Everything You Always Wanted About Lower Ball Joint Bolts

Is it possible the bolt coating is an effort reduce corrosion like antiseize?

 

This is the torque / clamp load chart I keep in my laptop as a quick reference. Data is from "Fastenal" handbook.



They don't show the M10 x 1.25 and it has a slightly larger stress area - but this is close. Mining equipment manufacturers I've worked for have all used the 75% of proof load as the upper limit for fasteners expected to be re-used. In remote foreign locations (been there) you often just have to run what you've got and keeping that sort of safety factor makes that practice possible. Interesting to compare your test values to the clamp loads shown here.

 

This data definitely helps us understand the superior characteristics of the Toyota 90119-10933 black bolt along with the other bolts tested. And it definitely appears to me that the black bolt is by far the best factory bolt to use on our LBJs regardless of whether the LBJ Protectors are used or not. I used the black bolts when I upgraded my new '97 Taco OEM LBJs to also use the later LBJ Protectors. Bottom line, we are still stuck with the FSM from different years giving us different torque specs for application of the black bolt. The 2001 FSM gives 2 different torque values based on whether or not the LBJ Protectors were used. But the 2003 FSM does not differentiate so gives only the higher torque of 59 ft./lbs. (below). And AFAIK, the LBJ Protector with black bolt attachment was standard by 2003.

2003 Tacoma FSM p246:

 

That table is a good guideline, and is "in the range" of my results. The problem is that we don't know how the 11T Toyota bolts compare to the more traditional 10.9 or 12.9 designations (we'll find out in the destructive test portion). And without testing, we don't know what nut factor (K) to use. That was the point of these tests.

With the data we now have, we know that the nut factor is around 0.15 to 0.2 for all bolts other than black, and about 0.1 for Black. So we can now find the best torque for each of these bolts by just looking at the data - 37 ft-lbs seems perfect for the Black bolt (exactly like the FSM said), 59 ft-lbs (also per FSM) is good for Zinc and Red. If using Green or ARP, I'd bump the torque up to 70 ft-lbs. All of these would give you about 9,000 to 10,000 lbs of preload.

 

I agree, the FSM's are at times inconsistent. I also have a very hard time deciphering which years had which bolts. I am mostly in the 3rd Gen 4Runner world, and they used (4) Green bolts until 2000, then (4) Red bolts (90105-10505) for 09/2000 to 03/2001 models, then (2) Red 90105-10505 and (2) Black 90119-10933 (bolt w/washer) for 03/2001 to 05/2001 (with introduction of the boot), then (4) Black 90119-10933 for 05/2001 to 08/2002.

 

watching this one with a lot of interest. Good info so far.

 

Had to review your thread again since I'm installing new LBJs. Since I am using the 'boot protectors', I did opt to order the black with washer/90119-10933 bolts. I used them the last time as well. Now one difference, I poured over every single tech reference/service bulletin/repair doc I could find on Toyota's TIS. Not a single one mentioned loctite/thread locker. Not even the original recall program's tech install procedure. Last time I installed them, I did use red thread locker but that was the consensus back then.

This time I'm leaning toward NO loctite at all.

One thing worth mentioning (even though its not bolt related), both LBJs were sticky all over right out of the bag. I thought I had a bad batch or something spilled all over them and the parts hub just packaged them up anyway. Wiped them down and now am thinking it might have been some type of anti corrosive or sealant applied (excessively, they were slathered in this stuff) on the part line. The damp paper towel wiped away blue. These were just ordered and are part numbers

$165 43330-39556 Suspension Ball Joint

43330-39556 &

$136 43340-39436 Ball Joint Assembly

43340-39436. Bags both carry the MADE IN JAPAN line and both came with castle nuts and cotter pins.

Hard to see in the picture of the 90119-10933 bolts but they do have a light coating of red on them from the factory.

If there is anything else you learned from the bolts you did test using Loctite, please update. Thanks again for the research Leon, awesome info!

 

Received the new LBJs from a different Toyota Parts Dept (Bosch). Same as the others. Slathered in that whatever it is. Will be using new 90119-10933 bolts at 37 ftlbs w/the dust cover.

 

Sounds like a plan! I'm hoping to finish the destructive torque testing soon, and also the effects of loctite.

 

The destructive tests are done, quite a good workout if I do say so myself.

I need to do some more data reduction, but one thing I can say already is that there will be a BIG surprise for you all (I was surprised too).

Just a teaser for now of the bolt "bloodbath":

 

Good stuff

 

Finally finished the Torque to Failure tests! In this series of tests, we determined the torque values at which the different LBJ bolts reach their yield, ultimate, and fracture loads.

We used the same test setup as for the load vs. torque testing – a tapped stainless steel plate to represent the steering knuckle, a flat stainless steel plate to represent the balljoint flange, and a load cell sandwiched between the two plates to measure the load. A torque transducer was used to measure the applied torque, and a digital multimeter (DMM) was used to read the load cell output. As a bolt was tightened, the display of both the DVM and the torque transducer were recorded and superimposed in the same frame, to provide a direct load vs. torque correlation.

More detail is available at the website at https://spikerengineering.com/lbj-bolts-torque-to-failure-tests, but here are the highlights.

Toyota Green (90080-10066)
The load vs. torque traces for the two Green bolts torqued to failure show that Bolt A reached its ultimate load at 97 ft-lbs (11,666 lbs), and fractured at 108 ft-lbs (10,769 lbs), while Bolt B reached ultimate load at 85 ft-lbs (10,691 lbs), and fractured at 100 ft-lbs (9,442 lbs).

A couple of other elements were added to the plot to illustrate how much margin exists above the Factory Service Manual (FSM) torque value. First, a dotted green line was drawn at 59 ft-lbs to indicate the FSM value of 59 ft-lbs. Then, an attempt was made to identify the torque range at which the bolts yielded (this is considered to be the failure load, rather than ultimate load). Because the yield load is not as easy to pinpoint as the ultimate load, a pink box was drawn to bound the range of estimated yield load torques. These boxes should be viewed only as an approximation.


Toyota Red (90105-10505)
Similar data was plotted for the two Red bolts. Bolt A reached its ultimate load at 92.6 ft-lbs (12,564 lbs), and Bolt B at 93 ft-lbs (12,173 lbs). As for the Green bolt, the FSM torque value of 59 ft-lbs was represented by the green line, and the range of Yield Load torques was approximated by the pink box.


Toyota Black (90119-10933)
Black Bolt A reached its ultimate load at 59 ft-lbs (13,461 lbs), and Bolt B at 67.1 ft-lbs (12,095 lbs). The FSM specifies a torque of 37 ft-lbs (vs. 59 ft-lbs for all other bolts).


ARP (673-1004)
ARP 673-1004 Bolt A reached its ultimate load at 97.7 ft-lbs (14,788 lbs), and Bolt B at 104.5 ft-lbs (13,812 lbs).


Belmetric 10.9 Yellow Zinc (BF10X1.25X30YLW)
The testing started with the same 30mm bolts that were used for the load vs. torque testing. However, the torque to failure tests of these bolts were unsuccessful – some bolts stripped at the threads before reaching the Yield Load.


Only one bolt (Bolt B) could be torqued to failure – it reached an ultimate load of 9,715 lbs at 76 ft-lbs, and broke at 86 ft lbs.


These bolts likely stripped because at 30mm long, they are about 2mm shorter than the Toyota OEM bolts (32mm long), leading to almost two less threads of engagement. This would not be an issue in an actual LBJ installation, but because we had a load cell sandwiched in the load path, there was not enough thread engagement to handle the higher applied loads.

So we procured some longer 35 mm Zinc bolts (BF10X1.25X35YLW) to continue the testing.

Belmetric 10.9 Yellow Zinc (BF10X1.25X35YLW)
The Zinc 35mm Bolt A reached its ultimate load at 170 ft-lbs (12,681 lbs), Bolt B at 150 ft-lbs (10,925 lbs), and Bolt C at 150 ft-lbs (10,535 lbs). The torque transducer that was used for these tests has a maximum limit of 150 ft-lbs, so above those values, we had to rely on clicking of the torque wrench. Therefore, results above 150 ft-lbs may be less accurate.


Because a significant difference in performance was noted between the 30mm and 35mm bolts, the two sets of results were combined into one plot for a closer examination.


It is evident from this plot that the preload generated at the FSM value of 59 ft-lbs was much lower for the 35mm bolts (4,500 lbs) than for the 30mm bolts (8,500 lbs). That’s only about half the preload at the same torque value!

To put this into practical terms, if we torqued a 30mm bolt to 59 ft-lbs, we could expect around 8,500 of load; but if we installed a 35mm bolt and tightened it to the same torque, we could end up with only half the load (4,500 lbs) – and we would not know this, because there would be no load cell in the setup. Also, it is notable that for some bolts, the installation torque of 59 ft-lbs is dangerously close to the yield torque of 70 ft-lbs.

There was significantly more scatter from bolt to bolt within the same batch of 35mm bolts – some bolts reached almost 6,000 lbs at 59 ft-lbs, others only 2,500 lbs. The scatter in 30mm bolts was smaller.

Finally, the 30mm bolts reached their Utimate Load at around 75 ft-lbs of torque; the 35mm needed over twice the torque (150 to 170 ft-lbs) to reach the same Ultimate load.

This tells us that the friction properties of the coating were vastly different between the 30mm and 35mm bolts. This is surprising and disconcerting, because the 35mm bolts came from the same supplier, and from the same family as the original 30 mm bolts, just 5 mm longer:


The 35mm bolts we received appeared to have different head markings than the 30mm bolts, despite being from the same family. While both heads are stamped with 10.9, the different head shape (flat vs. inset) and the different letters indicate that they likely came from different factories and different manufacturers.


Regardless of the reason, the performance of the 30mm and 35mm bolts was different enough to raise concerns. The testing clearly shows that when ordering these generic Class 10.9 zinc bolts, one would have no idea how much preload can be expected when installed to the FSM torque values – it could be as high as 10,000 lbs, or as low as 2,500 lbs. This uncertainty, and the risk of very low preload, would be highly problematic in critical installations such as LBJ.

Summary
The torque to failure data is very useful in determining how close to failure the different bolts tested are when installed per the FSM requirements.

There is one clear loser in this testing, and that is the Zinc bolt. The variability from batch to batch and bolt to bolt leaves us with no idea of how much load is being introduced during torquing, and how close that load is to failure. In retrospect, this perhaps should have been expected. All the other bolts are branded (either Toyota or ARP), with specific part numbers. One would expect that for those bolts, the manufacturer has taken care to control the surface finishes and friction coefficents to serve repeatably in the expected application. In contrast, the Zinc bolts are not branded – they are supplied by a reputable supplier, but there is no control on the coatings and friction, other than to specify “Yellow Zinc” and a strength. Our test results show the danger of using these bolts without testing each batch.

Based on these tests, our recommendation would be to stay away from any unbranded bolt for the LBJ application, no matter what strength class they are.

When it comes to picking the winner, these torque to failure results are not sufficient on their own to declare one. All of the bolts (other than Zinc) reached good preload levels at FSM torque values, and had reasonable margin to failure torque. Their yield strengths were also comparable (though ARP was clearly stronger).

There are many possible criteria for being a “winner” – highest strength, highest preload at FSM value, lowest scatter, highest margin between FSM torque and failure torque, lowest preload loss with ruse, etc. So rather than declaring a single winner, we will look at all the data collected to date, and will try to draw some conclusions about these bolts in our final article.

 

Gimme a tl;dr for why you're testing torque strength instead of shear strength? That pic you posted was shear. All load on the LBJ will be distributed vertically on the mounting plates and horizontally on the bolts (shear).

 

Not sure I understand your question. A torque applied to a bolt creates both shear and axial loads.

 

I just don't really understand the terminology used in this thread maybe.. not my field. From my understanding the failure of these 4 LBJ bolts occurs from a sideways load, not a stretch or compressive load, so what is testing torque limits giving us here? How does "ultimate load" translate to horizontal shear strength?

 

Gotcha. These tests were designed to identify which bolts are the strongest, and torque to failure is a good way to determine that. With this knowledge, you will be able to torque the bolts to the highest possible preload without damaging them.

When properly preloaded, the LBJ bolts do not fail in shear. What does happen is that bolts are inadequately preloaded, and start gapping under repeated loading. That puts the bolts in bending, where they can easily break. So the key is to provide enough clamping load to prevent gapping, and some method to ensure that the bolts will not back out.

 

Good info and looking forward to reading your final word/article. If you can, perhaps touch on the use of threadlocker as well if you have that data. Thanks again Leon, awesome write up.

 

Hmm fascinating, just so happens I have some of those ARP bolts from one of my Nissans too. @SpikerEng and thoughts on using an ARP stud and nut combo instead, my thought was that you'd have 100% thread engagement when torquing. I have many ARP M8 stud, nut, and washer kits kicking around too.

 

So far, it just appears that using the fasteners "as received" or "dry" only resulted in a close, consistent performance at specified torque by the Red and Green Toyota capscrews. It may not be at a high enough grip to do the job at 7500 pounds - but that clamp force was there. I have never believed that Zinc plating was a lubricant and always treated them the same as the black oxide coated parts when it came to clamping loads.

On the Black fasteners - maybe a mechanical effect is in there, caused by the separate washer having an effect on the surface grip of the bolt head on the base plate while torquing? There are many joints where I have specified the use of hardened SAE washers when using high strength fasteners on softer base metals. When those joints were opened - and the washer had been left out by assemblers - the decks were galled. That in itself showed that at least some of the torque was lost just displacing metal.

So is the next test to dip all of them in a known "thread lubricant" like colloidal copper paste and see if that narrows the spread? I'm pretty sure that Loctite isn't a thread lubricant, its just pressure activated glue - so that shouldn't do much.

 

Looking at your empirical data on the black bolts, 37ft./lbs. definitely appears to be the correct torque value for that specific bolt type. When I retrofitted my '97 Taco with the LBJ Protectors, I torqued the black bolts to 59ft./lbs. as listed in the 2003 FSM. That amount of torque puts them squarely in the "YIELD RANGE" which I am not at all comfortable with. So the current over-torqued black bolts will be replaced with new black bolts torqued to 37 ft./lbs. The Taco is on jack stands now, I have a set of new black bolts on hand, so the timing could not be better. We appreciate your data!