Measuring port velocities

[CraigAPD]

Hi Guy,

I presume you have facilities to measure port velocities whilst testing on your flowbench?

What display do you use? I want something digital so that rules out a simple manometer to give a differential figure (from which I would calculate vel.).
I have a velocity gauge in mind from the states, which I can use directly with my Pitot tube, but wondered if you have any other suggestions, preferably UK based?

Regards

Craig

[Guy Croft]

Nope, I don't measure velocities.

GC

[CraigAPD]

I see,

Any reason for this?

[Guy Croft]

Most likely, Craig, it began with my not being able to afford any add-ons at all when I bought the SF rig!

What it would tell me now, having worked without it (around it?) for 4 years or so I'm not sure really to be honest.

If there is anything you want to ask about my methods, feel free, it's a big subject, so you might have to prompt me a bit.

G

[chrislandy]

There are so many variables with velocity, because velocity is speed with direction it is different at every point in the inlet tract, in my opinion it is pretty much pointless trying to measure it.

Flow rate is the important measurement as this is the volume of liquid/gas related to time (l/s)

Chris.

[CraigAPD]

chrislandy..velocity is very much as useful as CFM.

volumetric flow rate is related to time directly through velocity. without velocity you have no flow.

And yes its different at numerous points, but that's half the reason for measuring it, to see if a CFM gain can be achieved by "balancing" velocity across sections of the port. Especially so on the inside radius, where separation is likely to occur, and a lower percentage of the valve area is being "used". If I can measure velocity I am in a better position to determine flow path behaviour through the port, and so modify accordingly in order to increase CFM.

Besides all that, velocity provides the fundamental part of cylinder filling...without it we wouldn't be making torque! The higher the better...in that respect. I've been porting heads using a flowbench, without measuring velocity for a few years now, and I believe I've reached the limits of what I'm testing, in that with the heads I currently work mainly on, I can modify them to yield best cfm AND power (the two don't always coincide).

My new aim is to corrolate velocity and cfm to performance trends on my engine dyno. to do this I need to measure velocity. from this I can better determine how port behaviour on the flow bench will alter ACTUAL performance data. As it stands, measuring soley CFM on a flowbench is pretty nonedescript in the scheme of things. Measuring flow at various lift values at ONE depression into an infinite volume is far from the actual dynamic and transient states encountered on an operating engine. Its a start at heading in the right direction, but I think there's more to be gained...

[Guy Croft]

Can't fault what you're trying to do there, Craig.

G

[CraigAPD]

Its all intresting stuff I reckon!

[chrislandy]

I openly admit that I am no expert by any means and I am trying to remember theory from my hydraulics classes years ago so forgive me if I make wrong assumptions. I do understand your reasoning for wanting to measure the velocity at numerous points within the port or chamber.

Surely the introduction of a pitot tube or other measuring device into the port would greatly effect the velocity in that area by creating turbulence, this could be reduced by decreasing the diameter or size of the measuring device but due to the relativly small port it would still effect readings considerably. Plus being able to re-measure at the exact point to allow comparison would be very difficult especially if you are trying to measure a point deep inside the head, unless you remove the valves.

Why not use a visual method to see whats going on e.g. drawing smoke through? You could see what was happening, where the turbulence is, where the air may stall etc

Chris.

[CraigAPD]

The head if the pitot tube is in un-interupted flow, due to the nature of the shape of it. Besides the head is only 3mm diameter.

Its standard practice in the lots of science based appliactions (measuring flow in small ducts) using the same method im applying, without any induced error.

A visual/wet flowbench is all well and good, but it has no quantative value, so I cannot draw a fair/reliable/repeatable comparison between various head types or port shapes. Though yes they are handy for "lookign at" flow patterns in general.

[robert kenney]

I don't want to argue just post my thoughts.
The psychics of velocity flow is always validated on a test fixture E.G. wind tunnel or in the case of engine design on a flow bench and dynamometer.

A impact pressure(pitot tube) works well on straight parallel flow but induce turbulence in the flow and the study becomes exponentially more complex then couple in valve movement and standing wave influence and how do you validate the information you would gleam from a velocity study. I am not by any stretch saying that one would not learn from a well engineered velocity study but ultimately the only way to quantify your results would be in a test cell and then on the track.

I would think Doppler Fluid Velocity Measurement would more fit the bill here. Is utilizes a non contact passive technique that will not influence the flow at all and the doppler effect to detect velocity. I don't own any such piece of equipment nor can I afford to buy one but I have seen one in action. Very impressive but again.

The life time of data GC has from his flow bench correlated with his dyno runs are far more valuable in the real world.

I suppose this is only my opinion as I can not under take such a study nor do I intend to.

Regards
Robert

[CraigAPD]

I don't want to argue just post my thoughts.
The psychics of velocity flow is always validated on a test fixture E.G. wind tunnel or in the case of engine design on a flow bench and dynamometer.

A impact pressure(pitot tube) works well on straight parallel flow but induce turbulence in the flow and the study becomes exponentially more complex then couple in valve movement and standing wave influence and how do you validate the information you would gleam from a velocity study. I am not by any stretch saying that one would not learn from a well engineered velocity study but ultimately the only way to quantify your results would be in a test cell and then on the track.

I would think Doppler Fluid Velocity Measurement would more fit the bill here. Is utilizes a non contact passive technique that will not influence the flow at all and the doppler effect to detect velocity. I don't own any such piece of equipment nor can I afford to buy one but I have seen one in action. Very impressive but again.

The life time of data GC has from his flow bench correlated with his dyno runs are far more valuable in the real world.

I suppose this is only my opinion as I can not under take such a study nor do I intend to.

Regards
Robert

All valid points, but testing on a flowbench is very much steady state compared to a live engine. There is no valve motion, there is no standing wave effect as we air simply measuring flow into an infinite volume. You are right about turbulence, as there will inevatibly be areas of localised turbulence in the port. However I do feel think that any induced by the probe itself will be minimal in that it wont effect what Im trying to "read". This is only confirmed by people in "our field" already carrying out this type of testing, with repeatable results.
As I said, the only validation i can do, which is specific to ME and MY work is testing my heads on my flowbench, then correlatiing to real data from MY engine dyno. This is definately enough data to form trends from and aid my development work, and really its the best an individual can hope to achieve...I think at least.

Im sure Guy has excellent experience and a large data base of dyno/flowbench figures, which he no doubt uses to his advantage in his work. But im not Guy, I do my own work, and therefore need my own data to draw upon.

[4v6]

I had also been interested in measuring velocity inside various areas of the ports for the reasons given by Craig.
Since my flow testing electronics were supplied from a company in the US, i contacted them with a view to measuring velocity, unfortunately they do not supply such equipment.
However, they were able to point me in the right direction ( i still havent aquired the equipment yet) and they suggested a Dwyer Instruments model 621-40 digital display which is used for such velocity measurements with pitot tubes.
Hopefully this may suffice for your purpose Craig.

Best wishes, Tony.

[Guy Croft]

This is an interesting thread, well done all contributors. I should like to write here - when I get some spare time.

GC

[CraigAPD]

Tony,

Its a Dywer display Im getting. Though the one you have given the part number for, just display pressure differential. This means you would still have to enter that figure into a spreadsheet to calculate M/s, or worse still write it down and calculate by hand!
The display I have ordered (or rather about to) will display velocity straight off, digitaly. Its marginally more expensive, but worth it in my opinion.

What hardware do you have from the States? Ive got an AudieTech based flowbench. Im guessing by you saying they dont offer any velocity probing equipment, you have the same hardware?

Guy,

Its a very intresting subject, im sure any input you give will be warmly recieved.