Ducati downdraught port modification and test

[superbike]

Hi all

I'm wondering on a heavily down draught port will the air flow more along the top of the port than the bottom ?

[superbike]

Hi again

I ask as iam at the start of a job i have a head that is at 170 at 11.5mm lift with 181 bpf.

i will be fitting 2mm over size inlet valves so i guess i need to increase the bpf as the port will not flow enough air for the 2mm oversize valves at high lifts.

i usaully just make the shoots a little bigger all the way round but this time i would like try enlargement through the splitter or make the roof higher.

so i guess i need to establish where most of the flow travels ie roof or floor ?

thanks

[superbike]

Hi all

I have been working on the 4v head again today and i widened the ports at the splitter up to 2mm on each shoot. The ports are now wider than the origials and there is pretty much no gain in bpf.
Is this because i have removed metal from dead no flow zones ?

[Guy Croft]

Chris, hi

Not ignoring you, just busy. I need to know:

Dimensions:
- horiz and vert at port entry section or perhaps circumferential measurement if not semi rectangular.
- diameter or horiz, vert and diagonal thru splitter
- valve throat diameter.

Just for now I can tell you this: In a good downdraft port the airflow distribution is quite well evenly distributed all across the section thru the barrels. They behave totally differently from sidedraft units and the whole region from splitter to valve is super sensitive.

Making the splitter vane thinner can work but if in doing so you generate an intrusive 'aerofoil-shaped' hump on the vane along centreline or at top and bottom where it meets the barrel it can actually reduce the flow (not being very technical here because I'm in a hurry). Splitters can be a bit tricky.

It may well be that the port is trying to flow more air because you made the splitter region bigger but the short side radius profile is wrong, setting off turbulence that is obstructing the airstream - this is very common. Correct the ssr and the gains become evident but not until. I have come across this many times.

I could do with seeing one of these, either the real thing or a very good quality close up of throat and ssr region, but you'll need to take care with lighting and image quality or I won't be able to tell anything useful.

GC

[superbike]

Ok

The port entrance is is 59 by 32mm
reducing to two 33mm diameter holes tapering out to 35mm at the bottom of the seat tapering out with a 70 degree cut to feed 40mm inlet valves.

All the pictures are of a standard untouched head.

One other thing i found measuring the bpf with this standard untouched head is that once the flow bench has been warmed up the bpf can jump anyware between 175-185cfm where as the head i have modified reads a steady 182-185 cfm ?

Thankyou.

[Guy Croft]

Chris, hi

I deleted two pics because they didn't come out well on dial up connection but the remaining one tells me all I need, thanks.

The fluctuation you saw with the std head is either turbulence in the valve throat - usually the short side (such as it is on your heads) - or a vortex forming off the splitter. I don't know what your final splitter form is but probably you got rid of it.

Those dimensions are bigger than what I'm used to so I hav no data around me to say what the flow problem is. I'd hoped to do some tests for you today but I have not had time but this is what you need to to.

1. Make replicas of the valve throat (smallest diameter of seat insert, that being the control orifice at the valve for bpf tests), one splitter and the entry section from thin sheet metal or plastic.
2. Put each in turn on the flow bench and record what they flow.

When you test them it is important to put a clay (plasticine) radius round the entry section, this needs to be clay moulded into about an 8mm dia section, and then stuck to the orifice on the outboard isde to reduce the sharp edge entry loss - ut not right on the edge of the hole, set it back about 1.5mm. You don't need to test both splitters/valve throats, just multiply the reading by 2 and you'll then have the total flow in that section.

Another test worth doing is blocking off one barrel with clay (be careful or it will get sucked into the flowrig) and seeing if both barrels (I think you called them 'shoots'?) flow the same. They should if you have got both splitter sections the same shape and dimensions.

You will find one section that flows the least and that is the controlling section for the whole thing. I don't know which it is, just test and tell me and I remark on the results. You will realise that just because, say, the mock-up valve throats flow 190 cfm it doesn't mean the port as a whole will. Conversely 200cfm coming in at the port entry section will be a lot less by the time it reaches the valve due to viscous losses. All we're trying to do is find the 'logjam' that is stopping you going over the 185cfm mark.

I could do with a picture of your finished splitter viewed from port entry if it's not confidential. If you want to keep it under wraps e mail it to me.

It may not influence the bpf figures but could well yield a gain with valve in - I would be using a 75 deg throat cut on those seats not 70. I have ever found anything else any good on downdraft heads.

GC

[superbike]

Really good help there Guy.

I have got somewhere now.

The smallest part of the seat being 35mm i opened up to 37mm with a paralell cut then blended it into the port (only doing this type of cut as later a 75/45/30 will be added when i install the larger valves.

Anyway i tested again and 200cfm was the result.

To make sure of my work i then did the same to the totally stock head and no gain followed staying steady at 180 with no turbulance

I guess the smallest part of the seat was the restriction even though it was not the smallest part of the port . Thats weird

[Guy Croft]

well done!

The loss thru the seat is due to the change in airflow direction, and the more acute the angle between the valve stem and port axis the better the flow - this is one thing that makes F1 engines so powerful.
It is usually the case with porting that due to losses from drag, turbulence and change of direction or section, that you have to make some regions much bigger than they would need to be if they weren't part of the port system taken as a whole.

GC

[Guy Croft]

Just to illustrate what I'm talking about, here I'm flow testing a section of Vauxhall XE port opened out to the absolute extreme limit, I mean within 1mm of breaking thru (not that I work that close to the limit on client property), to see that the outboard port section can flow.

Yes, looks crude, stuck together very hurriedly on the rig orifice plate with Blutack, sure, but it doesn't matter. You're in a rush to find something out that's holding you back in the search for more flow and there is no time for niceties! I've even flowed out heads, cut them up to find the metal thickness, and then glued them back together after modification to test them.

Cannot remember the exact amount at 50mm wide by about 26mm high but it was far, far more than you'd ever get from an actual (complete) port with valve throats and splitters. But at least it told me what I wanted to know which is that the logjam was the splitter region in that case. And yes where I can I cut up heads and test those regions too.

It's vital to keep things like that handy - and properly organised data files and sketches too - because when you do a lot of different heads and then go back to an old faithful, and when you are flowing 'at extreme limit' as I mostly do these days, it's often impossible to remember what you did, and if you guess at it - you're going to break thru somewhere..

GC

[superbike]

Ok

I have managed to get past the log jam at 180bpf like previously discussed. I have now cut 75 x 45 x 1mm x 30 to fit the 2mm oversize valve and I am now at 206 cfm bpf - was 180.

The thing is now at max lift I'm at 182 with 42mm valve in and oversized barrels 35mm wide opposed to 33mm round standard.

My figures so far:

stock 33mm barrels on 40mm valve and 35mm x 33 barrels on 42 valve

Lift (mm) vs flow (cfm)with 40/42mm valve
2 - was 47 now 50 (+3)
3 - was 66 now 69 (+3)
4 - was 87 now 90 (+3)
5 - was 108 now 110 (+2)
6 - was 127 now 128 (+1)
7 - was 139 now 145 (+6)
8 - was 150 now 157 (+7)
9 - was 158 now 167 (+9)
10 - was 164 now 176 (+12)
11 - was 169 now 181 (+12)

bpf
40mm valve 179
42mm valve 200



The thing that concerns me is the standard head was only 10cfm away from bpf at 11mm lift the modified head is over 20cfm away from bpf and I expected more from a 2mm oversize valve.
My last go at this conversion yielded a little more than this 5cfm more from 6mm upwards I was hoping to better it seeing as I been spending all my spare time on it.

Any advice welcome

[Guy Croft]

Chris, hi

If those figures are at 10" the flows are massive. I am slightly concerned about port velocity on the engine. I need to know what depression you are measuring please.

Yes a 20 cfm loss between bpf and flow with valve at 11m lift looks poor but it may not be the whole story. A couple of pointers:

1*. You may need to lift the bigger valve still higher (immutable fact - the bigger the valve the worse its discharge coefficient, ie: the factor by which flow past valve is reduced). Lift the 42mm valve to 12,13, 14, 15mm and see what happens to the flow. It should go up the more you lift.
2*. If the flow drops between 11mm and say 14mm (by 14mm the valve should represent no intrusion to the airstream at all I imagine, though I don't have any exp of the particular head of course!) you have got some port/throat turbulence effect upsetting the flow, common on Peugeot 205 GTi heads and very hard to get rid of, you risk losing all your gains so far.
3. Poor valve performance can be the result of shrouding or poor exit flow, if there is no shrouding, the top grind on the valve seat may need to be changed. I have seen a marked difference on some heads between 15 - 20 - 30 - 35 deg top grinds of varying width and depending whether the grind is smoothed out or left sharp. That aspect is hugely difficult to optimise.

*See videos.

Hope that helps,

GC

[superbike]

Yes the flows are at 10inch

I cut the 2mmover size inlet valves into the same depth as the originals(i can even use the same shims )

I know this post is going on a bit but i would like to show you something.
Its a standard head on the left (same as im working on) and some mystery work on the right . Can you spot the difference they look wider and have something weilded in before the non existant short side.

[Guy Croft]

Chris, hi

10", OK. But before I comment again I'd like to know what happens on the higher lift tests with the big valve.

GC

[Guy Croft]

Chris,

If can you give me bore, stroke and rod length, I'll take a look at the Mach index on that head with 2 splitter barrels of roughly 34mm dia.

GC

[superbike]

Ok

stroke is 61.5
bore is 100mm
rod length is 124mm