Hi All-

I had trouble finding a "best practices" post for a routine maintenance head re-torque all in one place, so I have taken a shot at creating one below. Comments and corrections appreciated.


How to Re-Torque a Head as Part of Routine Maintenance

1. For maintenance, do a “cold re-torque”, meaning the engine is completely cold. While it is called a “cold re-torque” it is actually involves a warm re-torque as the last step.
2. Remove radiator cap, then squeeze the top radiator hose down with your hand, then reattach the radiator cap while still squeezing the hose. This will provide some vacuum, which can help prevent water exiting the stud passage when you loosen the head nuts.
3. Back-off each head nut, one at a time, in sequence according to the spiral pattern below (see first figure below). Remove each nut and washer. Each nut will be both removed, then re-fitted before moving on the next nut in the sequence. If you remove more than one nut at a time, you will have a lot of coolant seepage.
4. Inspect where the head where the washer makes contact and ensure it is smooth and clean. If not, clean it with a brass/copper brush.
5. Ensure the washer is clean and smooth on both sides. If not, clean it with a brass/copper brush.
6. If nuts don’t remove smoothly, clean threads of stud and nut.
7. Lube (with motor oil) both sides of each washer before re-fitting it on the head.
8. Lube threads of each stud (not inside the nut, but the stud) before re-fitting the nut on the head.
9. Screw the nut down by hand until it reaches the washer. If nut doesn’t re-fit smoothly by hand-tightening, remove again and re-clean threads of stud and nut. If that doesn’t work, then chase nuts with thread tap to clean up the threads.
10. Torque each nut in 4 stages, performing each torque setting twice:
a. ~30% of the toque spec on the first pass
b. ~60% on second pass
c. ~80% on third pass
d. 100% on fourth pass
11. Repeat above process through each nut according to the pattern sequence diagram below (first diagram). Remember – one nut at a time!
12. Once all nuts have been removed and re-fitted, let the head sit for a couple of hours to let the gasket "relax" then re-torque again at 100%
13. Once all nuts have been re-torqued cold, then drive car until up to full operating temperature for a “warm re-torque” to finish the job. Torque each nut in either the spiral pattern illustrated in the diagram below (first diagram) or, optionally, in a traditional “criss-cross” pattern noted below (see second figure below). Some folks like to vary between the spiral torque pattern above for the cold part of the re-torque and a “criss-cross” pattern for the warm part of the re-torque.
14. OPTIONAL: If there is a coolant or oil leak present at the head that you are trying to remedy via a re-torque, advice is to perform another cold/warm re-torque sequence as per above (i.e. one nut at a time in a spiral pattern), but instead of removing the nuts, just back them off ¼ turn then re-torque to 2 ft/lb above the final torque specs listed below (e.g. if normal spec is 58, then torque to 60.)

Torque specs – final torque pressure by engine displacement. This applies to 105/115/116 Series “street” motors, not race-prepped motors.
• 1300, 1600: 52 ft/lb
• 1750: 58 ft/lb
• 2000: 65 ft/lb

you overlooked the part about having a headless chicken in the trunk and and making sure you are facing north and standing on one foot while holding the wrench in your left hand only...

why all the mythology ? and whats this about the gasket " relaxing " ? if your gasket is " relaxing " its already scrap.

all al heads are done cold ... always cold... never hot or warm . and the system on this car is the same as every other in the world. 3 stages ... spiral or crisscross depending on who trained you. if its a maint retorque then you crack the nuts loose one at a time noting what it took to release them and then pull it back down at a few lbs more than that number or the spec called for originally which ever is greater because once you release the clamp load on the gasket you need to exceed that when you retighten it to make certain it re compresses and seals...

there is no need at all for any magic process.

 

This is an urban myth.....
Rebuilt my '64 Spider in 1980 and never retourqued the head. Works fine.

 

not urban myth at all. if done properly ( torque properly, heat and stress cycle once , full cool and retorque ) the first time it shouldn't need doing again. but having said that, clamp load is a strict function of hardware preload. all materials " cold flow" under stress to some degree. in metallurgical terms its called intermolecular creep. that means you put enough stress on something long enough and it changes dimension. so... over time , if left to their own devices the initial preload will diminish slightly even if the materials are inside their elastic limit. combine that with the fact that the gasket is organic to some degree and compress-able and as the studs are subjected to high cyclic load and heat and all manner of thermal cycles and suddenly the notion of " gee it SHOULD be fine forever " becomes " gee for the sake of 30 minutes , i can be certain its good for another 10 years or so..."

my 79 had 30 k on it when i got it and had been sitting for 25 years . it only needed 35 ft-lbs to release the head nuts and almost a full additional turn to pull them up to 70 . sometimes being sure is better than what the theory says. in my case im sure it saved me a repair.

 

"in metallurgical terms its called intermolecular creep"
Perhaps my '64 hasn't heard of this.......

Do you have a graph (in feet) of how much creep per year will occur ?

Also, my GTV rebuilt in 1991 and my Super rebuilt in 2010 seem to be functioning fine with a little intermolecular creep.

 

its a function of the molecular bonds of the various materials as related to stress and temp. very brittle materials like glass are at one extreme of the scale and softer more compliant materials are at the other end. titanium creeps less than , say aluminum ... that sort of thing. in the case of cylinder heads it is far less of an issue than , say , the compress-ability of the gasket and how it will cold flow or even de grade over time depending on what it is made of and what it was subjected to. i don't think anyone is suggesting that retorquing should be on the same schedule as spark plugs... but every 10 or 15 years isn't such a bad idea. feel free to ignore anything you don't understand . if you are happy with and trust your assemblies then you have nothing to concern yourself with in regards to any of this.

 

My experience is that most head gasket issues are caused by poor workmanship and or materials and all of this retourqing is an attempt to delay the consequences.

 

well... if thats your experience then stick with what you know and understand.

certainly if a job is done badly for whatever reason then some patchwork later isn't going to be the answer but thats not what anyone is talking about and my car was factory assembled and had lost 1/2 of its preload over time.

being through and meticulous is not the same as trying to make up for what you did wrong in the first place and there is no downside at all to retorquing in exactly the same way as there is nothing wrong with doing a nut and bolt check over any other job you did. anal ? maybe . but no downside at all. some might say it shows genuine care and concern for a job well done.

 

well... if thats your experience then stick with what you know and understand.

I totally agree.
Not only should the heads be retorqued, but the Factory torque specifications are too low.
I know of NO professional engine builders using less than 10% higher torque specs on Alfa engines.
I use 75 lbft on the 12mm head studs/nuts on the 1750/2000/Montreal.

Low head torque is directly the cause of the Montreal engine`s headgasket issues. In a TSB (01.71.2.1) the torque specs were given as 8.1-8.2 KGM hot.

I used 75 lbft on mine (75 lbft = 10.4 Kgm)

FYI: 1 lbft + 0.138254954 kilograms m

 

I totally agree.
Not only should the heads be retorqued, but the Factory torque specifications are too low.
I know of NO professional engine builders using less than 10% higher torque specs on Alfa engines.
I use 75 lbft on the 12mm head studs/nuts on the 1750/2000/Montreal.

Low head torque is directly the cause of the Montreal engine`s headgasket issues. In a TSB (01.71.2.1) the torque specs were given as 8.1-8.2 KGM hot.

I used 75 lbft on mine (75 lbft = 10.4 Kgm)

FYI: 1 lbft + 0.138254954 kilograms m

Dont forget that factory torquing and retorquing manuals always talk about well lubricated threads and washers. Reaching x Nm or y lb-ft torque on a clean vs on an oiled screw will result about 30% different pre-tensioning of the same bolt. Lubricated screw will “use” less torque to win over friction and more to pull the screw in axial direction).
I think if an engine was in production from 1955 until the 90s, than maybe the factory engineers knew, what and why they did. Hm?

 

Excellent writeup, Scott!

I would however add to #7 that the flat side of the 8 thick washers need to be facing down, toward the head. Rounded side goes up. Doesn't matter for the 2 thin washers on the center lifting strap though.
#10. IMO, there's no need to do incremental pulls for a cold retorque. One pull to the final value is all that's needed (assuming sufficient swinging room).

 

Well, there's no shortage of threads on this topic. Everything I wrote above is just a collection from a variety of threads on this topic, mostly coming from THIS THREAD and its many offspring.

This question of this thread isn't "do you think re-torquing is a good idea". The question of the thread is "is this a good procedure if you are doing a re-torque".

Why the mythology? Because Alfas have a very long history of head gasket issues and conflicting/changing advice from Alfa itself (not to mention various non-Alfa manuals) regarding how often to re-torque, under what conditions and in what nut order. Given that, I am just trying to collect best practices provided across many threads over the years in one place in the event it might be helpful to someone and someone might have some updates or corrections to it.

That said - I personally don't think the "let the head gasket relax" matters, but guess it couldn't hurt.

 

Ah - thanks Jim. It's mostly advice you have written before across many threads just put all in one place. Thanks for the add'l info and with any other that comes in, I'll update the final flow.

 

If I can add one more thing I learned from Papajam was the sequence. I like it because last 2 nuts to be torqued are at the rear where most of the leaks are.

Thanks for the thread.

Vin

 

At least on the V6 I've fixed oil leaks twice by retorquing. YMMV.

I'm with Jim on the four-stage torquing. You can just torque each nut once smoothly up to the desired value and call it good. Probably a good idea to make sure you're using an accurate torque wrench too.

 

Ah - good point. It's always worth reminding about the accuracy of torque wrench.

 

My 30-year experienced Alfa mechanic (theres no substitute for having worked on so many cars over the years. in LA where there is so many) says that 'loosening' is nuts. he does a cold re-torque for me every two years, and ONLY TIGHTENS to spec.

I should mention that my motronic alfa has 200,000 miles and runs perfectly

 

on any fastener in any application, automotive or otherwise correct and standard practice is to release the preload and then re-preload. this is to insure that the threaded section is free to rotate without rust or corrosion or any other friction generating circumstance that will cause the torque being applied to be inaccurate. now... if your guy thinks anyone is talking about loosening ALL the head nuts and then retorquing , you have mis informed him and that would be the wrong thing to do. if he is simply cranking on fastners that are already tight and have been that way for any length of time then he is wrong and it doesn't matter how many years he's been doing it... its still wrong. this isn't magic and its not voodoo and it has nothing to do with alfa or ferrari or yugo or mercedes benz. it is standard engineering practice as to how threaded fasteners in ANY application are treated. the ONLY variable is how the factory calls out its torque spec whether done dry or with oil or with moly or any other specific lube whether in lb -st or angular deflection or total stretch... the raw methodology, however, is always the same and anythind done any other way is wrong to some degree. everybody wants to think that everything has to be special or magic or so secret that only the few can know the tricks... and thats nonsense. these are not formula 1 cars and they are not special or different in any meaningful way as to the techniques necessary to work on them properly.

 

Well,
thank goodness I haven't retourqued all these years. I would have been doing it wrong !

 

there is no doubt in my mind that that statement is true.

 

Have to agree with stevesxm. Having newly adopted "re-torque head" as part of my maintenance routine, I recently had nuts that were corroded on. I could've re-torqued to 100 ft/lbs without any outcome because the corrosion was the resistance, not the actual nut/washer contact with the head. This would be true with any corroded or otherwise binding fastener. You have to level set with a free moving fastener (by loosening it) to assure that your torque measurement is actually the pressure applied from the fastener to the item being torqued.

 

well thats the standard , isn't it ? whatever engineer specified what ever the preload was going to be didn't just say " 65 lb-ft or .025 stretch " ... he said "65 lb-ft or .o25 stretch using 30 wt motor oil assuming class 2 threads " or something like that. he specified quite clearly how that preload was achieved. so... if you are re torquing that piece of hardware, in order for you to get to what the true spec is ( assuming that is meaningful to you ) then you have to start from " scratch" and the conditions that were called for regardless of what they were. if it was dry then you have to do it dry and if it was oiled then you have to do it oiled and so on. if you don't do this then the preload you are creating is some OTHER number than what was specified regardless of what the torque wrench says. on critical hardware like rod bolts and head bolts and the like doing it any other way is asking for a failure. on oil pan bolts or valve cover bolts , maybe you use a speed handle and say " the hell with it " ... it all comes down to whether you care about doing something correctly or not. its what separates real mechanics from people who ' just work on cars "

 

Vin - thanks for adding that diagram. I was looking for a source to pull from when writing this, but couldn't find one and didn't want to guess at it. I'll add this in the final write-up. Given that any leaking head I have had also leaks from the rear, I would agree that this one seems to make the most sense. I like the idea of varying the pattern (this one, then spiral, etc.) on an aluminum heads just to ensure even pressure. I can't back that assertion with any metallurgic data, but seems to make logical sense.

 

Re-torque

there is no doubt in my mind that that statement is true.

I`m with you on that!

 

Note that the TSB is for post 1969 models, so I am good not needing it on my older models.

 

What?

Note that the TSB is for post 1969 models, so I am good not needing it on my older models.

Really! That`s the thinking that has made you such a knowledgable poster!

 

Yes,JR.....
and you know why they came out with the TSB?
And how many other car makers call out for repeated retourque of their head bolts?
Refer to my earlier post for a hint....

 

It is standard practice for every aluminium headed engine to have a cold retorque after the first heat cycle, or at least I am not aware of any engine that does not have a retorque procedure (even cast iron headed engines).

But yes normally the retorque is only done once.

With my Alfa, I will be retorquing once a year until I know it has settled.
Pete
ps: I am also aware of a customer that did not come back for the retorque on the Alfa Berlina that blew another head gasket ... next time the garage did not let the owner have the car back until the retorque was done, no further issues.

 

but the Factory torque specifications are too low.
I know of NO professional engine builders using less than 10% higher torque specs on Alfa engines.

Is this on lubricated threads or dry threads?

 

Lubricated with engine oil.

As an engine builder, I've always used Alfa torque specs, without issue, for some 40 years.

 

Hi all,

I have a quick question about the procedure here. I perform cold retorques as outlined here, and follow the alfa workshop manual maintenance procedure except I torque to about 90Nm. I have never however performed a hot retorque after this as the workshop manual doesn't, at least obviously, require it. Is it important to perform the hot retorque and if to what torque value? (2L engine). I ask as I have developed an oil leak below the exhaust manifold at cylinder 4 and hope a torque can fix it.

Thanks,

Stephen

 

again...standard practice is to retorque any system after it has heat cycled and been stressed a couple or few times. gaskets are not rigid. they compress and cold flow and release preload a bit from brand new so on cast iron heads you re-t hot and on al heads you do it cold. what I would do on every motor was torque during build , start and heat cycle and retorque ( either hot or cold depending on the system ) and then drive it for a day or two and do it again... when the client came back at 500 miles for a break in oil change and inspection I might do it then again as well just as a belts and suspenders thing. in your case you can try to fix the oil leak w/ a retorque but think about it... if the head isn't tight enough to controll oil pressure at 60 psi, what do you think is happening between the cylinders or into the water galleys ?

can't hurt to try but if it works then go buy lottery tickets immediately.

 

Was educating myself a bit with this thread after wondering if I should re-torque my head. just bought car. 130k miles. all original motor. Just have an accurate wrench, loosen, oil, and re-torque to spec according to pattern shown? Sounds pretty simple. Is this something that may open up a can of worms? lol...

 

I wouldn't.

Nothing to gain but a lot to lose.
Pete

 

I was looking for the torque values needed for my 89 Quad when I found this thread. Wow - There is a great deal to be said about "we have found from experience". I spent a large portion of my life as an engineer on 1200PSI (975 degree super heat) steam ships. I will keep this really simple - when you heat metal up it does strange things. When you put metal under stress and heat it up - it gets stranger. I had a situation where a shipyard supervisor swore his way was correct and it cost the government almost 6 Million dollars in new repairs. I documented his work - much to my great relief. They have never followed that supervisors instructions since. My bottom line is this - no matter what we all do and accept - we will find the correct way. Some of us are going to have tragic mistakes and by being afraid to talk about it we allow others to make the same mistake. I have learned so much form this BB!


My engine was not re-torqued after a visit to the mechanic four months ago. Could that be why I see a leak at the after end of the engine at the seal? I am willing to bet....

 

"Remove radiator cap, then squeeze the top radiator hose down with your hand, then reattach the radiator cap while still squeezing the hose. This will provide some vacuum, which can help prevent water exiting the stud passage when you loosen the head nuts."

This is complete nonsense ………… you cannot create sufficient vacuum to overcome the presence of antifreeze by squeezing the radiator hose. I and others tried this decades ago. It does not work! DRAIN the radiator below the level of the head!!!!!! One drop of antifreeze between the aluminum and the gasket and it is going to fail.

 

Back in the 80s, Fred DiMatteo told me about this pretty slick trick. Many years later, I came across the Alfa TSB linked below.
Knowing a bit about Fred, I'd wager I know who Alfa got the idea from.