[dataway]

After reading a recent topic on engine overheating I thought I'd simulate the air flow through radiator and fan openings.

Case 1: OEM 19" fan opening and approximation of an OEM fan shroud. Shroud is 4" from fan opening to inner surface of radiator.

Case 2: Dual 12" fan openings with flat shroud placed 3/4" from inner surface of radiator.

Simulation uses a 15 psi air pressure at the front of the radiator. This is slightly above static atmospheric pressure (14.6 psi) to simulate air pressure on the front of the radiator at highway speed. Fan openings are set at static atmospheric pressure.

Front what I can see ... it's all in the shroud. The dual electric fans could probably move as much air as an OEM engine driven fan (if powerful enough) but if the shroud isn't the right design they will pull a high volume of air through a much smaller portion of the radiator effectively reducing the apparent radiator size.

To simplify it, total volume of air flow is not as important as utilizing as much of the heat transfer surface as possible (ie the radiator). For instance you could use leaf blower up close to the radiator and blow a given volume of air through a small area of the radiator and the engine would overheat, you could move the blower farther from the radiator and blow the same VOLUME of air through a larger area of the radiator and the engine does not overheat.

In this study the electric fan shroud being flat, and so close to the radiator allows no organized flow to develop from any area except directly in front of the fan openings.

The arrows represent air flow direction, and the color of the arrows represent air pressure (in this case read back pressure) with RED being the higher pressure (indicating lack of flow to the low pressure fan opening).

Next maybe I'll simulate the electric fan scenario with a proper shroud.

Keep in mind, this simulation does NOT include any actual FAN drawing in air, it simulated flow with the vehicle at speed, no fan running. So it's really much more of a test of the shroud design than it is a test of OEM fan vs Electric fan VOLUME of air flow.

BTW ... anyone know how to embed videos directly into a forum post?


https://www.youtube.com/watch?v=XX4qVqz0dpY Dual Fan simulation

https://www.youtube.com/watch?v=mYMIuvz4KLo Single Fan simulation

[AG]

I calculated the pressure (above atmospheric) going 134 mph and it was about 1.2-1.5 psi if I remember correctly.

[nUcLeArEnVoY]

Great workup, and this reaffirms a notion I always had. You need a shroud that is CONICAL and that is not too close to the radiator core surface, or else the fan will only be transferring heat from an area equal to its exact circumference, leaving the rest of the radiator core untransferred. Some electric fans have a more efficient shroud design, but they're not as pretty on the eyes as the aluminum shrouds. The fan I always tout, the Flex-A-Lite 295 has a shroud with molded conical extrusions around each dual fan opening, giving a more area for the fans to actually pull more heat out of more of the core. I've yet to hear anybody speak ill of it, other than it sucks up a lot of amperage at full send.

In the case of highway speeds, the higher pressure around the fan openings on a shroud that is too flat and too close to the core will just cause the air to slam into and whip/slosh around the inside of the shroud without actually going anywhere.

[JLMounce]

Originally Posted by nUcLeArEnVoY

Great workup, and this reaffirms a notion I always had. You need a shroud that is CONICAL and that is not too close to the radiator core surface, or else the fan will only be transferring heat from an area equal to its exact circumference, leaving the rest of the radiator core untransferred. Some electric fans have a more efficient shroud design, but they're not as pretty on the eyes as the aluminum shrouds. The fan I always tout, the Flex-A-Lite 295 has a shroud with molded conical extrusions around each dual fan opening, giving a more area for the fans to actually pull more heat out of more of the core. I've yet to hear anybody speak ill of it, other than it sucks up a lot of amperage at full send.

In the case of highway speeds, the higher pressure around the fan openings on a shroud that is too flat and too close to the core will just cause the air to slam into and whip/slosh around the inside of the shroud without actually going anywhere.

And the Flex-A-Lite Black Magic 180 for F-body sized radiators. This is what I use and it cools sitting still and at highway speed.

The shroud is definitely critical.

[dataway]

AG, thanks for that info, I was just guessing at what pressure in front of the radiator might be at highway speeds.

I'll bump up the pressure to about 1 psi above static and see what it looks like.

[dataway]

Ran the simulation a couple more times using a positive 1 psi over atmospheric pressure at the front of the radiator. Flow vectors were the same ... just faster.

Added some sensors to get volume flow rate data. Ignore the fact that the values are negative, that is just an anomaly based on the direction of flow in or out of the computational space. Just picture the graphs inverted to the positive direction (still learning the software)

On the single OEM shroud/fan setup.
Total CFM: 35,478.5




On the dual fan/shroud setup.
Total CFM: 28,602.7



Results are not surprising considering a single 19" fan opening has more square inches of opening than two 12" openings. I would ignore the SHAPE of the graphs as the dual fan setup took more "iterations" for the simulator to figure out what was going on and for things to stabilize.

[dataway]

Here is something very interesting.
A dual electric fan setup with a 2" deep shroud actually flows worse than the same setup with a 3/4" deep shroud.

The 2" deep shroud gives the air just enough room to organize and flow perpendicular to the fan openings ... so as it tries to turn the sharp corner and flow into the fan openings it disturbs the straight flow coming through.

A chamfer on the inlet side of the fan openings would probably help a considerable amount.

What appears to be killing flow on these electric fan systems is the flat surface of the shroud more than anything else. And the fact that two 12" fan openings have about 30% more "edge length" where the air has to turn 90 degrees to flow, compared to a single 19" fan opening.

So even with a shaped shroud the same depth as an OEM 19" fan the dual electric fans are going to flow worse. Less square inches of flow area, more edge length, less organized flow because of two openings.

Only place I can see them being any help at all is sitting in traffic at idle rpm when you can run the fans at full rpm regardless of engine rpm ... although I don't know how long the battery would hold up to that.

Here is a Dual Fan with 3/4" shroud depth.





Here is a Dual Fan with a 2" shroud depth. Note how the straight flow through the fan holes is disrupted by the air flowing into the holes from the side. Leading to less total flow, very un-intuitive.
Leading to about 4,000 CFM LESS FLOW.

[nUcLeArEnVoY]

Is there a way you can simulate if conical extrusions around each fan opening like the shroud I mentioned? This is the Flex-A-Lite #295



Note how it's not perfectly flat like most of the shrouds are. Around each fan opening are conical extrusions similar to the fan clutch shrouds that encourage the flow of air smoothly out of the shroud.

I would also imagine the staggering of having one fan higher than the other also helps with your theorized issue of having two fan openings interfering with good air exchange.

[tstroud]

Does anyone make an electric fan large enough to be mounted in the stock radiator shroud in place of the clutch fan?

[nUcLeArEnVoY]

What's the car in question? The fan I linked above fits perfectly on a 2nd gen F-Body radiator.

[tstroud]

Well, the OP has a 1968 GTO so my guess is that is the subject vehicle.

[dataway]

Envoy ... do you know the diameter of the fan/openings?

Looks like the shroud depth is about 2.5-3" ?

My guess is the shape of that shroud makes a big difference. I'll work up a study of that, using the radiator size I used in the other simulations so it's more of an apples to apples.

[dataway]

Envoy, Here is a back of the napkin simulation.

Total flow is 29,344 CFM, so yes an improvement over the other dual fan setups but not a big improvement. In this graphic the flow "pipes" color is based on velocity ... blue slow (or zero), red fast.

A few things to keep in mind looking at these images .. again it's based on an empty hole with no fan .. so no obstruction from blades but ALSO no negative pressure created by a fan that would increase flow.

This is a simulation showing only the flow rate through those holes with a 15.6 psi pressure in front of the "radiator" and 14.69 psi (atmospheric) pressure at the rear edge of the fan openings. The results will only be effected by the shape of the shroud and the size of the openings. It will not take into consideration the flow restriction through an actual radiator (or condenser), blades obstructing flow when fan is off, or increased flow when fans are on.

Also ... when it comes to cooling it's NOT just about the flow, but where is it coming from, how much of the radiator is it using? If flow is 50% less, but it's 30 degrees hotter (removing more heat) ... will it cool better? Will it cool worse? That gets very complicated.

Air entering through the grill of a car and pressurizing the front side of the radiator is going to create a different flow dynamic than fans drawing a vacuum on the back side of the radiator in a vehicle that is stationary.

In the image that you posted the fans appear to be covering much more of the radiator surface, with less "dead air" than in my simulation. What I'm seeing is that the more radiator surface directly in front of the fans, the better it will utilize the radiator, and the more flow there will be.

With real world measurements of that fan setup the figures would be more accurate ... but again, what if a larger opening has a fan that blocks more of the flow when it's not running? All this simulation can do is tell you if the shroud shape and depth is making things worse or better than another specific shape and depth.

I think I'm going to do a simulation using the same fan openings but a ridiculous shroud that funnels air from the entire radiator back about two feet to the fans. A great way to get a grasp on a complex problem is to input ridiculous variables and see where it takes you.

[dataway]

For those that are still interested in this tedious subject matter.

For myself, I have come to a conclusion.
By FAR the the biggest factor for the volume of air flow is just what you would think ... it's the size of the holes

Secondary just how powerful the fan is, and how much flow it blocks when it's not on.

All the simulations I did with these two 12" fan openings hit a brick wall about 29,500 CFM, no matter how gentle the transition are, no matter how deep the shroud is.

Only way you get more air through them is with powerful fans that could pull a negative pressure behind the radiator at highway speeds. Or ... go faster

So while there are efficiently designed electric fan setups, and less efficient designs, they are not going to flow as well as an OEM setup unless they can match the square inches of opening, and the air moving capacity of a 19" engine driven fan.

That is NOT to say an electric fan setup will not work, obviously it does on many modern cars, but it's going to have to matched to radiator size (and capacity to remove heat), and matched to the general geometry of the vehicle (how much pressure is created in front of the radiator at speed), actual fan size, distance from radiator, shroud design, coolant flow rate etc.

[TheSilverBuick]

I've never had an issue keeping my cars cool. Even in stop and go Las Vegas traffic at 115ºF out, so I must be doing something right.

My '69 Firebird, I run a trimmed Chevy HHR fan off Amazon. The shroud covers the whole radiator core so its pulling air across all the fins. On mechanical fans, same deal on making sure the shroud covers the whole radiator, but something I think many over look is the fan should hang out of the shroud by like a ~1/4 the fan thickness so it can shed the air more efficiently as it pulls through the shroud. And don't under estimate the benefits of even a small air dam under the core support.

[dataway]

Silver, yep that looks like a decent setup. Single larger fan worked well in simulation ... and I see it's got a more conventional fan design with plenty of room for air to pass between the blades at highway speeds. Some of the electric fans I see have like 20 blades that look like they block a lot of flow when the fan is not running. Fits your radiator very nicely.

I also like the fact that the fan is centrally located so it also has the effect of blowing air directly on the engine.

Just checked out one on Amazon ... jeez, good deal for what appears to be a copy of an OEM part. I noticed Doorman makes one.

[nUcLeArEnVoY]

Originally Posted by dataway

Envoy, Here is a back of the napkin simulation.

Total flow is 29,344 CFM, so yes an improvement over the other dual fan setups but not a big improvement. In this graphic the flow "pipes" color is based on velocity ... blue slow (or zero), red fast.

A few things to keep in mind looking at these images .. again it's based on an empty hole with no fan .. so no obstruction from blades but ALSO no negative pressure created by a fan that would increase flow.

This is a simulation showing only the flow rate through those holes with a 15.6 psi pressure in front of the "radiator" and 14.69 psi (atmospheric) pressure at the rear edge of the fan openings. The results will only be effected by the shape of the shroud and the size of the openings. It will not take into consideration the flow restriction through an actual radiator (or condenser), blades obstructing flow when fan is off, or increased flow when fans are on.

Also ... when it comes to cooling it's NOT just about the flow, but where is it coming from, how much of the radiator is it using? If flow is 50% less, but it's 30 degrees hotter (removing more heat) ... will it cool better? Will it cool worse? That gets very complicated.

Air entering through the grill of a car and pressurizing the front side of the radiator is going to create a different flow dynamic than fans drawing a vacuum on the back side of the radiator in a vehicle that is stationary.

In the image that you posted the fans appear to be covering much more of the radiator surface, with less "dead air" than in my simulation. What I'm seeing is that the more radiator surface directly in front of the fans, the better it will utilize the radiator, and the more flow there will be.

With real world measurements of that fan setup the figures would be more accurate ... but again, what if a larger opening has a fan that blocks more of the flow when it's not running? All this simulation can do is tell you if the shroud shape and depth is making things worse or better than another specific shape and depth.

I think I'm going to do a simulation using the same fan openings but a ridiculous shroud that funnels air from the entire radiator back about two feet to the fans. A great way to get a grasp on a complex problem is to input ridiculous variables and see where it takes you.

Improvement for sure, but not as much as one as I thought. One thing to factor is electric fan shrouds (such as the one I showed) that have relief flaps that open up at RAM air speeds to allow the higher pressure air sitting in the "flat" areas of the radiator to be expelled. Several e-fan companies such as SPAL implement those flaps.

[kingbuzzo]

ifixedit...

[nUcLeArEnVoY]

Originally Posted by kingbuzzo

ifixedit...

Eff'n legend, dude...

[dataway]

I love it!
I have a dozen of those things lying around. Some with excellent heat sinks capable of dissipating over 100 watts. All of CPU or GPU cooling units.

Right now I'm building an electric bee smoker for the wife. Using a tiny 3D printer squirrel cage fan to replace the bellows and an M12 battery to power it.

Makes me wonder if a person could design an automotive AC duct booster fan that could increase air flow on systems with weak blower fans.

I've got a lot of bad ideas