RESEARCH & DEVELOPMENT  

Ambient Air Flow Testing

Most of this testing gives us data that helps us determine the best core specifications for each application -radiator, intercooler & oil cooler . As usual, it is a case of trying to balance all the factors that effect the cooling efficiency, because it's no good having a super cool intake charge if the intercooler fin mesh is that fine it's strangling the radiator for air, & the engine cannot be run at full power. Sometimes we close up the radiator finning if we know the customer is running a very efficient fan, as enough air can still flow through at low speed. We do the opposite if the vehicle has a 'cluttered' (usually 4x4's) front end, or a small air opening. Sometimes you have to loose to gain. A comprehensive knowledge of flow rates means that losses can be kept to a minimum.

The faster the air travels through a core, the higher the heat transference from the engine. Fans only help at very low vehicle speeds with an average of approx. 3 meters/ second, or 10.8 kph being common. Don't think that once this car has reached 11 kph, you can switch the fan off,  as there is a big difference in these readings with the air being pulled from behind, to the air hitting the radiator at the front & being able too "spill" away from the core. The amount of "spill" is determined by the ease in which the air can detour around the radiator. Trouble is that most cars can spend allot of time travelling slow. When stationary, the only cooling is the performance of the fan. Driving a performance car in  city traffic or a Rod at a show, places a big strain on the cooling system. Luckily, engines produce allot less heat at part throttle, but some 4x4's & competition vehicles operate under full throttle/ low speed conditions making air flow paramount, so it's impossible for an off the shelf radiator/fan combination to be right for every application.

6 fans tested in late '02

          The air speed meter we use.   Note the reference marks on the core so all readings are consistent.

            This stays in it's boxed packing & is treated very carefully 

Test Results

   This is a PDF file 4.4 kb.

Results of 8 fans tested in a real   world environment.

This is a copy of our spreadsheet of our own fan tests. Once again my disclaimer is that they were tested to our parameters, with our equipment for our knowledge & are totally independent. NOTE that the cfm.  figures are for air flow through the fan as we only measured inside the fan shroud case at the front of the core. If you want to compare them for actual cooling ability with the results from a car with a fan built into a shroud that encloses the whole core, you have too factor in the different surface areas.

They are 4 of the same Australian brand fan & 4 from an importer of fans, which must be noted that 2 were Chinese,  1  American & 1 un- branded, although they all had the same proprietary brand - (buyer beware). These cannot be directly compared with some of our other tests, as the conditions differ.  The most important aspect of consistency is the position of the meter in relation to the fan blade, as you can see in the charts, the distance in from the end of the blade gives very different flow rates. We mark the core with a grid pattern of a set size so the readings are repeatable. The next most important aspect to consistent results is the supply voltage, crank up the alternator output & the fan will pull noticeably more cfm.  Atmospheric conditions play a small part in the readings, so lately we have been recording them with a very accurate weather station, so we are as consistent as possible.

NOTE : the results of the test performed with the fan blade on the correct way - 1362 cfm & then incorrect - 883 cfm  - a loss of 65% flow !! A thermo fan comes as either a 'pusher' (usually) or 'puller' & if used in the opposite application, has to have the fan blade reversed. During our tests we broke one of the import fans because the blade was not removable, this was only on their 12" model.

These results can also be related too intercooler core thickness, or lots of products for that matter, even grille mesh. The thicker the core, the less air will pass through.

Our 'Sizing' Spreadsheets

      These are PDF files

This is our spreadsheet for air/air calculus.  29 lines of data 

This is our spreadsheet for air/water calculus.  61 lines of data. & 10 times harder to set up properly !!

Danny Irvine is a talented engineer who has (sometimes immeasurably) helped me with 'starting in the ball park' with many of our projects. He has given me the calculations & ideas that has sometimes short circuited heaps of wasted experimentation effort. These spreadsheets were developed by Danny for ARE & are proving an invaluable tool in providing us with a spreadsheet to be able to properly size an intercooler &/or radiator for any given application. We are still having to fine tune some of the formula (not conversion rates) as we build our own Data Base, but this is decreasing & becoming more accurate nearly everytime we plug a new set of figures in. It would not be possible without the expense of purchasing our test equipment, but to me it is well worth it as I know the product we sell is the best possible. We are getting a very clear picture of what to expect, & this is now visibly, accurately & unquestionably backed up by the Data Log results we are accumulating from street, track & dyno testing. 

Please note that some lines have been left out of both spreadsheets. We want to keep some of the data & calculus totally in house ! but we also want to let you see that we try extremely hard to have the best knowledge possible of what we sell.

 Data - Log  Testing

Over the years a few of our products have not performed as they should ( mostly radiators) &  many results of testing & prototypes have not given the answers expected, with some of these results being impossible to find a logical answer too. When dyno testing & in particular road testing, it is impossible to write down all the figures at a particular second of the test, so the results can be a little distorted, or worse, be recorded wrong. To assure our testing is as comprehensive, accurate, repeatable & easy to understand as possible, we have purchased a twelve channel (eight, seven or six for temperature, depending on whether none, one or two boost channels are used) FXL data logger off Graham                ph. no. (07)        who has been manufacturing these drag racing based loggers for 5 years. It is a matter of connecting the probes to the engine, running the leads out under the bonnet, in through the passenger window, going for a run on the dyno or road, recording the 15 seconds of information, downloading it onto the laptop, & I'm ready for another test, or back to work to analyse the results. I can leave them as line graph, or print them out on a spreadsheet, & have them for future reference or comparison. Apart from the obvious intercooler efficiencies, we will be testing the relationship of  engine output  to radiator water temps at different positions. The only radiators that have not performed up to our standards  have been three 57mm thick, large crossflow units, so I'm going to put all eight probes into one of these & see if my theory is right or not. It is also real graphic as too the operation of the intercooler - monstrous differences between out of the turbo compared to out of the intercooler. A real interesting area of testing that will open allot of peoples eyes will be under bonnet temperature testing & also the heat soak that can be picked up in pipework. I have seen over 80 deg. c. on a street car. With all eight probes in use I will be able to determine the "hot spots" in a cars engine bay.

 Notes 1)  This is a part of our own log & it is the first of many I have seen so many "claimed" figures for intercooler performances that not only are the people/occasional company & especially the magazines that print them - are plain dickheads. I have seen in car pictures of a hand held thermometer reading 4 - 10 deg. c. held beside a speedo reading 80 -110 kph. Sorry, but that means Jackshit. What was the ambient temp., was it under full throttle or coasting down? Intake temp. figures can make dyno figures look split atom accurate. Off the soap box.Water temps at different points in the cooling system are just as important to me, so it makes our investment worthwhile.

Too read our graphs ( I apologise for the blurred PDF) its easiest to note the position of the red curser line & the temp. no.1 to 8 figures underneath right are at that particular 1/50 th of a second. I will try too keep the same configuration all the time. In this file 146.8 

Note how both the charge air temps drop with the gear change & especially how quick. The charge out of the intercooler only varies just less than 50% than the pre 'cooler air.  This was an air water intercooler on our Skyline GTST running at 75% water pump speed, so it's telling me it needs full pump speed & a slightly larger intercooler.

                    xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 

Notes A) This is a prototype air/water intercooler we fabricated for Atlas Engineering's Forest mulching machine. It is powered by a Cummins         Straight 6 cyl. turbo diesel, rated @    kw.  We had never made anything like this application before, but with some guidance from Danny Irvine, our spreadsheets, air flow bench testing, experience, couple of guesses, fabricated a unit that we knew would meet the Cummins criteria. I wasn't prepared for just how much we exceeded it. As you can see by the results, our intercooler is extremely efficient, but the final result is let down by pour performing radiators - which we had no say in their selection.                                                        Have a look at the results of test 5 @ 14:45:40 (recorded in actual time) @ a flow rate of 126 l/m Radiator in - combined water from the i/c. into radiator ...........................................68.1 deg. c.  Radiator out - water temperature out of radiator, so it could only remove 3.1 deg!!...65.    deg. c.     R H. water out - temp. of the water out of the right side of i/c.  .................................74.   deg. c. Charge air (post cooler) - air temperature out of intercooler  .....................................64.7 deg. c.     L H. water out - water temp. out of the left side of the i/c. .......................................64.7 deg. c.       L H water in - water temp. going into i/c. note lost 0.6 deg in hose from rad. ............64.5 deg. c.   Charge air (pre cooler) - air temp. out of turbo .......................................................172.9 deg. c. Heat transfer to aftercooler water (KW) - the actual kilowatts of heat removed ........77.8 kw.       The intercooler to water efficiency is 99.8% !!! The intercooler to ambient efficiency is 70.8%. The radiator efficiency is 6.9% (good eh!).  As you can see, our intercooler removed 108.7 degrees (167%)  from the charge air, but the radiator could only remove 3.1 deg. c.(5%).  It was approx 750 x 750 x 75mm & had a 650mm variable speed & pitch fan, so don't underestimate the size & type of radiator needed. Our innovative design allows for one water outlet temp. to be 74 deg., one 64.7 deg. & the charge air 64.7 deg. (at this particular second). Imagine if this water was run through an 'ice box' ! Unfortunately Cummins specifications are inlet air temp. must be no higher than 30 deg. above ambient. The company is flying a radiator out from Italy to try & compliment our intercooler.   

 In test no.6 @ 15:04:40 @ a water flow rate of 55 l/m. the results are a little  different.                                                                                                                                      Radiator in - combined water from  the i/c. into radiator ..............................................65.1 deg. c.  Radiator out - water temperature out of radiator,  it  removed 12.4 deg!!....................52.7  deg. c.    R H. water out - temp. of the water out of the right side of i/c.  .................................106.3  deg. c. Charge air (post cooler) - air temperature out of intercooler  .......................................56.9 deg. c.   L H. water out - water temp. out of the left side of the i/c. ...........................................60.1 deg. c.   L H water in - water temp. going into i/c. note lost 0.6 deg in hose from rad. ................53.  deg. c.   Charge air (pre cooler) - air temp. out of turbo ...........................................................157.3  deg. c. Heat transfer to aftercooler water (KW) - the actual kilowatts of heat removed .............72.2  kw.   The intercooler to water efficiency is 96.3% ! The intercooler to ambient efficiency is 73.1%. The radiator efficiency is 27.5%.  The radiator could pull 12.4 deg. (23%) out of the water& the i/c. could  pull 100.4 deg.c (176%) with the water going in. Note the temp of the water out of the RH. side - 106 deg., a rise of  32 deg with a lower charge air temp. If our i/c. was undersize, it's efficiency would've really suffered at this low  flow rate.

Notes B) A whole bunch of figures are educational, but to me they don't have the impact of a graph. I often convert the data to graphs as it gives me a better understanding, quicker, & stays with me longer. Our data logger gives both on the one page - best of both worlds. For example, these snapshots are of a machine that operates nearly all the time at a constant throttle setting, with the load varying due to size of trees, but look how much the temp. out of the turbo varies. Note that at 14:55:00, the throttle was backed off.  Also note that the above data is from a hired mega dollar unit, our data logger has a couple more channels but only records for 15 secs., not 30 mins like this unit. 

                   The datalogger unit set up in the car ready for a test run.

                      Data logger leads connected  to the engine components.

  These are PDF files of  Data  Logs tests.

Snapshot of our own data log. See notes 1)

     xxxxxxxxxxxxx

Field test spreadsheet of a large Industrial air/ water intercooler.          7 kb file.      See notes  A ) on left

Line graph taken from above. 11 kb.file. See notes B) on left