View Single Post
galwaytt galwaytt is offline
Registered User
galwaytt's Avatar
Join Date: Feb 2005
Location: Galway, Ireland
Posts: 91
apologies if ye all know this already, but as it hasn't actually been stated yet, so if you indulge me a bit here.............

I sell Torque analysing kit and torque wrenches, both manual and air as well as electric, to a variety of industries. Some of my customers are subcontractors to automotive industry, so whilst not being a scientist, I have garnered a fair bit of info over the years. It never fails to amaze me the number of 'professionals' who have no comprehension of torque when they call me up to help with a problem they have, say, on the production line.

Torque readings on ANY wrench, are, in fact, useless!! No, really !! If you follow this, you'll see why:
The basic tenet of tightening a fastener, is to clamp two pieces together. The strength of the finished assembly (like a big-end cap on a conrod, for example), is contingent entirely on how tightly clamped together those two pieces are.

Now, at the design stage of the rod, someone determined that it needed a certain amount of clamping force to keep the thing working. But, to measure the clamping force under the head of the bolt required a special measuring device - a load cell washer, connected to a display unit - or to measure the stretch requires bolt-length measuring equipment. The dictates of production machining, space etc, laid down, for example, that they use an M8 bolt and nut. To achieve the required clamping force required to keep the two parts together, the bolt manufacturer has a table you can look up to give you a suitably strong bolt. What happens when you use it is that you are going to stretch the bolt. This is fine in the R & D dept, but out on the floor it's not going to happen - it'll interfere with the ability to make and sell cars. Not forgetting of course, that Porsche build a 911 engine in under 90 minutes. Complete.

So how are the guys (and girls !) out on the production floor going to know if it's clamped tight enough? Well, that's where torque wrenches come in. The bolt mfr and engine mfr conduct tests (using the load cell above) and determine that in order to get X amount of clamping force, the recommended bolt had to be tightened to X Nm of torque. That's all it is. It's a broad correlation between how much torque gives how much clamping force.

So where does angle come in? Well, like all things, science and R & D moves ever onwards. Originally, bolts were tightened to X Nm, you ran the thing to bed it in, and then you re-tightened them a second time. This is because the parts bedded in, and bolts which were somewhat stretched on original installation, would now have relaxed a little, and would therefore now have lower clamping forces. This is not practical for internal engine parts, so bolt mfrs and engine builders have used materials to build modern bolts from materials which are stretched at point of use, i.e. on the production floor. These are stretched so that they still deliver the maximum clamping force AFTER bedding and running in, in other words, after they have relaxed. Think of all the hot/cold heat cycles and the effects of them on new parts.........

The measurement of this stretch is by simple length, but which again, is difficult to measure - remember, Porsche need X engines a day out of ya ! - so we can't be hanging around with mic's and stuff - besides, they're, again, the R & D dept gets together with the bolt mfr and they come up witha certain number of degrees of rotation after a minimum baseline torque has been set which CORRELATES to the clamping force they're trying to deliver.

That's why you can't re-use the old bolts - their elasticity is practically spent............if you re-tighten them as per originally fitted, they will either break during your engine build..............or very shortly thereafter................and you don't want to know about that.

Other factors: lubrication. Generally, torque figures are based on assembly of new, unlubricated clean threads. Anti-seize pastes, threadlockers etc etc all affect the final result. And usually on the high side, and can be by a considerable amount. Locking devices such as crush and nyloc bushings also greatly add to measured torque without achieving clamping force, so make sure you're replacing non-nyloc nuts with the same non-nyloc nuts for critical parts.

If this is telling you something you already know, apologies, but if not.........maybe it'll save you talking the engine down a second time.............

And an angle adaptor to go on to your torque wrench is cheap - say Eur65-70. Less than 1hrs mechanic's fees.............considering you might spend 40hrs doing the engine.............

Finally, if using your torque wrench, you shouldn't 'pause' the wrench before hitting your preset torque. If you do, all your calc's are for the bin. You should start the wrench in one position, and make one full smooth motion all the way 'til it clicks, or whatever it does (beeps, vibrates...etc etc)

My 0.02
John M
Co Galway, Ireland
'87 911 Carrera Sport......not any more....00 TT quattro...oops, gone too....'94 968...for now..., add a '93 968 Tip to that. Hey, two cars are better than one, right ??
Old 02-17-2005, 05:08 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #20 (permalink)