Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Cam making

  1. Apr 3, 2006 #1
    Why are camshafts normally grinded? Is diecasting and then welding camlobes to a shaft too expensive and/or does it give less strength?
     
  2. jcsd
  3. Apr 3, 2006 #2

    FredGarvin

    User Avatar
    Science Advisor

    First two reasons that come immediately to mind:
    - Dimensional control
    - Surface finish
     
  4. Apr 3, 2006 #3

    brewnog

    User Avatar
    Science Advisor
    Gold Member

    Even if you die cast and fabricated a camshaft, and found some obscure way of ensuring that your ordinates were absolutely spot on (doubt you'd achieve this with any welding process), you'd have nowhere near the surface finish needed.
     
  5. Apr 3, 2006 #4
    What do you mean by dimensional control? Forgetting other factors for a movement, can't cam lobes be made in dies since cams only have a 2d profile, not a 3d one.

    Can't surface finish be done through grinding after the casting? That would save a lot of grinding.
     
  6. Apr 3, 2006 #5

    Cliff_J

    User Avatar
    Science Advisor

    I think just the wide variations in the design of camshafts (I mean more then just lobe ramp rates or seperation angle but basic dimensions like length or spacing and even center circle diameter) would make the casting process very complicated with little gain to offset the increased workload. Even if you could ship it out to less expensive overseas labor, it seems more expensive and error prone than just grinding down round blanks that can be checked for runout before grinding (and melted and recast if out of spec) rather than trying to assemble 2 dozen pieces. I can't see grinding time being that expensive.
     
  7. Apr 3, 2006 #6
    You mean that due to the wide possibilities using grinding (not to mention its uses elsewhere too), grinding comes out to be overall cheaper than dieing?

    How does one calculate the cost of grinding and the time taken?
     
  8. Apr 4, 2006 #7

    FredGarvin

    User Avatar
    Science Advisor

    Having a 2D profile has nothing to do with it. When I say dimensional control, I mean in the sense that most rotating surfaces such as those found on cams running locations and bearing locations are held to very tight tolerances that are not achievable (easily) any other way. You definitely will not hold four decimal places (english) out of a die or casting. It must have secondary machining operations after the bulk deformation process. The thing about welding, ESPECIALLY on rotating hardware is that it is difficult to maintain runout on all of the surfaces. You would have to go through rounds of annealing after the weld along with straightening. It would add more work. We weld shaft sections sometimes, especially for one of a kind R&D hardware. However, there is a real art to getting it right plus getting the rotating group to balance in the end.

    Yes. You are correct. I would assume that the final machining is a fine cut grind, akin to a very light lathe cut, to get the surface roughness required.
     
    Last edited: Apr 4, 2006
  9. Apr 4, 2006 #8
    Actually I'm not sure if a cam on an engine camshaft really is a 2D profile. If you pull a valve cover and watch the pushrods while the engine is running you will notice that they spin. There is only one way I know of for this to happen and that for the cam lobe to be ground slightly larger on one side. Usually the lifter has a slight convex shape to the part that contacts the cam. Come to think of it I've noticed the wear pattern on a camshaft and it is usually off to one side.
     
  10. Apr 4, 2006 #9

    FredGarvin

    User Avatar
    Science Advisor

    I have seen industry articles saying that cam machining is getting to be very difficult because of the 3D aspect. I am not sure if the 2D has been done away with. I guess in the context of the original question it doesn't really matter.
     
  11. Apr 4, 2006 #10
    Wait, if the cams are having a 3d profile, wouldn't that mean that the pushrod is at contact with the cam at only a point (assuming pushrods are used in the engine)?
    Wouldn't that make the stress unacceptable?

    EDIT: Just calculated, two spheres in point contact with radius 1 cm and force on valves = 560 N will have a contact stress of 260 MPa. Ofcourse the fact that one of the spheres is actually a rod and friction effects etc. would distort this figure, but my point is that point contact stresses with such high forces are EXTREMELY high not to mention the surface cyclic stresses caused by a spinning rod and a spinning camlobe. Wouldn't such pushrods which have point contact be expensive to manufacture and grind?
     
    Last edited: Apr 4, 2006
  12. Apr 4, 2006 #11

    FredGarvin

    User Avatar
    Science Advisor

    The Hertz stresses are higher than usual, which is why they rods and cams are usually surface hardened and then ground. This also gives you some insight into why things like bearings and gears can be very expensive. The hardness and resulting difficulty in machining is necessary to resist surface fatigue due to the contact stresses. This all also leads into why surface finish is important. I had to go back and look it up, but surface finish has a direct correlation to fatigue endurance limit (the [tex]C_s[/tex] Cs value in my book).
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Cam making
  1. Simple Cam Problem (Replies: 1)

Loading...