Valve spring pressure and boost 1

[yoshimitsuspeed]

Many highly respected engine builder/tuners will tell you that you need to increase spring pressure for high boost. There are even reports of fixing floating valves related to boost (However they determined that IDK) or performance issues by going to a stronger valve spring.

Now I don't believe this theory for one second. If the pressure differential is that great between the cylinder and the plenum then you are seriously liimiting the amount of air going into your engine to a point that I would find hard to believe. I mean if you had a 20 PSI pressure differential between the runner and the cylinder I would still expect a valve to close just fine unless the spring was already so precariously sized for the cam. On the other hand if you had a 20 PSI pressure differential between the two the amount of power that you are leaving on the table would be insane.

I'm confident enough in this that I'm not really even looking for confirmation although if I am missing something huge please tell me.
I am more here to talk about the stories of stiffer valve springs actually curing these issues allegedly related to the boost pushing the valve open.
What else could change with a stiffer spring? What would a logical explanation be that valve springs would actually fix the problem?
Or is it possible that they are running such an improper intake cam that they are actually preventing a massive amount of air to fill the cylinder?
In which case a properly sized cam would be the solution whereas stiffer springs are just going to help hide the problem.

Thoughts?

 

[Lou Scannon]

First of all, on a turbocharged engine with conventional valve timing, the pressure in the cylinder will never be appreciably less than the boost pressure, because the lowest it can get during the exhaust event is the pre-turbine pressure, which is typically similar to or higher than the boost pressure.

With early Miller intake valve timing, the cylinder pressure will indeed drop below the intake manifold pressure by several psi, depending on the severity of the Miller timing. Of course, with OEM engines engineered for early Miller timing, it can be taken for granted that the valvetrain has been validated to behave itself under all foreseen conditions. Also, obviously, this delta P across the intake valve will occur regardless of the intake manifold pressure, so in this case it doesn't matter if engine is boosted or not.

On the other hand, a mechanically supercharged engine, assuming its exhaust system is worth a damn, will see significant delta P across the intake valve toward the end of the exhaust event. However, this is about the time that the intake valve should be opening anyway, so if it should happen to "blow" open, who would notice? At any rate, I have difficulty envisaging a delta P large enough to overcome the spring force.


"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[yoshimitsuspeed]

Thanks for the input hemi that's exactly how I have been looking at it. Even with a supercharged motor like you say just before valve opening is the only time that that delta P should have any chance of doing anything and like you say at best maybe it would just crack the valve open just before the cam did. Even this is a long shot on all but the most heavily built motors. Fixed displacement superchargers don't like high pressure ratios and very few street driven supercharged cars would come close to the kind of pressures required to overcome valve spring force.
On intake valve closing one would hope that the pressure in the cylinder was very close to the pressure in the runner.

So the question remains though is there any possible truth to this myth and any way that stiffer valve springs could have any way of improving performance related to how much boost you are running?
People have reported the stiffer springs fixing whatever problem they were having.
Some of my theories.
Placebo effect. It has been pretty much proven that if you throw time and money at a car you will see, feel or find an improvement.
Something blamed on boost but maybe actually another issue. Maybe they raised their redline at the same time as boost and started floating valves which stiffer springs then fixed?

Any other possibilities?

 

[SomptingGuy]

Other possiblities?

Dynamic lift closer to kinematic lift? Not sure why that's more important for a boosted engine though. Just one more thing, possibly, if you're already in the eeking out more power mode.

- Steve

 

[yoshimitsuspeed]

Here is another question.
I have briefly thought about this before but never gave it a ton of my time but one of my customers goes to a technical institute and brought this up to me. His teachers are teaching that boost will cause valve float and that heavier springs are needed for more boost.
This bothered me because it means that either my theory is completely wrong or they are teaching bad theory.

So what are peoples thoughts on this being taught in a technical school?
Is this an old wives tale that rather educated engine builders can believe even though it can't be true?
Or am I still missing a very big piece that could make it bear truth?

My thought is that if it's possible for intake pressure to hold the valve open then the problem that really needs to be addressed is cams or cam timing to reduce or eliminate the delta P. Stiffer valve springs is not addressing the problem but just covering it up.

 

[yoshimitsuspeed]

Does anyone have any good combustion chamber pressure graphs?
I just spent a while searching and most only include the combustion cycle. Others that include 720 deg of rotation don't have enough resolution to really see what kind of pressures you would expect during the exhaust stroke.

 

[BrianPetersen]

Measure the seat pressure of your valves.

Measure the diameter where the valve is seated, where it actually contacts the valve seat.

Calculate the area of that circle.

Calculate the pressure difference that would be required to open that valve.

Ask yourself if that is a plausible situation and whether there is any foreseeable condition inside the engine which could cause that.

Supercharging (not turbocharging) trying to open the intake valve late in the exhaust stroke, just before the valve starts opening anyhow, is one foreseeable situation.

On a turbo engine, excessive exhaust manifold pressure during the intake stroke is another one.

 

[yoshimitsuspeed]

Still interested to see anything that anyone else has especially if there are graphs of combustion pressure on turbocharged engines.
I did just remember that I have a copy of Scientific Design of Exhaust and Intake systems and remembered that a good chunk of the book does focus on this very subject.
I snapped a few pictures of some of the graphs and thought I'd post them up here.

If we were to imagine a worst case scenario which would be a supercharged motor with an optimized exhaust we could guess that the intake and exhaust strokes would be similar to what they are here.
On a turbo the low pressure would be much closer to the pre turbine backpressure which is almost always higher than the boost pressure.


With a supercharger running extremely high boost on a motor with big valves and low seat pressure it does seem plausible that the valve could be pushed open by the boost pressure.

The 4AGE that I am used to dealing with has a 30.5mm intake valve or 1.2". Stock valve springs are about 35lb seat pressure so for boost to unseat the static valve it would require a delta of 30.9 PSI.
Granted these graphs are different motors and at much lower RPM than the 4AGE operates but all of those graphs show positive gauge pressure at the point the valve would be opening meaning your boost gauge pressure would need to be well over that 30.9 PSI to overcome the spring.

On the other hand if you had a big two valve V8 with a 2" valve and 35 lb seat pressure it would only take about 11 PSI delta to open the valve so in a supercharged motor with a high enough surface area to spring pressure ratio I can see this being plausible.

The same goes for the valve closing event. The stock 4AGE IVC is 44 deg ABDC.
The last graph shows the expected behavior of the pressure taking longer (IN reference to crank rotation) to hit atmospheric pressure. If the cam is sized anywhere close to proper the valve should be closing as the pressure in the plenum and pressure in the cylinder are the same. Unfortunately I believe these graphs are in reference to atmospheric pressure instead of port pressure but it starts to give us an idea.

Since these graphs are relative to gauge pressure I think it wold be pretty safe to say that cylinder pressure would never get lower than about 6 PSI below pre turbine backpressure.
If you have the perfect turbo with a 1:1 boost to PTBP ratio and are running 40 PSIG boost this means that cylinder pressure shouldn't go much below 34 PSIG leaving a 6 PSI delta between the port and the combustion chamber. This is at the point where the valve should be wide open. The delta would be much less at any other point.

 

[gruntguru]

Valve spring force is needed to control accelerations and keep the lifter in contact with the cam. Any pressure differential across the valve (times effective valve area) is subtracted from the spring force and reduce the force available to maintain that contact. For most engines there will be moments during the intake cycle (and possibly the exhaust) where this effect will reduce the rpm required for valve float to occur.

je suis charlie

 

[yoshimitsuspeed]

Thanks BrianPetersen
I was in the middle of typing the above doing just that when you replied.

On a turbo motor I supposed I could see the same issue on a motor with large valves and very light seat pressure.
I am less familiar with two valve motors but in a quick search I see seat pressure numbers over 100 lbs.
If you had a 35 lb spring on a 2" valve it would be quite possible for a small turbo to push an exhaust valve open.

Going back to the 4AGE that I know with a 25.5mm std valve and 35 lbs seat pressure you would need a 45 PSI delta to crack open the static valve or even stand a chance of stopping a closing valve.
The thing is that with a turbo motor the pressure in the chamber is going to be somewhere between intake pressure and pre turbine back pressure. It seems unlikely to me that you could ever achieve anything close to that 45 PSI delta even in the highest of boost applications.

 

[yoshimitsuspeed]

gruntguru
While this is true it is going to depend greatly on the pressure curve in relation to the valve travel.

It seems like we have narrowed down the likely points of float to be on the intake valve with supercharged motors and exhaust valve with turbo motors.


I am a little weak on fluid dynamics so my first question is whether the force being applied to the valve is directly proportional to the delta P when the valve is open and air is flowing past it.
As the air accelerates past the valve would it not create a low pressure zone around it?

On a supercharged motor if there is a 30 PSI delta between the port and the cyl as the valve is coming to a stop and then returning back downward while the valve is open and air is flowing past it is there 30 lbs of additional force being applied to the valve spring?



On a turbocharged motor on the exhaust valve there is the pressure wave drawing the exhaust pulse out and the intake pulse in. If your cam is timed ideally the exhaust valve will close as that pulse is moving toward the turbo or at least before it has started to move backward.
If that pressure wave has enough time to come to a stop and revert and start applying any significant pressure on the exhaust valve filling the cylinder with exhaust wouldn't it be safe to say that the problem that needs to be addressed is cam timing and not valve spring pressure? This should be long after the valve has hit max velocity meaning that even when the RPM is slightly off and a small degree of reversion hits the valve at the end of it's closing cycle it would not only be facing the spring pressure but the valve inertia.

This would take us back to my original idea that on a turbocharged engine valve float would be a symptom of improper valve timing and not insufficient spring pressure right?

 

[gruntguru]

Valve timing is optimised for somewhere in the region between peak torque and peak power - not redline and beyond which is where valve float begins.

je suis charlie

 

[yoshimitsuspeed]

Just sticking to the topic of the turbo and exhaust valve.
As the motor spins faster it will have less time for a reversion pulse.
At low RPM there will be more time and it will be more likely for the pressure wave to leave the exhaust and for a returning pressure wave to have time to bounce back in through the open valve. As revs get higher it will be more likely that the gasses will still be flowing out the exhaust valve as it closes. Therefore the higher the RPM the less force holding the valve open. Or if the gasses are still flowing out then it would help force the valve closed.


With supercharging and on the intake valve it would take longer in terms of crank rotation to fill the cylinders so it would make sense that the intake valve would start closing with lower pressure in the cylinder creating a larger delta P. The question is how much? It seems like if you are maintaining a reasonable VE that there should still be a pretty decent amount of pressure in the cylinder.
I wish I could find some better cyl pressure graphs for a motor and RPM more similar to the motors I am used to dealing with.

Does anyone have any further thought on how the pressure would behave with air flowing past an open valve. Like I said I'm not great with fluid dynamics but it seems to me that when the valve is open and when the air is accelerating past the valve that the pressure being applied to the valve would be less than the pressure in the port. Is there any truth to this?

I think it's pretty safe to say that specifically in regards to a turbo motor it is extremely unlikely that pressures will have any significant effect on forces applied to the valvetrain. Does anyone disagree with this?

 

[BrianPetersen]

Valve held open will significantly reduce any pressure difference across it, because the flow would simply try to equalize. Obviously there would still be some pressure across it because of the flow losses and the pressure wave effects - but the only valve in which the normal flow direction is in the direction of opening the valve is the intake valve, and with anything like normal event timing, there is no plausible way for there ever to be any meaningful pressure forcing the valve open at the end of the intake stroke. If there is ... you need a better camshaft; it would be better to simply leave the valve open longer and use that pressure to continue filling the cylinder.

 

[gruntguru]

A couple of points.
1. Valve float does not occur near the end of the valve-closing event. During this phase, the valve is approaching the camshaft and decelerating i.e. the valve is accelerating away from the camshaft - no spring force is required at all. Valve float occurs when the valve is accelerating towards the camshaft which occurs during most of the "middle" phase of the valve cycle. The cam profile is designed to take advantage of the available spring force as the spring length changes thus the peak acceleration towards the cam occurs at peak valve lift. (One advantage of pneumatic valve springs is the average acceleration can be much closer to the peak) So valve float is equally likely at any point where the acceleration (blue) curve is negative. 20 lbs of extra (gas pressure) force acting on the valve means 20 lbs extra spring force required to avoid valve float.

http://i.imgur.com/rVXDU0W.jpg[/img]]

2. Valve float occurs near the rpm limit (usually beyond) not the point of maximum VE where the camshaft is optimised. At the rpm limit it is very unlikely the valve has remained open long enough for pressures to equalise.

je suis charlie

 

[yoshimitsuspeed]

It will not have completely equalized but it should be close. That is my point. There will still be a differential but it should not be big. I feel like you would need to run a massive amount of boost on a supercharged motor to get to a 20 PSI delta as the valve is trying to close. If we are talking Top Fuel levels of boost and performance then yeah at 60 PSI there may be a notable force from pressure acting on the valve.
Then as BrianPetersen said and as I believe, the localized differential between either side of the valve face would be much lower and the overall forces acting on the valve faces should be lower than the differential as measured just a cm away.

 

[gruntguru]

What if i said "5 lbs of extra (gas pressure) force acting on the valve means 5 lbs extra spring force required to avoid valve float"?

As an example a small block Chev intake valve is approximately 3 sq in area, so 1.7 psi differential gives 5 lb force. Relevant differential would be the delta of total (stagnation) pressures measured either side of the valve.

je suis charlie

 

[Tmoose]

I think Hugh Macinnes in his book "Turbochargers" pretty much dismissed the theory of boost requiring increased valve spring pressure, and offered at least one alternative explanation for perceived valve float.

 

[Lou Scannon]

I have no firsthand experience, but I've read that the symptom of valve float on an engine with hydraulic lash adjusters, i.e. adjuster "pump-up", can just as easily occur with valve "bounce", i.e. when the valve does not seat cleanly upon closing. It is obvious that this behaviour is promoted at high rpms, when acceleration and/or any of the higher derivatives of the lobe profile might exceed the capability of mechanical system to control valve motion.
It is also obvious that a marginal system might be "pushed over the edge", either by excessive rpm, or excessive delta P, or both in some combination. Based on the above discussion in this thread, a mechanically supercharged engine might experience a significant delta P on the intake valve at IVC, and thus be vulnerable to this failure mode. By the same token, the exhaust valve of a turbocharged engine with an adverse exhaust to intake delta P would also be vulnerable to this failure mode. Interestingly, the adverse intake to exhaust delta P is (generally) a positive function of rpm (at least as redline is approached).
From a symptomatic point of view, I doubt if you could tell the difference between intake and exhaust hydraulic adjuster pump-up.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[gruntguru]

After a little thought I have decided there is probably no effect.

On the same engine with and without boost, there is essentially no difference in pressure drops throughout the system. This is because volumetric flow through the system is unchanged. Mass flow increases with boost but this is due to density increase. The same applies to throttling where VE and volumetric flow rate is constant at any given rpm - what changes is density.

je suis charlie