Ducati head development: flow reversal in barrel

[superbike]

I'm working on Ducati 997 heads at the moment and have achieved higher flow with smaller cross section.
For interest I decided to block off one barrel of the inlet port (each cylinder has a 4 valve head). Left is 120cfm at 15mm valve lift, right is 80 but at 11ish it's 95cfm.

Why is it reversing above 11mm lift as everything is pretty close in each barrel?

Chris

[Guy Croft]

Chris, hi

I don't quite follow what you did. Please present the lift-flow regime you refer to graphically with notes/photos to indicate the problem regions.

Don't forget we have many novice readers who would like to get a 'handle' on what we more exp members talk about!

Then I'll try & tell you,

GC

[superbike]

Ok yes

I basically flowed each barrell seperatly by leaving one of the inlet valves shut one at a time.

The left barrell flowed more air the more i lifted the valve as normal to a maximum of 110cfm at 15mm lift

The right barrells flowed more air the more i lifted the valve to a peak of only 90cfm at 11mm lift. From this point on the figures went backwards to 70 cfm at 15mm lift. This seems to be a huge discrepancy.

[Guy Croft]

Chris, hi

Both barrels should flow the same but yes, that is an oddity you see sometimes. Obviously we are talking 2 inlet valves on the Ducati head but I have seen this 'blocking' on heads with one inlet valve as well, notably the Peugeot 205GTi where the short side radius is wrong.

The cause is turbulence between the throat and valve, blocking the airstream though I must confess the true aerodynamics baffle me a bit.

The cause of the turbulence may be growth of a vortex in a barrel - where high pressure (low velocity) air starts moving towards the low pressure (high vel) airstream in a spiral direction around the port, or it can simply be tubulent drag off some port region near the short side radius.

You might like to try what I do sometimes, blow air thur the inlet from outer face to throat with cotton traces attached to the outer face. One photo shows a head with two inlet valves (Lancia Integrale) where a very rapidly rotating clockwise vortex is building in the right hand barrel at quite low pressure ratio - at full 10" pressure the thing is too blurred to photograph. There was no vortex in the left barrel at all, and I figure some minute fault in the aerodynamic profile of the port wall or splitter is the point of propogation. Air is super-sensitive to even tiny deviations in profile - I know this from my days at Napier Turbochargers, where a huge amount of aerodynamic and production effort was expended perfecting the curved profile of volute inserts. Get an imperfection in a critical place and it causes disruption to the flow out of all proportion! Sometimes you can't even see the flaw with the naked eye.

The other photo shows a Fiat Punto 75 head where, despite the two traces being affixed quite separate (port roof and floor) they both head for the short side radius at all blow-thru pressure ratios and at 10" twist themselves into a know, a sure sign of a vortex.

[superbike]

I did kind of come to the same conclusion. I worked on the splitter and the tiny curve onto the valveseat a little more but everything has stayed the same.

I then decided to test the rear head that i also modified at the same time.
The result was exactly the same on the other head. Iam now on monday going to see what happens on a totally stock head.

Wondering if it could be something to do with my bench as it is a home made contraption ?

[Guy Croft]

Provided there is a plenum between blower and mounting plate and a tunnel of about 5" height and bore diameter to mount the head on you can be sure it's not the flow bench, see photo.

GC