MEI/PVS Mid-Lift Rockers Systems

[Lynn Larsen]

All,

I spent a pretty good amount of time this afternoon reading about the theories behind the products of Miller Industries, Inc. The MEI and PVS (Precision ValvetrainS) lines offer products built in accordance with Mid-Lift® theories and design standards. There is a lot of technical information on this whole Mid-Lift thing and it makes a lot of sense. The prices for the shaft mount arms, the Precision Stand Series, is very much in line with most entry level shaft mount rocker systems.

Rather than trying to reinvent the wheel and explain Mid-Lift here, I'll give you the URL and let you read it for yourself. There is also a very good primer on cam terminology there: Mid-Lift.com

Lynn

 

[Tim Kay]

Man is that good reading /ubbthreads/images/graemlins/grin.gif Simple for the layman I am to understand.

As simple and logical as he makes it out to be it's a wonder it wasn't standard design from the beginning.

 

[Ron Earp]

Good find Lynn, and really not that expensive. Might have to retro fit those to my motor too since the cash outlay doesn't seem too bad. Nice description there for sure.

 

[Lynn Larsen]

Tim,

My thoughts exactly. There is no doubt about his assertions on how to set the rockers in relation to the valve. The tricky part, to me, is being able to do this and, at the same time, having the same relationship between the pushrod and the other side of the rocker. This would be highly dependant on block design, head design, how much surfacing has been done to the head and the deck, etc. etc. Then the pushrod length would have to be precise for each and every setup. But, again, there is no doubt about the advantages of having the rocker rotate from half way short of mid-lift through half way past mid-lift on both sides. The path of the rocker's roller across the valve tip is pretty easy to see. The efficiency of the translation of the linear motion through the rocker is also fairly easy to grasp. Less apparent is the issue of the valve's acceleration. You don't want it accelerating at the end of its travel (full open) because it will float more easily. You don't want it accelerating as it closes, for obvious reasons, either. So putting the greatest acceleration near mid travel is the only logical alternative.

The only question is the validity of Miller's contention that none of the other valve train vendors out there recognized this or designed their hardware with this in mind.

Regards,
Lynn

 

[Trevor Booth]

Quote "The only question is the validity of Miller's contention that none of the other valve train vendors out there recognized this or designed their hardware with this in mind"
His claims are a bit extravagant. It is a basic engineering principle when translating linear motion from rotary motion that you design for best geometry at the mid point of the linear motion. If you have a look at valve trains of the 20's & 30's, and in particular stationary engines, you will note that the geometry is correct at midpoint. The error occurred by not recognising that the line of action is through the axis of the roller that contacts the valves. This fact was recognised by the Brits and the Aussies a long time before Miller. I am amazed that Miller was able to patent a basic engineering principle. What Miller has not recognised, or in the least he makes no mention of, is the reduced side thrust on the valve guide. This is the biggest benefit of correct geometry. It is to be noted that if you alter your cam profile you need to also change your rocker setup to maintain correct geometry. The relationship of the pushrod to the rocker is not so critical but using an adjustable length pushrod for setting up saves a lot of time. His theory of 90 deg pushrod to rocker at mid point IMHO is not the best geometry. His method actually increases valve acceleration from mid point to full open. (Think about levers and conservation of energy and you will see what I mean). FWIW your valves are accelerating all the way to full open and close irrespective of rocker geometry. This is easily demonstrated by doing a graph of valve lift Vs engine rotation. Valve float is caused by the valve kinetics exciting the natural frequency of the spring. Incorrect spring choice rather than too much valve acceleration is the main cause.
Trevor

 

[Lynn Larsen]

Trevor,

I am not surprised at all that what Miller claims as his revelations were thought of long before he was even born. He does, however, describe how keeping the path of the roller on the end of the valve near the center throughout does lessen the side loading on the valve stem and the valve guides.

I have not studied the affect of harmonics on valve train operation, so I must accept your description of the major cause of valve float because I cannot refute it (assuming it was refutable.) You must admit, however, that if the valve is at its highest acceleration as it reaches the full open position, the valve's inertia will have its largest influence on causing the valve to continue its motion in the open direction and countering the springs ability to arrest the motion in that direction and start the acceleration towards the closed position. Will this inertia cause float? I can't make that blanket statement, but will it influence float? Yes, without a doubt.

Please, if you will, describe what you feel is the best geometry for the push rod to rocker arm? My guess is that you would want the push rod at 90º to the rocker arm as the lift began so that rate of change of acceleration (2nd derivative) would be minimized throughout the lift. But, wouldn't this also minimize the displacement of the (linear motion) of the rocker arm? I suppose this wouldn't be an issue if it were accounted for in the lobe lift, but I have no idea if the commercial cam makers take rocker geometry into account when they create a given cam grind or if engine builders assume one setup over another when specifying the cam to use. My guess is that they use the ratio of the lengths of either side of the rocker and do not consider the deviation caused by the change in the angles from closed to open and back again. My assumption was that Miller's reason for favoring 90º at mid-lift for the pushrod was for maximum fidelity in the information transfer from the cam profile to the valve motion as much as anything.

This is a great discussion for me because, for years, I have heard people talk about valve train geometry and could only agree that it was important in principal. All the time, I didn't really know exactly what they were talking about. At last, I feel I am starting to gain some understanding. So when I ask for your view on the best geometry, it is an earnest question, because I know that I still have a lot to learn.

Thanks,
Lynn

Edit: FYI to get the º symbol, while holding the alt key down, hit 167 on the numeric key pad. The symbols available can be found in an ASCII table available here.

 

[Trevor Booth]

Lynn,
My apology to Mr Miller, I did not see the reference to side thrust. My basic rule of thumb, open the valve as quick as possible, hold it open for as long as possible, close it as quick as possible all with the least amount of energy consumption. The acceleration of the valve may or may not be at its maximum at the full open posn and you are correct it is preferable that it is not. This depends on cam profile or more correctly lift Vs rotation. In terms of conservation of energy, max accel is best not to occur at start of lift nor at full open posn. ie get it moving, accelerate it quickly, slow it down. This is not always easy to achieve with flat tappet followers. It also exlains why two different cams, with the same timing and lift, can have vastly different results. My theory is that the rate of change of that acceleration is very important in conserving energy. Starting the valve to open slowly consumes less energy in overcoming the inertia of the valve and spring, and similarly with arresting it at full open (lighter spring) As you have noted, the position of the pushrod will affect the rate of change of acceleration. My theory is to have the pushrod at 90ºwhen fully open as it gives the longest lever arm at the highest load=less energy. Sure this increases the rate of change of acceleration but it is more than offset by the lesser load situation. Now all of this is not easily achieved, the cam lobes are too small to start with, and it all becomes a compromise situation. But if you start with correct theory, experiment and log the changes, some benefits can be obtained. These days computer simulation of valve trains gives you a head start and saves hours of experimentation. You may note that I have made regular mention of conservation of energy, to me this is free horsepower. Attention to internal losses is very critical, some people spend thousands of dollars for a 5% gain, when it could just as easily have been obtained for no dollars but time.
Cheers,
Trevor
PS thanks for the ASCII tip

 

[Lynn Larsen]

Today, I talked at length about this with the guy I am hoping will be my engine machinist. He explained Jesel's theory on this. (He builds a lot of NASCAR engines and has an engine shop to die for. He has another shop for his cars; poor guy Basically they want to minimize the roller movement across the valve tip when the greatest pressure is on them. Of course this is at or near full open. So they would want the valve stem and the rocker arm to reach 90º somewhere nearer to full lift. It would most likely be just before full lift, but in such a position that the lateral movement of the roller away from the valve tip center and back is an absolute minimum, maybe 3/4 lift? As you pointed out, computer simulation could give you this optimum location very quickly. There would seem to be some logic in this approach as well; especially, where longevity of the parts is concerned, but also with side loading: more load=more skipping or sliding of the roller. Miller did point out that, even with roller tips, the lifter does more skipping or skidding across the valve tip than anyone would think. A lot of motion at lower loads won't wear the parts nearly as much as it would at higher loads and would put a lot less lateral force into the valve as well. It would be interesting to compare the lateral distance traveled after mid-lift with the two setups since the "mid-lift at 90º" setup minimizes the total lateral distance traveled whereas reaching 90º near full lift minimizes the lateral component, albeit larger overall than the other approach.

Regards,
Lynn

 

[Trevor Booth]

Tell your prospective engine builder to hook up a strobe light to "freeze frame" the roller. Make the strobe pickup adjustable so he can freeze it at all positions of valve opening. It can be viewed and photographed, or video, via a good quality Borescope. There is little point in having roller rockers if they skip across the valve. It can be virtually eliminated and it is very simple to do. If your engine builder goes thru the viewing motion as above , he will discover what I did (by accident). When I did it Borescopes werent around so it was rocker off and oil everywhere.
I have not looked at the 90º posn at or near full lift but it may have some merit. Consider-- for a given valve lift and a given rocker arm length the arc so described by the axis of the roller is the same, therefore the lateral displacement is the same.(in the context of this thread) Your mans theory of having minimal lateral displacement at highest load is valid and good thinking, but , the roller should roll. Computer simulation of component parts is one of the reasons why we have longlife warranty and engines dont wear out like they used to.
Cheers
Trevor

 

When I went looking for the best available pushrods for my engine, I found that 3M makes Al-Al2o3 metal matrix composite pushrods. These are light weight and resistant to harmonics.

3M had a couple of PDF's about their pushrods. One was by David Vizzard. He discussed the potential advantages of these pushrods in a Winston Cup engine. On top of that he discussed a lot about current Winston Cup valvetrains.

These pushrods were so good that they could delay valve bounce by up to 400 rpm. Even with that huge advantage, Vizzard says they would cost power if you were to do a straight swap. Evidently, the Winston Cup motors "loft" their valves at high rpm. The roller lifter actually looses contact with the cam lobe and moves the valve further in to the chamber. With the 3M pushrods, this lofting was less, so there was less power.

I would not be too concerned with the acceleration near full lift, the worst that will happen is a bit of loft. On a street car the cam probably isn't big enough to get in trouble if you have appropriate valve springs.

The acceleration off and on the valve seat are the things to worry about.

Do a search for 3M AMC push rods to find that information. It has lots of spintron results and is very good reading.

Lastly, the price $80.00 per pushrod. That's $1280 for a v8.

I am currently looking for used 3M AMC pushrods.

 

[Paul Thompson 'Hooligan']

Great find Lynn - cheers! /ubbthreads/images/graemlins/grin.gif