Thermo cycle engine during build process?

[reclino]

Ok saw this video on YouTube a while back and it stuck in my head. John Edwards was pumping 150 deg coolant through an engine after replacing a head gasket to settle the head gasket, then re-torque.

https://youtu.be/m4IuhcrP7Bg

This would be easy enough for me to do with my 1986 911 turbo engine at this point in the build.



A little electric heater and a big cardboard box and I could take the entire engine to 150 deg or so a few times and then check torque on the head studs. These engines really don't get tight until they come up to temperature so this may help settle the cylinder base gaskets and the head gaskets.




Being air cooled it will come up to temp well I would imagine. Air-cooled small aircraft use forced air engine preheat during the winter so it's not totally out of left field. Though with the amount of thermal mass it may take a bit of time. I could bond a thermocouple on the case in a few locations to monitor progress.

David

[icarp]

As a matter of practice , I always do a second head re torque 24-48 hrs after the first one .
They seem to like a second go round. No failures , good practice .

Ian

You can't hurt the process by adding heat.

[Walt Fricke]

Certainly can't hurt, but adds more work for what I suspect is close to zero actual benefit. Leaving aside the very earliest 2.0s, there are no head gaskets as such The CE crush rings in their grooves get fully crushed well before you have fully torqued the head studs. As to the base gaskets, a retorque after some hours on the engine is a good practice (especially for one's first rebuild), though not necessarily essential.

One hears of "cold relaxation" on the stretch of CV bolts, though the precaution of a subsequent retorque isn't mentioned for, say, rod bolts (perhaps the most critical torque for our engines?) even though those are usually torqued well before the crank gets lowered into the case. A lot of these engines have been rebuilt by people of quite varying skill and experience levels without having base gasket leaks.

But certainly a retorque while the engine is still on the stand wouldn't add much time - you might not have the valve covers on other than loosely held with a couple of nuts to keep out dirt.

Doing a retorque on a warmer engine sounds like it would only catch a nut which was not properly torqued in the first place. The cylinders will expand more than the studs (how much depends on the stud metal), so the stretch (tension) of the studs will be higher than when first torqued cold. So you won't be adding any stretch to what a cold retorque would provide?

Retorque after running the engine isn't that big a deal if combined with checking the valves, or changint the oil, though. Goes fast. On the mag case engines, is kind of routine to see when a stud has pulled.

[reclino]

By thermo cycling, I mean heating the entire engine for maybe 3 hours to 150 deg F, letting it cool off, heating it again, then letting it cool off. This will tension and relax the studs and am sure there will be some relaxing of torque.
The picture shows me installing the head gasket, so it sure does have them now. I put in Mahle 3.4 liter cylinders and pistons and modified the 964 turbo head gasket to fit. Rods now have ARP bolts put on with a stretch gauge.
I am not trying to save time building this engine.
I have documented this build pretty well in the turbo forum, but don't often post on the engine side,. mostly just read and try to learn from the more experienced people here.
"Barn find" 1986 911 turbo

[PeteKz]

Reinforce what Walt said. For several mechanical engineering reasons, I doubt there is any benefit to heating the engine and retorquing the nuts/bolts. 150 degrees is insignificant compared to the maximum operating temperatures. Retorquing the nuts while on the assembly stand is a good idea to catch any mistakes you might have made in the initial torquing. But there shouldn't be any change due to just sitting, because the stresses are well within the elastic limits of both the steel studs and aluminum castings. If you need to add torque to any of them (beyond a slight movement), it's because you either didn't get the torque right the first time, or something is wrong (a Dilavar stud getting ready to snap, maybe?).

However, checking the torque after you have run the engine a few thousand miles is a good idea in case something has loosened up.

[reclino]

I am a mechanical engineer and can think of reasons this is a great idea..... Yeah 150 is pretty low, originally thought about using 230 deg F as I don't want the oil much above that..... On a running engine the heads will be hotter, the closer to the crank you get the cooler it will be so was trying to be conservative.
David
Performance EngiNerding

[icarp]

Pete Kz , a re torque 24 -48 hrs yields a min of one stud per engine that needs more.
Normal for 5-6 per engine , just my experience

Ian

[PeteKz]

Ian, hard to argue with experience. I wonder why that would happen. Maybe creep/compression of the copper base gasket? What are you using to lubricate the threads?

In order to do an engineering analysis, I'll need to remember to dig out a steel cylinder stud and take some measurements: working length of the stud (the center section between the threaded ends); the diameter of the stud, in order to calculate the cross section; and the diameter and pitch of the threads to calculate the conversion of nut turns to stud stretch and stress. I looked up the values for coefficient of thermal expansion for steel: .000007"/inch x F* and aluminum: .000013"/inch x F* (these are nominal as there are differences depending on the alloy). Anyway, if I get motivated tomorrow, I'll find a stud and measure it. Then I can calculate the change in torque value for a temperature change from 70F (room temp) to 150F. But I can say this without doing the calculations: Heating up the engine increases the stress on the studs because the steel expands significantly less than the aluminum head and cylinder it clamps together. I don't think it's a significant delta, but I (or someone else) will need to run the numbers to be sure.

[reclino]

I coated each base gasket in CurilT sealant, recommend by Henry at SuperTec. Head studs are stainless also from SuperTec, diameter is larger than the factory studs, don't have that measuremet handy.
I stacked each head and cylinder pair and measured combined height to the shoulder of the cylinder. 2 sets were .0005 low on 2 adjacent sides, one set was .0005 low all the way around. These are now cyl 1-3 with 3 being the lowest.
The other 3 cylinders measure the same within.00025, they will be 4-6.Did this in an effort to avoid deflection and binding of cams in the cam tower during assembly.
Unknown is the alloy Mahle uses for the cylinders, nor do I have a reference of CTE of RR350.
The base gaskets seemed to be die cut copper, stamping process left a rolled burr around the edges, I had both coated Victor-reinz ones and uncoated Elring ones, I udes the Elring as the thickness when checked with a micrometer seemed more consistent (ignoring /avoiding the rolled edge)
My thought was that this burr combined with the coating of CurilT would slowly compress during torquing of the head. Heating and cooling the engine a few times
will soften the CurilT and increases the tension in the studs due to the difference in CTE between the aluminum and the stainless. Eventually everything will settle down and become stable, but I don't believe a simple torque applied once is enough to do a good job.
My addition of the 964 shim head gaskets and hand lapped tops of each cylinder add to the complexity of this system.
I only want to build one 911 engine. I am not a professional engine builder, just a very careful engineer trying to do a better than average job.
Hoping to enjoy driving this engine for many many miles. As many miles as I can afford.

[reclino]

Well it's in a cardboard box, with a heater blowing approximately 200 deg air into the box. Been in there for about 2 hours and exit air temp has stabilized at 154 deg F. Dialed the heat back a bit and just going to let it soak for 5 hours or so at this setting. Will turn off and let it cool overnight and re-torque tomorrow.
The garage smells of Loctite and CurilT
EngiNerding

[reclino]

Now up to 158 deg F exit air temp, dialed the heater back to lowest setting.
It baked for about 6.5 hours, checked temps at the end.
The case was at 135 Deg F, heads were at 145 Deg F. Checked temps with IR temp gun not a perfectly accurate tool but close enough. So it had not quite equalized. Let it cool about 20 hours in the garage which was 70 deg.

[reclino]

When you make things complicated for yourself, sometimes they get complicated.

Checked torque on passenger side.

Reviewing notes looks like I torqued them to 24 ft lbs which was the factory spec prior to the 90 deg rotation new factory procedure.
Henry shipped the head studs with the following
TORQUE SPECIFICATION + or – 2lbs STANDARD TORQUE PATTERN IS MAINTAINED
a. 2.0 aluminum case 30 lb/ft
b. 2.0, 2.2, 2.4, 2.7 ( all magnesium cases ) 26 lb/ft
c. 3.0, 3.2, 3.6 aluminum case 30 lb/ft
d. 3.0, 3.3, 3.6 aluminum Turbo 32-36 lb/ft
THERE IS NO RETORQUING NECESSARY

I thought 32-36 seemed extreme considering these have fine threads at the top
Upon re-torque today took them to 28ft lbs, 9 moved at 26 ft lbs and 3 moved at 27 ft-lbs.

Will check driver's side tomorrow.
Need to make a note to take better notes.
Not sure if this added any value, but I had the time available.
David
EngiNERDing

[safe]

Originally Posted by reclino

I thought 32-36 seemed extreme considering these have fine threads at the top
Upon re-torque today took them to 28ft lbs, 9 moved at 26 ft lbs and 3 moved at 27 ft-lbs.

Follow the specs I'd say...