Cracked block or bad oil ring?

98 XJ. 250k.

When engine is cold, it makes a tonne of white (maybe extremely light blue) smoke/vapour. I really can't tell if it's oil, coolant, or unburnt fuel. Doesn't smell like anything other than normal exhaust. I added more pure coolant to be sure I wasn't just sucking in water and making steam. Still doesn't smell sweet.
Once it warms up, the smoke goes away completely, but comes back under any high vacuum conditions (eg: gearing down and decelerating, then hitting the gas, and everyone behind me says they see blue smoke (though I don't know if I trust their colour perception)).

Once it's warm, I can make it smoke again by disconnecting the spark lead, or fuel injector power to the #6 cylinder. Both will make smoke. Doing this to any other cylinder does not result in smoke. Only affects #6.

So we know #6 is drawing in oil or water. Problem is that I have a small coolant leak near the rad cap (moose did it) and I've always had to add a liter of oil every 4-6 months, but that doesn't seem excessive.

So now the question. Is it a cracked block, or a bad oil ring? If I was sure it was water vapour in the exhaust, I'd say cracked block for sure, but I just don't know. Doesn't smell like oil or coolant, looks white to me but buddy says blue when I'm driving.

Is there any way to figure this out with a block tester? It's my understanding that the block tester will only confirm a cracked block, as combustion gasses have to get into the coolant to test positive. So would a negative test result prove it's a ring? ANyone else think of any other possibilities?


More hints:
* recently failed emissions test. HC and HO were just barely over the limit, while NOx was 5x over.
* LTFT is around +3, suggesting the Jeep thinks it's running lean, and is trying to compensate. This makes sense as I blew 5x over on NOx.
* Roughly 30% loss of power and fuel economy. Went from 23mpg to 17mpg overnight.
* The only two things I did the day before the smoke started (it's smoked after vacuum for years, but never on startup till now) was add a liter of oil before a bit of wheelin', and removed the exhaust and pounded it back together.
* Checked for exhaust restrictions via vacuum. According to results, the exhaust has no restrictions (also exhaust was upgraded to 2.5" high flow).
* Not throwing any codes (other than the occasional TPS voltage, but only if I go near WOT).
* Fuel consumption at idle (cold) is 0.9 gallon per hour. Once warm, drops to about 0.66 gph. My ex's XJ reports about 0.45gph warm.
* Ocassionally get a cylinder misfire for about 30 seconds after start up. This only happens when I really push the motor for a while, then shut it down hot, and wait for about 10 minutes before starting it again.

*update*

Figured if the problem was coolant getting into the cylinder, and it was much worse when cold, that the thermostat must have something to do with it.
Reasoning was such:
Coolant warm, but not hot enough to open thermostat. Coolant pressure must build up until temp is high enough to open the thermostat. High pressure plus vacuum on #6 = coolant really wants to go into cylinder. Then when thermostat opens, the coolant pressure drops immediately, thus explaining why the smoke stops once the engine warms up. So, I figured to test this theory, removing the rad cap would decrease the coolant pressure to zero when the thermostat was open, so high vacuum conditions (like no ignition or fuel to the cylinder) wouldn't be as likely to draw in coolant. At the same time, with no way for coolant pressure to build, if there were exhaust gasses or even raw air/fuel being forced through a crack into the coolant, that'd increase coolant pressure further, so that it'd be more likely that coolant would spray back into the cylinder on vacuum.

Looks confirmed. Started up cold again, blew a tonne of smoke, removed rad cap and let it warm up and kept an eye on the thermostat. The second the thermostat cracked open, the smoke decreased, and when the thermostat was fully open, there was no smoke at all. Then disconnected the spark and/or fuel injector connector to #6 and still no smoke. With zero coolant pressure, I can't get smoke. So at least I know it's not an oil ring. Has to be cracked block or bad head gasket. Also, coolant in the cylinder also explains lean condition, as it'd destroy the O2 sensors.

Waiting for engine to cool down again with rad cap off, to test another startup. I figure with no pressure in the system when warm, by the time the thermostat closes again, pressure will be fairly low, so the next startup should have less smoke and no studdering at startup.

The only thing that I'm not getting, is why this started immediately after removing and reinstalling the exhaust. Coincidence?

Don't overthink this. Start with some basic testing and a compression test would be job #1.

Let testing guide the way for you here; a cylinder leakdown test is generally more helpful at determining "root cause" if the compression test tosses up a flag.

The compression spec for the 4.0 is 120-150 psi, with no more than a 30 psi variation between cylinders.

A simple compression test can tell you a whole heck of a lot about the internal condition of your engine. And it takes all of a half an hour. Good luck.

Whoops. Forgot to add that to the list of stuff I've already done.
Compression test is 160 +/- 4psi across the board. All good.

Anyhow, I just rigged up a better leakdown tester to confirm my suspicion about the #6 leaking into coolant.

Let the engine warm up to make sure the thermostat was wide open, then stuck an air hose in #6, gave it 10psi and removed the rad cap. Sure enough, water level rose, rad overflowed, and the slight pressure in #6 will continue pushing coolant out all day. Tested 1-5 as well, but no leakdown with them at all, and certainly not blowing into coolant.

So, now have to decide if it's a head gasket or block. Given I recall worrying about the coolant when I first got this Jeep, but I've since replaced the HG and head (brand new head), I'm goona go out on a limb and assume the block is cracked. Hopefully my line on a new longblock comes through this weekend

That said, I've put about 75k on this motor, with what seems to be a fairly sizeable crack in the block, and didn't slow me down enough to really notice

If the block is cracked it will leak into the crank case when #6 is at TDC and the cooling system is pressurized. I'd say head gasket or problem with the deck.

Talyn said:

If the block is cracked it will leak into the crank case when #6 is at TDC and the cooling system is pressurized. I'd say head gasket or problem with the deck.

Click to expand...

That's what I suspected originally, though I'm still a bit suspicious as the new HG and head only have a few 10's k on them, and I had the same issue before swapping HG and head.

I haven't been into the 4.0L block enough times to know the oil/coolant layout, but based on what I remember, I can't think of any reason to assume it'd be the HG instead of the block. Any reasoning behind your answer?

Thing is, I can usually grab a complete longblock for around $200. Given I'm in Canada, a new HG will cost me $150 locally, or $50 from the US, plus half a days work to swap. I figure if I'm not near positive its the HG, it'd just make more sense to assume the block and grab a new motor. Idea's?

I think you have a head gasket leak due to the deck needing milled. The issue is between #5 and #6 cylinders. 5 is sealing but 6 isn't

blistovmhz said:

I haven't been into the 4.0L block enough times to know the oil/coolant layout, but based on what I remember, I can't think of any reason to assume it'd be the HG instead of the block. Any reasoning behind your answer?

Click to expand...

Umm.. As I said, when the piston is at TDC most of the cylinder will be below the piston and will leak into the crack case. If the block was cracked there would be coolant in the oil. Cracked blocks also aren't common.

Sounds about right.
Trying to decide between HG or just waiting for another motor.

Would a small hole in the HG leading into the coolant cause a 10-20% loss of power?
Also, would coolant in the cat, cause the cat to fail fairly quickly, causing extremely high NOx, or is the NOx being caused by something else? My understanding is that NOx is usually a symptom of lean condition, but I shouldn't be running lean unless my O2's aren't reporting properly. Trying to figure out the cheapest (long term) fix for this. Don't wanna spend all day doing a HG and find out I'm goona need a new block

Yes, coolant in the cylinder could prevent that cylinder from firing. As well as pumping unburnt gas into the cat, which they don't like. Too much coolant can also plug a cat.

You are thinking of how an O2 sensor works incorrectly. They don't sense fuel, but rather extra oxygen. A cylinder that isn't firing 100% will have excessive fuel and excessive air, so the O2 sensor could be seeing a lean condition.

There could be a problem with the block surface preventing the heat gasket from sealing properly.

The +3% fuel trim is nothing to be concerned about. Its with in reason. I also doubt you were getting 23 mpg.

I WAS getting 23mpg constantly, city or highway. Measured both via OBD as well as at the pump, and checked GPS speed against VSS as well as distance travelled by map . I've always got about 22-24mpg in all my Jeeps.
That said, when I started blowing smoke during cold start (new), my fuel economy dropped significantly until it's fully warm. Usually drives around 17-18mpg highway for about 15 minutes, then comes back up to 22mpg.

The lean condition I'm still torn on. I don't really care what my sensors are reporting at this point, as I have to assume the O2 is wrong if I've been burning coolant. Fact is though, that my emissions test showed slightly high HC and HO, and extremely high NOx, which afaik means I have a lean condition. I guess unburnt fuel would increase HC and HO, but would the HG leak cause the entire engine to run lean?

You have one upstream O2 sensor, so one cylinder having a problem will effect the whole engine.

are you doing the math right from liters to gallons dude?

lol

eflores said:

are you doing the math right from liters to gallons dude?

lol

Click to expand...

Heh. Yup. Positive. Even my 90 xj got 23mpg. I typically get 600km before I fill up, and I must have some fuel left cause I've never put more than 65L in a tank (XJ is 75L iirc). No idea why. I even drive my buddies Jeep at 20mpg, while he's never managed to get more than 300km to a tank Also drove my roomies S1500 a few weeks ago and pushed it to 18mpg. I think he gets 16 at best.

I would jump on that longblock, stick it in, pull the old motor apart and figure out what's wrong with it. Then rebuild/reassemble it (assuming the block is in good shape) and keep as a known good spare.

It's been 14 years since the last non-0331 4.0 that fits our jeeps was made, and the junkyard donor pool shrinks every day. Cheap motors won't be around forever. My local yard used to have a dozen or more XJs at all times, these days I'm lucky to see 6 there and most of them have been there for weeks and are pretty picked over.

kastein said:

I would jump on that longblock, stick it in, pull the old motor apart and figure out what's wrong with it. Then rebuild/reassemble it (assuming the block is in good shape) and keep as a known good spare.

It's been 14 years since the last non-0331 4.0 that fits our jeeps was made, and the junkyard donor pool shrinks every day. Cheap motors won't be around forever. My local yard used to have a dozen or more XJs at all times, these days I'm lucky to see 6 there and most of them have been there for weeks and are pretty picked over.

Click to expand...

Yea, is what I'm debating right now. I PROBABLY just need a HG, and that's a simple job. I just didn't wanna put more time/money into the block until after I'd torn it down.
I'm thinking about that other long block right now. First one fell through, but a second one popped up for $300. Problem is that it was removed from the donor 6 months ago, and the owner took the oil out for some reason, 3 weeks ago. Says it ran perfect and just passed emissions, but now I'll have to fill it up with oil and stick a filter on, before I can even compression test it. Don't really wanna spend $60 to compression test someone else's motor and find out it's got blown out rings.
Just sorta arguing with myself here at this point, till I decide to just throw down $300 on an engine in unknown condition.

If the long block is the right price and you intend to stick with a vehicle with the 4.0L engine an extra engine would never hurt. Heads are easy to find, good blocks not so much.

Talyn said:

If the long block is the right price and you intend to stick with a vehicle with the 4.0L engine an extra engine would never hurt. Heads are easy to find, good blocks not so much.

Click to expand...

That's the thing. I love the 4.0 for getting me home safely, but I hate working on it. No engine should take half a day to change a head gasket. I've wanted to do an LS swap for a while, but apparently have to wait until 2015 because the LS will never pass emissions (even if it blows well under what the 4.0L would have... poorly conceived laws here in BC).

If I was going to stick with the i6, I'd go grab another LB, swap it in, and use the other for a stroker, but just not interested in doing another i6.


I'm just annoyed at this point that all my options are "spend money for unknown result". I can't find a running 4.0L that I can actually test run. Everything is "longblock, no oil, in pieces, unknown condition" or I just replace my HG and maybe that fixes it... or maybe then I find out I've got a warped head or block, or bad ring, or bad cam... and it goes from being a $30 fix to $1000 and I'm still not driving anywhere All this to pass emissions tests that will be outmoded next year.

Are you kidding me? The 4.0L is one of the easiest engines to do anything on. You think an LS swap would be easier to work on? Not to mention the actual swap part. Don't go all BimmerJeep on us.

Talyn said:

Are you kidding me? The 4.0L is one of the easiest engines to do anything on. You think an LS swap would be easier to work on? Not to mention the actual swap part. Don't go all BimmerJeep on us.

Click to expand...

Hahahha. It's not that changing a HG SHOULD be a lot of work, it's that the engine is slightly too far back in the engine bay, and the unibody is in the way of the rearmost head bolt, and you have to do a lot of shopping to find a torque wrench that'll fit in there. (I've actually never found one, which may be why my #6 is leaking).

Took me almost a full day to change my first 4.0L head, and I can do it in about 5 hours now. By comparison,my first v8 head swap took me 3 hours, when I was 14yo and I've never seen a torque wrench that wouldn't do the job. Hell, I did my Civic head in under 15 minutes (best time) from engine hot, to head and HG swapped, to test drive.

I REALLY like that the 4.0L had such great fuel economy and power (for it's day, it was a killer), and it's reliability is incredible. But always seems that when I DO have a problem, it turns into a ****ing $1500 goose chase