About time for the spintron

Valve bounce on seating, irrespective of rpm, can be the result of a combination of forces, main ones being the design of the closing ramp, and the oscillation of the spring coils.
In the case of the latter, first consider the movies of oscillating springs which I'm sure most interested members will have come across at some time or other. If on the seating of the valve the spring surge happens to be downwards, the valve can under this influence open slightly, to close again as the spring surge reverses. The factors influencing this are most often unfortunate choice of components that have very similat natural harmonic frequencies, which at certain engine speeds vibrate each in step with the other. It's rather like troops marching in step making a bridge collapse.
The best way to minimise this situation is to ensure that the components chosen for the valve train have natural harmonic frequencies that are as far apart as can be arranged. With a pushrod engine, main problem items in this respect are pushrods and valve springs.

Single valve springs can be problems because there is nothing to act as a damper on the oscillations. To overcome this coils can be progressively wound, and the springs made beehive or conical. Either way, as the spring is compressed and the bottom coils (usually) become compressed, the spring rate and the harmonic frequency of the spring changes. With dual springs, an inner spring that is an interference fit inside the outer has a similar effect.

The closing ramps of the PW3 are rather short and somewhat abrupt. The valve is returned sharply to it's seat at any engine speed, and the higher the engine speed the less time is available to ease the valve down to seat without bounce. Add to this the valve train harmonics which are always added into the equasion, and the best efforts to eliminate bounce can become a most frustrating exercise.

To investigate the posibility of overcoming some of the problems as stated above, I redesigned the PW3 to seek a solution to the bounce.
With the 10th attempt I ended up with a valve lift design that had absolutely no trace of bounce at 9500 rpm, but this was achieved assisted by a change to pushrod in both dimensions and material, the steel valve and the ovate wire beehive spring remaining unchanged.
Note:- this investigation was carried out for 9500 rpm only. It is very possible that at a lower rpm there is still bounce, as with a change in engine speed the harmonic frequencies of the valve train will also change.
What did I end up with?
A valve lift design that is no longer a PW3, and who wants to rev a 750 Commando engine to 9500 rpm anyway!

Jim's spintron may throw further light on this issue if he can marry up a picture of the valve bouncing with the spring oscillations to see whether the oscillations appear to be the cause of the bounce.

Snotzo said:

... dual springs, an inner spring that is an interference fit inside the outer has a similar effect.

Click to expand...

Is there a combination of springs and spring rates that could damp the ocillation or smooth the waveform without having the interference fit ? If the valve train has something rubbing, wouldn't a ( not sure how to say this) critically damped friction damper/spring system do the job ?

Snotzo said:

.... light on this issue if he can marry up a picture of the valve bouncing with the spring oscillations to see whether the oscillations appear to be the cause of the bounce.

Click to expand...

That's an interesting thought. A back of the envelope calculation says that valve would drop about 0.025" due to gravity at 7200 rpm. But it is not out of control the whole time. The spring can't push the valve off the seat, can it ? How elastic are valve stems ?

Amazing stuff. I hope that what ever the outcome of this is, you will let us know what worked.

Greg

Well, what do you think.
It doesn't look to me like the spring oscillations go along with the valve bounce. I would guess valve stem and head flex. Jim

[video]https://youtu.be/8M3DW8F2CA8[/video]

comnoz said:

Well, what do you think.
It doesn't look to me like the spring oscillations go along with the valve bounce. I would guess valve stem and head flex. Jim

Click to expand...

You beat me to it. I was just wondering if the valve head distorted. Empiricism rules. Now you need some kind of interferometer that can look at a large field.

Greg

Jim/Snotzo,

Apart from this being a fascinating thread all on its own I am engaged because I have a fresh 750 short stroke with a PW3.

I also have beehive springs, titanium retainers and steel tube pushrods. Perhaps the main difference to the spintron is that I have standard size inlet valves with a small backcut, so marginally less weight than those in the spintron I guess.

As far as I understand the exercise from Jim's perspective the issue is to investigate valve train performance at much higher rpm that typically used. As Snotzo says, who wants a 9500 rpm 750 Commando?, well maybe those heading to the salt! For my part I am happy with the potential for peak power at 7000 and being able to run over by 1000rpm in moments of need, relatively modest demands from 80.4 stroke!

I have only ridden a couple of track sessions so far, running up to 6000, and the highest rpm the engine has seen was when my wife exceeded my request and 7000 as I strobed it yesterday! I am off to the dyno tomorrow, but just for a basic set up/safety check on the jetting etc. So again no high rpms (or power figures) expected.

I shall be testing and then racing at Cadwell at the beginning of the month, but it will still be too early to provide any comments regarding power delivery characteristics, let alone possible valve train issues. (It always was a long term project....but sadly it has taken a lot longer than I had hoped to get here).

So my questions are;

Is the lighter valve likey to have any benefit in reduced bounce?

Does the 5" radius show any benefits in the simulation? (or perhaps later in the spintron)

Or should we look at another cam grind for a short stroke? and if so how is that grind likely to affect actual power delivery on the race track, bearing in mind that rider comments on the PW3 are generally very positive (though of course in general relating to 7000 rpm 89mm motors)

Would it help to use titanium hollow stem valves to reduce valve mass.
Ducati use such valves on their superbikes race bikes which rev to 12,000+ rpm. The valves are fairly large even in their 4 value combustion chamber. Obviously, titanium would be an expensive alternative, but it has the strength for the application.
A sodium filled stem would also improve heat transfer on the exhaust side.

comnoz said:

Well, what do you think.
It doesn't look to me like the spring oscillations go along with the valve bounce. I would guess valve stem and head flex. Jim

Click to expand...

If F really =-kx for a spring, there is no frequency term in the force applied by a spring. What happens if you try to pull a valve through a hole the size of the valve seat, using the same amount of force the spring generates at full lift ? Can you detect any change in length ?

Is there a way to calculate the harmonic frequency of the intake valve ?

Greg

SteveA said:

Jim/Snotzo,

So my questions are;

Is the lighter valve likey to have any benefit in reduced bounce?

Does the 5" radius show any benefits in the simulation? (or perhaps later in the spintron)

Or should we look at another cam grind for a short stroke? and if so how is that grind likely to affect actual power delivery on the race track, bearing in mind that rider comments on the PW3 are generally very positive (though of course in general relating to 7000 rpm 89mm motors)

Click to expand...

A lighter valve would likely have benefits although, especially with Ti, there may be trade -offs due to the additional flexibility.
With a Ti valve you can not have valve bounce. Failure will happen quickly.

I don't yet know if a radius would help with a PW3 -I have a feeling that it will not.

I would think there are better cams for a shortstroke than a PW3. The gains to be had from a shortstroke over a longstroke are not going to be seen until over 6500 or so. That is about where a PW3 starts showing problems. How much it will affect power -I don't know. I do know it will affect valve and seat life. Jim

Steve

very difficult to eliminate valve bounce completely, especially when an existing cam is in use. Some small bounce will most certainly be present somewhere in an engines operating speed range (we are specifically talking Norton Commando engines here, not other makes and/or types).
This small bounce is not going to be a problem in use, and would certainly not be noticed in the general running of a road machine. An occasional missed gear will possibly cause a valve to bounce, but unless the bounce is on the exhaust side, little harm is likely because the intake valve is well away from the piston. The exhaust side however is not free from problems, because if the bounce is severe, it increases the risk of valve clash, and piston clash.
Any bounce less than 0.001 thou will not cause problems, but 0.010 thou most certainly will if the engine speed at which it occurs is deliberately and continuously used.

The foregoing is generally descriptive of results arising from small design flaws in closing ramps aggravated by spring surge. A far worse bounce can occur when the engine speed is such that the valve train lags behind the cam during closing so the valve ends up crashing down onto the seat without the guidance of the cam closing ramp. Valve stretch and head deformation are the best one can hope for in such an eventuality. At worst the valve will quickly fail. Such can happen following a missed gear change, but very often an occasional trip into the red zone can survive.

The Titanium valve as an alternative to a steel original component is not worth the cost for a normal road going machine - even one driven in a sporting manner as some members seem to delight in telling us. Ti really comes into it's own when a race engine is built for use at higher engine speeds than a steel valve will permit. In such an engine any reduction in the weight of components on the valve side of the rocker pivot is highly beneficial. This comment also applies to a single valve spring, be it beehive or conical. The end of the spring nearest the retainer moves with the valve and a single spring is lighter than a pair of springs in this respect.

When Jim gets to the Ti valve he should be able to see if the bounce situation is eased compared to the steel valve, but ideally the movement of the valve needs to be measured , as visual estimation gets difficult when trying to evaluate small movement. Give Jim time, I'm quite sure he'll get there, but for the moment his films are fascinating to watch.

comnoz said:

SteveA said:

Or should we look at another cam grind for a short stroke?

Click to expand...

........

I would think there are better cams for a shortstroke than a PW3. The gains to be had from a shortstroke over a longstroke are not going to be seen until over 6500 or so. That is about where a PW3 starts showing problems....Jim

Click to expand...

The Axtell 3 cam lobe (same as Megacycle NR) had long duration just off lift with less nose. That way you still had a hot cam with less lift that allowed you to shim it tighter so there was more pressure with the valve on the seat. The longer duration cams with less lift could be revved higher without valve float. This is the cam Ron Wood used for his 84 HP short stroke. When talking to Axtell he said that at & above 8000 RPM the racing valve springs could bind the coils and heat up red hot and anneal/soften. All this was with stock lifters. Axtell and Wood were on to this and did not use higher lift cams. Now this thread is suggesting that such cams may produce more power at higher RPM than higher lift cams by avoiding valve bounce. For me it was all about keeping the valves from going through the pistons.

The lower lift Axtell#3/Megacycle NR is compared to the PW3 below

JS
I totally agree with you that high lift in a pushrod engine is not the be all and end all of a high power output. As I stated in an earlier post, my choice of the PW3 as an example was perhaps rather unfortunate, but as I had measurements from a new item I used the data to investigate the bounce possibilities at high rpm.
For those riders who use their machines for pleasure on the roads, the PW3 probably finds favour with many, but when working up a production engine towards top level race performance, one is to a great extent travelling through uncharted territory, and most folks would be limited to making use of components that are readily available, camshafts being a major item in this respect. If the engine is a long stroke 750, then the PW3 would most probably be a good choice to start with. Pushing the revs higher via a shorter stroke/larger bore combination, and one will almost inevitably end up searching for something better suited than the PW3.

To add something to this post that I forgot to include in my previous re double valve springs.
A pair of different springs (inner and outer) will each have its own unique mean harmonic frequency. In times previous the thinking was to have springs possessing different harmonic frequencies, the idea being that as one spring entered it's harmonic region, the other would be unaffected and would support the first through it's difficult period. Then at a different rpm the reverse would be the case. The problem found with this scheme was that the springs would be often operating below their maximum potential, and so an alternative arrangement was sought.
The most often used solution to this problem was to select two springs that had near identical harmonic frequencies, but frequencies that were outside the range of engine speeds the engine was intended to use. Add to this the damping from an interference fit inner spring, and the result can be a very satisfactory arrangement, provided of course that the wire stress remains within the springs safe limit for the intended use.
This does not mean that the harmonic frequency based oscillations will be magically removed from the springs. Lower order frequencies will still be present, but the major frequencies will have been 'managed' so they do not combine to produce unwanted effects.
But in all probability there will still be a little valve bounce occurring somewhere within the engines operating speed range, because it will be impossible to damp out everything, everywhere, all the time.

This thread is what it's all about! Thanks to all the nerds posting the good stuff on here!

I'd like to ask an 'elephant in the room' question re the PW3 cam...

Peter, as his father before him (Jack), was an outstanding engineer, particularly clued up on cam design. So, why did he get it so wrong with regarding to 'missing out' the ramps?

And why do so many people (inc top guys like Hemings and White) swear by it so much?

Ok, that's two questions! But they're closely related...

Fast Eddie said:

This thread is what it's all about! Thanks to all the nerds posting the good stuff on here!

I'd like to ask an 'elephant in the room' question re the PW3 cam...

Peter, as his father before him (Jack), was an outstanding engineer, particularly clued up on cam design. So, why did he get it so wrong with regarding to 'missing out' the ramps?

And why do so many people (inc top guys like Hemings and White) swear by it so much?

Ok, that's two questions! But they're closely related...

Click to expand...

First off, as a personal preference, I am not a big fan of the PW3 cam -not because of the grind but because of the construction. The cast iron cam deflects more and is somewhat brittle. The threads are weak and I have seen a lot of cases where the bearing journals had more wear than they should.
I prefer a steel cam and there are grinds similar to a PW3 that are steel.

As Snotso said, the PW3 is a fine grind for those long stroke engines that spend a good part of there life below about 6500 rpm. At 6000 rpm there is enough ramp to provide easy valve closing and they can make good power.

For a higher rpm shortstroke engine a cam with less acceleration and smoother ramps is going to be necessary.


I am getting to the point where I can get an idea of the higher rpm bounce I was seeing with the camera by using a laser sensor. The calibration is still rough but where I had observed what looked like .030 bounce around 7500 rpm the laser is saying between .020 and .030. It also shows there are more bounces than I can see by eye. Jim

Here is a better visualization of the Axtel #3 high rpm cam in comparison to the PW3. You can see the reduced lift and the better trailing ramp to ease the seat landing. Also shown is the stock Commando cam. [barley visible] Jim

As you can see here, there are many cams that are very close to the PW3 grind. This is the intake lobe from several different grinds. they are all close enough that the difference in power would be negligible.
You can also see the 86c cam [which is the grind I prefer] has a larger ramp on the trailing edge than the rest of the cams. That makes it rev easier with less seat bounce. Jim

Good way to illustrate why cam timing figures are sometimes given by the manufacturers at 0.040" lift.

For cam range & scope & pecking order character in one place - though not detailed as comnoz spoils us. Apparently all current valve trains will bounce in red zone so next step is toughening up to stand intervals of bouncing at WOT rpm 'extremes'.
http://www.nortonownersclub.org/support ... /camshafts

viva la difrencia

Here is the Webcam 86C with the same Honda valve springs, stock pushrods and lifters and lightened rockers from 7000 to 8000 rpm

Darn, audio problems again. The video starts at 7000 rpm [no bounce], then a pause and then 7500 rpm [tiny bit of bounce if you look close] and a pause and then 8000 rpm [still acceptable bounce].

[video]https://youtu.be/nH2oLU7vPTI[/video]


Here is the Laser sensor output at 5000 rpm



And again at 7000 rpm

Great stuff here Jim.

Can you show a plot of the Web 86C against the PW3 and Megacycle NR (Axtell #3)

Dances with Shrapnel said:

Great stuff here Jim.

Can you show a plot of the Web 86C against the PW3 and Megacycle NR (Axtell #3)

Click to expand...

I can't show that with the laser sensor in the spintron but the plots from the cam measuring system are on the last page.

I do not have an Axtel cam here to run in the spintron.

The spintron uses needle rollers in the quick change cam adapter. The cast iron PW3 cam I have failed where it was running on the needle bearing in just the few minutes that it run. I will not be able to run it again unless I sleeve the bearing surface. Jim