Thump…..thump. Thump..thump. Thumpthump. My heart began racing as I put all of my might into lifting the cast iron cylinder head from the engine that had “blown up” a few weeks prior, just 20 minutes after I’d sold its host — a beautiful 1991 Jeep Cherokee — to a nice lady. I had sent an oil sample to a lab to try to understand what had happened, but the results were inconclusive. But now I was going to find out for sure. Here’s what I saw inside that dead motor.
Last night, my colleague Jason Torchinsky and I spent about three hours tearing apart the 170,000 mile Jeep 4.0-liter inline-six in the 1991 Jeep Cherokee that I had sold to a nice lady who had driven all the way from Chicago for a Jeep that would only take her 20 miles towards home before overheating. I was initially worried that the “unkillable” Jeep engine was toast, since it had such a severe knock, but then I learned that the noise was just the water pump banging against the engine block.
Still, the engine didn’t run properly, and had no compression in the middle cylinders. So I sent an oil sample out to Blackstone Labs, who found elevated aluminum and iron content, indicating piston and piston ring/camshaft wear. The lab had found no indication of coolant in the oil, which was odd, as multiple adjacent cylinders without compression is almost always a symptom of a failed head gasket (which tends to cause coolant to leak into the oil). Maybe the head gasket wasn’t the issue, and those center pistons had somehow failed? This wouldn’t be the first time I’d heard of a Jeep 4.0 piston breaking to pieces.
I’ve lost a lot of sleep over this, since I really don’t feel like swapping an entire motor, especially not by the June 1 deadline that the buyer gave me (Yes, she still wants the Jeep, and she needs it soon). So to improve my nightly rest, Torch and I finally yanked the head off the motor yesterday. It’s a job I’ve done far too many times, as AMC’s most legendary motor is— if I’m honest — deeply flawed when used in an XJ, a vehicle with a marginal cooling system.
View this post on Instagram
This head extraction was easier than ever, since I had an extra set of hands and great weather on my side.
We removed the electric fan and airbox, undid the exhaust and intake manifold, took off the power steering pump, set aside the fuel injector wiring harness, popped off the spark plug wires, unbolted the thermostat and water pump (these weren’t strictly necessary to get the head off, but I’m replacing both), lifted the valve cover, and then loosened all the cylinder head bolts. From there, I slid a breaker bar into the thermostat hole in the head, and lifted up. After quite a bit of pushing, the cylinder head finally popped up, revealing my engine’s six cylinders:
Here’s a look at the engine just before head liftoff, but after we’d removed the intake and exhaust manifold (which you can see there along with the dangling injector harness), as well as the power steering pump, valve cover, thermostat, and airbox:
Here you can see the valve cover and intake manifold:
Once we had the head off, we faced the moment of truth: Would there be catastrophic damage, or was this truly just a head gasket failure?
By the grace of the car gods, the pistons all looked to be in good shape, though the bridges between cylinders two and three, three and four, and four and five were all covered in coolant. You’ll also notice that there’s coolant in cylinder five, sitting atop the piston:
With coolant getting into the cylinders, and oil getting quite hot, you may wonder if there’s any damage to the cylinder walls. Let’s have a look:
Cylinder six does show a bit of vertical scoring on the cylinder wall, but touching it with my fingernail reveals that it’s quite shallow (undetectable, really). Having opened up a 4.0-liter straight six before, this isn’t that abnormal, and I don’t think it’s something to worry about. The remaining visible cylinder walls all looked nice:
Let’s have a peak at the failed head gasket:
It’s hard to tell how much of the damage occurred during cylinder head extraction, but you can see some bent fire rings (that’s the silver ring around the cylinder bore hole). The image below shows how a new gasket should look. Again, it’s hard to see exactly where the failure occurred, but it clearly did, as coolant was able to pour into the cylinders and cover the bridges between the cylinders (and this didn’t happen during head-removal, as the head was pretty much completely devoid of coolant). I wouldn’t be surprised if the head is warped, allowing coolant to travel between it and the gasket.
Anyway, the cylinder head now sits in the back of my Jeep. Tomorrow I’ll dump it off at the machine shop, where a 350 pound dude will check the thing for cracks, find them, and then I’ll be removing the head from the spare engine I have in my garage. That head will be checked and then shaved flat. Then Torch and I will reinstall it and pray that this Jeep runs like a dream so I can finally sell it, and bring my fleet down to a reasonable number of cars. I’m not sure how many exactly, which is kind of an indication that maybe it’s still too many, but let’s just ignore that.
Just look at your hands young men! I started wearing Nitrile gloves when do all my car work. Makes life much easier…and I can just take pair off and go in for sandwich or use my cellphone without any mess.
For example: 400 count at Costco is $40 ! but that only works out to $.10 for each glove. Worth it!
https://www.costco.com/kirkland-signature-nitrile-exam-gloves%2C-400-count.product.100497804.html
The 9 Mil gloves @ Harbor Freight are my go to:
https://www.youtube.com/watch?v=cAMLa5ZC-B4
I see it has a (Dorman?) aftermarket exhaust manifold. The OEM ones always crack…always. My theory is the often-failing motor mounts. David, please please use those motor mounts I sent you to replace the likely-perished ones on that XJ. It’s such an easy job and most people ignore it, which I truly believe causes issues due to motor movement under the hood. I also believe that can contribute to head problems with the non-TUPY heads on later models. ‘Cause nothing good happens from an engine flopping around.
The videos really added to the explanation. Thanks very much! This was really interesting!
“and bring my fleet down to a reasonable number of cars. I’m not sure how many exactly, which is kind of an indication that maybe it’s still too many, but let’s just ignore that.”
LOL! Yep, there it is 🙂
Glad it was relatively good news upon teardown!
THE SAGA CONTINUES!!!
Son of a Monkey? Jimminy Cricket? Gee Whiz? Definitely keeping this family friendly I see
DT and JT need to make an official autopian t-shirt with a picture of a random oddball tail light and the random quote “son of a monkey” underneath it and sell it on the merch shop. i’d buy it instantly.
I don’t think the words “a beautiful 1991 Jeep Cherokee” should be next to each other.
Glad it’s getting fixed, hope it goes well from here on out. Echoing other commenters, you should do a writeup about how to improve the XJ cooling system. I remember buddy’s having repeated problems with theirs back in the day. It seemed we were always doing a thermostat, water pump, radiator, or getting balls-deep into it like you are now.
So, a somewhat off-topic question for our Editor-in-Chief: Have you ever thought about doing a stroker engine for one of your Jeeps? They seem to be pretty popular in the Jeep community, and should be a walk in the park for a man of your skills.
I know you love that inline 6, but could you love it more with more power? It might make for a fun project (cough, cough, CONTENT, cough, cough).
Glad to see it is on the mend.
Does the person who bought it before have any interest in it? Just curious if she wants it or has washed her hands of it.
“I’ve lost a lot of sleep over this, since I really don’t feel like swapping an entire motor, especially not by the June 1 deadline that the buyer gave me (Yes, she still wants the Jeep, and she needs it soon).”
This was right below a picture of Tracy and Torch mugging for a selfie with greasy hands, so likely you were reading in a trance at that point.
Most likely. I sometime skim to the details. 🙂
As a cooling systems engineer, what would improve XJ cooling? Does it need a larger radiator?
I have wired more than 1 XJ electric fan to a toggle switch because the thermostat controlled switch failed. This works best for people who understand the importance of watching temperature.
I put mine on a toggle and added a shutdown timer so I can run it for a couple minutes after I turn the Jeep off. That plus the hood vents really helps the heat soak.
As someone who fixed so, so, so many of these, I have a very differing opinion from David.
One, the worst cooling issues absolutely cannot be fixed. Because they’re inherent to the block. The 4.2/4.0 is a wet block mounted longitudinally. That means no matter what you do, as you move toward the rear of the engine, you are running hotter. Simply because the water pump and fan are at the front. That’s not touching on any other design issues (I’m sure they’re there, I just don’t know them.)
Two, the second largest problem is that the system is extremely prone to sludging, especially when they switched to HOAT. I :heart: Mopar 6977AB because of the 4.0’s. That sludging means a very poor coolant-metal interface and reduced flow through jackets. Many times, a slightly hot running 4.0 can be fixed by doing nothing more than a very thorough chemical flush. You can see on David’s pictures a mix of white corrosion (minor issue) and grayish goop along the inside. That’s Coolange(TM) Sluj(R); a melange of iron oxide, water, glycol, oil, rubber, and plastics. Awful, radiator killing shit.
Now, on the heat soak problems, I’m going to refer you to item #1. Once you’re shut down, fans or no fans, you cannot effectively remove heat from the rear of the engine. That’s why #6 is always scored. In order to fix the heat soak, you need to remove heat from the back of the engine. The water pump does a tolerable but not great job of circulating – when it’s running. So fans are largely placebo, because you’re not actually moving the heat.
If I was going to attempt to solve, my solution would be frankly, extremely drastic and wildly different from anything David came up with. Because I wouldn’t touch the radiator or fans at all. I’d yank the water pump from the front, the #5 (rearmost) freeze plug permanently, tap that freeze plug, and switch to a rear-to-front flow. Remember, the back of the block will always be hottest because it’s furthest from airflow and closest to exhaust. So that’s where I’d want the coldest coolant with the highest exchange rate.
It’s also a very clear illustration as to why I think it’s unsolvable. Sure, we’re gonna mount our water pump right where our intake and exhaust go. That’s definitely gonna work! Well, okay we’ll just put a 90 on it and… oh… we still have nowhere to put the pump because that’s where the power steering goes and we don’t have room. Well we could put it below, but then how are we getting water out? Oh, and how are we keeping the hot exhaust and catalytic converter from melting the water inlet?
I defer to your experience with this motor, however there’s a flaw in your logic. Not many straight six engines are prone to blowing head gaskets. In fact, the opposite is true, L6 engines are fairly tolerant of overheating.
Said another way, if the problem was caused by inline longitudinal engines with mechanical water pumps being prone to overheating, why dont they all blow? Slant 6. Cummins. Every semi truck.
I only ever owned one 4.0. It came to me with a blown head gasket, and I sold it with a blown head gasket.
What would happen if the 5 plug were tapped, run a water line to an electric thermo-switched pump, then continue to a Y fitting on one of the front hoses?
It needs a larger radiator. DT posted a video somewhere on social media showing the tiny radiator in the XJ vs the large one in the ZJ. He said the XJ radiator was designed for an I4 engine and the ZJ for a V8.
While I might not want to live on the same street as you DT (property values and all) I wouldn’t mind being a short drive away to come lend a wrench to such exploits.
He’s actually on a main road, so it’s not actually as much of an issue as you’d think. I’d wager the sheer noise from the four lanes (plus center turn lane) of rather constant traffic (which opens up into six lanes with boulevards just past his place) detracts more from property values than too many vehicles placed strategically around his property.
@MP81
“vehicles placed strategically around his property.” don’t you mean — vehicles placed haphazardly around his property.
Uh, something ain’t right here.
The head is definitely warped – head gasket didn’t fail. They warp at 3-4 because of clamping force issues. #6 scoring is a total non-issue. It’s the hottest cylinder (furthest from all cooling) and it always has scoring. As long as it passes fingernail and has hatch, it’s completely fine. The blocks were built to deal with it.
Problem is, based on your pictures, this thing was DUMPING coolant into #5 and #6. See how they magically have no carbon buildup? That’s where all the coolant went. Into those cylinders where it steam cleaned off the crud. Where you had compression.
That’s bad. Real bad.
Any time you’ve got a 4.2 or 4.0 with this kind of evidence, the head is no longer a mill job. Valvetrain is non-adjustable, and popping cylinders like that alone is enough push it out. And it’s a tight spec – 040 to 060 preload, spread preferred to be under 010.
Which brings us to the big problem: it was absolutely pouring the coolant into 5 and 6, and there’s no obvious indication as to why. Coolant had to come from somewhere, right? Yep. It’s the other fracture I hate on these. The one even good shops often miss because it’s only visible on x-ray, even when you pull the water plugs, and sometimes sneaks by pressure test. And it’s why you don’t have milkshake.
See the coolant plug at #5? Great. Go toward #6, now you see the casting mark between #5 and #6? That’s exactly where I’d expect to find the blowout. Inside the head. Pull the #5 water plug (carefully!!) with the head set up so that you can see the underside as well. (Yeah, it’s a little tricky.) It’s definitely best if you can block up the water jackets and let it sit a while. You’ll be looking for any water at all on the #5 and #6 valves. Exhaust or intake. Any trace of water on the valves, the head is a core. Don’t waste time welding.
“But I have no milkshake, this makes no sense!” Ah, but it does. The 4.0 runs hot, especially in back. Those cylinders got a steam cleaning, and because the block was extra hot due to a (likely coincidental) water pump failure, the pistons were sealing up very well. So the coolant was held in chamber well, and flashed to steam, giving you nice shiny pistons. And there’s absolutely no contact with oil except within the bore, where it’s being flashed to steam and out the exhaust.
Prep the other head now, come back to this one later. I suspect you can get an article on X-ray NDT of heads out of it.
Agreed. Most of my head failures have NOT had milkshake. BUT they have had trace amounts of coolant in the oil. Granted, they’d been driven more than 25 miles… So this I guess isn’t a surprise. I might have expected SOME coolant in the oil as was the case in all of my other head-crack failures, but fair enough.
I’ve got two heads ready to rock.
Yeah, it’s definitely an atypical failure on these where you do not have milkshake. Especially with warp. I’m chalking that up to the just 25 miles of driving though, since usually I got the ones going 25,000 between oil changes.
The more I think about it though, the more I’m suspecting the head was already fractured at #5-#6 and the water pump was incidental. (So hooray for that timing.) Those cylinders are not 25 miles clean, they’re 250+ miles clean. There’s also some evidence of steam in #2 so it might have two cracks. Plus the warping. Still, it was caught before hydrolock or thrown rod. So overall, the bottom end is better than fine. I mean hell, two clean cylinders. May as well grab a breaker bar and some brill-o while waiting and do the other four. 😉 (Joking. Get some Orison Piston Kleen though. Follow directions for mixing, lower pistons from TDC, let it soak, get as much out as you can after 24 hours, and add some oil to bores before rotating.)
There’s a shitload of gunk in the coolant channels though, and unevenly distributed as usual. I would absolutely do two flushes (disconnect the heater core) with Mopar 4856977AB – accept absolutely no substitutes, it’s literally the only one I use for all makes and models. Should be about $12-15 at any dealer. Because it works, and it works well. Plus at least one tap water flush after. Depends how bad the Jeep(R)-brand Coolange(TM) Sluj(R) coming out is. Very common problem on these, but if left unaddressed, can cause premature water pump failure and worse. Especially any jackets are fully clogged – believe me, it happens.
DT – you need to get this guy to write articles! This is the deep dive I come here for… and it’s in the comments!
100% agree- this is column-worthy stuff right here.
Right? I remember reading a dissertation-long comment at the old site talking about the nature of Eagle Hemi heads and found it very interesting. Hire this guy as a guest writer, I’d read whatever he puts out.
Thanks! I do try, but on these, I very much have the advantage of being a former Chrysler (and Dumber-Chrysler) certified mechanic who typically saw more Jeeps per day than David owns. Usually in equally atrocious shape. ;P Plus I’ve kept at cars on the side since I could drive.
But the dealer suffering is why I’m extra knowledgeable on these, and know where to find the things I don’t remember. Like the preloads – I didn’t remember the numbers off the top of my head. Soon as I saw them I was like ‘oh, yeah, within 010 spread and I like 005 but that sucks to do.’ (It really does. Unless you’re building a race motor, 010’s fine. Over 010 and you’ll have fuel mileage problems minimum, often issues getting into closed loop too.)
What does bug me about this one is that there’s evidence of coolant in the cylinders everywhere but where the head warped, and it’s long term. That tiny bit of rust there (doesn’t affect operation) is proof that water sat on the piston. And it’s on the high, not the low. But I’m going with the assumption that it’s from this failure and it’s head seep, since the car did sit for some time after. So the head might not mill at all.
So the question then is: where’d the coolant come from and why didn’t it produce noticeable white smoke? We’re burning coolant with nothing worse than cold winter start puffs observed, unless David’s also blind. And I doubt that. This also definitely isn’t a porous casting fault. Those fail sooner, and guarantee milkshake because they dump coolant straight into oil. But something has been causing coolant intrusion into multiple cylinders over a long period.
I mean, it’s better than my first concern, which was a fractured bore. Those very much do happen.
So where’s the damn coolant coming from!? Without inspecting the engine in person (sorry David, I can’t make it up tonight,) it’s all EWAGs (Educated Wild-Ass-Guesses) based on experience and knowledge. Like the head bowing. But the conditions here aren’t something I’ve seen before. #5 and #6 there, absolutely. But 1, 2, 5, 6? With no white smoke? Nope. I just have not seen that. Only thing I can guess is that the head has multiple failure points and has been silently seeping coolant into the bores for a very, very, very long time. Probably only while the system’s under pressure. Not enough to affect operation or drain coolant noticeably, but enough that over a very long period, it produced nice shiny pistons. We’re talking hundreds, possibly thousands of miles of little white puffs just slowly, methodically eating up the carbon. If you catch it very early, you’ll have ‘spotty’ piston cleaning. Nothing like these – I honestly was going to ask David if he replaced pistons in this car before selling it. They’re that clean.
But again: that’s 100% a guess.
This is awesome, and I’ve learned something new today. I was wondering why they looked so clean and shiny. Articles and the added insight like this are why this site is so great!
I’m stealing the phrases “Dumber-Chrysler” and “EWAGs”.
That’s less a phrase, more a statement of fact. Merge of equals my ass.
But really, come on guys, Jeep(R)-brand Coolange(TM) Sluj(R) gets no love?! See if I agree to fix your clown – ahem, cloud car next time it breaks. I mean after it runs for more than 30 minutes.
What? Yes, burning it down for the insurance money is absolutely an approved fix! It’s covered in TSB 03-002-51 for resolving customer complaints after replacing every computer module as a non-warranty item and having only made things worse. Not to be confused with TSB 03-002-50, negotiating reality with customers who insist their Sebring is a collectible because it’s a convertible; or TSB 03-002-52, ignoring customer AutoStick complaints without being murdered.
Very interesting
That’s exactly what I was going to say. You beat me to it. Anyway, don’t forget to check the pfetzer valve and have some extra 50 weight ball bearings available.
rootwyrm, you should be writing tech articles for The Autopian!
Where’s the YT video? You two could make a series of these videos. Call it American Mighty Car Mods.
I’d go with ‘Mighty Car Mooks’.
Yes, and nobody need worry, when they’re done fixing all the shitboxes between them, they can start fixing up winners from Shitbox Showdowns, to be sold or given away as charity.
Yay teamwork!
Yes, yea teamwork! It’s no fun trying to take off and replace that 65 pound cylinder head by yourself.
I do trim some of the excess threads off the top of bolt 14 so I can insert the bolt after the head is in place – I’ve never had any luck holding it up with tape.
Yep, I love an Inline 6, but maneuvering cylinder heads, intakes etc is a struggle. Long and heavy, no sexual pun intended.
With my current pickup, it’s set so far down in the gigantic engine bay meant to accommodate 454s, it’s really a challenge at times.
Try a redneck lifting rig. Ford 5.4 Triton heads are very big and surprisingly heavy for aluminum so I put a 2×4 across the fenders and used a ratchet strap to raise and lower the heads. In addition to being awkward they have to be lowered onto the locating dowels dead straight or you get head gasket problems.
Anytime buddys and (presumably) beer are involved, a good time is sure to be had!
Not bad 🙂
As i recall you predicted this outcome?
It got a bit confusing at the end.You’re planning to use another head ,but are getting the original
crack tested out of curiosity?
I read it as a pessimistic, “if it fails and cracks are found (and they will be) then I’ll be forced to use the other head, which will have to be shaved prior to using it in the XJ being sold”.
Id say it’s worth a few bucks to at least see if this head is salvageable. That beats tearing your spare engine apart
remember, the pack rat throws nothing out