Don Terrill’s Speed-Talk

Camshafts have always been a topic that has triggered a vigorous response from cam makers, engine builders and racing end users. Most discussions have centered about camshaft applications that are engine specific. I wonder if we could approach the issue of camshaft selection from a different approach—camshaft lobe design.

As we all know, the purpose of a camshaft is to open and close a valve at a certain time; the camshaft is defined by valve lift, and duration which is further defined as opening and closing points.

But through those three points (open, lift, close) an infinite number of curves can be drawn. Each curve (lobe shape) will determine how the valve train will behave dynamically. Some of the more common lobe shapes, based on their mathematical functions are:

Constant accel
Simple harmonic disp
Trapezoidal accel
Sine accel
3-4-5 polynomial disp
cycloidal disp
4-5-6-7 polynomial disp
poly spline functions

Each cam lobe shape has advantages and disadvantages that make it suitable for a specific valve train application. Some typical valve train applications could be:

Cam in block pushrod
Overhead cam tappet & pushrod (hydraulic or mechanical)
Overhead cam direct bucket (hydraulic or mechanical)
Overhead cam tappet & rocker
Overhead cam finger follower, center pivot
Overhead cam finger follower, end pivot
And variants of the above

Now the issue becomes-which lobe shape is appropriate for each valve train design? Which are completely wrong?

It would be great to hear from experienced cam makers, engine builders and racers as to which lobe shapes are best suited to specific valve train applications and their reasons as to why?

Well, most of you will tell me I'm nuts, but here's what I do.
When I design a profile for a specific engine, I use actual engine parameters to calculate the shape of the lift curve. If I design a 300 degree, .700" lift cam for a 420ci chevy, it won't be the same shaped profile as my 300 duration, .700" lift profile for a 327 chevy.
I believe that at every degree, there is an ultimate valve lift for each application. Not enough lift at any point, and you get choke. Too much lift at any point, and the air/fuel velocity gets too low.

I first design this lift curve, then if it's not dynamically or physically possible to produce, I change the curve as little as I can to meet the mechanical requirements/restrictions of the valvetrain.


OK, now fire away.

You are nuts.

Wouldn't lift at each degree be determined by the flow characteristics of the heads as well? I understand to a degree what you are saying, the velocity created by a 420 at each degree is different than a 327, but engine size alone isn't going to do it. With different heads/flow characteristics isn't it possible the same cam will work in both?

Quote by Camking

"When I design a profile for a specific engine, I use actual engine parameters to calculate the shape of the lift curve."



He is not saying that he takes CID along, the above statment would lead me to believe that he takes every aspect of the engine into acount. And the 420 to 327 compairison was just a generic example. of one small part of the build.


Camking, I have not heard of any other cam grinders talk about setting the cam up to keep velocity at X amount. That is a very interesting aproach and I woudl think it takes a lot of figuring to deside on a grind/lobe. Although I know there are many other things that come into play, but that is a interesting way of looking at it.


Jess

jmarkaudio wrote:
Wouldn't lift at each degree be determined by the flow characteristics of the heads as well?

Yes.

Jess is right.
I use the the engine parameters. That includes port flow.

We've been doing it this way since 1979.
The only thing that's really changed in 29 years, is the #'s we use for optimum velocity.

Not enough lift at any point, and you get choke. Too much lift at any point, and the air/fuel velocity gets too low.

Do you account for the change in port pressure from inertia and waves?

I don't see how this would be possible without a full-fledged simulation program.

SchmidtMotorWorks wrote: I don't see how this would be possible

Wow, that's hard to believe.

Maybe there'san SAE paper on it somewhere.

Wow, that's hard to believe. Laughing Laughing Laughing

Maybe there'san SAE paper on it somewhere. Wink Laughing

Yes there are papers on it and they do consider the inertia and waves. Do you think it could be that you have over simplified the problem?

SchmidtMotorWorks wrote: Do you think it could be that you have over simplified the problem?

Nope.
It's all done with one mathematical equation. We've had some very bright minds try and tear it apart, but it's withstood every test.

If it's wrong, or too simple, it's been that way for almost 30 years, and it's worked very well.
Maybe I'm just really lucky, but I've watched teams of engineers from GM with all the latest software fail to come within 20HP of my Indy car designs.
I'll take Lucky over smart, any day.

I'll take Lucky over smart, any day.

I'll take the luck when I can get it but even then I like to know what's going on.

BTW, check the balancing thread.

CamKing wrote:

jmarkaudio wrote:
Wouldn't lift at each degree be determined by the flow characteristics of the heads as well?

Yes.

Jess is right.
I use the the engine parameters. That includes port flow.

We've been doing it this way since 1979.
The only thing that's really changed in 29 years, is the #'s we use for optimum velocity.

Might you be willing to share just what that optimum port velocity might be, roughly speaking?

Larry Heath wrote: Might you be willing to share just what that optimum port velocity might be, roughly speaking?

In 1979, we used 300fps.

I guess I am stuck in the past as I shoot for around 325 fps.

Larry Heath wrote:I guess I am stuck in the past as I shoot for around 325 fps.

That's close to what I use now.

As you all know, Cam King and I have completely different views on cam design, as do almost all other cam designers that I know.
Even designers that use the same basic techniques design cams that are very different.
Harvey Crane and I use design techniques that are reasonably similar, but our results are completely different.
Harvey says "WEW!!", which is short for "What Ever Works!!!", and he's probably right. There is more than one way to skin a cat.

A short history of myself:
As Head Computer Operator for the State of Mississippi, I taught myself Fortran and computer cam design, buying books such as Harold Rothbart's "Cam Design", from 1956, and magazine articles like Stoddart's "Polydyne Cam Design" from Machine Design, Jan, Feb, Mar, 1953. These were my main references in teaching myself cam design by computer.
I started out with Polydyne cam designs, selling my first design to John Reed in December 1972. It is still made today.
Over the next 2 years, I began to modify the Polydyne equation, using the system spring coefficient and the dynamic deflection coefficient as tuning 'tweaks' in the equation. Such cams as the Reed R296-02-ULX were designed this way, and they are still decent cams today.
During this time, I was trying to sell unsymmetrical cam designs, but everyone was using symetrical designs, and that was what I sold.
In 1974, I went to work at General Kinetics, where I got involved with real, professional, racers, such as Grumpy Jenkins and Gapp and Roush.
In 1977, I joined the guys at Competition Cams as their original cam designer. I brought the concept of the unsymmetrical cam with me, and started developing it.
To fill in a hole in their inherited hydraulic cam line, I designed a cam we called the 268 High Energy, using the same technigues I was using on my roller cams, including unsymmetricalness.
It worked. Boy, did it work.
In 1980, I left Competition Cams and started UltraDyne.
I had a concept of using a blend of Constant Acceleration, Constant Jerk, and polynomial curves into what I called a "Multi-Segmented Polynomial" equation, or 'MSP'. Constant Acceleration and Constant Jerk curves are forms of simplified polynomials, so polynomial describes them all. The cams were unsymmetrical, with different opening and closing sides.
You guys all know the rest from there.

BTW, the port velocites and waves are caused by the cam and its design, not the other way around.

UDHarold