Silber’s 2002 Boxster S Engine Build
 

All,

Figured I would start a thread to document my project.

Was autocrossing, sudden loss of power. Still ran, but very poorly. Flatbed home. Drained the oil at home, most of the coolant came out the oil drain hole too. Drained the coolant, only coolant came out, so my water hoses are still good.

So, decision time, used motor or rebuild. Can’t find any good low mileage used motors below $5-6,000. Still don’t know exactly what you are getting. Decided to do a hybrid rebuild, steel parts from my engine all “new” low mileage alloy parts. Since I probably have head and/or cylinder damage causing intermix I have bought the following from reputable eBay sellers:

Pair of heads, 29,000 miles, undamaged donor engine.
Set of pistons 37,000 miles, undamaged donor engine.
Engine case, replacement part number, out of 64,000 mile car, assuming lower mileage, no scoring or cylinder damage whatsoever.

New engine parts include but not limited to:
All bolts
Piston Rings
Oil pump
Head gaskets
Rod/main bearings
All new rails
All new injectors
New chain tensioners

Basically building a frankenmotor from low mileage alloy parts, and 116,000 mile steel parts That’s my way of getting a low mileage motor where I know what’s in it. Total cost for all engine parts was right at $5K but I think I’ll end up better off than spending that on a used motor.

I’ve built a few air cooled vw boxers so have some experience going into this, looking forward to getting my car back on the road and hopefully back to autocross.

Where things sit now:

The transaxle is out, engine should come out tomorrow, engine tear down will start after that. I am going to take my time and enjoy the process and make sure I get it right. This thread may not have terribly regular updates, or may have daily updates, just depends on how much free time I can string together, but I will update as I make progress.

Regards,

Silber

I had intermix at 108k and had the head repaired. Still going strong at 156K then comes a 3.8 build. Cleaning out the chocolate milkshake water took a while.

My car had not the best maintenance before I got it, so at 116k I figured most everything inside the engine was tired and could use a refresh. I’ve already bought all the parts, so at least for this go-round, the die is cast.

Silber

Silver, please keep in mind that these cylinders are prone to wear into an oval shape and even by using new rings you may have issues..

I had considered that. I was planning to go with a 3.8 LN Nickies rebuild for that reason. Thing is, I just cannot afford that. So, I found an engine case that I think has less than 30,000 miles on it and figure it is going to break into the new rings all right, and people seem to get well over 100,000 miles out of a block. By that math this block should have maybe 70,000 good miles left.

If it doesn’t work out I made a gamble and lost but I am reasonably confident if I don’t screw up the rebuild it will give me enough miles of fun to be worth it.

Silber

The issue with these blocks is that the cylinders and rings wear together, leaving that as the cylinders "ovalize (is this a word?)" so do the rings. So, mixing and matching, even with a low mileage block, may be problematic as new rings will not seat properly.

Do you have access to a machine shop? The first thing I would do is have someone reliable to measure out your cylinder specs. If they're still round, proceed with build. If not, abandon all hope of success without bore and sleeve.

Sent from my SM-G970U1 using Tapatalk

+1. No sense building it if your cylinders are tapered and oval. You will end up with crappy ring seal and excessive crankcase pressure. Better to get the pistons (with rings) from the block halves, assuming they were numbered and use that than try to get new rings to seal

Engine is out!
https://adobe.ly/3BVvP3q
https://adobe.ly/3tscKmg

Beginning to reconsider my decision not to go 3.8. Might make sense to save up a while and delay the project in order to get a better result.

Going to start the tear down and see what I find, make final decisions once I see what I’ve got.

Silber

Upon close inspection, my low mileage case has a gouge in #3 cylinder rendering it unusable.

So, I am returning all the re-ring specific parts (block, pistons, rings, etc) now. Going to get the heads off and if there is no damage to my short block, put the heads on. If the short block is in any way damaged, then do a full tear down and save up for a 3.8 build.

Glad I posted this, the responses really took some wind out of my sails, but that seems to have been a good thing.

Regards,

Silber

yea, thats what really sucks about these engines. There is really no way to soundly rebuild them unless you find a unicorn with round unscored bores or you go to sleeves. The rest of the stuff isn't hatefully expensive (but damn there are a lot of ancillaries and gaskets), but having to drop $5k to sleeve them blows.

Quadcammer, yes sleeving the block is expensive, but I believe that the kit of the 3.8 LN Nickies, include the sleeves, new pistons and rings..

It does. But for instance, I recently rebuilt the S52 Inline 6 in our M3 endurance car, and a set of JE pistons, with Eagle rods, was $1250. Boring and honing the cylinders cost about $350.

So you're talking roughly $1200 for the BMW motor, vs $5k for the M96. Not earth shattering, but not chump change either.

And then there is the wait for the LN Nickies. Mine took 6 months. Waiting to see how long Husker Boxster's Cayman engine takes to get LN Nickies since his is currently there. The Nickies need to be plated after they are installed, that is what takes so long, very few platers in the country that can do this work.

Well, I’ve now got the engine on a stand and ready for disassembly. I have set aside this evening to pull the 1-3 head.

I’m fairly certain that something went wrong on #2 (melted spark plug boot, spark plug stuck in the head) so I’m guessing a 75% chance piston is melted or Dchunk, 25% chance head issue that left the short block usable as-is.

Will report here with pictures when it’s off.

Silber

Well that was quick. I pulled the intake and found all this stuff in there. Pieces of head, valve, and piston ring.

https://adobe.ly/2XhCdDa

Once I got the intake off I could look down and it was indeed number two that had failed. At first glance, it looked like the cylinder walls were not damaged. Now I’m not so sure, I got out my El cheapo borescope, and it looks like the wrist pin is missing so the rod was moving free in the cylinder smashing the remains of the piston. Mainly I’m concerned about the crank and crank carrier, if those are the only thing salvageable out of this engine I will be ok with that.

At this point I am absolutely 100% certain this engine is not re-buildable just by putting heads on it. Looks like it’s a 3.8 for me.

Silber

that case is definitely not going to be usable. Crank and carrier are crucial although I suppose you may be able to align hone the carrier to work with a different crank or vice versa.

The low mileage case I bought to rebuild had cylinder damage that made it unusable for a re-ring, but not enough damage it wouldn’t be rebuildable by LN. The seller let me keep it so I’ll probably give that one to LN.

Looking with the bore scope, the rod was not visibly bent, and there was no visible damage to the crank carrier. I am hoping I got lucky, but will definitely check everything!

Silber

I think I figured out what killed my engine. And it’s a new failure mode that I hadn’t heard of before.

If you look closely at one of the intake ports where the cylinder didn’t have an issue, you can see that there’s a bright line of clean aluminum where the gasket was.

https://adobe.ly/3lzIlzc

If you look at number 2, the cylinder that failed, the bright clean line goes all 3/4 of the way around but closest to the center of the engine there’s dirt all across the clean line. Looks like there was a vacuum leak from the plastic intake manifold warping. All the bolts were tight, nothing was loose. Yet there was not a good seal.

https://adobe.ly/3Cbidkw

Since I was a autocrossing the car, there was sustained high rpm usage so that vacuum leak caused cylinder 2 to to get hot enough to drop a valve.

How do I prevent this from happening again? Not sure yet, I’m going to see if the new seals are thicker than the old ones perhaps, I might run some extra sealant around outside the factory seals, not inside. I’ll probably trash that intake manifold and buy another used one but what’s to say the new one won’t be already warped or warp after I put it on.

Food for thought.

Silber

I would think that if what you describe really happened then exhaust valves would be burnt not dropped.
You can burn hell out of valves and never drop them. I would want more "proof" that a vacuum leak caused a valve to drop before I went that route.

You might be right, I don’t think it will hurt anything to make sure that the intake seals completely.

However, Something caused #2 to self destruct. The spark plug got so hot only on that cylinder such that the boot partially melted and stuck to it.

That much heat is usually caused by overly lean, caused by too much air or not enough fuel, or a combination of the two.

I already bought 6 new injectors, so covered not enough fuel. Making sure the intake seals should cover too much air. Think that should keep the new motor from suffering the same fate, any thoughts?

Regards,

Steve

Perhaps you could look at the exhaust valve seats and see if there is evidence of excessive heat. Or if there are other signs of excessive heat on the head, piston, or any area around cyl. #2. The piston should show signs.
My thought is that your engine went because of a mechanical issue versus a fuel/Air issue.
It is possible that a single cyl. went lean and burned up but it should have showed signs of that before imploding. You should have felt the cyl. dropping out or something.

This isn't entirely accurate. Lean burns do not cause excess heat. Lean burns can cause detonation, which forces the flame wave in unitended directions and consequently cause certain areas to run hot. Lean burning by itself, however, runs cool. If you don't believe me, grab a blow torch and max out the flame. Then start turning down the fuel knob and see what happens. A more objective proof of this lies in aerospace applications, where pilots have relied for many decades on cht and egt readings to monitor their engine status. Any pilot who understands these number will back up what I am saying here.

Sent from my SM-G970U1 using Tapatalk

I just RE-viewed the picture in post #15 you can see ring bits, valve head, and other bits.
If that hole got hot enough to cause a valve to drop you should see signs of excess heat on the ring pieces or valve bits. I don't see that.

The exhaust valve that is missing, the shaft is perfectly straight. I’m guessing the head fell off that one from heat, or possibly simple mechanical failure due to structural issues in the metal? The intake valves that are damaged, the valve shafts are bent, making me think this might be collateral damage from the exhaust valve?

I will get the head off this weekend and post more pictures of the carnage.

My main concern is ruling out anything that moves from old engine to new engine as the cause of the failure. Whether in the end it was the injector, the intake or a mechanical problem with the piston or head is not as important to me as keeping the new engine safe from a repeat performance.

Silber

Oh, and I read an interesting take on why people think an engine running lean is hotter.

A perfectly tuned engine usually runs richer than stoichiometric. So if you lean out from there it gets hotter till it hits stoich, then cooler after that.

So an engine running leaner than a slightly rich tune gets hotter but leaner than stoichiometric gets colder.

Does what I read make sense?

Silber

Silber,.

Fuel injected engines that run closed loop operations shoot for stoich. Nbo2 sensors actually only work around 14.7 because of their design. Always (unless manipulated ;) ). This is because of emissions.

Open loop operations are different. They run much richer (afr 12-13) for better power production, retaliation prevention, and cooler running. But open loop operations only happen under very few circumstances.

I have a wbo2 installed and can confirm that the 986/996 does run at 14.7 closed loop, 12.3-12.5 open loop)

Long before there were o2 exhaust sensors, pilots used cylinder head temp and exhaust gas temp readings for each cylinder to understand what was happening inside their engine at all times. It's a simple system that can tell you an exceptional amount of a running engine.

In terms of temp numbers, egt is representative of the combustion process and changes with each stroke. stoich is max exhaust temp. Max power is about 75deg lower than stoich (this is referred to as rich of peak, or ROP).

Cylinder head temps are different. They are closest you can get to measuring the actual cylinder temp. Much slower to react to changes because we're measuring the temperature of metal but still very important because it shows what's happening to the engine material.

For example, if you run lean and start to detonate, your head will often heat up because of improper flame wave propogation. However, less fuel causes a cooler burn, so you'll see a cylinders egt drop and cht rise.


Back to your problem though -

Ok looked at your pics. Was your gasket intact on cylinder 2? That pic is concerning but I don't know if that was the culprit.

The exhaust valves are sodium filled. Good luck melting one of those. A garage in ga with extensive m96 data who actually shares his data (NOT RABY) found the only valve related issues on the 2.7 and 3.2s to be related to lifter springs. Navarro mentions that on his website as well.

https://newsite.hamheads.com/2016/12/01/m96-m97-porsche-cylinder-head-specifications-and-general-info/

I guess I'm a bit confused though...How do you know what your valves look like if you have pulled your head yet?

Sent from my SM-G970U1 using Tapatalk

Thanks for the great explanation on lean versus rich in these cars. Very helpful!

I have pulled off the exhaust manifold and the intake manifold and I can look down in the ports and see that both of the intake valves are broken off and one of the exhaust valves is broken off. I found bent pieces of intake valve stem in the intake, but looking in the exhaust port the exhaust valve stem is in perfect shape other than missing the head.

Whatever happened it was isolated the cylinder number two, I don’t see any of the same problems on 1 or 3 leading me to believe it it was not a valve train issue.

The intake gasket looked relatively normal, but looking at the head it doesn’t look like it was sealing properly. Thus my theory that the intake was warped. I’ll post a picture of the gasket later.

Silber

From what I see in your picture in post #15 and what your describing here. (valves broken off). There was a mechanical issue. Broken wrist pin, broken rod, valve spring or lifter issue. will be interesting to see what you find with further disassembly.

I think you are onto the mark with the valve spring. One thing I did not mention is the car ran, but poorly, after it lost all power during a run. I started it maybe five or six times with no power and no expensive engine noise. But the last time I started it, it made all kinds of expensive engine noise.

So, if the valve spring failed partially maybe the valve would not hit the piston but if it later failed completely then it would and that would cause the damage chain to start.

Headed out to the garage in a few minutes will know more tonight hopefully.

Silber

yup with that info certainly sounds like a mechanical issue rather then heat.

:eek::eek:

uhh, sorry to hear that.

You and me both! When it had low power but would run, I was thinking some sort of easy to fix valve train issue. once I heard all the crunching I knew it was probably not going to be as easy as I had hoped LOL

On the good side the crunching happened basically idling in the garage so the damage is not as bad as it could’ve been if I were going at high RPMs.

I got the valve cover off the 1-3 head tonight:

https://adobe.ly/3kmd1Vc

I found a loose valve keeper, but the cams look to be in very good shape. May have the magnafluxed just to be sure.

My garage is kind of a mess I need to pick it up and build a workbench to keep all the engine parts sorted out on. Will probably come back to this Sunday morning.

Regards,

Silber

Broken Valve Spring.

https://adobe.ly/2XAyy3y

I guess the moral of the story is if I hadn’t keep running the car after it stopped putting out power, I might’ve gotten away with just a valve spring. Since I kept running the car until I heard expensive noises now it’s a rebuild. Maybe there’s a lesson in that I don’t know LOL

Edit: Kept going with the tear down and got the 1-3 head off. As expected there were three valves broken off, the head was kind of munched, Cylinder walls were cracked, and there was a small chunk missing out of the bearing carrier. now I have to try to source one of those.Got a lead though we’ll see what happens.

Silber

If cylinder walls are cracked, this isn't a rebuild any longer...



Sent from my SM-G970U1 using Tapatalk

I have a spare block and wasn’t really planning on using this one anyway, however when LN sleeves the block they machine out the entirety of the cylinder walls, and replace them, so cracks may not have disqualified it.

Interesting video:

https://youtu.be/ebxkhGTyL9w

Silber

Found a 28,000 mile crank/bearing carrier, on its way here.

Plans to travel to LN Engineering this week to drop off my spare engine block for 101mm Nickies conversion.

Finish tear down of bad engine hopefully tonight!

Regards,

Silber

Got the 4-6 head off. Cams are good, but lifters have some pretty obvious circular wear patterns. Lifter holder on this side looks pretty worn as well. Fairly surprised there is such a pronounced difference side to side. Almost seems like two different engines.

Ok, so I can’t sleep and I’m thinking a bit crazy:

https://www.fvd.net/us-en/100105986002/sport-camshaft-set-986-boxster-s-32-l-m9621-engine-schrick-hydraulic.html

My thought is, a 3.8 with 3.2 heads is going to be a real stump puller due to the large displacement and small valves. This kind of setup really builds low down torque, but chokes off the high RPM breathing.

Would the Schrick cams give me the best of both worlds? Sacrifice a little stump pulling to get some top end back, or would I end up with a flat low end and still be choked off high RPM, basically gaining nothing and losing my low end boost.

Any engine gurus want to weigh in? And, up front I should say that I know the 3.2 heads on a 3.8 isn’t ideal, but it is what I am doing and I’m going to try and make the best of it, So advice to go buy different heads won’t be very helpful.

Appreciate everyone’s time so far. You maybe wouldn’t think it would be, but tearing down a mostly dead flat six has been pretty fun so far.

Silber

EDIT: The cams on the FVD website are SUPER agressive, high lift. May have found a better set. First image is the milder set, second image is the set with the high lift, third set is a compromise with the same degrees as agressive set, but less lift. Not sure the stock valve springs could handle the high lift cam.

In the light of day my camshaft idea seem a little extravagant, probably just going to go with new lifters and my current cams.

Got the 4 to 6 case half taken off, going to lift out the crankshaft cradle tomorrow. LN Thursday. Confirmed the part number On my new crank box matches the part number on my current one, seller is shipping tomorrow.

Hopefully my ramblings are helpful, I’m going to try to post less frequent updates with more substance So as not to clog up the board with my rantings.

Silber

I seem to recall that the crank carrier and crank are matched to some extent. I'm guessing you would need to measure the bearing clearance and consider an align hone

Just thinking about it, main bearings are a standard size. They have to fit the crank on the inside, and the carrier on the outside within thousandths of an inch. If there was much wiggle room there the whole system would fail, so my expectation is I’ll be fine.

That being said, My plan is to:
- check the crank for runout with a dial gauge in old carrier
- Measure main bearing clearances in old carrier with plastigauge
- check rod tolerances of old rods with plastigauge
- Check crank endplay in old carrier with dial gauge
- have crank magnafluxed
- check main tolerances with new bearings in new carrier with plastigauge
- check rod tolerances of new rods with plastigauge
- check crank endplay in new carrier with dial gauge

If I pass all those tests I think I will be OK, if any of the tests fail then I will have to figure out why and get back to the drawing board.

Does anyone happen to know the main bearing and rod bearing tolerances offhand?

EDIT: apparently rod and main measurements on these engines is a big secret, added steps to measure what I have first. My crank carrier damage is out by the cylinder, and relatively minor, so the measurements at the journals should still be good.

Silber