2011 Mazda6 3.7 Headgasket replacement at ~69k miles

So, my wife's car died earlier today-she said it made a noise, and just stopped running in traffic.. I went to check it out, and got it started. It was running rough, but not terribly, so I moved it to a side road and noticed it was very low on coolant.

I went home, grabbed some I had in the garage, filled it up and drove the 1/2 mile home. The oil on the dipstick is still mostly oil colored, but it's clear to me from a bit of goop on the top of the stick and on the cap that it's probably a head gasket issue.

I found a ton of things to replace the 4 Cylinder head gasket, and not much for the 3.7 for Mazda when searching for that.

It SHOULD be the same process as for the Ford Fusion and the Mustang, right?

At that millage I would be asking Mazda for a little help. They may say no but it is worth asking nicely and see what they say. 

I'd be more inclined to think it is a water pump, which will puke into the oil if/when the seal fails.  Coolant mixing in oil is rather unlikely with this type of engine without the kind of head gasket failure that is easily seen with a compression test.  That is the good news.

The bad news, is that water pump R&R still requires lifting the car off of the engine since there is no way to remove the timing cover in chassis.

(but hey, the Internet says timing chains are better than belts, and the internet is never wrong)

Just make sure that the head gasket is actually the culprit and not condensation and oil mixing on the dipstick rather than coolant.l and oil.

Being low on coolant is certainly a clue but I’d hate to see you pull a head for nothing.

Im assuming there was no CEL...

In reply to ebonyandivory :

Plug in a scan tool and look for a large discrepancy between the fuel trims on bank 1 and bank 2.  Add up the long term and short term to get the total combined fuel trim, for each bank.

If it is something ridiculous like -20 combined on one bank and +20 on the other, something way out of whack like that where the signs are reversed, then the water pump is failing and causing the right (firewall side) bank to have massively retarded cam timing.  The cause of these readings is one bank is running a lot less efficiently than the other bank VE-wise, but there is only one mass airflow sensor, so the bank that is in-time is going to be getting most of the air, so it will be lean and fuel trims will be way positive.  The bank out of time is going to be not getting very much air, so it will be rich and fuel trims will be way negative.

It may be the water pump, there was a bit of a rattley noise. I didn't leave it running for long, for obvious reasons. Once it is daylight tomorrow, I will take a better look at it.

Sounds like this one might get farmed out. I have no desire to drop the engine.

In reply to Brett_Murphy :

A bad water pump might be a better outcome than a head gasket. Especially with so few miles

 then the water pump is failing and causing the right (firewall side) bank to have massively retarded cam timing

Are you suggesting that something like a failing shaft bearing in the pump is overpowering the oil pressure in a variable cam gear?  

Cheese on the dip stick may be bad venting? Does it smell like coolant?

A bit more research on the issue shows that as Knurled said, when the water pump bushing goes bad you get oil/coolant mixing and a rough running car. I'm betting that is it. Additionally, as he said, there seems to be about zero space to change the pump, so off to the outsourced garage it will go tomorrow.

Vigo said:

 then the water pump is failing and causing the right (firewall side) bank to have massively retarded cam timing

Are you suggesting that something like a failing shaft bearing in the pump is overpowering the oil pressure in a variable cam gear?  

I'm saying that when the water pump bearing fails, in a single belt/chain setup that uses the pump as the pulley in the V, the distance between the left and right bank gets significantly shorter.  This allows the chain tensioner to max out as it takes up the slack and it is generally the rear bank that jumps a tooth.  Even if it doesn't jump a tooth, the cams in the rear bank will be running signficantly more retarded.

I don't know how much authority the Ford cam gears have, but GMs with the 3.6 will run out of motion and get loss-of-control faults on the right bank when the chains are merely worn.  And this being with a multiple chain setup with separate chains for each bank coming from a primary chain.

All that said, if the engine was "very low" on coolant, and it went into the oil, the oil would be chocolate flavored mayonnaise, unless it was leaking out of the rear weep hole for a long time and only just now failed hard.  Or maybe it was just low for other reasons and the engine locked up/shut off because it overheated.

I'm saying that when the water pump bearing fails, in a single belt/chain setup that uses the pump as the pulley in the V, the distance between the left and right bank gets significantly shorter.  This allows the chain tensioner to max out as it takes up the slack and it is generally the rear bank that jumps a tooth.  Even if it doesn't jump a tooth, the cams in the rear bank will be running signficantly more retarded.

Ive seen water pumps angled enough for the impeller to start contacting the housing and making terrible noises, but i've never seen one cause timing to jump. Big cam timing variance between two heads is an interesting way to go looking for a bad water pump, but it seems to me the more useful side of this concept is if someone goes into the timing cover looking for the source of that symptom and DOESNT catch a loose water pump.. well, it'd be sad, but i wouldn't feel too bad for them. 

In reply to Vigo :

I have personally seen water pump failure cause the timing chain to jump on several Chrysler 2.7s.  Belt wise, I have seen water pump failure cause the belt to go off time on a Chrysler 4.0.  (Poor drivability, loss of coolant, and horrible noises as the pulley attempted to chew through the metal timing cover)  I also repaired an Audi 3.0 (the infamous AVK engine with the goofy cam drive) that jumped time and wore halfway through the timing belt when its water pump failed, leading to replacement of many valves, and didn't repair an Audi 2.8 that actually broke the belt.  Customer opted to scrap that car instead of repairing.

Big fuel trim variation between banks is not necessarily a test for a bad water pump, but it is a good test for a timing chain/belt problem.  (Have seen a lot of Hyundai 2.7s jump time on the rear bank, as well, easy to spot with a scantool without removing belt covers)  If the drivability problem coincided with a rapid loss of coolant, that's a clue as to the why there was a timing issue.

So, nobody will touch this car. My buddy won't do it at his garage because it's a pain in the butt and he hates this engine. The Mazda dealership won't do the water pump because it failed into the oil and they have a nice, remanufactured short block they want to sell me. No, they didn't test the oil to see if the bearings got washed.

So, I'm going to have to do this myself.

Can a mod change the title of the thread to Mazda/Ford 3.7 water pump replacement? I want to photo-document this ordeal. It will probably take me a month of Sundays.
 

Man, i wish we were all that rich that we could turn away work just because we don't feel like doing it.

BTW - Right now I have a 3.6 CTS apart.  Right side timing jumped six teeth due to lack of oil.  I'm thinking it was iterative.  While installing the right side (most outboard) chain, noticed that the exhaust cam bolt was finger tight.  Removed bolt to inspect cam/gear interface.  Found this:

That dowel is supposed to be part of the cam sprocket assembly.

I am thinking that the rattling around with a flappy chain made the bolt come loose, and then the world ended.

So, new cam and phaser later, discovered that the #3 exhaust valves are bent.  1 and 5 hold pressure but 3 blows right out the exhaust.  Off comes the head.  Find that all six exhausts are bent, but 1 and 3 were bent juuust enough that spring pressure would seal them.  Also find that the 3.6 has intake pockets on both sides, and the exhausts hit on the corner of the pocket rather than the tip.  Weird.

But, hey, it's work.  And after I pulled the head off, I did a heater core in a Silverado, that had half a Best Buy's worth of poorly installed stereo and alarm equipment that I had to navigate.

And THAT calls for less time than a water pump in a 3.7 anything....

Also, as much as certain aspects of the engine suck service wise, I grow to appreciate how good the design is in general, and am thinking of a 3.6 swap for one of my RX-7s.

This is known in the psychiatric world as Stockholm Syndrome.

Oh, you're just filling me with confidence, here, Knurled. 

In reply to Knurled. :

Is that a customer's car with the dashboard sitting right on the seat with no protection on it?! For shame! Get you some clean towels at least.

@Brett: Sorry to hear about this. I have a Taurus X that I believe has the same weakness. Ugh. Hope it doesn't need service under my ownership.

In reply to dculberson :

It's okay, the dashboard didn't get dirty. 

One of "those" kinda deals.