Weird oil level sensor issue

[andrewwynn]

"supposed to" different from "must do".

If you torque to spec the bolt well break after about 4-5 cycles. It will be noticeably damaged by the third.

However if you pre tension to say 50 N·m and torque to angle 45° vs 90° (I think 65 N·m + 90° is spec), you will achieve something like 80% of the spec clamping force without causing any additional damage to the bolt. (reminder bolt is damaged at the factory on purpose you can not undo torque to yield stretching)

Best to get some 12.9 bolts and torque to spec. Bolts will never wear out and will have similar torque as the 10.9 torque to yield

[jac]

Nice alternative! Thanks! :-)

I did see the https://www.fastenal.com/ and www.belmetric.com suppliers for the replacement M10-1.5x55 10.9 bolts/nuts on those other threads. One of them is local to me so I plan on getting one or two sets of hardware.

[andrewwynn]

Fastenal no longer allows off the street customers, you need to have a business account (though you can buy for local store delivery on the website)

[Fifty150hs]

Originally Posted by Overboost

Yes sir, I'm glad I checked. It's funny how you get used to the BMW nannies letting you know something needs attention and you stop checking things manually. That won't happen again.

I have the opposite failure mode. It will randomly tell me to check my oil level and when I do it's fine. Scanner says the sensor is bad. Replacing it Presidents day weekend along with engine mounts and sway bar bushings. Oil change too of course.

[Fifty150hs]

Originally Posted by jac

Nice alternative! Thanks! :-)

I did see the https://www.fastenal.com/ and www.belmetric.com suppliers for the replacement M10-1.5x55 10.9 bolts/nuts on those other threads. One of them is local to me so I plan on getting one or two sets of hardware.

I ordered 12.9 from Belmetrics. For $20 I can replace them every time if I want to. Beats the hell out of $150 OE replacements.

[andrewwynn]

That's what I think too but I haven't done the test and the math to determine the torque method.

10.9 @ 49 ft·lb = 7500#
Pushed to yield with 90°attritional degrees = 136*0.0888 12,000#

0.0888 is the tensile area in in.²

Spec on 12.9 is 8900# basically infinitely reused but will take more torque reach time to get to the same torque level so a method needs to be developed to get a reasonable compromise.

Bolts are spec'd to 75% of their no damage resuse so the 8900 spec means 12.9 can be loaded to 8900/.75=11866.667

Pretty stunning how close a 12.9 without damage almost identical to the yield of 10.9.

Yield on 12.9 is 160*0.0888=14,208

I would have to find a spring modulus of 12.9 to figure out the best extra angle to get to something between 10-12.000# clamping force and no damage.

That said, unlubricated torque spec is 60 ft·lb (you may have to brake clean them on reuse if your stiffening plate is as oily as mine).

That gets you to 9000#

To get to 12000# we need to stretch the bolt.

5cm almost exactky 2"

(12000*2)/(30000000*0.0888)=0.009

A stretch of 9/1000 on a 50mm M10 will get you 12000#

0.009*25.4=0.23 mm.

.23/1.5 =0.153*360=55.08 degrees


I just looked up the spec for the 10.9 to yield and it is 41 ft·lb plus 90°

41 ft·lb is 6275#
90/360*1.5=0.375mm stretch
(Way past yield: if bolt didn't yield you'd have closer to 20,000#)

The entire point of torque to yield is your are using the bolt as an automatic clamp force limiter.

So if we pre-load to 49 ft·lb, that will get to to 7500#. We need to add 4500# by torque angle

(4500*2)/(3000000*0.0888)=0.034

(Clamp force/ length)/(Young's modulus*bolt area)= inch of stretch

So 3.4mil of stretch will add 4500# let's see if that will damage the bolt.

0.0034*25.4=0.086/1.5*360=20.64

Preload of 51 ft·lb and 20° of additional angle works out to :

20/360=0.056*1.5=0.084/25.4=0.003

0.003*(30000000*0.0888)/2=3996

A little bit of rounding puts us at 7600# at 50 ft·lb preload and 20° additional adds 4000# =11600.

11600/0.0888=130630 psi

Proof load of 12.9 is 135000 that is just about perfect "normal use".

20° not a ton of angle but 22.5 is half of 45° I think you'd be able to eyeball that and be pretty darn close.

Make sure you use washers or flange bolts on the aluminum.

The bolts are on the order of 160ksi but the aluminum plate closer to 35ksi you don't want to simply broach a bolt head shaped hole through the plate.


22.5/360*1.5/25.4=0.004

.004*30000000*0.0888/2=5328

12000-5328=6672

22.5° is 5328# meaning preload of 6672.

6672/7531*49=43.411

So if you bump the preload down to 43 ft·lb you should get very close to 12,000# goal and half of 45° torque angle.

12.9 is 160ksi = 160*0.088=14.08k# yield so 12k is 85%. You should be able to reuse dozens of times with no ill effects.

You have to use a preload + angle solution because the threads of the nut and bolt meld a bit each use. In testing, when the goal was 11000 clamp force with a 1/2" bolt it took about 90 ft·lb the first time, 105 the second, 120 the third.

Don't be surprised when the torque required gets harder each time; when I did the destructive testing, on the third test, the torque required to get to 90° was over 100 N·m (75 ft·lb) but by the 4th and 5th the torque hit that binder l number at like 60 then 45° and was so strange when the torque meeter just stuck at one number as I turned the bolt.

The bolt snapped on the fifth reuse (6th presumable tightening including factory). They got spongy on the 3rd reuse. They felt quite normal on the 2nd re use (3rd tightening including factory)

So, you could use them to full strength as couple times, and if you use a digital torque wrench you can pretty easily tell when the bolt goes plastic and would have to bed retried.

I really like how the 12.9 solution works but it's not as perfect as torque to yield in that the clamping force will probably be more like 12,000 +1000 -2000 after they are used multiple times.

Compare to the usual DIY mechanic that just snugs them up and the clamping force drops by 9000#.

[Fifty150hs]

Originally Posted by andrewwynn

That's what I think too but I haven't done the test and the math to determine the torque method.

10.9 @ 49 ft·lb = 7500#
Pushed to yield with 90°attritional degrees = 136*0.0888 12,000#

0.0888 is the tensile area in in.²

Spec on 12.9 is 8900# basically infinitely reused but will take more torque reach time to get to the same torque level so a method needs to be developed to get a reasonable compromise.

Bolts are spec'd to 75% of their no damage resuse so the 8900 spec means 12.9 can be loaded to 8900/.75=11866.667

Pretty stunning how close a 12.9 without damage almost identical to the yield of 10.9.

Yield on 12.9 is 160*0.0888=14,208

I would have to find a spring modulus of 12.9 to figure out the best extra angle to get to something between 10-12.000# clamping force and no damage.

That said, unlubricated torque spec is 60 ft·lb (you may have to brake clean them on reuse if your stiffening plate is as oily as mine).

That gets you to 9000#

To get to 12000# we need to stretch the bolt.

5cm almost exactky 2"

(12000*2)/(30000000*0.0888)=0.009

A stretch of 9/1000 on a 50mm M10 will get you 12000#

0.009*25.4=0.23 mm.

.23/1.5 =0.153*360=55.08 degrees


I just looked up the spec for the 10.9 to yield and it is 41 ft·lb plus 90°

41 ft·lb is 6275#
90/360*1.5=0.375mm stretch
(Way past yield: if bolt didn't yield you'd have closer to 20,000#)

The entire point of torque to yield is your are using the bolt as an automatic clamp force limiter.

So if we pre-load to 49 ft·lb, that will get to to 7500#. We need to add 4500# by torque angle

(4500*2)/(3000000*0.0888)=0.034

(Clamp force/ length)/(Young's modulus*bolt area)= inch of stretch

So 3.4mil of stretch will add 4500# let's see if that will damage the bolt.

0.0034*25.4=0.086/1.5*360=20.64

Preload of 51 ft·lb and 20° of additional angle works out to :

20/360=0.056*1.5=0.084/25.4=0.003

0.003*(30000000*0.0888)/2=3996

A little bit of rounding puts us at 7600# at 50 ft·lb preload and 20° additional adds 4000# =11600.

11600/0.0888=130630 psi

Proof load of 12.9 is 135000 that is just about perfect "normal use".

20° not a ton of angle but 22.5 is half of 45° I think you'd be able to eyeball that and be pretty darn close.

Make sure you use washers or flange bolts on the aluminum.

The bolts are on the order of 160ksi but the aluminum plate closer to 35ksi you don't want to simply broach a bolt head shaped hole through the plate.


22.5/360*1.5/25.4=0.004

.004*30000000*0.0888/2=5328

12000-5328=6672

22.5° is 5328# meaning preload of 6672.

6672/7531*49=43.411

So if you bump the preload down to 43 ft·lb you should get very close to 12,000# goal and half of 45° torque angle.

12.9 is 160ksi = 160*0.088=14.08k# yield so 12k is 85%. You should be able to reuse dozens of times with no ill effects.

You have to use a preload + angle solution because the threads of the nut and bolt meld a bit each use. In testing, when the goal was 11000 clamp force with a 1/2" bolt it took about 90 ft·lb the first time, 105 the second, 120 the third.

Don't be surprised when the torque required gets harder each time; when I did the destructive testing, on the third test, the torque required to get to 90° was over 100 N·m (75 ft·lb) but by the 4th and 5th the torque hit that binder l number at like 60 then 45° and was so strange when the torque meeter just stuck at one number as I turned the bolt.

The bolt snapped on the fifth reuse (6th presumable tightening including factory). They got spongy on the 3rd reuse. They felt quite normal on the 2nd re use (3rd tightening including factory)

So, you could use them to full strength as couple times, and if you use a digital torque wrench you can pretty easily tell when the bolt goes plastic and would have to bed retried.

I really like how the 12.9 solution works but it's not as perfect as torque to yield in that the clamping force will probably be more like 12,000 +1000 -2000 after they are used multiple times.

Compare to the usual DIY mechanic that just snugs them up and the clamping force drops by 9000#.

Great info andrewwynn! My thinking is that for $20 I can buy new ones every time and not have to worry about the clamping force changing with use. I've put 100,000 miles on my X and never removed the stiffening plate. I don't know if it was removed before which is one reason I will replace the bolts. The likelihood of needing to remove it again for me is low.

[andrewwynn]

The need to remove grows exponentially nearing 20 years/ 200,000 miles. When you can buy normal bolts for $20 not a bad plan but when a set is $120 that’s just crazy talk.

[Fifty150hs]

Originally Posted by andrewwynn

The need to remove grows exponentially nearing 20 years/ 200,000 miles. When you can buy normal bolts for $20 not a bad plan but when a set is $120 that’s just crazy talk.

I'm at 245,000 so I guess I've been on borrowed time.

[aureliusmax]

How did this thread get de-railed to sharing opinions about fastener theory ?

I'm concerned about determining how the low oil level warning system operates. I replaced my oil level sensor, and tried to test it by filling up the oil sump, and with the ignition to ON draining it all out. The warning never came on once, leaving me wondering what happened. And how I can bench-test my old sensor to determine if it's working as I kept it.