Flow bench measurements

[motoguzznix]

A slightly damaged v11 head was lying around at our local guzzi dealer Pepo Seewald. Due to an accident some fins were cracked, but the head was otherwise in usable condition. Pepo ist an old tuning expert who runs his own Dyno in his shop.

So I went off with the head and had make some flow bench measurements from the intake port. The measurements were taken by a friend of mine who works at a local engineering company.

 

This are the results:

The vertical scale is the flow of the port, horizontal is the valve lift in mm (red figures)

 

As you can see, from 8 mm valve lift on the the flow does no more increase (blue line). Tha V11 camshaft lifts the valve to 10.5 mm. So the engine can not make use of the upper 2.5 mm of valve lift due to the bad breathing capabilitys. With a shorter lifting cam the valve train forces would be much lower contributing to increased valve train lifetime. Without any loss of power.

 

And vice versa, if the port could be modified to allow better high lift flow, a lot of engine power could be found.

 

For comparison the black line was the measurement of my LeMans 2 head. From approx. 8.7 mm valve lift onwards, the LM2 port flows better than the V11s, even though the LM2 sports 2.5 mm smaller intake valves.

 

I was somewhat surprised by the results, and I'm looking forward to modify this head to get better flow and consequently to modify the heads of my own V11.

[Guzzirider]

Thanks for posting on this subject which has not been covered too much on this forum. Makes a change to have something new to discuss- most subjects have now been done to death and the repetition gets a bit dreary after a while.

 

Please keep us updated- particularly interested in seeing the results from flowing your heads.

 

Cheers

 

Guy

[callison]

I have no clue where my chart is now, but when I sent my Sport 1100i heads off to Mike Rich, the stock flow was about 105cfm. After the Stage II porting, they were nearer 127cfm. Thats about 20% more flow.

[Skeeve]

I have no clue where my chart is now, but when I sent my Sport 1100i heads off to Mike Rich, the stock flow was about 105cfm. After the Stage II porting, they were nearer 127cfm. Thats about 20% more flow.

 

That's all fine & dandy, but what did that 20% improvement on top do to the bottom-end & midrange? Was this a "they should all come this way from the factory!" improvement, or "Guzzi got it right for around town, but yeeeeHAW this things great WFO!" sort of change?

 

I'm just sayin'...

 

WRT to the flow bench measurements:

 

That certainly is revealing that after more than 20 years of development, Guzzi hasn't been able to move the numbers much. Of course, some of the "if it ain't broke, don't fix it" axiom may apply...

 

[motoguzznix]

Guzzi got it right for around town

Guzzi sure did not get it right neither for the road nor race track.

For the road, smaller valves would be the better choice, as port velocity would be higher thus improving bottom end power. It is always better getting the same flow from smaller ports than just increasing port diameter. The engine with smaller ports and identic flow than another with larger ports will have more power from bottom to top.

 

So I'm interested to find the bottleneck in the port and not to increase the ports size. In the meantime i went one step further and did a molding of the port:

You can see that at the outside radius of the port behind the valve seat insert - see arrow - could be the bottleneck. Otherwise the port shape seems to be fine to me.

 

If you look at the graph above, a flow increase from 0.58 to 0.65 at the maximum lift should be possible when the fault in the port shape is corrected. This would represent an improvement of 12 %. Theoretically.

 

The bike the head comes from was tested at Pepo Seewalds dyno and put' out 75 HP on the rear wheel. The graph was posted by myself in the dyno thread. It is the bike with the lowest power output in the graph: Power graph

[mike wilson]

You can see that at the outside radius of the port behind the valve seat insert - see arrow - could be the bottleneck. Otherwise the port shape seems to be fine to me.

 

Er, no. The port shape is friggin' awful. By definition. You are trying to get air to do something you want, not what it wants. Ergo, all port shapes are bad. Some are worse than others, though. Classic bad things about this: firstly (and probably most important) it is not straight. Much, if not most, of the problems associated with airflow are due to change of direction. Secondly, it changes diameter. Not only where you note but before the bend also, where the valve guide is. Thirdly, it is of a fixed length. This means that it will work best at a certain rpm. All others are compromises.

 

Nowadays, there are probably computer modelling programmes that will design a port according the the inputted parameters. Previously, it was suckitandsee (not forgetting to take notes) and trialanderror. Some rules of thumb came from that, along with knowledge about the behaviour of flowing gases in restricted spaces. Some of those rules will be what drives the computer programmes but I suspect there will still be situations that defy computerisation.

 

Conventional wisdom would decree that the port shown needs the bend straightening as much as possible, the two obvious diameter changes removing and some shaping to direct the air somewhat as it enters the combustion chamber. All this needs to be done with knowledge of the rest of the induction and exhaust system (more generally agreed wisdom is that it is equally necessary to make sure the gasses can get _out_ at least as fast as they get in) including the camshaft and desired use. The above would be for outright power. Note that the waisting just behind the valve seat is sometimes put there deliberately to speed up the air when the valve is just opening. This helps cylinder filling at a time when the airflow is low.

 

Some decades ago, Cycle magazine ran a feature on reworking the cylinder head of a SR500 Suzuki. They had carefully decided the desired outcome and the parameters to change, spent numerous hours with grinders and epoxy (it's not always about making holes bigger....) and flowbench work. They came out with a noticeable increase in HP, related to a distinct increase in flow at all valve openings. The suprising result was that, messing around with one of the lolly (popsicle) sticks they were using to apply epoxy, they were able to exceed their best flow figures by a considerable margin by just shoving it in the hole at random. They couldn't work out how to hold it there.......

 

All this is not to decry your efforts. FWIW, there may be something you could try before you start cutting metal. I would see about getting rid of that awful bulge on the outside of the bend. It looks like an arthritic knee. Some epoxy, filling that, would make the port slightly straighter and might make an unexpected difference to flow rate. Without seeing the location of valve and guide it's difficult to judge but that would be my starting point on the evidence so far.

[helicopterjim R.I.P.]

 

All this is not to decry your efforts. FWIW, there may be something you could try before you start cutting metal. I would see about getting rid of that awful bulge on the outside of the bend. It looks like an arthritic knee. Some epoxy, filling that, would make the port slightly straighter and might make an unexpected difference to flow rate. Without seeing the location of valve and guide it's difficult to judge but that would be my starting point on the evidence so far.

 

This is exactly what an old friend and renowned Ducati tuner learned with Ducati bevel engines. He found that adding material (and removing a bit in other places) dramatically increased the power output. This bike went on to win the BOTT title in 1988 I believe.

 

Unfortunately there are obvious limitations encountered from adding material so not often are the gains very large. His only other success with this method was with Formula Ford motors in the late 1980's. Perhaps I shall send him a Guzzi head if I find one and see what he can uncover!!

[Paul Minnaert]

to hold the same diameter, should't the thing be wider where the valveguide is going through? I one looked in a 916 head, that has much more logical looking ports

[callison]

That's all fine & dandy, but what did that 20% improvement on top do to the bottom-end & midrange? Was this a "they should all come this way from the factory!" improvement, or "Guzzi got it right for around town, but yeeeeHAW this things great WFO!" sort of change?

 

I'm just sayin'...

 

WRT to the flow bench measurements:

 

That certainly is revealing that after more than 20 years of development, Guzzi hasn't been able to move the numbers much. Of course, some of the "if it ain't broke, don't fix it" axiom may apply...

 

 

I was trying to build an engine with a true 90-100 HP at the rear wheel. I wasn't necessarily concerned with the bottom end but Mike Rich's Stage II porting while hot, is not an unstreetable configuration. Stage III though, is what the Bonneville guys used and there isn't any sane way to run that kind of porting on the street. FWIW, I've only put about 220 miles on the Mike Rich heads as there are other issues with the bike but the bottom end seens untouched. The Norge sort of negated my desire to build a better Guzzi engine. Anybody want to buy some heads?

[Skeeve]

...So I'm interested to find the bottleneck in the port and not to increase the ports size. In the meantime i went one step further and did a molding of the port:

 

What did you use to make the mold? Just curious if we have it widely available here in the States...

 

[Skeeve]

to hold the same diameter, should't the thing be wider where the valveguide is going through? I one looked in a 916 head, that has much more logical looking ports

 

That was what I was going to point out WRT the "bottleneck" comment above. Actually, the "bottleneck" isn't at the neck just above the valve head indicated by the arrow in the pic of the intake mold, it's at the similar point just before the "bowl" where all the valve guide paraphenalia clutters up the port.

 

Ports should narrow gradually throughout their length to the point just below the valve head [as indicated by the arrow: this is exactly where the narrowest part of the intake port should be!]; that choke point just before the intake gets to the "bowl" needs to go, but only on the outside curve of the tract; the inside curve should certainly not be made any tighter! Smooth, gradual changes are what air likes...

[motoguzznix]

What did you use to make the mold? Just curious if we have it widely available here in the States...

I used a molding silicone consisting of 2 components called "Silicolene". It was obtainable in a shop for art articles.

[motoguzznix]

I would see about getting rid of that awful bulge on the outside of the bend. It looks like an arthritic knee. Some epoxy, filling that, would make the port slightly straighter and might make an unexpected difference to flow rate. Without seeing the location of valve and guide it's difficult to judge but that would be my starting point on the evidence so far.

Mike, thank you for your efffort.

I agree that the bottleneck might be in that area.

The reason for this is the fact that the intake port is out of center with the valve seat. The valve seat seems to be displaced to the short side radius of the port. Thus the 75° angle of the valve seat is positioned ideally at the short side radius and the bottleneck occurs at the outside of the port. The air stream is guided to the valve disk instead of to the gap where the gasses can pass freely.

[motoguzznix]

to hold the same diameter, should't the thing be wider where the valveguide is going through? I one looked in a 916 head, that has much more logical looking ports

Paul

 

The port ist in fact slightly wider in the area of the valve guide, which is not visible in the picture above. Unfortunately I took no photo from that direction, as that looked ok to me. Here is another picture of the port:

[Hoodrewoodre]

Paul

 

The port ist in fact slightly wider in the area of the valve guide, which is not visible in the picture above. Unfortunately I took no photo from that direction, as that looked ok to me. Here is another picture of the port:

 

 

Hello ernst,

sorry as i have not let hear of me within a very long time, if its up to me your port mold doesnt look that bad to me, even though i think it could be improved with either lots of special sense (meaning black art) or by means of a pitot/prandtl tube velocity tester adn perhaps as mentioned by means of a nice tube of devcon(even though there isnt too much material underneath the valve spring seat.

Too bad that the numbers by your friend (guess the one in the huge technology company AV...) dont mention the CFm numbers cuz then one could do comparisons, whereas numbers by that company allways seem to me a lil kryptic on purpose (maybe some of the marketing attitude of AV...)?.

For my guzzi i purchased a couple months ago some old LM1 heads which i plan on modifying majorly by means of valve reangleing, MC-x10 cam, 50mm valves, hi lift rockers etc. (lots and lots of work yuk brrrr)

Knowing also that article (htink by g. jennings?) bout the SR500 head which has been ported by J.branch one also has to mention that quite some water has ran down the river nile since then and also some new informations have been retrieved, nevertheless (after having ported (and also welded) quite a meaningfull and sucessfull number of Sr-heads mostly for races) i would think with the right camshaft, portlayout ,and intake and exhaust length, etc. numbers in excess of 100 horses should be possible. (thus not quoting mr. hofmanns of dynotec phantastic ideas of 119 horses as stated in the last Mo article (which made me quite a lil angry, for its unrealistic optimism), especially after having seen quite some (and different types) heads in the last almost 10 years of my operation with a flowbench.

If you ask me quickly over the counter i would make the SSR width sensefully as wide as possible dig a lil deeper behind the valve guide and try to gain the high airspeed in front of the valve guide by means of a pseudo velocity nozzle (laval style venturi).

everything else requires tons of work and time on a flow bench and a "proven dyno" which i both dont have of yet here since they are still in vienna.

Just my quick 10 cents...;-)

 

Sorry to all for not introducing myself (but im in severe lack of time currently), ciao a tutti da italia

 

Christian

 

PS: hope my rantings sounded a lil sensefull ;-)