Logo
FULL SITE MENU

Overheating Engine, Part 3

C

CaboJohn

Regular Contributor
I am finally back in Cabo with all the tools, gaskets, etc. needed for the tear down. My two past threads give all the gory details. We all decided (me, Chris, Rev Bob, Ricardo, etc.) that the head needs to come off. I did the test suggested by Chris where I looked for an immediate pressure rise on a cold engine. It was there, it went up proportionally to rpm, and it remained after shutdown. At lower rpms the leak is very small (2-4 psi), and at idle and below (off) it closes down entirely. At 3300 rpm it jumps to 6-7 psi and vibrates wildly, and outgasses over the pressure cap. And that was with the thermostat pulled and the temps never going over about 130.

The temp gun says the problem is on the left cylinder bank. Today I first checked compression. #1 170,#3 172,#5 178,#7 177. I don't see any obvious issues there. Then, with unbelievable difficulty, I managed to get the valve cover off and then loosened the valves. I rigged the compression tester hose as Chris suggested and put 140psi on each cylinder for 7 minutes. Nada! No pressure rise in the manifold whatsoever. The leak, if there, is very small indeed. But it sure seemed to leak seriously into the coolant in the past when running above 1500 rpm or so.

So now what? I had to remove so many wires, cables, plugs and brackets to get at the valve cover that I suppose I should keep going, drain the coolant, and pull it apart. I just wish I knew what to look for and exactly where to look.

My current "theory" is that if I look hard enough I will find a crack in the block that opens up slightly when the blocked experiences the torque loads it sees when running at higher rpm, particularly under load (not at the dock). The crack stays closed when the engine is not running and mere pressure is not enought to open it up. I am really hoping that is an incorrect theory and something cheaper is going on. But we shall see.

Question. Is it possible that at higher rpms there is enough exhaust gas pressure in the exhaust manifold to overcome the paltry 7 psi or less in the coolant system and gas could enter the coolant somewhere in the manifold or where is joins the head? It seems unlikely but then all of the stuff we are considering is unlikely. The boat had new manifolds put on just before we bought the boat, and that was done to replace an earlier replacement that was seeping coolant.

All comments and suggestion welcome.

Thanks,
CaboJohn
 
You are in new territory for me. At this point, it could be a crack in either the head or the exh. manifold. I tend to think it is not the block. Did we already ask if this engine was exposed to freezing conditions? Maybe a 100psi test on the manifold? I'd fill it with water first! If the manifold is not it, magnaflux the head, looking for cracks around the exhaust seats.
 
Last edited:
if the exhaust system is sized correctly, it should never happen. If laid out and installed correctly, the backpressure should never exceed 4" Hg....and the closed system's pressure is much higher.
If you pull the head get it checked thoroughly like DD suggested...don't know if the required skill is available locally. Don't forget to check it for flatness (warping). it could also just be the head gasket. As you suspect, the leak only exists after the heat has expanded things..
 
Thanks Makomark. I could not imagine there being enough pressure, but I had to ask. Also in thinking about the situation, I was not as surprised as I was at first that I could not get air flow into the coolant at 140 psi. I mean with those compression numbers the engine is pretty tight.

Moving on. I got the head off today. It is apparent that there is a combustion leak between #3 and #5 and there is also small staining that leads one to believe that the leak sometimes progresses to the water passage between and slightly below where those cylinders are near each other. This looks like the likely cause of the combustion leak into the coolant....but I would not call it a smoking gun. Still the combustion gas has clearly gotten behind the sealing rings in the gasket so this is likely the problem. Sounds better than a cracked block.

I have not cleaned the head up enough yet to take it to the kitchen and do the flatness test with feeler gauges on the granite. But using a carpenters straight edge I can see a slight bowing, which would fit with the leak in the very center of the head. So it is likely I will have to find a Mexican machine shop to flatten the head, and hopefully they can magnaflux it for cracks as well since everyone says it is a good idea to do that when you have it apart.

Incidentally, the inside looked awful with a big carbon buildup on all cylinders (it is kind of brittle and you can flake off big hunks) and light rust on the cylinder walls. I ran the engine for a minute or so two days ago when I was testing the port side engine (I did some minor work), but shut the starboard engine that is now apart down right away. I was surprised to see the rust but maybe it was there from the month off. Or maybe it happened in the last two days. The plugs looked like the engine was running rich, so if I manage to get it back together (ha-ha) I may need to look into a tune up. Or do what my Dad used to do....run his car flat out to "blow out the carbon".

Anyway, at least I found something that was clearly wrong which is a lot better than finding nothing.

CaboJohn
 
thermal expansion - almost as powerful as Ice is in the winter reshaping thing....

Glad you are going forward down there...I'd check the granite before using it as a reference. A good old fashioned metal square is usually "straight" - your eye can verify.
 
Head is at a machine shop to be flattened. No Magnaflux available.

Is that red spray-on high tack sealant so loved by mechanics to hold gaskets in position the right stuff to use when assembling with marine head gaskets? I understand some kind of sealer is mandatory.

Thanks

CaboJohn
 
My gasket is a Mallory 9-63706. Looks just like what I took off. I "assumed" that as a marine gasket it was classified as metal, and my Crusader service book said all metal head gaskets need a sealer. But it is an old book and I now see the gasket is some sort of metal/composite, and have read that many of these have a sealing coating already applied. Yet many experienced assemblers claim they spray seal everything. As Dave noted, Felpro does say do not apply sealant. Mallory, in their literature, is silent. No instructions came with the gasket. I will try to call them Monday unless someone has more definitive information on the Mallory product.

CaboJohn
 
John,

Good to see you back aboard, and well underway on the project! I would recommend just a clean, flat, dry metal on metal contact for the head gasket surfaces. The intake gaskets are another story, they'll need to be held in place (a shot of 3M adhesive spray on one side).
 
Last edited:
Yes on the clean and dry. Meaning all traces of old gasket removed,even if it means a steel scraper. Acetone or alcohol and lots of paper towels. Careful with rags, their link does,t dissolve like papaer.
 
I have new head bolts (with sealer). But I just read an opinion that if you machine one head flat, then you must machine both (the same amount), and also the intake manifold if you want it all to fit together again, particularly with the low torque, plastic GM manifold gaskets. Tell me it ain't so!

CaboJohn
 
John....It ain't so!!!!! On your engine anyway John. That gets into another can of worms involving deck height, don't worry about it! :cool:
 
I'll tell you "It depends", mostly upon where you were when you started and what the shop has to take off to get the head 'flat'.

Ideally, you would want to do both heads (block face) (usually not necessary) and then the intake gasket surfaces. Sometimes the end rails, too.

Best thing to do is check the intake to head gaps when they are done with the head. As a good rule of thumb, if the intake retaining bolts thread in w/o any interference, you should be fine.
 
Head gasket torque? All my data gives foot pounds of torque and proper order. But I saw a video where the method was to lightly torque the head bolts and then further tighten them in order a specific number of degrees (I think it was 40). I would guess this would insure that the head remains flat, and the head gasket is evenly compressed by taking out any variable caused by dirt in bolt holes or soft/hard spots in the gasket. Of course the mechanic, then "just to be sure", went back and tightened to torque specs anyway which defeated the whole purpose.

Is this tighten by degrees and not by torque becoming an accepted method? I don't plan to do so unless someone tells me it is needed. I am using an old book so I could be out of date.

CaboJohn
 
Engine is back together and runs smoothly. Took it out for a sea trial yesterday. Warmed it up, cranked it up to about 3100 rpm and in a few minutes the alarm went off. Exact same overheat problem as before. What a bummer. Successful surgery, but the patient died.

So, we know the head gasket failed between #3 and #5 and looks like gas worked its way to a coolant channel at that point. I would send a pic if I knew how.

But....I was never able to get any gas leak on a static engine, either in a compression test (all cylinders between 178 and 170) or into the coolant (140 psi air in cylinders did not leak into the coolant). So there are some conflicting data points.

I think I am left with only a few possibilities. A distorted block that need refacing (it seems unlikely but why did the gasket not seal?). Or a really out of spec Harbor Freight torque wrench (it said I put all head bolts to 66 ft. lbs. but could it be off?) Or a cracked block as an additional issue beyond the leaking head gasket. One mechanic does not think so, another thinks it likely.

I am tempted to crank the torques up to 100 ft. lbs. and try again, but likely that would just break the head or something else. Or maybe the thing would seal.

So at this point I plan to take it to a yard I have located that can pull the engine and tear down (at $85/hr.) and see what they can see. I am coming home on Sunday and will not be able to come back here a do that until mid July.

If the block is cracked should I just shop for a block, a short block, or a complete engine? It has 400 hours, runs great, good fuel system and mostly new or freshly cleaned parts (new pumps, cleaned heat exchanger, new exhaust manifolds and risers, etc.)

I am bummed.

CaboJohn
 
Was there any testing done on the exh manifold? Did you straight edge the area of the head and block that showed the gasket issue?

Torquing by degrees is done for yielding bolt designs, which is NOT the case with your engine. Three or four trips around (25% each increment) in the correct order is the correct method for Gen IV.
 
I am not sure what I could test with regard to the exhaust manifold. It is fairly new, runs cool and seems to "exhaust". I machined the head but did not straight edge the block (it looked fine, and if warped all could do is reassemble and hope, in any case). As the the head bolt torque, I did work up as directed and then even checked again the next day to make sure everything was correct.

I played golf this morning and tried not to think about boating. Now I need to go into the shop and work on my new harpoon so I will be ready for the 300 lb. swordfish this fall (assuming the motor get fixed).

CaboJohn

Cabo
 
Cabo I just went through a similar situation with my boat. I have had both heads completely replace along with manifold and elbows. I have traded rRuhr heat exchangers from one engine to the other and still have the problem. My boat never over heats but I getting a ton of bubbles in the reservoir and the one manifold registered 240 with an ir fun. I am actually sick of the problem... Good luck!
 
John:

you never had the head magnafluxed so the head is still suspect.

you can also do the "leaking head gasket test" using other approaches...can email some manual pages if desired.

regarding adding the pictures, there are two approaches: 1) post them to photo bucket or some similar site and then add the link to your post here, or 2) use the "go advanced" reply button and then scroll down and you'll see a "manage attachments" button in the additional options box. the forum help pages work ok, too.
 
@Mako,

Heck, everything is still suspect. But....I have gone back and looked at the many many past posts I made and received to get the history again clear in my mind. Then I wrote up where we are at in a 5 page document for my partner (I'll pm it to anyone interested). But the long and the short is I don't think I have a combustion leak. None of the tests support leak except for the famous "manifold pressure rise on cold engine start". And that one went up so little and then stabilized that I am now discounting it as waterpump/hose stretch residual pressure. That was further supported by the test of taking off plug wires in pairs of adjacent cylinders (no combustion gas now) and seeing if I still got the small pressure rise (we did). And the high rpm crazy pressure bounces observed could only be the pump cavitating because it did not have the thermostat to push against. And while the head gasket I removed is ugly, and clearly has had some failure in the past (perhaps during overheat), it actually tested tight by every other measure (compression, leak down,combustion gas detection, power generation, etc.) And the overheat immediately returned with a new head gasket installed. Usually they take a while to fail even if the block deck is not perfect. So I am kind of back to thinking (as I was 3 months ago) I have a coolant flow issue or blockage, causing a local overheat between #3 and #5, the engine's natural "hot spot" where low flow would cause an overheat. I plan to get the water pump off and look for issues, back flush the block/head, and open up the engine again if needed and bore scope the cooling passages and water jacket, or whatever I need to do to find it. It is that or a complete engine teardown. Any other suggestions?

CaboJohn
 
sounds like a prudent approach.

Regarding your use of "low flow" isn't likely as your hot spot is localized. That would suggest an obstruction limiting coolant contact area in a smal region. Same effect could occur due to aeration od the coolant. You might consider redoing some tests since the head work just to make sure you know what has/hasn't changed.
 
@Mako
Good suggestion about redoing some tests that might have changed since the head work....for better (if I actually accomplished something), or worse (if I screwed it up!). Right now I do not think anything changed (except that the gasket looks better), but one should be sure. In all these long term quests it is the assumptions that bite you.

As for the "low flow" being unlikely, it has more to support it than you might first believe. On a small block chevy #3 and #5 are closer to each other (with a smaller water jacket) than to the end cylinders, and both the exhaust valves (major source of heat) are located in between those cylinders (they are on the ends for #1 and #7). It is a design trade off no doubt done for good reasons, but it does mean the heat load is the greatest right where the cooling is the weakest. So in low flow that is exactly the area that will reach boiling first. And once it boils, cooling is greatly reduced, further heating occurs, and that will be a "hot spot". So the head gasket evidence supports that theory as much as the "local blockage" theory. I would say the jury is still out.

Do you know anything about trying to examine water passageways with a borescope? I presume the old fashion ones (optical with a very small tip) are the way to go.

Thanks for the input. Keep it coming!

CaboJohn
 
CaboJohn said:
@Mako,

But the long and the short is I don't think I have a combustion leak. None of the tests support leak except for the famous "manifold pressure rise on cold engine start". And that one went up so little and then stabilized that I am now discounting it as waterpump/hose stretch residual pressure. That was further supported by the test of taking off plug wires in pairs of adjacent cylinders (no combustion gas now) and seeing if I still got the small pressure rise (we did)... And while the head gasket I removed is ugly, and clearly has had some failure in the past (perhaps during overheat), it actually tested tight by every other measure (compression, leak down,combustion gas detection, power generation, etc.) .... So I am kind of back to thinking (as I was 3 months ago) I have a coolant flow issue or blockage, causing a local overheat between #3 and #5, the engine's natural "hot spot" where low flow would cause an overheat. I plan to get the water pump off and look for issues, back flush the block/head, and open up the engine again if needed and bore scope the cooling passages and water jacket, or whatever I need to do to find it. It is that or a complete engine teardown. Any other suggestions?

CaboJohn
Click to expand...

I never saw the direct evidence for the combustion leak, only the anecdotal evidence. I'm not sure I would have pulled the head with the compression numbers, lack of compressed air leakage into the coolant under static conditions, and engine performance you posted.

However what's done is done and you taken the head gasket out of the equation, which is a good thing. All along I thought this was an engine coolant performing a functional test and inspection of the internal coolant passages is the proper course of action. It may talk some more though and perhaps a little luck but you will uncover the problem.

Best wishes,

Tom
 
@Tom,
Thanks for the input. Unfortunately when I had the head off, while I did look into it and the block, the inspection was cursory. I could "see" the problem was the head gasket, so I did not need to look farther, right? So now that I am back looking for a blockage, I may well have to pull the head again. My "understanding" is that I can pull and inpect the water pump and perhaps look into the hole where the water enters in the block (if I can stand on my head). If I see nothing then I must pull the head. When it is off I understand I will be able to inspect all of the block passageways with a pen light. I am led to believe that I will not be able to completely see all of the head passageways, but I can insure they are clean with a coat hanger.....or I can get a small, optical (1/4" head) borescope. I did not notice any signs of clogging so if something is there we are likely dealing with a foreign object.

If you or any readers have another understanding of how to inspect, or have actually done one, I would appreciate input.

Thanks,
CaboJohn
 
Cabo,

The thing that amazed me about my situation is my boat never over heated.* This was true even when a large amount of the coolant had been replaced by air.* If you remember I had one of the manifolds register 240 and the boat still never had an alarm go off.*** The air pockets do seem to rise to some high point.* The hose which comes out of the fill riser to that manifold is higher than the other side. There is no doubt to me now that there was no blockage going into or out of the manifold that got hot it just collected air there.* Also the heat exchanger appeared to fill with air as well.* All this and the boat never overheated.*
So if I was in your situation I would really question the heat sensor.* Is it possible it is wacked and goes off too easily?* I know the boat runs well when the thermostat is out which to me means there is more of a level rise to temperature as opposed to one area getting hot and then the coolant moves into the heat exchanger.

Just a thought.* I hope you figure things out I sure know your frustration.

Bigpursuit.
 
Back
Top