Valve/Rocker/Pushrod Geometry

[#84Dave]

There is a rumor that Circle Track magazine will very soon have an article or two on extensive coverage of proper valve, rocker and pushrod geometry. Based on some of the proper methodologies of Miller Engineering. You pro's likely have been doing it correctly for quite some time. We hobbyists and weekend warriors are about to learn what many of us have been doing wrong...... some of us for 35+ years.

Dave

[Ed-vancedEngines]

Well, I have noticed that if you set your rocker arm geometry according to either T&D or Jesel and then also check it with the Miller theories that they will all hold to the same settings. I think that Miller was plagurized by Lunati Cams possibly but the Comp Cams instructions will not work on any serious engine. I do not understand why cam companies keep on giving bogus tech info but they do.
Ed

[#84Dave]

Ed..... your last sentence is the reason the CT Senior Tech Editor apparently decided it was time to put the issue to rest, once and for all. I have worked 10's of hours on the benchtop with the latest instructions from Dan on his shaft/stand system on my Brodix-10x's. Drawing pictures, moving things around. Close but no cigar. I have not seen the T&D instructions. ME and Lunati actually worked together a very few years to present the situation properly, but apparently that didn't work out for the total good either. I spent an inordinate amout of time moving stand heights around and measuring linear valve movement versus the crank shaft position. Quite frankly, I was stunned at how much 'area under the curve' can be lost with even a moderate misstep on the adjustment. You gents probably know all the shortcomings, but this weekend warrior had no idea how bad the situation could actually degenerate if it wasn't right. Now knowing what's required, hopefully 100 laps on a paved oval short-track won't break a bucket of parts. I have no current experience with stud-mounted systems, but I suspect the same percise requirements also exist for that type of valve actuating system.

Dave

[highVE]

Dave, you said you have no past experiance with stud mounted valvetrain? well consider youself fortunate! On a high end engine there SHOULD be a shaft set-up. but some still utilize stud mounted. and if not setup correctly.....there will not only be valvetrain failure mut evan more power loss that you can imagine. I've found more and more head castings don't evan have the rocker stud hole in the proper location. It's impossable to get a perfect Geometry! you'll need offset trunions when you shouldn't have to. Pushrod lengths that are only available if made custom and so on. the novest can't set up a perfect valvetrain (i use the word perfect with caution) they only now what to look for from what they've read in car mags or like Ed said, info availabe from comp! Some may take this the wrong way, but it must be said. some people just don't belong inside an engine! I'm not a tranny guy, and don't poke inside them for that reason! Leave the work to the pros, and pay for quality work and craftsmanship. Parts or company names get a bad rep from the novests whos motor blows up cause they don't know what there doing. But in the car mag it "seemed" ok. I may have got offtrack a little and i apologize but i often cringe when i here a mag is writing an article about somthing as important as geometry. Next you'll have the customer questioning YOUR work.

Mike

[#84Dave]

Mike...... I understand your points completely. I have been attacking the valve geometry situation lately like an ole dog shaking a shoe. Even talking to a very few professional engine builders who perform their services for their complete living. And many are going about geometry all wrong. Part of the problem I've run across, for instance? Renowned parts distributors who supply pushrods only in .050" increments. That's not good enough. Smith Bros. in Bend, OR, will manufacture the pushrod within .005" of your specified length. Hardened 3/8" for about $13/ea for a quantity of 20. That's the kinds of things I've been chasing recently. We all began at the novice level in this field. Except for a couple of mechanical engineers I know whose business is race engines. The tech mags can present it right if they get alongside somebody that knows some truth to the matter. Miller Engineering, for instance. And I wouldn't know Jim Miller if he delivered a pizza to the front door! I've taken a rabid interest in the subject, subsequent to working on assemblies someone else put together. A few of them race-winning oval track engines! But it's not their fault. They use the instructions provided by cam and/or shaft/stand manufacturers which can be, and often is, misleading and/or simply wrong. The same techniques that have been wrong for 40+ years. Your point is valid regarding a 'perfect' geometry. Manufacturing tolerances and lousy hardware can make it very tough at times keeping the geometry even in the ballpark. But when I find the stand heights on a shaft/stand system that were .123" too low for the cam installed in the engine, but the geometry matches the manufacturers instructions/gauges, something is wrong with the info being provided out there. That situation needs to be corrected. Irregardless of the correction method used.

Dave

[Ed-vancedEngines]

Dave,
I will give you a quick example;
The info supplied by any cam company will work and is a get-by with any mediocre engine with mediocre lifts. Even the stupid and assanine Comp Cams instructions will work.

Hmm' Let's position the roller so that it begins it's travel across the valve stem on one side of the valve stem biased toward the intake manifold side and at mid lift the roller should be right in the middle of the valve stem and as lift increases to full lift the roller should be then on the opposite side of the valve stem. That is so totally stupid of advice I have read about proper valve train geometry. That advice is really good for wearing valve stems, and valve guides. It will cause the valve to be pushed toward one side and then the other side except in the mid lift position. If that was tried with 1.000 inch lift or above it would result in instant failure.

Another problem I see is intake and exhaust valves that are too long in 23 degree sb chevy heads. That does create a valve train geometry nightmare because of the different angles of all the components that must work together. Off set rocker stands or trunions do help in these instances.

I use supplied guages from T&D and also measurements based on Miller's Mid-lift Principles, and also use marking devices to mark the top of valve stems and I also place a thin mark with sharpie at the end of the rocker tip and roller to watch also. I want the roller to begin it's travel just barely to the intake side of the stem and to roll across the center and when at maximum lift for the roller to be in the same position on the stem where it started. While it goes through the full range of travel I like a very tiny sweep pattern on the top of the valve stem and when watching the end of rocker-roller line I placed I like to see the roller to barely move. That will work and will give you a more accurate and full lift travel at any size of valve lift. That is how I set up in brief description, geometry for extreme lifts. Once you learn to set up geometries that will live at 1.00 inch lift and 10,000 rpm you will know you have it right.

If I am using studs for rockers It is still set up similar and you can buy studs of different lengths when necessary.

Gear marking compound works good when lightly applied to see the positions of sweep ot travel.
Ed

[#84Dave]

Ed...... your 2nd paragraph is a perfect example of 'over-arcing'. Valve guides take a tremendous beating with an over-arcing situation in the geometry. Your 4th paragraph explains proper geometry. Wherever the roller is positioned on the stem tip at CLOSED valve, that should also be the roller position on the stem tip at full OPEN valve, for proper geometry. Swinging through the arc between those points will cause the narrowest sweep of the rocker roller across the tip toward the exhaust side, which imparts the most of the cam lobe 'command' across the rocker ratio to the valve. Minimum lost linear valve movement and about as good as it gets. Your comment on valve length? Another comment, which you're likely aware of. On my typical paved oval Chev engines with a stand/shaft system. I'll have about .700" NET intake and .660" NET exhaust lift. Half that NET is .350 & .330 (Mid-Lift). So there's obviously a .020" delta @ mid-lift. I always set the proper geometry by having the exhaust valve .020" shorter than the intake @ full closed. That's why pushrod lengths in .050" increments are a no-no for my use. If I used them in the above example, I'd re-introduce error back into the valve geometry. And finally...... I have a whole bag of extra .025/.050/.100 Jesel shims. When required, which is most of the time, I'll machine the thicker shims to get the proper stand heights. I try to get within .005" @ final installation.

Dave

[andyf]

Minimum travel is one way to set up the geometry. The other way is to minimize the product of travel and load.

When you minimize the product of load & lift you end up with a geometry that travels sideways more at low lifts but as the lift gets higher the side travel is minimized.

I've set up rocker arms using both methods and haven't ever seen a difference but I don't build high speed engines either. Most I've run one of my big blocks at is 7500 rpm.

[#84Dave]

Andy...... if I understand you correctly, by 'sideways' do you mean travel of the rocker tip across the tip of the valve stem? If so, a lot of sideways at low lift and less sideways at higher lifts means you'd be 'under-arcing' the valve geometry. If my old brain is thinking right, you'd be losing opening valve rate/velocity at low lifts because some of the expected linear valve opening is wasted through 'sideways' motion across the valve tip end at low lift. Again, if I'm thinking correctly.

Dave

[Sir Yun]

interesting reading.

for my purposes i'm not quite shure what the actual powergains would be going from the "minimum sideways travel" to any other method.

but i never realised that you can basicly "alter" opening and closing ramps by adjusting geometry.

another bit learned.

[awesomebill]

There is a rumor that Circle Track magazine will very soon have an article or two on extensive coverage of proper valve, rocker and pushrod geometry. Based on some of the proper methodologies of Miller Engineering. You pro's likely have been doing it correctly for quite some time. We hobbyists and weekend warriors are about to learn what many of us have been doing wrong...... some of us for 35+ years.

Dave

I have been reading this post and can't help but to join in. We get a few engines in for rebuilds and freshen ups, more every year, and tend to see the 1 fits all style. Every engine we get in we dyno first with the customers tuning requirements to make sure we do not get stuck.

We then take the engine apart just as any one would a go back over everything just as if it were new. The shaft rockers from Jesel or T&D or even the roller rockers are at best close. When we check this area we use the black marker method along with the engine set up just as its run with valve adjustment included. We have numerous lenghts of p/rods to accomodate this procedure. .025" is as close as we get. When we introduce the real load situation, it is always a little wider than no load. When we do the no load condition and the mark is usually .020-.030 wide, I do not care what part of the valve it winds up, I look for the narrow line method. Because when I tell the customer he needs offset stands to move it as close to center as possible, they never go for it. When I show them this, they still don't go for it. So I have come to making the best out of a not so best situation.

Usually the push rod in the loaded position has to be at the very least .025 longer, and I always shoot for .050 longer when running conditions are in place. I find that with the little longer than shorter method or even right on buy the gauge, are not long enought when the engine is at 7000-9000 rpm because of deflection and just the way everything moves around.

With push rod flex, lofting, harmonics are all added in, the push rod never gets longer. It will straighen out and shoot the rocker out of its seat when it does go over max lift. Kinda like a pole vaulter. We always go with the best push rods like Trend, Manton, etc and this seems to help. My point is we always set up .050 longer and when we have had the heads come back, the guides, rockers, pushrods, lash caps, and most certain, the locks do not stick in the retainer. So, even if it does not set up correctly, you can make the best out of it. The customer seems to always wind up paying and most of the time, never sees the improvement for the $2000 bill for his new valve train set up if he goes with the perfect method as you have suggested. Those little increments, while we know they are corrrect, the customer does not see because of the ability of the car, driver, etc to maximize the perfect conditions we have now given them. I usually do not know what to say when I have suggested the repair and the repair does not yeild anything noticeable. I grew up poor so I know what it is to be sold something that I really did not need. I am the customer on the other side of the counter.

Even with the proper valve train set up, the engines do make little to none difference on the dyno. Unless ofcourse, it was so far out in left field. What I do find is everyone jumps on a band wagon about a certain area but really says nothing about the most important area with valve train. I can work around a rocker not perfect. What I can't do with out is good push rods. What we see on the dyno with the now standard taper wall very thick push rods is way more than any properely set up rocker system. I must add, it all belongs to be perfect, but the push rods by far are more important if a scale of whats worst is applied.

But when he runs 2 years and brings it back then we know. If I can get the engine to do at least what it does when it comes in, I know with a few runs, it will do better and better once everything gets honed and working again. This usually will make the customer happy. I find that with the more I read the less I know. This post will make some happy and some very sad.

As far as the customer doing it himself, we should have a uniform way to get them close. Most don't even care if its in the ball park to start with nor do they under stand why. So Ed's advise is pretty close. It is the one I use and tell my cutomers to use if I sell them the rockers or shaft systems. Its easy to use and will yeild the most consistant method along with some not so hard to read instructions. How could we mess up, " get the pattern as close to center and as narrow as possible with the engine together and turning the engine over 1 cycle of valve action, then take it off and look at the pattern", Then put the best pushrods in you can afford. Even with that being said, they will use the pushrods they were sold with the package. Someone will tell them and sell them what they think they will need. You guys know I'm right here. Customers call me all the time for pushrods needed for their combination. When I ask them what lenght they tell me their combination in detail and then ask me for my recomendation. Even if I did tell them what we did use for that particular combination, they usually by it somewher else.

So if we are trying to master this or just get a uniform method to keep the customer happy, let them continue what they are getting from the manufacture or give them the marker method. JMO

[Ed-vancedEngines]

Bill,
You are so right.

One thing I see far too frequently is people just blindly buying their pushrods before even assembling the long block. My next door neighbor is a wonderful mechanic and thinks he is a race engine builder too. He is putting together a new 555 cu in big block with a little of my help when he asks but if he doesn't ask does not want my opinions. I mean, after all, he has been building engines for over 20 years so he knows what he is doing and he does build a few street performance engines too.

What he calls and what most people call building an engine is what I call just doing an assembly and simply buying the parts and sticking them in. That is not engine building. That is waht most Crate Engine shops do too.

He mostly uses my help if he is doing a street performance engine for me to help him to check clearances and to maybe get opinions about camshafts.

Anyway for this 555 cu in Big Block he wants to build nothing is assembled yet but he does already have his new pushrods bought for it. I see this a lot. I even have a few customers try to get me to use pushrods they already have or have bought in thiers too. If I can I will but I do tell them that I will not know if they can be used until after I can can check the geometry after it is all built.

I find that a little off or out of whack geometry is doable in low performance and low lift applications.

In serious race applications I want to use the biggest baddest meanist pushrods I can get when I can. I will open up the head in the pushrod hole areas if possible just to get a bigger pushrod into the hole if I am using big springs and it is a serious output engine. I, too will opt for a pushrod that is appx .025-/050 longer than my measured ideal length. They sure don't get longer when running except for maybe a few thousandths of heat expansion but they do get shorter under compressive loads. With a rocker stand shaft system I don't think you can get very much longer without getting critical with the adjusters and the way oiling is done through them. Maybe some of you have gone longer but I haven't.

With a shaft style rocker system the stand ht and the distance from the fulcrum to the valve are the critical areas. With a stud rocker system the stud length and the pushrod lengths are critical.

If the valves are too long to work right we can chop the stem length some in some cases or buy an offset trunnion like Dave Crower is selling.

I try to use a good quality spring that will stack tight with some applications in small blocks instead of having to use a long valve. You can get some good springs with a 1.050 coil bind and use them instead of using a spring that coil binds at the 1.150 which will require a longer valve to get the right installed ht.

Speaking of valve train geometry fun;
I did a one off 6 cylinder Ford 300 a couple of years back with a highly modified Ford 240 cu in cylinder head, using ARP Studs for a Chevy and Roller rockers for a 460 Ford, and 2.05 and 1.625 valves for a SB Chevy that were .100 over length. Man those pushrods were long and I had to cut the crap out of the pushrod holes to even get a 3/8 .138 wall pushrod into it. Can you say 6,000 plus rpm inline 6 cylinder. lol. We also set it up with the late EFI manifold and a custom Throttle body with Haltech EFI brain, 44 lb injectors.
Ed

[Wolfplace]

When we do the no load condition and the mark is usually .020-.030 wide, I do not care what part of the valve it winds up, I look for the narrow line method.

Usually the push rod in the loaded position has to be at the very least .025 longer, and I always shoot for .050 longer when running conditions are in place.
I find that with the little longer than shorter method or even right on buy the gauge, are not long enought when the engine is at 7000-9000 rpm because of deflection and just the way everything moves around.

My point is we always set up .050 longer
What I can't do with out is good push rods. What we see on the dyno with the now standard taper wall very thick push rods is way more than any properely set up rocker system. I must add, it all belongs to be perfect, but the push rods by far are more important if a scale of whats worst is applied.

As far as the customer doing it himself, we should have a uniform way to get them close. Most don't even care if its in the ball park to start with nor do they under stand why. So Ed's advise is pretty close.
It is the one I use and tell my cutomers to use if I sell them the rockers or shaft systems. Its easy to use and will yeild the most consistant method along with some not so hard to read instructions.
" get the pattern as close to center and as narrow as possible with the engine together and turning the engine over 1 cycle of valve action, then take it off and look at the pattern",

Then put the best pushrods in you can afford. Even with that being said, they will use the pushrods they were sold with the package.

Hi Bill & Ed
Good info, been using the "pattern" for as long as I can remember & have posted about this before.
I like to start with checking pushrods & a light spring & the beginning pattern just to the intake side of center,
Then I get the pattern as narrow as I can while still on the valve tip. then recheck with full pressure
And I always err to the longer side on pushrods.

In short terms, my opinion is if the rocker is sweeping across the tip it is not pushing the valve open,, wasted motion not to mention guide wear.

I do not care where the pattern ends up with regards to center on the tip within reason as long as it is on the tip & I feel the valve does not really care either.
I want a narrow pattern as close to center as possible.
In the real world you have to make compromises here without spending a lot of money especially with stud mount stuff.

Here is an old post on another board I did in answer to a fellow that was advocating light weight pushrods

[QUOTE=GOSFAST]
P.S. One tip here, we are starting to deliver a few units using 5/16" pushrods, seems to be excellent advantage for weight-saver. The only issue as of now is there are no 5/16" guide plates available from the aftermarket. We order them direct from G.M. And seeing how we use to use them in SB's with roller-spring pressures, we figured to try it here. One unit's been out a few months running at 6500 and appears fine. Any reciprocating weight (in any area of the overall unit) you can keep down to a minimum always improves the RPM range.
[/QUOTE]

=
Gary,
On this I have to disagree also.

As Roger stated, could be that with lighter springs you are getting away with it but there is more to a pushrod than a stick of steel,,,

Pushrod & roller weight has been found to be a very small issue in regards to valvetrain control from any testing I have seen

The valve, retainer, spring, rocker mass on the valve side is a different story.

Give me a big fat treestump lookin pushrod over some spindly lightweight part anytime

Here is some info I have posted before,,

"Consider the pushrod as a piece of spring steel complete with harmonics & resonant frequencies for a moment.
Now consider what happens when your spring goes through it's harmonics & dances clear off the head & retainer.
And at some point in your rpm range I can just about guarantee it does.
An 800lb spring has about 1400 lbs on that little piece of spring steel.

All pushrods deflect & rats are one of the worst offenders with their goofy angles.

In my opinion, a pushrod cannot be to stiff to control valve movement.
If you consider it as a piece of spring steel you can see why the thick double taper ones are the pushrods of choice in most professional hi rpm applications."

[andyf]

#84 - yes I'm sure that allowing the rocker arm to travel sideways a bit more at the beginning of the cycle does slow things down a bit. But, with very high spring loads you are now reducing the sideways travel under load.

If you draw a couple of arcs on paper you'll see the difference. One arc is the classic over and back minimum travel setup. The other arc that I'm talking about just moves the arc so that at the end of the lift cycle the rocker arm tip is moving more vertical than sideways.

To do this trick you typically have to lower the rocker shafts.

The classic method shoots for the rocker arm to be perpendicular to the valve at 1/2 lift. The minimum travel under load method drops that down so that the rocker arm is perpendicular at about 66% of lift.

Here is a picture of the setup that we use to measure side travel. It is a little more accurate than the classic felt marker on the valve stem approach.

[Ed-vancedEngines]

Andy,
I like it.

Is that a fixture to hold an indicator welded to a retainer?

That is what it looks like to me and I can do that if it is.

I do like that idea.
Ed