Gas Flow through VERY small pipe

[downybear]

HI,
This site is wonderfull - great work - Congratulations!!!
My problem - I am an electronic engineer, not "pipe-engineer", so I have had some additional work to be closer pipe-problems.
After few weeks I hope that I can ask no stupid questions here .

So, currently I am trying to calculate LPG flow through the injector, which is in fact the valve, not "pump".
Injector and something like Venture nozzlee is connected between car's intake pipe and pressure reducer installed in car.
Pressure reductor has almost constant pressure on their output and is connected to injector valve input, but intake underpressure changes from 20 kPa to athmospheric when throttle is fully open.

I have to calculate current LPG flow using current value(s) of differential pressure(s), because while fast closing throttle on high RPM, pressure difference is very big and LPG load is too big.

I have read many times formulas here to implement it in my computer, but there are limitations and notices about diameters.

My typicals diameters are 8-12 mm (2*r) on the reducer output and 1,5 - 4 after nozzle (Venturi type, but not good calculated I think).

My question is - what I should change in formulas to use it for so small diameters.
Available data are:
D-input [mm],
D-output [mm],
pressure (kPa) on input,
(under)pressure on output,
LPG temperature.

Thank you in advance for any help
Regards
Mariusz

[admin]

Some picture will tell more ...

Regarding the pipe sizes,... All calculations in "basic" fluid dynamics are based on the assumption that fluid is continuous, which is actually not true, but if the diameter is big enough than you are OK. Big enough means that diameter is bigger than the free path that one molecule of fluid travels before it bumps to another one and in 1.5 mm there are many of them.

[downybear]

WOW,
very good news for wake up for me.
Now I have rememebered this condition from high school...

But I have some questions more:
Here
http://www.pipeflowcalculations.com/pip ... ifices.htm
are equations with L1, and L2.
Conditions for this equations are away from my situation.
How change this equation or which set of L1 & L 2 should be choosen.
Regards
Mariusz

[admin]

Can you draw some image about your injector?

[downybear]

Hi,
I have prepared two pictures attached.

There is no details about particular valve because I have to do something universal as good as possible for many injectors on the market.

Remember that I am not "pipe-engineer" , and please answer if my thinking is ok.
So,
I understand problems with unregular shapes on the fluid or gas way.
This is very difficult to analyse, but the fundamental question is - if we can evalute problem to:

Input -> tank -> output.

"Input" - 8 mm input from reducer, relative long pipe (about 1 m) with diameter ca 1,5 times bigger than input nozzle).

"Tank" - this is a injector rail and valves where diameters and gas-transfer possibilities (kg/s) are much bigger than through output nozzle.
"Something" in our bing tank provides opening output way for LPG.
"Output" - small Nozzle (1,5 mm) and after nozzle pipe ca 4 mm diameter and length 10 cm or more (should be as short as possible).

Does we can ignore gas flow through tank because obstacles there are smaller than on output nozzle?

Additionally I would like to inform you that accuracy not must be very good.
I hope that about 3-5 percent will be very ok.
Why?
Depends of seasons LPG mixture may be vary - from 40/60 to 60/40 Propane/Butane and ro [kg/m^3] and R aren't constant as well.

I am planning to do some data evaluations for vary parameters as mentioned above, but at first I need a trusted formula .

I hope that I have explained everything clearly - english isn't my native language .

Thank you in advance
-------------
Hmm, I have a problem with uploading PDF's here, TXT extension is not alloved as well
I will send PDF's on mail which is provided while notyfing about posts.

Regards
Mariusz

[admin]

To be able to upload you have to put it in .jpeg or similar.
Here are images with my scheme for the problem.

[admin]

Can you tell me if the scheme on the top is good representation of the problem?

[downybear]

Hi,
Yes, your image shows problem good.
Only difference is that between input nozzle and output nozzle we have "ideal valve" without obstacles comparable with output nozzle.

Regards
Mariusz

I have answered fast, but later I have discovered, that your very good model may be not usefull because of no data available.
Therefore I have prepared new image which shows available pressure and underpressure data.
These values are available from two places shown on image.
On the opposite site to input of the injectors (matrix and Valtek) there is a place where pressure sensor is applied.
And this is everything what we know about pressures in the system.

Additional infos:
Reducer is connected to intake mainfold as well, and simple spring-pneumatic feedback loop in theory should stabilize pressure "visible/available" over the output nozzle.
Parameters of the input pipe from reducer to injectors have to be enough big to do this condition true.
So - I think that everyting what me may find is the flow trough output nozzle from a "tank" with stabilized pressure...

And now comes next question:
What is the "bigger" diameter? - Is it 8 mm of the input nozzle or something others?
or
Another formulas have to been used when we know only diameter of the output nozzle?

Am I right?

Regards
Mariusz

[admin]

Here is the model that might give us results - it is the outlet nozzle:

image - injector 2.jpg (9.41 KiB) Viewed 12441 times

Now, pressure at pos. 1 is the pressure measured by "pressure sensor" and pressure at pos. 2 is the pressure on the "pressure sensor intake" - right?

Diameter at pos. 1 and 2. are needed to do the calculation.

With these parameters we can get flow rate for different pressures.

Can you now give actual: p1 - absolute, p2 - absolute, D1 and D2?

[downybear]

Hi,

Now, pressure at pos. 1 is the pressure measured by "pressure sensor" and pressure at pos. 2 is the pressure on the "pressure sensor intake" - right?

MD: Yes, of course

Diameter at pos. 1 and 2. are needed to do the calculation.

MD: In nozzle which I have in my hand right now D1 is 4 mm, D2 is 2 mm. Output to intake is 4 mm as well.
Input looks like on your image.

With these parameters we can get flow rate for different pressures.

Can you now give actual: p1 - absolute, p2 - absolute, D1 and D2?[/quote]

MD: Absolute pressure p1 is about 100kPa more atmospheric (ca 2 at) - screw on reducer controlled this level, but may be vary during thousands of km.
Absolute pressure in the intake mainfold changes from 25kPa (closed throttle) to atmospheric (ideal - full open throttle without air filter and pipes obstacles) in engine without Turbo or
to about 200kPa over atmospheric while Turbo is installed.
Reducer's regulation loop should stabilize pressure difference between p1-p2 to value about 100kPa (depends of reducer type/model and HP of the engine).
This is theory when everything is new and good tuned - therefore I would like to do a step more in my LPG Electronic Control Unit, comparing to market solutions.

You are great .
While contacting with you my oblique is to think better and more about this new for me engineering area .

Finally, please let me know which kind of formulas I have to use, to be as close as possible to solution.

I have attached sample ignition map - I am feeling that it will be interesting for you .

Green grid - this is underpressure: -100kp (left) to 0 kPa(atmospheric - right side, till red grid), or 0 to 100kPa absolute.
This small engine (1100 cc, Fiat Panda) has a maximum underpresure about -75kPA (25 kPA absolute).
For example - bigger engine in toyota Verso works in the range -50kPA - 0 kPa (50 to 100 absolute).
RPM is clear - revolution per minute, Pulse time is injection time measured on gasoline injectors.
Regards
Mariusz Dec