Valve seat angles and effects

[Chris S]

Guys hi

I was just thinking about what a good discussion would be . Would anyone like to comment on there experiences of seat angles and the effects produced by angle changes .

Regards

Chris s

[Guy Croft]

Chris,

Before I get engaged here - have you read the articles in GC Q&A (head prep) 'all the power is in the head' and 'influence of valve shape', 'Ford CVH prep'
and 'comparative inlet flows, 8V Fiat TC in 'results and feedback' and others?

There is a lot written on this in those features.

G

[RedLexus]

My local engine tuner has led me to believe that proper 3 angle seats on modern 16 valve heads are possibly more important than porting! He does a lot of Suzuki 1400cc and 1600/2000 Vauxhall XE engines for Tarmac rallying . Would you agree with that Guy?

[Guy Croft]

The short answer is you just can't make those hard and fast rules with heads in terms of whether port/manifold mods or valve seats/valves will yield more benefit. The valve region, however well optimised - will only flow what's fed to it by the port. So, in short, it depends where the controlling section is.

I want to point out that the massflow (be it 10", 25" at test depression or on-engine realtime pressure ratio - which is far higher) thru the valve region is nowhere near what it would be if there were no losses in the adjacent port region. I'm talking about the effect of the turn (that exists in all production heads) that the airstream has to negotiate from port axis to valve throat axis. With a pretty basic design eg: sidedraft with narrow valve angle and deep-drop from port floor to valve region, the flow loss going round that 90 deg turn can be 50% or more of what the parallel outer port section would flow on its own. It's easy to evaluate the various losses by sectioning a head and flow testing those regions. And of course the losses are all due to viscous effects (drag at the port surfaces) and - primarily - turbulence. So the valve's behaviour is governed to a great extent by its surrounding architecture. The real-world or 'practical' flow potential of a valve setup - during head development - means the valve region 'taken as a whole'.

You can encounter heads where two totally different flow regimes exist.

The 20V Fiat Coupe head is one such. At low valve lift the air flows stably round the short side radius and the distribution around the valve head is excellent - as it should be for a well-downdrafted head with a commensurate valve angle. However, owing to the height of the the port floor from the valve there is a rather adverse profile on the short side radius which you just can't get rid of by modification; you'd have to lower the port floor by about 4mm. The consequence of that is that at 6mm lift the turbulence from separation around the curve of the SSR becomes so extreme that it lifts the flow completely away from the SSR and directs across to the furthest regions of the valve. That problem doesn't exist on the 16v Vauxhall XE series, the 8v Lotus Twincam, the 16v Sierra Cosworth and a few notable others. So, considering that, you can see that the gain you're going to get from your very best efforts at valve-seat work depends entirely on how that particular head is laid out.

Now, on a head with very big ports, like the Honda VTEC featured in GC Virtual Workshop, the outer port region is huge and the controlling section that governs what goes into the cylinder is definitely the valves and a yes, a fair proportion of flow gain (with valve in) comes from seat work. And of course there is a limit to how big you can make the inlet valves and that is usually set by the diameter of the cylinder. And on that one, you're definitely not going to get more flow (as if you'd need it) from making the port bigger.

Whereas - on the 2L Vauxhall XE you can alter the valve region for given valve size as extensively as you like - and even fit much bigger valves - but the controlling section is the splitter barrel and it will consistently underflow (even at borderline break-thru diameter) what the 33/34/35mm valves could achieve if tested as an entity with no port attaching.



GC