Cam lift - value of overlifting?

[Christopher_205Rallye]

Hi Guy Happy new year to you!!

I have a question that I would like your comments on.

There are some schools of thought that maintain that say that it is beneficial to lift a valve over the peak flow rate.

Because: The value has to accerelate faster then through this lower flow, lift region and therefore result more area under the lift curve and more torque.

Is there any truth to this in your experience?

Best Regards

Christopher

[Guy Croft]

Chris hi!

Overlifting (did I coin the term? Is it a 'Croftism'?) is when you lift the valve higher than point X even though the flowbench (or other test method) tells you that there was no increase in flow beyond that. Why? reasonably simple mechanical explanation, nothing to do with flow.

Fundamentally this is about springs more than cams. The spring has to be able to cope with the acceleration imposed on it by the cam during the lift phase. In basic terms if the spring is too weak it cannot stop the valve coming off the profile.

For highest torque and power you always want the biggest lift integral on the cam, in other words the most area under the cam lift-degree curve. What you can do is totally constrained by the valve springs. The diagram showing cam areas (3 overlaid) explains that bit. Now refer to the cad dwg I just did. Two cams (red, yellow) of identical duration. Ignore detailed features like radii, ramps and base circles etc. It's just a very basic sketch to illustrate overlifting.

The red cam has a higher rate of acceleration than the yellow one, and albeit that it has a very big lift integral and a lot of dwell (I've exaggerated for emphasis, dwell holds valve open at peak lift) let's suppose that its rate imposes far too high a load on any available spring. So it just can't be used. It is a common problem with cam design and I have come up against this from time to time. You don't 'overlift' unless you're supertuning, believe me, things get complicated. Manufacturers of production engines incidentally, rarely do it. The adoption of air valves in F1 engines solved the problem of very high spring rates with very high lifts. Apart from anything else, wire-wound helical (coil) springs are very prone to fracture if you run them too close to their 'known' metallurgical stress limits.


Accelerative rate is derived form differentiation of the lift-time function. To keep our big lift integral, we first reduce the rate (see 'A' showing that the red cam lifts to a given height several degrees quicker than the yellow one). Then, by reducing the flank angle we can increase the lift (assuming a higher lift spring is available, maybe one with a lower poundage, probably with thinner coils - or fewer of them). The dwell period suffers, but we hope by a bit of playing-around that the increased flow around peak lift compensates for the loss of flow on the opening and closing flanks of the red cam. Unfortunately even a dyno test won't prove you're right because with software simulation you'll never know what the 'red' cam would_have_done.

BTW - you just left-click on the pics to enlarge them, we haven't figured a way of indicating 'thumbnail', seems to be a forum glitch..

I hope this is helpful,

GC

edited for clarity by GC 1652hrs 32rd Jan 08

[Christopher_205Rallye]

HI Guy,

Firstly thanks very much for taking time to answer my question in such detail and with helpful CAD diagrams. It is very much appreciated.

I must say that I have read this quite a few times to try and understand the concepts you are explaining correctly. Its really quite complex.

Let me try and make some deductions and statements. I probably all wrong of course.

This seems to be more about the forces at play on the valve train then flow as you say. What your diagrams also show is that one has absolutly no idea what a camshaft is like based on the duration or lift figures. One can only know that the lift will lift that much and the is off the seat for that many crankshaft degress. How fast or how long it stays there is a secret guarded by the grinder?

If I try and now turn the situation around. If you look at the modifed diagram of yours (apologies for the bad diagram I have no AutoCAD). If we say that the red curve is a high rate of acceration. However not the highest that the valve train can take. Then we could install springs of a higher specification and in this case if we accerlerate the value faster to the peak flow level (Curve D) but also allow the value more time to slow down (so resulting in lift over peak flow during this slow down period) then the lift integral also increases?? Would this not result in a better torque spread perhaps?

I know this is all theory and it's probably far more complicated then this.

Also if we lift higher (Curve E) but keep the accerelation the same but lift a little higher there is little gain.

Cam Lobe.jpg (40.36 KiB) Viewed 5990 times

Best Regards

Christopher

[Guy Croft]

Chris, hi

the red curve is already assumed to be the highest rate the valvetrain can take,

G