This is pretty involved in my head, so bear with me:

On every pushrod OHV engine, the valves are held in place by the spring. Conventional wisdom says why change it. Welp, that's where the most reciprocating motion occurs in the valvetrain, meaning more spring stresses and fatigue. Would it be practical to design an engine with "valve" springs acting on the pushrod side of the rocker arm? There would be less motion due to the spring only compressing the equivalent of the cam lobe lift, not the valve lift(unless you're running an air-cooled VW motor), along with less overall weight in motion, which should translate into more stable valvetrain, better reliability, and higher operating potential.

Of course, you'd have to develop a way of fixing the valve to the rocker, and the rocker to the pushrod, etc. I'm just trying to think of why this wouldn't be an improvement. Y'all's thoughts???

 

Well there is something called a rev kit that uses an extra spring for the lifter so there's less stress on the valvespring. What you're talking about wouldn't be an improvement because of this

and because the spring would have to be stronger to still control the valve's inertia with the disadvantage of being on the short side of the lever (rocker arm).

 

Hey Rick, good idea, here are my two cents.

Lightening the valvetrain through head mods is probably simply too much labor (putting the springs around the rocker area or pushrod instead of valve stem). It's only going to matter if you rev it up high, and you'll run into driveability/idle issues way before then with a lumpy cam. If there's some type of variable valve timing, maybe, but until then...

 

I think the best way to use what we already have is to make it as light as possible. This is why the LS6 uses hollow stem valves and the LS7 uses titanium.
The other effect you can do is make the valvespring lighter, ie beehive springs!!!

The rev kit is really not good because every spring has a harmonic region and unfortunately, the small springs come to a harmonic region somewhere around 2750-3250RPM, which is terrible for our engines.

The best we can do today it having a behive spring, lightweight valve and ti retainer. The other effect besides the lighter weight of the spring itself, is the fact that because it has different diameters, the effective harmonics of the spring are way up in the rpm range.

 

"desmodromic" mechanism eliminates springs entirely. As complex as it is, it might give you ideas for attaching things to each other. Try Googling it.

With the spring at the lifter valvetrain stability would be worse, not better. What we are trying to do is control the valve motion so that it pretty much follows the cam lobe profile and doesn't bounce on the seat. With metal or pneumatic springs acting directly on the valve stem this is still a challenge. As we move the closing force farther from the valve things get sloppier and much more complicated.

Now remember that everything is a "spring", and the pushrod is the second weakest spring in the valvetrain. Next is the rocker then the rocker mounting. All of these springs would be working in series and need joints which worked in tension rather than in compression as we now do it. Every joint would have some minute clearance which would destabilize the valvetrain even more. Friction would be much higher so durability would suffer.

As was pointed out, the loads on the cam side are higher than on the valve side if rocker rato is above 1:1. Springs have to be much stiffer so there really isn't a free lunch with less travel but more load. The overriding problem is the multiple springs in series. That's a nightmare.

If you can find copies of the Brit magazine RACE ENGINE there has been a great continuing series on software called 4sthead (I think) by Blair and Associates. The last in the series was on pushrod actuated valves. Wow!

Sorry to throw cold water on your idea, but you asked. Keep thinking out of the box. Not every idea wins a prize.

Good luck!

 

Where the valve stem tip meets the rocker arm, the rocker to pushrod and the pushrod to lifter would all have to be some form of connection allowing the valve to be both pushed open and pulled back to close. This connection would introduce added friction to the mix, as well as some added play. And the biggest hurdle would be that since you no longer have pressure from the valve itself all the way back to the cam, you cant use a hydraulic lifter to take up the slack (lash).

 

Hey, don't worry bout it, SStroker, others. I have more ideas in my head, ones that would have to be engineered to the point that retrofitting would be impossible, I'm afraid. The desmodromic operation... I have an idea that runs those lines, but uses more like a push/pull rod mechanism that could be adjusted with preloading springs on both the push and pull sides of the rocker. The cam would basically be of a design that opens and closes the valves... kinda like a desmodrimic I guess, eh?

 

Hey GuitsBoy, what kind of axe is that in your avatar? I play an Ibanez JS.

 

What about a pneumatic valve spring? Something similar to the F1 style nitrogen compressed valves. You could have a 2 chamber "spring" assembly. Air forced into the top chamber would open the valve and air in the bottom chamber would close the valve. Each chamber holding small amounts of air to prevent the valve from opening too much or too little. All the solonoids wired through the ecu could potentially eliminate the need of a camshaft. Think of the tuning possibilities...lift, duration, and all cam specs could be controlled via the ecu. Almost unlimited power and driveablility options. I don't know if this is even possible...but it sounds good. It would be a nightmare to design on both mechanical and electrical standpoints, and probably way to expensive to be practical...but it sure does sound good.

 

<threadjack>
Its a guitar i built myself. Heres a few pics:
http://www.ihearyou.com/~tony/pictur...ar/page_01.htm
http://www.ihearyou.com/~tony/pictur...ar/page_02.htm
</threadjack>

 

IMHO.... The best thing to do is get rid of the cam, push rods, rocker arm, spring , and poppet valve. Reciprocating motion is inefficient. Rotary valves would be the ticket if people would do more development work on them. Smokey Yunick had an engine that used rotary valves. A few others have to.

http://www.dself.dsl.pipex.com/MUSEU...aryValveIC.htm

http://www.lortim.demon.co.uk/aspin/index.htm


EDIT: You might be able to get a set of heads with rotary valves http://www.coatesengine.com

 

Thoughts:

Reciprocating poppet valves actually do a good job mimicking what the engine demands for airflow due to it's reciprocating pistons. Flow thru a rotary valve into a cylinder (not a 2-stroke crankcase) is quite disruptive. Perhaps that's one of the reasons rotary valves have not been able to outperform poppet valves in terms of power. Of course sealing and having the rotary valve part of the combustion chamber also adds to the rotary's problems. It's not that there is not $ to develop rotary valves, it's that they are inherently not as good, all things being considered.

Pushrods aren't the greatest way to actuate valves even though they have been developed to do a fine job at least up to 10K rpm. OHC with finger followers and pneumatic springs are doing just fine at 20K+ rpm. Who'd a' thunk it when poppet valves appeared with the first engines over a century ago.

My $.02

 

Airflow through a rotary valve can't be all that disruptive....

"A stock 302ci 5.0L Ford small-block was chosen as a test engine to demonstrate the awesome characteristics of CSRV. The original, stock engine was dynoed at 260 hp and 249 lb-ft of torque. After the CSRV heads were installed, and with no other changes to the short block, the same engine churned out 475 hp and 454 lb-ft of torque! This is due to the reduction in frictional losses and improved airflow. Furthermore, the lack of valvetrain limitations allowed the engine to spin to 14,750 rpm. "

http://www.carcentral.net/content/ar...taryvalves.php


Of course the phrase "with no other changes to the short block" makes me suspicious. Alot of components don't count as being part of a short block.

 

14750 on a stock shortblock 5.0? 82.6% more power and 82.3% more torque and 225 psi BMEP with a "head and cams" change?.

YGBSM!

With those kind of claims, put me down as "slightly skeptical".

Barnum was right.

 

One other way of operating the valves is magnetic.

There are examples of motors running with valves operated by solenoids.

No valve train whatsoever.

Here is a link:

http://lees.mit.edu/public/In_the_News/LATimes_5-17-00.pdf#search='auto%20engine%20uses%20electrically %20operated%20valves'

 

Well if you notice they say no changes to the "short block" other than the heads. That means they can change intake and exhast manifolds as well as the carburation and add velocity stacks. I'm sure they didn't go with forced induction because with forced induction volumetric efficiency goes up quite a bit with ordinary valves and they want to stand out.

The 14750 rpm number is probably the point the engine fragged. An engine will with stand extreemly high rpms for a very short time(on the order of seconds). With a conventional valve train, engines couldn't hit these high rpms before valve float limited them. At their maximum attainable rpm they will usually last for more than 20 seconds before breaking.

 

14750 rpm on a 5.0 is about 11,800 piston gs, or about twice the gs a Cup engine sees @ 10,000 rpm, and about 18% more than an F1 engine sees. The Ford piston is about 130% of the wieght of a Cup piston and about 260% of the F1 piston. If F=ma, the forces on the rod were about 2-1/2 times what a Cup rod sees. IF it could get that high (over 350 friction hp @ 14750 to do that), 20 seconds might be about 100 times longer than it would actually last, IMO.

Sorry about the rant, but claims like those bother me because some mechanically-naive folks believe them. No offense intended.

 

The engine only has to last as long as it takes to rev up to 14750 rpm for them to make the claim. Mabey two seconds tops, and only about half that time would be at extreem rpm. On the coates website they are now claiming 14850 rpm. They must have had a second go at it.

As far as the cup and and F1 engines go, they are designed to produce work at these high g loads and last a couple of hours. I'm sure that when they were going for the high rpm test on this stock engine they had it running with no load, and didn't let it warm up to keep the strength in the metal.

 

Thoughts:

A high-end engine sim program said the friction hp that a 4.00 x 3.00 V8 with lower friction ring package needed to overcome just to rev to 14,750 was in excess of 350 hp. The problem is that that same engine didn't have enough breathing capacity to produce 350 indicated hp (hp produced from burning the mixture from which friction hp is subtracted to get "brake" of measurable hp at the flywheel). That means the engine won't free rev to anywhere near that rpm even under no load.

Of course if you put the block and valvetrain on a Spintron you could spin the rotary valves up to 7375 or 7425 rpm (1/2 engine speed for valves) and I'm pretty sure the rotary system would live for a bit. The seals might get a little warm, however. So I really don't think they ran a complete engine to 14K+ even though the text leads to that conclusion.

There's just not a whole lot of use for a 14K SBC or SBF. The rest of the systems don't like the 73-7400 ft/min mean piston speeds. There's not a lot of need for SBC, SBF or DC or Toyota pushrod Cup engines to go over 10K, but the valvetrains will go that high, or higher. The valve motion is a bit less aggressive at those revs, but they can live with current pushrod poppet valve technology.

FWIW, F1 engines need to run about 1400 km in anger for the two-race requirement that's been in place for a season or so. That approaches 4 million revs, or nearly 3 times as many revs as a Cup engine runs in anger. That's impressive. Of course, some of the new V8s aren't making that goal yet.

I've beat on you too much already Conn. Just don't believe everything you read...even from me. There are a few folks around here that that will agree with that.

Stay cool!

 

Does your sim program take into account the better breathing capacity of the rotary valves? The engine was dynoed at 475hp. On a flowbench the stock poppet vavle flowed 133cfm. The rotary valve flowed at 319cfm.

http://www.coatesengine.com/engine_of_the_future.html

Its not the high rpms that make the rotary valves so attractive. Its the fact that they breath better and have less parasitic drain on the engines output.