Opinions On These Pics?

[Normski 1]

I put a post up a few weeks back on this subject. I now have the engine apart and have taken some pictures of the damage for the engine experts among you to help diagnos the cause.

 

I would be grateful for your input.

 

Details are as followed, 1.9 Mi16 motor used only for trackdays, standard rev limit, not over reved, valves replaced 15ish trackdays before this happened (~1000miles).

 

 

 

 

 

The broken valve is in 3 bits (valve is an AE v91469). It seems to me that the valve failed just below the collet allowing the rest to fall and hit the piston (although there is not alot of damage on the piston surface). As I said in my other post this has happened to me before (with a different head and valves).

 

Thanks.

[ALEX 98 1 Cars]

A sticking valve would cause it.

Has the head been skimmed much?

[Normski 1]

Not by me. I checked the head height and it was 131.5mm where a standard head should be 132mm.

Is that right for a skimmed head?

Can any more be taken off if needed?

 

This seems to me almost exactly the same as the other valve failure that I had with another head. Both times the valves failed just under the collet grooves like in the above picture.

 

Has anyone else had any thing similar?

[B1ack_Mi16 67]

Seem like the friction weld has failed.

 

Collet groove area is much harder material than the valve-stem itself, and is friction-welded onto the stem.

 

EDIT: Didn't realise it broke at the lowest groove.. then I really can't see why it would break there. Maybe collets had a little rough surface which made small scratched on the valve in the groove and led to a fatigue crack.

Edited October 25, 2006 by B1ack_Mi16

[Normski 1]

Yes, that makes sense. I'll renew them.

 

The collets as standard butt together allowing the valve to rotate (to help clean the seat area). I've read that the collets should be ground down slightly so they don't butt together and therefor hold the valve tighter. Anyone got any experience of this?

 

[PumaRacing 1]

Seem like the friction weld has failed.

 

That would be a bloody miracle seeing as inlet valves are one piece on this and nearly all other engines and don't have any friction welds in them.

 

Collet groove area is much harder material than the valve-stem itself, and is friction-welded onto the stem.

 

No it isn't. It's just heat treated in that region to make it harder.

[PumaRacing 1]

Yes, that makes sense. I'll renew them.

 

It makes about as much sense as the moon being made from green cheese seeing as the valve broke below where the collet touches it and in any case how much bigger of a 'crack' do you think you can put into a valve stem than the collet groove itself?

 

The collets as standard butt together allowing the valve to rotate (to help clean the seat area). I've read that the collets should be ground down slightly so they don't butt together and therefor hold the valve tighter. Anyone got any experience of this?

 

Yes lots. The triple groove collet problems are common to many engines including Ford (Pinto, CVH, Crossflow), BMW, VW and many others. The theory is that the collets not abutting helps the valve to rotate and the practice is that it's a steaming load of horse poo which the designers ought to have realised is pointless decades ago. Where it actually causes a problem is with 21/4N valves which can't be heat treated so the collet groove area is soft and starts to distort. On OE heat treated valves it makes little difference although I still see distortion (lipping of the groove edges) on some valves that haven't been made to the highest standards. The TU engine is very prone to it. I've linished the collets down on triple groove race engines in the past and that certainly stops the valve distortion problem but that isn't what made your valve break anyway.

 

As to what did make it break I'm afraid there are no definitive answers. AE valves are well respected and OE spec although I find their quality is not what it used to be. Perhaps the heat treatment was too excessive and made the valve brittle in that area. Maybe there was a flaw in the material under the surface. From your photo it looks like the 'beachmark' stress propagation mark starts about 1mm under the surface.

 

Coincidences happen. It could be that you've just been very unlucky on two separate heads.

[Normski 1]

So it's most likely it was a valve fault then.

 

What other brands would you recommend to be better quality?

 

Is it still worth grinding the collets down do you think?

 

Thanks for you input.

[petert 554]

I use custom made single groove valves to hopefully avoid such problems. They cost approx. AUS$50 each and use a Kawasaki Z900/1000 collet.

 

eg

http://forum.205gtidrivers.com/index.php?a...post&id=876

Edited October 26, 2006 by petert

[Rippthrough 98]

Just looks like a casting/hardening defect to me, propagated a crack across the boundary and off it goes, I'd give them a bell if they are that new and see what they say about replacements.

 

-Phillip

Edited October 26, 2006 by Rippthrough

[PumaRacing 1]

What other brands would you recommend to be better quality?

 

Other than OE Peugeot at god knows what price nothing else I can think of.

 

Is it still worth grinding the collets down do you think?

 

I doubt it.

[Normski 1]

Ok, thanks all of you for the input. I'll stick it back together (it's getting a new seat and guide fitted as well as a check over) and see how I get on. Surely it can't happen a third time to me.

[richsmells 2]

Longmans supply their own single piece valves I think? Probably of a good quality.

[Mattsav 1]

Look at the top of the valve tip (where the followers sits).

 

If the wear/polishing is only on one side then there's a good chance the valve springs is not seating properly.

 

This type of failure is very rare but 20v Audi's are prone to it (due to the springs and collets being a real bitch to fit) and I've recently seen a VW Diesel engine with the same problem.

 

On the diesel you could see where the bottom of the spring had been sitting on the casting in the head and you could plainly see where the top of the valve tip had only been in contact with the follower on one side (ie the valve tipping due to the spring pushing it over).

[newdean0 0]

Did you do anything to the valve seats on both heads which failed? i.e. get them re-cut. If the seats weren't cut central to the valve guides then the valve would fail.

 

It wouldn't explain the failure at the collet end but I guess if the remnants of the valve are being rammed back up into the port by the piston the remaining stem is going to break at its thinnest part.

[Normski 1]

Look at the top of the valve tip (where the followers sits).

 

If the wear/polishing is only on one side then there's a good chance the valve springs is not seating properly.

 

 

I've just had a look at this point and I can see a concentration of wear in a circle in the very centre. That's how is should be, yes?

 

 

newdean0, the first head no. The second one yes, but I don't think that was the case. I didn't see any of the side affects of poor valve sealing e.g. loss of preformance. And I'm confident in the quality of work that the engineer did for me.

 

Cheers.