I built my own flowbench and it works.
- Thread starter Pablo
- Start date
Welcome!
By registering with us, you'll be able to discuss, share and private message with other members of our community.
Mike Licht
I was here first
- Joined
- May 23, 2001
- Messages
- 4,642
Audie Technology makes a inexpensive unit called the flow quick, they also have some pictures of an air supply (vac source) that you can make form some centrifigal fans that you can get at Granger. I built an air supply like that to power our bench we use to flow maf sensons. We needed something that would move lots of air to max out some of the sensors. I have 8 motors it draws almost 60 amps full out. Nice work on yours it is a great tool for learning.
Mike
Mike
Another option is to see if you can find an old bosch jetronic flow meter or one out of an older jap car. They use a trap door linked to a pot and it should give some simple results. As far as a way to check flow, how about some 1/4" or 1/8" clear tubing and some sort of smoke. It would give you a better idea of flow because the smoke dissapates and you can see turbulance easier.
pablo---i actually have a couple of superflow benches and they come with calibration plates------you could easily build a copy------it is a plate made of 18 ga steel (.062") -----it has two holes-----one is 0.312" and the other is 1.875"------it calibrates at three levels------small hole, large hole or two holes at the same time------at 20.4 inches it flows 225 CFM with both holes open------i'll have to look in the manual to see what it does with each hole individually open------calibrate at one point and other points on the curve are simple math------i'll send you a pic tomorrow------i also fabricated some valve actuators that are sort of like custom rocker arms------ if you see them you might get an idea on how to modify an old rocker arm set to actuate the valves...........RC
Well I've learned a lot of interesting things in the past couple of days
First off, I need to add a restrictor to the manometer line from the spark plug hole. The small pressure fluctuations that occur make reading the manometer almost impossible. Thankfully I saw someone mention a restrictor and it immediately clicked. I was reminded of the Y restrictor to the wastegate on our cars stock to delay the air flow. So I just bought a small needle valve so I can adjust the dampening I need.
The other really cool thing that I learned is that there is an engineering standard for accepted amount of flow through a given straight orifice. (No need to get those picts for me Rich, but thanks!) Apparently a general rule of thumb for orifice flow is that it has a coefficient of between .58 and .62 when compared to a perfect venturi of the same diameter.
A perfect venturi flows around 134 cfm per square inch of area at 25" pressure drop. So if you had a 1 inch area orifice plate you would flow about 80 cfm through it at 25inh2o. There are quite a few rules regarding how the air needs to reach the orifice and where the manometer readings need to be tapped but they are pretty simple.
What makes this all very easy is there are a few calculators online that will correct for any pressure drop, air temperature, pipe configuration, etc. It's pretty awesome. So making the bench read actual CFM numbers is pretty easy and very cheap. I will probably do that soon.
As for flow velocity in the port- what people are doing is making small pitot tubes that they move around in the port and this gives them flow velocities for different areas. What's neat about this is that you can apply some fundamentals about air flow speed to the port pretty nicely... apparently at 350 fps, air tends to seperate from a short side radius... things like that can be accounted for and adjusted.
Anyway this is all very cool stuff, thanks for the suggestions guys (good call on honing the schedule 80 pipe Carl)
once I get the bench nailed down I think im gonna do a step by step port/flow of a port to show where and what the biggest gains are for 445 heads. For ex. based upon what I've seen so far, gasket matching probably does almost zilch for you.
First off, I need to add a restrictor to the manometer line from the spark plug hole. The small pressure fluctuations that occur make reading the manometer almost impossible. Thankfully I saw someone mention a restrictor and it immediately clicked. I was reminded of the Y restrictor to the wastegate on our cars stock to delay the air flow. So I just bought a small needle valve so I can adjust the dampening I need.
The other really cool thing that I learned is that there is an engineering standard for accepted amount of flow through a given straight orifice. (No need to get those picts for me Rich, but thanks!) Apparently a general rule of thumb for orifice flow is that it has a coefficient of between .58 and .62 when compared to a perfect venturi of the same diameter.
A perfect venturi flows around 134 cfm per square inch of area at 25" pressure drop. So if you had a 1 inch area orifice plate you would flow about 80 cfm through it at 25inh2o. There are quite a few rules regarding how the air needs to reach the orifice and where the manometer readings need to be tapped but they are pretty simple.
What makes this all very easy is there are a few calculators online that will correct for any pressure drop, air temperature, pipe configuration, etc. It's pretty awesome. So making the bench read actual CFM numbers is pretty easy and very cheap. I will probably do that soon.
As for flow velocity in the port- what people are doing is making small pitot tubes that they move around in the port and this gives them flow velocities for different areas. What's neat about this is that you can apply some fundamentals about air flow speed to the port pretty nicely... apparently at 350 fps, air tends to seperate from a short side radius... things like that can be accounted for and adjusted.
Anyway this is all very cool stuff, thanks for the suggestions guys (good call on honing the schedule 80 pipe Carl)
once I get the bench nailed down I think im gonna do a step by step port/flow of a port to show where and what the biggest gains are for 445 heads. For ex. based upon what I've seen so far, gasket matching probably does almost zilch for you.
drunkenwhippet
New Member
- Joined
- Aug 24, 2008
- Messages
- 1
Looks great! wouldnt the intake manifold be a great entry for the air into the head? it'd be cool if you could flow multiple cylinders to get more flow through the intake and see how that goes.
First a nit: the restrictor in the wastegate y does not delay the boost in any kind of time sense, it provides a pressure drop due to the flow through it so the wastegate sees a lower pressure than the compressor output. Without that flow there is no pressure drop which is why you have to have the bleed valve (such as the stock wastegate solenoid, why you can't cap the vent on that solenoid 
), and why you only get 12 psi boost if you put the y in backwards.
How about adding a small tank between the manometer and the head? Something like a glass Ball jar with a couple of fittings in the lid for the tubing. That way if there is an accident all the liquid in the manometer will get caught in the jar instead of squirting into your head and plumbing and shop vac, and the jar volume will give you damping. You might still need the restrictor (which doesn't cause a pressure drop in this case since there is no flow at steady state).
As for your comment on gasket matching, I've always been of the opinion that it probably hurt, not helped, since unless you open up the manifold runners all the way up and give them a little taper you wind up with a flow path whose diameter starts small, goes up, then goes back down at the throat under the valve. That flows worse than a flow path that steadily decreases in diameter from inlet to the throat. Anyway, Dave Vizard shows some data from a Chevy head he sawed apart in part 6 (I think) of the series on gofastnews.com that makes it clear that the port opening already flows nearly double what the valve throat does. Lot of good points in Dave's series.
), and why you only get 12 psi boost if you put the y in backwards.How about adding a small tank between the manometer and the head? Something like a glass Ball jar with a couple of fittings in the lid for the tubing. That way if there is an accident all the liquid in the manometer will get caught in the jar instead of squirting into your head and plumbing and shop vac, and the jar volume will give you damping. You might still need the restrictor (which doesn't cause a pressure drop in this case since there is no flow at steady state
).As for your comment on gasket matching, I've always been of the opinion that it probably hurt, not helped, since unless you open up the manifold runners all the way up and give them a little taper you wind up with a flow path whose diameter starts small, goes up, then goes back down at the throat under the valve. That flows worse than a flow path that steadily decreases in diameter from inlet to the throat. Anyway, Dave Vizard shows some data from a Chevy head he sawed apart in part 6 (I think) of the series on gofastnews.com that makes it clear that the port opening already flows nearly double what the valve throat does. Lot of good points in Dave's series.
pablo----ive attached a few pics that may be of interest-------since i have two flow benches i do find lots of uses for them-------you may want to experiment with your vac to see if you can use the output port since you will want to flow the exhaust ports the opposite way------my superflow benches will flow either direction for use on intake/exhaust heads--------you can see in the pictures how i made custom adapters for various V6 items-------for cylinder heads i bored a hole in a giant block of delrin and placed studs so they would match where they are in the head----that way it is always lined up------its long enough to support the entire head no matter which cylinder i want to measure-------for intake and exhaust ports i machined delrin guides with holes that allow attaching to the same holes that hold the intake/exhaust manifolds--------those transition ports really do change the flow numbers a lot---------also shown is the calibration plate--------i designed a set of custom rockers to actuate the valves and hold a dial indicator on the spring retainer-------you could easily adapt an old set of rockers to do a similar task--------i made mine heavy duty enough that i don't have to remove the original springs but i have included a picture of the Goodson low pressure test springs designed just for use on things such as flow benches if you don't have the necessary heavy duty actuators------to test things such as MAF's etc i just attach a simple scan tool and power up the MAF and hook it to a GM computer so it reads flow--------makes it easy to test calibration of MAF's----------its interesting that on commercial benches they use both vertical and angled manometers and to increase range on some of them they use some kind of fluid that is VERY heavy-------don't know what it is but its over $130 a pint---------if you can make it out in the pics its the ones with the blue fluid--------the red fluid is a lot lighter and is used in the longer vertical units.................RC
Attachments
How about adding a small tank between the manometer and the head? Something like a glass Ball jar with a couple of fittings in the lid for the tubing. That way if there is an accident all the liquid in the manometer will get caught in the jar instead of squirting into your head and plumbing and shop vac, and the jar volume will give you damping. You might still need the restrictor (which doesn't cause a pressure drop in this case since there is no flow at steady state
if you look at my pictures closely you will see how superflow deals with "accidental overpressure"----------they put a cork plug in the exit of the manometer that floats up against a small opening causing it to close and prevent the escape of fluid-------simple but effective............RC
Richard,
Thank you for the great photos and information. You've given me a lot of good ideas.
A question, do you use anything to seal the head to the cylinder adapter?
Also, when you use the calibration plate, do you place the plate on the bore adapter opening by itself or do you put a tube extending above it?
Thank you for the great photos and information. You've given me a lot of good ideas.
A question, do you use anything to seal the head to the cylinder adapter?
Also, when you use the calibration plate, do you place the plate on the bore adapter opening by itself or do you put a tube extending above it?
An update on my bench...
I can't seem to figure out what is making it inconsistent. I get the same readings within about a 5-10% range but that is not good enough to be useful. The only thing I can figure is that the vacuum itself is inconsistent. I checked the voltage to the vacuum and it isn't fluctuating more than .1 volt so it's not that.
I think the only solution is to use the orifice plate so that I have a direct knowledge of how many cfm the thing is flowing. I just made one at work today which I will hopefully be able to use in conjunction with the calculator on this website Flowrate Calculation for an Orifice Flowmeter fyi, there are charts showing air density vs temp. At room temp the air density is less than the number they have in the calculator.. iirc its about 1.18. I'll probably have to tweak the flow coefficient a little to get the numbers to line up with something of a known flow rate.
BTW i calculated theoretical flow through that calibration plate using the above calculator at about 216 cfm. The only variables I guessed at were the temperature (i used 25c which iirc is 77 degrees f) and the flow coefficient. From what I have read the flow coefficient can vary from .58-.62. I used .60. I was able to use .62 on the big hole and then that website locked me out. The big hole came up to 217 cfm so I think if i was able to do .62 on the small hole it would be real close to 225
I can't seem to figure out what is making it inconsistent. I get the same readings within about a 5-10% range but that is not good enough to be useful. The only thing I can figure is that the vacuum itself is inconsistent. I checked the voltage to the vacuum and it isn't fluctuating more than .1 volt so it's not that.
I think the only solution is to use the orifice plate so that I have a direct knowledge of how many cfm the thing is flowing. I just made one at work today which I will hopefully be able to use in conjunction with the calculator on this website Flowrate Calculation for an Orifice Flowmeter fyi, there are charts showing air density vs temp. At room temp the air density is less than the number they have in the calculator.. iirc its about 1.18. I'll probably have to tweak the flow coefficient a little to get the numbers to line up with something of a known flow rate.
BTW i calculated theoretical flow through that calibration plate using the above calculator at about 216 cfm. The only variables I guessed at were the temperature (i used 25c which iirc is 77 degrees f) and the flow coefficient. From what I have read the flow coefficient can vary from .58-.62. I used .60. I was able to use .62 on the big hole and then that website locked me out. The big hole came up to 217 cfm so I think if i was able to do .62 on the small hole it would be real close to 225
J-B Racing
Forced Induction
- Joined
- Sep 16, 2003
- Messages
- 2,466
Wonder if you are seeing air density differences due to weather - temp, humidity, and barometric pressure? If you run the vacuum cleaner off of a variac so you could turn it up or down a little, could you put your calibration plate in and adjust the variac to set the vacuum for one orifice hole, and then check the other holes to see if everything is all okay now? (Or you could make a crude slide valve so you could slightly choke off the vacuum cleaner for normal operation so you could open or close the valve slightly to compensate for weather changes.)
Update:
I did some more research and decided to add an orifice plate to the "bench". I put a 90 deg elbow on the end of my bore adapter and added ~5' of 4" pipe to it, at the end of it I put a coupler and inside the coupler I added a plate with a 1.25" hole in the center. I added another section of pipe behind this - maybe 2 feet- and then the vacuum.
On each side of the plate, about two inches away, I added taps for a second manometer. The second manometer is pretty much just like the first. This second manometer allows me to correct for changes in my vacuum efficiency due to weather, voltage, etc.
I take the two manometer readings and divide the one at the orifice plate by the one at the head. I then find the square root of this number and multiply it by a factor determined by what my orifice plate should flow at 28" depression. Since I don't have a precision cut orifice plate I had to play around with the discharge coefficient of the plate in the flow equation.
I flowed a stock head and changed the coefficient to match what others have gotten out of a stock head. I settled on a coefficient that gave me 154 cfm at .5 lift on a stock head.
From there I tested my ported head. I was pretty dismayed to find that my ported head according to my equation was only flowing 164 cfm. I was able to repeat this number to within 1 cfm numerous times.
From here I tried all kinds of things in the port to decrease or increase flow. I was able to kill off flow pretty easily but surprisingly, adding bits of clay up in the port area didn't affect flow that much.
Trying to increase flow was a lot harder. Grinding all kinds of area in the port pretty much netted me nothing. I measured a 1 cfm gain. So I took a step back and started reconsidering everything I had done.
When I looked at the throat area I found that most gurus online suggest a throat area of about 90% of the valve diameter. This is significantly larger than I had.
Just so turns out that an untouched exhaust valve is just a hair under 90% so I used this as my gauge. I opened up the throat area a ton to get to 90% and then blended the short side radius and all the other edges back into the port.
This took awhile (I was being cautious)
Back on the bench, I could immediately tell there was a flow improvement just in the raw numbers. After doing the math I found that I am now up to 175 cfm.
That is still a lot less than what I expected. Still, 21 cfm more than stock isn't all that bad. Maybe my setup is reading low.. or are these heads that hard to get flow out of?
Gurus, does getting past 175 cfm require magic?
Also bear in mind that this is with the stock valve job and completely stock valves (no back cut). Can I expect another 10 cfm out of that?
I did some more research and decided to add an orifice plate to the "bench". I put a 90 deg elbow on the end of my bore adapter and added ~5' of 4" pipe to it, at the end of it I put a coupler and inside the coupler I added a plate with a 1.25" hole in the center. I added another section of pipe behind this - maybe 2 feet- and then the vacuum.
On each side of the plate, about two inches away, I added taps for a second manometer. The second manometer is pretty much just like the first. This second manometer allows me to correct for changes in my vacuum efficiency due to weather, voltage, etc.
I take the two manometer readings and divide the one at the orifice plate by the one at the head. I then find the square root of this number and multiply it by a factor determined by what my orifice plate should flow at 28" depression. Since I don't have a precision cut orifice plate I had to play around with the discharge coefficient of the plate in the flow equation.
I flowed a stock head and changed the coefficient to match what others have gotten out of a stock head. I settled on a coefficient that gave me 154 cfm at .5 lift on a stock head.
From there I tested my ported head. I was pretty dismayed to find that my ported head according to my equation was only flowing 164 cfm. I was able to repeat this number to within 1 cfm numerous times.
From here I tried all kinds of things in the port to decrease or increase flow. I was able to kill off flow pretty easily but surprisingly, adding bits of clay up in the port area didn't affect flow that much.
Trying to increase flow was a lot harder. Grinding all kinds of area in the port pretty much netted me nothing. I measured a 1 cfm gain. So I took a step back and started reconsidering everything I had done.
When I looked at the throat area I found that most gurus online suggest a throat area of about 90% of the valve diameter. This is significantly larger than I had.
Just so turns out that an untouched exhaust valve is just a hair under 90% so I used this as my gauge. I opened up the throat area a ton to get to 90% and then blended the short side radius and all the other edges back into the port.
This took awhile (I was being cautious)
Back on the bench, I could immediately tell there was a flow improvement just in the raw numbers. After doing the math I found that I am now up to 175 cfm.
That is still a lot less than what I expected. Still, 21 cfm more than stock isn't all that bad. Maybe my setup is reading low.. or are these heads that hard to get flow out of?
Gurus, does getting past 175 cfm require magic?
Also bear in mind that this is with the stock valve job and completely stock valves (no back cut). Can I expect another 10 cfm out of that?
One of the simplest restrictions you can remove is the valve stem. Instead of just trimming it you can shave it at an angle. If you plan to put inserts in you can adjust the height of the insert to cut down on the restriction. The bowl and throat just above it are the most critical to flow.
