It's going to be easier to measure compression in a manufactured wood material (as imperfect as it is) and work backward from there (action=reaction) then to actually measure any actual micro changes in the metal since obviously this is huge dollars since to do it right really requires dynamic analysis with lots of sensors to determine deformation zones.
Actually a better item would be to compress against springs with predictable compression rates. With known compression force you have to have the same tension in the bolt.
But I'm not talking about cranking something to 500 ft-lbs either. I'm talking about 20 to 80 ft-lb ranges.
My "acid test" on whether something is too tight is when you see thread deformation or damage (stripping). As long as I don't screw up threads, I'm really not that concerned about exerting the maximum tensile strength of a bolt. Screwing up a thread often is worse than even breaking a bolt in the hole.
I've never even come close to having a problem with lubing threads and then being able to take things apart later. I grew up in the rust belt and it's often a huge headache to remove what's left after you broke something that was frozen. I still do the same thing in California because it works. At the very least I put on something that's going to displace moisture (like WD-40 or the like.)
Too, having been in engineering I realize that a spec is only as good as the motives of the designer. Could be the reason that something is specked a certain way is only because of rule of thumb or an arbitrary value that was in an appropriate range for the fastener size.
But you're right engineering is fun. It's all about figuring out how to get from point a to point b successfully and with safety factors built in for errors in design calculations and errors in actual application. Rules of thumbs go a long way though. They became rules because they work over time.
Alan
Quote:
Originally Posted by DJ
Try putting it through a piece of steel. Make sure it fits the hole quite well, use hard (grade 8) washers under the head and the nut, and make sure the mating surfaces are flat and burr-free. Make sure the bolt is only slightly longer than necessary, perhaps two threads or so.
Then, use a micrometer to measure the length of the bolt at a given tightening torque on the nut. Lubricate the threads and do it again. The additional stretch of the bolt with the lubricant on the threads is a sign of additional tensile stress on the bolt.
If you want to have even more fun, look up the strength of the steel bolt, compute its Hooke's Law constant, and compute the additional stress in psi. Finally, measure the decrease in thickness of the steel the bolt goes through and take that into account.
That's what the project I talked about did, over and over. I wasn't there -- I heard it 2nd hand. Their goal was to tighten the bolt to a specified stress, which they determined by measuring its elongation. They found that they could not depend on the tightening torque at all.
My "acid test" on whether something is too tight is when you see thread deformation or damage (stripping). As long as I don't screw up threads, I'm really not that concerned about exerting the maximum tensile strength of a bolt. Screwing up a thread often is worse than even breaking a bolt in the hole.
I've never even come close to having a problem with lubing threads and then being able to take things apart later.
[...]
A wheel stud is a particularly hard bolt. A hard bolt can break from brittleness long before the threads strip. Again, my concern is not that you can't take it apart later or that it will come loose, it is that it could break, without deformation, because of being overstressed.
Part of this comes from an experience I had when I was about 16. I was driving my mother's "second car" (which my two brothers and I all drove as our "first" car). A rear axle seal began leaking and I had it repaired at the local Sears shop. The buffoon who put the wheel on used a 4-way lug wrench to tighten the lug nuts. He put all five on finger tight, then put the lug wrench to one and just cranked on it until the lug snapped. To say that I ate his arse out is an understatement. He repaired it at no expense.
The amazing result was a clean break through the stud just inboard the nut. The stud itself was clean and dry, and I simply unscrewed the lug nut from the snapped-off piece. So, there was no deformation of threads under the nut, just a clean overstress of the stud until it broke, and he did it just by cranking on a lug wrench.
So, with lubricant on the threads, it takes a whole lot less torque to put the same stress on the bolt. Do you see my concern? It's your backside at risk, not mine, and you really don't know just how much additional stress it puts on the bolt in your case. So, the fact that it hasn't failed looks like success, but I wonder what your safety margin is? Perhaps the risk is small, but perhaps it's not. Absent real measurements, it's not engineering, it's just guesswork. That's why I said that I hope your luck holds up.
With only a couple of exceptions, I've done all my own service and repair work since then. This episode is the reason why.
It's going to be easier to measure compression in a manufactured wood material (as imperfect as it is) and work backward from there (action=reaction) then to actually measure any actual micro changes in the metal since obviously this is huge dollars since to do it right really requires dynamic analysis with lots of sensors to determine deformation zones.
Actually a better item would be to compress against springs with predictable compression rates. With known compression force you have to have the same tension in the bolt.
[...]
I think you're going the wrong direction. You would not be simulating the condition that occurs on the vehicle.
The idea is not to produce the same tension in the bolt with less torque, it is to produce higher tension in the bolt with the same torque. If the higher tension can lead to thread deformation or shear, then the simulation has to work at that higher tension, not lower tension. Your suggested simulation leads away from that, not toward it. That's why I suggested using steel.
It doesn't take any expensive equipment, just the steel, the bolt, nut, and washers, the torque wrench, oil or anti-sieze, and a caliper or micrometer to measure the length of the bolt.
I think you're going the wrong direction. You would not be simulating the condition that occurs on the vehicle.
The idea is not to produce the same tension in the bolt with less torque, it is to produce higher tension in the bolt with the same torque. If the higher tension can lead to thread deformation or shear, then the simulation has to work at that higher tension, not lower tension. Your suggested simulation leads away from that, not toward it. That's why I suggested using steel.
It doesn't take any expensive equipment, just the steel, the bolt, nut, and washers, the torque wrench, oil or anti-sieze, and a caliper or micrometer to measure the length of the bolt.
Deformation is a pretty complex subject. There is no way that I'd have the tools to accurately measure where the bolt was deforming - if it actually did (which is a huge assumption). I'd actually think that it would/should not except temporarily. If it does to a degree that I can measure with the crudest of mechanical measuring tools then I'd have an issue. Unless I used a bolt that was 10 times longer than the actual use length so that any stretch would be exagerated time 10. Deformation is non-linear so I'd have to be able to track it in all the zones as it happened.
But then again, I'm not concerned with deformation, I'm just comparing tension A vs tension B with corresponding torques. I don't need to know the tension if I know the resulting compression. One has to equal the other.
But your experience of the snapped stud is pretty common and I have seen the same thing. I'm not sure if they case harden the studs. I would imagine they do. I do know they do it for many type of bolts after cold rolling the threads. They usually don't just cut threads in raw material and call it a day for technical applications.
In general, you don't want brittle material for structural members subject to shock where you don't have to - for example it's NOT good to use a class 8 bolt where a class 6 would be fine. Sure the former will be stronger in absolute terms, but they'll be more prone to catastrophic failure as in your stud example. Whereas a class 6 would just stretch.
I'm not sure of the equivalent class of bolt for the studs, but that would tend to describe it's behavior and I'm sure the manufacturers could say what they were.
I'm more along the lines to see what what kind of pressure is generated with dry threads vs lubricated thread by examining the reaction in another material since it would be too hard to do it directly (and be as accurate as I would like to be). Wood might be too soft but it would exagerate the pressure difference. Not sure how much squeeze 80 lbs in a screwing motion would exert.
As I think about it, I could try it on a brass plate or washer since the deformation rates would be more predictable. Steel might be too hard.
Any test I can do at home is going to be more subjective than objective and I'm not kidding myself about that. A whole lifetime could be spent on accurate measuring and testing and people do exactly that.
All I'm looking for is equivalents and determine base "rules". I do the same thing when I extend my oil intervals. I extend because I'm able to determine my own rules for *my* oil usage - regardless as to what the manufacturer tells me to do. I'm happy with my results and I'm the one that's going to have to live with them, not the manufacturer.
For example: I believe that if I were to measure the number of turns of a nut to get to a 60 ft-lb torque, it would probably be different if did it "all at once" (completely dynamic friction) vs snugging to 50, then to 55, then finally to 57 and then to 60. Maybe it's the same, but due to static and dynamic friction differences I think there would be a measurable differerence. In the end the number of turns should be exactly the same regardless as to which method you use, but I don't think they would be.
So how you get to a torque is as important as the absolute value. Certainly if I'm deforming there is difference because now time comes into play. It may be a while until I get to a steady state and any deformation ceases. That could be 24 hours, maybe longer. Could be 24 milliseconds.
Actually, now that I think about it, tightening on a brass plate should give me a pretty accurate sample of equivalent torques. Equivalent deformation should give me an indication of equivalent torques to get to the same deformation (of the brass). I would hope any deformation in the fastener (if any) is purely from normal elastic behaviour.
If the ranges vary by as much as 5 times as your buddies found out, then I'd be comparing apples to oranges since permanent deformation is occuring and the whole test is useless (other than to find out I'm comparing apples to oranges. There has to be some extremes coming into play that you didn't mention to get that kind of a difference.
If the ranges vary by as much as 5 times as your buddies found out, then I'd be comparing apples to oranges since permanent deformation is occuring and the whole test is useless (other than to find out I'm comparing apples to oranges. There has to be some extremes coming into play that you didn't mention to get that kind of a difference.
Alan
I think the situation is a whole lot simpler than you do.
Part of the torque you apply to the nut overcomes friction between the nut and the threads, and the rest of the torque produces tension in the bolt by stretching it. Remember Hooke's Law?
The point is that the oil or other lubricant on the threads reduces the friction between the nut and the threads, and so, if there is a lubricant on the threads, then a given torque on the nut will produce more stretch of the bolt and so a higher tension in the bolt. The "extremes" is simply the tremendous reduction in friction between the nut and the threads that the lubrication can provide. The surprise is not that the lubrication makes a difference, rather the surprise is that the difference it can make is so large.
I just did a short and simple experiment that I suspect you'll find quite interesting.
I rummaged through my bolt bucket and came up with a square head grade 5 bolt, with 3/8 x 16 threads, about an inch and a half long. The nut I found is softer, but I don't know what grade it is. I selected it because it is unused and has a square head, which is suitable for chucking in a vise. Then I found a stack of 3/8" washers such that, with all the washers and the nut on the bolt, there was about one thread showing. I cleaned the threads of the bolt and the nut, and cleaned the washers, all with solvent, then I cleaned off the solvent with Birchwood Casey Gun Scrubber (which is a REALLY good degreaser), and finally blew it dry.
I began by measuring the length of the bolt with calipers. The head is slightly rounded and the thread end is uneven, so I measured from the round head to the "high spot" at the end, with the shaft parallel to the caliper axis. I was careful.
Then I put the head in a vise, stacked on the washers, and tightened it to 55 ft-lb. The torque wrench clicked quickly and easily.
Then I measured the length again, using the same procedure as before.
Then I removed the nut and washers and measured it again.
Then I stacked on the washers, oiled the threads of the nut and the stud using 3-in-1 oil, and torqued it again. Actually, I "tried" to torque it again. I could not get it to 55 ft-lb. It felt "soft", as if the threads were stripping. So, I backed off the torque wrench to 20 ft-lb and tried again. I increased the torque wrench setting until I couldn't get it to click, which happened at 41 ft-lb.
There are TWO interesting results.
First, here are the length measurements:
1.523" starting
1.526" at 55 ft-lb with dry threads
1.523" after relaxing
1.529" at 41 ft-lb with lubricated threads
As you can see, the results with oil was TWICE the stretch of the bolt, meaning TWICE the internal tension, with 25% LESS torque on the nut.
Second, the nut and bolt are in fine shape. The "softness" I felt was the washers being compressed and pushed onto the square head of the bolt, meaning it was trying to pull the square head of the bolt through the round hole of the washer. There is 0.243" of thread now showing past the nut, which means that the additional bolt stretch was caused by less torque using less of the bolt length. Unfortunately, I can only approximate the amount of thread showing when it was tight with dry threads, as I didn't measure it and now the washers are buggered. My best estimate is about 0.08" was showing.
Here's my simple analysis:
The bolt length that was elongated with dry threads was 1.523" - 0.080" = 1.443". The elongation was 1.526" - 1.523" = 0.003".
The bolt length that was elongated with oily threads was 1.523" - 0.243" = 1.280". The elongation was 1.529" - 1.523" = 0.006".
The "stress improvement ratio" due to the use of oil is computed as the ratio of "stretch per unit length per unit torque" of one case vs the other case:
Your mileage may differ, but I think that's quite significant. Ordinary 3-in-1 oil on the threads produced three times the stress in the bolt, other things being equal, compared to clean, dry threads.
I've attached a picture of the result.
Addendum: My original post had a really dumb goof in the computation. Since nobody had posted past it yet, I edited it. Sorry, it's late and I'm tired. Mea culpa.
Though I don't buy the three times assumption as deformation isn't necessarily linear so you can't assume the tension is linear. That's the problem with measuring length. In this case it *might* be correct to assume the deformation linearly applies to tension but often and more likely usually, that's not the case. Only the bolt manufacturer is going to know.
In general the proof load characteristics for a 3/8ths bolt (16tpi) with a .0775 in. squared cross sectional area measured at the minor diameter:
Class 2 bolt
Nonetheless, with 3in1 in the tension is definitely more than without it. I don't know that you can accurately quantify how much more, but it's more. You find the equivalent dry torque that did the same damage as the wet torque.
That's a pretty good shot though. I'd be curious as to your results with a silicon based lubricant and an aluminum based anti-seize that is applied with the excess wiped off the bolt threads.
Too, I'd be curious as to how the spring quotients that you observed changes after work hardening.
So you'd have to run 9 tests in all.
Alan
Quote:
Originally Posted by DJ
Hey Alan:
I just did a short and simple experiment that I suspect you'll find quite interesting.
I rummaged through my bolt bucket and came up with a square head grade 5 bolt, with 3/8 x 16 threads, about an inch and a half long. The nut I found is softer, but I don't know what grade it is. I selected it because it is unused and has a square head, which is suitable for chucking in a vise. Then I found a stack of 3/8" washers such that, with all the washers and the nut on the bolt, there was about one thread showing. I cleaned the threads of the bolt and the nut, and cleaned the washers, all with solvent, then I cleaned off the solvent with Birchwood Casey Gun Scrubber (which is a REALLY good degreaser), and finally blew it dry.
I began by measuring the length of the bolt with calipers. The head is slightly rounded and the thread end is uneven, so I measured from the round head to the "high spot" at the end, with the shaft parallel to the caliper axis. I was careful.
Then I put the head in a vise, stacked on the washers, and tightened it to 55 ft-lb. The torque wrench clicked quickly and easily.
Then I measured the length again, using the same procedure as before.
Then I removed the nut and washers and measured it again.
Then I stacked on the washers, oiled the threads of the nut and the stud using 3-in-1 oil, and torqued it again. Actually, I "tried" to torque it again. I could not get it to 55 ft-lb. It felt "soft", as if the threads were stripping. So, I backed off the torque wrench to 20 ft-lb and tried again. I increased the torque wrench setting until I couldn't get it to click, which happened at 41 ft-lb.
There are TWO interesting results.
First, here are the length measurements:
1.523" starting
1.526" at 55 ft-lb with dry threads
1.523" after relaxing
1.529" at 41 ft-lb with lubricated threads
As you can see, the results with oil was TWICE the stretch of the bolt, meaning TWICE the internal tension, with 25% LESS torque on the nut.
Second, the nut and bolt are in fine shape. The "softness" I felt was the washers being compressed and pushed onto the square head of the bolt, meaning it was trying to pull the square head of the bolt through the round hole of the washer. There is 0.243" of thread now showing past the nut, which means that the additional bolt stretch was caused by less torque using less of the bolt length. Unfortunately, I can only approximate the amount of thread showing when it was tight with dry threads, as I didn't measure it and now the washers are buggered. My best estimate is about 0.08" was showing.
Here's my simple analysis:
The bolt length that was elongated with dry threads was 1.523" - 0.080" = 1.443". The elongation was 1.526" - 1.523" = 0.003".
The bolt length that was elongated with oily threads was 1.523" - 0.243" = 1.280". The elongation was 1.529" - 1.523" = 0.006".
The "stress improvement ratio" due to the use of oil is computed as the ratio of "stretch per unit length per unit torque" of one case vs the other case:
Your mileage may differ, but I think that's quite significant. Ordinary 3-in-1 oil on the threads produced three times the stress in the bolt, other things being equal, compared to clean, dry threads.
I've attached a picture of the result.
Addendum: My original post had a really dumb goof in the computation. Since nobody had posted past it yet, I edited it. Sorry, it's late and I'm tired. Mea culpa.
Though I don't buy the three times assumption. Deformation isn't necessarily linear once you hit the plastic region so you can't assume the tension is linear. That's the problem with measuring length. In this case it *might* be correct to assume the deformation linearly applies to tension but often and more likely usually, that's not the case. Only the bolt manufacturer is going to know. I wouldn't assume it, but I'm generally conservative on assumptions and facts I don't know to specifically be true. Generally, a 0.2% offset from the linear region is considered plastic deformation and you clearly hit that.
FYI, In general the standard SAE proof load characteristics for a 3/8ths bolt (16tpi) with a .0775 in. squared cross sectional area measured at the minor diameter:
(With 24 tpi specs being a bit higher. Almost 1,500 lbs higher for grade 8 tensile)
At these known limits you could indeed calculate the stress in the bolt. Other than that, good luck....... But you could work backward from a stress vs. strain chart. Since you can measure the strain. It's obviously a lot easier to work with "real" tools. By real, I mean electronic.
Nonetheless, with 3-in-1 in the tension is definitely more with than without it. I don't know that you can accurately quantify how much more, but it's more. You find the equivalent dry torque that did the same damage as the wet torque.
That's a fairly decent shot though. I'd be curious as to your results with a silicon based lubricant and an aluminum based anti-seize that is applied with the excess wiped off the bolt threads. The other thing I'd be interested in is if the bolt you had indeed meet grade 5 specs. But again, it's certainly a good whack at it. Not enough for me to change my ways since I've not seen a failure, but still good.
Too, I'd be curious as to how the spring quotients (hooke's law) that you observed changes after work hardening. You'd have to bump down the torques and chart it before and after the plastic deformation which would likely work harden but thin the cross section.
So you'd have to run 9 tests in all. You've piqued my curiosity though.
Personally, I've never had and issue on anything and that includes a 30 year old vehicle that has seen so many miles I'll have to retire it due to rust finally wearing through thick major metal components (like the frame). I have only used anti-seize on lug nuts and spark plugs and I just coat the threads and remove the excess. But I use WD-40 on most threads.
Here's something I came across for a like mind that "really want's to know". I mostly use a light lube like wd-40 or the like to lube threads\displace moisture. But it supports your contention (sort of.....)
I just did a short and simple experiment that I suspect you'll find quite interesting.
I rummaged through my bolt bucket and came up with a square head grade 5 bolt, with 3/8 x 16 threads, about an inch and a half long. The nut I found is softer, but I don't know what grade it is. I selected it because it is unused and has a square head, which is suitable for chucking in a vise. Then I found a stack of 3/8" washers such that, with all the washers and the nut on the bolt, there was about one thread showing. I cleaned the threads of the bolt and the nut, and cleaned the washers, all with solvent, then I cleaned off the solvent with Birchwood Casey Gun Scrubber (which is a REALLY good degreaser), and finally blew it dry.
I began by measuring the length of the bolt with calipers. The head is slightly rounded and the thread end is uneven, so I measured from the round head to the "high spot" at the end, with the shaft parallel to the caliper axis. I was careful.
Then I put the head in a vise, stacked on the washers, and tightened it to 55 ft-lb. The torque wrench clicked quickly and easily.
Then I measured the length again, using the same procedure as before.
Then I removed the nut and washers and measured it again.
Then I stacked on the washers, oiled the threads of the nut and the stud using 3-in-1 oil, and torqued it again. Actually, I "tried" to torque it again. I could not get it to 55 ft-lb. It felt "soft", as if the threads were stripping. So, I backed off the torque wrench to 20 ft-lb and tried again. I increased the torque wrench setting until I couldn't get it to click, which happened at 41 ft-lb.
There are TWO interesting results.
First, here are the length measurements:
1.523" starting
1.526" at 55 ft-lb with dry threads
1.523" after relaxing
1.529" at 41 ft-lb with lubricated threads
As you can see, the results with oil was TWICE the stretch of the bolt, meaning TWICE the internal tension, with 25% LESS torque on the nut.
Second, the nut and bolt are in fine shape. The "softness" I felt was the washers being compressed and pushed onto the square head of the bolt, meaning it was trying to pull the square head of the bolt through the round hole of the washer. There is 0.243" of thread now showing past the nut, which means that the additional bolt stretch was caused by less torque using less of the bolt length. Unfortunately, I can only approximate the amount of thread showing when it was tight with dry threads, as I didn't measure it and now the washers are buggered. My best estimate is about 0.08" was showing.
Here's my simple analysis:
The bolt length that was elongated with dry threads was 1.523" - 0.080" = 1.443". The elongation was 1.526" - 1.523" = 0.003".
The bolt length that was elongated with oily threads was 1.523" - 0.243" = 1.280". The elongation was 1.529" - 1.523" = 0.006".
The "stress improvement ratio" due to the use of oil is computed as the ratio of "stretch per unit length per unit torque" of one case vs the other case:
Your mileage may differ, but I think that's quite significant. Ordinary 3-in-1 oil on the threads produced three times the stress in the bolt, other things being equal, compared to clean, dry threads.
I've attached a picture of the result.
Addendum: My original post had a really dumb goof in the computation. Since nobody had posted past it yet, I edited it. Sorry, it's late and I'm tired. Mea culpa.
Though I don't buy the three times assumption. Deformation isn't necessarily linear once you hit the plastic region so you can't assume the tension is linear. That's the problem with measuring length.
[...]
There was no plastic deformation of the bolt. It returned to its original length when I backed off the nut, and the nut spins easily on the threads. That means the stress in the bolt was elastic.
As stress in the bolt increases from zero, the strain in the bolt is proportional to the stress and its elongation is elastic. That's the range where Hooke's Law applies. As the stress increases but before plastic deformation of the bolt sets in, the strain becomes less than proportional to the stress. That's the non-linear region. Finally, as the stress in the bolt becomes very high, plastic deformation of the bolt occurs and the bolt will not return to its original length when the nut is backed off, but the stress in the bolt was never that high, as no plastic deformation occurred. So, during this test the stress in the bolt was always in the elastic range and so was at a minimum proportional to the elongation, and so it was at least three times with the oil on the threads what it was with clean, dry threads.
You can nit pick the details until the sky falls. The test simply proves what my concern is, that using a lubricant on the threads can cause a very much higher stress in the bolt with a given torque on the nut than would be the case with clean, dry threads, and the difference is quite significant. Try it yourself -- it's easy, it's cheap, it takes less time than you've spent writing about it, and it works.
And, I would not use WD-40 as a lubricant. It's a "water displacement" product, which is what "WD" stands for. It's mostly kerosene, and it attracts and holds dirt and other contaminants. It's for use to displace water and act as a solvent, not for lubrication. That doesn't mean that it doesn't lubricate, and lots of people use as such, but there are much better lubricants available.
Sorry, you're right, DJ. I misread your "oiled" numbers. You're right. Hooke's law definitely should have applied. Without a stress-strain graph I can't say, but it probably did.
I guess I anticipated plastic deformation due to the excessive torques you used. I'd have to look them up to be positive, but your torque figures are pretty much out of the range of what would be considered "normal" torque specs - even for a grade 8 bolt - which you didn't have. Darn near double for the actual bolt you used.
Your testing would seem to confirm that it is acceptable to use 3-in-1 for the 3/8ths bolt that you had since the torque on the should be something like 30 ft-lbs and definitely not 55. So you you have quite a bit of safety factor built in. At least from the bolt's perspective. Can't say from the clamped material's point. Optimum loading is 70-75% of the yield strength. So your bolt was much stronger than a class 5 spec if you didn't deform. But that's also the problem with torque specs for tightening. They aren't really reliable. Just most reasonable for giving *some* method to tighten for the average joe. There have got to be at least 5 to 7 other ways to spec how to tighten a bolt and how you get there does matter. More ways now since they have better tools to analyze tightening results and better/superior materials with computer analysis available to everyone.
There is NO question that about 85% of the torque value is merely overcoming friction. That's both the friction under the head and on the threads. Even swapping a washer (type/material) can make a huge difference. The reason that I use WD-40 is not to make it easier to tighten a bolt, but to prevent galling and locking. Displace moisture, and I'll displace potential rust, which also means less likelihood of breaking during loosening.
It's mainly due to the work I've done on my own decades old vehicle's that I do what I do. Breaking fastener's when loosening really sucks. (So I don't do it.)
I'll admit I'm a bit obsessive about performance and in spite of the cost, I even swap stainless for steel bolts frequently and it's usually wiser to lubricate stainless in particular than not. But to me, it's a lot easier to pay a little more now than pay a lot more down the road in headache alone.
But you have kind of piqued my interest to validate what I've already been doing for a long time with no issue. As long as I'm out of plastic deformation, I'm happy. My knock-off torques (I do measure those) are more than acceptable to me and I usually measure that on at a couple fasteners when loosening something just to double check expected behaviour.
If I don't damage a cone seat aluminum wheel with lubricating studs with anti-seize, it's highly doubtful that I'm going to over stress the studs. But that's only been my personal experience so far and I'm aware of what I'm doing. As soon as I see different, then I might change my mind. Thus far I haven't seen any reason to change.
But the point of NOT assuming that a lubricated thread is the same as a dry thread is very, very valid. And your experiment definitely shows that, though it was at least a bit out of range for the size bolt you used. Quite a bit I think, but I'm not going to swear to that. I would have thought it would have broken. Maybe it would have if you didn't deform the washer.
DJ, I do have a question though based on your past experience. What have you seen for the roundness quality of Nokian tires in general? It won't actually be convenient for me, so I'm doing if for the performance, but I'm thinking of getting a set of Nokian tires for summer even though it's going to be the most expensive option for me by far - even over Michelin's which really aren't that much considering how long they last. I'll have to get them cross shipped from Denver (due to winter tire popularity there) to here in California since they're really hard to order here. And I'll have to buy 5 since replacing an unrepairable tire is going to take at least a week (plus).
I know you love Michelin and they are a great tire. I'm looking for something that has the traction of my Revos for summer use (never even rains), but with slightly less noise and unfortunately LTX m&s aren't it unless they've changed something recently. Been there, done those several times already. Though they are a wonderful tire. I've kind of gotten spoiled with my Revo traction. But if Nokian aren't the answer I'll probably with Michelin again for summer use even though I'm not really happy with the dry traction - even with additional siping. I try to save my Revo tread depth for driving in slushy Sierra cement so I want something for non-winter weather.
Alan
Quote:
Originally Posted by DJ
There was no plastic deformation of the bolt. It returned to its original length when I backed off the nut, and the nut spins easily on the threads. That means the stress in the bolt was elastic.
As stress in the bolt increases from zero, the strain in the bolt is proportional to the stress and its elongation is elastic. That's the range where Hooke's Law applies. As the stress increases but before plastic deformation of the bolt sets in, the strain becomes less than proportional to the stress. That's the non-linear region. Finally, as the stress in the bolt becomes very high, plastic deformation of the bolt occurs and the bolt will not return to its original length when the nut is backed off, but the stress in the bolt was never that high, as no plastic deformation occurred. So, during this test the stress in the bolt was always in the elastic range and so was at a minimum proportional to the elongation, and so it was at least three times with the oil on the threads what it was with clean, dry threads.
You can nit pick the details until the sky falls. The test simply proves what my concern is, that using a lubricant on the threads can cause a very much higher stress in the bolt with a given torque on the nut than would be the case with clean, dry threads, and the difference is quite significant. Try it yourself -- it's easy, it's cheap, it takes less time than you've spent writing about it, and it works.
And, I would not use WD-40 as a lubricant. It's a "water displacement" product, which is what "WD" stands for. It's mostly kerosene, and it attracts and holds dirt and other contaminants. It's for use to displace water and act as a solvent, not for lubrication. That doesn't mean that it doesn't lubricate, and lots of people use as such, but there are much better lubricants available.
DJ, I do have a question though based on your past experience. What have you seen for the roundness quality of Nokian tires in general? It won't actually be convenient for me, so I'm doing if for the performance, but I'm thinking of getting a set of Nokian tires for summer even though it's going to be the most expensive option for me by far - even over Michelin's which really aren't that much considering how long they last. I'll have to get them cross shipped from Denver (due to winter tire popularity there) to here in California since they're really hard to order here. And I'll have to buy 5 since replacing an unrepairable tire is going to take at least a week (plus).
I know you love Michelin and they are a great tire. I'm looking for something that has the traction of my Revos for summer use (never even rains), but with slightly less noise and unfortunately LTX m&s aren't it unless they've changed something recently. Been there, done those several times already. Though they are a wonderful tire. I've kind of gotten spoiled with my Revo traction. But if Nokian aren't the answer I'll probably with Michelin again for summer use even though I'm not really happy with the dry traction - even with additional siping. I try to save my Revo tread depth for driving in slushy Sierra cement so I want something for non-winter weather.
Alan
I use Michelin because, of all the thousands of tires I've seen on balancers, and rotated during compensation of alignment sensors, Michelin was the ONLY tire that was consistently round. I don't mean just "reasonably round", I mean round, including its lettering and "decoration" as if it were turned in a lathe.
That means, unfortunately, that ALL other brands, Nokian included, were sometimes pretty nasty in terms of roundness. However, I don't have any specific memories of Nokian tires.
The advent of the GSP9700 strengthened this opinion. Michelin tires aren't just "round", they are "round while rolling" under load.
There are five other reasons why I use Michelin tires:
1) Michelin, to the best of my knowledge, has NEVER had a recall. Can you say "Firestone"?
2) My brother's Silverado has over 40K miles on LTX M/S in the same size as mine, P265/70R16, and they look NEW. Mine have 51K miles on them and are showing a lot of wear, but most of those miles were on mountain roads and so the were nearly always on curving roads or turning corners. The roads are usually pretty straight here in Oklahoma, so I'm getting better wear now.
3) A good friend bought my '89 Toyota truck when I bought my '00 Tundra. He put over 100K miles on the Michelin X radials that I had on it. That's TWO sets of tires, including the originals, in almost 200K miles of driving. (I don't remember the exact figures. He lives in Maryland now, so we don't converse often anymore.)
4) A group of engineers from Michelin came to Hunter with pedigreed tires for use in testing the new GSP9700 road force balancer. It's measurements matched theirs and so they bought the first six machines for their engineering labs. They care about quality.
5) I've never had a defective Michelin tire, nor have I had one that I was disappointed with. The don't slip or slide, they do fine miles offroad in heavy snow during deer season, and they are fabulous on the highway in all types of weather. I have no doubt that you can find a tire that is better on dry pavement, one that is better in snow, one that is better in rain, and so on. But can you find ONE tire that is better than the LTX M/S in ALL, or even in MOST of these categories? It would take me a lifetime to try.
I have great confidence in Michelin and I'm very satisfied with their performance, so I stick with what works and that I know I can depend on. It's just transportation.
So, for those of us who persevered to the end, if torque values were computed by the manufacturer with dry threads, its best to mimic those conditions during reassembly.
__________________
Ivan Stewart Body Kit; IS Wheels; Clear Corners w/ silverstar 3157A bulbs; Snug Top SuperSport cap; Marathon Super Hides seat covers in Shadow Grass; Husky floor liners; Toyota self-dimming rear-view mirror w/ temp & compass; Yakima Track Racks; debadged; Colgan front mask (only when going off-road); Cornfed 2 inch lift; Fabtech add-a-leaf; Total Chaos 1 inch differential drop kit; silver/blue TRD Offroad decal from Street Graphics.net; Line-X spray-in liner; 2005 tail lights; tailgate seal; Challenger Step Bars; Hellwig sway bar w/ Frontrange disconnects; SolidWiring Technology FOG603 foglamp mod & DRL601 switchable DRLs; battery manager; Prodigy brake controller.
DJ/Alan...Did it just get 'nerdy' in here, or is it just me? (j/k guys...)
If you want to make sure you don't snap a lug nut, the solution is simple: tell the tire shop guys to torque them properly (with a Torque Stik) I've had tire shops spin the lugnuts on with an airgun and they NEVER come off easily, most are seized before the lift even hits the floor. If you torque them to the proper spec, they not only come off more easily--they'll be less likely to seize. Not sure about anti-seize, because the lube would change the torque specification a bit and cause you to overtighten...sounds like a good idea, but I don't use it and I ALWAYS watch the tire shop guys like a hawk, I've never busted a wheel stud or had a lugnut seize up. That also could have something to with the fact that the wheels seem to come off of my truck for repairs/upgrades about every other week.
Upgrades: PA 3" Body Lift, Powertrax No-Slip rear, Detroit ezLocker front, Fabtech/Fox front, Wheeler's AALs, TC Front Diff Drop, 16 x 8 MB Wheels, 285/75R16 Goodyear Wrangler MT/Rs, Pioneer Head Unit, Infiniti Reference speakers [front and rear], 150 W bridged amp to bandpass enclosure, 6 CD changer with IR remote, iPod mini and xm hardwired into stereo, Toyota drop in bedliner, and Tonneau cover
DJ/Alan...Did it just get 'nerdy' in here, or is it just me? (j/k guys...)
[...]
Mike, you know me better than that. I'm a retired engineer. If you can't express it in numbers and show that those numbers are correct, then it ain't science and it ain't engineering, it's opinion. I prefer science and engineering.
Shouldn't I raise a red flag if people suggest doing something that could result in many times the stress in their lug bolts than torquing it to specs should provide? How much safety margin is left for the stresses of driving, hitting potholes, and bumping curbs? How many people get injured because they don't know what they don't know?
DJ/Alan...Did it just get 'nerdy' in here, or is it just me? (j/k guys...)
If you want to make sure you don't snap a lug nut, the solution is simple: tell the tire shop guys to torque them properly (with a Torque Stik) I've had tire shops spin the lugnuts on with an airgun and they NEVER come off easily, most are seized before the lift even hits the floor. If you torque them to the proper spec, they not only come off more easily--they'll be less likely to seize. Not sure about anti-seize, because the lube would change the torque specification a bit and cause you to overtighten...sounds like a good idea, but I don't use it and I ALWAYS watch the tire shop guys like a hawk, I've never busted a wheel stud or had a lugnut seize up. That also could have something to with the fact that the wheels seem to come off of my truck for repairs/upgrades about every other week.
You are right, Mike. You'd never want to torque by airwrench. And the shops I go to (granted they are good shops) never do. They measure with a half inch torque wrench.
I also pull my wheels off a lot (or used to - been too busy recently) and that's probably one of the other reason I tend to lube my studs and many fasteners in general. It's a lot easier to lose the wrench and spin the lugs off with the entension and socket alone. I've got an air wrench, but I can do it just as fast by hand by the time you pull out the hose and the wrench.
But what DJ was trying to point out is that even with torqueing by hand, you could overtighten if you lubed something and ate up the design safety factor. Personally, I've never run into an issue and I've been doing it for quite a long time. But I also use reasonable amounts (at least reasonable to me). But I also think about what I do and I even measure knock off torques from time to time. It's in the same vein of changing oil at 10,000 plus miles like I do. I've tested and it works for me.
DJ did get my curiousity up though to see what the actual pressure differences might be with and without lubricant for reasonable applications/torques. I'll have to try some testing and see what they are within acceptable spec for a couple fastener sizes. But I'll have to find someone around here that can test my digital torque wrench's accuracy limits before I even start. Then I'll have to test it again when I finish so I know it's still the same. I can't just assume it's accurate even though for what I use it for it's good enough.
The problem with tests are that they're like surveys - unless you know the boundaries of the assumptions, the summaries could end up being misleading. That's one of the reason I do a lot of things myself. I know what was done so I don't have to assume anything.
The end goal of a spec for fastening is make sure that the load on the fastener only is 70% to 75% of the yield load (generally). Increasingly, pure torque specs are becoming a bit old fashioned because they really aren't very precise, but for many things it's still the best someone has as a measure since it is somewhat linear since friction increases linearly with pressure. But too, unless you regular test the measuring device (torque wrench), you have no idea that the measurement is accurate. But that's where safety factors come in - accepting stacked errors so that you still don't get failure.
(So I don't do it.)
Linear Mode
