[roccitycafe]

question: do newer bikes use an asymmetrical valve motion profile, i.e. is the valve acceleration rate greater on the opening ramp than on the closing ramp? I've measured a few older bikes and the motions are symmetrical, but dynamically speaking, you could use a lighter valve spring and probably have a more durable valvetrain if the lift was asymmetrical. From what I understand though, solving for the BC's of a function that gives asymmetrical valve motion is very computationally intense, and I wonder if bikes from the 60's and 70's didn't design cams that way because the time required to design an asymmetrical system would have been too long.

please actual answers from experience or research only, no speculation.

 

[brad black]

i designed a cam at uni, think it was a third year project. that would make it 1990. we had one cad machine in the whole mech eng dept. used to get used a bit for flight simulator from memory. so it was all hand calcs. wouldn't have been that hard computationally.

the max velocity is based on the lifter diameter. not sure how that relates to curved rockers whcih kind of have an infinite radius i expect - the ducati i deal with all have curved rocker faces, and that alone can change the actual profile as the cam moves across the rocker from open to close.

also, the cams on horizontal and vertical are the same valve profile, but because they come onto the rocker from opposite sides due to the positioning the physical profiles are different.

there's some 4v cam profiles down the page a fair bit here - they used a lot of assymetrical cams. the strada and a profiles originated in the early 80's at least. BikeBoy.org - Ducati Cam Specs

2v profiles here: Ducati 2V cam profile comparisons

the desmo cams can have wacky rates due to the valve control though. there was a 999rs cam that was an oval and one side was concave. wacky as.

sure none of that is actually helpful though.

 

[roccitycafe]

interesting that you designed a cam, what was your process? what kind of function did you use to generate the lift? what functions remained continuous and differentiable? I managed to cob together a matlab program to generate the lift curve by drawing a set of lines for acceleration, then integrate back for the displacement. I am thinking about piecing another function in to round off the corners of the acceleration and make the jerk both continuous and differentiable rather than just continuous... lots of things to consider, I'll probably design a cam for the cb350 first and toss it into a beater bike to see how things perform

 

[brad black]

that was 30 years ago. no idea now. we had a text book about it i guess.

 

[Teazer]

It makes sense to make valve lift symmetrical. If you take an existing cam, plot the lift against degrees of rotation and work out the acceleration and jerk rates and you can tweak those results to see what the lift looks like. That crap used to take forever to measure and calculate but there are measuring and computational tools available now that make it much quicker. I have done that on several race cams to see if I could work out why some grinds had worse wear or stellite chipping than other supposedly similar designs.

Of course, cam profiles in an OHC rocker design are always asymmetrical to allow for where on the curved rocker arm the peak is touching as it rotates.

What would be the purpose of a non symmetrical lift curve? You could use a different closing acceleration rate c/w opening acceleration. You could also keep the valve closer to peak lift at a more advantageous relationship to piston position I suppose. I have not seen or measured any cams that did not have a conventional symmetrical design. Might be worth pulling cams for a 2019/2020 high performance bike and put them on a Cam Doctor and do the calculations to see what's up.

 

[jcw]

question: do newer bikes use an asymmetrical valve motion profile, i.e. is the valve acceleration rate greater on the opening ramp than on the closing ramp? I've measured a few older bikes and the motions are symmetrical, but dynamically speaking, you could use a lighter valve spring and probably have a more durable valvetrain if the lift was asymmetrical. From what I understand though, solving for the BC's of a function that gives asymmetrical valve motion is very computationally intense, and I wonder if bikes from the 60's and 70's didn't design cams that way because the time required to design an asymmetrical system would have been too long.

please actual answers from experience or research only, no speculation.

answer: yes.
why is beyond me, but the simple answer would be a fast opening is desirable for air flow and a slower close to ease the wear on the valve seat.
put it is probably much more complicated than this.

actual experience...
bmwk1300s cam profile

 

[roccitycafe]

yeah but does that k1300 have a bucket follower or rocker arm? you can't tell the valve motion by the cam lobe shape. cb350 has a symmetrical valve motion profile but the cam is shaped a lot like that one

 

[jcw]

yeah but does that k1300 have a bucket follower or rocker arm? you can't tell the valve motion by the cam lobe shape. cb350 has a symmetrical valve motion profile but the cam is shaped a lot like that one

Not sure how that works on a cb350, but on the K1300s, the cam presses on a short rocker arm that directly opens and closes the valves eliminating the weight of the bucket and shim design and some of the sideways forces.

The asymmetric cam lobe directly translates to faster valve opening and more gradual valve closing.

Most of the modern literbikes use this now. Sometimes they are called finger followers.

What is a ‘Finger Follower’?

Kawasaki’s ZX-10R Project Leader on how a Finger Follower reduces weight and increases power | BikeSocial

www.bennetts.co.uk

 

[jcw]

I do not see that asymmetry with the cam in this video...

 

[roccitycafe]

it's not a direct translation though, if the follower is rotating, the motion at the valve stem is different... I could model a symmetrical valve motion back through a rotating follower and show you an asymmetrical cam lobe that's required to generate that motion, if the follower is translating in one axis only, then a symmetrical valve motion would generate a symmetrical cam profile. I've got a program I wrote to model the pitch curve of the cam lobe with a rocker, I'll see if I can change things around to model a short finger follower like that

 

[jcw]

It can be done on solidworks motion as well. Dont know if I can model it but I should try.
I have not doubt this set up has asymmetric valve opening and closing.

 

[roccitycafe]

I just sketched something in SW approximating the setup, the origin is where the follower pivots, if you copy this and play with it, you should be able to see that the contact angle where the cam would be changes quite a lot, meaning the instantaneous lever ratio of the cam to the follower also changes a lot, meaning an asymmetrical cam profile may indeed generate a symmetrical motion at the valve

 

[roccitycafe]

and I'm not saying it's not an asymmetrical valve motion, I'm just saying the lobe shape itself doesn't make it so

 

[jcw]

LOL, it'll take me a week just to remember how to open the program...

 

[roccitycafe]

I'll try to draw a better approximation of that setup, I should be able to put the dimensions into my program and spit out a pitch curve fairly quick

 

[jcw]

and I'm not saying it's not an asymmetrical valve motion, I'm just saying the lobe shape itself doesn't make it so

Then why are most cams lobes symmetrically ground?

 

[jcw]

And if you are saying that a lobe profile somehow changes the speed/distance a valve moves on the "opening" phase on the cam vs the "closing" phase of the cam, then wouldn't it not reason that running a cam forward and backward on the same open OR close phase would create different valve motion which it clearly wouldn't???

 

[jcw]

I can clearly see that to achieve a particular velocity or rate of valve movement, the slope of the cam on the lobe itself will not necessarily be directly related to the valve motion given the geomtery of the cam bearing and bucket surfaces. But whether you are opening or closing, the valve motion itself should be the same for a given cam lobe profile.

Of course, everything in mathematics and physics is not always intuitive...

 

[TrialsRider]

What's the goal here? You guys trying to figure out how to advantage variable valve timing on a budget, or come up with a better way to open and close a poppet valve
You want to get more performance out of an ancient 2 valve per cylinder CB350, put fuel injection on it, that would be money way better spent!

 

[roccitycafe]

it's definitely not intuitive, and the cam lobes that are symmetrically ground are for bucket followers. the hard thing to visualize is that the contact point between the cam and the follower radius moves along the follower radius as the valve goes up and down, so depending on valve lift, the cam has a different rate of valve lift/rate of angular change, the lever lengths of your effective linkage system are changing, it's very impossible to visualize without an animation really...