Well you would be correct, I'm not an expert in fluid dynamics. I take it from your post that you do understand fluid dynamics. I'd like to understand it, so can you explain to me exactly why there is a difference?

This is a hugely debated topic in the automotive and the electronics world. So far I haven't found any proof positive information either way other than lots of opinions.

For what it's worth, I prefer the fan to pull. But I truly wondering when all things are equal, is there a real difference?

 

Bolded part above is correct EXCEPT you forgot that with a puller the cooling air has already passed the radiator core and therefore it doesn't matter if it now gets disrupted by the fan motor and brackets ... who cares ... it's already done its job. Hope you can see the difference?

 

Maybe explaining it this way helps:
-Objective: keep the CPU running cool.
-Solution part A: attach a heatsink. Why? Because a CPU has no fins.
-Solution part B: attach a fan to the heatsink to cool it so that in turn cools the CPU (a device that has no fins).
-Solution part C: fan should be a pusher. Why? because it makes no sense to use a puller to draw air past a hot CPU (no fins) in order to cool the next item attached to it, the heatsink (the device with fins) ... that's counter productive. AND take a look at that heatsink pancake pusher fan, you will notice that the motor is on the side opposite the heatsink (or at least it should be mounted that way). Why? because placing a fan motor adjacent to a hot heatsink doesn't promote adequate MTBF.

So why did I say that the heatsink is the focus? Because that is REALLY the device that you need to focus on with the push fan. Cool the focus point (heatsink) and accomplish your objective: "keeping the CPU running cool."

Make sense now?

 

7T1vette has it correct when he mentions that fluid dynamics principles are key to understanding pusher and puller fan designs and, most importantly, when is it most advantageous to use either.

What you're trying to cool and how that device is mounted to other stuff around it is important. lionelhutz and his computer CPU example is an apples & oranges comparison because a radiator has nothing adjacent to it on either side of the fins ... a computer CPU has a system (mother) board on one side and a heatsink on the other. Hence for one, a puller is more efficient (and practical) and, for the other, a pusher is more efficient (and not counter productive).

Fluid dynamics and thermodynamics are required undergraduate courses for mechanical engineering majors. I didn't want to use the phrase "fluid dynamics" in my posts above for fear of being tagged as an "uppity-up" poster. Trying to stay

 

Blade design, as has been said, will make a difference. Every thing I have ever seen from fan companies recommends they are mounted to pull. I've always run them that way, but I have had friends try them both ways and they said they cooled better pulling.

 

My apologies to the OP as lost in my posts to further this discussion was my answer and that is to configure his fans to PULL. AND I would follow the advice from wilcar about making a new circuit. That's exactly what I did on my '82 although I did not use the same solenoid part number that he recommended. I used this one: The color matched my vette

 

No, it does not help. I have never seen a CPU heat sink that pulls the air past the actual CPU. The side of the heat sink against the CPU is flat with no way air could be pulled past the CPU. Every CPU heat sink is designed so air goes in one side and out another side (or sides). The fact the air may have to change direction is irrelevant.

As for you claim that the CPU would heat the air.

1. You have claimed it has no fins and can't be cooled directly. Well if it can't be cooled by the airflow then it also won't heat the air flowing over it.

2. Even if it did heat the air, that is cooling that the heat sink doesn't have to do. It would all balance out.

As I and others have posted before, the fans blades for radiator cooling fans are typically designed to be pullers so they are not as efficient when run backwards and used as pushers. The S curved fan blades are directional. When used in a shroud, the shroud design to allow better free flowing air (fan off) is easier when it is behind the rad (rubber flaps in the shroud). Also, the fans are often sold to replace mechanical fans which are all pullers. These all trade-off possible best efficiency of the fan for packaging or better operation when the fan is off.

As I posted before, power electronics designs always blow with the fans. Pullers are only used to supplement the pusher airflow after running out of room on the heat sink for more pusher fans. A radiator is a heat sink just like any other heat sink.

 

I don't really care for the comparison of the radiator to a CPU heat sink. Mainly because a radiator is designed for the airflow to go completely and straight through whereas a heat sink is typically not through but on or around. This is apples and oranges in my mind. Each have their own design parameters.

 

Both are designed so that air passes over cooling fins. The air certainly does pass through the fins of a CPU heat sink. Not sure how that that is apples to oranges. It's a mistake to believe that having a design where the air has to turn changes what it's doing. Besides, many power electronics heat sinks are straight through flow designs.

 

If you have room, a pull fan is best because you leave the front of the radiator open to clean air. Putting a fan on the front, blocks to much surface area. Only use the front if you have no other option. You need all the cooling you can get.

 

You're I answered the OP's question with my .02 so I'm

 

You guys are hurting my brain!

Push or pull, as long as the air moves front to back, it is a minimal difference.

Yes, there is complex science involving thermodynamics, but in this case it's a moot point.. ( "moot" spelling? ! Lol !)


Fascinating, but overkill!





.