thanks,
John

 

[analogdesigner]

Fea

Jay, nice job on the bolt FEA analysis.

150Ksi at the blot head fillet might be a little high even for a good ARP bolt. Was wondering-
What thread coefficient of friction did you assume to arrive at the bolt load? The friction coefficent that I used was 0.08. This should be a good number since I would be using ARP's thread lube, I would run the fastener in and out 5 times as recommended by ARP and their Stainless 300 series has a natural lower friction coefficient due to the dissimiar alloys.
Or what bolt preload was used? 6,654 lbs. (29,597 N), this would occur when the fasteners are torqued to 22.9 lbs/ft.. This represents 85% of the fastener's yield strength. I could of used 90% though...
The model indicates almost a point reaction at the washer OD to react the bolt load. Is the circumfrential reaction lip at the washer OD really that small? I would think a bit more reaction shelf would allow less washer flexing and allow a more uniform load into the bolt head (ie more equal loading left-to-right of the head) as shown in the results figure. However the design is what it is if that is the extent of the reaction lip. The retaining washer will deflect about 0.008 inches at the center. This huge amount of distortion creates a huge stress peak at the outer sides of the fasteners, just as you mentioned. The original fasteners break exactly where my software predicts and this is even when doing the analysis without the threads in the fastener. The threads will obviously create a huge stress riser, however it bogs down the analysis time, so I supress that feature for now.

I also agree with the safety wire imbalance calculation, it will be no-factor. The low radius essentially on the centerline (or very close thereto) of the axis of rotation makes the wire mass inconsequential in its contribution to vibration.

Also, the software that i am using is BOLTCALC v5.83 and Solidworks 2006 with COSMOS Designer. Any ideas, thoughts or criticism is completely welcomed! Jay

 

[BlackICE]

Since the washer deflects 0.008", wouldn't the idea of a shim between the washer and the axle help the situation? Something like a custom shim for each side to limit the deflection to something like 0.002". I know this is easy to say, but harder to do, but is this an idea that would work by limiting the deflection, while still retaining the clamping force on the flange?

 

[analogdesigner]

Shimming...

Since the washer deflects 0.008", wouldn't the idea of a shim between the washer and the axle help the situation? Something like a custom shim for each side to limit the deflection to something like 0.002". I know this is easy to say, but harder to do, but is this an idea that would work by limiting the deflection, while still retaining the clamping force on the flange?

BlackICE,

Exactly! I am going to machine my own retaining washers with that feature built-in. The problem is that it would not be a part where one-size-fits-all... I will have to custom fit each part to the particular clearance. I made my washers from 17-4 ph stainless (15-5 ph is another excellent alloy) and I am going to heat treat them myself, as soon I come up with the final size. On a "mission-critical" design, it's best minimize the number of interfaces. Meaning, if a shim can be eliminated by adding that amount to the part which contacts it, the toughness of the assembly will improve.
Jay

 

[BlackICE]

BlackICE,

Exactly! I am going to machine my own retaining washers with that feature built-in. The problem is that it would not be a part where one-size-fits-all... I will have to custom fit each part to the particular clearance. I made my washers from 17-4 ph stainless (15-5 ph is another excellent alloy) and I am going to heat treat them myself, as soon I come up with the final size. On a "mission-critical" design, it's best minimize the number of interfaces. Meaning, if a shim can be eliminated by adding that amount to the part which contacts it, the toughness of the assembly will improve.
Jay

This should be a sound fix that will last. Unfortunately, this is difficult and expensive for Ford to do, and more so for the average backyard mechanic. If you can publish an easy procedure for us owners to get the proper measurements, then maybe you can craft custom washers for us GT owners, or maybe set series of "off the shelf sizes" would work if we knew the range of sizes needed. I would like to limit down time, but custom parts shipped around the country may take some time.

 

[nota4re]

So Jay, I guess you are agreeing now with my theory (see post #431) and intend to implement a resolution?

 

[BlackICE]

I looking at Gimal's drawings, another idea is one could machine the flange to get the washer/axle clearance to the correct range. I don't have a flange to look at, but I one could chuck it to a lathe square it up and face the flange's washer mating surface.

 

[analogdesigner]

So Jay, I guess you are agreeing now with my theory (see post #431) and intend to implement a resolution?

Kendall, yes, your solution is excellent. :thumbsup There is still this potential problem for the outside edge of the retaining washer to bend outward depending on the tightness of the spline engagement of the flange hub. Just in the past few days I have learned that the "hardened" washers can also permanently bend. Also, when the flange hub is pulled half way out, it wobbles up and down by several degrees. It's just a poorly toleranced part. A thicker retaining washer would be a great help, only if there was some additional room.
Kendall, nice going! Jay

 

[analogdesigner]

I looking at Gimal's drawings, another idea is one could machine the flange to get the washer/axle clearance to the correct range. I don't have a flange to look at, but I one could chuck it to a lathe square it up and face the flange's washer mating surface.

Yes, that could work too! You just have to be very careful to remove just the proper amount, since the tolerance should probably be +/- 0.0005".

One other note to everyone. Hopefully I will have time tonight to run some more finite element analysis, so if any of you have an idea, I could give it a try. Jay

 

[AMB]

Halfshaft Bolt

You could get more Halfshaft Bolt CLEARANCE by making a SPACER between the Transaxle Output Flange Hub and the and the Axle Flange!!!

AMB

 

[analogdesigner]

You could get more Halfshaft Bolt CLEARANCE by making a SPACER between the Transaxle Outpurt Flange Hub and the and the Axle Flange!!!

AMB

AMB, yes, that has been considered also. Good thinking..
Jay

 

[tpraceman]

Jay & Gimal nice drawings, what is the material/Heat treat of the flange the new harden washer lip edge mating with?

I would suggest a W-1 shock tool steel @ 56/58 RC for high strength but if the retaining washer is harder than what it is mating with wont that just wear instead?

 

[Gimbal]

Yes AMB good thinking! We could get 1/8-1/4 inch of room without jeopardizing anything, probably.

tpraceman...The flange looks like cast steel only. If everything is nice and tight nothing should be rubbing to wear things out. If there is movement between the hard washer and the cast flange that could be a sign of the fundamental problem. I did not notice any wearing in that area on my car but I only had about 800 miles when I replaced the 2 bolts. All seemed pretty good. The only thing I noticed was the bottom of the screw heads seemed a bit too shiny and the flange splines were tight in rotation but dange loose in the "rocking" direction. Only the washer being tight would try to prevent the flange from rocking. That would be a bad design if that is the case.

 

[911teo]

OK I have just finished reading the whole 52 pages... I feel exhausted and still confused. Sorry I am no engineer...

Basically there is a problem with no final solution. The problem may or may not affect my car, but the patch is cheap insurance as there are very few (one if I remember correctly) documented failure after the upgraded bolts/washers were used.

I am about to pull the trigger on the Accufab kit. Am I correct in understanding that this is not the final solution but a very very good upgrade from the stock one and it should reduce dramatically the chances of me being left stranded in the middle of nowhere?

Thanks again

 

[Gimbal]

Hello 911teo,
In my opinion there is no real fix yet.
I was told there were more than one upgraded failures but I have no idea how true that is.
I would choose the Accufab over the Ford "upgrade" kit. In fact I have already installed the Ford kit and now plan on buying the Accufab kit and installing it. I might even design my own custom washer. Not quite sure at the moment, I'm still toying with a few ideas.

 

[Shadowman]

Hello 911teo,
In my opinion there is no real fix yet.
I was told there were more than one upgraded failures but I have no idea how true that is.
I would choose the Accufab over the Ford "upgrade" kit. In fact I have already installed the Ford kit and now plan on buying the Accufab kit and installing it. I might even design my own custom washer. Not quite sure at the moment, I'm still toying with a few ideas.

This thread certainly has legs

I had the opportunity to chat with Jay over the week-end and then again today; I shared that IMO before the space between the stub shaft end and the base of the coupler securing washer is filled in I would verify that there is not a required preload established on the inner bearing race during the installation process which removing this space would be altered are removed.

Furthermore; my left coupler has very slight axial roation and almost no end wobble...and no oil leak... However my right side has what I consider to be excessive axial movement and excess wobble when fully pushed in..... add to this the right side was leaking oil.

All the best

Shadowman

 

[Gimbal]

Here is a layout showing a new custom hard washer design. The washer is being attached to the flange hub with 6 screws, then both are attached to the axle with the 4 screws, if I am able to drill and tap 2 more M8 threads into the axle otherwise only 2 screws. Machining the 6 screw threads into the flange hub should be easy. Matching the custom washer to the flange hub will also be necessary along with machining a flat surface for the washer. This should not be too difficult.
The unkown is whether this is really much better. It's not practical to try a new design out on a few thousand cars and put on some miles and see what breaks when. So only hunches, experience and some analytical calculations is all we have to go on. I feel like a giunea pig! Hey, no risk = no fun!