Bericht: Motoruberhohlung E28S M5

[alpinab10biturbo]

I received a small update from the engine rebuilder who just started working on the engine by taking it apart and measuring the bearings, cranckshaft, pistins etc.

I have to wait untill Monday before I can take detailed pictures, but he informed me that the cranckshaft and bearings are OK, but the clearance between the pistons and cylinders is between 0,12mm and 0,16mm. According to him this is twice as much as regular.

I did call my dealer for factory specifications, but they have no such data for the M88 engine, only for the S38 that apart from the B38 engine has the same bore. The chief-mechanic send me the data by fax so will compare them with the obtained results before drawing conclusions.

According to the engine rebuilder, the measured wear is twice as much as what is normally accepted. He suggests going oversize and use new oversize pistons.

[alpinab10biturbo]

I had the subframe bead-blasted and powdercoated. Looks much better then before.

[alpinab10biturbo]

The TIS specs for the S38B35 are:

Bore: 93,40mm (+0,01/-0,00)
Bore-midsize: 93,45mm (+0,01/-0,00)
First oversize: 93,60mm (+0,01/-0,00)
First oversize: 93,80mm (+0,01/-0,00)
Surface roughness: 0,2 ... 0,4 (Ra [mu])
Max allowable ovality: +0,005/-0,005 (mm)
Max allowable conicity: +0,01 (mm)

[alpinab10biturbo]

Whilst on my way to work this morning, I made a small detour to see the fully disassembled engine from my M88/3 which was carried out by member Javier who coincidentally works for the engine shop that we selected to carry out that work.



The engine block in disassembled state.



The six pistons.



The internal parts layed out on a table.



Just a randomely selected cylinder (this case #6). clearly visible is the machining pattern (grooves). Clearly visible is [d1] The brown tinted surface on top of the cylinder and [d2] the cleaner looking surface from the bottom up until one third of the height. The surface in between [d3] is used by the piston-springs and wear out over time. With other words, there diameter of the bore that is 'used' by the piston springs wears out in time and with heavily worn engines, a ridge can be felt on the cross-over points between these surfaces that may prevent the removal of the piston without damaging the cylinder walls. If that is the case, the cylinder needs to be bored anyway so that doesn't matter. Javier did make a report with measurements that he carried out to determine the wear of the engine.

Javier was not present this morning so I had to do with the measurements for [d3] that he wrote down.

cyl #1 [d3] = 93,48mm
cyl #2 [d3] = 93,47mm
cyl #3 [d3] = 93,47mm
cyl #4 [d3] = 93,48mm
cyl #5 [d3] = 93,48mm
cyl #6 [d3] = 93,48mm

I discussed these results with his collegue and additionally measured [d2] that turned out to be 93,45mm for all six cylinders. The bore wear is determined by substracting [d3]-[d2] or 20umtr to 30umtr for all six cylinders. The original diameter is what is strange. According to the BMW specifications this must be 93,40mm ('size 0'). Coincidentally, 93,45mm matches the so called 'size 00' in the technical information. Above 'size 00', there is 'size 1' and 'size 2'. Both are oversize (93,60mm resp 93,80mm). Unknown to me is were 'size 00' is used for but it does list a matching piston sized 93,40mm.

The pistons in my engine measured.

piston #1 [d] = 93,35mm
piston #2 [d] = 93,33mm
piston #3 [d] = 93,31mm
piston #4 [d] = 93,32mm
piston #5 [d] = 93,32mm
piston #6 [d] = 93,35mm

There are no inscriptions anymore what the original construction size is, but I assume it is 93,35mm, the matching piston for 'size 0' since for 93,40mm the wear must be visible on the piston. So someone must have decided in the past to machine the bore according to 'size 00' whilst using the pistons for 'size 0'. This equates to 50umtr or 30% of the allowed piston-to cylinder wall clearance (0,15mm) !! I have no clue who is responsible, but it sure is strange. Sounds like some cost cutting measures by a dealer who had a warranty claim (or something like that).



A randomely selected main bearing journal on the cranckshaft. Their diameter measured 59,97mm for all but one that measured 59,98mm.



A few (lower) main bearings still in the main bearing caps. After 155k km these still look excellent.



Some other (upper) main-bearings. Also these are still excellent.



A randomely chosen piston rod with the rod-bearings still installed. Also these bearings didn't show a sign of wear. Their journals on the cranckshaft measured 47,98mm for all six.



One of the chain guides. Well worn and thus in need for replacement.



The timing chain sprocket on the cranckshaft. The tooths are a bit too sharp so also these will be replaced.

[alpinab10biturbo]

The disassembly and analysis of my M88/3 engine is almost complete, so it is time to draw some conclusions. The facts at a glance:

1: The valve clearances for all intake-and exhaust valves were to small.
2: The cylinder head gasket has S38 printed on the topside.
3: The engine wiring loom has been damaged on two spots.
4: The original bore diameter of all six cylinders is 93,45mm, matching ‘size 00’
5: The pistons in diameter are close to the value for ’size 0’

Facts 4 and 5 combined led me to believe that the engine needed serious repairs at some stage in the cars life. Normally, if an engine needs to be bored, a matching set of pistons must be used as otherwise a significant part of the engine-life is wasted. From the factory, BMW manufactured the M88 and S38 engine with a 0,05mm clearance between cylinder wall and pistons. As maximum clearance BMW specifies 0,15mm for a worn engine. This has not so much to do with oil consumption, but more with the maximum tilt of the pistons during cold start.

I rule out excessive piston wear, as these are not designed to come in contact with the cylinder wall. There are no (visible) signs of scoring either, so I can safely conclude that the pistons are originally 93,35mm in diameter, matching the criteria of ‘size 0’ in the TIS. The difference in measurement recordings can be explained by measurement tolerances and metal expansion / retraction due to temperature cycles.

The main question is: What caused the 93,45mm bore size what should have been 93,40mm? The difference seems to be minor, but bare in mind that the maximum specified piston-to cylinder wall clearance is 0,15mm. Subtract the initial manufacturing clearance of 0,05mm (the tolerances are ignored for simplicity) results in a mere 0,1mm as maximum wear. As a rule of thumb, one can assume that an M88/S38 engine lasts 300k to 350k km before needing a rebuild (abuse excluded). With these figures, the slope of the engine wear can be determined to a value between 0.00000029 and 0.00000033 mm/km.

Compared to the original bore size, the cylinder wall wear on my engine is determined at 0,03mm. Combined with the very good condition, i.e. minimal wear on crankshaft, bearings, camshafts, the true life of my M88/3 engine can be assessed to between 91k km and 104k km. Granted, this is an educated guess, but I can safely assume that this is within a ten percent accuracy, setting the limits to 82k and 115k km. However, the mileage indicator on #231 shows 155k km, meaning that the engine must have been opened once before. Corrected for the calculated true life, this must have occurred between 40000km and 73000km. Chances are that the factory warranty had already expired and that #231 was still in the hands of the first or second owner and the repair was covered under goodwill or a used car warranty. In that case, it was the seller’s responsibility, most likely a main dealer.

The logic is that if the pistons are original, they measure 93,35mm nominal. By increasing the bore to ‘size 00’, the piston to cylinder wall increases to 0,10mm! Although still within limits (0,15mm max), roughly halve of the engines operating life expectancy is wasted, meaning the engine wear reflects that of an engine with 150k to 175k km. A dealer would call this a successful repair with the problem shifted outside their liability or possibly outside the cars technical and economical life. The customer would not notice that the trade-off in engine life is at least 77000km just to save on a new piston set!

One may argue that this occurred at the factory already in 1985, but I think that is unlikely for a variety of reasons. First, if this engine block has been rejected initially due to a tolerance issue, it is very unlikely that the engineers would have enlarged the bore to ‘size00’ without using a matching piston set. Second, the engine wiring loom has been damaged and this would not have occurred at the factory. Lastly, the cylinder head gasket that came out of this engine has S38 printed onto it. This designation did not exist at the time #231 was manufactured in October 1985, let alone the manufacturing date of the various single parts some of them being stocked months before.

[alpinab10biturbo]

This evening, I started with one of my secundary goals that is so easy with the engine out of the car. When I removed the subframe, the mating surface of the chassis-beams were a bit corroded. Just some 'fly-rust' and nothing serious, but something that needs attention.



With some light sanding, I could remove the corrosion and the old primer coat.



More serious are some welds deep in the engine compartment. A dremel came in handy and allowed me to remove the rust between the sheet metal plates on the borrom of the following image. The weld right next to it was attacked by corrosion as well, but nothing serious.



P.S: The bracket for the AC condensor will be replaced.

[alpinab10biturbo]

The cracks between the valve seats are not uncommon to cylinder heads of the S14/S38 and M88 engine families, but in most of the documented cases, these cracks appear only between the exhaust valve seats. In that respect, the condition of #231’s cylinder head is more serious with cracks between the intake-and exhaust valve-seats and cracks between the exhaust valve-seats and the spark plug holes. I have doubts about the maintenance being carried out by the previous owners before I took over ownership of #231. Already in the spring of 2001, I had to replace two expensive parts that were manually modified for some reason.





Both facts combined led me to believe that the condition of the cylinder head is not necessarily related to the piston to cylinder wall clearance being out of specification. The cause of the cracks between the spark plug holes and the exhaust valve seats may be over tightening of the spark plugs combined with high temperature rises in a short period of time (dT/dt) or with other words, plain abuse by ignoring a proper warm up phase.

With the valve seats affected by these cracks, their stability is lost and valve clearance affected. Even if I had adjusted the valves in between, an engine failure could not be avoided and would have occurred sooner or later. With the removal, disassembly and failure mode analysis, the second phase of this project is almost completed. The third phase, the actual rebuild will formally start next week after I have given the approval to proceed. The third phase will start with determining the usability of the cylinder head. After the visual intake inspection, Thomas Hoop from the ‘Twentse Motoren revisie ‘ gave the preference to using my original cylinder head, as it had not machined before. The overhaul of the cylinder head is a difficult route, but necessary as using the cylinder head ‘as-is’, is not an option and replacement cylinder heads are not available through the BMW dealer network anymore. This is my main worry so my project planning is as follows:

1: Confirm the usability of the cylinder head.
2: Purchasing the parts that have a time-critical impact.
3: Machining and overhauling the cylinder head with new valve guides and valve seats.
4: Machining the engine block.
5: Assembly of the engine with new oversize pistons (93,55mm), matching piston rings, intake-and exhaust valves, new timing chain and associated parts etc.

With step one being confirmed, I placed the order for six new pistons today. My planning of the completion of the third phase by the end of Mai still stands, but this depends on Thomas and Javier’s operational planning. I’ll try to post some updates from time to time.

[alpinab10biturbo]

Before installing the revived subframe , I had to do something about the surface rust on the front chassis beams. Member 'richardbaxter' brought an American marine product to my attention, which is known as POR15. POR stand for Paint Over Rust and is a primer coating that can be painted over rusted surfaces. It goes to far to describe the differences to other etching coatings in detail; for that I refer to the POR15 website .

Since I wanted to try it out first and don't need a large quantity, I was seeking for a starter kit that contained all the substances. It was available on the US website, but not from the Dutch resellers who prefer to sell the normal containers for way too much money. I couldn't order the kit directly in the States, but was refered to their German distributer, Ronald Hoeseler who offered the starter kit for 18 Euro's which is very reasonable compared to the US$ price. The DHL shipping cost almost outweighed the value of the kit itself, but the kit was delivered in four working days.



The kit contains a bottle with 'marine clean', which is a highly concentrated degreaser that can be dilluted with water up to a 1:10 rato. The second bottle contains 'metal ready' fluid, which leaves an etching layer on the cleaned metal. The metal container contains the POR15 primer coating. It is available in black, silver, grey and transparrant, but since #231 is salmon silver, I ordered the silver primer coating. Two latex gloves, a pencil and an instruction leaflet are included as well.

Since I already cleaned and degreased the affected areas some time ago, I only used the 'marine clean' to remove the last contaminations. I mixed the 'marine clean' with water in a 1:10 ratio. It takes some time to dry, after which the 'metal ready' can be applied.



The last step is the application of the primer coating.



I added three thin layers which is recommended for industrial applications. The curing time is two to five hours. I'll leave it for overnight so that will do.

[alpinab10biturbo]

The cylinder head has been blasted to remove the debries and get a full insight in the depth of the cracks. Clearly visible is that the crack (indicated by the green arrow) runs deep into the exhaust port. The true extent of these cracks was only reveiled after the blasting process. The other exhaust valve port (yello arrow) has similar damage, although not that bad as the crack is limited to just underneath the valve seat itself. The other exhaust valve ports have similar damage. It was only a matter of time before these extended to the water galeries.



The last step for preparing the cylinder head for the metalurgist (alloy specialist) is the removal of the valve seats. The first one was done in my presence.



With a dremel tool, an incision is cut in the valve seat to relieve the tension so it can be removed.



After the removal of the valve seat, the crack can be seen in its entity. The crack extends from the combustion chamber, alongside the mating surface of the valve seat, well into the exhaust port.



The valve seat from another exhaust port. The purple arrow points to a similar crack of cylinder #2. Also visible is the difference wear inflicted difference in the valve-seats slope. Due to this, the valve seat does not seal off properly and affects valve clearances as well.



The remaining twenty-three valve seats will be removed today after which the cylinder head will be dropped of at the metalurist's. The following picture shows a cylinder head from an Opel/Vauxhall Ecotec engine that suffered similar damage after a mere 150k km on Liquid Petrolium Gas (LPG). The valve seats have been renewed, but need to be machined before installing the valves.

[alpinab10biturbo]

Last thursday evening, I painted the lower front cross member with POR15 to neutralise the corrosion underneath. The primer paint fills the gap between the spotwelded plates rather nice.



I also installed the front subframe earlier this evening. As I have no complaints about #231's handling and the visible inspection of the bushes and control arms reveiled no defects, I decided against replacing the control arms and bushes. I did use new nuts and bolts though.



I also did not replace the pitman arms as there is no need to do so.

[alpinab10biturbo]

About two weeks ago, a small parcel, containing six pistons, twelve intake valves and twelve exhaust valves arrived at my dealer.



The six boxes with pistons. To reduce the damage risk, I only unpacked one to make some detailed pictures.



The inscriptions on top. With a 93,60mm bore, the piston to cylinder wall clearance will return to 0,05mm (the last 0,005mm is for the honing pattern). Please note the S38 inscription; it appears that BMW does not distint the M88/3 from the S38 anymore. Judging from the date code, these pistons have been manufactured 12 years ago. By the time I enquired about the availability, BMW had an inventory of twenty-three of these pistons; now just 17.



Comparison between one of the new-and one of the old pistons.



The piston crown!



The bill.



One of the intake valves.



One of the exhaust valves.

[alpinab10biturbo]

Whilst we are waiting for the cylinder head being returned from the metalurgist, the engine shop started with the engine block. The bores have been enlarged to 93,60mm to obtain a clearance of 0,05mm with the new pistons. In the mean time, the bores have been honed as well.



An important part of a bottom end overhaul is the selection of the bearings. These are available in standard size (60,00mm) and three oversizes (59,75mm, 59,50mm and 59,25mm) in case the crankshaft needs to be machined. As this is not the case with my engine, the recorded measuremets determine the required bearing size.

However, the bearings are avalable in two color codes (red or blue). This is a classification that depends on the production tolerance from crankshaft and crank-case. With other words, a cranshaft is supplied with matching bearings. This is determined by the colored dots on the crankshaft that in this case is red.



Fortunatelly, the colored dots were still visible on this particular crankshaft. If they weren't the BMW repair instructions state that the crank-case determine the required bearings type. In my case, the crank case has red dots as well, thus matching those on the crankshaft.



With other words, the main-bearing selection is unambiguously determined by the crankcase, cranshaft, but also the markings on the old bearings. This is a bit more difficult for the piston rods. As clearly can be seen, these have red color dots as well. However, matching bearings are not available anymore.



Only rod-bearings with the yellow color code can be bought. This is worrying, but I don't have a choice so I have to do with these. Given the fact that the tolerance of these bearings is less then 1/1000 of a millimeter and the recorded journal diameters indicate the clearance ending up within specification, plastigage measurements must provide the empirical proof.

In the mean time, I have ordered all the parts that are needed to finish the short engine block. This includes the gasket sets, sprockets, guides, timing chain, bearings and all the bolts and nuts to complete the short engine. I also am preparing three scientific documents that decribe the failure mode analyses and the rebuild with underlying proof of all the choices that have been made, including the overhaul proces.

[alpinab10biturbo]

After boring and honing the short-engine in the week before, the machine shop removed the water-channel coverlids and cleaned the engine thoroughly, after which it received a finish with application specific black paint. In the mean time, my dealer received the three separate batches of parts that I ordered for the assembly of the short engine.



To manage this project successfully, I have decided to separate the parts ordering to match the different rebuild stages. This not only gives a better overview of the deliveries, but also reduces the risk of mixing up parts between the few stages. Fortunately, BMW still offers gasket sets for the M88/S38 engine family. Two sets are required; one for the cylinder head and one for the short engine. The gasket set for the short engine is relatively cheap. The stealer wants ~65 Euro ex VAT for it which is a bargain if one considers that both crankshaft seals are included.



Much more expensive is the cylinder head gasket seat that retails for around 320 Euro ex VAT. But it contains all the required gaskets including the cylinder head gasket (120 Euro ex VAT) and the valve stem seals (45 Euro ex VAT).



Clearly visible are the two valve stem seal repair kits. In total the cylinder head gasket contains 62 separate parts. Many of the O-rings have a minimum order quantity (MOQ) when ordered as single parts so the gasket set makes a lot of sense.



The sump has been cleaned and painted in application specific silver paint.



The same applies for the timing chain cover. Clearly visible is the front crankshaft seal.



The new pistons have already been installed on the piston rods awaiting the delivery of the new bearings shells.



All the timing chain related parts will be replaced with new parts. The following picture shows the new crankshaft sprocket for the timing chain and the oil-pump chain.



The following picture shows the rear crankshaft seal. It is mounted in an alloy cover that is bolted to the short engine. The bolts and washer plates are replaced as well.



I ordered a new hub as well. The old one contained a small groove, that could be removed by machining, but who bothers at 75 Euro ex VAT. Actually, this is an obligatory part when performing a full engine-rebuild.



The remaining parts needed for assembly of the short engine have arrived at the dealer today. Monday morning, the engine shop will continue with the assembly of the short engine. The cylinder head has not yet been returned from the metallurgist. The expected delivery time will be three weeks, thus the second week of July. This delay means that I won’t be able to install the engine before my summer holidays at the end of July.

[alpinab10biturbo]

The following picture shows the old and new chain guide-rail (tensioner side). The upper part is the original part, presumably 23-24years old. The lower parts shows the replacement part. Clearly visible are the wear patterns caused by the timing chain.



A close up of the cracks in the rubber lining.



The old crankshaft sprocket (left) with the replacement part (tight). I now see that the hub-side of the old sprocket appears to have some damage, but I am not sure about this as the picture only shows a 2D perspective.



The old parts will be kept for the time being. Clearly visible are the dismissed main crankshaft seal (flywheel side), the old chainguides and the timing chain.



One of the old main-bearing shells with a date code that prooves that they were manufactured before #231 left the assembly line at the Preussenstrasse in 1985. This also prooves that the crankshaft still is the original part (new crank comes with new bearings). This unlike a statement from a BMW representative who doubted that this was the case for this particular car.



Suffice to say that the chain tensioner will be sourced from the S50B32 as is described in this reference article.

[alpinab10biturbo]

With teh arrival of the last missisng parts such as the rod bearing shells, the short engine is now in the middle of the reassembly phase. The six pistons and rods have been installed as can be seen on the following picture.



Last week, Ronals used the old crankcase bolts, but with the arrival of the new parts he could complete that part yesterday as well. I am not that keep on the black paint on the pipe for the dipstick, but since I want to replace that part as well, I don't care.



The number of replaced parts increases slowly but steadily.



By advise of member Stevie, I asked Ronald to soak the new chains in a bucket with oil for a week or two. This will allow oil to creep alongside the pins between the shackles for proper lubrication. This will significantly reduce the start-up wear. If all goes well, TMR can complete the short engine assembly soon after.