E: Engine Camshaft

Flo

Well Known and Active Forum User
VOC Member
Bruce, your reference to constant accelleration should have been answered in my previous post. Attached is a graphic to illustrate four traces, lift, velocity, accelleration and jerk. This clearly shows a fairly long duration where accelleration is constant, and because accelleration is constant, so is jerk.
The data is design data, not measured, and John Andrews original design data would have been similarly smooth

Hi All,

this Forum seems to concern primarily the analysis of cam profiles by measuring. The methods reported so far will always stuggle with the quality of the measuring apparatus (exemplified by the Grey One's data of a professional rig) and the method of interpolation (although at 21600 P/r this is obviated by the data density) all splines have a tendency to oscillate (usually smotheness is defined as continuity and differentiability for 2 derivatives. Whether applying a low pass filter might be helpful is outside my personal experience (I have always insited to see 'raw' data i.e. what comes out of the sensing element).
It is obvious though that the precision requirements are much less for design data, where one would put the profile together from known functions (always making sure that at the transitional points the 3d derivative (jerk) is continuous and differentiable) so the relation of successive surface points is obeying the maths and not the absolute value. I am sure that Mr. (Dr. ?) Hubbard has something to say about this in his Handbook.

For a basic understanding of valve trains I always recommend Wolf Bensinger's classic text Die Steuerung des Gaswechsels in schnelllaufenden Verbrennungsmotoren (unfortunately in German). Hence the quieting ramp is really important for the opening flank (when the effective clearance is reached on the closing flank valve and spring have come to rest and do not contribute to valvetrain inertia and forces. And of course when acceleration is constant jerk is zero (which I am sure has contributed to the German misconcepton of Ruckfreier Nocken). While a world free of jerks is certainly not achievable a cam profile free of jerk is not even desirable.

F
 

roy the mechanic

Well Known and Active Forum User
VOC Member
Ever since the industrial revolution there have been prototype engineers and toolmakers who take pride in making the most accurate and best finished parts possible. In the real world for an air cooled motor around .001" will give good service. But you must not tell the guys making your parts, they get upset. On a personal level it's nice to do a build with the best quality parts.
 

vibrac

Well Known and Active Forum User
VOC Member
I dont know about Maughan & Sons but I bet Roger Moss could give it a good go I visited his place a year or so ago when |I wanted some advice on a Scott engine I was building. He has certainly has a lot of fancy machine tools and measuring equipment.
 

Bill Thomas

Well Known and Active Forum User
VOC Member
Hands up, All the Vincent owners that have seen an 1/8" tappet clearance at some time in there life !!.
Even if you got it spot on , It would not stay like it long. Cheers Bill.
 
D

Deleted member 3831

Guest
MK 2 cam profiles meaured from a new cam in 2012. The intake only is shown, with velocity and accelleration traces overlaid.
Unfortunately the manufacturer is unknown as the cams carried no identification. This is data measured at the cam, not via the follower. This data would be the basis for constructing a manufacturing file to have copies ground up.
Attached is a strip of closing data to 8 places of decimals, measured at 1 degree increments. With such data it is possible to determine the roughness of the profile being measured. Less accurate data at greater degree increments makes it difficult if not impossible to determine whether roughness is the resjult of the data format, or the cam surface. Rough data can be used to check the cam installation, but is worthless for any meaningful comparison purposes.
 

Attachments

  • Mk 2 In lift plus vel plus acc.jpg
    Mk 2 In lift plus vel plus acc.jpg
    133.3 KB · Views: 50
  • Close measured data.jpg
    Close measured data.jpg
    39.6 KB · Views: 49

timetraveller

Well Known and Active Forum User
VOC Member
It is always interesting to communicate with an expert and someone who has designed cams which won TT races etc. is certainly an expert. However, referring to the above graph from Grey One which is taken from the face of the cam with no cam follower and no allowance for Vincent timing gear geometry, then it seems to this untutored producer of 'worthless data' that although this might be the correct way forwards to manufacture cams, it lacks something vital when it comes to valve movement etc. And before Grey One asks, yes I have designed cams and had them manufactured and, as an aside, I found it difficult to get people to work to one tenth of a thou (fourth decimal place) never mind an eighth decimal place. My worry about the relevance of the above graph to the behaviour of the valves in an engine is the geometry of the Vincent valve gear. The flat based lever follower imparts a variable lever ratio between the cam face and the valve movement. Once again, writing as a non professional cam person, it seems to me that it is the movement of the valve which is of importance, not just the cam profile. I have taken the liberty of tracing the profile above on to tracing paper, inverting it and then putting it on top of the original to see just how asymmetrical this displacement curve is. The answer is ' not very'. This seems strange to me as Vincent cams are clearly very different from one side to the other.
Below I am going to show a graph of the valve motion and the cam displacement for a Vincent cam.
26283

The blue points are the valve motion and the red the cam displacement. One can clearly see the differences in the ratios as the cam surface moves along the flat lever follower, giving a variable leverage ratio. The horizontal scale is cam degrees and the vertical scale is inches.
The next graph shows the way in which this ratio varies Between about 1:1 and 2.8:1.

26282

Please note that the shape of this graph would change if one were to index the lift profiles at a different point. Moving the point at which the two maxima coincide makes a big difference to the ratio values but they always vary considerably. Consider that the inlet valve starts to lift when the cam is contacting the lever follower towards its pivot, high ratio, and the exhaust starts to lift when the cam starts to contact the lever follower away from the pivot, low ratio.
What I do not understand and which it would be a pleasure to be educated about, is why the velocity and acceleration curves from the cam surface look sensible when the above ratios need to be allowed for when it comes to valve movement, which also has to look sensible.
 

roy the mechanic

Well Known and Active Forum User
VOC Member
Having spent 30+years building race car engines (a lot of them winners) cams are still a black art. I have worked with two well known cam makers. To the best of my experience, they are all full of "bull cookies". It is no importance what it looks like on paper. The only parameter is the numbers on the dyno, and the numbers on the stopwatch. that's the only way forward. Or go back and spend more money!
 

greg brillus

Well Known and Active Forum User
VOC Member
Yep.......Look at modern engines and the trouble manufacturers have gone through to get engine performance and tractability. Variable valve timing, exhaust port/ pipe valves to vary back pressure, variable length inlet trumpets............ All the stuff we are playing with is stone age.
 
D

Deleted member 3831

Guest
Having spent 30+years building race car engines (a lot of them winners) cams are still a black art. I have worked with two well known cam makers. To the best of my experience, they are all full of "bull cookies". It is no importance what it looks like on paper. The only parameter is the numbers on the dyno, and the numbers on the stopwatch. that's the only way forward. Or go back and spend more money!
Roy,
The Americans found this out some years back when they devised the Spintron equipment and discovered there could be a big variation between static and dynamic valve actuation. Originally designed to observe the flex of pushrods, it was quickly expanded to cover all components of the valve train, and the variations that were found to occur between the static and the dynamic valve actuation were a big eye opener.

TT
I attach a graphic to show overlaid traces of measured cam data, the MK 2 profiles and those of the 105.
The intake lobes are both aligned exactly one on top of the other, and in addition to the cam lift, the graphic also includes the velocity traces. You just cannot get this kind of graphic from rough data. The traces show where discrepancies occur between the two cams, even though the 105 was said to be a copy of the MK 2 but with the lobe separation increased. If required a section from anywhere on the graph may be enlarged (zoomed), by four times if needed. All the top cam grinding companies now have CNC grinders although the older Churchill and Berco machines still handle a lot of the run of the mill grinds The CNC machines run from data files which contain data to 8 decimal places.

Greg
I agree with your 'stone age' comment, but it doesn't need to stay that way. I guarantee the Horner brothers didn't operate that way when deloping their Vincent racers.

Anyone sufficiently interested in this subject would do well to obtain a copy of Jeff Williams book "Introduction to the Analytical Methods for Internal Combustion Engine Cam Mechanisms".
Jeff designed the cams for the ILMOR 265E engine (V8 pushrod) that won the INDY 500 in 1994
 

Attachments

  • MK2 and 105 overlay.jpg
    MK2 and 105 overlay.jpg
    241.9 KB · Views: 29

Robert Watson

Well Known and Active Forum User
VOC Member
I have just had a recall on a 2012 Subaru 2.0L because they are breaking valve springs. They don't have the tooling or the parts yet at the dealers to do the repair. If one breaks you may feel a vibration or hear abnormal engine noise and the engine may stall and be unable to restart. If this should occur please pull over and call your nearest Subaru dealer.

It amazes me that things like this still occur, and why they still are unable to make many disk brakes that will not warp in normal driving conditions.........

Never mind cam design!
 
Top