Camshaft lobe design

[gazunk]

Cam design is after all a science and not a black art.

UDHarold would beg to differ. His old posts are very much worth reading. viewtopic.php?t=6807

After reading the referenced posts I am not sure I conclude that his position contradicts my input. The only "black art" really left in cam design is knowing where to start. The next "best cam" is usually based on something that went before it. What I find so fascinating about camshaft design is that, due to the multitude and complexity of system inputs, I could give another cam designer a camshaft and he would struggle to determine the rank order of the relevant parameters used to design it and in some cases would not be able to calculate the rational behind the design. Like a very complex,multifaceted, jigsaw puzzle.

[OldSStroker]

After reading the referenced posts I am not sure I conclude that his position contradicts my input. The only "black art" really left in cam design is knowing where to start. The next "best cam" is usually based on something that went before it. What I find so fascinating about camshaft design is that, due to the multitude and complexity of system inputs, I could give another cam designer a camshaft and he would struggle to determine the rank order of the relevant parameters used to design it and in some cases would not be able to calculate the rational behind the design. Like a very complex,multifaceted, jigsaw puzzle.

Much like Ma Nature, who, in the end, determines how our engines run.

Jon

[SchmidtMotorWorks]

I input Bore, Stroke, Rod Length, Comp Ratio, Valve size, Choke diameter, Desired max HP RPM, and a few other things.

I think I have figured out how the different theories can coincide to similar results and how analyzing wave action doesn't need to be much of a factor in the flow demand based approach, it might come by coincidence of using motion within the boundaries of reason.

If I understand CKs idea from the partial description; it is based on the idea of targeting valve curtain area to result in a flow velocity of one or more values throughout the stroke. There obviously has to be more to it than that though.

There are at least two missing secrets:

1. If you made the valve follow flow demand you would only need the valve to be open while the piston is moving away from TDC (in the intake stroke). Obviously the duration is longer than the intake stroke so that means that there is some other part of the routine that designs the part of the lift that would be an on-ramp and off-ramp to the part he described. This is the one part that you would have to know to understand how it works. I haven't charted it out yet but it might be that the duration has to be longer than 180 degrees just to get on and off the flow demand designed lift curve with reasonable rates of acceleration and velocity at the ends.

2. Even if #1 is approximately close; that still leaves the position of the lobe center unexplained. If it were purely based on matching flow demand I guess the lobe would be at about 90 degrees. Unless it just happens to coincide with max piston velocity, that would be a funny coincidence to discover. Maybe there is some kind of offset involved.

So if this guess is close then how can designs of such different theory coincide? The difference is in what part of the lift the different approaches concentrate on. One method concentrates on getting the curtain to match the flow demand during the middle of the curve while the other focuses on catching the waves that happen to be trapped or started when the piston is moving in the opposite direction of the primary purpose of the stroke. Seems like a lot of coincidence but it might work out that way.

[AROT]

Jon,
Lobe centers are pretty much a function of precieved tdc lift and duration. regards

[SchmidtMotorWorks]

Jon,
Lobe centers are pretty much a function of precieved tdc lift and duration. regards

Sounds interesting, could you explain it in more detail?

[Larry Heath]

So I have read through four pages of how people roughly arrive at a given optimum valve displacement curve while controlling all the various derivatives up to and including Ping and Puff, using everything from polynomials, B Splines, C Splines, trigonometric series, and simple math up to and including chickens pecking at X’s and O’s in the chicken coop out in back of the shop for all I know.

All that aside, we have now arrived at, through what ever methods necessary, an exquisite valve displacement curve needing to be accurate to millionths of an inch to work as designed, ON PAPER! Do we really have the machine tools to be able to translate, ACURATELY and REPEATEDLY, what is dreamed up on paper into a reality in cold hard steel? A ten thousandths of an inch mistranslation in the interpolation going from one ¼ degree displacement point to the next puts a hiccup in the displacement curve that sends jerk and or higher derivative values to the moon, not a good thing for sure. Given that my understanding is that most cam designs today are made up of multiple independent displacement segments of a varying velocity and acceleration all spliced together and that at any one of these splices things can go badly awry. It begins to strike me as if cam grinders are trying to do the equivalent of eye surgery with a chain saw, in translating one of these exquisitely designed and controlled curves from paper to steel, even given the strides made in CNC machining devices of today. It further begins to strike me that it is more by chance than intension that I as a small time consumer of racing cams get a cam that is actually a true representation of the latest and greatest design on paper, in the steel that I receive.

So cam grinders and designers, any of you hereabouts willing to have a go at the following questions?

Over the racing cam industry as a whole, but limited to drag racing roller profiles, as that is the world that I live in and I think represents the ragged edge of design, what are the chances of an average Joe, like me, getting a single cam that truly represents what is designed and advertised as the latest and greatest wiz bang racing profile? Or simply getting what I pick out of the Master Lobe catalog. Let me ask the question another way, out of 100 cams of a given design how many actually go out the door as designed in all respects? Still further, how many of those 100 cams are actually checked to see if they “are” an accurate representation of the design intended, before they go out the door. What kind of quality control is done on the machining centers, to qualify each of them as precise and accurate enough to produce the intended displacement curves and at what intervals are these quality control checks done, daily, weekly, monthly, quarterly, ever at all after initial setup?

Any of the cam grinders or designers around here, willing to answer the above questions for their shops in particular? Give themselves a report card, so to speak?

Later, Larry

[OldSStroker]

Larry, who grinds YOUR cams? The answer may shed some light on the QC procedures used to make your cams. Different manufacturers have different abilities to "cut metal" and/or measure the results.

More importantly, who designs (specifies valve events, lobe shape or lobe number as well as all the related valvetrain components) your valvetrains? A perfectly made lobe with the wrong valve events may leave a lot of power/torque on the table. If a cam has (measured) valve events or lobe centers within +/- 1/4°, but the valve event(s) are 10° off from optimum (20 x the cam accuracy), does the engine even care if the lobes are +/- 1/4° or +/- 1°?

Unsteady gas flow in an running engine, especially if it is accelerating at many hundreds or thousands of revs per second, is an extremely complex operation. The better one's understanding of what is happening in your particular engine, the better the chances of effectively choosing valve events and everything that makes them work.

Some folks have a better idea of what makes air move through an engine than others. It IS science and it's basically very simple, but like many "simple" things in Nature, knowing exactly what is ocurring at every crank degree at every rpm is a daunting task. Our ability to actually measure such things and even to calculate them has improved an amazing amount in the last few years. Sometimes understanding lags the data gathering.



Jon

[Larry Heath]

Larry, who grinds YOUR cams? The answer may shed some light on the QC procedures used to make your cams. Different manufacturers have different abilities to "cut metal" and/or measure the results.

More importantly, who designs (specifies valve events, lobe shape or lobe number as well as all the related valvetrain components) your valvetrains? A perfectly made lobe with the wrong valve events may leave a lot of power/torque on the table. If a cam has (measured) valve events or lobe centers within +/- 1/4°, but the valve event(s) are 10° off from optimum (20 x the cam accuracy), does the engine even care if the lobes are +/- 1/4° or +/- 1°?

Unsteady gas flow in an running engine, especially if it is accelerating at many hundreds or thousands of revs per second, is an extremely complex operation. The better one's understanding of what is happening in your particular engine, the better the chances of effectively choosing valve events and everything that makes them work.

Some folks have a better idea of what makes air move through an engine than others. It IS science and it's basically very simple, but like many "simple" things in Nature, knowing exactly what is ocurring at every crank degree at every rpm is a daunting task. Our ability to actually measure such things and even to calculate them has improved an amazing amount in the last few years. Sometimes understanding lags the data gathering.



Jon

Jon I think you have misunderstood my question entirely. I am not concerned with picking a cam that will be the best cam for any given application or make the most HP. I know the pitfalls of the whole process of selecting a cam with the correct gross valve event numbers. As well as at least some of the physics related to air movement in an IC engine and how a cam shape controls this.

What I am trying to elicit from the people here that design and build cams is whether the extreme precision necessary to execute a very sophisticated extreme performance profile that is controlled at many places and to many derivative levels are in reality able to be reliably produced on the current cam grinding machinery generally available in the industry.

Again I feel that at this point in time the cam designers are quite capable of designing exquisitely controlled shapes, on paper or in the computer, that if correctly executed in steel are capable of extracting higher and higher power levels. But if any one of the hundreds or thousands of critical points on that curve is misshapen by even a few tens of thousands of an inch it could produce very undesirable affects in the valve train as a whole not the least of which could be a catastrophic failure of any one of the component parts of that whole.

The question I ask is can the machines generally used in the industry, produce the cam shapes accurately and reliably that the designers intend? Or, have the designers level of sophistication, at least in extreme performance profiles, outstripped the machines capability to faithfully produce the shapes they intend? And if the machines are at the ragged edge of performance, how often do they get the job done right.

[CamKing]

Before we had CNC, we machined the profile 5 points past the decimal, every degree.
Now we input 8 points past, every 1/4 degree. we could go more, but I don't think it matters. The machine's are only so accurate.

I don't spline anything together, so I don't have to worry about that.
My Valve lift curve, is one continuous curve.
If I measured a .0002" spike in the curve, the cam would never leave the shop.

[CamKing]

OK, I just went to my Audie Cam Pro Plus files and checked for spikes in the "jerk" curve on a couple of my cams, and a couple of other companies cams.

On the "jerk"(difference in acceleration) curve, my cams had spikes of .000013". My micro-polished cup cam had spikes of .000009". The Big Production cam company's cam had spikes of .000018". Another custom cam company's cam had spikes of .000016".
A 60 year old Winfield 64 had spikes of .000020"
A very expensive "all CNC" custom cam company's cam had spikes of .000015".

I also went and looked at one of my NASCAR designs. As designed, it has a max "jerk" of .000018". that's pretty close to the machining tolorences, and that's why I don't focus beyond "jerk".

[AROT]

Jon,
Just looked up a lobe and in this case, over the nose the proposed tappet lift increase was .0035--" in 6° cam. Much the same for the closing 6°. To illustrate how little movement there is around lobe C/L.
The curves are laid out from the center to both sides.
Generally a narrow lobe with a fixed tdc lift will have a narrow C/L and so on. regards

[Cammer]

The "Black Art" is the constant changes taking place in an operating internal combustion engine.

Everyone out there needs to understand that a great deal of what we know about internal combustion engines is based on theory!

Pull out any automotive engineering text and note all the "ideal this and ideal that" and the constant reference to theory.

All of this is important to note as we vigorously defend our positions on subjects presented within this forum. Engineers would not get far by constantly attacking the work of others. Engineers produce their own work by using science and the work of others for reference. Engineers publish their findings and others continue to work the cycle.

As is referenced above, theory abounds in the study of the ICE, and this makes the study of engines both interesting and frustrating!

"Camshaft design is just math and science until you turn the key and it becomes an experiment."

[OldSStroker]

Jon I think you have misunderstood my question entirely. I am not concerned with picking a cam that will be the best cam for any given application or make the most HP. I know the pitfalls of the whole process of selecting a cam with the correct gross valve event numbers. As well as at least some of the physics related to air movement in an IC engine and how a cam shape controls this.

What I am trying to elicit from the people here that design and build cams is whether the extreme precision necessary to execute a very sophisticated extreme performance profile that is controlled at many places and to many derivative levels are in reality able to be reliably produced on the current cam grinding machinery generally available in the industry.

Again I feel that at this point in time the cam designers are quite capable of designing exquisitely controlled shapes, on paper or in the computer, that if correctly executed in steel are capable of extracting higher and higher power levels. But if any one of the hundreds or thousands of critical points on that curve is misshapen by even a few tens of thousands of an inch it could produce very undesirable affects in the valve train as a whole not the least of which could be a catastrophic failure of any one of the component parts of that whole.

The question I ask is 1)can the machines generally used in the industry, produce the cam shapes accurately and reliably that the designers intend?
Or, 2)have the designers level of sophistication, at least in extreme performance profiles, outstripped the machines capability to faithfully produce the shapes they intend? And if the machines are at the ragged edge of performance, how often do they get the job done right.

Larry,

My short answers:

1)Yes, most can.
2)It depends on the designer, but for the good ones, no, I don't believe so.


Longer answer 1): The ability of a machine tool to make a "good" part has a couple of facets. One is the ability to follow the program (CNC grinder) or master cam (manual grinder) accurately. This is designed into the machine. As the moving parts of the machine wear accuracy could be compromised, but there are software things that can minimize this, and machine maintenance can minimize wear.

Accuracy needs to be, and is, not within a few "tenths" (.0001 inches) but within microns or perhaps a portion of a micron. The thing that can blow all of this accuracy might be an "unbalanced rock" or a grinding wheel that might not be of a uniform construction. This might cause almost unmeasurable variations in the lobe surface which could show up as valvetrain instabililty at some rpm.

Has this ever been a problem? Sure. Has it been a problem to most of us who rarely run our flat tappet or even roller cams to the high 9000s with the most aggressive lobes we can make live for a million or more engine revs? Probably not if we are having our cams lobes ground on well maintained state-of-the-art CNC machine tools, or similarly cared-for manual grinders. It may be that a well maintained manual (non-CNC) grinder run by a very skilled grinder operator using extremely well made large size masters could actually produce a truer lobe surface than a CNC grinder. Perhaps machine tools have come far enough that this is a moot point. I suggest that only the running engine or perhaps the Spintron could tell.

As far as having a hiccup in the digital design of a lobe, I guess it depends on what software you use. The more sophisticated software that I am aware of, which costs more than a modern sports car (maybe even more than the new ZR1), does not depend upon the operator to input hundreds of numbers, any one of which might cause a problem if entered incorrectly. If you are not familiar with lobe design software, you might search it out.

2) As far as the lobe designer asking for motion that the valvetrain can't achieve, or a design like an inverse radius smaller that the smallest grinding wheel available for the lobe grinder, the design software pretty much keeps that from happening if you input the correct machine tool parameters.

I am a believer in specialists when it comes to engine building. While lobe designers need to have a great deal of understanding of how engines work, they don't have the time to thoroughly analyze every engine design that needs a new cam. The engine designer is the one who determines what the engine wants. He then either chooses from existing ( tested) lobes available from his cam house, or has a new lobe designed and hopefully tested. My impression is that most "new" lobes are just slight variations in existing lobes. The maximum cam velocity, acceleration, jerk, etc. limits are quite well known by lobe designers, so going outside of those parameters usually doesn't work. Perhaps that's why there are families of lobes with duration and lift difference of a few degrees or a few .001.

If you choose the right company to manufacture your cam, you don't have to be as concerned as you seem to be, Larry. If someone tells you that he has a new radical, never-before tried lobe profile that will be capable of extracting higher and higher power levels, I would be skeptical. You just can't rape Mother Nature.

IMO, which you may have noticed is never humble, getting the valve events correct for a given engine/vehicle combination is WAY more important and WAY more difficult to do than choosing a lobe or having a new one designed which will compliment the design. Engine design is getting more sophisticated, but there are still way too many RE valve event selections going on.


As always, your opinion may differ.

Jon



RE = Rectal Extraction

[Larry Heath]

Thanks Mike, that answers my questions quite nicely. To my unwashed eye what Mike seems to have said is that yes in deed the current equitment, his and others, can be controled well below the level that might induce unwanted vibratory responce in the rest of the valve train. Am I wrong in this assumption?

If you don't mind, you say that you design as one continous curve, this includes clearence ramps as well?

You refer to your "micro polished cup cams" is the polish process used to remove the surface asparities or fair the curve or both? If a guy had a cam that he would like to increase the smoothness of the surface finish of the lobes and had a very fine flannel buff and red jewlers rouge or even some very fine mesh diamond compound, would he be helping or hurting himself by doing so, even if he was very light and even handed with the buff? Given that these lobes are pretty darn hard, 60+ HRC, can someone with a buff and polishing compound harm the desired shape? Is it even worth the effort in a roller lobe to roller lifter joint, is it going to help the joint live a longer life?

Thanks for the answers.

Later Larry

[CamKing]

If you don't mind, you say that you design as one continous curve, this includes clearence ramps as well?

No. I design a curve from valve opening point to valve closing point. The clearence ramps are added on both ends.

You refer to your "micro polished cup cams" is the polish process used to remove the surface asparities or fair the curve or both?

We do it to make the finish on the lobe less abrasive against the lifter.
We normally only polish the NASCAR Cup cams.

If a guy had a cam that he would like to increase the smoothness of the surface finish of the lobes and had a very fine flannel buff and red jewlers rouge or even some very fine mesh diamond compound, would he be helping or hurting himself by doing so, even if he was very light and even handed with the buff? Given that these lobes are pretty darn hard, 60+ HRC, can someone with a buff and polishing compound harm the desired shape? Is it even worth the effort in a roller lobe to roller lifter joint, is it going to help the joint live a longer life?

I wouldn't.
I'd be very careful with any type of polishing.
If you remove material from the wrong place, you can do a lot of damage.
Look at the numbers I gave on the spikes you'd be trying to smooth.
.000013" isn't very much, and you could make it a lot worse then that.
Back when we were doing the cams for the Buick V6 Indy engines, someone on the project decided to send our cams out to be micro-polished. The problem was, the polishing equipment couldn't apply the same pressure through the inverse as it did over the nose. This caused spikes an the transition points in and out of the inverse. The polished cams floated 400-600rpm sooner.