Analyze this valve train Please

[Unknowho]

Has anyone else seen this valve train system? The claims I have read state it can rev higher than a OHC engine and out breath it as well due to the valve remaining at peak lift for up to 120 crankshaft degrees. Thanks

 

[Flyboy]

Hmmmm. I don’t understand what advantage there is to be gained from this valve actuation approach. Seems like a more complicated approach than your traditional pushrod setup, only with more inertia and lash.

 

[jrmichler]

Compared to a conventional pushrod valve train, this concept:
1) Has higher peak force on the cam. Cam lobes are already a weak point.
AND
2) Pushrod is loaded in bending, so both inertia and stress are higher.
AND
3) The pivot bearing is highly loaded.
AND
4) More parts and failure modes than a pushrod engine, so more expensive and less reliable.

Because of the inertia and bending, it is not possible for this design to rev higher than an OHC engine, or even higher than a pushrod engine.

 

[Unknowho]

Hmmmm. I don’t understand what advantage there is to be gained from this valve actuation approach. Seems like a more complicated approach than your traditional pushrod setup, only with more inertia and lash.

More inertia? The only time the rockers, pushrod, lifters and valves and springs move is on the opening and closing ramp of the camshaft. Notice the camshaft lobe is very, very small. So small in fact that it has a base circle. Due to the minimal movement and very small lift inertia is minimized.

Elaborate on lash please.

 

[Unknowho]

Compared to a conventional pushrod valve train, this concept:
1) Has higher peak force on the cam. Cam lobes are already a weak point.
AND
2) Pushrod is loaded in bending, so both inertia and stress are higher.
AND
3) The pivot bearing is highly loaded.
AND
4) More parts and failure modes than a pushrod engine, so more expensive and less reliable.

Because of the inertia and bending, it is not possible for this design to rev higher than an OHC engine, or even higher than a pushrod engine.

1) Has higher peak force on the cam. Cam lobes are already a weak point.
How so, the rocker that is actuated by the pushrod pivots away from the fulcrum of the rocker that interfaces with the valve. This would reduce the rocker arm ratio thus reducing the torque needed to open the valve.

2) Pushrod is loaded in bending, so both inertia and stress are higher.
The pushrod is only around 2.5" long and only move the same distance as the cam lobe which is .08" The response to answer one addresses your stress statement.

3) The pivot bearing is highly loaded.

The pivot bearing load does not exceed the spring pressure at peak lift due to the pushrod rocker bringing the valve rocker to a 1:1 ratio.

4) More parts and failure modes than a pushrod engine, so more expensive and less reliable.

A DOHC coyote has 4 more cams and double the valves, springs, lash adjusters, locks, retainers and it does NOT face any reliability issues.

You might want to go and study COMPOUND LEVERS for a refresher to arrive at a logical conclusion.

 

[Flyboy]

More inertia? The only time the rockers, pushrod, lifters and valves and springs move is on the opening and closing ramp of the camshaft. Notice the camshaft lobe is very, very small. So small in fact that it has a base circle. Due to the minimal movement and very small lift inertia is minimized.

Elaborate on lash please.

Linear inertia is being traded for rotational inertia, and a lot of it. Think how much effort it takes to slide a pencil along its length, then how much is required to wiggle it from one end. It’s significant higher.

Lash in this case refers to the mechanical “slop” in the system. Most of the engines I an familiar with use some sort of hydraulic lifter system in the tappet/cam follower to maintain zero lash across the operating envelope, but I don’t know how well they work with smaller movements like what you’re proposing for your cam follower.

Finally, @jrmichler pointed out something that I missed: the forces at play. You’re sacrificing a LOT of force in this system, and valves need a lot of force, especially at higher speeds. Are you sure that you fully understand the loads on that valve when running the engine at full power? I have severe concerns that the swinging arm system will not be strong enough, even with the ramp mechanism, and will get bent. Hell, I have seen straight pushrod tubes get bent under normal wear and tear at under 3000 rpm.

In my opinion, informed by my history with aviation piston engines, I don’t think this is a viable design. If it was, I suspect we would have already seen one hit the market.

 

[jack action]

Notice the camshaft lobe is very, very small.

Yes, but notice also that the "primary" rocker has an insanely huge ratio. The end of that rocker also applies force laterally and longitudinally. I suspect a force analysis on this part might show that it requires stronger and/or bigger parts than expected. So the "potential for lighter components" claim seems to be an illusion.

I'm also doubting the "increased valve lift" claim. I'm pretty sure the valve motion itself - together with the spring force maybe - is the limiting factor for the maximum valve lift in the current valvetrain designs.

The "reduced friction" claim is also hard to believe with adding more parts interacting together.

Reference for CLVTS claims: www.youtube.com/watch?v=uHm5YRQZPig

 

[Ivan Nikiforov]

As I understand it, the main point of this idea is to reduce the height of the cams. This is important when operating at high rotational speeds. If the cam is too high, the pusher will not be able to return to the lower position at high rotation speed. The use of a low-altitude cam naturally necessitates the use of a lever mechanism to convert force into distance. It seems to me that in general this is a good idea. At the same time, of course, as the participants wrote earlier, there are problems with backlash, loads and friction. However, these problems can probably be solved by using suitable materials and the design of the mechanism.

 

[Unknowho]

Linear inertia is being traded for rotational inertia, and a lot of it. Think how much effort it takes to slide a pencil along its length, then how much is required to wiggle it from one end. It’s significant higher.

You are overlooking the fact that the pushrod rocker changes the valve rocker ratio to a 1:1 ratio. Therefore the spring pressure applied to the components on the intake side never exceeds the spring pressure. A fixed ratio rocker arm on the other hand multiples the pounds of pressure needed to open it. For example. A open spring pressure of 300 lbs with a 1.7:1 ratio would require the cam, lifter and rocker to provide 510 lbs of pressure to open that 300 lbs spring and valve.

Lash in this case refers to the mechanical “slop” in the system. Most of the engines I an familiar with use some sort of hydraulic lifter system in the tappet/cam follower to maintain zero lash across the operating envelope, but I don’t know how well they work with smaller movements like what you’re proposing for your cam follower.

It would work just as the lash adjustor on OHC engines do. Look at the lash adjustor on a DOHC Coyote for example.

Finally, @jrmichler pointed out something that I missed: the forces at play. You’re sacrificing a LOT of force in this system, and valves need a lot of force, especially at higher speeds. Are you sure that you fully understand the loads on that valve when running the engine at full power? I have severe concerns that the swinging arm system will not be strong enough, even with the ramp mechanism, and will get bent. Hell, I have seen straight pushrod tubes get bent under normal wear and tear at under 3000 rpm.

Again, you are overlooking the fact that the pushrod rocker pivoting away from the fulcrum of the valve rocker brings it to a 1:1 ratio. That means less "FORCE and Torque" are applied to the components. That means lighter valve springs and pressures are needed on this system.

 

[Unknowho]

Yes, but notice also that the "primary" rocker has an insanely huge ratio. The end of that rocker also applies force laterally and longitudinally. I suspect a force analysis on this part might show that it requires stronger and/or bigger parts than expected. So the "potential for lighter components" claim seems to be an illusion.

I'm also doubting the "increased valve lift" claim. I'm pretty sure the valve motion itself - together with the spring force maybe - is the limiting factor for the maximum valve lift in the current valvetrain designs.

The "reduced friction" claim is also hard to believe with adding more parts interacting together.

Reference for CLVTS claims: www.youtube.com/watch?v=uHm5YRQZPig

Yes, but notice also that the "primary" rocker has an insanely huge ratio. The end of that rocker also applies force laterally and longitudinally. I suspect a force analysis on this part might show that it requires stronger and/or bigger parts than expected. So the "potential for lighter components" claim seems to be an illusion.

Are you taking into account the fact that the pushrod rocker arm pushes away from the fulcrum of the valve rocker. This brings the ratio to a 1:1 ratio. Thus, the torque and force applied to the camside components never exceeds the open spring pressure at full lift. A fixed ratio rocker arm on the other hand multiples the pressure on those parts by the given ratio.

I'm also doubting the "increased valve lift" claim. I'm pretty sure the valve motion itself - together with the spring force maybe - is the limiting factor for the maximum valve lift in the current valvetrain designs.

A simple understanding of compound levers and the strategic placement of the inclined plane where the 2 rockers(levers) interface proves the lift is increased by the high ratio you sited.

The "reduced friction" claim is also hard to believe with adding more parts interacting together.

The reduction in friction is due to the fact that the only time all the valve train components are moving is when the lifter roller is on the opening and closing ramp. The lobe has a base circle that remains at the same lift for 120 crankshaft degrees. That means less inertia, less movement, less friction.

 

[Unknowho]

As I understand it, the main point of this idea is to reduce the height of the cams. This is important when operating at high rotational speeds. If the cam is too high, the pusher will not be able to return to the lower position at high rotation speed. The use of a low-altitude cam naturally necessitates the use of a lever mechanism to convert force into distance. It seems to me that in general this is a good idea. At the same time, of course, as the participants wrote earlier, there are problems with backlash, loads and friction. However, these problems can probably be solved by using suitable materials and the design of the mechanism.

This system allows the camshaft lobe to be reduced drastically. This enables the lobe to be ground with a base circle to keep the valve at peak lift for up to 120 crankshaft degrees. This means the only time all of the valve train components are moving is when the lobes ramp engages with the lifter roller and when it engages the closing ramp. This means the non moving components are NOT creating inertia, nor friction.

 

[jack action]

This enables the lobe to be ground with a base circle to keep the valve at peak lift for up to 120 crankshaft degrees.

This is impossible with a spring/poppet valve. I don't even think you can do that with desmodromic valves.

It doesn't depend on the camshaft/rocker/pushrod/whatever actuating the valve. Just the valve itself. There are maximum accelerations - and especially a continuity constraint on, not only, those accelerations but the jerks and further time derivatives - to respect on the ramps going up and down to/from maximum lift. To keep these accelerations low, you either have to reduce the maximum lift or reduce the time spent on maximum lift. I never heard of valve timing on an engine with "idling" at peak lift, no matter how small the duration of that idling period.

I want to see the position/velocity/acceleration/jerk curves of that valve before discussing this further. Then we can look at force/inertia analysis on those components to produce those curves.

 

[Unknowho]

This is impossible with a spring/poppet valve. I don't even think you can do that with desmodromic valves.

It doesn't depend on the camshaft/rocker/pushrod/whatever actuating the valve. Just the valve itself. There are maximum accelerations - and especially a continuity constraint on, not only, those accelerations but the jerks and further time derivatives - to respect on the ramps going up and down to/from maximum lift. To keep these accelerations low, you either have to reduce the maximum lift or reduce the time spent on maximum lift. I never heard of valve timing on an engine with "idling" at peak lift, no matter how small the duration of that idling period.

I want to see the position/velocity/acceleration/jerk curves of that valve before discussing this further. Then we can look at force/inertia analysis on those components to produce those curves.

The fact that the cam lobe is only .08" of lift enables the lobe to be ground with a base circle on it. This means the lobe can be ground to stay at the peak lift for up to 120 crankshaft degrees. It's not difficult to understand. Imagine you take a typical camshaft lobe which has a LCA (Lobe Centerline Angle). Then you mill it down from the opening ramp to the closing ramp at .08" of lobe lift.

 

[jack action]

Then you mill it down from the opening ramp to the closing ramp at .08" of lobe lift.

Not that simple.

Let's do some back-of-the-envelope calculations.

Say your peak valve lift is 0.600", and with your typical cam, you reach it in 120 crankshaft degrees, and then the valve goes back to its seat in another 120 crankshaft degrees, for a total of 240 crankshaft degrees from seat-to-seat.

The average valve speed is $\frac{0.600\text{ in}}{120\text{ deg}} = 0.005\text{ in/deg}$.

Since we know the speed when the valve is seated and when it is at maximum lift is zero, there is some peak velocity somewhere between those points. This will depend on your cam design, but it should be somewhere in the middle at around twice the average velocity, so $0.010\text{ in/deg}$.

Therefore the average acceleration - not peak - from the seated position to half the peak lift is $\frac{0.010\text{ in/deg}}{60\text{ deg}} = 0.00017\text{ in/deg/deg}$.

For your cam which keeps the valve at peak lift for 120 crankshaft degrees, you take 60° to reach peak valve lift of 0.600", then 120° idling at that position, and then another 60° to go back to its seat, still for a total of 240° from seat-to-seat.

Now, your average valve speed is doubled, thus also your peak velocity, now at $0.020\text{ in/deg}$. So your average acceleration of the valve is now $\frac{0.020\text{ in/deg}}{30\text{ deg}} = 0.00067\text{ in/deg/deg}$, i.e. 4 times larger than with a conventional valvetrain.

Assuming you do this, you will require a much, much stronger spring to fight back against this now 4 times greater inertia, which will throw away all the gains you think you've made elsewhere.

With a lever, whatever you gain on one side, you lose on the other.

Compare the position/velocity/acceleration/jerk curves of your design to the ones from a classical valvetrain design and you will be most likely deceived.

 

[Unknowho]

How did you draw the conclusion that the ramps require 60 crankshaft degrees each to open and close the valve? Due to the fact the valve remains at peak lift on the "LOBE base circle" for up to 120 crankshaft degrees and is only .08 inches means the ramps will can ground much smaller than in a larger lift cam. With a 1 inch base circle cam at .08 inches of lift for 120 Crank degrees. The ramps can smoothly transition from the base circle to the lobe base circle in as little as 30 crankshaft degrees. The larger the cam base circle the ramps get even smaller because the transition is smoother.

And, how can the valve while seated at peak lift for 120 degrees create inertia? It creates no inertia there anymore than it would on the regular cam base circle.

You forget that when an equal force or torque is applied to a lever at equal distance you reach equilibrium. So, there is no give and take. That is what the pushrod rocker does as it interfaces with the valve rocker. It brings them into equalibrium.

 

[Ivan Nikiforov]

In fact, this is an ordinary lever mechanism. If you reduce the cam height, then you reduce the stroke length of the rocker arm. In order to maintain the stroke length of the rocker arm, you need to change the ratio of the lever lengths, that is, place the pusher closer to the axis of the rocker arm, and make the second end of the rocker arm longer. Why don't you make a system with a single pusher? The use of an intermediate rocker arm reduces the reliability of the system, as additional backlash, inertia of the intermediate rocker arm, and increased friction appear at high load points. Intuitively, it seems that due to backlash and inertia, such a system will not allow to increase the speed of rotation of the shaft. By the time all the backlashes are triggered and all the levers are accelerated, the shaft will have already turned significantly and the engine will stall. Of course, to get an accurate representation, you need to do a kinematic calculation.
Alternatively, perhaps it makes sense to simply increase the valve diameter with a reduced rocker stroke?

 

[jrmichler]

Too much hand waving, and not enough calculations (except for @jack action in Post #14) in this thread. To @Unknowho, it's time for you to do some calculations to support your ideas. Start at the valve, and define the position, velocity, and acceleration curves for the valve at the maximum engine RPM. Simplify the problem for now by looking only at the peak acceleration opening the valve. Then proceed as follows:

1) Including the valve spring force, calculate the force between the valve and rocker arm.
2) Calculate the inertia and angular acceleration of the rocker arm.
3) Calculate the force to accelerate the rocker arm and valve and compress the valve spring.
4) Calculate the force between the rocking pushrod and rocker arm. It will be larger than the force in Step 2 because of the ramp angle.
5) Calculate the bending moment on the rocking pushrod.
6) Calculate a cross section for the rocking pushrod using realistic materials and working stress.
7) Calculate the torque needed to accelerate the rocking pushrod plus deliver the force to the ramp on the rocker arm.
8) Go back to Step 5 and iterate including the angular inertia of the rocking pushrod. Repeat until the rocking pushrod is strong enough to deliver the necessary force to the rocker arm.
9) Now that you have the torque at the pivot of the rocking pushrod, you can calculate the force from the linear pushrod and the size of the pivot bearing. This might require going back to Step 5 and iterating again.
10) The force on the cam follower is the force from Step 9 plus the inertia of the linear pushrod.
11) Compare the force on the cam follower of your design to that of a standard pushrod engine with the same valves and valve motion profile.

At this point, you have completed the concept design. If it looks good, you can proceed with the actual design and refine the calculations starting from Step #1. This is the Mechanical Engineering forum, and this is how mechanical engineers look at ideas. No engineer would consider showing a concept design idea to management until the first pass of the above procedure through Step 11.

If you are truly interested in evaluating this idea, start by defining the valve motion profile as indicated above. I suggest that you use the motion profile from an existing engine. We can help you with this.

 

[Unknowho]

"Why don't you make a system with a single pusher?"

You mean like this?

And, some racing rules require a "pushrod" be used.

 

[Unknowho]

Too much hand waving, and not enough calculations (except for @jack action in Post #14) in this thread. To @Unknowho, it's time for you to do some calculations to support your ideas. Start at the valve, and define the position, velocity, and acceleration curves for the valve at the maximum engine RPM. Simplify the problem for now by looking only at the peak acceleration opening the valve. Then proceed as follows:

1) Including the valve spring force, calculate the force between the valve and rocker arm.
2) Calculate the inertia and angular acceleration of the rocker arm.
3) Calculate the force to accelerate the rocker arm and valve and compress the valve spring.
4) Calculate the force between the rocking pushrod and rocker arm. It will be larger than the force in Step 2 because of the ramp angle.
5) Calculate the bending moment on the rocking pushrod.
6) Calculate a cross section for the rocking pushrod using realistic materials and working stress.
7) Calculate the torque needed to accelerate the rocking pushrod plus deliver the force to the ramp on the rocker arm.
8) Go back to Step 5 and iterate including the angular inertia of the rocking pushrod. Repeat until the rocking pushrod is strong enough to deliver the necessary force to the rocker arm.
9) Now that you have the torque at the pivot of the rocking pushrod, you can calculate the force from the linear pushrod and the size of the pivot bearing. This might require going back to Step 5 and iterating again.
10) The force on the cam follower is the force from Step 9 plus the inertia of the linear pushrod.
11) Compare the force on the cam follower of your design to that of a standard pushrod engine with the same valves and valve motion profile.

At this point, you have completed the concept design. If it looks good, you can proceed with the actual design and refine the calculations starting from Step #1. This is the Mechanical Engineering forum, and this is how mechanical engineers look at ideas. No engineer would consider showing a concept design idea to management until the first pass of the above procedure through Step 11.

If you are truly interested in evaluating this idea, start by defining the valve motion profile as indicated above. I suggest that you use the motion profile from an existing engine. We can help you with this.

The title of this post is :

Analyze this valve train Please​

So, practice what you preach.

 

[renormalize]

The title of this post is :

Analyze this valve train Please​

So, practice what you preach.

It's not our job to help with your analysis until you show a good-faith effort to at least set-up and try to solve the problem mathematically yourself.

 

[Unknowho]

It's not our job to help with your analysis until you show a good-faith effort to at least set-up and try to solve the problem mathematically yourself.

Speak for yourself. This post is for those who would like to "Analyze this valve train." I've done the analysis myself and am seeking others to do the same. If you are not interested then change the channel.

 

[jrmichler]

And on that note, this thread is closed.