| Things went well and uneventful at the flowbench today. The owner was excited to see the type of testing I was doing and after everything was said and done, the testing was a freebie. :) I know many of you out there understand the principles behind flowrates and the mechanics of flow properties, but for those of you who dont have a grasp on what the information provided here is telling you, I'm happy to explain it here: In order to produce flow of a fluid (the fluid being air in this case), you must have a pressure difference. That difference in pressure, in simple terms, would be analogous to blowing through a straw. You are exerting a higher pressure air into one end of the straw, which causes the air to flow through the straw to lower pressure (that being the atmosphere). The amount of pressure differential you produce directly relates to the quantity of air you will be able to move. The harder you blow, the more air you will move. So how much air does a straw flow? The answer to this simple question would be, an infinite amount, given that no pressures were defined and the straw's exact size and materials weren't defined. For all you know, the straw is infinitely large and our planet is just a spitball waiting for that pressure differential to come about and heave us into oblivion, where-ever that may be. Coming back though, when a flow measurement is made, it is quantified by a standard set of parameters, namely by the pressure at which the test is performed. The standard test pressure is 25 inches of water. With this pressure limitation in place, and considering a standard McDonalds straw, there is a finite volume of air you will be able to move... Obviously it wont be much, but this is why there is a standard test pressure. A flowbench is very simple: it generates a known amount of airflow, has a manometer (fancy word for a pressure gauge) to reflect the pressure that the test is performed at, and a flowscale showing percentage of flow. Say you set the flowbench to move 600CFM and flowtest a component, pulling 25 inches of water, and the flowscale reflects 50% flow. Simple math will tell you that you are flowing 300CFM (50% of 600CFM). The reason there is a standard pressure is simple: if you flowtest something at greater pressures, you will ultimately move more air. It is a simple factor of standardization, analogous to using an SAE standard correction for dynochart results. However, in some cases, it is not possible to create 25 inches of pressure differential; like that of flowtesting something that simply outflows the testing equipment. Fortunately for us, enginerds have taken the time to provide a conversion table that allows us to analyze flowrates at different pressures and still derive a true value. This was put to use when flowtesting the MAS simply because it will far outflow the testing apparatus. 
I have a text document with the table information that this graph was produced with - email me if you really want a copy of it. The method I used to generate the data was based on a three-point flowtest. I did not perform 64 different tests as that would have been unnecessary as well as impossible. The reason it is impossible is for the fact that the MAS unit can move such a large quantity of air at relatively low test pressures. You need at least 5psi of test pressure to calculate the CFM but the MAS is capable of flowing up to 280CFM at under 5 inches of pressure. With the machine set to a baseflow of 460CFM, we pulled exactly 5 inches at 61.5% flow for a total of 282.6CFM, and a measurement was made at 3.65 volts. This test was performed once again, but the machine was switched to a base of 610CFM flow and run at 5 inches, yielding a 0.46% for a total of 280.8CFM, for a value of 3.63 volts. This test was performed once again with the machine set to a base of 457CFM and we pulled 10 inches of pressure this time, providing a flow of 88%, yields 402.4CFM at a voltage of 4.08V. Now some of you may question how we can get correct flow information when we are using different pressures. Remember, there is a correction factor that is applied to the measurement when using pressures other than 25 inches. You might also question the use of the velocity stack and presume that it will make the MAS flow different amounts of air. While this is true, the velocity stack will improve flow, we are measuring exactly how much it is flowing and taking a reading on the MAS's voltage output. With the velocity stack removed, it wont flow quite as much at the same pressure, but we would also see a decrease in the MAS output voltage. I only needed one measurement to build the table as I can use the VQ table in the ECU to provide the rest of the information. But I went with two because we were able to (using two different flow pressures and rates), and because I wanted an additional data point to verify accuracy. In the ECU's code, there is a 2-d table that represents an arbitrary flow value to the MAS voltage. It is a lookup table that the ECU uses to convert the MAS voltage into an airflow value, which is used in 90% of the equations governing the ECU's operation. This table is called the "VQ Table". It contains 64 values and each value represents 0.08volts from the MAS. Unfortunately the values aren't in any unit - they are simply a spread of flow values over a 16-bit scale. However, there is a simple multiplication factor that can be applied to these values to convert them into a known unit. That is what this test today was all about. Verification: At 283CFM, the MAS voltage was 3.65V
In the VQ table, the airflow value that represents this voltage is in position ~45.5 (3.65 / 0.08). At position 45, the value is 20953 and at position 46 it is 22462. 22462-20953=1509; 1509/2=754.5; 754.5+20953=21708. 21708 / 283 = 76.7 <----- this is the divisor of our values in the VQ table to convert that arbitrary airflow value into a CFM value. This can be applied to all points in the VQ table now (which is why I said we only need one datapoint to create the graph). However, since we took two measurements at different flowrates, we can now compare our results to another known value to see if it fits (whatever differences are seen in the final numbers will be a result of variances in the testing equipment only). So, we look at our test performed at 402.4CFM with a voltage reading of 4.08 volts.
In the VQ table, the airflow value that represents this voltage is in position 51 (4.08 / 0.08). At position 51, the value is 31267. There is no need for interpolation here as this voltage is exacting to position 51 in the VQ table. 31267 / 408 = 76.6 <------- How close is this for accuracy? :) I averaged the two for a factor of 76.65 and applied this operand to the VQ table to generate the results in the graph you see above. Here is the testdata collected:
 Now that we have this data, we can do all sorts of nifty analyses of a vehicle's performance, namely volumetric efficiency of the system as a whole. We can't direclty calculate VE of the engine with this information, but we can review efficiency of the system as a whole simply by calculating the theoretical airflow based on engine RPM and boost pressure, and then compare that value to what the MAS is telling us is actually going into the motor. The end result will be a volumetric efficiency value for the entire powerplant from the filter to the exhaust tip. So, expect to see more of this sort of information coming together in short time. :) I will be posting the flowrate of the filter and the intercoolers in seperate posts so that they can be easily searched for in the future.
[ ashspecz.com ]
[ agpowers@bellsouth.net ] Enthusiasts soon understand each other. --W. Irving.
Are you an enthusiast? If you are out to describe the truth, leave elegance to the tailor.
Albert Einstein
|
FLOWBENCH Extravaganza: MAS Voltage to CFM Chart - 
