Gordons Pitot - Very experimental bench so far...

[GordonE]

[color=#000000]Hi!
I started to think about building a bench in the autumn of 2008 when I had a valve seat that let go of my cylinder head.
I wasnt satiisfied with just replacing the old head with a similar one. So I decided to build a flow bench so that I can see how much differance my porting does.

I thougt that electric mesurements would be nice, so i decided to try to build some electrical device for capturing the pressures.

I live in Sweden and there are some good forums and homepages here that discuss porting and some with pictures of benches that people built.
So after a couple of months of thinking about it and trying to understand the different principals of orfice and pitot I decided the obvious, to build a simple pitot bench!
It seemed simpler and more effective than a orfice bench, and I liked the fact that I didn

[49-1183904562]

Gordon,

What type of smoothing algoritham are you using in your code?

Aslo you may wnat to look at power supply stability in your circuit.

Just some thoughts.

Rick

[GordonE]

Hi Rick!

I use the power supply built in the computer. I took ground +12v and -12v from the power supply to drive the trancducer and amplifier card. I have mesured the 12v signal and its ripple free. And I guess it should be quiet stable.

The only smoothing I use is software based in the program that comes with the Labjack, its called DAQ factory.
Its a averaging function that is adjustable.

I dont have the skills to write programs, so I just use the Daq factory and apply formulas that I calculate with my Texas TI-83 calculator. Its not as good looking as your program, but I can change it as I want right now.

Do you now if adjustable resistors can change from day to day? When I tested yesterday the idle voltage had drifted of a little... Annoying!

Gordon

[GordonE]

I have dedeicated som more time on the electronics this weekend.
I have had some help with some extra filtering that didnt do very much. Now I have capacitors on the supply from the computers. We also added some more capacitors on the amplifier card. Better safe than sorry...

Before I had the trancducers and cards just laying on a shelf, but i noticed that the offset voltage drifted a bit when I moved them. So now they are mounted in a little box. Now I felt that I could have concictency.
I also replaced the variable resistors with a better type that I have to turn many turns to change the value.

I also have made a better reducer from 110-75mm. I glued a grinded down 3" Aeroport that usually is for base speaker use in the 110mm pipe.

I also pop-riveted the plastic pipes to the steel mesuring pipe. Before I just glued them, but stress was to high for that, so they werent 100% in place.

I also have redone the static pressure fitting in the pitot tube. Now its just a 1mm hole drilled through the wall.

My friend helpt me with machining of the first cylinder adapter that I shall use for the upcoming flow test of my mercedes head.

I also tested a home made averaging pitot versus the original middle mounted pitot.
The conclusion that I finally made was that the standard pitot was a little better for me.
I mesured/calibrated the gauges for the three plates that Ive got, the 50cfm, 150cfm and 300cfm.

8-hole averaging pitot:
50cfm plate=50CFM 150cfm plate=150cfm 300cfm plate= 321cfm

1-hole pitot:
50cfm plate=50CFM 150cfm plate=150cfm 300cfm plate= 315cfm

This is with one mesuring element, the 60.4mm Id 2.5" exhaust pipe.

I wonder if I shall be satisfied with this accuracy?
Is it possible to have better results with just one measuring pipe?
I am glad that I have come this far, but im not 100% happy yet!

[GordonE]

Soon its time to test a head or two. Sadly I havent found a solution yet that can rectify the reading that is off with the 300 plate.
The sensor is still drifting away some, but it is easy to reset the offset with the variable resistor.

Does anyone have a good solution for having mor linear readings? any ideas is appriciated!
I think that the lower readings doesnt follow the root function to 100% and therefore the 300 value drifts off.

[GordonE]

I found a plate calculator on this page

This shows different values of the plates i have. I will try this to see if I can make it more linear.

The new flows are(with 103mm pipe underneath)
50 plate:46.6; 150 plate: 141,5; 300 plate: 292.3 CFM

Does anybody hace a better calculator for plates? My plates are made out of 3mm steel plate, the calculator i mensioned dosent have this as an input.
The plates are simply machined as a straight hole, two sharp edges.

Regards Gordon

[GordonE]

Now I have made may first flow test of a standard Volvo 531 head.
It seems that I didnt handled the calculator I talked about erlier good. A second try gave me other, more apropriate numbers.
I flowed the head before that, so the numbers are off around 5% to the lower side.

I have tried to mesure the three plates and then doing a linear curve of the readings from the three plates. I am now going to use only the biggest plate, the lowest readings screw up the linear function so that the linear graph dont pass through zero (x,y)(0,0). This leads to that the no flow reading is around 7cfm or negative. I am going tu use just 0,0 and the voltage that the biggest plate gives from now on. That should give me a better reading around zero flow.

I have struggeled many hours with the three plates, and I always got a high reading on the biggest plate. I now have reached the understanding of that the bore of the connecting pipe affects big plates, and makes the flow higher the closer to the bore diameter the plate gets.
The calculator is good when trying to figure out how this works.

I have also made a excel list of the three plates, compensating for the air temperature(10-25C degrees). Im not sure that I should take this in account, but it makes a rather big differance on the big plate...

I am now closer to my goal of linearity. I am satisfied if I reach 3%, I think that is enogh, Its a rather big turndown ratio to mesure.

This Is what I flowed, A Volvo 531(swan neck as we call it in Sweden) head. The retainer bottomed out in the valve stem seal at 12mm lift.
The readings Is rather low, around 5% lower than other tests that i have seen.
The test is just made on the intake, I havnt tried to reverse the direction of flow in the bench yet. I am glad that my girlfriend havnt kicked me out yet. The weekend hours dissapears too quickly.

For me the functionality is of the highest importance. I dont have time to build a proffessional looking cabinet right now. Im satisfied that I have come so far with so little money. I also have lernt a lot.

I am soon going to flow test a Merc 2.3-16 standard head. It will be interesting to compare this to other 16v heads(volvo and so on).
I have just once seen a flow graph on a mildly ported 16v merc head and it flowed over 300cfm at 10mm lift. I want to see this trough my own eyes, Its hard to now other peoples accuracy.

Here is some pics:











Regards Gordon

[bruce]

A few quick questions;

Where are you taking your static pressure from?

The plenum area under your test part is rather small, you need a large area for the air to settle in to get an accurate static pressure. If the air is moving you are not going to be getting a very stable static pressure reference.

Calibrate your bench without the head adapter and see how your plates flow.

The main thing to remember it's your bench and you are the one who has to be happy with the level of repeatability and accuracy you strive for. No bench is ever 100% unless you spend some serious $$$'s to achieve it. As you said you've come a long way in a short time and probably don't have a lot of $$$'s spent and that is the goal!

WELL DONE!!!!



Edited By bruce on 1234617186

[GordonE]

Hi Bruce!
Thanks for your comments!

As you suspect, I measure the pressure in the pipe wall around 20cm from the adapter top. This may notbe optimal, and the static pressure goes up and down somewhere around +- 0.3. Would I get a more stable reading if I built a 6inch pipe underneath? I got the Idea to take it directly in the pipe from flow performance benches that I have seen!

Isnt there a smarter way to build a static pickup? Someone mensioned multiple pickup points? Cant I just glue a 10cm long 2" pipe in the 110mm pipe 90 degrees from the flow direction and take a mesurement in the end of that? Maybe the flow/turblence will affect this reading...

Sadly I had to work some hours this morning, so I just had around two left to spend before it was time for Valentines day major sucking up
Anyhow I recalibrated with the (0,0) and the reading for the big plate in conjunction with my excel table that compensates for temp. When I used the big plate (302CFM) I managed to get the 152cfm plate to read 151. The little 48 plate flowed 46.5.
Ill get back with the exakt numbers. My numerical memory isnt what it should be... Maybe its for my own good
I will also try to have some kind of a log, so I can write down the different readings between days. I use barometric and temperature correction in my flow calculation formula.

The Volvo head now is up around 3% all over.

Now I am getting somwhere!
I am beginning to feel satisfied now that it works out as it should. Next step is trying to flow reversee direction.

I still use my diy averaging pitot. It seems to work rather ok!

Bruce, is it necessary to have a compensation for temperature when flowing the test plates? It makes sense for me, colder air is heavier and should have more resistance through the orifice. Does this make any sense to you?

[GordonE]

Here are the work i did today, I flowed the intake on a mildly ported Mercedes 16v head against a standard head.
It is really good values for a stock 16v head!
It will be interesting to start grinding some soon!





Here is the Volvo 531 head with a better picture!



Regards Gordon

[bruce]

[GordonE]

I will try to make a plenum underneath the test piece later on.

Today I was at the scrap jard, I was looking for thermostat tubing. I also found two suitable tubes, I mesured one of them and the OD was 1.9mm. The hole inside seems to be around 0.2mm. The tubes happend to find their way to my pocket instead of beeing chewed down in metal recycling plant.
I also am trying to get a hold of 1m of 4x0.5mm steel hydrualic tubing. Probably I will have that next week.

I am going to try to do my own pitot probes for port velocity mesurement.

I also redid my valve opening fixture tonight. I wasnt satisfied with the old one, as it was a quick solution that opened one valve more than the other. Now the opening bolt is between the valves. I lost som low lift flow beacause of this, wich is better in a accuracy perspective!

[GordonE]

Hi!
I have reworked my box for the measuring board. The wiring was a disaster before, I have shorted the lines as much as possible to reduse noise. But the noise was quiet good before. I saw no difference at all. The AQ is 10 bit, so it can only read down to 4mV in resolution.



This installation give me more accuracy and there is less risk of that the readings should change.
I have moved a varistor that adjusts the pitot board offset, so that I can access it without opening the box.
The DAQ is glued inside the box, so there is no longer any chance that the wires will bend and alter the readings. I am much more satisfied with this than before.

The next step is to get push-in connectors for the pressure hoses!

I also took the time to spray paint the measuring pipe. The last thing I did was to change the depression control from restriction of the flow to bleeding the flow instead. This must be much better for the motors and the temperature when its time for blowing backward when I try the exhaust testing!

[GordonE]

I have to build a setteling chamber soon. I will try this to see if i can get a little more steady depression!

I am also a little unsatisfied with the flow capability of the bench. I have six used vacuum motors and they seem to only go to around 350CFM at 28in.
I have tried them out by running the bench flat out at 28in, then I tried unplugging one motor at a time so that I could see how much depression I lost for each motor. For every motor the depression was lowered to around 20-22in.
They all lost about the same depression, then they must be ok!
I thought that it at least should be sufficient for 450CFM. It looks that I must install two or three motors more! Crap...

How much do you other pitoters get from your motors?

[FPV_GTp]

GordonE nice work

cheers