Cam profile design 10

[EngJW]

Hello,

Just discovered this forum. I have been searching for a forum on camshaft design. That is, actual design like ramps, profiles, materials, machining, and so on, as opposed to "what cam should I use?" I found a gear forum but no cam forum. Anyone know of one, or is this forum the appropriate place?

Thanks,
John Woodward

 

[EngJW]

I have many old SAE papers and the polydyne papers from Machine Design. I worked with Harvey Crane for a while on a new aircraft cam, and also know the guys at Comp Cams as they are across the street from where I used to work. I eventually wrote my own program but I still do not know as much as I want to.

Thanks,
John Woodward

 

[RacerRick]

Is there a basic formula for the valve spring pressures need on flat tappet hyd. cams. I'am interest in using as little pressure as possible for max 4500 rpm truck engine. Some cam manufacture recomend 400lbs/in springs and others recomend 250/spring. I liketp find a formula that takes into account the following,,,, valve, retainer and 1/2 the spring wieght, rocker ratio, and lift per deg of rotation. Am I out in left field here?

 

[EngJW]

To RacerRick,

There are equations scattered about in the literature but nothing basic. I don't know of any that you can use as is. They usually take a lot of manipulation to put them in a workable form and then they are too time consuming to use unless you write a program.

You have to know the cam profile to get the valve lift and acceleration at every degree. You already have the weight of all your components. For the spring you need to know the compressed load at one point and the spring rate. Basically the formula starts with

Force = Mass x Acceleration

You calculate the inertia force on the valve train created by the cam acceleraton, using the above equation. Plot it for 360 degrees. Next, calculate the spring force applied to the valve at every degree and plot it on top of the inertia force.

There should be more spring force than inertia force at every degree. It is not enough to do the calculation at max lift, as there could be points in the early stages of lift where the curves get close together.

Exhaust pressure also needs to be considered. The exhaust valve has to be opened against cylinder pressure, which can cause high forces, and in a turbo engine the exhaust port pressure cam overcome the spring seating pressure.

Of course, none of this has answered your question. To meet your immediate needs it may be best for you to go by the cam manufacturers recommendation, especially if they offer a kit.

John Woodward

 

[ProgressiveRacing]

Racer,

There is a formula in Joseph Harralson's "Design of Racing and High Perforance Engines" that might be of some help to you. If you're not familiar with the book (ISBN 1-56091-601-X) it is basically a collection of SAE papers. Here's the excerpt which deals with abnormal valve motion:

"VALVE MOTION- Abnormal inlet and exhaust valve motion is a major obstacle in the development of high-speed engines. The main factors influencing valve motion are:

1. The degree of ridgidity in the valve train.
2. The equivalent inertia weight of the moving parts in the valve train.

For the tappet-type and rockerarm-type of overhead cam valve mechanisms, we determined the engines speeds at the initiation of abnormal valve motion (such as jumping or bouncing) and plotted these data as a function of the equivalent inertia weight of the valve train.

It can be seen that the abnormal motion begins at a higher engine speed for the tappet-type, which can be attributed to the higher ridgity of this system. The equation generally used in valve mechanism design is expressed as:

Ne = 2C SQRT[(Fs·K)/(W·(-y))]

where:

Ne = engine revolution speed
Fs = valve spring force at maximum lift, kg
K = G(30/pi)^2
W = equivalent inertia weight of valve train, kg
y = maximum negative acceleration, mm/rad^2
g = acceleration of gravity 9.8x10^3, mm/sec^2
C= correction factor

In order to predict the engine speeds at which valve jumping and valve bouncing will begin, we empirically determined the value C as follows:

C = Cj =0.85 (for valve jumping)
C = Cb =1.0 (for valve bouncing)"

Sorry for the length but wanted to get everything into context there.
I post this in hopes that someone can maybe explain it. I'm a sophomore in college working towards an ME degree so I'm kinda stuck waiting til I have a better understanding of kinematics and dynamics before I can extrapolate anything from this on my own. I'm sure someone can shed some light on this though.

Best of luck.

 

[EngJW]

Mr Prog.,

Can you supply the SAE number of the paper? I don't have the book but might have the paper.

I usually don't bother with empirical formulae if one can be derived on a scientific basis. You should soon be able to do this, although you may know enough already. It is just a matter of writing F=ma for the valve train inertia, but writing it as a function of cam degrees. Acceleration is from the cam profile. You can get it by differentiating the polynomial equations if known or if you have a lift table find the difference between 2 lift points for velocity and the difference between 2 velocity points for acceleration. The mass comes from the weight of the moving parts in the valve train, taking into consideration the rocker arm ratio. I don't remember the definitions; I can look them up if you want.

Now you have valve train inertia force vs. cam angle. Next you calculate the spring force at each cam angle. If you know the spring force with the valve closed, you then add to it lift x spring rate at each cam degree.

If someone would tell me how, I would like to post an illustration here that would take the place of all these words.

John Woodward

 

[ProgressiveRacing]

Thanks for the help John, I need to digest this before I can ask any pertinent questions.

The paper I referred to is, "Research and Development of High-Speed, High-Performance, Small Displacement Honda Engines" SAE #700122

Valve motion is dealt with in the mechanical efficiency section of the paper but only in brief.

Another paper that you might be interested in is, "Modeling and Measurement Techniques for Valve Spring Dynamics in High Revving Internal Combustion Engines" -Ford Motor Company SAE #930615

Of course most of the information in that paper is well above my head at this time but it is very informative.

Thanks for your time,
Allen

 

[victorymotorsports]

John,

Specifically, what do you want to know?

 

[RacerRick]

John You mention writing a program. Do you know of any good programs that will do the math for me? I looked around the web and there are a few. Now I need to know which one to buck up for? Thank

 

[EngJW]

Rick,

Let me know where you found the programs. I would like to look at them. My searches never turned up anything.

The program I have is written in Visual Basic. I obtained the equations from some old SAE papers but they required considerable modification to make them usable, and also to be able to solve them. I added lots of things like ramp calculations, plots, lift tables, etc. Other programs in the group calculate valve spring forces and valve train inertia throughout the range, and cam and lifter stress. The main program is for the design of the profile. A smaller program will generate plots from a lift table.

At one type I worked with Harvey Crane and used his program. My program gives the same results, but to me it is easier to use. It is similar in principle to what they use at Comp Cams but not as sophisticated.

I have half a mind just to give it away to anyone that wants it. Trouble is, I am not sure it will run on another computer without VB being installed.

John Woodward

 

[RacerRick]

John Heres one

http://www.integralcams.com/springmaster.htm

and another with a demo

http://www.soft-engine.com/pagine.web/inglese/cams0.htm

 

[rrohrich]

Don Hubbard's "Camshaft Reference Manual" is a very detailed reference on the low level details of camshaft design. Pricey but worth it, and it comes bundled with some very good software.

Professor Gordon Blair of Queen's University of Belfast wrote a very sophisticated cylinder head design package that covers very detailed cam and spring design details.

Details at

http://www.profblairandassociates.com/GPB_Products_Main.html

Some good cam design references can be found here:

http://www.audietech.com/CAMBiblio.htm

I hope this helps.

Rich Rohrich

 

[hitzinger]

Dennis,
I would be interested in your cam design sources as well.
If you could e-mail infos, design software and so on to me would be great.
Thanks
Alex

 

[victorymotorsports]

I should have posted more the first time...

I work for a major tier 1 supplier of valves and valve train parts. We do not design camshafts, but we are very knowledgable on their design, materials, dynamics, etc.

What specific questions do you have?

Chris Hill

 

[EngJW]

Chris,

Lots of questions as I think of them, but let me try this one:

I know how a flat tappet cam is ground, for example, using a Norton grinder and a large wheel. The same lift table can be used for a flat or roller cam, but the shape of the lobe is very different. The large radius of the large grinding wheel approximates a flat lifter. However, it is not very clear to me how a roller tappet cam is ground. Since the profile depends on the roller diameter, what size grinding wheel would be required, and how is it determined?
Equations would be great so that I could add them to my program.

Many thanks to anyone who can provide this information or direct me to it.

John Woodward

 

[ivymike]

jlwoodward,

I've never actually seen a cam being ground, but I do know that different manufacturers will have different concavity limits (minimum allowable amplitude of negative radii), which I assume are determined by the size of the grinding wheel that they use. A program such as Camspring (Ricardo Software) can calculate cam cutting ordinates for just about any combination of valvetrain geometry and follower diameter. Sorry that I can't get you any closer to writing the equations yourself, but if you happen to get a copy of the user manual for Camspring, the equations might be in there (not sure). If you're doing this work for your boss, perhaps you should push him to get a trial copy of some off-the-shelf cam design software. It might be worth the money.

Regards,
Isaac

 

[victorymotorsports]

John,

Sorry for taking so long to get back to you...

I'm not real strong in this aspect of cam design, but here goes.

Basically, you have to determine what the radius of curvature of the cam needs to be for a particular lift profile vs. roller diameter. I'm not sure how to explain it though. But I do think some calculus books address this issue.

If you have some dynamics questions I'm better with those.

 

[facty]

John & Chris,

The cam profile depends directly on the geometry of the roller follower. You will know/notice that the cam profile for a roller follower may have a concave portion and it is the radius of curvature of this portion that determines the diameter of grinding wheel that can be applied. The max cutter radius is limited to the min absolute value of the rad of curvature in this portion. (otherwise it would physically not fit into the concave!)

 

[RacerRick]

Ther are very few cams with IR flanks that can't be ground with a std. wheel and machine. We lay a straight edge across the cam flank to see if its concaved. If it is we use a 14" wheel. You can buy attachment to run a 7"wheel but I have yet to come across a cam that needed it. I checked with a cam doctor. I guess some wild ultradyne or IR crower might need a smaller wheel.
John when grinding any steel cam, we slow the cam rotation way down to prevent chatter. Its no different than grinding a cast cam except for the time it takes.

 

[ShaunT]

Get away from Hydraulic cams. If you want to make power, it is time for a clearance ramp cam. Do Not fall victim to the statement: You can't run solid lifters on the street because they have to be adjusted constantly. This is a wide spread mis-guided criticisim of clearance ramp cams. They are constantly being adjusted because the locking method used to lock the rocker arm down is either not utilized or is totally installed wrong by the ameteur. 99% of the cams I personally sell for street/marine engines and 100% for race engines (accept where rules ban me from doing so) are solid flat tappet or solid roller cams. If you take a .025 lash cam and set the lash at .019, the valve train is almost void of any solid lifter noise. Makes a Great Sleeper! The reason you need to run from hydraulics is that they are slow and lazy opening cams compared to clearance ramp cams. For example: a solid with 264degrees advertised duration @.020 is 235 @.050 VS a hyd with 280 advertised @.0045 is about 230 @.050. My point is the 264 opens later thus smoking the much earlier and slower opening hyd at slow engine speeds, and also will beat it up stairs too. You see it is caught the nearly 20 degree bigger hyd by .050, so what do you imagine it did to it up @.200? It is actually bigger!!! It is smaller in the motor before TDC, but bigger after. (It lies to the motor) And since I am ULTRADYNE, I'd like to say that we are the only ones that will show our cams at.200. It's kind of funny when I speak with other cam manufacturers(wanna be cam designers that are largly cam copiers)when they tout their seemingly cool duration @.050 numbers, They are dumbfounded when I ask them what the .200 is. We are the only true mathametically non compromised cam design. Dont be fooled by a so called computer designed cam. It is only as good as the programmer.