This is about my 2002 Porsche Carrera AWD Tiptronic Convertible barn find, arriving on the fladbed. Cosmetically a beauty, a well-optioned car but on the downside featuring a destroyed engine (of the type for which repair you get $15-20k+ quotes from the (typically not even a P-dealership) specialist workshops to fix). However, priced as a rolling chassis, I was motivated enough to attempt what I always wanted to do - namely to build a boxer engine of a Porsche 911 from scratch before I can enjoy this engineering marvel in the back of this iconic car . Here a little docu. of my project, which took me about 4 weeks to complete - with almost half spent on finding/waiting for parts, and the rest on daily (well better - "nightly") 3-4 hour sessions in the garage
Fig. 1 :
Car jacked up - here the look under the skirt… The previous owner had the head taken off - revealing the cracked and scored #4 Sleeve…
Fig. 2 :
Disconnecting the motor from periphery - electrically, secondary air pump, drive shafts C2 and C4, coolant hoses @ pump and under body, fuel-line; The AC compressor was already removed - so I could skip that part; I removed the exhaust - which turns out to be unneccessary at that stage - but this is what you learn in the full process.
Fig. 3 :
Of course I do not have a proper shop lift. So, Homedepot provides with Paver stones, and plenty of 2x4's and 4x4's. Stacking up under Oil pan and trans end; Losening the 4 nuts which hold the entre engine …
Fig. 4 :
And then gravity does the rest of the work to get the whole engine effortlessly down in a couple of iterations (lowering front and rear engine ); on the way down I disconnect the hydraulics and remove the coolant reservior.
Fig. 5 :
In the final stage I settle the oil pan on a 4x10x10 block, with haevy duty metal rollers mounted to (extremely simple , but what will turn out to be higlly effective tool ); The engibe is now ready to roll out …
Fig. 6 :
Here is the flat-six M96 bad boy on the floor - What a monster !!! How can this have been in this tiny car ???
Fig. 7 :
Tear-down time (I took many pictures for reference purposes (manuals leave you in the blind here)- here a few @ some of the mile-stones): taking off Intake, wiring harness, pumps , starter, alternator, etc.
Fig. 8 :
Look at the oil pump - the massive Oil-coolant intermix was evident
Fig. 9 :
The engine is stripped bald by now; taking off the head cover, cams, tappets & guide, head, and transmission,
Fig. 10 :
Sonic bath starts for some of the parts
Fig. 11 :
Taking off flywheel and housing , chain guides, IMS bearing (was advertised as ceramic - but turns out to be a std. (steel) roller bearing), oil-pan cover, internal AOS's etc. , all case bolts
Fig. 12 :
Pulling off 4-5-6 case half, unbolt the cradle from the other case half, and taking out the crank in cradle , finally getting down to the bare crankshaft and connecting rods
Fig. 13 :
There are quite a few bolts holdingthe stuff taken apart together ; logical grouping reveals that there are approx. 400+ pieces and identifies the belongings of each in a pretty straight forward way (manual only in case of doubt needed)
Fig. 14 :
The empty engine casing
Fig. 15 :
By this time I got the case insulated the fisrt parts are nice and clean (you can eat of them …)
Fig. 16 :
More bits get perfectly clean through down to the last pore…
Fig. 17 :
BTW-here the picture showing the cylinder heads and pistons; The #4 Valve seal ring is cracked (The root cause of this particular engine failure) ;Piston #4 is smashed (cracked sleeve is the likely consequence of the seized rings) ; #1 piston got damaged too (Seal ring debris pushed back through intake) , but only superficially; Left head and #1 piston would work again, the #4 piston is junk; rhs head surface needs restoration and re-valved ; well-the engine case is of course blown, but an excellent base for upgrade with Nickies
Fig. 18 :
After getting the replacement parts it is time to put things back together; I did not want to wait for the correct lubricants - so I have chosen a Moly-based assembly grease from the local auto store;
Fig. 19 :
The crank shaft is put back together first; installing rods for 4-6 , installation of chains, guide, tensioner slide and intermediate shaft on to crank
Fig. 20 :
By the time the parts keep coming from the sonic bath and the box gets packed to the top ; Well, they don't look 15 years old for sure… (a true pleasure to move on); BTW - for those who may tend to think this stacking as careless - well the opposite is the case all bits treated as raw eggs, and I just was after more space and getting an awesome picture on the side
Fig. 21 :
Interesting fact on the side- the linear reciprocating mass is about 6 pounds in this engine
Fig. 22 :
After a few attempts to use a ring compressor the regular way (from top) , I decided to install the pistons for 1-3 from inside as you would do on a two-stroke engine (This engine has indeed vent cutouts , and after all my process appeared much more gentle to the rings through the slightly tapered lower sleeve, and relativel easy manipulation by bare fingers with literally no axial force) ; The merger of the crank and left case half was easy with a helping hand (to keep the rods upright while lowering cradle as well as guiding chains into drive tunnel)
Fig. 23 :
Piston-clip installation tool fabricatoon starts at the homedepot (in a country where you don't find much metric half-products, I was running through the isles with caliper, and by coincidence Electro-dept, for purpose of surface-installed-wiring they have zink-plated steel piping with OD just below the desired 24 mm ); small cutout (ca. 5x5 mm was done opposed to each other on the tube perimeter; a socket and some just above OD 10 mm metal tubung in sequence acted a pushrod; Note: very helpful is to put a marker on the perimeter, which identifies the proper relative azimutal orientation of the c-clip opening (+-5 degree is fine, as if you go beyond, your clip will not be placed properly, and you will have to remove the clip again (which is doable (I did develop my technique), but not neccessary with a simple sharpie marker on the tool - and of course not to forget to pay attention in the process )
Fig. 24 :
I did one dry-run of piston-clip install for all 3 pistons (yes - until the plug) before I joined and sealed the 4-6 bank cases with Dreibond 1209 sealant and bolting together with proper torque; Final re-iteration - the pistons got installed (here I had no choice but to impove my skills on using the reing compressor tool), staring with #6 & crank near UT correspondingly; a rare-earth magnet served to center the rod to adjust the rod eye pos. azimutally (ref lower eye), and visual in-situ feedback with inspection scope (simple stuff from HD I used already for my last home-remodling project) to check the line-up of rod and piston wrist pin bore (this was also used as feedback to bring the piston radially in place, by knocking the pistom bottom down with plastic hammer)
Fig. 25 :
"Once the rod-piston core line up (the coaxial camera view will show a small dark ring of uniform thickness - which represents the nearest gap between both parts), the wrist pin gets installed: I had no proper sized bar with shoulder etc. , so I simply lowered the pin holding by 2 fingers into the sleeve, and pushing with the bar you can find in the every Carrera to losen the wheel bolts. The wrist pin, pushed all the way to the hard stop is now double-checked with scope ; Finally the clip is installed in the tool, inserted into block until it hard-stops at the piston tool recess - then a few hits on the push-rod until the impact feels stiff and metallic sound (a partial (unsuccessful install feels resilient and acoustically dampend) ; Post mortem the seating check with scope, and higher res image (iphone inside cyl -bore…)); Repeated this for cylinders 5 and minor alteration to #6 (easier as it is closer to case plug) ; Once you have the case plug closed , and all 6 pistons installed you can be proud for a moment . I think this is this piston installation process is complicated, takes skill, and is for sure not economic on the large scale - one of the many reasons I think P-cars cannot be cheap when it comes to the
implementation of the fundamental advantages of a boxer engine into a production car"
Fig. 26 :
Installing crank shaft main seals; IMS-bearings - front and back (Part of oil pump housing);
Fig. 27 :
Assemble Oil pump (Drive pin, gears, cover) ; install water pump; crank pulley wheel; Chain guides and misc spigots, coolant temp sensor , internal AOS's and sump components & Oil-pan
Fig. 28 :
Cylinder heads: I had no clue what people were think when they developed valve spring compressor tools you can find in autostores (what a waste of money); I ended up taking the pusher from such tool only, and installing into a drill-press , and lining up valves and pusher axis by tilting the cylinder head via wood piece (Which also supported the valve when pushing on the valve) ; this simple tool set allowed a seamless diss-assembly and re-assembly of the valves; thorough sonic bath for all head parts thereafter
Fig. 29 :
Note - I am using 997.1 3.6l heads as one of the original 996 heads is smashed and I did not want to spend time to get them rebuilt at this stage. However, my biggest concern was the missing fins / splash guard in the cast below the exhaust cam - fundamentally though, this seemed no problem as my thorough comparison of the 997.1 and 996long block parts let me conclude that the risk of lubrication issues in the head is contained, and likely be satisfactory remedied by afterward upgrades from outside (particularily the upgraded exhaust valve cam covers - particularily 4-6 - which seems to improve the pumping capacity of the shared standard oil pump for 996/997.1 short block), So, the biggest risk seemsd manageable by peripherals, but there were 5 more differences in the head designs identified, where-off where-off the exhaust header mount hole-pattern had to be taken care of at this stage ; Drilling M8 thread holes into the pre-existing casted stud. ; Thereafter the heads were cleaned once more and the valves re-installed valves into original places
Fig. 30 :
Installing the cylinder heads; Lubricating the new head bolts; To meet the factory torque requirements, the 2-step 60 deg final straining was done via simple tool (sheet of paper with 60 deg angle drawn onto and a hole for the socket drive shaft - super primitive, but the 2 deg tolerance can easily be met)
Fig. 31 :
Once the heads are properly installed, the Cams 4-6 are being installed; I started checking the straighntess of the cams by installing them without tappets (the cam should spin smoothing around its own axis (I learned the hard way, as one of the cams from the original 4-6 head turned out to be bent - and had to be replaced);
Fig. 32 :
Once proper spin was checked, the Crank @ lock pos U6 - before TDC (essentially - to put least strain on cams during install…), the tappets and cams were installed (by workshop service manual) ; I did not use a special clamping tool, but I used the clamps of the bad original heads to temporarily hold the cams in place until valve cover was installed;
Fig. 33 :
To adust the cam timing, the crank was rotated CW into TDC; cams did not move as the drive gears were loose yet; what followed was a slightly modified technique to adjust timing, but ensuring that the half-moons are within ca. 1 deg from perfect alignment (a straigt 4 mm Allen wrench, which provided a uniform gap of ca. 1 mm to the head along ca. 100 mm distance - see image); to apply slightly torque to the cams (at the target pos the valves are open...) I made a simple tool from flat iron (1/4 x 1) in from Homedepot, bent 90 degree ca. 10 mm from the bar's end - this served a a perfect wrench to adjust cam angles through the cam notch interfaces - see image) ; This tool was also used to apply the initial torque to the cam drive gears after adjustment; Needed 2-3 iterations (measuring cam angle-adjustment-verification) to get within the <1 deg window
Fig. 34 :
The latter tool was also used to apply the torque to the exhaust cam drive gears ; The initial tightening was applied similarily to intake cam gear, but where the final screw straining by 90 deg took place by releasing the rhs of the cam shaft , and exclusively holding the cam drive gear via the 30 mm hexagon (I was sceptical w/o speical tools as of the chances to slip of the short hex, but a very tight c-wrench, w/o and chamfer did indeed allow to apply to final 110 deg straining w/o slipping off);
Fig. 35 :
In this cam position the cam housing was populated (chain guide between cams, main chain tensioner cyl 4-6, filter, gauges, actuators, gaskets) ; The temporary cam clamps were removed; valve cover and peripherals on bank 4-6 installed (Cam pos sensor, cam covers , vent studs, muffler brakets, intake channels, Spark plugs , coil packs, exhaust heat shield)
Fig. 36 :
Now , all the cam shaft work was repeated on cyl 1-3; the procedure is identical, yet the final crank position is +360 deg CW, and the process starts again at locking the crank at the next U6 CW to install the cams @ least strain; fast forward, the picture shows the state once done and muffler brakets and AOS attached (not final, but it got in the picture...)
Fig. 37 :
More peripherals were installed (oil pump, starter, Knock sensors, oil temp sensor, starter, secondary air supply plumbing, cooler, hoses, main engine bracket)
Fig. 38 :
Vac lines, wiring harness, intake plenums
Fig. 39 :
At this stage three more modifications had to be done to accommodate the 997.1 heads: 1. The exhaust cam covers had to get one hole relocated to fit the head casting 2. The larger 997 crank case vent aperture had to be adopted to the 996 vent-line spigot (I planned to use 997 vent line to AOS, but there were a number of other constraints on the hose/wiring side, which made me decide to retrofit the spigot) 3. the cam pos sensors were replaced by 997 models as the head casting features a larger mounting aperture ; Pictures show mods to both cam covers (4-6 got minor cut-off & special stand-offs to uniformly apply pressure by now offset screw & 1-3 got simply one extra hole on perimerter) , vent spigot (an aluminum adapter plate), 996 cam pos. sensors were replaced by 997 type sensors)
Fig. 40 :
Installing torque converter, trans & exhaust; (wow - this assembly starts to looks just huge - hard to imagine all started with a crank shaft a few days ago…); the engine gets starter gear , torque converter, transmission (note: had to remove starter once more to access the bolts to torque converter) & exhaust system installed
Fig. 41 :
Engine back into car; Porsche factory coolant 50/50 mix; Fluids in (runnning cheap STP mineral oil 10W40 for the first flush (My local store was also out of M1 for that day anyway) ) , Cheap filter (to have sth), and the proper Pentosin hydraulic oil )
Fig. 42 :
Starting 1st time: manual cranking ca. 10 turns; removed fuses for ingintion & fuel to build up some oil ciculation during 2 sets of 5 s jump starts (my battery was dead and the 1000A jump starter gets exhausted after ca. 10 s and needs re-charge) ; With fuses the engine started pretty much instantly ; Runs & sound smooth ; Finishing Coolant addition ( was not entirely empty as after the whole procedure I added ca. 16 liters - which means more than a gallon remained somewhere in the car somewhere); There was initially minor smoke from 4-6 bank coming @ throttle push, but this settled after 2-3 miles (likely one of the oil rings which were cleaned and re-installed, and needed some run-in in its new placement - which is certainly not perfectly in the place as original );
Fig. 43 :
Checking fluids after 10 miles: the magnetic oil pan plug shows accumulation of minor ferrous wear accumulation - not much concerning at this stage; oil and coolant sample shows nothing visually concerning;
Fig. 44 :
First 100 miles: Final check of coolant levels (perfectly stable level); Pelican package got in, which allowed to replace Oil&Air filter (now Mahle) and the oil by Mobile 1 10W40. The car drives and sounds fantastic for its ca. 65k miles on the odometer ;
Fig. 45 :
Update after ca. 500 miles: Observing ocassional hot idle oil pressure just below 1.5 bar; increasing hot-viscosity to 50 now (M1 5W50 was urprisingly locally available at Napa). I am getting in average above 1.5 bar @ temp gauge slightly left from center even after a harder drive 5-6 k rpm for couple min autocrossing local empty parking lot (exploring the 4wd - which I realized is such an asset when it comes to traction on this car over the C2); This weekend (29 Oct I participated in my first Autocross (PCA GGR Round 8 ) at Alameda place . Fantastic event, people and the car ran flawlessly. I am certainly not used to shifting on the steering wheel, but every lap gave me more confidence in the elasticity of this engine at the moments when I my attentiuon is elsewehere than on the engine speed gauge. It will be just fine , and I am looking forward to any upcoming alike events.
Fig. 46 :
Update end Nov ; ca. 1000 miles: 2nd Autocross at Alameda (PCA GGR Round 9 ). Front trans mount gives way, engine purrs like a cat.
Fig. 47 :
Dec 2017 appraoching 2000 miles: Solid (elastomer-) trans mount does great job already for over 500 miles. Adds audible noise and virations, yet allows to push the car in turns as desired. Otherwise, maintaining engine oil level within 1 bars below max level at rate 250ml/250mi is all this car needs.
Fig. 48 :
Sept 2018 appraoching 8000 miles, and going strog....
Oct 2019 approcaching 12000 miles... Car is now a wonderful family member for two years since it had its heart surgery. How vital it is, I decided to test on one of the most beautiful natural race tracks. I booked it for the Halloween weekend HOD driving event at Laguna Seca. It was a very well organized event, and with an excellent instructor on the side, I realized that this is the environment the car reveals its true potential. As for the engine - ran fantastically smooth at over 6k rpm and long tunrns did not casue any lubrication issues reported by some owners prior upgrades. I was fully focused on mastering the trail braking concept, and indeed, I start getting this fantastic feeling in the end. It is addictive, and my smiles extended from ear-to-ear and days after.... I am sure this wasn't the last time and I am already looking forward to another track day. Here a few visual impressions (credits to the trackside photo shoots of the car go to www.gotbluemilk.com):
June 2020 16000 miles...third track day - with extended oil sump, improved baffling (X51), thicker (M1 10W60) oil and playing with telemetry via OBDII; Getting the car to slip in most turns; engine runs cool, with stable oil pressure
December 2020 approx 18000 miles since the diy engine rebuild...fourth track day - just bled brakes, applied my bore scoping method before and after the track day, and experimented with scrubbing into turns; NCRC held a fantastic 7 session track day, whereoff I drove 6 session until my front left brake caliper failed
Through to Sept 2022 the car is at approximately 24000 miles since engine bild. Diven less as it shares miles with another car, but runs with nothing but normal maintainance to this day. Just passed the smog test for another two years to come through '24. However, I got tired of the boring exterior color. So- I tried myself in vinyl wrapping, and changed the color to something i thought looks so much more inspiring - like the sky here in California
Car continues serving as a solid family member... Participates in other passions for biking, fashion, participted in vacation trips, high school graduation, Monterey rennsport reunion or just helping to make cool science on and with it
Serving science ...Here I mapped the mechanical properties of the original cylinder wall and the pistion of this car. Its amazing to see that the piston skirt is made mainly from a soft material, comparable to what a keyboard keys on a computer are made from. The cylinder wll is an Aluminum matrix with up to 100um large Silicon chips embdded to it. With hard particles, comparable to the piston clearance in size, and lacking perfect adhesion, the plastic film on the piston plays a vital role in this amazing tribo-system as it absobs the eventually falling out Si particles, and its high compliance maintaines low contact pressure to the wall, maintaining low chances of scratching the cylinder wall (a.k.a. bore score). It acts like a polishing cloth. Keep it structurally sound, soft, and lubricated and then Lokasil is a wonderful material to have in your car...
For those who are into cars with M9x boxer engines (Boxsters, Caymans, 996 and 997 Carreras) - once the dust around the haevily discussed intermediate shaft bearing is settled, the most important remaining problem are the relatively soft cylinder walls made from LOCASIL on these engines. Bore scoring will eventually affect all of the more than 400k engines produced, and will likely keep those world-wide handfull of highly specialized machine shops busy for quite some time to keep those fine driving machines in service. While there is typically plenty of speculation about whether already scored or not, it typically takes either advanced skills, or constly pre-/post- purchase inspection (because this happens mostly in case the vehicle changes owners) to assess this problem directly by visual inspection. The cost of such is typically high because plenty parts have to come of the car to perform this either via the spark plug holes (top part of cyliner ) or crank shaft side (bottom part of the cylinder). This is typically not sth. a diyer can do routinely - such as fluid-level check at a gas station. Since my own engine build isn't a good candidate for a precise indirect assesment via oil condition or, engine noises either, and yet I do like to use my car the way it was intended to be, this was motivation enough to come up with my own simple methodology to get the bore score status minimal-invasively. The entire process focusses only on the most vulnerable parts of the surface in the upper- and crank-shaft facing cylinder surface. The process does can be done without draining engine oil, and can yield imagery of the surfaces in question on all six cylinders within less than 30 minutes. In the following video I am demonstrating such process from start to finish. Its is worth noting that image quality of the bores and the bores themself could be better. However, the path for improvement on those is clear, while my main focus is to share a methodology, easily applicable by low-to-moderately skilled diyer.
FrequentlyAppliedProcess 1 :
One main disadvantage of these wet sump engines is the relatively large oil surface area, and which causes the vapor pressure to increase at elevated temperatures, and that the crank case needs to be ventiated to maintain correct pressure (vacuum.). For emission-regulatory purposes this ventilation takes place back into the intake for disposal by combustion, while the oil shall return back to the crank case. The device tanking care if this fluid mixture separation is the so called Air-Oil-separator (AOS). It is technically a condensator of a distillation device, which has a cooling chamber (coolant floating around micture), and a membrane, which opens or closes the port to the intake, depending on the pressure differential between crank case and hydrostatic pressure in the intake venturi. Because both, the cooling chamber (plastic) and membrane (rubber) tend to fail, replacement is required in order to prevent the engine from either leaking coolant or supplying the intake with too much oil. The technical challenge for replacing this AOS is that it is located ahaed of cylinder 3, and typically requires engine removal (costly professional service). Engine-in variants exist ("Pelican Way"), but having tried both ways prior DIY myself, would not want to repeat at least the latter, so that I tried a way with engine in, more convenient, and quicker along. This is a video, showing the removal of the AOS w/o removing the engine, and how this can be done within a bit more than an hour. By no means this shall demo how to do it right, but it shall be encouraging the DIYer to actually do and anjoy it a bit more than typically represented.
