Air flow optimiyation through radiator

Feynman81 · Apr 11, 2023

Hello everybody,
I'm running into a cooling problem at work on our UAV, essentially the engine is running too hot and I've been tasked to find a solution.

At the moment we have an automotive configuration with the radiator in front of the aircraft and an axial fan right behind the radiator that sucks air in (pull configuration). The fan actually cover only 1/3 of the radiator area and it's basically attached to the back of the radiator with very little gap.

The cooling issue happens when we are on the ground with the fan at max speed, there is simply not enough air going through. We do not expect any cooling issue when in forward flight (fan turned off) since we're gonna have plenty of ram air being pushed through the radiator.

Basically my points are:
- Use two fans instead of one (this alone is not enough though)
- Air flow distribution through the radiator should be as uniform as possible, to achieve it:
- Use some sort of plenum that cover the entire radiator and direct the air to the fans intake.
- The fans should be located as far as possible form the radiator.
- Use guidevanes to compensate for sharp bends (actually used in the past and they work great).
- Redesign the radiator to get a lower pressure drop on the air side.

An additional feature I'd like to implement is to differentiate the air paths when the aircraft is on ground with the fans on and when the aircraft is in forward flight with the fans off. This is both to avoid the fans windmilling during forward flight and to avoid the additional drag from the same fans.

My idea on how to solve this is schematically shown here below.
In the figure is shown the top view of the aircraft, only half of the UAV is shown since it's symmetrical.

My doubt is if something like this would work, especially the point about the two different air paths. The easiest thing would be to put a louver system that close or open the path to the fans but I cannot add any additional weight/actuator so I came out with this "fluid dynamic" solution instead of a mechanical one.

Before starting doing CFD/prototypes I'd like to know your opinion or if you had/know any similar situation. Is the airflow I indicate in the picture realistic for the ground/flight situation?

To summarize my (wishful) thinking is:
- during forward flight no air should go through the fans outlet.
- during ground operation no air should be sucked from the ram air outlet port.

Thanks in advance.

EdStainless · Apr 11, 2023

Are you sure that the air flow is the limiting issue and not the coolant flow?
I don't think that this will work without a damper of some sort.
Could you use the air flow to actuate it?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

IRstuff · Apr 11, 2023

Is the airflow I indicate in the picture realistic for the ground/flight situation?

I don't think so; air is a pretty low viscosity fluid, and the resistance of the radiator is substantially higher than the open mouth of the ram air outlet.

But, it seems to me that blocking the ram air outlet on the ground might work, since that would force the air to be drawn through the radiator.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

verymadmac · Apr 11, 2023

I will guess your airframe config is a pusher prop and a venerable geometry exit & or entry isn't possible & you want a degree of thrust recovery (more likely cooling drag off set)?

Another exit duct seems like a bunch more custom parts / cutouts etc. These problem has been commonly encountered in the homebuilt aircraft community and numerous solutions applied. The simpler the better seems to have best outcome. There is also lots of lessons learn papers on radiator installation design written by NASA and their UK counter parts (1940 mostly).

Why not just go completely automotive and mount 2 fans directly to the radiator. They can always be powered in the climb phase which may let you reduce the size of the ducting.

Have you considered just adding a spray bar to the radiator for the taxi and having not fans. The spray bar can be set to dump the remaining water at the start of takeoff.

WKTaylor · Apr 11, 2023

This appears to be a very typical ground cooling problem... air cooled or liquid-cooled.

Most automotive engines have a fan for pulling [or pushing] air thru the radiator, in static [no vehicle movement] ops. As speed picks-up, the fan has a 'clutch Assy' that disengages the fan motor, when ambient air-flow is able to provide adequate cooling. The fan is simply allowed to 'free-wheel/windmill' by the natural flow-thru the radiator... and only re-engages for greater airflow when coolant temperature rises and demands more flow-rate. NOTE: there are usually separate radiators for the coolant and the engine oil.

Air-cooled engines have a similar problem... low airflow thru the cylinder-fins and the oil-radiator fins. On very hot days, there are strict temperature limits which demand that the engine be shut-down-to-cool-down [on the ground]... or quickly/expeditiously taxi/take-off and enter a reduced-power cruise-climb until temperatures settle into the safe range where increased-power generated heat can be dissipated to safely climb-out.

NOTE. In some case, piston engine exhaust can be used to 'pull' a modest airflow thru a radiator [or thru the cylinder-fins, air-cooled engines] using the 'ejector-pump' phenomena.

Regards, Wil Taylor
o Trust - But Verify!
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation, Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", HBA forum]
o Only fools and charlatans know everything and understand everything." -Anton Chekhov

MintJulep · Apr 11, 2023

No matter how many arrows you draw, air will flow through the path of least resistance. That path is unlikely to be through the heat exchanger.

3DDave · Apr 11, 2023

Vanes can deflect flow with momentum, but in this case that momentum will be low from the fan induced flow. The vanes will act mainly as a restriction.

As mentioned, you need to block off all fan sourced air except flow from the radiator - I'd use a plate hinged from the top, far side of the plenum/radiator that redirects to the fans. If it is arranged right a very light weight/counterbalanced panel could close under gravity, but be deflected by ram air through the radiator in flight; it may be that passive operation by an external flap that is forced flat against the fuselage by flight air could add some more force to hold the panel open.

When on the ground the weight of the panel would allow it to block the exhaust and the low pressure generated by the fans would seal it.

Similar flow control worked on the P-51 but used a pilot to operate the vent.

An additional oddness - in normal flight the air passing over the edges of the vanes will add drag and may generate vortices that might whistle. It will not pinch down as diagrammed.

Compositepro · Apr 12, 2023

Guide vanes have zero effect on flow upstream of the vanes.

rb1957 · Apr 12, 2023

you're using the heat exchanger on the ground ... so why not a vane that closes the heat exchanger in flight and closes the engine inlet on the ground ?

there are obvious safety issues with this (vane closing the engine inlet in flight) but nothing that "can't" be solved ...

or maybe multiple vanes which close the heat exchanger inlet in-flight but open to allow the heat exchanger to "suck" whatever airflow it wants on ground ? This way you don't obstruct the engine inlet.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

Sparweb · Apr 13, 2023

We do not expect any cooling issue when in forward flight (fan turned off) since we're gonna have plenty of ram air being pushed through the radiator.

There's no reason that has to be true. You could find that in flight, full power from the engine(s) produces more heat than the radiator can radiate no matter how much air flows past it. The radiator could be undersized for the engine. Perhaps so much that so no fan or duct can fix it.
You can use thermodynamic principles to estimate the heat rejection from the radiator to verify that it's properly sized or not.

CWB1 · Apr 13, 2023

The fan actually cover only 1/3 of the radiator area and it's basically attached to the back of the radiator with very little gap.

If the HX is appropriately sized then your fan is likely undersized by more than half. If its also lacking a well-sealed shroud then cooling issues should've been expected by every engineer and technician worth their paycheck.

Before starting doing CFD/prototypes I'd like to know your opinion or if you had/know any similar situation. Is the airflow I indicate in the picture realistic for the ground/flight situation?

Stop thinking of this as an airflow issue, stop trying to reinvent the wheel, and please put down the CFD for a few years. You have a very basic heat transfer design problem that any 1st-2nd year undergrad should be able to reasonably optimize. If you dont have one already, you need a basic spreadsheet for the heat transfer calcs to understand each variable and ensure that you have excess cooling capacity under every operating condition. You then need to wire a few thermocouples to validate the spreadsheet and troubleshoot the existing design.

Your sketches are leading you to a hugely inefficient (read: oversized/HEAVY) cooling system for no real benefit. Shutters are lightweight, simple, and cheap, even light-duty automotive uses them to control airflow through radiators.

rb1957 · Apr 14, 2023

hummm, radio silence from the OP.

why won't you put the ground fan in the ram air outlet ? I'm guessing the blockage of the fan may (??) not affect the airflow in flight too much ??

how about adding the fan aft of the ram air outlet for use on the ground ?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

WKTaylor · Apr 14, 2023

RB... Your idea...

Ground-launcher/test-station with built-in forced-air engine-cooling... to push or draw air thru the radiator for added cooling when running-up... that is left behind at 'launch' [un-needed in flight] = KISS!!!

Regards, Wil Taylor
o Trust - But Verify!
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation, Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", HBA forum]
o Only fools and charlatans know everything and understand everything." -Anton Chekhov

Sparweb · Apr 16, 2023

That sounds like a toy I played with as a kid. Hand-held launcher used to "pump up" the toy helicopter, then release the heli to fly away. It didn't have to carry the launcher because that stayed in your hand.

CWB1 · Apr 17, 2023

I'm guessing the blockage of the fan may (??) not affect the airflow in flight too much ??

IME HX/radiators are the major flow restriction hence the importance of shroud sealing and fan sizing, and shutters are used for warmup and other thermal purposes not aero. A windmilling fan is a comparatively tiny restriction.

rb1957 · Apr 20, 2023

but windmilling requires beta control on the blades, no?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

3DDave · Apr 20, 2023

Feathering requires control. Props will windmill all on their own and generally create a higher drag than if locked in position.

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Air flow optimiyation through radiator

New member

Recommended for you

Materials

Aerospace

Mechanical

Active member

Mechanical

Aerospace

Chemical

Aerospace

Aerospace

Mechanical

Aerospace

Active member

Aerospace

Mechanical

Aerospace

Aerospace

Similar threads

Log in

Part and Inventory Search

Sponsor