Register now to get rid of these ads!

Technical Homemade Roller Cam for Flathead Six

Discussion in 'The Hokey Ass Message Board' started by 27troadster, Aug 30, 2013.

  1. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    I started a thread about 6 mo. ago on a two piece aluminum head for a flathead 6 cly Plymouth. In that thread, I mentioned my plan to make a roller cam from scratch which generated a lot of ideas and questions from fellow HAMBers, thanks guys for all the input and great ideas, so I decided to start a thread on my Cam Project.

    The goal of the project is not ultimate HP (as in open-a-Summit-your-check-book-and-start-ordering ultimate HP, now pushing the ultimate HP out of a blown and rollered 201 cuin flathead, well, that's a different story!), nor is the goal to order custom parts from any one of a number of excellent cam grinders in the country. The goal is to have fun, learn new concepts and skills and enjoy the challenge of it.

    So the basic idea: build a cam and all the associated machinery, electronics and software from scratch.

    1) write a software program to design cam shafts
    2) design and build the electronics that will interface between the computer program and the grinder.
    3) build the grinder.

    I've got the software program about 70% complete, but haven't touched it in about a year. Attached are some very early screen shots I'll upload some newer versions of the program when I get a chance. There are four windows, the upper left is the cam profile, this happens to be a radical roller design with the center of the cam located where the axis cross each other, thus the base circle is on the right at 0 degrees and the centerline of the lobe is at 180 degrees in standard x-y graph fashion, the upper right is the lift as a function of cam shaft degrees, the lower left is the velocity of the lifter, and the lower right is the acceleration in in/sec/sec. (latter versions read out in pounds of force). The specs on this particular lobe are: Base circle radius - .5", roller diameter - .750", lift - .450", duration - A LOT! ~330 degrees of crankshaft rotation at .050" lift, vlv is dwelled at its max lift.

    To design the cam one clicks on the squares on the profile and drags them, the computer then calculates a smooth curve through all the square points for the new profile and calculates the lift, velocity, and acceleration. The profile in the first attachment is a relatively smooth roller cam profile (albeit a very radical profile). In the second one I dragged the the square that is highlighted blue in a little and we can see the step in the lift graph and the rather huge accelerations, which one would expect with a rather large dent in the cam shaft lobe.

    Incidentally, When I started writing the program, I assumed the hard part would be the behind-the-scene calculations, but by far the hardest part is getting the graphs to be displayed correctly!

    To answer NEXXUSIAN's question from the other thread: "I don't suppose we could get links to the threads that gave you your ideas about grinding cams?
    I would love to read more bout inverse radius, I remember hearing about that in the late '80s somewhere."

    Sorry to disappoint, but the fact is I've learned very little about cam shaft design via the internet, rather my initial knowledge came from reading hot rod magazines and thumbing through Crane and Comp cams catalogs over the past 25 years. For example I remember years ago reading about Isky's noisy cam shafts in an article about Isky and his company in the early years. It was in the 1-2 lines of this article that I learned about lead in and out ramps necessary to gently take up the vlv clearance on the one end and gently set the valve on it's seat on the other end. Otherwise you have mushrooming vlv stems and vlvs that bounce off their seats and beat themselves to the point of failure. The vast majority of my knowledge came from writing the cam design program, there's no better teacher than just getting in there and doing it. Then about a year ago I got to the point where I needed experimentally determined information, for example, just how fast can the vlv contact the seat without bouncing, in the program I made some assumptions and mathematically calculated it, but that's not the same as what others have learned from experience, and just how much jerk (how fast the acceleration is changing, this is basically a measure of how much we are beating the components, for example a hammer striking an object causes a large change in acceleration and therefore a large jerk and as we know, beating things with a hammer tends to break them) can a cam and vlv train withstand, etc. So about a year ago I bought a book by Don Hubbard called the Camshaft Reference Handbook. I recommend it to anyone who wants to learn in-depth cam shaft design. I was happy to see that the graphs and limiting parameters that my program was outputting matched the graphs and information in the book. That was a relief because it validated that I was on the right track and thus I should continue.

    Don Hubbard did point out one thing that I wish I would have done, he said we really don't care about the profile of the cam, the only thing that matters is what the vlv is doing in relation to the piston. Therefore a cam program should allow the cam designer to input what he wants the vlv to do and the program should output the cam profile necessary to achieve that result. I wrote my program in the opposite direction, in my program the designer fools around with the cam profile until he gets the vlv movement he is seeking. I've been thinking about re-writing the program to do what Don suggests, but I know that will take several months of writing code....

    Inverse radius flanks:
    Or whatever you want to call it, the sides of the cam lobe curve inward towards the center of the lobe.
    So first big picture, in a perfect world the the vlv would instantly snap open to it's maximum lift at piston TDC, stay there until piston BDC then instantly snap close. The maximum lift would be the lift at which air flow is restricted by port size instead of vlv lift. Obviously mechanical things can't move instantaneously. But the more we can approach this ideal the better off we are, and more HP. With flat tappet cams we are fairly limited, there is a finite rate at which the vlv can be opened, and stay within the dimensional limits of the camshaft and engine block. Thus the higher the lift the longer the vlv is open for and the longer the duration. Which as we all know is fine for high RPM HP, but sucks at idle, hence that awesome idle we all love, esp. a big block loping at the stop light about to quit every other revolution. Anyways, with a roller cam we can invert the flanks of the profile and the small diameter roller (compared to a flat tappet with a ~40" radius on its face) can follow the the profile, this allows large lift with less duration. A draw back of a roller cam is the large side loads on the lifter, (a flat tappet has no side loading of the lifter, neglecting the small amount of friction between the cam and the bottom of the lifter) and by inverting the flanks that side load increases quickly. If one were to imagine a line drawn verticaly through the centerline of the lifter and a second line drawn from the point of contact of the cam and the roller through the center of the roller, there will be an angle between these two lines as the lifter is being raised or lowered, from what I've read 30 degrees is the maximum that can be tolerated before the side loading is too great and the lifter gets stuck in the bore...bad day. So that's the limiting factor in inverse radius flanks. Also they require a special cam grinder because a conventional grinder uses a large 24" or so grinding wheel which cannot grind an inverted curve into the cam lobe.

    Anyways more to follow, this is a huge subject. I'll post more as time goes on. I need to figure out how to share the program with any one who is interested.

    All ideas, comments and concerns are welcome, I'm still very much in the learning phase and there are a lot of you who know much more than I do, especially when it comes to on-the-track, on-the-dyno experience and various things tried over the years.

    cam prog early ver.png

    cam early ver 2.png

    Last edited: Apr 6, 2018
  2. Fullraceflathead
    Joined: Apr 27, 2012
    Posts: 23


    Great Project to Learn from.
  3. krooser
    Joined: Jul 25, 2004
    Posts: 4,585


    You should talk to Harvey Crane about this deal... he still does consulting for some cam grinders and OEM's and may give you some advice.
  4. coryw
    Joined: Nov 4, 2005
    Posts: 225

    from Omaha, NE


    Harvey Crane passed away in June.

  5. Ambitious but very cool!
  6. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Hi all, I'd like to share more of what I have learned about camshaft the topic of:

    Max Valve lift and Dwelling the cam.

    When I began this project, I nievely thought there's no need to lift the vlv more than what the port can flow. In other words, as the vlv begins to open more air can flow through the port and vlv and into the cylinder. The more the vlv opens the more air that can flow. But there comes a point when the vlv is no longer the factor that limits the air flow and rather the port becomes the limiting factor. Once that happens, no matter how much the vlv is lifted beyond that point, there will not be an increase in flow. For this example I'll use a Ford flathead. For a ported, polished, undercut, relived flathead Ford this occurs around .430" So make the max vlv lift .430, right? Nope...well maybe...depends on the situation.

    Lets say our cam had a max lift of .430" That means that maximum air flow is only reached for a brief moment just as the vlv reaches max lift. It would be much better if maximum air flow could be had for more than just a brief moment. So we need to keep the vlv open to at least .430" for a significant amount of camshaft duration to realize more HP. There are two ways to do this, first we can simply open the vlv more, say to .515" max lift, that way the entire time the vlv is open greater than .430 there is maximum air flow through the port. The second is to dwell the vlv at .430". There are several problems with dwelling the vlv, larger crankshaft duration at .050, larger loads on the cam and stiffer springs are required (esp to keep crankshaft duration down).

    To illustrate these points refer to the 3 attachments.

    The first shows a flat tappet cam with .515" max vlv lift and 270 degrees crankshaft duration. The vlv is greater than .430 for 54 camshaft degrees, note the blue lines on the lift graph, the horizontal .430" line crosses the profile at -27 degrees, and there is another 27 degrees past the lobe centerline, designatied as 0 degrees, that the vlv remains open greater than .430". (we have to keep camshaft degrees and crankshaft degrees straight here, crankshaft degrees are the industry standard when describing duration, taken at .050" lifter lift. When referring to this duration I will use crankshaft degrees, for all other "durations" I will use camshaft degrees)

    Some graph explanation first, the lower right graph is Acceleration/Force, the values are in pounds - force (lbf). Positive values denote the force that the cam has to apply to the bottom of the lifter and negative values denote the force that the spring needs to exert in order to keep the lifter on the cam lobe. The blue curved line on the lower half of the Acceleration graph is the force that the spring applies, as long as this value is larger in magnitude than what is required to keep the lifter on the lobe, then the lifter will follow the cam lobe, else it will leave the cam lobe. The force depends on RPM and the mass of the vlv train. I chose 300 grams for vlv train mass and 4500 RPM, the exact values won't affect the outcome of this discussion. Also, the spring shown is stiffer than what is required for this particular cam and vlv train combination, but again, it won't affect this discussion.

    For the first attachment, the largest load on the cam is 265 lbf and, though not shown, the spring would need to apply a force of 90 lbf when fully open.

    By comparison, the second attachment shows a flat tappet cam that dwells the vlv at .430" for 54 camshaft degrees, therefore, the same amount of time for maximum airflow as the first cam. But the crankshaft duration is now 330 degrees, at .050". (I'm not even sure if that will run?!) All other parameters where kept the same. This cam too would require 90 lbf of spring pressure at max lift but notice that the cam now has to exert ~600 lbf on the bottom of the lifter, a significant increase. In order to reduce the crankshaft duration, the time spent at .430" would have to be reduced on this particular cam.

    Therefore, for a given crankshaft duration, a cam that has a max vlv lift greater than the vlv lift required for maximum air flow will allow the port to achieve maximum air flow for a longer period than a cam which simply dwells the vlv at the vlv lift required for maximum air flow, thus more HP.

    Back to the cam that dwells the vlv. By switching to a roller cam, (see the third attachment) we can invert the flanks of the cam and get the crankshaft duration back down to 270 degrees. However, notice that the spring force required to control the vlv train has more than doubled, the spring on this cam would have to exert ~200 lbf at max vlv lift, as opposed to 90 lbf with the first cam.

    Therefore, comparing the first and third cams, both have the same crankshaft duration and both keep the vlv at or above .430" for 54 camshaft degrees, but the one that dwells the vlv requires a spring that is over two times stronger than the first cam.

    So, that is why it is better to have a cam that exceeds the maximum that the port will flow. The only reasons to switch to a cam that dwells the vlv would be in some sort of racing that limits max vlv lift or, as in our case, a flathead engine where max vlv lift is limited by the head. An overhead vlv has no restrictions on max vlv lift so that is why we see .650" plus cams. But on a flathead, the more we relieve the head to get greater vlv lift, the more we loose compression.

    A roller cam that dwells its vlv seems to be the best option for a flathead engine. And, fortunately, the vlv train is lighter and more robust than the overhead vlv engine, therefore, we don't have to worry about the higher spring pressures.

    Incidentally, 270 degrees duration at .050" is a competition only style cam. For example, 270 dur in a SBC has a power range of 4000 - 8500 RPM. This is not what I'm planning to run, it was just for illustrative purposes.

    Other things that limit the "dwelling cam" more than the "excessive lift cam" that I haven't discussed here are the ability to get lead in and out ramps that will allow the cam / vlvs to live and satisify the radical geometry of the dwelling cam and the limits imposed by jerk.

    Any thoughts? things I've forgotten? tip? please let me know.

    flat tappet 430 for 54 deg.png

    flat tappet 430 dwelled for 54 deg.png

    430 54 deg dwell 270 dur.png

    Last edited: Apr 6, 2018
  7. The rate at which a roller opens and closes also impacts how much spring is required. As the rate increases, so must the spring rate.
  8. OldColt
    Joined: Apr 7, 2013
    Posts: 504


    Wow, being a mechanical enginering major in college, I find your stuff very interesting. I bet the current pro stock engine builders have cam profiling etc ... down to a pretty exact science. I also bet they wouldn't be willing to tell any details.

    --- Steve ---
  9. Rusty O'Toole
    Joined: Sep 17, 2006
    Posts: 9,404

    Rusty O'Toole

    Have you thought of doing a flat tappet cam, hard facing the surface with hypereutectic welding rod?

    Or welding up a stock cam and regrinding it?

    Isky used to grind a groove in a stock cam and fill the groove with weld, to make a hard face cam. They were recommended for racing not for street use, I don't know why they were not recommended for street use. Except they were made for extreme high spring pressure which you would not need in a street driven car.
  10. CNC-Dude
    Joined: Nov 23, 2007
    Posts: 929


    I bet some of the hot Isky grinds for the Flathead Ford would work good in this engine also. You should find some of those specs and profile them.
  11. loudbang
    Joined: Jul 23, 2013
    Posts: 32,131


    Awesome thread. Love reading "stuff' internal combustion engine related that makes you think. And in language everybody should be able to understand.

    Gave you a 5 star rating for your efforts.
  12. Ned Ludd
    Joined: May 15, 2009
    Posts: 4,147

    Ned Ludd

    One thing whose absence from most elucidations of the theory has always struck me as odd is the ratio of maximum lift at the lobe and base circle radius. It would seem to me that where ramp angles are the limiting factor the situation will get easier the bigger the base circle is relative to the lift. That is, if we take a flat flank tangential to the base circle as the hypothetical extreme case, maximum lift would be achieved in a smaller angle of cam rotation the bigger the base circle relative to the lift.

    Of course that makes the entire cam physically bigger, so finding space inside the block or head might become a problem. The other way is, of course, to use a rocker or such as a motion multiplier. In the case of a flathead design - assuming it is physically possible to incorporate such things where the lifters usually reside - the rockers from any of a number of modern OHC engines should give a useful multiplication. One would also have to angle the valve seats and guides to allow for the offset from the cam axis, but this might actually have advantages for the combustion chamber shape.
  13. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Ned Ludd, Kid you not, I was poddering those exact thoughts this morning on the way to work.

    It occured to me last night that for an inline flathead motor, one could fabricate a cam "box" about 4" tall and 4" wide and the length of the block that would bolt to the outside of the engine block where the lifter covers go. The cam would be inside this cam "box" and use modern rocker arms from an OHC engine that uses cam-on-rocker set up. These rockers have a roller that rides against the cam. The rockers would be installed upside down to actuate the vlvs.

    Then, not being confined to the OEM block restrictions on the cam, would there be a benifit to a larger cam, and thus a larger base circle? I don't know, I will have to play around with the idea in the cam program and see.

    The end goal being to dwell the cam at max lift for as long as possible while simultaniously keeping the crankshaft duration of the cam down to a reasonable 230-240 degrees, maybe smaller if the ramps can be steeper. I believe the cam-on-rocker should be able to handle more side loading of the roller than a roller lifter as well, thus a more radical profile should be possible.
  14. nexxussian
    Joined: Mar 14, 2007
    Posts: 3,240


    If you are thinking of running cam under rocker (upside down OHC stuff) the Ecotec stuff is also available from Jesel I believe.

    It should take any conceivable beating you care to throw at it. :)

    As for a flathead grind, my vote is for a Max 1 Isky, they would likely even grind them on for you, but that defeats the purpose of this project. ;)
  15. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Kerry, I would agree that is generally a true statement, but not always.

    The rate at which a roller opens or closes refers to the velocity of the vlv train. A large peak velocity doesn't necessarily mean a large acceleration is required. For example, lets say we wanted to drive 50 miles in one hour. Now lets look at two different methods to reach this goal. Method 1: gently accelerate for the first 30 minutes to a peak velocity of about 75 MPH then gradually decelerate reaching 0 MPH just as we get to the 50 mile and 1 hour point. Method two: stomp on the gas pedal and accelerate as fast as possible to 50 MPH hold that for one hour then slam on the brakes and come to a screeching halt at the 50 mile and 1 hour point. Both methods went the same distance in the same amount of time, however, the acceleration and deceleration for method one was much, much smaller than for method two, therefore the amount of force to produce method one was much smaller than the force required to produce method two. In the vlv train, it is the spring that supplies the force to decelerate the vlv while opening and accelerating the vlv to close. So if we were to go from a cam profile similar to method two above, and say .400" at a given duration and RPM, to a cam with a profile similar to method one above with say .450" at the same duration and RPM, we may actually need less spring force even though the average velocity increased.

    Spring rate: The spring has two parameters we need to look at, the force at a given compression and the spring rate, these are related but different (I'll admit I will often use the term spring rate, or stiffness, when I should be saying spring force). Mathematically they are related by the equation: (known as Hooke's Law)

    Spring Force = (Spring Rate) X (free length - compressed length)

    Lets look at a scenario, I'll use large unrealistic numbers just to make the math easier. Lets say we had a cam that had a 1" lift and required a spring force of 200 lbf at max vlv lift to keep the lifter on the cam. And lets say the installed height was 3", thus at full open the spring height is 2". Now lets say we choose a spring that has a spring rate of 100 lbf/inch and a free, uncompressed, length of 4". Using the above equation, with the vlv fully open, we get a spring force of 200 lbf (100 lbf/inch X (4" - 2")) and, with the vlv closed, a force of 100 lbf.

    Now lets say we increased how fast the vlv opens and closes by increasing the RPM, thus we have the same max vlv height and same duration, but at the higher RPM the cam now requires 300 lbf at max vlv lift. We can use a spring with the same spring rate it would just have to be longer. The free length would be 5". With the vlv full open the spring force would be 300 lbf (100 X (5 - 2)). But the spring force with the vlv closed would be 200 lbf.

    If we were okay with the increase in spring seat pressure to 200 lbf then great, we can use a spring with the same spring rate even though we increased the rate at which the vlv was opened and closed.

    However, if we want to reduce the seat pressure back down to 100 lbf then we need to increase the spring rate to about 170 lbf / inch with a free length around 3.7" thus a seat pressure around 100 lbf and a full open pressure of around 300 lbf.

    So back to Kerry's statement, in general the cam profiles are not as radically different as the driving scenarios above, thus, generally speaking an increase in the rate at which the vlv train is opened or close will increase the force required to keep the lifter on the cam, thus requiring more spring pressure. And when it comes to springs, we wouldn't have to increase the spring rate unless we wanted to keep the seat pressures down, which we generally do. So yes its a generally true statement, but not always, just depends on the specifics.
  16. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Its my understanding that the reason for using cast cams with steel flat tappets has to do with the bearing properties of the two metals. As we know a good bearing consists of one hard metal and one soft metal. This allows any foreign matter to be embedded in the soft metal and prevent galling of the surfaces. It also prevents galling during start up, before an oil film is established. I would venture to guess that they were having problems with the cast metal holding up under high spring pressures and therefore hardened the cam to keep it from catastrophically failing, but I would bet they wouldn't last very long on the street. I would further venture to say they may have purposely made the cam harder than the lifters, causing the lifters to become the sacrificial wear item. Then the lifters could be replaced every couple of races (if not every race) which would be easier and cheaper than the replacing the camshaft.

    RichFox on the homemade two piece head thread stated "Those cams are to be used with "Chilled Iron Lifters" Which I seem to recall reading about some time ago. I do not know the material properties of the Chilled Iron, but iron is generally softer than steel which would support my above ideas.

    Anyone know the hardness of these lifters compared to the hardened cams?

  17. Devin
    Joined: Dec 28, 2004
    Posts: 2,352

    from Napa, CA

    Wow! This is a very ambitious undertaking. I can't wait to hear how this turns out.

    Posted using the Full Custom H.A.M.B. App!
  18. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Did some playing around today with the notion of larger base circles to test the idea that a larger base circle will allow for the same lift with less steep ramp angles and thus shorter duration.

    The lobes in both of the attachments use a .750 roller diam, dwell the vlv between .430 and .432" for 54 camshaft degrees, and use the same spring.

    The cam in the first attachment has a .5" radius base circle. The smallest crankshaft duration that I could achieve, and still have reasonable forces (at least axially to the lifter) was 252 degrees (from the lift graph, the blue vertical line that crosses the .050" line is 63 cam degrees before the lobe centerline. To calculate crankshaft degrees, 63 X 2 (for both sides of the lobe) X 2 (because the crank turns twice as fast as the cam) = 252) We can also see that the flanks are substantially inverted, I very much doubt this would work, way too much side loading of the lifter.

    The cam in the second attachment has a 1" radius base circle. With the same lift and dwell characteristics, this cam easily achieved 228 crankshaft degrees and the side loading of the lifter is not nearly as bad as it is with the .5" base circle radius. As a side note, this lobe would require slightly more spring.

    So, it appears that a larger base circle will allow either higher lifts with the same duration or shorter duration for the same lift. Although this is a roller design it probably holds true for a flat tappet as well.

    It should be noted that the cam in the second attachment would have to have cam journals of nearly 3", quite an increase for an extra 24 crankshaft degrees. I wounder what the increase in torque would be and how much this would shift the power curve down, and the increase in dynamic compression ratio? interesting stuff.....

    With a flat tappet the linear speed at which the cam slides across the bottom of the lifter is greatly increased with the larger base circle, not sure how much of an effect this would have on heat production or wear characteristics. I have always wondered why the cams are so small, I mean when we think about the amount of force from all the vlv springs and the forces due to accelerating the vlv trains, we're talking hundreds of pounds stock and 1000s of pounds for high perf, exerted between cam journals on a cast shaft whose small diameters (between lobes) is ~3/4"!!! Why didn't the manufactures make larger diam cams? I wonder if it has to do with the linear speeds against the lifter and the wearing of the materials, esp. in the early years. Anyone know why cams are the size they are?

    half inch base circle.png

    1 inch base circle.png
    Last edited: Apr 6, 2018
  19. PackardV8
    Joined: Jun 7, 2007
    Posts: 890


    The best defense my own kids had was the really, really bad things other kids did. Their transgressions paled by comparison.

    It's so wonderful when a guy like you shows up. When my wife or my friends give me grief about getting too far out on a project, I can say, "You think I'm doing it the hard way with obsolete engines? Here's a guy on the H.A.M.B. who's designing and building his own flathead 6-cyl roller cam! Now that's really nuts!"

    Packaging and cost. The OEMs use as little iron in as small a space as it takes to live a reasonable life span. FWIW, the Gen I OHV8s most had problems because the increased slop of pushrods and rocker arms and larger valves and higher RPMs loaded the cam/lifter interface more than they were used to. All the '51 Stude V8s had to have the cams replaced. The first Y-blocks didn't have enough oil to the rockers and the passages plugged up. Chevy had to redesign the rear cam journal. Even the '65 Chevy big block had valve failures. The large, heavy canted valves required completely rethinking the valvetrain.

    By comparison, flathead valve trains are cake. There's no slop, no going around corners, valves are small diameter, tappets are often large diameter, RPMs are low, most are gear driven.

    jack vines
    Last edited: Sep 6, 2013
    gas pumper likes this.
  20. ronnieroadster
    Joined: Sep 9, 2004
    Posts: 802


    This is good stuff.
  21. CNC-Dude
    Joined: Nov 23, 2007
    Posts: 929


    The smaller the base circle is with the profile getting more aggresive it is harsher on the valvetrain. Thats one of the advantages to the Nascar engine builders and others using 65MM roller bearings for the cam journals because they can then make the same profiles or larger on a bigger base circle and having smoother motion on the valvetrain for 500 miles and be more easier on parts. While your on a roll, go ahead and bore you cam tunnels out in your block and use roller cam bearings.
  22. gary terhaar
    Joined: Jul 23, 2007
    Posts: 656

    gary terhaar
    from oakdale ny

    Just a thought,when I built race Harley engines we would change the valve train around so much we needed to phase the cam events completely over. Many instances there were no cams available.
    We would order blank lobes of preferred profiles and have the existing lobes ground off the cams.
    Press on the new lobes and phase them in and weld them at the base circle to the shaft.
    Now you can change Lsa as well as opening and closing events without having several cams ground.
    Look at a 4.7 Chrysler,pressed lobes on Dom tubing.
    Now find a lobe you like,ud Harold or comp then cut the cam up then bore and hone the lobes wala a cam kit.
    A little time consuming but you will degree each cylinder independent allowing for lifter variables and other things we have little control of.
    Alter timing events for cam twist,now we are getting carried away.
    Last edited: Sep 6, 2013
  23. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Thanks for the info, those are new to me, I have thought about running larger cam bearings.

    Anyone know just how thin the cast iron in the block can get when over-boring the cam journal holes and still work? I'm talking about the radial thickness of the casting out side of the cam bearing.

  24. CNC-Dude
    Joined: Nov 23, 2007
    Posts: 929


    Post some pics of the bulkhead around the tunnel area and lets take a look.
  25. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    Can't, my shop is packed up for my upcoming move to VA.

  26. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    For all who want to play around with my cam program: (I haven't written a "Help" section yet so I've included instructions here)

    Attached is the cam program thus far. It is a Java program that can be run from the command prompt (I still need to learn how to package it up and make running it user friendly, any Java guys out there that know how to do this? pls let me know so I can share this program and make it user friendly) Most computers come with the necessary Java Runtime Environment (JRE) already, but in Windows you need to "tell" Windows to "link" the command prompt to the JRE. There is plenty of documentation on-line of how to do this. I don't run Windows anymore, but did it years ago.

    When you unzip the file there will be a directory called "Cam Project." Inside of it there are two directories, "Cam Files" and "Cam Program." Cam Files has various files that I've created just as test files for lobes, springs, etc, none of the values, ie mass, spring rate, etc. are actual values. (except SBC roller lifters have a .750 diameter)

    The attachment "Command Prompt" is a screen shot of the Ubuntu version of the command prompt and the instructions I had to type to run the program. Note, when there is a space in a directory name or file then a "\" is used before the space to tell the computer that there is a space in the name.

    There is a lot of documentation on-line of how to run Java Programs from the command prompt so you should have no problems googling it (now being able to follow what some of those guys write....well...)

    Once you get it running, the "Engine and valve train specs" tab will be on the screen first. When the program opens, all options are available to input data (OHV, triple springs, different springs on intake and exhaust, different rocker arms, etc) In general, the engine won't need all the various options, under the File menu you can select various options that will reduce the amount of stuff on that first tab. For example, if you check "Flathead" all the rocker arm information goes away, if under "Spring Files" you check "All Springs Are The Same" then the exhaust spring data goes away and the spring data just becomes "Spring Specs." The same is true for F-head, single, dual, triple springs, same rocker arms, etc. Note, if under "Cam Lobe Files" you select "Single Pattern Cam" there will still be an Intake and exhaust tabs, that's because, in general, at least the valves will be different and therefore the graph of Acceleration/Force will be different for the intake and exhaust. In this mode, making changes to the lobe in either tab, Intake or exhaust, will automatically update the lobe in the other tab.

    To start, open an existing intake or exhaust lobe from File -> Cam Lobe Files -> Open Intake (exhaust) Lobe then navigate to where you put "Cam Project" , "Cam Files" , "Cam Lobes" and select one of the lobes.

    To start a lobe from scratch, click on the menu "Intake (Exhaust) Options" -> "Create Circular lobe." Then click a point on a lobe and drag it outward to make the cam lobe.

    Some pointers / options / known bugs:

    - To shape the lobe, click on a point in any graph to highlight it blue. In the Cam Profile Graph, click and drag the point for course movements, usually when starting a new lobe. Then you can use the arrow keys for fine adjustments. You can change the amount by which each key stroke moves the point under "Tools" -> "Set Arrow Key Increment." I usually start at .010" then .001" then .0005 then .0001, as I'm tweaking the lobe. The arrow keys move the point the direction you would expect, up, down, left, right. Even if you highlight the point in a different graph, say in the lift graph, you are still moving the point in the Cam Profile Graph, up, down, etc which has an effect on the same point in the lift graph, but you are not moving the point in the lift graph directly.
    - Zoom in and out with the mouse wheel (easy way) or under "Tools" -> "Zoom settings" -> "Zoom in (out) with mouse click." (a left click in the fields of the graphs gives this option as well but it is currently broken) Once a point is highlighted in the Cam Profile Graph, then zooming in and out will zoom the other graphs in and out so that only the portion of the graph that is displayed on the Cam Profile Graph is displayed on the other graphs. This is true even when zooming in and out on the other graphs. If the highlighted point goes out of view, in the Cam Profile Graph, the other graphs go to their default zoom until the highlighted point is brought back into view. All the graphs can be moved by clicking on any white portion and dragging the graph around.
    - center the lobe at 180 degrees not 0 degrees (standard x-y graphing degrees), in other words, to the left of the vertical axis. (the mathematical model is a linear representation of a circular object that wraps around the lobe 2 1/2 times thus minimizing any mismatch at the 0 degree point, but I've had troubles with placing the lobe on the right side at 0 degrees) It shouldn't really matter where you put the lobe center line because the computer calculates the lobe centerline (that's the green line) by making the area under the lift graph equal on both sides of the lobe center line.
    - Acceleration/Force graph, positive values are the forces the cam exerts on the bottom of the lifter. Negative values are the forces the spring has to apply to keep the lifter on the cam. (Strictly speaking, the positive values displayed is only the force, along the axis of the lifter, required to overcome the inertia of the valve train, it does not include the extra force from the spring, nor does it include the extra force on a roller lifter that comes from the side loading of the lifter, latter I'll have to include graphs of these other forces, so we will know how much force the cam will really have to exert)

    - play around with right clicking on the graphs and their header bars and on the blue lines that you or I have created. There will be options to change the units, where appropriate, to insert, delete, and move the blue reference lines, etc. On the Acceleration/Force graph there is a curved blue line for the spring force, this will only come off the horizontal axis, and thus be useful, if the units "lbf" are selected.
    - In the window "Cam Profile Graph," if a section of the cam lobe turns red, that means the lifter, or roller on the lifter, cannot follow the cam there, (the inverse radius of the lobe is smaller than the radius of the lifter) On the rest of the graphs, this area will disappear and there will be sharp spikes on the Acceleration and jerk graphs, as one would expect with a large "dent" in the cam lobe, its like hitting pot holes in the road. If a section turns orange, that means the grinding wheel cannot follow that section of the cam lobe.
    - If the spring is not strong enough, a curved red line will appear on the Lift graph. This is the actual path the lifter and valve will follow as they leave the cam lobe.
    - Select which graphs you want displayed in "Intake (Exhaust) Options" -> "Displayed Intake (Exhaust) Graphs"
    - You can have a symmetrical or asymmetrical lobe. Under "Intake (Exhaust) Options" -> click on "asymmetrical", "Match lobe below X-axis to above X-axis" or the other way around. There is a glitch, you have to be in Asymmetrical mode when opening a lobe file, open the lobe file, move a point a little bit, then switch to one of the symmetrical modes.
    - you can add and delete the "control" points, those are the squares on the graphs, by right clicking the cam lobe line to add a new point or right click the point to remove it.
    - If you want to change the base circle, you can select points to be "on the base circle" then change the base circle on the Cam Specs tab. Currently the base circle is the only thing that can be changed on this tab, the cam lobe separation angle is set at 110 and the advance at 0 (to be changed latter). The rest of the info is calculated from the lobe / lifter combination. Glitch - it only displays one crankshaft duration not separate Intake and exhaust, so the last lobe to be modified is the crankshaft duration that is displayed.

    Its not finished yet but I hope you guys enjoy this, comments and recommendations are welcome.


    Attached Files:

    Last edited: Sep 11, 2013
  27. 27troadster
    Joined: Mar 14, 2013
    Posts: 105


    I thought about doing this but for a different reason...On sixes the dist gear is in the center of the cam. I don't have a gear cutting machine.

    Some options I've come up with are:
    - two piece cam that is pressed together with a cast gear in between
    - turn the rear half of the cam to less than the base circle, then press a cast gear onto the cam, press all the cam lobes and journals aft of the dist gear on, welded them in place, then grind the cam.
    - relocate the oil pump and dist. and drive it off the front of the engine. But in a cool early Franklin kinda way, NOT a modern cog belt.
    - break down and have a cam manufacture cut the gear into a steel billet for me. Then find a bronze gear that can be modified to fit the Ply.

    Anyone got ideas on the best way to go about this?

    Gary, what is Dom tubing?

    Last edited: Sep 11, 2013
  28. You know all this cam tech and don't know what drawn over mandrel tubing is?
  29. Pete1
    Joined: Aug 23, 2004
    Posts: 1,838

    from Wa.

    Does a brain surgeon know how to put a set of railroad locomotive wheels in a lathe and turn them round??
  30. If he's going to be building the tender, yes.

Share This Page

Register now to get rid of these ads!


Copyright © 1995-2020 The Jalopy Journal: Steal our stuff, we'll kick your teeth in. Terms of Service. Privacy Policy.

Atomic Industry
Forum software by XenForo™ ©2010-2014 XenForo Ltd.