Tear Down and Rebuild of My R400 Duratec Engine

[aerobod - near CYYC]

Yes, I feel quite lucky in having these vehicles, James. I had a deposit on the M2 from Oct last year before they announced the spec and prices in Nov, then I managed to get first custom order allocation at one of the local dealers. Besides the carbon roof (which is an option in Canada) I avoided most add-ons, possibly making it the cheapest M2 in the world at CAD$85,600 / £50,800 OTR. Perhaps the last new BMW model available with a manual transmission, the main reason I bought now.

[John Vine]

#21:  Absolutely.  I'm following this with more than a little interest. Best of luck with the repairs.

JV

[aerobod - near CYYC]

I removed the pistons and inlet valves today. Cylinder #1 gudgeon pin is bent, as there is a lot of resistance to pivoting the piston on the connecting rod, looks like to be safe I should replace the connecting rods.

More intriguing is the fact that every inlet valve spring is broken, all very close to 1.5 turns from the spring seat on the valve seal  top hat. After finding two like this, I wondered whether I was somehow breaking them when compressing for removal, but as you can see in the photos, there is a witness mark on the top hat that could only be caused if the spring was broken, this mark is on all 8 valve seal top hats. There seems to be either a tempering problem with the springs or a fatigue life that I have exceeded to cause them all to fail in the same place.

I wonder if this is an endemic problem with other cars having undiscovered broken springs?. Without the second break near the retainer in cylinder #1 "B" valve spring, the engine probably would still be operating fairly normally, as the break at the valve seal top hat end just causes stacked coils and a spring about 5mm shorter than spec, due to the top hat keeping the coils centred.

Using a rule of thumb for 10x wear for hard track use where the rev limiter is being reached every 500 metres or so compared with average road use, my car has about 10,000 track kilometres and 30,000 road kilometres, for a wear life of about 130,000km / 80,000 miles.

It will be interesting to see if there are any broken exhaust valve springs, hopefully I'll have time to remove those tomorrow.

[Shortshift]

Remind me, please, James - what was the build date of your engine?

James

[aerobod - near CYYC]

My car is a 2012 model, James. The cylinder head casting is marked as cast in Dec 2010 and the cam cover has a 17 Mar 2011 engine production date on the sticker.

[simon_h]

The springs from Kent had a spate of breaking from that sort of year. We used Cosworth springs on the 485 and I think that went across the board although that may have changed since of course

[mic]

Cosworth are no longer supplying the springs and have not for some time, not to Caterham. As Simon said there was quite a few that broke in the earlier engines. Kent parallel springs and caps were also used for a short period of time.

[aerobod - near CYYC]

I'm thinking Piper springs and cams as a kit (BP300 profile), as recommended by 7wotw / Neil, due to the Cosworth conical springs not being available anymore.

All exhaust valve springs are in perfect condition visually, compared with every inlet valve spring being broken. I expect inlet valve train resonance due to the different valve weights compared with the exhaust ones, with higher forces leading to fatigue failure of the springs.

Have the block, head and flywheel loaded in the SUV to take to a local machine shop for assessment tomorrow. I was going to go this afternoon, but with 30C here today (much like in the UK), there are storm cells heading this way and the hail seeding planes are circling, so best not to risk hail damage to the vehicle by going out.

If the bore scratching is too much for a 0.5mm overbore, the liner looks as though it can take 1.0mm to 88.5mm without significant loss in strength, Supertech, Wossner and other do pistons in this size, but I'm not sure what compression ratio I will go with yet, was thinking around 11.8:1, but that CR seems to be more rare than high tens or mid twelves.

[aerobod - near CYYC]

Abnormally high forces due to coil resonance in the spring seems to be the failure mode and the reason the inlet valves with a different inertia reacted differently than the exhaust valves. As indicated in this (quite old) paper, the failure is normally 1 to 1.5 coil turns from the spring end: https://www.jstor.org/stable/44434016

I think this means that any use of these springs will lead to inevitable premature failure that is likely linked to the amount of high rev use. One thing I noticed when checking the valve clearances before disassembly was the ease that an oversized feeler gauge went in between some of the valve tappets and the cam, but didn't link this to the inlet valves at the time. You can't directly see the broken valve springs without cam removal, even then the valve needs removing to confirm. I think a preventative maintenance check could be done by removing the valve cover and measuring the force to depress the tappet on a given valve.

[7 wonders of the world]

Are these definitely Cosworth springs and not OE Ford versions which are also Beehive versions...? 

Had you been experiencing any high rpm slight missing - as with these you were well in the realms of valve bounce......

I know Cosworth offer 2 spring options but i thought i had data on them but cant find it, be interesting to compare seat pressure and poundage,

Clearly these were not capable of withstanding your sustained driving regime   

Though to be fair race engines would normaly see new springs every season certainly assuming an engine lasted that long between builds.

CR figures on the pistons normally assume a 42cc chamber (stock is circa 44cc) - please check with individual manufacturers before ordering, I would aim for top side of 12.5 certainly to see the true potential.

Also some come with specific rings and other need OE rings - extra cost I think I gave you the Supertech part number for standard bore...?

[aerobod - near CYYC]

I don't think my springs are the Cosworth ones, Neil, more likely Kent I think, but I can't see any markings on them that indicate manufacturer.

No high rev misfire at all to the redline, but going by the top-hat valve seal witness marks the springs had been broken for a while. I would say they had half the tension when the valves were closed with one coil collapsed, compared with an intact spring. My knock sensor would typically pick up a misfire (as it did immediately when the #1 cylinder "B" valve broke into 3 and lost all closing tension), but had no warning whatsoever of a problem. The car was pulling cleanly to the redline and matching in the normal way in acceleration to friends' 911 996 Turbo and Exige 260 right before the failure.

I'm thinking now of a build that is close to an R500 build (as it seems safer to do the forged rods and pistons) but a bit lower redline (say 8250RPM) and power output in the 240 to 250 range to try to give at least 10,000 track kilometres before a rebuild is necessary. Hopefully the Piper cam and spring kit with the BP300 profile can handle it, but a slightly lower lift may be a bit safer.

[22daz]

If you're revving to 8250 I would consider the R500 timing chain tensioner modification if you haven't already. Removing the ratchet and replacing with a bolt to adjust tension more accurately and prevent backlash (timing misalignment) when coming off the power at high revs. It still uses oil pressure to tension the chain but the 'stop' is more finely adjustable. 
22daz

[aerobod - near CYYC]

I was planning to do that mod as part of the rebuild, thanks for the confirmation of it's value.

[cobar]

Regarding the mechanical tensioner, be aware of a warning Ammo Castellani, who started Raceco, shared with me several years ago.  New Duratec timing chains stretch.  Therefore, if one is used, a mechanical tensioner will require a re-tension which requires removing the engine cover -- i.e. major job. For this reason, he always installed a used chain when swapping to the mechanical setup.  Obviously this won't matter if you are keeping your current chain, but wanted to share the info just in case you end up with a new chain.

-John

[aerobod - near CYYC]

Thanks for that info John, I inspected the current chain and guides, they all look in excellent condition, so I will reuse. I'm thinking of modifying by putting an aluminium spacer over the ratcheted slipper shaft, as opposed to a clamping screw, to restrict any movement of the ratchet.

[Alan Reeves]

surely the £2700 420 engine for sale on here is the way to go even with shipping and taxes ?

good luck

[cobar]

The spacer should work.  If it's helpful, here are Ammo's instructions for setting the chain tension correctly which will determine the length of your spacer:

"Attached is a drawing that explains how to tension the chain. You basically fit it without the spring and plunger first and adjust the screw until the chain has 10mm movement between the sprockets. It must not be slack anywhere else. I press down hard with the heel of my hand on the chain between the sprockets so all, the slack is there. When the engine heats up you end up with 3mm movement which is about right."

Also, if interested, this post shows side-by-side photos of the stock and mechanical adjusters as well as other related information shared with me by Ammo, Raceline, and Esslinger:  

[aerobod - near CYYC]

Re #41 - Hi Alan, the £2700 engine with shipping, taxes and duty will be about a £4,000 engine when it arrives in Calgary and still may have what seems to be an inherent inlet valve spring resonance fatigue failure problem, that my engine died from. My current engine rebuild plan will be about £5,000 to give significantly upgraded parts besides the repairs, which I would probably want to do on the now £4,000 420R engine, likely costing £3,000, making it a £7,000 engine once I'm finished.

[aerobod - near CYYC]

Well, news on the head is not good, machine shop says the damage around and in the #1 cylinder "B" inlet port would be very difficult to do a proper repair on, with the narrow land to the "A" inlet port damaged and the seats mangled. They say the labour involved to do a delicate build up of the metal before machining new areas to fit new seats is something they couldn't do for me economically.

Block seems good for repair, though, but still don't know if 88.0mm pistons or 88.5mm ones will be needed until they do a full bore measurement, now I have asked them to continue with block refurbishment.

[aerobod - near CYYC]

Just ordered a 2.3 3S4G casting high port head off a Ford Ranger from Clearwater Cylinder Head in Florida. They seem to be the best source of refurbished heads in North America and should solve my original head damage problem, guaranteeing a minimal skim (typically 0.003", max 0.005") to ensure timing cover height match and a 3-angle valve grind with Ford valves. Just have to throw away the cams, springs, etc and do a port matching fettle.

[John Vine]

Thanks for these fascinating updates, James. Do keep them coming!

Do you plan to have any other port work done, such as gas-flowing?

JV

[aerobod - near CYYC]

Hi John, I don't think I will have any other work done such as gas flowing other than port matching on the head, my understanding is that my 240 to 250bhp target would not benefit significantly from it, for the $2000 or so additional cost. Also, smoothing out the slightly rough surface of the 3S4G head casting inlet ports may lead to less even mix of the intake charge, to the detriment of hot spots in the combustion chamber. I will polish the exhaust ports, though.

[aerobod - near CYYC]

It is a bit of a waiting game at the moment. The reconditioned cylinder head left Florida this morning, FedEx ETA 10 July here. Will wait to see if it is correct and meets my needs before ordering additional top end parts such as the springs and cams. The engine machinists confirmed 88.0mm should be OK for a rebore, so have now got the feelers out on availability of Wossner and Supertech 88.0mm 12.5:1 pistons with a local supplier here in Calgary

[aerobod - near CYYC]

Reconditioned cylinder head arrived today, looks to have been nicely done, especially for the USD$620 it cost with core charge and shipping from Florida. Casting number is the all important RF 3S4G high port version that seem to be difficult to obtain in general. Pity I have to discard the stock cams and valve springs it comes with. It is a 2005 casting compared with my original 2010 one, but the surface finish is close to new, especially compared with some of the manky and corroded ones I've seen for sale at higher prices.

 

Now if only the piston suppliers I've contacted would give me answers to my questions, I could get them ordered and the block rebore underway, together with the other parts ordered. Picking up the phone instead of it going to voicemail or answering emails seems to be hard for many suppliers these days.

[7 wonders of the world]

Glad to see it arrived safely and in good order too  

There always seems to be a market to recover so of your expenditure, albeit a small amount, for the stock cams etc on fleabay