Thread: Thermofan over Viscous?

g'day, let's see if we can drag this old topic kicking and screaming back into the daylight


I recently revisited this topic in my head since I was working on the build of my own car and was searching for PWM controllers for fans. That long story I'll keep very short: it's not easy if you want to figure out what the PWM frequency should be to run a DC motor optimally. My homebrew setup with an arduino has a "nice" 500hz tone coming from the coils of the AC fans... Anyway, I digress as usual.


The reason I started thinking about "to viscous or not" had to do with the above experiment. I want/need the A/C fans to be there, no choice really, and I want to use them whilst stationary. In order to get the thing as quiet as possible I wanted the RPM to be variable, hence the PWM since it is more efficient than regulating the voltage through a beefy transistor. Since I need two fans anyway for this purpose and the two already on the car take up about half of the radiator I began my research and thought I'd share my thoughts here.


The last we discussed in detail was the power requirements of the fan. I seem to recall that I did an experiment with a drill and the actual range rover fan. at around 1500RPM it consumed about a kilowatt and if air is linear (which it is not entirely) at around 3000RPM, the speed a hard working engine is usually running, it would consume double that. It was also mentioned that the VC eats around 11hp. I did some searching and found a video of a bloke who had a belt driven fan (not clutch like ours) on a pickup with a tuned V8 to 400HP. He had a flex fan installed, ie one of those models that bend when speeding up which is about the same as our clutch driven fans, they ask less of the engine when the RPM really go up. On a dyno run he lost about 8HP. This translates to 6KW which is A LOT in terms of 12V and I think we concluded quite definitively that we do not need that much power.


I have seen tests of spal vs contour vs some other brands and what stuck out the most is that indeed some fans work better than others but the most important take away, I think, was that the guy could not get all the shrouds to fit as well as they could and thus lost air flow. Another thing that I noticed is that the general consensus is that a push fan is 80% less effective as pull fans.


That last part had me thinking: the way the myriad of radiators is mounted in our P38's make air flow a bit difficult. At the front there is the A/C condensor (and technically 2 fans on the upper part), then there is 2 oil coolers (I have 3 since I run my winch hydraulically so need cooling) and then there is the actual engine coolant radiator. The shroud for the coolant radiator fits snugly but the stuff that comes in front of it has barely any ducting whatsoever, in other words air can slip through between the radiators. It seems the engineers either did not care or did not find it important to do so and I think all this combined explains why the AC fans are not that useful for lowering ECT.


The AC fans push, which is less effective to begin with, through the condensor and do that job fairly well. Behind them sit the oil coolers which will have some benefit of the extra air as well but there is enough space for the air to escape around them and blow out the sides. Whatever air is left pushes through the radiator at the back. Since they were never designed and meant to cool the radiator, they are indeed only a small added bonus. As for the VC fan, it pulls through the radiator with the help of a nicely fitting shroud and thus cools ECT quite well, at least when it is spinning hard enough. Even though it then starts sucking air from wherever it can for the purpose of cooling the engine it does not matter since it is from outside of the engine bay at that point and thus "cool" air. With its copious amounts of over capacity it will do the job of the rest as well, even if it sucks in air from the sides and what not (between the radiator, oil coolers and condensor), whatever is left will be quite enough under most circumstances. The only part that needs the extra cooling is the A/C when driving very slowly, thus the thermo fans.


With this in mind and eager to see if I can find a way to get my car "full electric" I started browsing fan solutions. Since Spal is a well known brand and available just about everywhere I started with their product offerings first. It turns out that they make a number of nice 11" fans that are actually quite cheap. With the dimensions of the radiator I could fit 4 11" fans over the entire area. with 1100 CFM a piece they would add up to 4400CFM in total. They use around 10,5A a piece which would add up to 42A in total. At 12v that is 500watts of power. Even with your average alternator being around 40% efficient at most that means you rob the engine of 800 watts of power which equates to just about 1hp.


The figures say that it SHOULD work out and I am tempted to try this. Since I could ditch the two front A/C fans and the VC fan and shroud in terms of weight it should be close to the same. With 4 fans I would have plenty of redundancy to get out of trouble should one fail, I might have to slow down but that's ok and it would suit my stationary silent condensor cooling as well though it might be less efficient. I could however improve the ducting around all the radiators and since the oil coolers are on top, I could switch on the lower 2 fans for A/C duty only and make it more efficient.


The only real downside I can see is that the VC fan is much farther back and combined with the shroud does not really block the radiator in any way, ie. the full surface area of the radiator is cooled since the air can pass freely. When using 4 11" fans the shrouds would block a significant area of the radiator since they are not as deeply mounted, not to mention 4 electric motors that can block the flow. Even so, from the front point of view, ie. the air being pushed into the radiator whilst driving, we actually increase the aperture by 35%. Not sure how much of that is taken up by the fan motors though.


I think that is what has gone "wrong" with Keithy's experiments; the fan probably does not fit well enough, the shroud for so far it has one sits to close to the radiator and blocks it a bit and potentially it is not quite powerful enough to do the job. It would be "cool" (pun intended) if he were still on the case because I am quite curious of it can be done


these are the brushed fans I had in mind, I would have to make my own PWM controller for them: Spal va04-ap70/ll-37a | Spal Lufter 12V 280mm | Spal Axial fans (brushed)
Spal also make a brushless PWM fan, though they are about 3x as expensive and I would need to make a shroud: Spal va99-abl315p/n-101a/sh | Spal Lufter 12V 280mm | Spal Brushless axial fans


Also, the PWM version has a MUCH higher CFM rating at around 2400 a piece which would add up to 9600CFM in total, if that ain't enough... The current draw at full tilt would also be quite the thing 18Amps a piece for a total of 72A. My guess is that would be overkill.


When I read about the disco 2 experiences and the classic rangies with only 2 fans I tend to believe I should be able to make this work, or not?


Cheers,
-P

Originally Posted by prelude

g'day, let's see if we can drag this old topic kicking and screaming back into the daylight ............................

...................When I read about the disco 2 experiences and the classic rangies with only 2 fans I tend to believe I should be able to make this work, or not?


Cheers,
-P

Hi, I wrote a long piece about why you shouldn't try this on a P38 but got timed out.

In short, thermos work well for a RRC coz there's a common Ford set up with the shroud and 2 fans that is almost a perfect fit.

This set up has cooled older Ford 4 and 5 lire engines for decades now.

It works because of the stock tight tolerances. It's way more efficient than a VC fan, so needs way less 'power'.

These things don't come close to fitting a P38 radiator.

There are pics of mine on the previous page. Keith looks like he gave up............... I'm just trying to save you some time consumed, expensive disappointment.

cheers, David L

Yeah, that time out is quite annoying at times What I do these days is select my entire reply, copy it, press submit, log back in, get a blank screen, reload the site, get back to the topic, press reply and paste the entire thing in the text editor and press submit again. Of course things like lists, quote's and links need to be redone but I usually manage within the time limit

In any case, thanks for your reply. Indeed I think that you are right that the RRC solution will not work for the P38. That is why I was looking into using 4 fans with shrouds that will cover the entire radiator and be a near perfect fit. You think there are other factors that might not make it work? like airflow through the engine compartment or something?

Cheers,
-P

session cookies are set to 3600 seconds (60 minutes)

if you get logged out prior to that then something is playing with your session cookies

cheers

David, I agree that the RRC setup would not work on a P38 since we do not have similar radiator designs. The P38 has an almost square radiator whereas the RRC and the discovery 1/2 have a more rectangular one at least that I could see. This helps a lot with fitting fans since it is much easier to place two fans next to each other covering nearly all of the core. It also helps with the shroud since you can have a better angle which helps airflow, more on that later.


This weekend I have been going over all the data once again and have also been double checking my numbers and as it turns out, I made a bit of a mistake. The space for the radiator in the P38 is roughly 600mm wide and something like 560mm in height. Whilst four 11" fans would fit that area, there is of course the small matter of the tanks and in/outlets. The dimensions of the core are more like 580mm by 490mm. It would be harder to mount 11" fans this way, although fans by their very nature are circular we could modify the shroud where we would need the space for the in/outlets whilst still presenting a very large surface area to the fan. The biggest downside I reckon would be that you lose a lot or surface area of the fans that simply suck at parts of the radiator where no air can flow anyway, unless one would make a shroud that is fairly deep.


As an alternative I started looking into a setup that would fit the core as perfect as possible, I ended up with four 225mm fans. I also tried fitting the largest fans that would fit diagonally and ended up with two 385mm fans. The largest single unit fan(s) I could find were around 18" whereas the "perfect" size for our radiators would be around the 19" mark. It seems though that 18" fans are already quite rare and may not be readily available. Examples from a BMW X5 or 750 V12 came along and even spal has one, although not on their european site. I could not even find places to buy them other than in south africa and on ebay. 19" find from a reputable brand I have yet to find but then there is the question; would we want a single fan? Let's look at the powered flow numbers first:


4x225mm fans : 743cfm a piece which is 2972cfm for all four.
2x385mm fans: 2016cfm a piece which is 4032cfm for the both
1x465mm fan: 3600cfm


The last unit we should be comparing is the VC fan to begin with of course. Let us ignore the power usage for now, we have gone into that theory a lot in previous posts, and try to figure out it's performance. We know that a good VC bi-metal unit achieves at least 85% of input RPM when hot but it will also start slipping at higher RPM to prevent the blades from flying off. It seems 3000RPM is a much quoted figure for top speed in regards to VC operated units. Makes sense I guess since most of us will not run above 3500RPM for any extended periods of time? In any case we know the diameter of the VC fan to be around 450mm in a shroud that has an opening size of around 500mm. There is quite a bit of tolerance between the both since the engine can move in it's mountings but the radiator and shroud are fixed so we need those tolerances. This no doubt decreases efficiency but I do not know by how much. For now let us assume 3000RPM at max and compare it to the closest contender in our list. It can produce 3600cfm at almost 3000RPM. I am sure the VC can spin a bit faster still but then again the thermofan is of a more efficient blade design so I guess the bottom line is that we can expect around 4000CFM from the VC unit under most conditions, no matter how we load the engine up since RPM = RPM.

I did not take into account pressurized operation at this point (ie. the resistance through all components) but suffice it to say that one would pick the best functioning unit at higher pressures since we already concluded there is a lot of resistance with 3 radiators stacked in a row and considering the original fan, as pointed out above, has it's limits in performance as well. What I mean to say with this line is; not all fans are created equally and looking through the available fans in different size you get different performance specs. Some are more quiet, some push more air. We'll go with more air.


Anyway, let's step back a bit and see what Keithy's problem was in running the thermofan: he could not maintain decent temps with 40km/h at a steep gradient. From the top of my head I would think that engine RPM at that speed to be around 2500? My car is in pieces so I can't confirm at this time His conclusion was however that where the VC fan had no problems his thermofan was not up to the task but that is just part of the story as I mentioned to David above; it's just as much a case of fitment and airflow as anything else. Let's take a look at the surface area numbers for air to flow through.


I calculated the effective "open" area assuming a perfectly fitting shroud for all the options and also deducted the surface area of the motor.


4x225mm fans: 398cm2 a piece with a motor that is 98cm2 resulting in 1200cm2
2x385mm fans: 1162cm2 a piece with a motor that is 124cm2 resulting in 2080cm2
1x465mm fan: 1700cm2 with a motor that is 314cm2 resulting in 1386cm2 (this figure is also very close to the VC fan area, it would be slightly larger, say 1500cm2 at most)



Mind you, the total theoretical surface area of the radiator, ignoring grilles, bumpers and what not is 58x49 roughly and that makes for 2842cm2. Looking into Keithy's setup and designs using thermofans in stead of the VC fan in general found all over the net I am beginning to see a pattern: Most thermo solutions use a near square and tight shroud, and with that I mean the angle and distance between the radiator and the back of the shroud, whereas the VC solutions generally do not since the fan is close to the engine and the radiator is reasonably far away facilitating a more sloping design. I have also noticed that in some applications of thermofans the radiator shroud has one way flaps that open up when driving to let air through the shroud but they close when the fan is producing more air flow. This got me thinking: what is the air flow at 40km/h such as Keithy tested?


Again, I did not take into account all the stuff that is in the way but just calculated the maximum theoretical. with a frontal area of the radiator of about 3 square foot (since all fans have a rating in CFM I calculate it in feet) and an air speed of 2178 feet per minute (or 40kmh) we come to a staggering 6561CFM. I guess there is much in the quote "do not underestimate cooling from driving" that has also been uttered in regard to alternator cooling elsewhere, a subject I am into as well at the moment.


So, finally, for those of you who hung in here and had the patience to read trough my ramblings: I believe the reduced airflow of Keithy's setup to be the biggest problem. Not just the fact that his fan can not do more CFM at full speed but that the entire flow of air through the radiator is severely restricted due to the shroud design.


Perhaps as a final note I should clarify my own desire to have an electrical setup in stead of a VC one. Apart from the fact that I wish to reduce components on the one hand, ie. the push fans for the radiator condensor should I not need them, I also want to make the rest of the cooling system more reliable. Arguably a belt driven system is one of the most reliable to be had however genuine landrover parts are drying up for some vehicles (such as the P38) and in the end a single fan is still but a single fan. If it breaks for any reason, no cooling. Having multiple electric fans would certainly increase redundancy since if one breaks the other(s) will continue to work in most likelyhood. Yes you increase components and possibly complexity but I believe it to be manageable.


At this point I might be trying to do too much at the same time but I have found this thought experiment worth while and I'll ponder on it some more to see what I will do going forward.


Always interested in hearing from you guys!


Cheers,
-P