DC's 1756VK-assisted B4

Transmission
A couple weeks ago I pulled the transmission to 'fix' the spring to keep the release lever in the right spot on the pivot ball since it seemed like it broke. Good thing I did because after about 2 years on the stage 4 clutch, it had already started to bend. Since the deflection was perfectly visible for me, I was more than happy to replace the lever with a strengthened one before this one broke.
Click to embiggen


Decided to install a brace for the output shaft too. Required a little bit of dremel work on the housing and the plate itself to clear the release lever.
Click to embiggen


Turbo
Here is the vane gap I speak of below:


While I'm doing a once over on the turbo, I noticed some decent amount of play and associate wear on the shaft where the spherical bearing mates to the vane actuator.
Click to embiggen


The wear on the bearing was uneven but not too bad. It had an ID of 7.75 mm and from what I've seen, the new ones have an ID of 7.6mm.
Click to embiggen

Click to embiggen


That being said, I didn't want to spend $10-20 for a new one so I went for the next best thing, 2 layers of metal tape. The differences measure out to 0.75mm of play. The most immediate fix I could think of was a couple layers, 2 in the case, of metal tape and some general purpose bearing grease on the bearing in the rod and on the pivot.
Click to embiggen


The grease may not like the heat by the turbo, but there is still some play in the bearing's seat so I'm not too worried about it freezing.

Intake Pipe
Now, I'm not sure where it is from and I cannot find my post with the original photos, but the TIP was half stock, half from another non-vw car made from rubber. It has a coupler for the 2.25" OD of the GTB1756's compressor inlet, but the other end of it was meant to be clamped down onto a hard pipe or sorts. However, the ID of the stock B4 accordion pipe is larger than it's OD and has been best served being secured over it with a worm gear clamp. I've gotten it well secured in the past and it's caused no concerns. A look at the compressor turbine shows no added wear from a possible non-flush mating between the two.

Well, since its made of rubber and I had spare pipe sections around from doing exhaust work, I cut off about a 5/8ths section to fit into the rubber pipe so a flush mating surface and higher pressure clamp could be used. Had to pop is out and trim off about 4 mm over 2 test fits, but they both mate cleanly now with a standard constant tension hose clamp from VW.
Click to embiggen

So, its been about 8 months since my last post. About 2 weeks after that post I joined up in an electronics design competition, which helped me greatly advance the progress of the work I'd been doing even though I didn't place. I've put over 15K miles, including a nice trip to Death Valley, on the engine and its still performing admirably. I've fixed a few quirks here and there such as leaking injectors, broken spring for the clutch release lever, tightened down [with new] head bolts, new though unfinished downpipe, and some updates to the tune. There are still a few minor oddities I'd like to investigate and 'correct', but I digress.

While on vacation away from work during our busiest week at work, I opted to pull the turbo and look over things for a hot second. Why? Even though the car drives very well, the turbo spools so fast that it nearly snaps your neck right on the nose of 2100 RPM but before that the engine just feels bogged down and choked. My economy has been no been averaging 44-46 mpg with 1 tank barely at 50. Given I used to be able to get 52-55 mpg tanks, this didn't feel quite right even if I didn't use the power available during a whole tank. No extra smoke and overall the tune produced a nice puff on spool up. Previously I swore I set the vane gap to 1.2 mm and though I've not had time to install an EMP gauge, as I've generally had the drive-more-worry-less mindset with how busy life is, checking things out to their fullest extent hasn't been feasible. Now, I figured that with such a small vane gap that the fast spool and sudden onset of torque @ 2100 was mostly a function of that small vane gap so I wanted to pull the turbo and double check everything. I've had my economy go up since the engine has been mostly broken in and oil consumption is very minimal with maybe 2 of the 9 dots on the dip stick every 2500 miles, which seems to be about 0.1 quarts at its worst and with the CCV vented to atmo since oil in the intake clogs the MBC this feels very minimal. All in all, I've been guessing the engine is being choked a little with the vanes being set too close too much.

Pulling the turbo today, I noted that I have an exhaust leak that only shows on the mating surface of the exhaust housing and adapter on the exhaust manifold, but there were no external soot makes on anything else. I'll grab a slice of copper and make a custom gasket to fix this. However, when I measured my vane gap it didn't show up at 1.2mm... I'm measuring it by taking my calipers, using a miniature screwdriver set, and seeing how thick these drivers are. They range from 0.7mm all of the way up to 6mm for a small hex set. After futzing around with the vanes a bit, I noticed that they have a play in their mechanism so they can be closed down to the 1.2mm that I set it to, but they can also open up to ~2.3-2.4 mm with the light press of my finger. the actuator level didn't budge and was securely pressed against the stop screw then I took these measurements.

My general question is this: What doe the vanes prefer to do in the presence of exhaust? They are airfoils so I don't know if they will prefer to open up completely or close down somewhat, let along to which point would the prefer to close down to? Basically I'm asking if the range of the gap being 1.2 to 2.4 mm with the play in the mechanism, to which extreme will they be biased? This is a rebuilt GTB1756, but from just the logs my tuner says that the turbo spools faster than even a VNT-17.

The tune can be adjusted to compensate by pulling less vacuum, or the stop screw can be adjusted. Despite the play that I notice now, the turbo show no signs of wear with neaxt to zero radial and axial play in the shaft and I have no hesitation trashing the car so the vanes are clean and move completely freely.

D.C, I always enjoy the thoroughness of your approach.

Because of the feedback nature of the turbo boost control, I think some
'free-play' in the linkage will not affect the spool-up or output as long as the
limits can be reached. I have no data to support this, though.

I like the quick spooling of the 1756 even with my mild tune and fueling.
This may be the year that changes, too!

flee,

Thanks for your comments. If I'm wrong in anything I post, I like to know why, though I admit I can be a bit verbose. I am happy that others are getting something from this if they travel down similar paths.

If this was still a 'vk' turbo, I'd agree with you. Feedback in a control system means that the system is aware of what physically happened when it told this one thing to do something. That would mean that the ECU knows the position/length of the actuator rod on the turbo when x amount of vacuum is applied. This is what those "smart" actuators do on the newer turbos.

That being said, the rod moves what? About 5/8" full travel? This is would be ~15.9 mm. The 0.75 mm is then ~4.7% so yeah, it's not too much, but I want to be a little neurotic.

All in all I like how the turbo is responding. I need to drive it around a bit more to see how it's behaving as a whole though I can definitely say that 2 turns of the stop screw, providing a minimum vane gap of 1.6 mm is much better. The car still kicks a bit @ 2100 in 2nd, but not breaking traction all of a sudden and the power is noticeably better below that spool point.

I think I'm having some lifter clatter. It developed in the latter third of the trip out there. All components have equal wear hours and about 14K miles on them. Coming back it got worse when driving at 60 mph. Went away after about 5 minutes at 70 mph. Came back about 30-40 min thereafter. Stopped to make a pit stop near Phelan for about 10 minutes and playing more with the power band after the Cajon Pass, I didn't hear a thing.

Anyhow, I still need a shorter oil line so I'm okay pulling the turbo again to tweak the stop screw with a 1/4-1/2 turn longer if need be. I'm highly interested in the correlation of the vane gap and EMPs. That and working on the exhaust are the last couple crucial things I need to complete before I consider additional support upgrades like fuel lines, aux pump, and intake manifold. Those will come after I make it to a dyno to see what the rebuild did.

Edit: What else are you plans?

vanbcguy said:

Digital Corpus said:

There is one itty bitty problem with going with this manifold, the turbo is positioned closer to the firewall. I guess there is supposed to be a heat shield that one bolts onto the hotside, but I don't have one. The upper mount nub allows for about 1/2" clearance with the firewall. Unless you plan on using it, cut it off, imho, and give yourself more clearance.


If you don't, your firewall could end up looking like this:

Click to expand...

I wish I had listened to this - dropped my engine in yesterday. I had this in my head as a 'to do' but never actually got around to it. That nub definitely is going to rip up my firewall insulation over time.

Click to expand...

So I got around to this 2 days ago before the 16 hr day, half of which was driving, out to Henderson, NV yesterday to visit a friend from work 'cause another friend was in town.

Anyhow, the list of changes were remembering that there are 6 bolts that hold the hotside housing on, not 5. I also picked up longer screws and shortened them a touch in order ot have more threads interlocking. However, despite having only 5 screws holding on the housing, there was no exhaust leak . I finished cutting off both nubs and with a 1"x30" belt sander I curved the offending edge a bit more.

That was accompanied by the stop screw being adjusted. I opted for adjusting the stop screw to account for the minimum vane gap and went for effectively 2 complete turns. Confirmed that this yielded a 1.6 mm vane gap. Turbo feels noticeably different and the powerband is smoother as I noted.

Replace the driver's side sway bar link/bushing. Remember, to squeeze a 20 mm ID rubber bushing over a 24 mm OD steel bar, Dawn dish soap and water are your friend. It'll take you 2 minutes. Try motor oil or WD-40 and it'll take you 45.

Lastly on the assembly, I used a milled-flat block as a guide and sanded the mating flange of the turbo hostside inlet and downpipes and fabricated a 0.025" copper gasket for it to mate to the adapter plate on the exhaust manifold.

I had trouble originally getting the turbo off the car because 2 of the studs were a touch too long; and the nub hadn't been cut off. I had to double-nut the studs and remove them to get the turbo off. To ease the job in case this had to happen again in the future, I cut the studs down a bit. Measuring the one that didn't have to be touched, I noticed that it had near 20 mm height from the surface of the adapter plate so I cut the others down to size accordingly.

Now, I had exhaust leaks. Again. Hence the copper gasket fabrication. But from that I noticed that I had a stud come loose. After quite a bit of research, the adapter I had was made from cold rolled 1018 steel. The manifold was some alloy of cast iron. I had guten-tight hand torqued all of these down before and that obviously hadn't worked so a more technical approach had to be taken.

The threads being torqued down were M8x1.25. Since this was a grade 12.9 stud, I looked up torque values which turned out to be directly associated with the hardness and/or tensile strength of the material. Long story short, the best information I could find and thus values I could "guess" at were torque values of 8-12 ft*lbs else I'd risk stripping the steel of the adapter or manifold. Long story short, from the feel of the metal when torquing the bolts down, I limited myself to 10-11 ft*lbs. One of the I attempted going to 12 ft*lbs, but it began to feel too plastic like I was about to strip the threads. With cleaned threads and a dose of 2200°F Stainless anti-seize, I torqued the nuts holding the turbo down to 9 ft*lbs assuming they'd partially act as if oiled.

I moved to stud+nut because I sheared off two, grade 12.9 M8 bolts when I pulled the turbo while the block was being worked on last year. I don't know if anti-seize was used on them and I don't know what they were torqued to, but they were damn tight and it all still leaked. When I pulled the return oil line for replacement, I may spend the time to examine if the torque values I used were enough. For anyone else educated in material sciences and can lend advice if this was good enough or not, definitely if this proves to have not been tight enough, I have no qualms to buying a new adapter or an appropriate grade of steel from a yard near by and getting a few adapter fabricated that can withstand the required torque.

One thing I realized today that I never detailed was the 4-bar MAP sensor upgrade. Some have moved over to the MKIV sensor upgrade, but I decided to stay with stock plumbing. There are upgrades available from Malone, but I didn't want to spend that much for a PCB and a $10 sensor. Going from kooyajerms' post from back in the day, I had added in a MPX6400A to my ECU when the turbo was installed.

The first PCB I made for this was equivalent to fugly and flawed, but I made it work and work it did for 2 year until I replaced it last night.
Click to embiggen:


Well, as mentioned, last night I decided to redo this in a cleaner manner with a revised PCB I had made. The dimensions make it so it fits in the holes left over from the stock 2.5 MAP sensor that was originally present.

The back is supposed to be laid out so you can add a dual DIP switch for the manufacturer suggested decoupling capacitor, C2, and an RC pair for a low pass filter set for a frequency of your choosing using US 0603 components. However, that portion of the layout is flawed though you can still wire the sensor in directly without any of those present. I'll probably update it in the next week or so in order to make it fully functional...

Click to expand...

Yeah, well ignore this bit... I ended up using 0402 components which are a tad too small for most. Though I though the PCB design was flawed, i did have it match the datasheet for the suggested layout:



Anyhow, the component layout is now 0603's, the silkscreens are aligned better and should be more legible, and I altered the body of the PCB a tad to get rid of hard corners, let alone grip the MK III ECU better. The new design is still that old link up there



The downside to using this sensor is that the nipple is tiny so I had to make a JB Weld ridge to help out.
Click to embiggen:

The ridge is fugly because after working on my axle for so long, I forgot how how PCBs get when soldering and dropped it when I tried to pick it up. However, I found out that the heat from the hot air rework gun was enough to affect a shortened cure time to the JB weld

Profile shot, click to embiggen:


If you're considering this yourself, keep in mind that the 5V through hole contact on the ECU's PCB is likely connected to a massive power plane. Why do I say this? I had to crank up my iron to 420°C just to flow the solder after a couple minutes of contact. Ground is similar, but I'd still suggest ~400°C. the output signal line was fine with 330°C.

Axle info coming later...

Here's some news. When I have anything finite I'll let y'all know. It happened a little over 7 hrs ago and I'm still at work.

http://forums.tdiclub.com/showthread.php?p=4765900

My worst nightmare. Good luck.

Abacus said:

My worst nightmare. Good luck.

Click to expand...

Yup. It's the only thing in my mind that will make it so I cannot drive my car.

I, m finished my vacuum conversion on gtb1756vk, also noticed that vanes have some play, I set stop screw aprox 2.0mm gap, and I will start from there, gonna use emp and boost to adjust from there.
it willbbe pita to adjust while on engine... but it will have to be done that way. I belave that gass pressurenforse vanes to open more so under pressure its tight control.

So, over the past 5K or so, I've been monitoring debris in my coolant. I just finished pulling the head and I have a definative breach in the back left corner right towards the coolant galley. Aside from my futzing around with general curiosity, I'm about done for the evening so I'm cleaning up and going to bed soon. I'll take photos tomorrow and reassemble everything.

Pistons and head have quite a bit of carbon on them, but I also put fuel timing back to center to see how the engine performed. Subsequently, I got quite a bit smokier.

Side note, the 1Z cam sprocket weighs in at 741.7 grams and the ALH weighs in at 645.5 grams. At least the 2 I have to sample from.. I'm sticking with my heavier one...

So, why did my head gasket fail? I think it comes down to 2 factors:

• Surface rust in/on 2 to 3 threads in the block for the head bolts
• Thread engagement

Stock head bolt:

M12 x 1.75 x 115 mm, Class 10.9

Click to expand...

Off the shelf "head bolt":

M12 x 1.75 x 110 mm, Class 12.9

Click to expand...

Yes, there are ARP studs but those will/would set me back $170 at least and if there is a less expensive, dig-into-my-pocket-for-a-brand alternative, I prefer to use it. My time is my time and I don't equate a $$ value to it so changing a head gasket or head bolts comes down to timing and cost. So, for ~$20 I get a stronger bolt, which I've opted to go for per the advice of a trusted member of the forums here. How much stronger though?

I was originally pointed to the tensile yield strength for a ballpark comparison. A class 10.9 bolt is about 1040 MPa and a class 12.9 bolt is about 1220 MPa, or a ~17.3% increase in minimum yield strength. ARP studs are 200,000 PSI. Sounds impressive, but that's ~1380 MPa, or ~13.1% higher than class 12.9, ~32.7% higher than class 12.9... for 8x the cost.

Anyhow, I spent a portion of the evening last night looking up pre-load and clamping force, found a detailed but rough calculator, and though I didn't want to break out my calculator, I came across another factor that appears to impact how well these bolts work in our intended use, proof load/strength. Between class 10.9 & 12.9, 830 MPa & 970 MPa, there is a ~16.7% increase in strength. And, as you torque down the bolt, you get a greater pre-load. If yo uhave more thread engagement, you get a greater pre-load. It wasn't setting well with me that I was using a strong bolt with 5 mm less engagement, though I didn't know how much thread was in the block.

Well, the head is off and I've measured it out for a 1Z head, which was shaved before I unstalled it last January, and I'm getting a little over 79 mm from the washer seat to the surface. The washers I'm using are ~3 mm thick, but can compress a bit so I removed 0.5 mm from this addition. Because of the guide in the head to ensure the head gasket is lined up, each of the, I don't know what to call them so "ports" will be used, ports for the bolts has 12.5 mm to 13 mm of depth before the threads begin. The port bottoms out as 45.75 mm in all 10, but the bolts stop threading 40-41 mm from the block deck depending on port. I'll add a nominal head gasket thickness of 1.5 mm to the math.

In summary:

• Head thickness: 79 mm
• Washer thickness: 2.5 mm
• Head gasket thickness: 1.5 mm
• Depth before thread engagement: 13 mm
• Stock thread depth available: 27 mm
• Theoretical thread depth available: ~32 mm depending on the tap

This means that a 115 mm bolt will have about 19 mm of thread engagement. So my 110 mm bolts only get 14 mm, which is nearly half of the available threads. Having no knowledge of engine blocks aside from my 1Z, I say oil pooled in all 10 ports means they're not connected with anything, with a dry engine, so going with a longer bolt, 120 mm, would put me at 24 mm depth. I could go with 130 mm bolts and grind off 5 mm, but I'd rather not. I have a tap that I ground a bit of the tip off and used to chase the thread on all 10 ports and to give me another ~2 mm, 360°+45°, just to be safe.

All in all, despite some claims of class 12.9 bolts not being strong enough to warrant their use as an upgrade seems to me, in my oh so ignorant opinion, to be invalid. I believe it stands to reason that past attempts at using these off-the-shelf bolts not being successful was that there was so much less thread engagement.

I promise to eat my words if I come back 20K miles from now and have to upgrade to ARP studs because I blew another head gasket.

Anyhow, cleaning things up and taking photos now before I put it back together again.

So what head bolts were you running?

I think those are the same head bolts I ran when I was running a GTB2056VL, I pulled 235hp/380ft-lbs IIRC using that setup with no issues. I swapped over to H13 head studs when installing the compounds, swapped them one by one without pulling the head, still holding strong.

A trick I picked up from the diesel truck guys - copper spray-a-gasket from permatex. Spray on a few coats onto each side of the head gasket letting it dry between coats. It seals up quite nice.

How does one determine if the block to head sealing is adequate ? My daily driver @199whp doesn't push/loose coolant however I know that combustion pressure is entering the cooling system.When the coolant smells burnt its a pretty good indication that there is leakage.

I was using the 110mm 12.9's. I'm guessing the slight rust in the threads was the culprit. I don't recall if I posted on here, but I put a 2nd set of 12.9's in and torqued them to 125 ft*lbs oiled. but that didn't stop the leak.

When I took these off yesterday, 2 of the bolts came free rather easily. Even though it's not suggested anywhere here and I'm being rather paranoid this time, should I torque them down and preload them and then go around round robin style and loosen and re-torque them?

I actually have a can of the permatex copper stuff. It doesn't like the heat of exhaust sealing or the surfaces of the turbo were not flat enough. I've had to lap them twice to flatten them. Made my own copper gaskets last time around I checked both and they seems to be holding well.

andy2 said:

How does one determine if the block to head sealing is adequate ? My daily driver @199whp doesn't push/loose coolant however I know that combustion pressure is entering the cooling system.When the coolant smells burnt its a pretty good indication that there is leakage.

Click to expand...

I'll take a photo to post up later, but you'll noticed hard, black, sand-like debris in your coolant reservoir if that is where it's escaping too. I've not lost more than 2 oz of coolant since I've been monitoring this.

JFettig said:

A trick I picked up from the diesel truck guys - copper spray-a-gasket from permatex. Spray on a few coats onto each side of the head gasket letting it dry between coats. It seals up quite nice.

Click to expand...

On second thought, I'll follow though with this suggestion with a couple additions to re-installing the head. Thanks, John!

Progress is slowed because I'm house sitting and tending to a mellow mixed breed small dog and a hyperactive terrier that puts a 2 yr old on a sugar rush to shame. That and with the heat in the region, the ants like the dog food of which the dogs are self feeders :/ so I have to make sure it's safe and go to town with a spray bottle of cleaner to kill the invasion a couple times a day when I'm not at work.

Photos are forthcoming, though may be a little longer since I'm bouncing between 3 computers are 5 OS's trying to organize my personal data, but I digress. A thread from 2008 over on a powerstroke forum linked up to an article about fasteners which has these two paragraphs in it:

The next step up from 4340 or 8740 steel is 1722 (AMS 6304). Manufacturer ARP calls it ARP 2000 and it’s all the same material, a 220 ksi material. The next step would be H11, which ARP calls L19. It too is the same material and also comes from Carpenter. Those are getting up to 240-250 ksi. You can go higher than that but it becomes brittle, according to Trapp.

The next step up include two materials: Custom Made 625 and Aerospace Material Specification (AMS) 5844. Trapp says A1 Technologies gets both materials from Carpenter as well. “The AMS 5844 has a trademark name MP35M, which means it’s multiphase,” says Trapp. MP35N is an age hardenable Nickel-Cobalt base alloy that has a unique combination of properties – ultra high strength, toughness, ductility and outstanding corrosion resistance. MP35N resists corrosion in hydrogen sulphide, salt water and other chloride solutions. ARP calls this material ARP 3.5.

Click to expand...

I don't know if this old information can be corroborated, but it'd be nice to know.

As for how torque values can change, man that has been a research project that I'm sure I don't even get the half of the complexity. Anyhow, according to the thread and a linked PDF, ARP suggested 77 ft*lbs for a class 12.9 M12 fastener. Fast forward to today and they advise 98 ft*lbs.

I'm using a dual washer setup for a couple reasons that probably caused me to spend $2 in in parts with little return, but I've lubricated the surface between the 2 washers with ARP's Ultra Torque. If this holds true, then the use of 0.1 for the coefficient shouldn't be too far off. In fact, ARP's general torque specs equate to coefficients 0.1 and 0.15 for thread and head coefficints. At 70% the yield strength, I'm at ~107.7 N*m, rounding to 80 ft*lbs, of torque. If the two washers weren't lubed, then that puts me at ~132.6 N*m, ~98 ft*lbs, which matches ARP's approximate torque value.

Preload isn't dependent upon torque but upon how much the bolt has actually rotated and it's thread pitch. Torque is just a measurement of the amount of friction that exists with the threads and the bolt head. However, if the coefficient of friction is the same, then a torque value can be computed for a set preload. Anyhow, from all of my reading and using the saying of "an ounce of prevention is worth a pound of effort" a sprayed down head gasket and anal retentive torque cycling of all 10 bolts with a rest period between each for letting air/excess lube squeeze itself out of the way, I don't think I'll have to do this again in 20K or even 100K miles. I'll post up the math later, but 4xM12, class 12.9 bolts with 60 kN of preload per cylinder give enough theoretical preload for PCP's of 460 bar and current tests show that a GTB2260VK and Firad +120 with a SOI of 22˚ yeilds about 250 bar. This might be a half of a pound of prevention, but w/e. Anyhow, back to resealing my IP while I wait between torque cycling the bolts...

Digital Corpus said:

Preload isn't dependent upon torque but upon how much the bolt has actually rotated and it's thread pitch.

Click to expand...

Bingo!!

You can make a career as to all the details behind this statement. I've personally demonstrated that something torqued in by angle with lubrication that only ended up at ~70 ft-lbf will stay together when the prior method of torque only at 125 ft-lbf wouldn't and the excess torque was causing other issues. Nobody believed until I ran 12 samples of each with zero failures and the high torque only method had 5 of 12 fail during the same test.

The trick of all this is knowing how much angle to apply

*** This ratiometric information is relevant only to the M12x1.75 class 12.9 bolts used in conjunction with ARP Ultra Torque applied to the threads and between the two washers ***

I had a chance to run comparative numbers to FUB's back in the old ARP Head Studs thread. With the a length of 79 mm vs the 133 mm used, it appears that the additional preload added by thermal expansion to 200°F nets ~2800 lbf, ~12.45 kN. This would bring the preload of the bolt torqued down to 85 ft*lbs, with lubed threads and washers, to about ~81 kN, of 90% of the minimum yield strength of the bolt. Backing each one out and retorquing to 80 ft*lbs nets to ~76 kN, or ~85% yield. If by chance the coefficient of friction for Ultra Toruqe is closer to 0.08, I would've already yielded the bolt and caused stretch, but by just barely.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In my angst to get things done, I neglected the oil return line as I noticed this morning. I don't have time to attach it and do the fuel pump before work so the car is going to sit another half-day until I get home.

Bolt yield strength only matters when you're fastening them up to and past the yield points of the bolts (ie. using TTY bolts). And even then, using a higher YS bolt but using the same tightening torque will, all else being equal, not yield one iota more clamping force.

In order to have a stronger clamping force (or differently, less stretching deflection), you need:
- more torque
- less friction through lubrication or surface treatment at a given torque
- larger cross-section
- higher Young's modulus of the material.

Regardless of a 8.8, 10.9, 12.9 or proprietary tensile strength grades, all ferrous-based materials will have Young's moduli within a few percent of 200 GPa, regardless of yield strength, heat treatment or alloying. Assuming you employ the same torque to remain within the elastic region of the bolt with any of the above grades, it makes not one iota of difference in the clamping force.

Short bolts are also less affected than very long ones because the stress = young's modulus * strain, and strain is the deflection normalized by the total length.

TDIMeister said:

Bolt yield strength only matters when you're fastening them up to and past the yield points of the bolts (ie. using TTY bolts). And even then, using a higher YS bolt but using the same tightening torque will, all else being equal, not yield one iota more clamping force.

In order to have a stronger clamping force (or differently, less stretching deflection), you need:
- more torque
- less friction through lubrication or surface treatment at a given torque
- larger cross-section
- higher Young's modulus of the material.

Regardless of a 8.8, 10.9, 12.9 or proprietary tensile strength grades, all ferrous-based materials will have Young's moduli within a few percent of 200 GPa, regardless of yield strength, heat treatment or alloying. Assuming you employ the same torque to remain within the elastic region of the bolt with any of the above grades, it makes not one iota of difference in the clamping force.

Short bolts are also less affected than very long ones because the stress = young's modulus * strain, and strain is the deflection normalized by the total length.

Click to expand...

So the additional pressure created by thermal expansion cannot or will not provide additional tensile stress that would cause the bolt to yield if the bolt is tightened/torqued down to "a hair's breadth" from it's yield strength, at least in this cause since we're dealing with dissimilar metals? Is this because the material causing the expansion, the aluminum will deform first?

From back in the day, I forgot to post glamor shots of the pistons and rods married in their respective pairs before installing them in the car. Here is the updated post.

Anyhow, the photos of which I promised last week or so. Before I get to the head, I removed the EGR block off plate that had been installed since it showed signs of leaking. And using my caliper as a the closest straight edge:

Yup, it's a warped sucker



A little propane heat and bolting it upside down in the same port, but with a socket between it and the manifold and some additional torque applied gave me these result:

Click to embiggen:


Tell me, how long do I have to wait before it warps again? I ran across the steel OE EGR plate and want to track that sucker down...

I also tinned the tip, with standard silver/tin electrical solder, of my oil return line in case it's surface was the cause of my forever leaking line

Click to embiggen


Click to embiggen

Digital Corpus said:

...
In summary:

• Head thickness: 79 mm
• Washer thickness: 2.5 mm
• Head gasket thickness: 1.5 mm
• Depth before thread engagement: 13 mm
• Stock thread depth available: 27 mm
• Theoretical thread depth available: ~32 mm depending on the tap

This means that a 115 mm bolt will have about 19 mm of thread engagement. So my 110 mm bolts only get 14 mm, which is nearly half of the available threads...

Click to expand...

How much more thread engagement is a 120 mm over a 110 mm bolt? I don't recall why I set my caliper to 91 mm when I had the above numbers, 'cause it should be ~96 mm, but here is a 120 mm bolt next to a 110 mm bolt:

Click to embiggen:


Now, I'm not sure this next part is, I doubt it is, but the cost was less than a cup of coffee, as some say. I noticed when I pulled the previous set of 110 mm bolts out that the washers were scarred for life:

Click to embiggen:


And with the clamping force on them, they bowed out a bit like putty. Even the seat on the head for the washer showed stress cracks not-dissimilar to the negative of the pattern you see here

Click to embiggen:


So to distribute the load of the bolt on the washer in a more even manner, I sandwiched a hardened washer between the extra thick on and the bolt head. I bought these from my local hardware store called McFadden-Dale, and they only had SAE sizes but 7/16 washers are a great fit for M12 bolts

Click to embiggen:


I ended up flipping the extra thick washers upside down relative to that photo since the other side was flat without the curve edge.


Installation bling. Yeah, I used blue bolts because I could

Click photos to embiggen:

Since one possible cause of having a head gasket breach is not having a perfectly clean surface of the head of the block and the gasket being ground against that debris. Since the 1Z head only have 1 roll pin for guidance during the had-to-block mating, I not only used the previous 3/8 aluminum rods from before, but I added some 1/2" or 13 mm tube to the mix. Cut out in ~30 mm lengths and then modified a touch, I just needed 2 to help guide the head in and keep it from wiggling. Now, if you choose to do this, keep in min that the 4th and 2nd bolt hole from the right and in the front have channels for coolant and oil so you only ever want just a bolt here. I picked the front two corner hole to make these guys with.

Click to embiggen:


Cleaned & Installed. Click to embiggen:


BUT WHAT IS THIS!?!? As I mentioned, I headed JFettig's advice, though I may have applied a little too much. By the way, this stuff doesn't dry so once it's tacky, just lay it down where it needs to go and don't move it.

Purdy!

Click to embiggen:


Click to embiggen:


Click to embiggen:

Now, the keen will note that I cleaned the pistons some but there are some carbon deposits around the crown in that last photo above. There were matching deposits on the head that I cleaned off with solvent and the help of thin, soft aluminum scrap. Photos are in order of cylinders 1, 2, 3, & 4 respectively.

Yes, I confirmed that I had exhaust reversion in cylinder 2. It was notable in the runner.

Click to embiggen:


Click to embiggen:


Click to embiggen:


Click to embiggen:


I don't know if the deposits are from running rich, because that was the last few thousand miles and a result of fuel timing adjustment: middle line of timing graph with an IQ of 5 mg/str. Or if it's from fuel additives, which are 1:128 2SO, ~1:500 of Stanadyne (suggested ratio)? Or if it's a combination thereof?

Will running straight DieselPurge clean it off both?

Could this change in squish volume been enough to bump the head to cause the head gasket failure? If it wasn't my bolts, then it was something else. After all, I want to know why I had the failure.

Calling in a favor with temporaptor to assist, we managed to get onto the same stretch of tarmac we previously used. Inspection of the coolant bottle thereafter a good 30 minutes or so show no signs of a leak. These bolts are a holdin'! Oh, clutch is 80% fixed too, but need to find a better grommet...

In other news, my IC programmer is dead. I'll be buying a new one here soon and I'll test out the updated tune.4

If you buy things online with any frequency, unless for some reason Amazon doesn't have them in stock, buy Prime. New programmer is working like a charm. Already burned 4 chips to verify:

The three main seals on my IP are now replaced. Diesel HPR exasterbated an already leaky pump. In my eyes, 19 yr old seals don't necessitate raising a stink over compensation due to a fuel change. That and 2 head seals cost me ~$16 shipped from DieselGeek. Since I had to reset pump timing and I intended to look at my pump's IQ more closely, that's just what I did.

The default value for adaption is always 32768, though depending on channel, you have a variable range of +/- 200 or less. For channel 1 the range is from 32686 to 32784 for my tune, effectively a GH ECU. THis means a range of 98 only so small changes here had a profound effect and the default value is at ~83.7% in the range. From refreshing my memory of what robonitro and others had posted here and here, I wanted to take a whack at the range I had available. An IQ of ~8, voltage of ~1.8 makes it fairly difficult to start my engine. Bolts for the head cover and the QA housing were torqued to 9 ft*lbs from seal replacement and from stumbling across it in the Bently, only 7 ft*lbs is called for. My loose assumption is that the properties of the Diesel HPR effectively "leaned" out my IQ.

Anyhow through oversampling, I took the 8-bit values for IQ and upped them to 12-bit. Logging for just about a minute and twenty seconds provided the necessary samples, 256. At 12-bit resolution, my QA voltage rests within 1.717-1.721 V before making any adjustments and the current maps I have put this is at ~5.212 mg/str. I pulled these samples at 0%, 33%, 67%, & 100% marks in the adaption range for channel 1 to see how the respective IQ lookup is plotted:
Upon a little math it was evident that the curve is similar to exponential, but I couldn't get a good fit. Even when I distributed 10% range differences linearly and logarithmically, things didn't quite line up so I'm not sure what math is going on in the background from the 4 samples I pulled. Anyhow, since I adjust my B4's IQ to acceptable smoke level for the summer, I find it more beneficial to be able to "turn up" the fueling when it's colder out, thus I decided to go for 20% in the adaption range, 32706, and then hammer the QA housing of the pump to my new IQ. With hammering, I adjusted beyond the IQ I previously had to see if prior parking-lot-speed drivability issues, and from-stop clutch engagement issues would re-present. I ended up with an IQ of ~4.813 mg/str and a QA voltage of 1.624 V.

I drove over to top off the tank and see how things were behaving and I'm rather pleased with the results. Idling in 1st and 2nd behave just as they did when stock and tapping the clutch to kick off cruise control creates a notably softer transition into engine braking and shifting was actually more pleasant. It's about the little things in life. Anyhow, once I fab up a steel EGR block off plate, or buy a threaded variant for EMP, I'll do a proper test for economy between the Colt Stage II cam and the OE cam.