19 replies to this topic

[hydra]

This is something I've been thinking of for a while now. Searched all the camshaft books at my local university library without much success... An internet search proved just as fruitless... So just how are cam lobes ground? Here is my (limited) understanding of the process

I assume one would have to start with a lobe design in mind (I've gotten this far already) and a target base circle.

The way I understand it, the next step is to make a master lobe. In the old days, this used to be made from plate with a series of discrete plunge cuts, and then smoothed/dressed by hand. Nowadays, a CNC does the job faster and easier

Once the master lobe is finished, it can be installed in the cam grinding machine, which then proceeds to duplicate the lobe profile, and eventually spits out the finished cam...

This leaves me with a few questions:

What about grinding wheel diameter limitations (especially with modern high-performance concave/inverted flank lobes?

I JUST read about a CNC cam grinder that doesn't need a physical master lobe to be created. Is that fairly commonplace tech, or is it still an emerging technology ($$$)?

You can't just CNC a cam lobe directly (machine it as opposed to grinding it) can you?

Hopefully somebody here will be able to shed some light on this subject and give us a chance to learn a thing or three!

[rhm]

As far as I know (and I'm far from being an expert on machining)...

You grind cams and other things like crankshaft journals because they're hardened, not because grinding is necessary to achieve the required smoothness. You can achieve the required smoothness using turn-milling - that is using an end-mill to machine the surface while the shaft is turned slowly (and the head moved up and down in the case of a camshaft).

See http://www.wfl.at/pr...&UID=&Lg=2&Cy=1

I've read that if you can machine the part accurately enough in advance you don't need to re-grind it after hardening. Alternatively you can apply a surface coating like DLC.

Automotive parts will likely still be ground because it's cheaper to go striaght from a heat treated casting or forging and grind it than it is to machine it first.

[McGuire]

Originally posted by hydra
What about grinding wheel diameter limitations (especially with modern high-performance concave/inverted flank lobes?

Inverted or re-entry lobes are done one of two ways: on a machine with two grinding wheels (big one and little one) or with an abrasive belt fixture (a sort of high-precision belt sander) that can alter its radius on the fly.

I JUST read about a CNC cam grinder that doesn't need a physical master lobe to be created. Is that fairly commonplace tech, or is it still an emerging technology ($$$)?

That is a very common technology anymore. Most decent-sized cam makers these days will have both types of machines... you can see how the masterless process would be ideal for prototypes or short runs. Of course, an automotive cam grinding machine is essentially a specialized cylindrical grinder.

[McGuire]

This is what a typical master cam looks like, if anyone is interested...



Back In The Day, cam lobe profiles were generated by combining three circles of varying radii: a base, a nose and a flank, which were then turned onto a master or ground directly onto the shaft. Fairly primitive but it worked. Then "polydyne" and "splinedyne" profiles evolved, and polar coordinate methods. Today it's done with software and the computer does all the heavy lifting.

[Hans Derbe]

Here is a short video of modern cam manufacturing.....

video


Hans

[Joe Bosworth]

Hydra asked:

"So just how are cam lobes ground?"


Hans, thank you for your reply, we all really did know the obvious answer:

On a cam grinding machine!!!

Duh


Regards

[Stian1979]

Originally posted by rhm
As far as I know (and I'm far from being an expert on machining)...

You grind cams and other things like crankshaft journals because they're hardened, not because grinding is necessary to achieve the required smoothness. You can achieve the required smoothness using turn-milling - that is using an end-mill to machine the surface while the shaft is turned slowly (and the head moved up and down in the case of a camshaft).

See http://www.wfl.at/pr...&UID=&Lg=2&Cy=1

I've read that if you can machine the part accurately enough in advance you don't need to re-grind it after hardening. Alternatively you can apply a surface coating like DLC.

Automotive parts will likely still be ground because it's cheaper to go striaght from a heat treated casting or forging and grind it than it is to machine it first.

I think I have a week memory of being told that the hardening proces of the cams are done after they are machined.

[hydra]

Stian,
Heat treating usually causes a bit of distortion (unless its relatively low-temp nitriding) and so should not be left to the very end.

Joe,
Took you all of 5 days and 5 posts to come up with that? Wow, I'm impressed!
Don't get me wrong, I'm perfectly capable of taking a joke, its just that your comment wasn't even in the same ballpark as funny... Please, do us all a favor and try harder next time

[McGuire]

Camshafts are typically induction or flame hardened, carburized, nitrided or tufftrided (molten salt-bath nitriding). These processes will tend to make the part grow a bit, so this heat treating is done before grinding. However, if there are gears integral to the shaft, these may get their own additional heat-treating step after the hobbing and grinding phase. Additionally, camshafts may get some kind of final surface treatment such as Parkerizing to help with break-in.

[darin]

Electronic control of of the grinder is a fairly straight forward. The position of the grinder is commanded by the angle of the cam . The control system is called a Electrocam. About 20 years ago I designed the controller for what I think was the first electronically controlled system using a sanding belt. General Motors had invented the belt sander technique because they wanted to put a reflex in the cam allowing the valve to close faster. A traditional grindstone has to large a diameter and so ( I can't remember his name) had 3M develop the belts and abrasives. Great idea, I don't think GM gets half the credit it deserves for engineering excellence .

Dick

[Stian1979]

Originally posted by hydra
Stian,
Heat treating usually causes a bit of distortion (unless its relatively low-temp nitriding) and so should not be left to the very end.

Joe,
Took you all of 5 days and 5 posts to come up with that? Wow, I'm impressed!
Don't get me wrong, I'm perfectly capable of taking a joke, its just that your comment wasn't even in the same ballpark as funny... Please, do us all a favor and try harder next time

I ben doing some work in gear manufacturing and normaly heat treatening is posible to compensate by the meshurements you machine the components preor to heat treathening.

I don't say this is the rule since it's marine and indestrual gears that I'm exsperienced with.
Machining preor to heat threatening is cost cutting because it decrease the wear on the tool's

[hydra]

Remember, the max rpm of the engines you work with (if I remember correctly) is very low, on the order of a couple hundred rpm, this reduces the importance of camshaft accuracy as you're not even close to being valvetrain limited.

Secondly you've got scale effects coming into effect here as well. A given error on a slow-speed industrial/marine camshaft will have much less of an effect as the same error on a high-speed automotive camshaft

[Stian1979]

Originally posted by hydra
Remember, the max rpm of the engines you work with (if I remember correctly) is very low, on the order of a couple hundred rpm, this reduces the importance of camshaft accuracy as you're not even close to being valvetrain limited.

Secondly you've got scale effects coming into effect here as well. A given error on a slow-speed industrial/marine camshaft will have much less of an effect as the same error on a high-speed automotive camshaft

60-120RPM or 600-900RPM is the engines normally involved.

[hydra]

Okay,
I'm back with another crazy suggestion. A machinist I know seems to think that automotive-grade cam lobes can successfully be EDM'd (Electrical Discharge Machining), and that the process has excellent accuracy and surface finish. Good enough to make molds and gears in fact... What do you guys think?

[Halfwitt]

Originally posted by hydra
Okay,
I'm back with another crazy suggestion. A machinist I know seems to think that automotive-grade cam lobes can successfully be EDM'd (Electrical Discharge Machining), and that the process has excellent accuracy and surface finish. Good enough to make molds and gears in fact... What do you guys think?

It might be OK for industrial cams, but I doubt it would be good enough for serious automotive work or racing cams.

I spoke with a racing gear manufacturer earlier this year and bought this subject up. He said he couldn't see a situation where it would be worth making gears like this, unless you were in a massive hurry and went the through hardening route. But then again, a through hardened gear is unlikely to be found in a serious racing engine or gearbox if minimum weight is a priority.

[hydra]

I don't mean to disagree with you, but what makes you think it wouldn't cut the mustard?

[McGuire]

Originally posted by hydra
Okay,
I'm back with another crazy suggestion. A machinist I know seems to think that automotive-grade cam lobes can successfully be EDM'd (Electrical Discharge Machining), and that the process has excellent accuracy and surface finish. Good enough to make molds and gears in fact... What do you guys think?

I suppose you could, the question being why you would.

[Halfwitt]

Originally posted by hydra
I don't mean to disagree with you, but what makes you think it wouldn't cut the mustard?

I don't believe that the metal removal rates would be anything like fast enough for series production where I imagine finished cams are produced in a couple of minutes, and I don't believe for racing you could achieve the surface finish or accuracy required. Furthermore, EDM produces a very thin recast or remelt layer which can have poor properties and is often in a state of tension and would thus require grinding.

Perhaps EDM might find application on cams which are nitrided or carburised as an alternative to the roughing operation on the lobes, but I doubt it can compete with the established methods.

[Engineguy]

Originally posted by McGuire
I suppose you could, the question being why you would.

I think you're being generous with the "could":

With wire EDM you could cut a cam lobe but not a camshaft, for obvious reasons.

With plunge EDM you need to machine a carbon negative of the surface you want to generate... typically that means a male carbon electrode to plunge into the metal to create a cavity. To make the carbon for a camshaft you'd have to machine two female cavities in two carbons (very difficult for the shapes involved) and plunge one side of the cam then the other, maintaining perfect registration accuaracy and a perfect 180° index throughout the process. Good luck with that.

I suppose you could plunge a (lobe width) square electrode in a machine that has X, Y, and Z travel and a precision rotary 4th axis to index the cam, and program it to work just like a cam grinder. the end of the electrode could be convex'd to mimick any size grinding wheel. Probably a custom built machine and a lot of trial and error to get the process parameters right.

[McGuire]

Originally posted by Engineguy
I think you're being generous with the "could"

True. It could be done - that is, it's hardly impossible - but I can't imagine how it could ever be made better, cheaper or faster than grinding.

The cheap way to make camshafts these days is composite or assembled construction. The lobes are sintered or coined from sheet and then fixed on a shaft or tube.