Torque To Yield, Angle

[Don D]

Why doesn't Cometic recommend TTY for the use of their head gaskets?

[Admiral Akbar]

Because their heads gaskets don't collapse over heat cycles.

[Don D]

Considering the application of TTY why doesn't Harley recommend the replacement of bolts after one use as other manufacturers do? Or do they? I know they do on M8

[Coyote]

You're suppose to replace them.

[cheech]

FWIW and to differentiate, there are torque to yield fasteners (TTY). And the torque plus angle method of tightening fasteners.
That method may not necessarily just be used with TTY fasteners.

Harley of course has been using it since the Evo days on the heads.
And even though they suggest or require replacing the "head bolts".

Is there any published info in service manuals or otherwise that they are indeed TTY and shouldn't be reused because of that?
Given Harley's head fastener arrangement, the "head bolt" is basically a nut.
One would think the TTY component would be the cylinder stud itself.
And would require replacement any time they are un-tensioned.
Do M8 manuals require the studs to be replaced?

[Don D]

I hear you and my concern is the lack of consistent stretch for the torqued fasteners even when using a calibrated torque wrench, click stop + or - 3%? and torque to value VS torque to angle. I do know the oil or grease used under the heads of the bolts makes a large change. ARP has a product they recommend that will change things a lot. Seeing blown head gaskets from time to time on M8 and high compression big bore twin cams has got me thinking.
RA finish is right and parts are flat and clean. Cometic MLS head gaskets

[cheech]

It would be a good read as to why Cometic doesn't recommend the torque plus angle from someone in the know there for sure.
Everything you read on the torque plus angle, it's a more accurate way to achieve a specified clamp load.
Hence why it's the norm in the automotive world. And I'd guess Harley at the OEM level.
Lubes as you said for sure change the load when only using torque.
I wonder what Cometic's tightening specs for MLS are in high performance automotive apps.

[hrdtail78]

Why doesn't Cometic recommend TTY for the use of their head gaskets?

I always thought it was for generic application reasons.  Wouldn't hardened ARP cylinder studs stretch less than stock?  Also mentioned above.  Wouldn't some lubes on threads and under bolt head allow more tq than other lubes?

My question for you would be tq plate set up.  Do you TTY or shoot for specific tq?

[Don D]

I torque to gasket recommended values. M8 is the exception. I recommend 52 ft lbs for the M8 engine with cometic head gaskets, S&S bolts. I use motor oil on the threads and under the head. I inform my customers to duplicate that for their builds. I use a calibrated CDI digital torque wrench.

[kd]

I think you first have to understand the make-up of Cometic MLS gaskets.  They are 3 layer stainless with the outer layers having embossed ridges for sealing contact and coated with a thin layer of high temp Viton.  The center layer is of varying thickness, non coated and flat for even distribution of pressure.  The varying thickness is used to make up .027, .030, o036 etc up to .120 gaskets.  In short the gasket is much more stable and distributes the torque much more evenly.

Harley uses a 2 step torque program plus torque to yield 45* (not including the first finger tight set up stage) It goes 120 - 144 inch lb, then 15-17 ft lb + 90 degrees and then finally 45 degrees.

Cometic (with a more refined MLS gasket, uses 5 stages closer together for smaller more even increments.  With lightly lubed threads and bolt flanges, alternately snug the head bolts finger tight. Next, in proper sequence, tighten to 9#, then in sequence 14#, then 22#, 35# and finally 42#.    It seems to me the 5 stage procedure is much more able to draw down the head more evenly giving a higher quality even result.  It is also claimed to not cause cylinder distortion that results from installing heads and cylinders. 

Although I don't have the evidence handy, I do recall this conversation here a few years previous with respect to 45* yield compared to Cometic's 42 "# of torque.  The 45* turn resulted in the same as 42"# of torque.  I believe rbabos was one of the contributors.

Bottom line, you don't hear of a lot of Cometic head gasket failures (if any) that can be confirmed to be installed as Cometic instructs.

[Norton Commando]

I think the Cometic gasket is more forgiving in terms of accommodating variations in cylinder/head preload compared to the HD gasket.

The Harley torque followed by angle procedure will result in a more accurate preload compared to torque alone.  This is because small variations in friction during the torque process produce large changes in preload.  In comparison, rotating a nut a specified angle removes the uncertain friction variable.

The torque method has an error of +/- 25% whereas the angle or turn-of-nut method has an error of +/- 15%, per the 26th Edition Machinery's Handbook.

Jason   

[motorhogman]

I think the Cometic gasket is more forgiving in terms of accommodating variations in cylinder/head preload compared to the HD gasket.

The Harley torque followed by angle procedure will result in a more accurate preload compared to torque alone.  This is because small variations in friction during the torque process produce large changes in preload.  In comparison, rotating a nut a specified angle removes the uncertain friction variable.

The torque method has an error of +/- 25% whereas the angle or turn-of-nut method has an error of +/- 15%, per the 26th Edition Machinery's Handbook.

Jason

I have to ask.  The +/- 25% number. Is that just for the head gasket application or any TTY number ?

[kd]

I think the Cometic gasket is more forgiving in terms of accommodating variations in cylinder/head preload compared to the HD gasket.

The Harley torque followed by angle procedure will result in a more accurate preload compared to torque alone.  This is because small variations in friction during the torque process produce large changes in preload.  In comparison, rotating a nut a specified angle removes the uncertain friction variable.

The torque method has an error of +/- 25% whereas the angle or turn-of-nut method has an error of +/- 15%, per the 26th Edition Machinery's Handbook.


Jason

I do understand the engineering behind how TTY 1/4 turn will give a measurable stretch to the studs by calculating that amount of travel using the TPI of the studs.  What bothers me in the Harley head bolt sequence is only 2 first steps, the last of which includes a 90* (1/4 turn) then followed by the 3rd and final 45* (1/8 turn). It almost seems to me to be some sort of time saving method by reducing the number of clamping steps. I trust that Cometic has done their homework through rigorous testing of their 5 step sequence (as stated in their literature).  The rest is up to the installer to have a known good torque wrench and the skills and patience to use it properly.

added later:

Corrected info thanks to Buglet's eagle eye.   

[Buglet]

  I think I'm missing something isn't 90* a 1/4 turn and 45* 1/8 th turn or this this the new way.

[hrdtail78]

I torque to gasket recommended values. M8 is the exception. I recommend 52 ft lbs for the M8 engine with cometic head gaskets, S&S bolts. I use motor oil on the threads and under the head. I inform my customers to duplicate that for their builds. I use a calibrated CDI digital torque wrench.

Thanks, I think that is most important.  Try to duplicate how they were honed and bored.

[Buglet]

 S&S has there own sequence and there finial torque ends at 45 lbs

[Admiral Akbar]

I think the Cometic gasket is more forgiving in terms of accommodating variations in cylinder/head preload compared to the HD gasket.

The Harley torque followed by angle procedure will result in a more accurate preload compared to torque alone.  This is because small variations in friction during the torque process produce large changes in preload.  In comparison, rotating a nut a specified angle removes the uncertain friction variable.

The torque method has an error of +/- 25% whereas the angle or turn-of-nut method has an error of +/- 15%, per the 26th Edition Machinery's Handbook.

Jason

I wonder how much yield you get out of an 7/16 inch bolt?   It's 36% higher than a 3/8 bolt based on diameter / cross-sectional area.   

I suspect that the composite head heads have more give than Cometic.   One thing that seem to be ignored is that aluminum expands more than steel when heated.   The newer motors do run a little hotter.  Here is an interesting discussion comparing the 2 torque processes.  These guys seem to be doing a lot of motors.

https://www.youtube.com/watch?v=5KS9uk1vguU

According to them TTY causes problems. 

[kd]

   I watched that video some time ago. It's interesting how much deviation occurs when you have larger steps in the sequence.  It's part of what sold me on the Cometic instructions for a 5 step torque sequence.  Smaller steps cause a more even tensioning on the studs. Cometic also states their research and testing has proven it prevents distortion in the bore of the barrels. (I wouldn't doubt the same would hold true when mounting a cylinder in torque plates)  I expect the culprit is outed in this video where #3 is over spec after drawing down on #4 and rocking the head on the gasket.  The fact that the other 3 are then under spec would throw the next sequence off even worse if it wasn't noticed / corrected.  I have always gone back over them to confirm the torque before moving to the next sequence because it really doesn't take any extra time and confirming is a feel good thing. It sounded like they were using Cometic MLS gaskets in the video.  You will notice that the 15 minute check-up showed no change in any of the 4 bolts. The softer Harley gasket is probably another story and the reason Harley uses torque to yield.  I expect after getting the engine hot they are looser.  I know my shovel using similar gaskets always was.  They always got a little tap up after a bit of run time.

BTW, the fact that the TTY came out to the 45# supports the findings in the thread we had going here that I mentioned in post #9.

[Admiral Akbar]

The torque method has an error of +/- 25% whereas the angle or turn-of-nut method has an error of +/- 15%, per the 26th Edition Machinery's Handbook.

Jason

Not sure that this is true. Maybe it's miss-interpreted.  I have 26 can you post the page?

[Admiral Akbar]

I'd like to bring up a challenge to even if HD cylinder studs are even torqued to yield.  From what I see they are not.  Here is how I got to the conclusion.  I'd post pictures of the calculated results but the admins would likely delete due to the copy right restrictions at the bottom of the page on this link. You can go there and run the calcs yourself. 

https://www.futek.com/bolttorque/american

TC cylinder stud.

Bolt size 3/8-16
Grade 8  Yield 130ksi
k Factor:0.2

For receptacle, I used 4130 which has a over a 200ksi

Repeat for 7/16-14.

Results for 3/8-16:

Suggested Tq    / Resultant pressure. 

36.32 fp / 5812 lb  - Recommended
48.43 fp / 7749 lb  - Max

Results for 7/16-14:

Suggested Tq    / Resultant pressure. 

58.14 fp / 7973 lb  - Recommended

TC Analysis

Tq to 42 fp is dead center of the recommended  range so the average of the 2 pressures would be about  6780 lbs

The diameter if a TC cylinder stud is 0.336 (measured). The area is 0.08867 square inches.

The stress on the bolt material is  6780/0.08867 = 76463 psi

M8 Analysis

The TQ on a 7/16 bolt for the M8 is 45 fp.  Since the results list 58.14 fp  = 7973 lb, I'll use proportions  to come up with final pressure.

Since I don't have a M8 stud I figure the center diameter to be 0.395 or about 0.12255 square inches.

(45 / 58.14)  *  7973 = 6171   (notice that it actually less than TCs)

Stress =  6171 / 0.12255 = 50,355 psi.

With grade 8 having a yield strength of 130000 PSI, neither of the 2 cylinder studs are torqued to yield. M8s aren't even close.

[Coyote]

I'd post pictures of the calculated results but the admins would likely delete due to the copy right restrictions at the bottom of the page on this link.

You're a smart guy.    https://imgur.com/upload

[Admiral Akbar]

I'd post pictures of the calculated results but the admins would likely delete due to the copy right restrictions at the bottom of the page on this link.

You're a smart guy.    https://imgur.com/upload

Well maybe not.. If move the images to the link, I can post links to the images here? 

[Coyote]

You can use the link to the image to embed it in your post using the IMG tags. Like this one.

Code Select Expand

[img]https://i.imgur.com/K2g30vZ.jpg[/img]

[Don D]

I am not trying to link this torque discussion to other discussions I have been in regarding cylinder distortion.  Neither the TC or M8 engines barrels have significantly different bore dimensions in plates when torque values are raised or lowered, + or - 10%. Of course this all happens in a static state at room temperature. 🤔

[Norton Commando]

The torque method has an error of +/- 25% whereas the angle or turn-of-nut method has an error of +/- 15%, per the 26th Edition Machinery's Handbook.

Jason

Not sure that this is true. Maybe it's miss-interpreted.  I have 26 can you post the page?

Look at page 1480; I posted it earlier but someone removed it.

Jason