Stiffening plate removal / reinstallation

[andrewwynn]

on the 2013, they call it the 'Structural reinforcement panel' and it's part number 31106786568

The bolt is: 33306772888

The stiffening plate on the 2013 model does not also share the bolts with the sway bar which is good, it uses the same diameter bolts just shorter, so i have a strong suspicion that BMW will have spec'd the same rules for install but maybe with different values since it's a shorter bolt (35 v 55 mm).

I haven't found the directive yet but in the link above it should show.

[andrewwynn]

Originally Posted by bcredliner

WARNING---
Do not reuse self locking fasteners. They are designed to be only used once and may fail if reused. Replace with new"

I missed this earlier; that's the NUT they are describing here. So as I surmised, they should also be 'single use' as they are crushed locking nuts def. not meant to be re-used.

[andrewwynn]

Originally Posted by cn90

oldskewel is correct,

I have removed the bolt many times and compare it against a brand-new bolt.
There is no deformity of the old bolt from repeated tightening and loosening.
In other words, the old bolt does not lengthen when compared with a brand-new bolt.

The fact that a bolt is angle-tightened does not mean it is a "use-once-and-discard" type of bolt.

This has been a funny thread, especially if read by BMW or indy mechanics, as they do not replace these bolts.

This will only be the case if the bolt was never fully tightened. The bolts i've taken out and measured are ALL longer than they started and grew each time I tightened. (it will not look longer it's only 0.08mm longer per cycle or so). The picture i took above shows clearly that the bolt is 36.25mm long in the thread count that should be 36.00.

Tightening to spec 4 times plus an unknown times it was done on the vehicle took the bolt to internal failure and one more time broke it in half.

It will not distort below maybe 70-80 N·m which will net you about 2/3 of the design clamping force. Clearly that is enough to not cause cars to suddenly start flipping over, but it also is a lot less than design.

When i reused my bolts i looked up the advised torque for M10 class 10.9 and would have set them to something close to that (67 N·m spec).

Torque'd to normal spec should be ok in my humble opinion, but until somebody actually does the research to determine for sure nobody knows. if the plate moves a tiny bit under certain suspension dynamics that will affect performance to some extent. Is it something anybody will ever notice, that's clearly up for debate.

I like the idea of maintaining the design clamp force but not have to replace the bolts so i'm upgrading to 12.9 bolts to achieve that.

measure the torque it takes to remove your 'seat of pants' tightening, i would be curious; when i removed my wife's they ranged from 68 down to about 20 ft·lb, which I would regard as not nearly ok. (one was 45, also not ok).

[andrewwynn]

The *proper* method to re-use if you can't TTY would be to use a torque wrench and use lubricated torque values for M10; you'll have a decent clamp force and won't damage the bolt, it will be on the order of 7000 vs 9000# clamping force, which clearly has been working 'ok' for 'most everybody'.

(the reason for using lubricated torque, other than the bolt will likely be auto-lubricated on a 17 yr old car, is that without lubrication the torque to clamp force changes each re-use, so you either have to add torque or you'll lose clamp force).

My personal estimation is that the 9000# spec is more than needed and that if everybody just torqued the re-used bolts to normal spec or close, maybe a little higher, but not close to yield.

In my testing, I didn't get to yield until about 105 N·m so it would probably be pretty safe to use a dry torque of 80 N·m or such (again, i plan to experimentally derive some useful numbers).

Here's the idea: until you get to yield there will be a little bit of internal deformation but not much; if you pull to say half the way to yield strength, (10% over normal torque), especially if you start with a new bolt, coincidentally the math just worked out to 54 N·m to achieve that level of tension. (almost identical to the 56 N·m from design spec).

I'm very curious about the 56+45° solution it's a great hypotheses and definitely deserves testing. I'm sold on the just use bigger bolts properly torqued though it's too nice of a coincidence to get the same clamping force as design with normal torque, i call it kismet.

[andrewwynn]

Originally Posted by cn90

Could you do the same measurement for a wheel bolt?
We re-use wheel bolts all the time and wonder what the wheel bolts look like after 15 years of service...

did you see the picture of the lug bolt that was over-tightened? I've recently removed over-torqued lug nuts and bolts that were more than double the spec.

Fortunately there is a decent amount of overhead that a normal person using a lug wrench is not likely to over-torque them but could by stepping on it pretty easily, but with impact it's quite easy.

Advised torque for M14-1.5 10.9 bolt is 202 N·m and BMW spec's 130 on at least my x5, not sure if different on the M sport or E70s.

Each bolt can handle 16,200# force, but that's at full torque, it's 'only' 14,169# per lug bolt at 130 N·m. that is 70,845# per wheel when properly tightened.

yield is 20% over proof so you will damage a lug bolt at about 180 ft·lb, (242 N·m)

[andrewwynn]

On MY car, when i put the bolts back in, i think i used 4 of the original bolts and 2 new SAE 3/8 i had laying around. I can't remember how much i torqued them but will measure when i remove them to get an estimate. I did NOT TTY/TTA as I was unaware of the procedure when I did it, My 'M.O.' however is to look up the torque spec on my reference app and use that value, so that's almost surely what i did.

There is a problem with the 56 N·m plus 45°, which is that at enough applications it will still damage the bolt to the point it can break apart (from a big bump in the road for example); it's not 'mission critical' even if the one of the bolts holding the sway bar breaks (which would likely cascade to the other), you'll know immediately if it happens when you get the 110dB morse code of the sway bar slamming into the subframe. (experimentally confirmed).

What i noticed during my re-tighten to destruction experiment was that the yield not only was met at a lower N·m torque, it ALSO too less degrees of turn, and on the 5th and final tighten it was already at yield by about 45°, now the previous were all 90°, so it would take more times at 45 and if the 45 only gets you from 7500# to 8000# what's the point might as well just stay at proof load and advised torque of 49 ft·lb.

For holding the sway bar the bmw spec torque is actually less strong, since the bolt is pre-stressed to 9000#, it needs less of an impact to hit the tensile strength and break, and if you torqued them normally, it would be a stronger joint.

The TTY is all about the diagonal strengthening of the subframe.

Here's the thing: the front subframe has ZIP for rigidity as a parallelogram as made pretty obvious if you look at the drawing ;



The subframe is like the letter "U" it's open on the rear side and has nothing to keep it from twisting as a parallelogram other than the stiffening plate.

Here is where it comes into play: hit the brakes hard and have the right tires hit a patch of ice with the left tires on dry pavement; with a reasonable braking force of 0.5G now you have about 3000# force pushing backwards on the LEFT side of the car only. This is the condition for which this plate is designed. There are TWO circles about 1 sq in. on the back and one in the front that will get yanked with that force. It would also work to keep the back of the "U" from pulling apart but i'm not sure what forces would get applied there.

I finally remembered exactly what it felt like when i drove my X5 without the subframe stiffener; i likened it to a minivan wobble above but a couple days ago i remembered exactly what it felt like:

When i removed my stiffening plate, it was *exactly* like when i took the top down on my Z28 Camaro! every time i hit a bump i could feel a little extra jiggle like a harmonic; think tuning fork; perfect analogy considering the sub-frame is a big "U" like a tuning fork.

Others may not feel it, but just like I felt the difference every time i drove top-down with the Z, it was immediately obvious to me when i drove the X without the plate. (and oops when i forgot to put the rear bolts in, that was a very loud few miles, it sounded like i was driving a jackhammer).

So, at least a LOT of the mystery is resolved. I don't have the exact specifics but i've found out most of the 'why' and that's what was driving me nuts. I have a plan that suits me and have put out plenty of info that will let others find their comfort zone with respect to how they abuse their bolts. (not meant as an insult, the bolts are abused by design).

-awr

[andrewwynn]

Originally Posted by cn90

In my 1998 528i, FRONT wheel bearings bolts...BMW manual says replace them.

All BMW mechanics I know re-use them. I re-use them, they are as big as the wheel lugs.
That was 10 years ago, the bolts still hold the bearings just fine.

This is a very interesting point and also greatly added to the mass-confusion on the stiffening plate bolts; many of the bolts that are NOT overly stressed are also recommended for replacement, in what seems to be a clear-cut case of CYA; there is no logical reason to replace a 16mm bolt that isn't stressed to deformation.

I had been fairly certain that the bolts weren't stretched until i actually measured a couple of them. When i did the first test of re-tighten to spec and actually measured it grow by 0.08mm then after 4 more it snapped.

FYI; the advised torque for M10-1.5 10.9 is 49 ft·lb = 67 N·m and I am willing to bet that the 56+45° will not cause a loss of strength for well into the double-digits because there is a 20% margin in the design of bolts and that may be right in the middle.

For those like me with a digital torque adapter with live torque reading, it's darn simple to determine when the bolt goes plastic; the torque number just stops climbing it's bizarre! Like i mentioned right after my experiment, I noticed Also since i did one after the other after the other, that each successive time, the angle where the plastic deformation kicked in was sooner at each time, i will pick up one of the TTA tools for my detailed experiment i was only using 'start vertical, end horizontal' with the breaker bar for the initial test.

If you use a torque adapter, and follow the TTA specs, As long as the torque gets to 100 N·m before 90° There is still going to be life in that bolt. One of the bolts I put back in was over 115. I'm going to do a handful of new bolts to get a better scatter plot of what numbers to use, but with the right knowledge you can definitely use the bolts a few times at least.

Too bad the OEM bolts aren't $3.53 like real-oem says. The thing is, even if they were $3.53 + 78¢ (nut) that's still over $25 that need not be spent just to check on the oil leaking status.

I just thought of another biggie; in the example of braking while on ice, it only generated a sheer force of 3000#, in that example but it would be more than that because the lever ends are the tires, the bolts are in the middle there will be leverage factor so it could be double that in moderate braking and quadruple if you get near a 1.0g panic stop.

HOWEVER; bump into something on ONE side of the bumper and now you could easily get into the 10g range; if you put a 10g deceleration on one side of the bumper, now you are putting 50,000# force twisting the car into a parallelogram!

NOW it's coming together to make sense. When i did my early math to determine that the sheet was thick enough to hold back 50,000# of force it felt pretty serious and 'why that much' but it will GREATLY reduce the risk of totaling your car from a fender-bender.

Now i wish i had a couple end-of-life X5 to bump into something solid at 10 mph. The reason people aren't having problems with not tightening their bolts properly is they are good drivers :-D

Get into a fender-bender with your plate removed, there is a better chance you will total your car.

So, it's becoming clear that a large part of the point is to keep the car from twisting into a pretzel from an offset impact, the shear strength of an M10-1.5 is just about 8000#; but hardened aluminum has a tensile strength on the order of 40-50 ksi, so by using a strong 'pinch' they can get 10s of 1000s of # of resistance. just 'snug' that bolt and especially in the case like my wife's where one corner was 20 ft·lb, it doesn't drop to 8000# force, it drops to the sectional size of the cross of the bolt and the aluminum. That works out to 1/16 of a sq. in. so approximately 45/16 = about 4000# force to pull the bolt sideways through the aluminum.

I have to measure the clamping point on the car, but it was maybe 1.75", so that is about .27 sq. in. multiply by 45k and get 12,400#.

These are obviously a lot of educated guesses but they have sound backing in science. basically the idea is that with 9000# of clamp force, the raised donut in the steel about 1.75# in diameter will lock the alum. plate in place so the aluminum would have to tear out sideways it can't slip since it will actually be slightly 'dented in'. Don't have enough force to keep the alum from slipping and then you drop down from 0.27 to 0.062 sq. in. of cross section of alum holding things back. a nearly 80% drop in the capability.

Anybody else starting to see why BMW decided to use a TTY solution to ensure the clamping force? The design (to me is clear now) is to protect the car/frame from damage in an uneven front impact (though would also help from rear impact as well).

I only have gear that can pull 1T, maybe with a sheave (pulley) i can double that, i might see if 2T will pull the plate sideways at what torque on the bolt. I also can't wait to examine the plates on mine and wife's car to see if there is a permanent 1.75" dent from the clamping force. (if I was bmw I would stamp in a dent and then that greatly reduces the need for the extreme clamping force).

There is more function to the plate than above, clearly since i can feel the wobble in the suspension when i don't have the plate on it tightens up the front end, but i'm pretty confident that for that purpose the bolts can be FAR FAR less tight; even if the bolts slipped under stress of cornering etc, they will act like a shock absorber taking energy out of the equation.

So BMW could have just explained in a sentence "we require replacing the single-use bolts to ensure the maximum clamping force required to ensure the design strength is maintained during an accident".

Minor irony alert: the plate makes your life more at risk; anything that makes the car more rigid to protect the car pushes that force onto the contents of the car; deformation of the frame is the best way to save people. So now ya don't know what to do! loosen the bolts to make you safer in an accident or properly tighten them so you don't have to replace your car from a minor bump into a post.

[andrewwynn]

Originally Posted by jcattletown

You guys talk about bolts and nuts. For my 2013 X5 E70, there are no nuts per se. All the bolts go directly to the subframe steel beams. Are these still TTY bolts? Thanks.

It took a few days to track down that answer; they are in-fact TTA/TTY just like the E53 and I believe for the reasons mentioned above; mostly to protect the car from dynamic parallelogram loads. The forces quickly 'go off scale' when even a minor bump is involved.

The full answer is here

The exact same procedure is called for: 56 N·m + 90°

The mechanics, either pro or semi-pro that 'just reuse' without the understanding of why are doing their customers a disservice. That is not professional in my opinion. They don't know why they are supposed to be so friggin' tight and make the assumption BMW just over-engineered it and re-engineered the car to have 80% less capability in that one little thing.

I'm going to see if my local junk yard has any crashed E53 and take some measurements. It's not a stretch of the imagination to say that any crashed 15 year old X5 has almost certainly got re-used under-torqued bolts, i can even test for that if i can find one.

If my contention is valid, I should find some parallelogram shift on just about any offset front impact X5 (with the expectation the bolts have been reused and not properly torqued).

Any holes in my logic guys? It does happen, but I think i've stumbled into some solid arguments and basically 'the only reason' for this thing to be so damn strong. I was confused when i came up with 3000# twisting force from a real-world driving example, but when i looked at the diagram of the front subframe and saw it looks like a "U", things started to become clear.

[andrewwynn]

Originally Posted by Freestyler

That prediction that made you do that edit, is spot on.

It is only the bolts that are supposed to strech when tightened to torque spec. not the nuts. So i would assume that you could re-use the nuts.
That been said, if i were convinced that i needed to replace the bolts, i would go all the way and replace the nuts too.

I thought it might be interesting to go back to the first posts of this thread, and found this query. I found a specific directive from BMW during my search that mentions the NUTS are also SINGLE USE.

paraphrasing: "since the nuts are self-locking type, they can't be trusted to engage and lock on reuse"

however, there is a more important reason, covered by this document from Fastenal, that describes how the NUT stretches during installation and the threads engage differently the second time. (this greatly throws off the torque to tension numbers).

So IMHO, it's actually more important to replace the nut than the bolt (at least for the first five uses; I would not use the bolt more than 5 times, and I will be actually using 12.9 bolts that don't need to be stretched to yield to get the same force).

[andrewwynn]

15 pack of clone bolts are here

The first few threads of the nut will support the majority of the load. Research has shown in some cases involving UNC threaded nuts that the first thread will have to support nearly 35% of the load. The second thread will support about 25% of the load, and the third thread about 18%. In this case the first three threads support 78% of the load.
To allow this distribution, nut threads are designed to be softer than bolt thread and will conform to the contour of the bolt threads when tensioned. If a nut were reused, there would no longer be a "ideal" thread match. This will create more friction between the threads during installation, which will significantly alter the installation torque.

So, probably should replace the nut even if you keep the bolt or it throws off the torque to tension ratio dramatically.

n a demonstration with a 1/2-13 zinc plated SAE J429 Grade 5 hex cap screw and zinc plated SAE J995 Grade 5 hex nut with an installation torque of 70 ft-lbs to obtain a clamp load of 9000 lbs (without any added lubrication). On the second installation, this torque had increased to 95 ft-lbs to obtain 9000 lbs. By the fourth installation, we required 145 ft-lbs to reach a clamp load of 9000 lbs.

So it would seem the factor that changes the torque required the most is the nut. Learning new things every day. So, i'm thinking i'll be using NORMAL nuts and locktite and i'll replace the nuts rather than lubricate them, but the final solution will depend on the with or without lubrication and using new nuts.

On the other hand, i just found the T-slot nuts that are 12.9 to match my upcoming 12.9 bolts, i'm hoping i can get them to fit up top without room to turn so they are self-stopping no need to reach up to hold the nut.