Rocker contact concepts in practice?

I'm confused, [shush it] about best geometry to seek on where to set rocker contacts. On first glance and general consensus, the practice is to center rocker motion to evenly spread contact wear across the stem face. On second glance the valve is easiest to press against the spring when the rocker first contacts stem to open valve so least need of straight down near center high force vector. Yet closed valve stem is sticking up beyond guild the most so is most unsupported for lateral force at that stage. Is there significant sideways friction of rocker on stem or cap to matter? I don't think so but I don't know so asking. One would think the highest load of rocker on stem would be at highest lift against highest spring resistance, but Jim and others tell us that at hi rpm's, where this issue may really matter the most, the valves are thrown open over the cam peaks more than they are pushed open by the rocker at that point. Then spring pulls valve back hard on rocker contact but rocker is also running away from the valve too, so I don't see a need of centered contact on valve closing. But I'm confused not knowing all the physics and geometry involved > in the range our heavy twins enter their elastic cartoon character. What do the God's do?

hobot said:

One would think the highest load of rocker on stem would be at highest lift against highest spring resistance,

Click to expand...

Until spring coils join a spring has the same resistance all along when pressed , (that's why you use progressive springs on shocks or forks) so same resistance at high lift that low lift . i know this is counter intuitive but it's the way it works

Until spring coils join a spring has the same resistance all along when pressed , (that's why you use progressive springs on shocks or forks) so same resistance at high lift that low lift . i know this is counter intuitive but it's the way it works

Click to expand...

Hm, maybe so in some spring applications, but not in our heads or suspension as I understand it at the moment. Those that run Greg Fault's Roadholder upgrade have a good sense of valve spring character, ie: they spike in resistance as they near coil bind, thank goodness. Assuming my ass-sumptions are wrong please expand on your remark and how it might apply to rocker contact. If spring pressure and side loads are nil issues then centered for even wear makes most sense, as is most common aim, I think.

Not sure I know which kind of springs you have in your head . not being Dixieland native , my Hobot's language ability is sometime limited.

In a coil spring resistance is linear : you need 2 time more pressure to depress it by 2 mm than to to depress it 1mm, but as soon as you depress it by 1mm, the additional effort to depress it by another mm is the same as the effort you already put to depress it for the first mm.

Regarding valves I think the most important part is to increase the surface between valve stem and rocker to avoid wear , and try to keep effort even , so as often on one side than on other. meaning the rocker should be centered on the stem when mid opened. the spring is pushing straight on the valve , no uneven wear due to the spring.

Im offering this as a suggestion - not a law.

On a race bike the rocker should be centred close to valve closing because you want the valve to seat nice and square and not bounce on closing. You are going to maintain it often anyway and heavy wear is the price you pay for going racing.

On a road bike you want it centred in the mid range to minimise wear.

Im opened to be corrected !

Huh JRD, I thought 'our' heads was plain enough I wasn't just talking about my head, sheeze.
I understand constant and progressive spring rates, either way to get springs to compress fully one keeps adding up the incremental increase in force till a maxed out total load at or on the end.

johnm, appreciate the mechanical insight reasoning on hi speed valve closing helped to center in seat with rocker baring on stem center at that point. That was the new factoid I was hoping to learn about here.

So best thots at this point imply, center contact at mid stroke for most wear tolerance and center contact on valve seating for higher rpm tolerance. I plan to run hardened caps on Peel so can just replace as worry indicates. Love learning more and more about less and less on Commandos.

hobot, you centre on valve seating and you'll pay the price in wear and friction as the adjuster ball drags excessively across the valve end. Centred at mid lift is the best compromise. And compromise it is, because the ball end of the adjuster is far from ideal, it amounts to little better than a highly loaded point contact, which is where much of the scoring of the valve end is coming from. Talking solely about the rocker end contact, the ideal is a radiused end to the rocker, and going one stage further, the radiused end would be so designed that the absolute minimal shift takes place across the valve end throughout the total valve lifting and closing sequence. Of course, some other means would be required for controlling and adjusting tappet clearance, but a predetermined pushrod length is a must, then with valve end ground to suit, tappet shims would fill this requirement admirably.
Before slagging off the original designer of the rockers, bear in mind this engine was never intended to be modified in the way some folks do today, and some indeed have very effectively gone a long way towards making a silk purse out of a sow's ear. Long may they continue.

Gentlemen
having re read my last post it occurs to me that I should offer further explanation re the centred rocker. My comments about the centred rocker at half lift still hold good for the standard ball ended adjuster. If this ball ended adjuster is set so that it is centred when the valve is seated, then for the exhaust rocker to lift the valve 9.5 mm, the ball end would drag 1.6mm across the end of the valve. This equates to approx. 20 degrees of arm movement.Centred at half lift it moves approx. 0.4mm
The solid end I mentioned earlier incorporates a specially centred radius, designed so that at all times the contact, which will be across the width of the valve stem end, will remain virtually dead centred on the end of the stem. THEN the rocker can be positioned so that it is centred with the valve closed. Anyone going to this much trouble and still looking to go a touch further - get the rocker pad DLC coated.

Alrighty snotzo, that sure filled me in till over me head. Thank goodness for disposable hardened stem caps.

Valve/rocker geometry must be one of the least understood parts of engine building.Anything you read about roller rockers can't be applied to a Norton.Everyone seems to know about roller rockers,but they seem to know nothing about solid rockers.
There are 3 different types of rocker in common use:
* Solid (including a fixed adjuster screw)
* Swivel-foot (like Porsche or T160 Triumph)
* Roller-tipped rocker

With a fixed (non-swivel) adjuster,the last thing you worry about is the angle of the adjuster.
The second last thing you worry about is the contact point on the valve-tip.
These things can be fixed AFTER you get valve/rocker geometry right.They can even look wrong,when side-thrust on the valve guide has been minimised and optimised.They can look perfect,when side thrust is terrible.Pushrod length can change these things,but it won't affect the side thrust.Only valve stem length can change side thrust.

At some point during valve lift (hopefully),you can take a line from the valve-tip or lash cap square to the valve-stem and it will pass through the centre of the rocker spindle.At this instant there is zero side-scrub between the rocker and valve-tip.Side thrust on the guide has stopped for an instant.At less lift,the thrust is in one direction and at more lift it is in the other direction.If that point is about 1/3 valve lift,all is well for normal touring.At higher rpm it would be better to have it at lower lift,where acceleration loads become higher as rpm increases.

There are some good comments near the bottom of this page from Speedtalk.com
http://speedtalk.com/forum/viewtopic.php?f=1&t=20333&hilit=geometry&start=30
and more here,about 1/3 way down the page:
http://www.thesamba.com/vw/forum/viewtopic.php?t=167378&postdays=0&postorder=asc&start=160

JohnM,

Heinz Kegler agreed with you. He set his 500 Daytona 88 race bike up so that the rocker arm contact was as centered as possible when the valve was closed. His logic was that the highest stress on the valve was on the initial lift off when acceleration was the highest. He said that once the valve was moving and had inertia then the need for the rocker contact to be centered was less critical. So he said the guide was under less stress on initial lift and while slowing down to a stop, having it maybe more centered while seating is another benefit.

Snotzo said:

Talking solely about the rocker end contact, the ideal is a radiused end to the rocker, and going one stage further, the radiused end would be so designed that the absolute minimal shift takes place across the valve end throughout the total valve lifting and closing sequence. Of course, some other means would be required for controlling and adjusting tappet clearance, but a predetermined pushrod length is a must, then with valve end ground to suit, tappet shims would fill this requirement admirably.

Click to expand...

This idea won't work on any non-swivel,non-roller,rocker arm.The rocker must roll across the valve tip,starting closer to the rocker spindle and finishing on the other side of the stem at full lift.The rolling action does not exactly equal the arcing action,but it's close.
The bigger the radius at the rocker tip,the further it rolls.The only advantage of a pallet style rocker tip is the increase in contact with the valve tip (line contact vs point contact for an adjuster screw).

JRD said:

hobot said:

One would think the highest load of rocker on stem would be at highest lift against highest spring resistance,

Click to expand...

Until spring coils join a spring has the same resistance all along when pressed , (that's why you use progressive springs on shocks or forks) so same resistance at high lift that low lift . i know this is counter intuitive but it's the way it works

Click to expand...

Not exactly right. A strait rate spring means its progression is linear.

IE, A 100 LB/IN spring requires 100 lbs to compress it 1 inch- 200 lbs to compress it 2 inch -300 LBs to compress it 3 inches and so on.

A progressive spring with a 100 lb/inch initial rating could mean it requires 100 lbs to compress it 1 inch -250 lbs to compress it 2 inches - 450 lbs to compress it 3 inches and so on.

The maximum load on a pushrod is just before maximum lift and just after maximum lift. At the point of maximum lift the inertia of the valve train tends to unload the valvetrain for a moment as the parts are reversing their direction. Jim

comnnoz and beng touched on two important factors, spring rate and acceleration. Another factor to consider is range of engine speed you should be designing to.

Generally, valve spring design is to ensure performance throughout the range but also to protect against valve float at peak rpm (presumably with some rpm safety factor).

From what I can tell, with lower engine speed ranges (street use), spring rate would be more of a factor whereas higher engine speed range (race use) acceleration becomes more of a factor (and spring rate remains a factor).

So the sweet spot (appropriate location of the contact) should be based on where the engine speed range will be.

comnoz said:

IE, A 100 LB/IN spring requires 100 lbs to compress it 1 inch- 200 lbs to compress it 2 inch -300 LBs to compress it 3 inches and so on.

A progressive spring with a 100 lb/inch initial rating could mean it requires 100 lbs to compress it 1 inch -250 lbs to compress it 2 inches - 450 lbs to compress it 3 inches and so on.

The maximum load on a pushrod is just before maximum lift and just after maximum lift. At the point of maximum lift the inertia of the valve train tends to unload the valvetrain for a moment as the parts are reversing their direction. Jim

Click to expand...

What you say is true at very low rpm.
For example,a valve spring could have 100lbs seated load and 200 lbs at full lift.
At high rpm,close to valve-float,that 200 lbs force is decelerating the valve from about .090" lift up to full lift and there is almost no load on the cam and tappet.There will still be almost no load on the cam until the valve returns to 0.090" lift.The cam will then have to decelerate the valve gear.That final deceleration is approximately 3.5 times as much as the valve spring could manage with only 200 lbs force.

At that point,the load on the pushrod is much higher than 200 lbs.It could be 500 lbs or more.At high rpm,the biggest forces on valve gear happen near the opening and closing points.

Further to x-files and my post on the matter, the forces due to acceleration vary with the square of the engine speed so it is tricky to determine best geometry, especially when designing for over an rpm range.

Interesting stuff.