OK, first time using the torque angle method on these cars.... I torqued to 15 ftlbs, went to the store bought the gauge came home.. turned 90* on each. At about 75* on the first pass the bolts got tight (had to actually try to to torque them) and now trying the 2nd pass any more then 30* feels like I will damage something..

I put a beam style torque wrench on the 2nd nut after the first was difficult after 30* and it showed 60 ftlbs, while a few showed 45-55 ft lbs.


Every other head gasket i've done was on n/a and I torqued to 15, 37, then 66 ft lbs and it was significantly easier. Do 951s have a torque value that much higher?

I know I did not exceed 15 ft lbs on the first pass, and no more then 90* on the 2nd pass.. the 3rd pass calls for another 90 and I don't see that happening..
thanks
mike

 

Were they new studs? Were they lubed well? What HG are you using? New nuts? With my new ARP studs it's by torque value, not angle. If they are stock studs I have to wonder how many times they have been used before and they might be stretched a bit. I used the torque angle on my BMW and thought I would twist them off on the final pull; it really was hard even with a 24" bar and only ended up at 75 ft-lb when I checked them.

 

Different engines obviously, but the tacoma engine I did is by torque angle.. and it had me go to a torque, then 90 degrees, then another 90. With a breaker bar on a half inch socket I could only get about 80* on the last sequence pulling with all my mihgt... Nothing broke.

 

Cool, thanks guys

EDIT: So I just did the last 90* pass.. nuts 1 and 2 were at approx 75 ft/lbs.. the rest were 95+ ! All of the nuts were tightened in order 15 ft lbs then 90* 90* ?? Could having some new nuts/ washers (5 new, 5 used) cause that? Only reason I'm not sure is because only 2 nuts were torqued less but 5 have new nuts/washers..

I don't attribute this to inexperience as I've done head gaskets and everything else you can do with an engine many many times on 944 and other cars... built engines for old Jags and formula atlantics etc.. but I've always had a ft lb spec.


Could the HG be crushed from the extra torque? it was done in correct order
thanks

edit x2: would it be worth doing a compression check before finishing the assembly and putting it in the car?

 

..anyone know the max torque you can have on the head gasket before failure? from research online is says too much torque on the head studs can cause bore distortion..

 

I don't use the 90 degree method anymore after doing it twice more than 10 years ago and never had failures going with 90 ftlb final torque for the last 10 years and many 944/951 head gaskets later. I am not sure why they use this method and think its pretty stupid personally. I torque mine down with a snap on torq-o-meter with a dial gauge instead of the clicking type since they are pretty inaccurate. Use anti seize on the studs and I never torque them over 90 ftlb on the final turn. When you have clean well oiled threads, this is the most accurate way to ensure proper torque. By measuring the inconsistent torque specs of your head, you pretty much proved how inconsistent the 90 degree method is.

 

I dissagree, Lego, and I know this subject has been covered before. The torque angle method is more accurate to get a consistant clamping force on the head... Essentially, what happens, is the steel studs need to be pulled in tension so they put a compressive force on the head. By knowing the tensile strength of the steel and the diameter of the studs, we can calculate how much force would be required to "stretch" a stud, let's say, 0.5mm. (I put stretch in quotes, because this stretch is the "elastic deformation" kind, where the studs will spring back to their original shape -- or course if you exceed the stretch rating, you'll get into the "plastic (permanent) deformation" part of the curve.)

The most precise way to measure this bolt stretch is to put a dial indicator on the top of the stud while you tighten the nut with an opened wrench -- this way you can physically measure the studs ellongation while you're turning the wrench. However, you need a pretty snazzy setup to do this properly. So, the alternative is to CALCULATE the stud stretch. We know the thread pitch of the nut (and stud), so we can calculate how much nut rotation is needed to ellongate the stud. Porsche has done all of this math for us -- this is where a little trust on our part come in.

Using a torque wrench, be it beam, click or dial, isn't very accurate, because, a torque wrench really measures the FRICTION the nut is feeling. This can change do to many variables (thread interface, washer interface, lubrication, etc.), and as Mike's data suggests, the torque values can be quite different, although the stud ellongation is pretty much the same.

As you say, being as careful as possible with a torque wrench (new nuts, new washers, everything lubed, like the factory manuals originally instructed) will REDUCE the friction variables, but it is still not as precise as the torque angle method... which is why the factory manuals were updated with the torque angle method.

And Mike, congrants on getting your head back on!

 

I would tend to disagree as the first measurement is based on a torque reading so the whole premise of friction making it inaccurate makes no sense to me. If you have equal friction (clean thread and lubed), the head would be clamped consistently. Yes I have taken statics, dynamics, and material science so I am very aware of HRB hardness testing, dislocation energies in lattice, and different types of deformation and stretch for my mechanical engineering school so I am no stranger to Young's modulus. If friction varies so much, then the torque wrench would be unnecessary as you will trust your rod, crank cradle and all other internal bolts based on the 90 degree method. I am sure we can both draw a force diagram and calculate how much force is applied to the stud vertically knowing the angle of the thread and how much torque is applied to the nut. I would tend to say the friction for a clean stud with anti seize pretty UNIFORM for all the studs, which is the desired result. This is why most cars I work on are torqued with a torque wrench and not a torque wrench to the the first 15 ftlb of torque (which includes the "inconsistent" friction you objected to) and then angular rotation of a wrench. I am sure you are familiar with error propagation. If the studs are already stretched, then the elasticy will vary much more as the diameter of the studs are not equal creating different clamping force.

I believe there are more cars made where the manufacturer uses and recommends the torque method than the angle method.

 

You may find this article interesting: http://e30m3performance.com/installs...rods/rods5.htm

And, here's ARP's stretch gauge, which they recommend be used whenever possible: http://www.arp-bolts.com/Catalog/Cat...talog_0088.pdf

You are also correct that stud shaft diameters must be consistant between head studs -- this is why people pay a fair amount of money for precise products. Another jem from the ARP catalog: http://www.arp-bolts.com/Catalog/Cat...talog_0036.pdf -- their head studs are centerless ground - a process that can yield repeatable results to ten-thousandths of an inch.

The 15 ft. lb. preliminary torque is just to "snug everything up" so your relative starting points for each head stud are consistant.

I'm not saying it's WRONG to torque head studs with a torque wrench of some kind -- I'm saying that the torque angle is a more accurate, more precise method. And I'm also saying that measuring bolt elongation is the most precise method.

Spend some time in a race engine shop (where engines are built to withstand twice the HP over a stock engine), and you'll see how much precision and measuring plays a role.

 

This may sound stupid but the ARP bolts are designed to stretch to a fixed length correct? I can see that measuring the finished stud length with a dial indicator would be the most accurate method by far. What I don't understand is why the torque angle method usually ends up with nice neat 90* pulls. It seems that if precise elongation was so critical then you would see some angles that fall some where in between? Is the initial torquing to a precise value adjusted so that the 2-90 deg. pulls work out even?

When I replaced a newly machined head on the BMW using new head bolts and chased block threads, I used the recommnded 22# initial, 90* torque angle, rest, 90*torque angle. The head leaked slightly at two locations until I used a torque wrench and added between 5 and 12 ft-lbs to each stud to even them all out at 80 ft-lbs.

I have to do the ARP's on my new block in the next couple weeks and the ARP recommended 65# final value has resulted in leaks for several people here. Many have ended up using 80#as the final pull with good results. Is there anyway to calculate, not measure, what the added 15 ft-lb does to the stud length or if it's detrimental?

Thanks

 

Steve, I have used 90 ftlb on my head nuts with no problems for my street and track cars. The measurement process and calculations will depend on many things including the steel used in the head stud and how its annealed. Its pretty hard to do that without a shop/lab with specialized equipment.

Van, my philosophy is going with what works. I don't live at my friend's race shop but have been around when 1500 to 3000 hp engines are built. Per your links to e30s they mentioned "friction in the threads of the bolt can greatly affect the torque that is required in order to achieve the correct pre-load. Thus it is critical when torquing a fastener that you use the same thread lubricant". This is exactly my point. They never mention anything about torque according to 90 degree turns. In fact, they recommend 50 ftlb torque for their rods, which hints at torque measurement rather than angular measurements.

 

Here at work, there are a lot of strech bolts used, The Cayennes subframe up front takes 2 good sized guys to get them to strech. Porsche takes it to such an extreme that If we dont replace a stretch bolt, meaning there isnt one billed on a Warantee bill, when one was supposed to be undone, Porsche will kick the claim. Sorry but like was stated before, if Porsche is this passionate about this spec, there has to be a reason. I understand you arent going to change your ways because of this post, but i wanted to get my point across as to how important Porsche says this is.

Mark

 

I never said anything about ignoring stretch. Let me ask you a question, do you torque those bolts down with a torque wrench or preload a bolt and turn 270 degrees?

 

All these posts are intresting and informative.. thanks

Do any of you guys have any idea of how much torque is too much ? as in the head gasket will be damaged and I will have low compression or bore distortion etc

I'm trying to decide I am replacing the head gasket now before the engine is in the car or waiting and seeing if it blows.

The studs were lightly oiled (put my finger in a quart of oil and wiped the threads for each stud) , seemed to work for the past few 944 head gaskets, do you guys think its too little?

thanks guys
mike

 

What kind of head gasket is it? Perhaps, for the piece of mind, you want to change it again.