61 replies to this topic

[primer]

BMW S1000RR Technical Video - Valves in Motion High RPM

Great stuff.

[Canuck]

Are those springs turning the direction it appears they are, or is it an optical illusion?

[primer]

It is difficult to say for sure because the video does not show the engine coming to a complete idle and stop, but I'd say yes they are rotating. The stroboscopic effect should not make them appear to rotate since the motion is happening in a perpendicular direction.

[SteveCanyon]

Yes they're rotating but I'm not sure at what speed.
You can see some marks on the springs & retainers going around and around, but due to the frame rate of the camera you can't really tell how far they rotate for each frame. The only thing I can think of that suggests that they really are rotating at the speed you see in the video is that the revs pick up slowly and the spring assembly seems to start to rotate smoothly in the higher revs, and doesn't jump around which is what you'd expect as it synchs & unsynchs with the frame rate.

[Canuck]

Sorry - what I was referring to was the apparent direction change at the higher end of the scale. You can clearly see the paint marks turning, but they appear to change direction at some point.

[Tony Matthews]

Sorry - what I was referring to was the apparent direction change at the higher end of the scale. You can clearly see the paint marks turning, but they appear to change direction at some point.

Yes, it could be 'wagon wheel' effect - how many times have I seen four sweating horses, whipped mercilessly, accelerating a covered wagon only for the wheels to suddenly go into reverse rotation. All that effort for a nose-bag of oats... I know odd things happen, but for the direction of rotation of the springs to change does seem odd, rather than just a change in speed of rotation.

[McGuire]

Unless the valves are equipped with rotators, valve/spring rotation can vary, start, stop, or change direction in response to speed/load. Thus the term positive or active rotators. A typical positive valve rotator might employ a set of concentric ball ramps so the valve ratchets one "step" each time it is opened. In a release or passive valve rotation setup, the valve is free to spin when the assembly is in equilibrium at some point in opening, subject to valve follower offset, retainer key interference angle, etc. Positive rotators are usually integral to the valve spring retainer, but not always -- they may be installed at the bottom of the valve spring, as with some BB Chevrolets for example. The passive system is often adequate these days for passenger vehicles. On many engines positive rotators are employed on the exhaust valves with passive rotation on the intakes.

[zac510]

That would be great with a high fps / slow motion camera

[primer]

That would be great with a high fps / slow motion camera

Indeed, Time Warp this.

[McGuire]

That would be great with a high fps / slow motion camera

http://www.youtube.c...feature=related

[Terry Walker]

More jiggle than the Spice Girls - fantastic sequence, who would have thought the spring coils would bounce so much around between lifts?

[Tony Matthews]

Yes, amazing. Just think what would be going on without the inner coils!

[gt40jim]

One of my jobs is taking care of a Camping world truck. We run an SB2 Chevrolet motor. It is a real low budget team and have to run parts a little long. The amount of wear that occurs on the bottom of the valve retainers from the relative movement of the spring is astounding to me. The retainer will wear to paper thin if not changed out. Just throwing this into the conversational mix.

[gruntguru]

http://www.youtube.c...feature=related

Love this clip. (Probably done with a strobe rather than high speed camera.)

Much more useful than the clip in the OP which appears to show the valves constantly closed at high speed. (Camera is only capturing the valves effectively when they are stationary)

Edited by gruntguru, 21 February 2010 - 04:36.

[Lee Nicolle]

I ran a roller cam SBC with a hole cut in a rocker cover up to about 6000 rpm. You can watch the valves, springs and retainers all turning [as they are supposed too] The rockers look very busy but they never seem to be moving 600 thou. With 3/8 studs you could see delection which was not there when changed to 7/16. This was the purpose of the exercise.
While I am sure 7/16 deflect too it is defenitly less with that rate of spring and lift.

[gruntguru]

I ran a roller cam SBC with a hole cut in a rocker cover up to about 6000 rpm. You can watch the valves, springs and retainers all turning [as they are supposed too] The rockers look very busy but they never seem to be moving 600 thou. With 3/8 studs you could see delection which was not there when changed to 7/16. This was the purpose of the exercise.
While I am sure 7/16 deflect too it is defenitly less with that rate of spring and lift.

Did you use a strobe? Timing light with advance-check knob does the trick. Then you will see the 600 thou. Unless you are running a cam with 550 thou valve lift.

[McGuire]

Love this clip. (Probably done with a strobe rather than high speed camera.)

Much more useful than the clip in the OP which appears to show the valves constantly closed at high speed. (Camera is only capturing the valves effectively when they are stationary)

Nope, that's high-speed imaging. I believe you are thinking of high-speed still photography. Regardless of the means of illumination used, a high frame rate is required to obtain moving images like this. Note the play rate of the video above. The engine is reportedly running at 7000 rpm, while the video shows shows the valve events slowed down to around one per second or so. Obviously that can't be done with a standard rate in the range of 20-40 frames per second.

The first video of the BMW motorcycle valvetrain apparently employs a standard frame rate or thereabouts. Thus the valves appear constantly closed, due to the illusion known as persistence of vision -- the valves are indeed closed the majority of the time, so that is how they appear to the eye, or with a normal motion picture camera (film or video). If you were looking in on this scene in person, that is what you would see. If you were to illuminate this scene with a strobe light, you would see the valvetrain apparently "frozen" in position at the instant the strobe fired. (How a timing light works.) In order to show fluid motion, the strobe would have to be synced to the camera's frame rate. And with a standard, non-high speed frame rate, you would see the same thing as this video, only more brightly illuminated.

Strobes are sometimes used in high-speed motion work, mainly when high illumination levels are required that could adversely heat the subject. A strobe's intermittent illumination creates less heat.

[McGuire]

One of my jobs is taking care of a Camping world truck. We run an SB2 Chevrolet motor. It is a real low budget team and have to run parts a little long. The amount of wear that occurs on the bottom of the valve retainers from the relative movement of the spring is astounding to me. The retainer will wear to paper thin if not changed out. Just throwing this into the conversational mix.

Thanks for the interesting input. Sorta goes to show how good springs have gotten. Once upon a time when you saw a retainer like that, there was a broken spring under it.

[McGuire]

Here is another fun high-speed video -- Jay Brown's reproduction SOHC Ford V-8. Key on the rate of valve operation, and then note how quickly that bolt unscrews and jumps out of its bore.

[Tony Matthews]

Key on the rate of valve operation, and then note how quickly that bolt unscrews and jumps out of its bore.

Looks like about one and a half valve cycles from done up to gone! Astonishing.

[gruntguru]

Nope, that's high-speed imaging. I believe you are thinking of high-speed still photography. Regardless of the means of illumination used, a high frame rate is required to obtain moving images like this.

Yes you are right and I was wrong to say it doesn't need a high speed camera. The point I wanted to make was that it doesn't require a high frame rate (and the clip in question was almost certainly not recorded at a high frame rate and almost certainly used a strobe for illumination.)

What is required is a short exposure either by using a high shutter speed or brief intense illumination (strobe) during a somewhat longer shutter event. The illusion of slow motion is achieved simply by synchronising the exposure to the cyclic event being recorded and changing the phase of the synching at the speed desired for viewing.

[gruntguru]

Here is another fun high-speed video -- Jay Brown's reproduction SOHC Ford V-8. Key on the rate of valve operation, and then note how quickly that bolt unscrews and jumps out of its bore.

A good illustration of the point in my previous post. The video is playing at actual speed so the bolt is not unscrewing in slow motion. The "apparent" slow motion of the valve is created by synching the exposure (and probably a strobe light) to each cycle of the engine.

[zac510]

Is there any benefit to controlling the speed of the spring after the valve has closed or not?

[McGuire]

A good illustration of the point in my previous post. The video is playing at actual speed so the bolt is not unscrewing in slow motion. The "apparent" slow motion of the valve is created by synching the exposure (and probably a strobe light) to each cycle of the engine.

There are no high-speed cameras or methods that operate in that manner. However, you are entirely welcome to think so. If you like, you could build such a rig to test out your theories but it won't work.

Edited by McGuire, 23 February 2010 - 15:48.

[desmo]

Isn't that essentially what an old timing light does? I must be missing something.

[McGuire]

Isn't that essentially what an old timing light does? I must be missing something.

Actually, it does rather the opposite -- a strobe light "freezes" or eliminates the appearance of motion.

[DesWeller]

If your strobe was synced to the engine, all you'd get is a very boring video of the apparently completely stationary assembly. You'd have to run your strobe at a frequency slightly above or slightly below the engine speed to get the appearance of continuous motion.

[desmo]

Obviously.

[gruntguru]

There are no high-speed cameras or methods that operate in that manner. However, you are entirely welcome to think so. If you like, you could build such a rig to test out your theories but it won't work.

OK - are you saying that the videos you posted are high frame-rate videos, shot under intense steady illumination, played back at slow speed and then a regular speed sound track dubbed over the top? If so you are wrong - there are numerous pieces of evidence simply watching the videos.

Edit. BTW, this statement is pretty big ("There are no high-speed cameras or methods that operate in that manner.") The concept I described is a very simple one and I don't find it credible that such a system would never have been built.

Edited by gruntguru, 23 February 2010 - 23:24.

[Lee Nicolle]

If you read the Smokey Yunick book he mentions putting a light inside the motor and then perspex panels in oil pans and timing covers to watch the action.
That inspired me though with valve train it is fairly easy cutting a hole in a rocker cover.I simply mounted a small globe down low near the head surface. And I had very little mess.

Edited by Lee Nicolle, 25 February 2010 - 22:14.

[gruntguru]

There are no high-speed cameras or methods that operate in that manner. However, you are entirely welcome to think so. If you like, you could build such a rig to test out your theories but it won't work.

It does work. Every time you see the "wagon wheel effect" in a video, that's exactly what you are seeing - near synchronisation of the frame rate to some multiple of the cyclic frequency of the subject.

[imaginesix]

It does work. Every time you see the "wagon wheel effect" in a video, that's exactly what you are seeing - near synchronisation of the frame rate to some multiple of the cyclic frequency of the subject.

In the case of the video McG linked to, the image is too crisp and the movement too fluid to be explained as a stroboscopic image, IMO.

[gruntguru]

In the case of the video McG linked to, the image is too crisp and the movement too fluid to be explained as a stroboscopic image, IMO.

On the contrary.
1. The image is too crisp to be produced by ultra-short exposure time. The light intensity required would be enormous.
2. Look at movements other than the valve train eg oil splashes. Rather than appearing as slow motion, they are stacato, jerky - indicating there are big chunks missing from the video record.
3. The sound track is playing at correct speed. This could be dubbed onto a slo-mo video but in this case I think not.
4. The light intensity varies frame to frame . Some frames are very dark. This does not happen in a HS video with steady illumination. The video was produced using a strobe running at slightly above or below half engine speed. The camera was probably running at the same speed as the strobe but with an exposure time significantly larger than the strobe duration.

[gruntguru]

http://www.youtube.c...feature=related

Sigh. . . . . Inspecting this video more closely, and accepting the claimed speed in the video title of 7000 rpm, it is most likely that the video was shot at 30p with a strobe speed of 29 Hz. These settings will give the observed combination of events as follows:
1. Engine speed 7000 rpm, valve opening frequency 58.3 Hz
2. Apparent valve frequency on video 0.3 Hz
3. Dark frames on video 2 Hz
4. Video format as presented on YouTube is 30 fps

The strobe is running at valve frequency (or half valve frequency) +/- 0.3 Hz so the possibilities are 58, 58.6, 28.85, 29.45
The camera is running at strobe frequency +/- 2 hz so the possibilities are 56, 60, 56.6, 60.6, 26.85, 30.85, 27.45, 31.45

All numbers are approximate and rounding needs to be applied (but wasn't to avoid masking the calculation method)

Edited by gruntguru, 26 February 2010 - 02:22.

[cheapracer]

Sorry if it's posted above i can't get Youtube but this is obviously the same engine from the chamber view.

http://v.youku.com/v...U0Mzk5NzAw.html

[gruntguru]

Sorry if it's posted above i can't get Youtube but this is obviously the same engine from the chamber view.
http://v.youku.com/v...U0Mzk5NzAw.html

Interesting vid but not even close. I have downloaded the YouTube vid at work. I will try to post it somewhere for you on Monday.

[McGuire]

Sigh. . . . . Inspecting this video more closely, and accepting the claimed speed in the video title of 7000 rpm, it is most likely that the video was shot at 30p with a strobe speed of 29 Hz. These settings will give the observed combination of events as follows:
1. Engine speed 7000 rpm, valve opening frequency 58.3 Hz
2. Apparent valve frequency on video 0.3 Hz
3. Dark frames on video 2 Hz
4. Video format as presented on YouTube is 30 fps

The strobe is running at valve frequency (or half valve frequency) +/- 0.3 Hz so the possibilities are 58, 58.6, 28.85, 29.45
The camera is running at strobe frequency +/- 2 hz so the possibilities are 56, 60, 56.6, 60.6, 26.85, 30.85, 27.45, 31.45

All numbers are approximate and rounding needs to be applied (but wasn't to avoid masking the calculation method)

If you record and play the image at 30p it won't be in slow motion.

[cheapracer]

Interesting vid but not even close. I have downloaded the YouTube vid at work. I will try to post it somewhere for you on Monday.

There may be some confusion here, the Vid in the OP post is also at Youku and I have seen it, the one I posted just above was right next to it.

Thanks for your kind offer though

[gruntguru]

If you record and play the image at 30p it won't be in slow motion.

Correct.

(Except for the components that are operating just above or below the strobe frequency - ie the valve train. It is not "true" slow motion - its actually a precision version of the "wagon wheel effect".)

[gruntguru]

There may be some confusion here, the Vid in the OP post is also at Youku and I have seen it, the one I posted just above was right next to it.

Thanks for your kind offer though

One interesting thing about the vid you posted is the stop/slow-motion observed at several engine speeds. This is also "wagon wheel effect" occurring when engine speed is close to a multiple of video speed or vice versa.

Edit. Had another look at YouKu and didn't see the video. Oh sorry - its not the OP link - its McGuire's first link - this one
http://www.youtube.c...feature=related

Edited by gruntguru, 28 February 2010 - 03:53.

[McGuire]

Correct.

(Except for the components that are operating just above or below the strobe frequency - ie the valve train. It is not "true" slow motion - its actually a precision version of the "wagon wheel effect".)

That's why your idea would look nothing like the video -- far too little visual information would be collected. And that's why nobody uses your idea. High-speed motion pictures like these are done with a high-speed camera.

[Powersteer]

It looks similar in design to the Ferrari F1 V10 illustration way back. Is this where BMW F1 tech has channelled to because it certainly look like it other than the pnaumatic valves. Video has been removed.

[McGuire]

OK - are you saying that the videos you posted are high frame-rate videos, shot under intense steady illumination, played back at slow speed and then a regular speed sound track dubbed over the top? If so you are wrong - there are numerous pieces of evidence simply watching the videos.

Yep, that is essentially what I am saying. However, I would not get too hung up on the illumination or the sound. That's not an engine you hear running. That is a cylinder head assembly mounted on a test rig built by Bill Conley, which he calls an Ottodyn. The video camera peers in through the two tubes you can see on the far end of the clear plastic enclosure.

[cheapracer]

It looks similar in design to the Ferrari F1 V10 illustration way back. Is this where BMW F1 tech has channelled to because it certainly look like it other than the pnaumatic valves. Video has been removed.

which video? The OP's Vid? if yes I know where it is at other places.

[gruntguru]

Yep, that is essentially what I am saying. However, I would not get too hung up on the illumination or the sound. That's not an engine you hear running. That is a cylinder head assembly mounted on a test rig built by Bill Conley, which he calls an Ottodyn. The video camera peers in through the two tubes you can see on the far end of the clear plastic enclosure.

The video in question is not produced with a constant light source.
1. Have a look at it frame by frame - there are dark frames at regular intervals (2 Hz).
2. Oil splashes are present but not slowed. The 200:1 frame rate ratio would slow the oil dramatically. (same for the unscrewing bolt in your second video.)
3. The sound track is obviously at full speed (7000 rpm). It doesn't matter if its a running or motored engine.

I suggest you ask Mr Conley about these videos before you attribute them to his technique.

[GrpB]

I would not get too hung up on the illumination or the sound.

This is true. The specifics of the video are not as important as the illustrative nature showing the characteristics of relative motion and lubrication. They give a general sense of what is happening inside the cam cover, and therefore are useful for diagnosing a specific durability or wear issue. If you were trying to infer something like valvetrain stability at various rpm without the other requisite data collection you certainly wouldn't use video, you would just close your eyes and listen for the changes in sound as the rig was motored through a range of rpm.

I don't know about specific video techniques, but I know that I would prefer cell phone video of valvetrain motion as compared to anything with a strobe for the things that the video would be useful for. What does it matter anyway?

[cheapracer]

What does it matter anyway?

Ban him.

Had another look at YouKu and didn't see the video.

http://so.youku.com/..._video/q_engine

Edited by cheapracer, 28 February 2010 - 15:31.

[gruntguru]

Ban him.
http://so.youku.com/..._video/q_engine

Yeah that's the OP video. I will try to post the other one tomorrow.

[gruntguru]

I don't know about specific video techniques, but I know that I would prefer cell phone video of valvetrain motion as compared to anything with a strobe for the things that the video would be useful for. What does it matter anyway?

Perhaps I am too much a stickler for technical accuracy. On the other hand this is a technical forum - I am sure there are others here?

Regardless, when a poster claims that a valid, popular, real-world technique doesn't exist and "won't work" I feel compelled to correct them.

The video posted by McGuire is high quality footage and illustrates a number of things you would not perceive using your ear, your mobile phone camera or both.

Edited by gruntguru, 28 February 2010 - 22:03.

[GrpB]

Perhaps I am too much a stickler for technical accuracy. On the other hand this is a technical forum - I am sure there are others here?

Regardless, when a poster claims that a valid, popular, real-world technique doesn't exist and "won't work" I feel compelled to correct them.

Well that's fair, I added nothing technical and I have nothing more to offer than what I already said. But pursuing ad naseum the way the video was captured doesn't seem like a technical discussion to me either. If they took the video to help diagnose some characteristic wear or failure issue they see in durability testing or competition then certainly the appearance(s) of motion relative to the frame rate must be tempered by the failed/worn components they have inspected as well as the other data collection and analysis that would no doubt have come first. Video isn't measurement, they would use other methods to determine actual valve or spring motions, stresses, loads, phasing, etc in the normal course of development. If those things are not sufficient to produce a satisfactory explanation for the issue, you take video. As such, you want unadulterated video with maximum light and frame rate because "something else" is going on. Because valve motion is entirely about the interaction of a bunch of parts of various materials, stiffnesses and natural frequencies usually with sporadic lubrication, you absolutely don't want to miss some visual clue because the lights are turned off.

So regardless of the camera frame rate, given the choice between keeping the light source turned on, or intentionally turning it off periodically, well, that doesn't seem like a deep technical question requiring discussion. Whether to take video for valvetrain as shown on this thread, or to diagnose an accesory belt getting tossed, or to look at the characteristics of tire delamination, I have seen/requested a decent range of high speed video and have never seen strobes being used. If anything, the shadows as a result of the surrounding componentry/rig hardware/etc, are the bigger problem in terms of lighting.

In any case, tell me what you saw, or what you were looking for, otherwise I stand by my question: what does it matter how the video was taken?