How can I accurately measure the mass air flow in my intercooler heat exchanger?

AI Thread Summary
To accurately measure the mass air flow in an intercooler heat exchanger, various methods are discussed, including the use of mass air flow (MAF) sensors and Pitot tubes. The effectiveness of MAF sensors can vary based on their design, with some providing easier data interpretation than others. The project involves using a hair dryer for airflow, but concerns are raised about whether it can generate sufficient heat and flow to simulate real engine conditions. Calibration of MAF sensors is suggested, potentially using compressed air to establish a baseline for readings. The discussion emphasizes the importance of understanding the limitations of the equipment and the need for accurate temperature and flow measurements to ensure reliable results.
LeeK9
Messages
11
Reaction score
4
Hey guys, long time reader first time poster.

I'm undertaking a project in partial fulfillment of my Mechanical Engineering degree with regards to the cooling of an intercooler heat exchanger with an exterior water spray. basically i have an intercooler and i am using a hair dryer to supply a flow rate and heat input through the inside of the exchanger.

One of the things I'm really stuck on is how to measure the flow rate of air internally through the exchanger. I could probably find out from the data sheets of the hair dryer but i don't think these would not representative of the real flow rate as the downstream effects will reduce the flow rate.

i have been looking at mass air flow meters but can't seem to find any useful information as to how to correlate their voltage readings with a flow rate. I am aware there is graphs for some of these online but id need an accurate reading and any mass air flow meters i have seen for automotive applications do not come with a graph, or do they?

ive looked into using the pressure difference from a venturi setup but not sure which way is best to go with or if there are any other methods i am overlooking?
 
Engineering news on Phys.org
Could you fabricate something to calibrate your MAF? If you had an engine using a known quantity of air, that would be one way.

I think a lot of the readings will be dependent of the type of MAF sensor you have.. some have a moving cone or vane connected to a potentiometer.. I think they add a considerable restriction, but they're probably also quite easy to read and interpolate data from, and with attention in design, can be linear.
Others use a pair of wires, one with a minuscule current (just to measure resistance), and an identical wire with a significant, regulated current through it with the objective of keeping the wire at a certain temperature above ambient.. the more current through, or voltage across this wire, the faster the air must be moving.
There are "grid" type MAF sensors too, but am not quite familiar with how they work..

I don't know the size of the intercooler you're working with, but bear in mind that it doesn't take a large turbo (I'm guessing that's what is going to be generating the hot air in the 'real world') to produce many KW of heat... a pressure ratio of 3:1 can easily yield exducer temperatures of 200C and higher.. I don't know if a hair drier will be generating enough heat and flow to see the benefit of intercooler spraying. Perhaps a heat gun would be capable.. less flow, but much higher temperatures, and a more forgiving overtemperature safety.

I'll let someone else chime in... Welcome to PF, and keep me posted.. I enjoy this.
 
would air flow in miles per hour (MPH) be satisfactory for your readings?
 
I agree that you will probably need more hot air flow if this is a real engine intercooler that you are working on .

A standard experimental method of determining air velocity is to use Pitot tubes facing into the flow .

Ideally have several at different radial locations across the exhaust plane or less accurately have one central one .
 
Rx7man said:
Could you fabricate something to calibrate your MAF? If you had an engine using a known quantity of air, that would be one way.

I think a lot of the readings will be dependent of the type of MAF sensor you have.. some have a moving cone or vane connected to a potentiometer.. I think they add a considerable restriction, but they're probably also quite easy to read and interpolate data from, and with attention in design, can be linear.
Others use a pair of wires, one with a minuscule current (just to measure resistance), and an identical wire with a significant, regulated current through it with the objective of keeping the wire at a certain temperature above ambient.. the more current through, or voltage across this wire, the faster the air must be moving.
There are "grid" type MAF sensors too, but am not quite familiar with how they work..

I don't know the size of the intercooler you're working with, but bear in mind that it doesn't take a large turbo (I'm guessing that's what is going to be generating the hot air in the 'real world') to produce many KW of heat... a pressure ratio of 3:1 can easily yield exducer temperatures of 200C and higher.. I don't know if a hair drier will be generating enough heat and flow to see the benefit of intercooler spraying. Perhaps a heat gun would be capable.. less flow, but much higher temperatures, and a more forgiving overtemperature safety.

I'll let someone else chime in... Welcome to PF, and keep me posted.. I enjoy this.

thanks for the reply, from your username i assume that you like the odd wankel every now and again. As my second year project i designed, machined and assembled a model rotary engine for demonstrating transformation of motion, check out LYIT Wankel engine on youtube, i have a project report available as well as a full thread on a automotive based forum that i could link you to.

In regards to the MAF, a student from the previous year successfully gathered data of flow to produce a graph for the given hot wire MAF, so i need to investigate whether the results are satisfactory enough to use them as a base for my readings. I'm not sure how this was done but compressed air was used through a valve from what i gathered. This needs further investigation on my part but any ideas of how to do this would be welcomed.

From the research carried out I've seen inlet temperature of the intercooler in common applications to be in the range of 120-150C for standard boost levels not exceeding 2Bar ABS pressure( 1 bar gauge ). Typical heat values from a hair dryer device seem to be in around the 75-80C so a little over half the temperature, however the device i will be using does have variable heat control so doing several experiments at different heat levels i could possibly make predictions as to the results at these higher temperatures? But health and safety would dictate that i would not be allowed to run such high temperatures and furthermore my apparatus would need to be designed with these high temperatures in mind.

Ranger Mike said:
would air flow in miles per hour (MPH) be satisfactory for your readings?

It may be satisfactory as i have a known cross sectional area at the outlet but i feel it may lose accuracy. assuming your talking about a wind anemometer.

Nidum said:
I agree that you will probably need more hot air flow if this is a real engine intercooler that you are working on .

A standard experimental method of determining air velocity is to use Pitot tubes facing into the flow .

Ideally have several at different radial locations across the exhaust plane or less accurately have one central one .

Yes correct about the air flow as seen in real engines being quite a bit higher, similarly to a range of heats available there is also a range of fan speeds from the device so possible predictions could be made in relation to real world flows from my data.

if my bench top apparatus yields an improvement i am not against testing my theory out of my own car as it has a rather large intercooler and two turbos running at 0.6 bar gauge. i would be simply adapting my sensor circuitry to retrofit it to my car ( R32 GTR skyline )
 
Yes, I've had my share of rotary engines :)... about 15 of them, the last one was my dream car, an 3rd gen FD Rx7.. much fun... I put a whole ton of money into it back in the days where I had some to spare, full bridge port, 400 RWHP @ 11 PSI (.8bar).. I sold it to my friend, he now has over 500rwhp. I also worked at a rotary engine specialty shop on 3 rotor drift, drag, and road race cars... it was a lot of fun.

Circumstances changed for me, I moved back to the farm, so I got a truck, a '94 Dodge cummins diesel, so I went from a screaming demon to stump pulling torque.. The project I'm involved in is "Lil' black box", which is an aftermarket Holset VGT turbo controller powered by an Arduino.. it's pretty cool.. I now have 1.5 bar @ 1500 RPM, and 3 bar at 2500, and at that pressure it uses about 60kg of air/minute and the compressor temp is about 200C.. I'm still working on getting a thermocouple amplifier to more accurately measure my temperatures, since I want know where my optimum boost points are.
I'll try and remember to look up your work when I'm on a youtube capable internet connection.. There was just a thread on here about how to model the epitrochoid and rotorI don't know what sort of tools you have available to you, so it's hard for me to think of ways to improve accuracy.. Pitot tubes would certainly be one way, but again at low flow rates they may not be much use.
 
Thread 'Turbocharging carbureted petrol 2 stroke engines'
Hi everyone, online I ve seen some images about 2 stroke carbureted turbo (motorcycle derivation engine). Now.. In the past in this forum some members spoke about turbocharging 2 stroke but not in sufficient detail. The intake and the exhaust are open at the same time and there are no valves like a 4 stroke. But if you search online you can find carbureted 2stroke turbo sled or the Am6 turbo. The question is: Is really possible turbocharge a 2 stroke carburated(NOT EFI)petrol engine and...
I need some assistance with calculating hp requirements for moving a load. - The 4000lb load is resting on ball bearing rails so friction is effectively zero and will be covered by my added power contingencies. Load: 4000lbs Distance to travel: 10 meters. Time to Travel: 7.5 seconds Need to accelerate the load from a stop to a nominal speed then decelerate coming to a stop. My power delivery method will be a gearmotor driving a gear rack. - I suspect the pinion gear to be about 3-4in in...
Thread 'Calculate minimum RPM to self-balance a CMG on two legs'
Here is a photo of a rough drawing of my apparatus that I have built many times and works. I would like to have a formula to give me the RPM necessary for the gyroscope to balance itself on the two legs (screws). I asked Claude to give me a formula and it gave me the following: Let me calculate the required RPM foreffective stabilization. I'll use the principles of gyroscopicprecession and the moment of inertia. First, let's calculate the keyparameters: 1. Moment of inertia of...
Back
Top