<VV> Fuel Leak

[Hugo Miller]

A couple of brief responses; 

1) You say; “Precession may not be the correct term for this but it is good enough for now.”
Precession is indeed the correct term – here’s a link to good old Wikipedia to explain how it works.  https://en.wikipedia.org/wiki/Precession_(mechanical)

Going back to wheels & lug nuts, the effects of precession will tend to loosen the wheel nuts on the left side of the vehicle. The effect is obviously minimal, since practically all modern cars have right-hand threads all round, and they manage ok. BUT ... car wheels are located by the lug nuts. Up till the 1980’s, commercial vehicles had the same set-up; conical lug nuts, relatively coarse threads, and left-hand threads on the left side. And wheels NEVER used to fall off. But since then, we have gone over to ‘spigot-mounted’ wheels, which are located on a central register, using fine threads (= easy to strip) and right-hand threads all round. I own one vehicle, a 1989 Leyland, which has spigot-mounted wheels with left-hand threads on the left side. That is the only time I have seen that hybrid arrangement. Clearly precession is indeed a problem with spigot-mounted wheels that rotate anti-clockwise and have right-hand thread nuts. Or else they wouldn’t keep falling off. 
As I said earlier, I am not aware of this being a problem with US commercial trucks. I have seen trucks pull into a truck stop to have tires fitted, watched the fitter do the lug nuts up DRY with a 3/4” air gun, drop the jack & drive straight out. I asked the guy what torque he was setting them to, and he muttered something about 120 psi & just shrugged. That would almost be enough to get you arrested in the UK, but it doesn’t seem to be a problem in the US. The mounting system is identical, threads are similar if not identical, and European wheel studs are further out from the centre than American ones, which theoretically puts the US ones at a disadvantage. So that remains a bit of a mystery.   

2) You say; “We also agree that these type of fastening systems are dependent on friction... “....”The friction derived from tightening the bolt creates the friction which prevents the threads from backing off.”
I’m not sure that we do agree on this. A stud or bolt or set-screw is effectively a spring, a very strong spring, that stretches when the nut is tightened and holds everything together literally by spring force. The more friction that exists between the threads, the more torque is going to be absorbed by this friction, and the less the ‘spring’ is going to be stretched. Obviously the less the ‘spring’ is stretched, the less the clamping force between whatever it is you’re trying to hold together. It is not, or at least it should not be, the stiffness of the thread that stops the nut unscrewing; it is the friction between that flat surface of the nut and whatever it is pressed up against (ideally a spring washer of course). 

3) You say; “Forgive me but the only reason your system doesn't result in disaster is because the threads are capable of withstanding the increased stress to which you subject them.  Other fastening applications would not survive.  By the way, if that crazy guy with an impact wrench works on your greased lugs, you may have an accident waiting to happen.”
What makes you think I subject my wheel studs to excessive force? I do them up by feel. In any case, although I would never use an impact wrench to fully tighten a wheel nut (or any other nut for that matter), it wouldn’t hurt the threads – as I said earlier, the vast bulk of the friction occurs between the nut and the wheel, not between the threads. And it is that friction which stops them unscrewing. The problem is, if you do them up dry, you may not have enough clamping force, so that as soon as there is any wear between the mating surfaces (due to heat cycles, dirt/rust/paint wearing off) you have lost whatever clamping force you had, and consequently there is not enough friction between the nut and the wheel to stop it all unravelling. If the stud is properly stretched, it can withstand a lot more wear between the mating faces before it becomes slack. 

4) you say; “If applied to attaching heads to our Corvair motors, your method would likely result in multiple studs ripping out of the crank cases.  The additional stretch of the nut/stud interface would transmit to the stud/crank case with disastrous results.  Never lubricate these threads.”
Most British engine manufacturers advise oiling the threads before fitting the head nuts. I fully accept your point about the other end being in aluminium, so I wouldn’t argue this point. I do have a related experience however; I have a Bristol 2 litre, with a cast iron block and aluminium head. Because of the bizarre valve gear (it’s based on a pre-war BMW) it is impossible to get a torque wrench anywhere near the head nuts. The threads are ‘cycle thread’, which may not exist in the US – it is a VERY fine thread. Somebody did suggest an appropriate notional torque would be 25 lb/ft, which gives you an idea how fine it is. 
Anyway, I had the head off mine once, duly refitted it, and went to re-torque the head after a hundred miles or so. I was surprised to find one of the nuts barely finger tight. It turned out that the thread had stripped, solely from the expansion of the aluminium head when it got hot. And of course, it couldn’t be the nut that stripped – it was the stud, so the whole damn lot had to come off again!  


From: roboman91324 at aol.com 
Sent: Thursday, September 27, 2018 7:47 PM
To: Hugo at aruncoaches.co.uk ; virtualvairs at corvair.org 
Subject: Re: <VV> Fuel Leak

Hugo,


Thanks for the interesting story and your theory.  I believe we may agree on more than you think.

First, a bit of info.  Chrysler products used to have left hand threads on the driver's side wheel lugs.  That's the American driver's side, not the British.  This ended, I believe, with the 1970 models.  There may have been a reason for it "way back when" but Mopar vehicles joined the rest of the car manufacturers without adverse effects in 1971.  Precession may not be the correct term for this but it is good enough for now.

Where we agree...  Like you, I believe there must be a technical reason for "wheel loss syndrome" and there must be a technical solution to the problem.  I would say that a simple phone call between Engineers of the British coach manufacturers and an American one might resolve the issue.  Professional courtesy would likely prevail and information would be shared.  It is probably an issue of thread pitch and/or thread profile and/or materials used.  In any case, with billions of perfectly happy lugs out there, somebody has probably solved this problem and that knowledge exists.  Of course, I may be wrong.  Despite the appearance of simplicity, the interface of so many dynamic and static parts in a wheel system makes the issue very complicated.

We also agree that these type of fastening systems are dependent on friction which is, among other things, a function of the contact force between surfaces.  As you torque a bolt or nut, the force between the thread surfaces increases.  The simplified formula is F = (mu) x N where "F" is the sliding force required, "mu" is the coefficient of friction and "N" is the normal force between the surfaces.  Further to this topic, we agree that the stretch/compression of the threads' contact surfaces is what increases the "N" factor in the equation.  As "N" increases, so does "F" if "mu" remains steady.

As an aside, I worked on a project whose purpose was to more precisely fasten components together,  Our customer was designing equipment to bolt heads to motors.  The way it was done previously was strictly with torque.  The new system would turn bolts/nuts to a preset torque to save time and then further tighten the fastener to the required stretch point.  The automotive industry has adopted this method universally.

That all has to do with the "N" part of the equation but you evolve into the "mu" part where we may not agree.  You state, "The problem with that is that a large part of the tightening torque is absorbed in overcoming the friction of the thread, and the stud doesn't stretch sufficiently to clamp everything together." The friction derived from tightening the bolt creates the friction which prevents the threads from backing off.  It is a good thing, not a "problem."  As stated above, "the stretch/compression of the threads' contact surfaces is what increases the "N" factor in the equation.  As "N" increases, so does "F" if "mu" remains steady."  I believe what you are doing is reducing "mu" and then compensating by increasing the stretch of the system.  The increased stretch increases the "N" which compensates for your reduced "mu."  "F" is the critical factor.

Bringing this together ... The lug/lug nuts on your coaches are over designed for strength.  They designed them this way to guard against stripped threads when some guy goes crazy with an impact wrench.  Forgive me but the only reason your system doesn't result in disaster is because the threads are capable of withstanding the increased stress to which you subject them.  Other fastening applications would not survive.  By the way, if that crazy guy with an impact wrench works on your greased lugs, you may have an accident waiting to happen.


As other examples:

1.  If applied to attaching heads to our Corvair motors, your method would likely result in multiple studs ripping out of the crank cases.  The additional stretch of the nut/stud interface would transmit to the stud/crank case with disastrous results.  Never lubricate these threads.

2)  The application that started this thread has to do with fuel leaks.  Randy suggested that grease be used on the threads when tightening this interface.  He cautioned that you should only tighten finger tight to avoid stripping the threads.  Threads in softer metals like copper, aluminum/aluminium, malleable steel, etc. are easier to strip than the lugs on your coaches.

As an aside, there are many types of "thread-locker" products available.  Their uses and attributes vary wildly and care must be taken to select the right one for the application.  In any case, they are not lubricants.  In addition, the use of one of these products that is too viscous might give you poor fastening results.

Doc


In a message dated 9/27/2018 2:20:31 AM Pacific Standard Time, Hugo at aruncoaches.co.uk writes: 

  An interesting hypothesis, but I'm going to disagree with most of what you 
  say - apart from the fact that if it is proving so difficult to get a seal 
  when there's only a couple of pounds pressure, there is clearly something 
  fundamentally wrong. It just should not be that critical.
  A bit of background - I am a coach operator based in the UK. we have a thing 
  called 'wheel loss syndrome', whereby the left-hand rear wheels tend to fall 
  off buses. At least that's what the politicians call it - I call it simply 
  bad engineering practice.
  Some years ago we switched to the European system of having right-hand 
  threads on all wheel fastenings. British vehicles used to have left-hand 
  thread lug nuts on the left side of the vehicle (as did quality cars also). 
  This is to counter the effects of 'precession', which will tend to make them 
  unscrew otherwise. A bicycle pedal will have left-hand threads on the left 
  pedal for the same reason, which seems counter-intuitive at first, till you 
  understand the principle of precession.
  Anyway, back to bus wheels falling off - the coach and bus industry as a 
  whole is paranoid about this 'syndrome'. But I am not. The main reason that 
  wheels keep falling off is because people adopt the procedure you have 
  outlined - they fit everything dry. Let me try to explain - a bolt or 
  set-screw is like a spring. When you tighten the nut, it will stretch, and 
  clamp everything together like a strong spring. On old British Triumph 
  motorcycles, they didn't give a torque setting for the big-end bolts (conrod 
  bolts) - you just used to tighten them till they stretched 1/8" or something 
  like that.
  On the back wheel of a bus, with twin wheels, you have a total of six mating 
  faces. Any dirt, rust or paint on these faces when they are assembled will 
  inevitably wear off in use, leaving the whole assembly loose. If the studs 
  are not stretched sufficiently, the whole lot will fall to bits. So the 
  first thing I do is clean all mating faces. Then I grease them lightly. I 
  never put two metal surfaces together dry. Then I use copper grease 
  (anti-seize grease) on the studs and lug nuts. That is to minimise the 
  friction on the threads, and ensure that the stud is properly stretched and 
  applying sufficient clamping force. Many if not most operators will fit the 
  lug nuts dry. The problem with that is that a large part of the tightening 
  torque is absorbed in overcoming the friction of the thread, and the stud 
  doesn't stretch sufficiently to clamp everything together. The nuts are 
  tightened till they are stiff rather than tight. These are the people who 
  have to check their wheel nuts with a torque wrench every day and fit little 
  plastic indicators so they can see when they are coming loose. These are the 
  people who end up in traffic court when their wheels have fallen off, and 
  everybody scratches their heads because "all the proper procedures have been 
  followed". I never check mine from one year to the next, I never use a 
  torque wrench, and they never come loose. The only remaining mystery is why 
  American trucks & buses don't seem to suffer from this problem.
  I realise this diatribe is of limited value when discussing fuel lines. 
  Power-steering lines will hold a couple of thousand PSI without leaking, so 
  as I said earlier, it should not be necessary to adopt all sorts of exotic 
  practices to stop gasoline leaking out at a couple of PSI. I reckon you are 
  missing something somewhere.