20Nm + 90 degrees, Renew = Torque to yield bolt?

Little Dog

Little Dog

A2OC Donor
European-Union
Not necessarily is the answer and, while best practice would be to follow the manual and renew, money can be saved by reusing screws that have not been stretched beyond their elastic limit. I noticed this some time ago when buying bolts for my A2 and Golf. Both the same part number, golf reusable with a new lock nut, Audi renew. For the Golf toques are specified as a number of Nm, the Audi torques are more commonly specified as a small number of Nm plus an angle.
I looked into this and discovered torque plus angle reduces the risk of over torquing. The theory is that the average mechanic doesn't stop pulling when the wrench clicks especially if the required torque is large. However the same person is good at pulling 90 degrees. With torque plus angle the error due to over torquing occurs at a small torque value so is a small error and the final torque is pulled accurately being specified by an angle. I've also noticed that torque plus angle reduces tooling costs. The measured torque is applied by a small torque wrench and the angle pulled with a breaker. Large expensive 1/2 and 3/4 torque wrenches are no longer required. Torque plus angle is less likely to result in over torquing compared to a single larger torque. Providing a bolt has not passed its elastic limit it can be reused.
A single torque figure cannot be specified for and applied to a torque to yield bolt. The bolt necks then the application of single torque figure, above the torque at which the bolt reached its elastic limit, is likely to leave the mechanic pulling away at the bolt feeling it get no tighter, with no clicks on the wrench until it breaks. Torque plus angle is required to ensure the mechanic stops pulling at the right point in the plastic zone. So torque plus angle may indicate a torque to yield bolt. A single torque figure indicates the bolt is not torque to yield.
My advice elsewhere has been to visually inspect the bolt to see if it is torque to yield. Dark Side offer, using their jargon, stretch (torque to yield) and non stretch bolt sets for PD injector clamps, cam bearing caps and rocker shaft. I bought stretch for my engine before discovering the bores are shot. I decided to keep the expensive stretch, circa £90, for my engine rebuild. I ordered a set of non stretch about £40 cheaper and even cheaper still if you get them direct from fastener suppliers.
The difference is easy to see, 12.9 high tensile on the left, torque to yield on the right.

A2 Torque 1.jpg



The high tensile bolts are marked 12.9. Torque to yield have no tensile marking and have a reduced diameter section where the bolt will yield and neck.

Dark Side recommend torques of 15 Nm for the M6 injector clamp screw and 44Nm for the M8 rocker screw. A check of maximum torques for 12.9 screws shows 19.1 Nm for the M6 and 46Nm for the M8. Both are within their elastic limit so can be reused in the future. the M8 is a bit close to its limit for my liking so I'm going to experiment and determine if the non stretch bolts result in over torquing compared to stretch bolts.

I'll report my findings.
 
Last edited:
The first thing I have concluded is that if you know you have stretch bolts buy stretch bolts. I bought a set of stretch bolts for the top of my cylinder head then discovered that my bores were shot. I decided to save my expensive stretch bolts for my replacement engine and bought some much cheaper non stretch substitutes. I've only investigated the M8 rocker shaft bolts, I was a little concerned that the recommended 44 Nm into aluminium was a little high. I also thought I didn't pull such a high torque when removing the factory stretch bolts.
So I have to decide what torque to pull to replicate the clamping of the stretch bolts. The manual says 20Nm + 90 degrees for stretch bolts but my gut tells me that will be too tight for non stretch bolts. Sorry no pictures I had to sort this out.
The bolts that came out of the engine were narrower at the narrowest point and 0.2 mm longer that my unstretched new stretch bolts. I don't have access to manufacturing tolerances so I have to assume the whole 0.2 mm is due to the narrowest point necking.
I have reservations about using 12.9 non stretch bolts for two reasons, within the elastic zone they may be stronger and resist elongation more than stretch bolts. Secondly because they are unlikely to yield the bolt may pull out the threads in the head. I'll explain in a little while but first the stiffness of 12.9 vs stretch bolt.
I torqued both, stretch and non stretch, to 7Nm then measured the angle to to apply at total torque of 30 Nm:
Non Stretch 82 degrees
Stretch 92 degrees
Stretch bolts are soft compared to 12.9 so there is much potential for over torquing.
The stretched, used, bolts are 0.2mm longer than new bolts. I shimmed a 0.2 mm gap under a bolt the measured the angle to the matting surface, 50 degrees. So if I have got this right when I pull the 90 degrees the bolt yields at 40 degrees and then is pulled 50 degrees into the plastic zone.
Checking with theory the thread pitch is 1.25 mm and I need to stretch 0.2 mm 0.2 divided by 1.25 mm equals 0.16. Multiply 0.16 by 360 degrees and I get 57 degrees. Very close to the empirically measured 50 degrees. This just adds weight to my experiment.
So I have a stiff 12.9 bolt that will without doubt over torque if I apply 20 Nm + 90 degrees. Next I had to work out what the torque should be.
First what it shouldn't be, I know I get 23 Nm divided by 82 degrees or 0.28 Nm per degree. Multiply that by 90 degrees and add to 20Nm and I get 45 Nm. Very close to the recommended 44 Nm but over torqued. I'll pick this up again tomorrow.
 
Last edited:
I now have confirmation that, as I thought, that 45 Nm is more torque and compression or clamping than would be applied by the stretch bolt. I now need to work out what torque would be applied by the stretch bolt.
20 Nm is applied then it is pulled a further 90 degrees. I know what I have calculated but I will assume that the the bolt yields at 45 degrees. So the theoretical total torque is 23 Nm divided by 92 degrees multiplied by 45 degrees plus 20 Nm. 31 Nm, gut feel tells me that is a little low, also a small amount of torque is added in the plastic zone. I'm going to apply 35 Nm to the non stretch bolts.
I was hoping to pick up a replacement engine today and I would then have verified this torque when removing the factory stretch bolts. Unfortunately it is an older AMF and I can't use some of the parts purchased for my engine. Instead I will have to pull 35 Nm and see how it feels.
I've got the same concerns with the injector clamp screws. 15 Nm on a M6 into aluminium. That doesn't feel good, I have stripped M6 screws out of the head at 10 Nm. But the specified torque is 12 Nm plus 270 degrees. Some of that angle must be compression of the brass washer so 15 Nm may not be unreasonable. I'll check the depth of the thread in the head.

When the non stretch bolts come out they can be reused so will go in my spare fasteners box. If I'm right the rocker bolts could have be cheaper 10.9 grade.
 
I'm following your calculations with interest. My only comment would be: might it not be more benficial to calculate (or better still measure, empirically) the tension in (or pressure applied to mating faces by) the bolts, rather than focussing on the torque values? Once you've stopped turning the bolts, the torque will be zero. The tension in the bolt thread/shank will remain, maintained by the friction between the screw thread and whatever it is the tapped hole is made of.

Once the bolt shank starts behaving elastically, the torque values are going to move around and become more difficult to predict. Not impossible, but difficult. Setting up a simple rig with a strain gauge would allow direct comparison of the clamping pressure applied by the 12.9 and the stretch bolts incrementally in, say, 5 degree steps of the bolt rotation. You'd be able to see exactly what the differences are between the two once the stretch bolts go elastic. That setup would also alow comparison of different thread types. The thread ramp angle on metric fine is different to that for metric 'coarse', so the clamping pressure they generate will differ for the same torque applied.

Like you, my gut feel is that M8 44Nm figure in aluminium is getting a bit uncomfortable.
 
@TFG thank you for that comment your points are very very valid. I’m restricted by the tools I have and my calculations are based on an assumption that elongation in the elastic zone is a function of angle turned. Also clamping is a function of elongation which is a function of the torque applied. I think I got that right and I appreciate I’m assuming other variables, plate finish, lubrication etc are constants. Things are very different in the plastic zone.
However I think it has all worked. Based on the assumption that the bolt yields at 45 degrees and stretches up to 90 degrees I decided to put 20Nm on a used bolt and the see if I could pull it past 45 degrees with a torque wrench set at my calculated 35 Nm. It passed 45 degrees without a click and with ease and was heading for a full 90 and a break. I concluded 35 Nm on non stretch would clamp the same or more than 20 Nm plus 90 degrees on stretch.
Non stretch bolts installed;

D34D282B-04E6-45EF-8540-0FA1FC5F1FC6.jpeg
 
Some good experiementing here. I was under the impression main reason for Torque + Angle was actually due to it being a more reliable and robust way of attaining a specified clamp force. Because when you torque using Nm, friction plays a huge role, hence why ARP fasteners are supplied with a specialised lubricant. You also need a correctly calibrated torque wrench. However, Torque + Angle means the inital low torque value is laregely uninportant, it just needs to be close to make sure the faster is snug and there is no slack. Then the angle strenches the bolt by a known amount, and thus exerts a known force on the joint. This moves the requirement for tight tolerances away from the person using the bolt, and back to the manufacturer, where the process is already very consistest and well specified. Obviously, pulling 90 degrees is also fairly easy, even without a tool and its harder to overshoot as you mentioned.
 
@NateS2 good point well made. I’ll write a summary to wrap this up and include your point re friction.
Not sure why I didn’t find it when researching stretch bolts. Friction will make a difference.
 
You must log in or register to reply here.
Back
Top