Head flow testing.

Status
Not open for further replies.
As an added note, Dave Nourish at one time supplied his engines with aluminum rods but no longer does for some strange reason....hmmmm; makes me think.
Click to expand...
As mentioned above I'm not quite sure why conrods are getting a bunch of press in a headflow post but here goes anyway.
I discussed conrods with Dave Nourish since I've run one of his 750s for quite a while. The only reason he stopped using/making his own aluminum rods was because he couldn't get the rod bolts any more. His comment was that he preferred to use the aluminium rods since the little end was generally lighter than the steel ones, even if the overall weight wasn't much different. My conversation with him was sparked by the broken rod shown in the picture. Its out of a 750 engine (70 -75 ish rwhp and a fair bit of detonation ) and it broke after 6 seasons. When asked, Dave said that yes aluminium rods do fatique after a while and he recommended changing them at regular intervals. What that interval was he couldn't really say. I figured that if mine lasted 6 years then I'd change them every 3 to be on the safe side. Given that he used the same rods in his 950 cc 95 hp engines the strength of the rod wasn't really an issue - rather it was just its fatique life. I carried on using the aluminum rods, changing them out evry 2 -3 years until he stopped making them and I switched to Carillos . The 2 rods on the left are slightly different versions of the ones he supplied, the Carillo on the right is out of a shortstroke engine that had a seizure after the oil feed line broke and damaged the little end bush ( and the piston and the bore)
Head flow testing.


Its a while since I've looked at a standard Norton rod But i believe the Nourish ones are herkier. Hence I'd believe Steve Maney when he says its not a good idea to use Norton alu rods in his 1007s but I would think a suitably sized alu rod would work just fine (but with the understanding that it would have a limited life)

Nourish makes a range of engines sizes from 500 to 950, all using the same components. An intersting question for the maths whizzes out there would be which rod sees more relative stress, the one in the 500c 60 hp shortstroke engine turning 10,000 rpm or the one in the 950cc 95 hp longstoke turning 7000rpm. Strokes are 58.75mm and 93.5 resp
 
TRIUMPH :p Lifed the Thruxton Bonneville race rods ( std Tri 650 rods ? ) at SIXTY race HOURS . About a season for them .

One needs to avoid the ' vicious circle ' where the rods heavier , so the piston needs to be stronger , so the rod needs to be heavier ,

Head flow testing.
 
Jim,

Can you provide some general dimensions on this head (ex intake valve dia, seat dia., throat dia etc)? Curious minds want to know. A side by side picture with say and RH4 would be interesting.

Care to comment on why the Steve Maney Stage 3 head flow plateaus whereas the subject head keeps on giving (dirty filthy pun in there somewhere).
 
Ok, after spending some time playing with the new airflow portmapping setup I found that there was no way of making the information it gives comparable to the airflow curves I had been posting. It is a really neat setup that takes readings at 9 places inside the port and paints a picture showing where the air is moving fast and where it has turbulent or slow flow. I will definitely be using it more.

I did a new test on DWS's head with the old system and posted it along with the graph from the Maney head. As you can see the airflow is quite slow due to the extremely large ports. The flow is slow enough that turbulence is not a factor in this port.

The valve sizes are 44mm intake and 37mm exhaust. The valve seat is radically bell shaped from the edge of the valve to a bowl that is nearly stock sized.

Head flow testing.
 
So what is your take on the break point of the cyan curve at 0.350 lift.
Is turbulence coming into play? It seems rather abrupt.
Could the difference be in the absence of the conventional guide protruding in the port?
 
Dances with Shrapnel said:
So what is your take on the break point of the cyan curve at 0.350 lift.
Is turbulence coming into play? It seems rather abrupt.
Could the difference be in the absence of the conventional guide protruding in the port?
Click to expand...

The break in the flow curve is caused by turbulence. When the airflow reaches a critical speed it begins to lift from the port floor on the short side radius and the rough air plugs that part of the port. It is pretty common especially if the valve has been re-angled. There is not enough metal in a stock head to get a proper short side radius and it gets worse when the valve is re-angled. The airflow also gets very noisy at that point.

If the port is widened where the guide is, then the airflow slows and the guide creates very little turbulence. I have seen very little change in flow with or without a 1/2 inch or smaller guide. The larger guides in the 850s do cause some flow problems.
 
comnoz said:
The valve sizes are 44mm intake and 37mm exhaust. The valve seat is radically bell shaped from the edge of the valve to a bowl that is nearly stock sized.
Click to expand...

Sounds like Bernoulli's Principle at work here. A goal of intake port design is to get the velocity (kinetic energy) converted to pressure (potential energy) as efficiently as possible. This is what comnoz has been a proponent of for quite some time. So if you have high velocity and your port is efficient you get high pressure for greater combustion chamber filling.

http://www.princeton.edu/~asmits/Bicycl ... oulli.html

This conversion from velocity (kinetic energy) to pressure (potential energy) occurs where the port throat begins to dialate through the valve seat and where the flow is entering into the combustion chamber.

The other pitfall or constraint is the practical speed limit (often referred to in engine tuning books as Mach number or Mach index). As I understand it, at sonic velocity, you need substantially more energy to get higher velocity. This is a bit of a wall in normally aspirated engines. So it is a balancing act between velocity, "hitting the wall" and the characteristics you want out of an engine.

Although a simplistic view, consider the Norton piston on intake stroke at peak velocity (not mean piston speed but badass peak piston speed) at say 5,000 rpm and compare its cross sectional area (air demand) to that of the cross sectional area of the port and you will wonder how the engine breaths at all at speed. There are lots of neat things happening at speed that make this work rather well and even allow up to around 120% cylinder volume filling (this occurs at peak torque, not peak horse power)
 
lcrken said:
but with the Norton's long, flexy, camshaft and small lobe size, I can't use something like a Chevy small block racing profile. I'm pretty limited to just going with more duration and more total lift, and a little bit more acceleration, and that's pretty much what the N480 has. In a perfect world I could put in a center cam bearing (sound familiar?) and make room for larger lobes.
Ken
Click to expand...

If you look at a Norton rocker arm, you will see that it has enough material to move the pushrod ball inwards enough to increase the rocker arm ratio up to about 1.3:1. This would let you run a camshaft with much less lift at the lobe, and it could be ground on a larger base circle making the cam stiffer, plus it would all fit into the standard cam tunnel.

The high rocker arm ratio would give high acceleration to the valves that would not have to be ground into the lobes anymore, you would need a much less aggressive lobe.

Also you could run lighter springs because now without the almost 1:1 ratio of standard Norton rocker arms, the only inertia the valve spring would have to control is that of the valve and retainer because the increased leverage the high rocker ratio had over the weight of the lifter and pushrod would make it's inertia moot. PLUS the lifter and pushrod would now be moving only .25-.3" much less than on old Norton profiles and have much less inertia anyway.

Smokey Yunick tested this all on the small-block Chevrolet and with it's 1.5:1 rocker ratio said that the weight of the lifter and pushrod absolutely did not matter at all as far as it's effect on valve float went, all that made a difference was the weight of the valve and retainer.

I have a drawing and math figures done by the late Heinz Kegler on converting the Norton rockers to 1.3:1 ratio.

The stock Norton rocker ratio is a joke for high performance and time spent on working with it and around it is not well spent.

Maybe one of the sharp cookies here with machine tools can TiG up the old rocker ball holes and re-drill them inwards and have someone grind a cam with about a 1" base circle and a chevy lobe to use with them. It should flex less, allow better timing, valve acceleration and rpm before float.
 
beng said:
lcrken said:
but with the Norton's long, flexy, camshaft and small lobe size, I can't use something like a Chevy small block racing profile. I'm pretty limited to just going with more duration and more total lift, and a little bit more acceleration, and that's pretty much what the N480 has. In a perfect world I could put in a center cam bearing (sound familiar?) and make room for larger lobes.
Ken
Click to expand...

If you look at a Norton rocker arm, you will see that it has enough material to move the pushrod ball inwards enough to increase the rocker arm ratio up to about 1.3:1. This would let you run a camshaft with much less lift at the lobe, and it could be ground on a larger base circle making the cam stiffer, plus it would all fit into the standard cam tunnel.

The high rocker arm ratio would give high acceleration to the valves that would not have to be ground into the lobes anymore, you would need a much less aggressive lobe.

Also you could run lighter springs because now without the almost 1:1 ratio of standard Norton rocker arms, the only inertia the valve spring would have to control is that of the valve and retainer because the increased leverage the high rocker ratio had over the weight of the lifter and pushrod would make it's inertia moot. PLUS the lifter and pushrod would now be moving only .25-.3" much less than on old Norton profiles and have much less inertia anyway.

Smokey Yunick tested this all on the small-block Chevrolet and with it's 1.5:1 rocker ratio said that the weight of the lifter and pushrod absolutely did not matter at all as far as it's effect on valve float went, all that made a difference was the weight of the valve and retainer.

I have a drawing and math figures done by the late Heinz Kegler on converting the Norton rockers to 1.3:1 ratio.

The stock Norton rocker ratio is a joke for high performance and time spent on working with it and around it is not well spent.

Maybe one of the sharp cookies here with machine tools can TiG up the old rocker ball holes and re-drill them inwards and have someone grind a cam with about a 1" base circle and a chevy lobe to use with them. It should flex less, allow better timing, valve acceleration and rpm before float.
Click to expand...


At the risk of going too off topic it would not take too much to increase the rocker ratio. Cam loading is not a moot point as now the cam has less of a mechanical advantage (leverage) to accelerate the valve mass (regardless of the lighter spring mass). With the 500 ultra short stroke we needed to put a center bearing support on the cam due to the extreme inertial loads and greater spring loads required.

Keep in mind that with generic V8 engines and wedge heads you cam really slam the valves open as there is really nothing in the way whereas with our lovely Nortons there is valve nip so there are other considerations. I am not saying greater acceleration is not desireable but other things need to be considered. We went through a whole valve train design excersize using 4StrokeHead with the assistance of the late Professor Gordon Blair to get to a point where the valve train would not self destruct every race weekend.
 
I think the best way would be to lengthen the valve end of the rocker arm. That would increase the ratio plus allow you to move the valve away from the centerline of the chamber and fit a larger valve without changing the valve angle and messing up the flow. I have put some time into achieving that but I have not tested it yet. Jim
 
Yeah, I like that idea although it Increases the load on the other side of the rocker and on our relatively tiny lobes and amplifies any inaccuracy in the grind but all that is academic compared to flow.
 
I went ahead and did a test on my carved from billet head that was on my crm900 bike. It uses a 42mm valve. This head was designed before I had the use of a computer flow bench and made just north of 100 horse on 890 ccs.

It is the lower [blue] trace in comparison to the Maney stage 3 and the DWS head.

Head flow testing.


Now comes the real surprise. I had tried to keep the velocity high but I had never had a good way to plot it before. Looks like I could trade a little velocity for some higher flow numbers. The velocity was so high the probe was causing turbulence and messing with the reading.

Head flow testing.
 
This is a great thread and I just want to throw in a little Tid Bit that may be of use to some one improving the Norton head. I found that the over sized valves that Paul Dunstall used in his reangled cylinder heads were from a certain year/ model BMW car engine. I dont remember which one, but I know I bought 2 at my local auto parts store back in the mid 70's and they were a perfect match.
 
comnoz said:
I think the best way would be to lengthen the valve end of the rocker arm. That would increase the ratio plus allow you to move the valve away from the centerline of the chamber and fit a larger valve without changing the valve angle and messing up the flow. I have put some time into achieving that but I have not tested it yet. Jim
Click to expand...

Eccentric adjust Needle roller :? rocker arms , as the Pin would be offset from the bores , could be done with a higher ratio.

:lol: 8) :mrgreen:
 
Matt Spencer said:
comnoz said:
I think the best way would be to lengthen the valve end of the rocker arm. That would increase the ratio plus allow you to move the valve away from the centerline of the chamber and fit a larger valve without changing the valve angle and messing up the flow. I have put some time into achieving that but I have not tested it yet. Jim
Click to expand...

Eccentric adjust Needle roller :? rocker arms , as the Pin would be offset from the bores , could be done with a higher ratio.

:lol: 8) :mrgreen:
Click to expand...

Now your talking, with a roller tip to take the side load off the valve. I will put them on my list......
 
I have a RH7 on Bike #1 that I use all the time, there is a RH10 on Bike#2 thati haven't done anything with yet. Would I be better to put the RH10 on Bike#1 with the JS Keihin carbs?
Rgards Mike
 
comnoz said:
Matt Spencer said:
comnoz said:
I think the best way would be to lengthen the valve end of the rocker arm. That would increase the ratio plus allow you to move the valve away from the centerline of the chamber and fit a larger valve without changing the valve angle and messing up the flow. I have put some time into achieving that but I have not tested it yet. Jim
Click to expand...

Eccentric adjust Needle roller :? rocker arms , as the Pin would be offset from the bores , could be done with a higher ratio.

:lol: 8) :mrgreen:
Click to expand...

Now your talking, with a roller tip to take the side load off the valve. I will put them on my list......
Click to expand...

We were initially running them in the 500cc Norton Domi Seeley. Herb Becker fashioned a set from billet aluminum and they worked nicely. We never collected any data on what they did for us but our rider at the time noticed the difference when we changed back to the standard lightened steel Norton rockers. He said before "it seemed to rev more freely". The rockers had needle roller bearings at the valve tip and on the eccentric spindle. Maybe the mechanical drag is significant when you are spinning +10,500 RPM.

The reason we went back to steel was we broke one of the aluminum ones and when we collected data for valve train analysis by Prof. Gordon Blair we realized that the aluminum roller tip rockers were about as stiff as warm Wisconsin Swiss Cheese. Clearly it was a design (flexure and durability) issue but we never got back to them for one reason or another - we were chasing demons on that engine. It is certainly worth a revisit.
 
I have tried to use aluminum rockers before- not on a Norton- They worked great for a while and then started breaking one after the other.
Aluminum rockers would need to be very large to avoid flexing and the fatigue factor. I would only consider steel or Ti in the limited space available in a Norton head. Jim
 
Brooking 850 said:
I have a RH7 on Bike #1 that I use all the time, there is a RH10 on Bike#2 thati haven't done anything with yet. Would I be better to put the RH10 on Bike#1 with the JS Keihin carbs?
Rgards Mike
Click to expand...

Gooday Brooking 850.

You sure its RH7 ?. This the short stroke head designed to run with omega pistons. It would be tough to get mid range with this head.

Standard RH10 is the best std road head for sure. (My opinion).

Cheers
 
Status
Not open for further replies.
Back
Top