Alfa Romeo Forums banner

21 - 40 of 47 Posts

·
Premium Member
Joined
·
15,431 Posts
"No one on this thread seems to have water leaking into the oil. All external leaks including mine"

True. However, it is a common ailment in high mileage Alfa engines, both the four banger and the V6. Retorquing the heads can forestall gasket replacement, sometimes for many thousands of miles, as it has in my 91S. Nothing to lose when the choco mousse appears.

Having said that, our 75 Alfetta sedan, as well as other four banger Alfas we've owned, always seemed to leak oil down the outsides of the block after a while, even with a fully flat head and proper torquing, evidently the gaskets just not that good to account for an aluminum block and head, and steel liners, design.
 

·
Registered
Joined
·
1,076 Posts
always seemed to leak oil down the outsides of the block after a while, even with a fully flat head and proper torquing, evidently the gaskets just not that good to account for an aluminum block and head, and steel liners, design.
That will likely not happen when modern head gaskets are used and the gaskets gets a proper done re-torque procedure. All engines which I've dismantled after a HGF (oil leakage) suffered either from low torque figures or not properly done re-torque procedure. With Reinz gaskets you even can re-torque before first firing up.

As well as adding extra sealant to specific points on the block surface. But the most important thing to check is correct liner protrusion when changing a head gasket. That's the key to get a Nord/TS to run without any leakage for years.
 

·
Registered
Joined
·
2,191 Posts
I had the same leak on mine and retorqueing the head as outlined above did the trick. Twinsparks only have two oilway with the little o-ring seals, one just above and behind the distributor and one on the opposite side above the dipstick.
Very encouraging. Seems to be my case. I'll try soon.

Thanks,
 

·
Registered
Joined
·
411 Posts
I come from the world of Jaguar XK engines.

In the late '60's Jaguar cars issued a service bulletin that said Barr's Stop Leak was to be added to the coolant every time an engine's cooling system was flushed & refilled. The XK head was notorious for warping.

Small coolant leaks, AKA weeping? Then add Barr's.

Does anyone on this forum use Barr's?
 

·
But Mad North-Northwest
Joined
·
9,859 Posts
Alfa TSB 07.93.01 suggests using a coolant additive "if seepage is not the result of improper hose clamp or fastener torque." If memory serves the stuff they recommended was an alumaseal-type product.

Personally I'm not a fan of putting sealants in the cooling system, but YMMV. I've never had persistent coolant leaks on any of the Alfas anyway. The head gaskets I've had issues with typically leaked oil to the outside.
 

·
Richard Jemison
Joined
·
7,077 Posts
Heads & leaks

Most engine builders (including the factories) use Barrs leak or similar in every new engine built.

Critical in this re-torquing leaky head gaskets is the use of Barrs leak as a preventitive.

Torque values from the "Factory" are too low by at least 15%. The smaller 10MM head studs need 50 to 55 lbft of torque hot.

There`s a bunch of mis-information posted here on the BB re warped heads from people that are less than experts on the issue.:frown2:
 

·
Premium Member
Joined
·
13,081 Posts
I have checked a few cheap torque wrenches against an electronic torque indicator and they all indicated a few ft-lb higher than the measured torque so I always set mine a few ft-lb higher than what I want.
 

·
Registered
Joined
·
2,191 Posts
Would you expect a high quality torque wrench to deviate much?

Edit: I just read about using the digital device in a bench vice and snapping the manual wrench into it and set the value and read it off the device when it clicks. Duh. I'm gonna do that.

I in fact retorqued my TS this weekend. There is still an oil drip but it is less than before. I followed the instructions using grease and 60lbs then 70lbs. Used my manual one as the battery was dead on the digital one.
 

·
But Mad North-Northwest
Joined
·
9,859 Posts
Would you expect a high quality torque wrench to deviate much?

Edit: I just read about using the digital device in a bench vice and snapping the manual wrench into it and set the value and read it off the device when it clicks. Duh. I'm gonna do that.
That's how I did it. All my wrenches were pretty much within 5%, which is a typical clicker spec. Generally they got a bit worse at the high and low ends of the range and were like 2% in the middle.

(Interestingly my cheap-o Harbor Freight ones did better than my Craftsman ones...)

Good to check the wrench calibration once in a while but don't overthink it too much. Couple % isn't going to make a difference and I'm sure it's within the noise of operator technique...you just want to make sure they're not really out of whack.
 

·
Registered
Joined
·
2,191 Posts
Some updates for my attempt. As mentioned, I went through the sequence at 60 and then 70. I now have a new battery for the digital unit - a Harbor Freight Pittsburgh brand. I had not checked my torque wrench against the digital one prior to doing it, and it still leaks a little.

I put it in the table vice and had my son film the screen on his phone and we looked at it after in slow motion. I set the digital one to beep at 70 lb/ft. The torque wrench was set to 70, in the same position as my final torque on my head. When it clicked, the digital did not beep and the screen in slo mo only showed 62 lb/ft. I set the wrench to 75 and the digital meter read about 66 max and still did not beep. I set the wrench to 80 and the digital meter went to about 72, but still did not beep. I set it to 85 and the digital meter beeped, and the digital reading went to about 74 or a little more.

The odd thing is that you can hear the two clicks of the torque wrench while watching the video in slo mo and the highest value is after the first click and before the second click. The 72 lb reading recorded this after the first click. I don't know what that means with regard to the beep/no beep.

I think I should probably set it to some where between 75 and 80 and re-torque my head again. That may or my not beep, but the torque value will be closer to 70. If I do that, should I just go through the torque sequence at this setting one more time, or loosen them all again and perform the whole thing in steps again? I've driven it maybe 7 days total, 10 miles each day.

Thanks,
 

·
Registered
Joined
·
4,136 Posts
Alumaseal will work to stop some kinds of seepage. Back in the 80's Alfa recommended it to dealers.
 

·
Registered
Joined
·
114 Posts
Discussion Starter #35
No, the problem was on the exhaust side.
Still no more leak after about 300 mls now...
Knock on wood... :hammer:
 

·
Registered
Joined
·
2,191 Posts
should I just go through the torque sequence at this setting one more time, or loosen them all again and perform the whole thing in steps again? I've driven it maybe 7 days total, 10 miles each day.
Anybody have any input to this in the next few hours? I have a window of opportunity to work on my car this morning. Maybe not tomorrow.

Edit: I read somewhere on this site (but can't find it now) that someone says in cases like these - re-torque of a re-torque - to back off 1/4 turn and re-torque. I plan to do this and not drain the coolant. Hope that's OK.

Thanks,
Stefano
 

·
Registered
Joined
·
4,136 Posts
That is for the coolant. Unless you are suggesting to put that in the oil?
Alumaseal is added to the coolant. It is intend to stop minor coolant seepage around gaskets. Alfa felt that it worked well enough to recommend it to dealers who were having 2 liter head gasket problems. Scheech.
 

·
Registered
Joined
·
2,191 Posts
Oh yeah, the original poster had a coolant leak. I hijacked the thread for my oil leak. Sorry guys.

So I just re-re-torqued the head. I tested the torque with the new setting and it clicked right away. So I backed off the nut 3/4 of 1/4 turn. What's that 3/16th? I did not drain the coolant so I just wanted to break the friction. It actually wasn't hard to back off the nuts. It torqued more than I backed off. The re-torque went at least 1/2 turn more or 3/4 turn. So I now have more torque. I checked with the digital very slowly and got about 71-73 on each nut.

So if this still leaks, I have to decide if I want to live with the leak and drive and fill or do a head gasket. Winter is coming, so it's not a bad time to do it. It was only external oil. Internal was still OK. Unless I broke the seal more....
 

·
But Mad North-Northwest
Joined
·
9,859 Posts
I put it in the table vice and had my son film the screen on his phone and we looked at it after in slow motion. I set the digital one to beep at 70 lb/ft. The torque wrench was set to 70, in the same position as my final torque on my head. When it clicked, the digital did not beep and the screen in slo mo only showed 62 lb/ft. I set the wrench to 75 and the digital meter read about 66 max and still did not beep. I set the wrench to 80 and the digital meter went to about 72, but still did not beep. I set it to 85 and the digital meter beeped, and the digital reading went to about 74 or a little more.

The odd thing is that you can hear the two clicks of the torque wrench while watching the video in slo mo and the highest value is after the first click and before the second click. The 72 lb reading recorded this after the first click. I don't know what that means with regard to the beep/no beep.
Technique is a reasonable part of getting consistent performance out of a torque wrench. You don't want to go too slow or too fast: you just want to be swinging it at a moderate rate and stop immediately when it clicks.

If you want to check calibration don't use the beep. Your digital gauge should have a "Peak Torque" or maybe "PtoP" setting. Basically a setting where it stops at the highest torque reported: that's what you want to use. Set your wrench at the lowest setting, take three readings or so to get a good average, and write the numbers from the digital gauge down. Then step the wrench by 5-10 ft-lb and do that all the way through the range. Again, don't go super slow or super fast...nice, consistent swings of the wrench with a consistent hold on the handle, and make sure you can stop before the wrench hits the other end of the "click."

When you're done, plot out the data in Excel (setting vs. measured) and you should get a nice, linear plot close to X=Y assuming your wrench is decent. I haven't looked at the numbers in a while, but almost all of my clicker wrenches were like within 3-5% in the middle of the range, increasing to 5-8% at the ends. That's your calibration curve: if it's close enough you can just use the wrench as-is, otherwise you can adjust based on the curve.

(e.g. if your wrench is set to 70 ft-lb but the gauge says 67, you can set the wrench to 73 ft-lb when you need 70. But at some point you're splitting the proverbial hairs: we ain't building the space shuttle here.)
 
21 - 40 of 47 Posts
Top