Engine/Transmission (1998.5 - 2002) Cracked manifold

[isb360]

I just posted to a similar thread where a member describes a crack in the manifold as shown in the photo here, except he had a pre-turbo pyro (probably crack initiation site).



I've read nothing about manifold cracks, then in the past couple of days, they're popping up out of the wood work. What gives?



Thoughts?

 

[2-ND TIME]

Are you letting your engine cool down for 5+ mins before you shut it down?? Have been doing this for alot of years and have good secess. I have the dreaded 53 block and can't afford a problem. All this with about 300,000 kms.

 

[curatchko]

A sure fix for bolts that are backing out is to have the bolt heads drilled for wire. Wire all the bolts together across the manifold and put twist in the wire between each bolt, and that'll solve the problem.



Chris

 

[Hohn]

Originally posted by curatchko

A sure fix for bolts that are backing out is to have the bolt heads drilled for wire. Wire all the bolts together across the manifold and put twist in the wire between each bolt, and that'll solve the problem.



Chris

If I may add:



Safety wire pliers are available from Eastwood company and other places. They are really a cool tool and under $20.



Justin

 

[RustyJC]

I really don't want to be accused of hijacking this thread, but since the last few posts have dealt with bolt loosening, it might be worthwhile to interject a little bit of fastener theory and application information.



If the manifold bolts in question are being torqued to yield, by definition any further bolt stretch will not be in the bolt's elastic deformation region (where the bolt will act like a spring and return to its original length when everything cools down), but rather the bolt is being stretched in its plastic deformation region. In the plastic deformation region, if the bolt is stretched by the expansion of the exhaust manifold at operating temperature, it will remain at the longer length when it cools down - therefore, it will be loose (or at least below its original prestress) when cold.



If this is the case, safety wiring the bolts may indeed keep them from backing out, but it will do nothing to ensure correct fastener prestress after a few thermal cycles.



Rusty

 

[Rman]

Rusty... . whats the answer then? How can these guys solve the problem? I used to safety wire tons of stuff (aircraft maintenance) and am very good at it, but i also hate it .

 

[RustyJC]

1. If the bolts are currently being torqued to yield, reduce the initial torque (yes, that's right) to keep any thermal stretching in the bolt's elastic deformation range. Of course, any fastener that's been torqued to yield should be replaced before doing this.



2. Use longer bolts with spacers (take a look at the 2002 exhaust manifold arrangement. ) This reduces thermal stretching per unit of bolt length. For instance, if the exhaust manifold grows . 012", a 1" bolt (that's 1" from the bottom of the bolt head to first thread engagement in the bolt hole - the working length of the bolt) must elongate . 012"/inch, but a 3" bolt only elongates . 004"/inch. Again, this elongation must take place in the bolt's elastic deformation range.



3. Use a device such as a Belleville washer stack under the bolt head to absorb the exhaust manifold expansion without overstressing the bolt.



Now, the bolts can be safety-wired or have the Cummins-type retention lock plates installed to keep them from vibrating loose, and the bolts should maintain the desired prestress.



Rusty

 

[Rman]

that's exactly what i was thinking. One of my friends is a structures guy on aircraft and i've had some discussions with him about similar things. I thought by not torqueing them down so much you'd keep the elasticity of the bolt, however i didn't know if you'd risk the chance of backing out due to vibration.



Good answer, i'm sold!

 

[rweis]

This is exactely what I am about to test.



On the 1st attempt I used the cap bolts that came with the manifold. I torqued them to 32 ft #, rechecked torque and retorqued after 5 heat cycles or so, rechecked torque and retorqued again after 5 more heat cycles and mine was not maintaining torque. Two of the cap bolts even unscrewed themselves completely out, lost one overboard, other was resting on the top of the manifold when I was under the hood for another reason.



Hence I knew there was something not quite right.



I first bought 10x1. 5x47mm studs. Put a manifold washer, split ring lock washer and a nut combination on each. The split ring lock washer could not handle the temperature and lost all its "spring" allowing the nut to freely turn in about 3 weeks. On the 1st attempt I also torqued the studs to 32 ft # which takes them to their plastic limit, not good. Then I started trying to figuer out what to do. In the interim I locktite the nuts on to keep them in place.



When I decided the do the Belleville route I measured the exposed stud length for the 47mm stud, figuered the washer stack depth, the Belleville stack depth, the nut depth, the lock nut depth, plus 3mm and I bought studs 10x1. 5x52mm (the next regular stock length) regular grade 8 steel studs.



Then I found a source for the Belleville washers. The Belleville's I got can withstand 1200*. I know the stud system in the open air is not going to be a hot as the actual exhaust gas, but I don't want to do this project a second (actually 3rd) time. I have seen exhaust manifolds cherry red hot. How hot is that? I know for sure it is not 1200* and that is why I went that route.



Someone on this forum did the calculations for how much the stud would be elastic, then figuered how many Belleville washers would be needed per stud, to match the growth of the ATS manifold and the whole thing MAINTAIN 16 ft # torque throughout the range.



Then you have to put a lock nut, or safety wire, or retainer of some sort, or locktite?, to keep the nut from backing off. I chose a stainless steel Drake locknut as the retainer.



Then as the Belleville's become more and less conical they need to ride on something that can be sacrificed. I did not want them riding on the ATS manifold boss so it would not damage it or cut into it. I am puting a stainless steel fender washer between the Belleville's and the ATS manifold boss. The reason I am using a fender washer is all the stainless washers I could find were very thin (reseller keeping the price per each down I think), or barely would provide a bearing surface for the Belleville cold. As I understand it the Belleville stack is going to flatten (and I would bet grow slightly in diameter) a small amount as it accounts for the ATS manifold boss growth. So I got 3/8" SS fender washers and carefully ground the eliptical shape of the ATS manifold boss into each washer so they fit correctly and flush to the ATS boss.



Then I got 12 regular SS 3/8 washers for the nut end. Since the Belleville's cone small end is at the nut end, nothing special needs to be done with these. The washers just keep the cone end of the Belleville's from cutting into the nut as the Belleville's flex.



I am going to screw the studs into the block just snug tight and probably put some locktite BLUE (ie removable) on the thread in the block to insure it stays put.



Then the exhaust gasket. I could use someone to chime in here if the gaskets need to be changed after being "over" torqued. Yes? No?



Then the ATS manifold. And I think put the fender washer, Belleville stack (3) and buffer washer and start the nut on as I put the ATS on because there may not be enough clearance to work the washers and nut between the stud and the ATS casting if the ATS is fully against the block. I think ther will be about 3mm clearance of the stud to where the heater hose bracket area of the manifold casting is (the center ports of the manifold).



Then work it all together torqueing the lock nut to 16 ft #.



My Drake has a lock nut for the nut. I am going to torque the locknut 10 ft # against the main nut once I get the main nut at 16 ft #. Might put some locktite BLUE on the lock nut too.



Has been an interesting ride, but seems logical to me. Calculations bear out the necessary flex of the Belleville's and stud to match the manifold boss growth, and the stud remains elastic.



Bob Weis