Dyno Testing on Fans, Alternators, and Oil

Dyno testing on fans, alternators and oil…..all here!
– – Back – –

  Home   >   Ramblin   >   Dyno Testing Fans

by Steve Jack
Concept One Pulleys
Jackstands at Old Cars OnlyWell guys…the dyno results are baked and ready to be served up. It has been a really interesting project to look at and took more time than I had anticipated…I still have a regular job to keep two kids in college you know!!

The idea just came up when a group of us started talking about the results of a fan test that ran a couple of months ago in one of the industry periodicals. Coincidentally, I had a 350/350HP that was sitting just ready to be broken in and tuned on the dyno (it pays to have drag racing buddies with dynos…a local pro-stocker to boot) ….and the owner said that I could borrow it for a week or so…..just stay out of his way in the shop (they know me too well there…dang!).

The set-up is easy and we broke the engine in at 2300RPMs for 30 minutes and let cool for a day. Then came back to tweek the carb (a jet change and rods as well on the Edelbrock 750)to peak specs. I ran the engine at differing RPMs for 4 hours and let cool. The sucker put out an expected 358.6HP and 385 ft-lbs of torque. I made ten pulls and all the numbers where within a half percentage point so I figured everything was pretty consistent and stuck together!

The goals of the other testing was to fool around a bit with some salient issues that perplex many a rodder and once an for all, see for myself what the heck is happening. It’s the engineer in me that forces my curiousity to amble on I quess. I set out to do fan/hp loss testing and cfm measurements (really a secondary project in and of itself), alternator load losses, and a last minute thought of seeing if synthetic oil would produce any hp in my engine.

The set-up for the fan/hp loss thing was pretty simple. I used a stock plastic shoud (for a camaro radiator) and built it on a screen box (a frame from which a fine mesh metal screen is stretched to simulate some air resistance like that through a radiator) and used two aircraft pitot tubes (opposed in the shroud) and an aircraft airspeed indicator from both to give me a starting point for calculating CFMs…..it worked beautifully! I actually calibrated the radiator/box simulator, as we called it, with an known good source of CFMs to see how close our calculations really were with two differing electric fans. The calculations came in at around plus or minus 5%, and we can’t figure if the simulator is off or the calibration has error, but the important thing is that we established a yardstick for CFMS that was consitent. We already had the dyno for measuring the hp and torque thing. Another feature of the radiator simulator “rads” for short, was that the frame was built so that we could adjust the opening of the rads in three plains! This was a time saving feature and made it easy to fit each fan.

All of the fans were directly mounted on the 350. The 350 was fed from a separate water source at a constant pressure. The water pump was a Howard Stewart Stage II (an excellent pump albeit pricey…I guess you get what you pay for). The shroud actually measures 18 inches and is perfect for 18 inch, 17/16 inch (used flex moulding on the opening to bring the edges in for a good fit…hard to explain but worked). Each fan was tweeked to its maximum reading moving the rads in/out to give a good base start. Once the fan was calibrated at base the rads was not moved again. It just happens that we tried to move the rads at higher rpms and the base setting was usually the best performer……and so represents the real and fixed world.

The fans in testing were the Chevy repo big-block fan (956N from Yearwood), seven blades and pitched with taper ends, the same fan with a GM repo thermo-clutch (9189N from Yearwood), the Aspen/Volare six cyl fan that Skip etal likes (and is darn difficult to find….I had alot of mosquito bites after that trip!)and I lovingly refer to as “Skipper”, and last but not least the popular Flex-o-lite stainless 17 incher (from NAPA #823-1240). To get the thermo-clutch to work we heated the clutch with heat guns to replicate a hot situation. It really worked pretty good.

The testing was to determine the CFMs at idle (800rpms) and at a slow engine speed (1600 rpms) and hp losses at 4500rpms as well as, and this is an interesting catagory….engine spin-up time from idle to full throttle at 4500rpms (a test that measures the ability of the engine to spin itself up to the given rpm and is directly proportional to the stuff you have hanging out front….this would tell give us a torque expenditure that is dedicated to doing nothing but getting the fan mass up to speed!). A figure of 100% was used for the engine with no fan or accessories involved. We checked all the fans for balance and corrected where necessary! The figures below for this measurement reflect an additional time taken per apparatus!…… And awaaaaaaaaaaaaaaay we go!


FAN TYPE      IDLE (800)     IDLE (1600)    HP LOSS (4500)   TIME LOSS %

 BB18          1680           2355            40.8              8%

 BB18C(heat)   1580           2205            35.6             13%

 BB18C(COOL)   1104           1400            19.3             11%

 SKIPPER       1503           2020            31.5              7%

 FLEX 17       1420           1814            20.2              5%


Having presented this here, I want to say one thing. It isn’t a perfect scientific test…but I think it represents a close proximity to real world.The conclusion I reach here is that the flex fan is a poor performer and certainly many of you have raised safety issues with this design as well that should be taken seriously. A better choice is a Skipper from the stand point of idle air flow but gives up hp on the other end (if you are interested in that…it may be a moot point), and the fact that it is alot cheaper if you can find one. The Skipper is a fine performer at idle and moves alot of air for its size, but ours had to be rebalanced whereas it was out of shape some how…but you couldn’t tell it by looking at it….must have been hit or something. The real air mover and hp eater here (and there is no surprise to me) is the BB18. It performed the best and of course has the fattest, longest and highest pitched blades. When coupled with the themo-clutch it gives up some CFMs depending on temperature etal, but doesn’t eat up hp on the other end as if fixed. Notice the torque percentage difference with the clutch attached. This is due to the work it takes to get the mass of the thermo-clutch up to speed. One other note here….I played with one fan at several differing RPMs and found that there is a performance envelope that the fan is designed for which will give linear results until the fan starts to have losses due to cavitation and or non-productive eddy losses at the ends of the blade. Some losses are at the center axis as well, but the engineers from flex-o-lite predicted this phenomenom before the testing started. Just remember….Being that we didn’t have access to the design specs for these fans, the performance envelopes will vary with application.

The best choice here may be to combine the Skipper with a clutch arrangement to get the best of both worlds…a high mover and diminished hp losses on the other end.

One other note….the themo-clutch was difficult to get to perform consistently with what we set up. We would heat it up and everything was fine then all of a sudden it would lock up under high rpms. We never could get it to stop all the way….even after putting it in the refrig over night. The results posted here took the most time to get the consistent 4 pulls and all of the results were averaged anywho! I would suggest that it would work more consistently in a car with a larger heat source etal.

Of course if its pure air movement that you want the BB18 is the choice for the mechanical fan but is a beheamoth massive structure.

Enough about fans and let’s move on to the alternator test. I simulated a test with a Summit one-wire, 100 ampere alternator and tested it at several RPMS. I loaded the alternator down with what I thought a fair representation of typical application of 65 DC amperes output at 13.65VDC. At 4500RPMs the alternator averaged over the 4 tests a measily 1.28 hp loss. It was really consistent with the high being 1.29 and the low 1.26 over the testing pulls.

Last but not least…..is the oil change out. I did a series of pulls after all the testing that we had done and got a reading of 360.4 hp, a slight and expected increase due to friction loss improvements and ring sealing (compression improvements)! I changed to Valvoline Synthetic 10W30 and did it all over again. The test revealed a 363.5, a slight improvement. But there is one other thing that I noticed that I can’t explain. The regular oil was measured at a temperature during these testing cycles consistently. The synthetic showed a reduction of 9 degrees??? What?…….the thermostat was the same…the water temperature was the same. My estimation of this is that the surface/friction loss reduction diminishes the heat production and therefore the oil is cooler? Hmmmmmmmmm? I have to think on this some more. Anybody got an answer to this?

Anyway it was a fun and informative project. I had to give the dyno back or I would have run the Skipper with a clutch to see the results. I also wanted to change my roller rockers out to see the difference with my own eyes on my own engine but didn’t have time either. Darn……

Steve Jack
Concept One Pulleys
Jackstands at Old Cars Only

Last Update: 10/04/00 Hits:

40 Ford Tudor Sedan 40 Ford Tudor Sedan HomeLinksEventsStoreVendorsForumsSpecsPics
George Carnut McDowell
est. 1996