I've been tryin to talk a buddy into droppin my sr20vet in his brz for over a year... Now that I'm putting it in the b14, amd after seeing this series, he wants it 😆
So nice to see a duct created correctly. See so many people making ducts that go from bigger to smaller thinking that it is going to work better. The smart guys figured this out decades ago...
Love the fan work, this kind of stuff take forever. Minor technical nits, you want the highest mass flow possible through the heat exchanger for maximum cooling. The problem is a duct you put in front of hx is likely going to stall (flow.separation) at some range of Reynolds numbers, this makes mass flow.really drop. To help keep it from stalling, put at least a 10-20 mm radius on the inlet as you won't reattach in a duct like that if you get inlet separation (look at F1 duct inlets).
@defsr7 Help me wrap my mind around something: if you want more mass flow through the heat exchanger why is it beneffitial to have a smaller inlet? I'm not saying they should go the opposite way. I trust this is proven experimentally and theoretically. But I don't understand what would happen if your inlet was 100% or120% of the frontal area of the heat exchanger? Will the air "spill" around the inlet? And what determins the "optimum" ratio of inlet to heat exchange area? Is it a function of the drag induced by the heat exchanger? A function of the speed? All of the above?
@@TrackDaysMX The smaller inlet helps expand the air, so it does indeed convert more of the total pressure (which is Ptotal = Pdynamic (the moving part) + Pstatic (what pressure you see referenced 90 degrees to its streamline) to static pressure. Essentially you slow flow down from a high speed, small inlet at some given Ptotal, to hopefully the same Ptotal (not possible, there are *always* losses due to friction/separation to various degrees) but at a higher static pressure and lower dynamic pressure. This Pstatic is what you really need to drive flow through the radiator core. Why does a giant inlet of say 1.2x the radiator area not flow more? Well, the radiator exit probably has a pretty high Pstatic (slow moving flow there) even if it's evacuating to a fairly low pressure area, so it needs to build up some Pstatic to actually flow across the radiator core. A giant 1.2x radiator face duct inlet is going to be seen as incoming flow as a giant "brick wall" of almost no airflow, so most the air hitting this giant brick wall of a duct will generally spill out and around it. This is going to create a ton of drag, as you're pulling this giant area of generally high Pstatic along and doing nothing but forcing a ton of air out of the way in a generally turbulent/uncontrolled fashion. A well sized duct will expand the air just enough to get the Pstatic high enough to flow at an optimal amount depending on your radiator fin density/exit Pstatic etc. Expand it more than this and you've throttled the inlet down and you just can't get enough flow into it (Ptotal is fixed in airstream). Go above this perfectly expanded inlet size and you do tend to just stall flow in the duct as it chaotically and not uniformly slows down. It also tends to be highly unsteady, which creates tons of vortices and buffeting (more of that drag thing). Here's a good paper going on the basics of duct design. I wouldn't put too much faith in this undergrad level of CFD, but they get the general idea right, and there's a good graph showing that for their application, an inlet area ratio of about 0.8 gave them peak mass flow through the radiator: openscholarship.wustl.edu/cgi/viewcontent.cgi?article=1101&context=mems500 I imagine if they looked at what is happening after the HXer in a bit more detail, they'd probably trim that down to a 0.6-0.7 area ratio for peak mass flow.
You're right, some large radii would definitely be a good addition in future. Like many things when you're in a hurry like we've been, you've got to draw the line somewhere in order to speed up manufacture time - Tim
Awesome man!! I built exactly this for my racecar, and not only did it provide exceptional cooling - it produced as much approximate downforce as my front splitter. Once I redesigned the splitter with tunnels - front downforce was insane. Look forward to seeing this build in person.
One thing doesnt make too much sense, Tim-sleepy-guy mentions at about 2:00 that slowing the flow means the air molecules will be more time in contact with the radiator, thus absorbing more heat, and that this is beneficial -> actually, you want the air to get out of there asap, thus the higher the velocity, the better, because you have a bigger airflow and thus more cooling capacity, everything else being equal.
Ya exactly. I don’t know why they think creating a bottleneck at the intake is going to do anything besides limit your cooling. They are just limiting the degrees of angle that the surface of the first heat exchanger can pull in air. If the electric fans are properly shrouded and the radiator and inter-cooler are properly sealed together that intake bit is useless besides a flow restrictor.
@@dsauce8780 the whole point of the bottleneck is to expand the air and increase air pressure on the radiator front face. You are wrong in your assumption that they are restricting flow through the radiator with their duct, especially at high speeds: the air will always follow the least resistance path, and for sure that path is not through the radiator, that's why they need to create a high pressure area and seal it against the radiator his only mistake was to assume that there will be more heat exchange because the flow is slower.
Filthy uh I don’t follow you. That is most definitely a restriction at any speed considering they have a properly shrouded electric fan pulling air and idk how you can argue against that.... but I’ll trust you know what your talking about for what ever application restricting airflow increases cooling in lol. If they are concerned about sucking hot air at idle, evacuating hot air under power, or the aero being soooo bad that the radiator is starved then that’s a clear reason to add a bottle neck somewhere but seeing the project that doesn’t seem to be the case.
@@dsauce8780 well, you are either trolling or never heard about the mad lads Bernoulli, Reynolds, Prandtl and Nusselt. Your comments show you don't even have a basic knowledge about fluid dynamics. Here's a paper that proves a bigger inlet is not always more efficient, look at table 2, page 8: jestec.taylors.edu.my/vol%206%20issue%201%20february%2011/vol_6_1__094_108_charitha%20ds.pdf
Filthy wow I guess I’m an uneducated troll? Did you even read that paper? Are you trolling? The conclusion is that the largest throughput with the least restriction produces the most cooling. The whole paper is examining the air throughput based on the engine’s cooling requirements and the drag effects they produce in order for them to understand the acceptable trade off between cooling and drag for their application. The column in the chimney or inlet does not need a bottle neck, compression, or expansion to increase cooling. Just needs to be non turbulent. Sounds like you are asserting a bottleneck or restriction for pressure differential = laminar flow which is completely incorrect. The largest possible inlet with the least possible restriction(restrictions including turbulence) is the most effective. You cannot get around that. The main idea being reduce all sharp compression to avoid the vacuum that inevitably comes with it accompanied by air rolling and creating turbulence. Go look at Steve Morris’s pro-volute centrifugal super or turbo intake piece. Entirely built to create uniform flow upon entry to the compressor avoiding unnecessary turbulence and therefore restriction. A Cardboard cutout with reinforced walls with no volume change would do more than the inlet in the video. You sound super educated tho. Even liked your comment for typing Bernoulli lol because idk anything about thermo er I mean fluid dynamics. Just remember low velocity is the problem here(your reference, 3.3 grills and chimneys). Look up gradient optimization for more info. Doubt ya will. Based off the tone of your last comment you have already been educated.
Good that there is so much free space bt radiators and engine, which in fact is quite strange, remembering that this was originally place for a short length boxer!
Engine was mounted as far back as possible without having to cut the firewall, which certainly helped. Still not as much space as we would have liked! - Tim
Cool video. Do you know how to use a simple water filled manometer to check the pressure gradients through, and across the ducting and matrix, respectively? Might be useful for the fine tuning, especially if using a Gurney flap on the top edge of the exit?
We have actually considered measuring the pressure at different points through both ducts at some point in future, would certainly be interesting the check the effectiveness. The outlet on the bonnet does have a gurney-like feature on the front to help create a low-pressure area to help draw the air through both ducts. You'll be able to see it in one of the next updates, good suggestions! - Tim
Logical, sound theories there guys. If only the manifolding sat the turbo further out. Never ending with saloon cars, single seaters are easyer to work with. Worth checking that the exit duckting isn't too far back on the bonnet. I.E the air over the top does not reattach itself after the low pressure area behind the bonetts leading edge. Hope all is more or less on schedual 👌
@@hpa101 Sorry to learn there were overheating issues in testing and the first event. If you read this, is would be worth checking there is a head of water above the top of the engine. Looking forward to future episodes.
It's true using strakes in ducting can help keep the flows cleaner, but if we were going to use these anywhere in this setup it would be in the exhaust as that's where things tend to the be the most turbulent. Good call though - Tim
For sure can be a good way to do things. For us that wasn't an option with both manifolds and turbo taking up so much space at the top/rear of the engine bay - Tim
Can this same concept apply to ducting it out the bottom of the car? Heat would theoretically create a high pressure zone under the car causing less max grip? My car is primarily a street car and don't plan to be pushing limits on the track for anything. 91 300zx
Thanks for reminding me about air flow out. I literally forgot about it I did directing air in but not out. Where's my cardboard cheap and easy to work with. Has there been any don't on installing dry sump oil systems like conversating from wet to dry?
Glad you found it helpful, there's definitely a time and place for cardboard and tape! Here is a recent video we did on the dry sump system on this car if you're interested - Tim th-cam.com/video/m8ofSWgtWKQ/w-d-xo.html
As a guide line, what should we be aiming for in terms of surface area ratio % of intake, heat exchanger and outlet. Should it be something like 40:100:50?
I’m curious what the reason for not making the bonnet exit narrower (side to side)but deeper (front to back - to maintain your 40% ratio with the radiator) to avoid having to do all the extra scolloping and shaping to clear the intake? I’d have thought just as you’re squishing the air top to bottom you could apply the same principles squishing it side to side?
Not sure there was enough room to make the exit larger front to back with the rocker cover in the way? Not sure if making it larger towards the front of car would muck up the gradient of the exit and how that would affect it though
Main reason is packaging, there's simply no more room with the engine in the way. But another thing to consider is that in general you don't want to go too far back on the bonnet (especially in the middle) as you tend to get a high pressure area in front of the windscreen that will degrade the performance of the duct - Tim
Great content guys! Would an inlet still be beneficial if an outlet isn’t achievable due to engine and intake configuration (k swapped Integra) We have bonnet venting already but ducting the rear isn’t viable
With this style of duct, for sure the outlet is just as important as the inlet, they're designed to work together. Packaging is always a challenge with this stuff, there are certainly plenty of compromises in this setup of ours as well simply due the space restrictions! In saying that, if there's simply no room then yes it is probably still worthwhile to have the inlet only. For the outlet, just do the best you can to try direct the air exhausting from the heat exchanger to the lowest air pressure area you can, good luck! - Tim
Hi guys, great work!! Is the ducting in the engine bay sealed or is it drawing from the bay also? Just trying to get my head around the physics involved.
Thanks! The inlet duct is sealed to the intercooler, the outlet is sealed to the radiator and the intercooler and radiator are sealed to each other. The idea is to keep the flow through both heat exchangers as separate as possible from anything else like the engine bay. This style of system only works properly when you can control the airflow - Tim
Thanks HPA. Cool video. I recently made something similar to this for my Mazda RX7 racecar. Due to class rules, I have to run a standard bonnet with no vents. The back of my radiator has no ducting, but the front is very similar to your 86. The oil cooler sits in front of the radiator and the exit ducting goes underneath the radiator. My question is: will I decrease water temperate by creating a duct behind my radiator to increase the exit speed? And, if I am forced to duct that exit air towards the ground, will that give a negative measurable effect on aero? All help appreciated.
Certainly worth putting the effort into an outlet duct, it's just as important as the inlet to maximise the cooling and efficiency. If you get the outlet duct right, you will definitely end up with more cooling capacity for less drag. Ideally try and exhaust the air in a place of relatively low pressure and high speed flow to help draw the air through the outlet duct. Fundamentally, flow is driven by a pressure difference, so you should try to put your duct exhaust in a place with low pressure. Finding a place to fit everything and rules will almost always mean you have to compromise on this, you just have to do the best with what you have. The effect on your aero will depend on what sort of aero elements you have on the underbody, just remember to aim for low pressure! - Tim
Could you possibly use fins inside where the angle gets steeper to try and keep the air connected to the surface rather than detaching and causing stalled air?
What are the benefits of bringing the outlet over the top of the vehicle instead of under? Or if space permitted, out the fenders? More downforce? Faster air over the car acts like a venturi pulling hot air out of the duct?
The two main things to consider are packaging and aerodynamics of the outlet areas you have available. Most often, unless you are building an entire car from scratch there limited options of the places you can actually fit the ducting. Once you know the possible places you can actually fit it, you need to look at the likely pressure and velocity of the potential outlet areas. To make the outlet as effective as possible you want to exhaust the duct into a low pressure area, this helps "drive" more flow through the whole system. For us, the bonnet was a sensible place to package the duct. With the outlet (which isn't shown in this video but you'll see in later ones) we use there is a small gurney-like feature on the front of the outlet. This works by taking the high velocity clean air at the front of the bonnet and disturbing it, which creates a low pressure area in the duct outlet which helps drive the flow out through the duct. This arrangement gives us more cooling for less drag, therefore giving us a higher efficiency - Tim
Why not make the outlet more square to get horizontal clearance? The flow out of the radiator has to deflect up pretty aggressively regardless so I wouldn't think the change in profile would change the air flow to significantly, particularly compared to kinking the inlet duct and maybe making a turbulent section beyond the kink.
For our case, we really maximized the amount of space we had available in front of the engine. Packaging proper ducting in cars like this always has massive compromises, this example is no different. We were trying to balance the outlet cross-section area we needed with the space in front of the engine and trying to make the outlet angles as gentle as possible, this pretty much defines the aspect ratio for you. We have to work with what we've got! - Tim
Sizing the heat exchangers is best done together with the help of the company that builds them. This is because they are (hopefully!) the ones that have the heat rejection data for their products. In the absence of any proper data - which let's be honest can be difficult to come by unless you're working with a high end product, it's usually just experience and testing. We have a fair bit of instrumentation on this setup which we can use to understand the effectiveness of both heat exchangers in this car which is helpful - Tim
@@hpa101 Hi, thanks. Im building a 550hp Porsche 951, where i need to get a new cooling package. i can fabricate endtanks and other parts, but to calculate the cores, i need help. Thanks. - Erik.
@@ealoken If you are going to build your own radiator then the best bet is to buy a core from a reputable manufacturer. Any reputable core supplier should be able to supply heat rejection information for their cores. Using some generic efficiency and heat rejection assumptions for an internal combustion engine, for a given power output they will be able to help you to size it for your application - Tim
Is an outlet duct really necessary? I the ducting is more pressure behind the radiator than if it were simply open to the engine bay, and then the hood vents could help heat escape from the engine compartment. Turbulence is the only con I can think of.
Packaging, if you look at the jgtc supra, it has a 3s, but its pushed way back into the fire wall, that allowed them to go v mount. V mounts work well with fd3s rx7s too. So it all comes down to packaging. Those cars work because of where the engines are located, in local series like these guys run most sanctioning bodies dont allow fire wall modification
Unfortunately you do need to be careful with the expansion ratios, going too steep will lead to separation inside the duct with gives you dead recirculating areas of flow - Tim
Would it be beneficial, or possible, to relocate the radiator to the rear of the car to increase airflow to the front? I’ve seen it done on modern GM cars.
Certainly possible, there are other reasons apart from packaging that people tend to shift the radiator to the rear like helping with weight distribution and damage tolerance. For us in this case, keeping things at the front suited us best - Tim
With this style of car (something still relatively close to OEM) some air will always find its way into the engine bay and inevitably remove some heat as it moves through and on its way out. However, this isn't the dominant heat transfer method for the engine, the bulk is being rejected via the water cooling system. It's true a little bit less air will find its way to the front of the engine with this ducting setup, but the gains in efficiency (cooling divided by drag) should far outweigh any loss of cooling form the ambient air swirling around the front of the engine, fair question though! - Tim
@@ReubenHorner The problem is the fan shroud, the round fans will never fit the square hole. A shroud is needed at low speed to make sure the fan is moving air from the entire radiator, but at high speed the shroud is guiding the air through the fan instead of just passing by the fan more efficiently. Ideally flaps should be mounted beside the fans, at low airspeed the fans suck them closed, but at higher airspeed the flaps open to bypass the fans for free flow. But the complexity starts to go up. Another problem with highpower fans are that suddenly you end up with 1-2 kW fans that needs to be powered. 13V and 100Amps is only 1300Watts, not much for a high performance fan, a normal car has around 100Amps generator, so there is nothing left for the fuel and spark system.
We have fans mounted in the outlet duct, but these are only intended to be used with the car stationary. Once the car is moving on track the ducts will be far more effective than the fans are capable of keeping up with! - Tim
Why dont you just make the outlet narrower (in the transverse direction of the car) so that you clear the intake? I would imagine that it's OK as long as the outlet is the same area.
For sure this is a possibility which we thought about. In the end there was a target cross-sectional area we wanted for the outlet and the space back to the engine sets the depth of the outlet for us. That then defines the outlet width and not allowing the inlet tube to pass through the duct would have meant the duct would have been a long way off center. It's certainly a compromise, both in terms of adding complexity to the duct manufacture and potential separation at the duct outlet. This is simply our choice of compromises - Tim
Packaging is always a challenge, is was no different for us in this setup. The rule of thumb for the length of the inlet duct is for it to be at least equal to the height of the heat exchanger. So if your heat exchanger is 300mm high, you would aim for an inlet duct length of at least 300mm. For us this wasn't possible without the duct length getting ridiculous, as it is it already extends forward of the air dam. If moving the intercooler backwards is not an option, then you just have to make the inlet as long as you can get away with. Make the transitions as gradual as possible and avoid any sharp corners or steps - Tim
In theory you can exhaust the ducting anywhere with relatively low pressure. For our case with a circuit car and especially when you're doing endurance racing like this car is designed for, the brakes need a good supply of cold air so feeding them with air that's already been warmed by the intercooler and radiator wouldn't be a good option - Tim
There are different types of pressure, and the type he is referring to is static pressure (note this does not mean the flow is static, but you can think of it as internal pressure). The higher the static pressure, the more dense the fluid.
I'm with you on that one. The higher STATIC pressure doesn't mean that the air changes density. Normally the assumption for gasses and liquids flowing at low Mach numbers (low speed) is that they are incompressible and the simplified Bernoulli's equation assumes there is a constant density througout the flow. Not that this theoretical technicallity will change any of the performance of the setup.
To a negligible extent. Its important that they are suction fans, i mean behind radiator. Some ppl are using blowing fans, in front of radiators - this might block the air flow.
@@jareknowak8712 After a quick google I came across this article: www.verus-engineering.com/blog/informative-8/post/radiator-fans-in-depth-explanation-and-information-32 In here it says that after 20-30mph airspeed and pressure alone are enough to outflow and electric fan's capabilities. To me that says they are only really necessary in road cars, and are sub optimal and an added complexity in a race car.
@@flyingfox09 Decent rad fans work above 20-30 mph, not sure where they got that number. But generally they're not going to do much above say 40-50 mph on most setups. If you have a thick radiator core with a high fin density (lots of aftermarket radiators fall into this category), the fans help out at higher speeds. I did testing to show they helped on my S13 with an SR20VET on track (stock Altima fans on high) with a non-ducted hood (about 5-10 deg F lower coolant temps). Once I ducted the hood they did lower coolant temps a tad, but not much, just a few deg F, so like 1-2 C.
@@jacobfannon4295 Interesting. I think hood ducting is probably a big factor. So you did testing with and without fans? I assume also on a track or at highway speed?
so when you have a volume of gas, and then you expand it, you end up with a lower pressure, think of it like how a ram air system works, you start with a bigger inlet, and compress it down to a smaller outlet, this increases the speed and density of the air, effectivly boosting the engine (like on a road bike). If you reverse this process and expand the air, your going to slow it down and reduce the pressure. So you should make your inlet as large as possible to catch as much air as possible, the rest comes down to packaging. (im not saying his setup isnt going to work, just his assumptions about fluid dynamics is wrong)
Here, I think you're getting confused between static and dynamic pressure. When the velocity of a fluid decreases, it's true that it's dynamic pressure also reduces. The dynamic pressure can be thought of as the kinetic energy in the flow. When the velocity of a fluid decreases, the static pressure increases. The reason this is true (at least in an idealized case) is due to conservation of energy. To put it casually, the total energy within an idealized flow is constant. So if the flow goes through an expansion, the velocity drops, this means the static pressure must increase for the energy in that flow to stay constant. Dynamic pressure is what you feel when you put your hand out the window with a car driving at high speed, you feel a high pressure, this is the kinetic energy of the air. It is the static pressure we are interested in manipulating with this style of ducting - Tim
That's how prototyping goes .Once you find what works best then you make it out of permanent materials. I suppose those prototype cars out of clay or foam are silly as well? Please lol
Here's the thing - Everyone has their own opinion on what is the best engine in the world and that's fine. What you obviously don't understand however is the reasons for our decision. First of all a K24 would be absolutely useless to us since it would put us in a different class so it's not even an option. 'Use a K20 instead' I hear you say? Yup, could have done that and granted it's a great engine - Perhaps one of the best N/A 4 cylinder engines I've tuned. Problem is we didn't have a K20 and couldn't have sourced, built, installed and wired one in 4 weeks so again, simply not an option. On the other hand we did have a complete built SR20VE sitting in our 350Z. Next let's address this 'shitbox' SR you speak of. While I wouldn't have used the term 'shitbox', I also don't have a lot of love for the run of the mill SR20DE/DET. The head flows poorly and the valve train is a disaster. This however is NOT an SR20DE/DET as it uses the VVL head from the P11 Primera. The P11 head flows exceptionally well (probably pretty similar to a K20) and uses Nissan's take on Honda's Vtec with high and low rpm cam lobes. It also uses a rock solid shaft-mounted rocker system fixing all the woes of the DE/DET heads. Ultimately there's not many 2.0 litre 4 cylinder engines I've tuned that can produce 600 whp on 24 psi and have a power band from 4500-9000 rpm - Hard to refer to that as a shitbox in my humble opinion but perhaps that's just me - Andre
![Engine Break-In, 9,000RPM Limit | 650HP SR20VET Swap | SR86 EP 5 [#BUILD]](http://i.ytimg.com/vi/w8dLpNSmOi8/mqdefault.jpg)
![Engine Break-In, 9,000RPM Limit | 650HP SR20VET Swap | SR86 EP 5 [#BUILD]](/img/tr.png)
![Sensor Solutions and 3D Scanning | 650HP SR20VET Swap | SR86 EP 4 [#BUILD]](/img/n.gif)
I think I'll use CAD (Cardboard Aided Design) to create an Intake box like you did for your duct.
There's certainly a time and place for old-school CAD, great for mockups - Tim
I'm thought this was for the intercooler I don't understand this I m with this guy
as an 86 owner and also an ex sr20 guy this is the coolest build I've ever seen. Literally and figuratively!
I've been tryin to talk a buddy into droppin my sr20vet in his brz for over a year... Now that I'm putting it in the b14, amd after seeing this series, he wants it 😆
Me too!!!!
Sr20ve 20V is the best sr engine
This series is some of the most informative content being put out on TH-cam at the moment. Great work everyone.
Finally someone doing a video on air baffling for flow
Finally..have been wondering how you're gonna do the ducts...
Let me watch it
#notificationgang
So nice to see a duct created correctly. See so many people making ducts that go from bigger to smaller thinking that it is going to work better. The smart guys figured this out decades ago...
Exactly.....even worse when a massive intercooler is sitting out in front on it's own, with no bumper😑
2:00 I never even thought about that. Incredible.
Perfect timing for this video for me. You guys are rad, keep it up 👍🏽👍🏽👍🏽
Love the fan work, this kind of stuff take forever.
Minor technical nits, you want the highest mass flow possible through the heat exchanger for maximum cooling. The problem is a duct you put in front of hx is likely going to stall (flow.separation) at some range of Reynolds numbers, this makes mass flow.really drop.
To help keep it from stalling, put at least a 10-20 mm radius on the inlet as you won't reattach in a duct like that if you get inlet separation (look at F1 duct inlets).
@defsr7 Help me wrap my mind around something: if you want more mass flow through the heat exchanger why is it beneffitial to have a smaller inlet? I'm not saying they should go the opposite way. I trust this is proven experimentally and theoretically. But I don't understand what would happen if your inlet was 100% or120% of the frontal area of the heat exchanger? Will the air "spill" around the inlet? And what determins the "optimum" ratio of inlet to heat exchange area? Is it a function of the drag induced by the heat exchanger? A function of the speed? All of the above?
@@TrackDaysMX The smaller inlet helps expand the air, so it does indeed convert more of the total pressure (which is Ptotal = Pdynamic (the moving part) + Pstatic (what pressure you see referenced 90 degrees to its streamline) to static pressure. Essentially you slow flow down from a high speed, small inlet at some given Ptotal, to hopefully the same Ptotal (not possible, there are *always* losses due to friction/separation to various degrees) but at a higher static pressure and lower dynamic pressure. This Pstatic is what you really need to drive flow through the radiator core.
Why does a giant inlet of say 1.2x the radiator area not flow more? Well, the radiator exit probably has a pretty high Pstatic (slow moving flow there) even if it's evacuating to a fairly low pressure area, so it needs to build up some Pstatic to actually flow across the radiator core. A giant 1.2x radiator face duct inlet is going to be seen as incoming flow as a giant "brick wall" of almost no airflow, so most the air hitting this giant brick wall of a duct will generally spill out and around it. This is going to create a ton of drag, as you're pulling this giant area of generally high Pstatic along and doing nothing but forcing a ton of air out of the way in a generally turbulent/uncontrolled fashion.
A well sized duct will expand the air just enough to get the Pstatic high enough to flow at an optimal amount depending on your radiator fin density/exit Pstatic etc. Expand it more than this and you've throttled the inlet down and you just can't get enough flow into it (Ptotal is fixed in airstream). Go above this perfectly expanded inlet size and you do tend to just stall flow in the duct as it chaotically and not uniformly slows down. It also tends to be highly unsteady, which creates tons of vortices and buffeting (more of that drag thing).
Here's a good paper going on the basics of duct design. I wouldn't put too much faith in this undergrad level of CFD, but they get the general idea right, and there's a good graph showing that for their application, an inlet area ratio of about 0.8 gave them peak mass flow through the radiator: openscholarship.wustl.edu/cgi/viewcontent.cgi?article=1101&context=mems500
I imagine if they looked at what is happening after the HXer in a bit more detail, they'd probably trim that down to a 0.6-0.7 area ratio for peak mass flow.
You're right, some large radii would definitely be a good addition in future. Like many things when you're in a hurry like we've been, you've got to draw the line somewhere in order to speed up manufacture time - Tim
Very inspirational guys!
looks awesome! carboard CAD is my favorite method! 😁
Awesome man!! I built exactly this for my racecar, and not only did it provide exceptional cooling - it produced as much approximate downforce as my front splitter. Once I redesigned the splitter with tunnels - front downforce was insane. Look forward to seeing this build in person.
One thing doesnt make too much sense, Tim-sleepy-guy mentions at about 2:00 that slowing the flow means the air molecules will be more time in contact with the radiator, thus absorbing more heat, and that this is beneficial -> actually, you want the air to get out of there asap, thus the higher the velocity, the better, because you have a bigger airflow and thus more cooling capacity, everything else being equal.
Ya exactly. I don’t know why they think creating a bottleneck at the intake is going to do anything besides limit your cooling. They are just limiting the degrees of angle that the surface of the first heat exchanger can pull in air. If the electric fans are properly shrouded and the radiator and inter-cooler are properly sealed together that intake bit is useless besides a flow restrictor.
@@dsauce8780 the whole point of the bottleneck is to expand the air and increase air pressure on the radiator front face.
You are wrong in your assumption that they are restricting flow through the radiator with their duct, especially at high speeds: the air will always follow the least resistance path, and for sure that path is not through the radiator, that's why they need to create a high pressure area and seal it against the radiator
his only mistake was to assume that there will be more heat exchange because the flow is slower.
Filthy uh I don’t follow you. That is most definitely a restriction at any speed considering they have a properly shrouded electric fan pulling air and idk how you can argue against that.... but I’ll trust you know what your talking about for what ever application restricting airflow increases cooling in lol.
If they are concerned about sucking hot air at idle, evacuating hot air under power, or the aero being soooo bad that the radiator is starved then that’s a clear reason to add a bottle neck somewhere but seeing the project that doesn’t seem to be the case.
@@dsauce8780 well, you are either trolling or never heard about the mad lads Bernoulli, Reynolds, Prandtl and Nusselt.
Your comments show you don't even have a basic knowledge about fluid dynamics.
Here's a paper that proves a bigger inlet is not always more efficient, look at table 2, page 8: jestec.taylors.edu.my/vol%206%20issue%201%20february%2011/vol_6_1__094_108_charitha%20ds.pdf
Filthy wow I guess I’m an uneducated troll? Did you even read that paper? Are you trolling? The conclusion is that the largest throughput with the least restriction produces the most cooling. The whole paper is examining the air throughput based on the engine’s cooling requirements and the drag effects they produce in order for them to understand the acceptable trade off between cooling and drag for their application. The column in the chimney or inlet does not need a bottle neck, compression, or expansion to increase cooling. Just needs to be non turbulent. Sounds like you are asserting a bottleneck or restriction for pressure differential = laminar flow which is completely incorrect. The largest possible inlet with the least possible restriction(restrictions including turbulence) is the most effective. You cannot get around that. The main idea being reduce all sharp compression to avoid the vacuum that inevitably comes with it accompanied by air rolling and creating turbulence. Go look at Steve Morris’s pro-volute centrifugal super or turbo intake piece. Entirely built to create uniform flow upon entry to the compressor avoiding unnecessary turbulence and therefore restriction. A Cardboard cutout with reinforced walls with no volume change would do more than the inlet in the video.
You sound super educated tho. Even liked your comment for typing Bernoulli lol because idk anything about thermo er I mean fluid dynamics. Just remember low velocity is the problem here(your reference, 3.3 grills and chimneys). Look up gradient optimization for more info. Doubt ya will. Based off the tone of your last comment you have already been educated.
Fascinating viewing
Great info, thanks for the easy to understand explanation👍
Great Work Ande
Good that there is so much free space bt radiators and engine, which in fact is quite strange, remembering that this was originally place for a short length boxer!
Engine was mounted as far back as possible without having to cut the firewall, which certainly helped. Still not as much space as we would have liked! - Tim
Cool video.
Do you know how to use a simple water filled manometer to check the pressure gradients through, and across the ducting and matrix, respectively? Might be useful for the fine tuning, especially if using a Gurney flap on the top edge of the exit?
We have actually considered measuring the pressure at different points through both ducts at some point in future, would certainly be interesting the check the effectiveness. The outlet on the bonnet does have a gurney-like feature on the front to help create a low-pressure area to help draw the air through both ducts. You'll be able to see it in one of the next updates, good suggestions! - Tim
Logical, sound theories there guys.
If only the manifolding sat the turbo further out.
Never ending with saloon cars, single seaters are easyer to work with.
Worth checking that the exit duckting isn't too far back on the bonnet. I.E the air over the top does not reattach itself after the low pressure area behind the bonetts leading edge.
Hope all is more or less on schedual 👌
The outlet ended up being quite close to the front of the bonnet, so yes, not too far back! - Tim
@@hpa101 Sorry to learn there were overheating issues in testing and the first event.
If you read this, is would be worth checking there is a head of water above the top of the engine.
Looking forward to future episodes.
good stuff HPA 💪
Awesome video
Should add vertical fins in the inlet scoop to reduce turbulence when the air hits the inter cooler.
It's true using strakes in ducting can help keep the flows cleaner, but if we were going to use these anywhere in this setup it would be in the exhaust as that's where things tend to the be the most turbulent. Good call though - Tim
I'd split the duct exit in middle in a Y and run it out the cowl around engine for a longer travel and less back pressure to the flow path.
For sure can be a good way to do things. For us that wasn't an option with both manifolds and turbo taking up so much space at the top/rear of the engine bay - Tim
Great content thanks
Can this same concept apply to ducting it out the bottom of the car? Heat would theoretically create a high pressure zone under the car causing less max grip? My car is primarily a street car and don't plan to be pushing limits on the track for anything. 91 300zx
Thanks for reminding me about air flow out. I literally forgot about it I did directing air in but not out. Where's my cardboard cheap and easy to work with. Has there been any don't on installing dry sump oil systems like conversating from wet to dry?
Glad you found it helpful, there's definitely a time and place for cardboard and tape! Here is a recent video we did on the dry sump system on this car if you're interested - Tim th-cam.com/video/m8ofSWgtWKQ/w-d-xo.html
@@hpa101 that video was very helpful and lost of information.
I bet strategically biased dimpling in the ducting can help manipulate the direction of flow
And people run with NO bumper at all 😂
As a guide line, what should we be aiming for in terms of surface area ratio % of intake, heat exchanger and outlet.
Should it be something like 40:100:50?
I’m curious what the reason for not making the bonnet exit narrower (side to side)but deeper (front to back - to maintain your 40% ratio with the radiator) to avoid having to do all the extra scolloping and shaping to clear the intake?
I’d have thought just as you’re squishing the air top to bottom you could apply the same principles squishing it side to side?
Not sure there was enough room to make the exit larger front to back with the rocker cover in the way? Not sure if making it larger towards the front of car would muck up the gradient of the exit and how that would affect it though
Looks to me they've already maxed the front to back area, unless they cut more vents into the bonnet.
Main reason is packaging, there's simply no more room with the engine in the way. But another thing to consider is that in general you don't want to go too far back on the bonnet (especially in the middle) as you tend to get a high pressure area in front of the windscreen that will degrade the performance of the duct - Tim
Great content guys! Would an inlet still be beneficial if an outlet isn’t achievable due to engine and intake configuration (k swapped Integra) We have bonnet venting already but ducting the rear isn’t viable
With this style of duct, for sure the outlet is just as important as the inlet, they're designed to work together. Packaging is always a challenge with this stuff, there are certainly plenty of compromises in this setup of ours as well simply due the space restrictions! In saying that, if there's simply no room then yes it is probably still worthwhile to have the inlet only. For the outlet, just do the best you can to try direct the air exhausting from the heat exchanger to the lowest air pressure area you can, good luck! - Tim
Hi guys, great work!! Is the ducting in the engine bay sealed or is it drawing from the bay also? Just trying to get my head around the physics involved.
Thanks! The inlet duct is sealed to the intercooler, the outlet is sealed to the radiator and the intercooler and radiator are sealed to each other. The idea is to keep the flow through both heat exchangers as separate as possible from anything else like the engine bay. This style of system only works properly when you can control the airflow - Tim
One turning vane in the top of the front to get air into that corner.
Thanks HPA. Cool video. I recently made something similar to this for my Mazda RX7 racecar. Due to class rules, I have to run a standard bonnet with no vents. The back of my radiator has no ducting, but the front is very similar to your 86. The oil cooler sits in front of the radiator and the exit ducting goes underneath the radiator. My question is: will I decrease water temperate by creating a duct behind my radiator to increase the exit speed? And, if I am forced to duct that exit air towards the ground, will that give a negative measurable effect on aero? All help appreciated.
Certainly worth putting the effort into an outlet duct, it's just as important as the inlet to maximise the cooling and efficiency. If you get the outlet duct right, you will definitely end up with more cooling capacity for less drag. Ideally try and exhaust the air in a place of relatively low pressure and high speed flow to help draw the air through the outlet duct. Fundamentally, flow is driven by a pressure difference, so you should try to put your duct exhaust in a place with low pressure. Finding a place to fit everything and rules will almost always mean you have to compromise on this, you just have to do the best with what you have. The effect on your aero will depend on what sort of aero elements you have on the underbody, just remember to aim for low pressure! - Tim
@@hpa101 thank you so much, lots to think about.
Could you possibly use fins inside where the angle gets steeper to try and keep the air connected to the surface rather than detaching and causing stalled air?
What is sleep?
What are the benefits of bringing the outlet over the top of the vehicle instead of under? Or if space permitted, out the fenders?
More downforce? Faster air over the car acts like a venturi pulling hot air out of the duct?
The two main things to consider are packaging and aerodynamics of the outlet areas you have available. Most often, unless you are building an entire car from scratch there limited options of the places you can actually fit the ducting. Once you know the possible places you can actually fit it, you need to look at the likely pressure and velocity of the potential outlet areas. To make the outlet as effective as possible you want to exhaust the duct into a low pressure area, this helps "drive" more flow through the whole system. For us, the bonnet was a sensible place to package the duct. With the outlet (which isn't shown in this video but you'll see in later ones) we use there is a small gurney-like feature on the front of the outlet. This works by taking the high velocity clean air at the front of the bonnet and disturbing it, which creates a low pressure area in the duct outlet which helps drive the flow out through the duct. This arrangement gives us more cooling for less drag, therefore giving us a higher efficiency - Tim
Why not make the outlet more square to get horizontal clearance? The flow out of the radiator has to deflect up pretty aggressively regardless so I wouldn't think the change in profile would change the air flow to significantly, particularly compared to kinking the inlet duct and maybe making a turbulent section beyond the kink.
For our case, we really maximized the amount of space we had available in front of the engine. Packaging proper ducting in cars like this always has massive compromises, this example is no different. We were trying to balance the outlet cross-section area we needed with the space in front of the engine and trying to make the outlet angles as gentle as possible, this pretty much defines the aspect ratio for you. We have to work with what we've got! - Tim
How do you calculate the sizes of the radiator and intercooler?
Nice video as always.
Sizing the heat exchangers is best done together with the help of the company that builds them. This is because they are (hopefully!) the ones that have the heat rejection data for their products. In the absence of any proper data - which let's be honest can be difficult to come by unless you're working with a high end product, it's usually just experience and testing. We have a fair bit of instrumentation on this setup which we can use to understand the effectiveness of both heat exchangers in this car which is helpful - Tim
@@hpa101 Hi, thanks. Im building a 550hp Porsche 951, where i need to get a new cooling package. i can fabricate endtanks and other parts, but to calculate the cores, i need help.
Thanks. - Erik.
@@ealoken If you are going to build your own radiator then the best bet is to buy a core from a reputable manufacturer. Any reputable core supplier should be able to supply heat rejection information for their cores. Using some generic efficiency and heat rejection assumptions for an internal combustion engine, for a given power output they will be able to help you to size it for your application - Tim
Is an outlet duct really necessary? I the ducting is more pressure behind the radiator than if it were simply open to the engine bay, and then the hood vents could help heat escape from the engine compartment. Turbulence is the only con I can think of.
Is there a benefit on going a v mount setup as opposed to mounting the inter cooler and radiator inline?
Packaging, if you look at the jgtc supra, it has a 3s, but its pushed way back into the fire wall, that allowed them to go v mount. V mounts work well with fd3s rx7s too. So it all comes down to packaging. Those cars work because of where the engines are located, in local series like these guys run most sanctioning bodies dont allow fire wall modification
I mean even if you have a bigger curvature, won't at least some of the air cling to it because of the coanda effect.
Unfortunately you do need to be careful with the expansion ratios, going too steep will lead to separation inside the duct with gives you dead recirculating areas of flow - Tim
Would it be beneficial, or possible, to relocate the radiator to the rear of the car to increase airflow to the front? I’ve seen it done on modern GM cars.
Certainly possible, there are other reasons apart from packaging that people tend to shift the radiator to the rear like helping with weight distribution and damage tolerance. For us in this case, keeping things at the front suited us best - Tim
Interesting
First think I do is separate the intercooler from the radiator and a/c ..
What the trade off with airflow over the engine itself, or is it just not really something to worry about with such an efficient cooling set up?
With this style of car (something still relatively close to OEM) some air will always find its way into the engine bay and inevitably remove some heat as it moves through and on its way out. However, this isn't the dominant heat transfer method for the engine, the bulk is being rejected via the water cooling system. It's true a little bit less air will find its way to the front of the engine with this ducting setup, but the gains in efficiency (cooling divided by drag) should far outweigh any loss of cooling form the ambient air swirling around the front of the engine, fair question though! - Tim
What if I have hardly no room on either side to get air in or out I have maybe 4 inches on both sides any suggestions
Wish it was possible to have two more fans in the bonnet to help pull the air out.
This will not help at the speeds that the car will run, i mean above 50kmh.
It might only help when the car is standing still.
@@jareknowak8712 maybe that's true with standard oem stuff but you can get some insane brushless fans nowdays that have crazy airspeed
@@ReubenHorner The problem is the fan shroud, the round fans will never fit the square hole.
A shroud is needed at low speed to make sure the fan is moving air from the entire radiator, but at high speed the shroud is guiding the air through the fan instead of just passing by the fan more efficiently.
Ideally flaps should be mounted beside the fans, at low airspeed the fans suck them closed, but at higher airspeed the flaps open to bypass the fans for free flow.
But the complexity starts to go up.
Another problem with highpower fans are that suddenly you end up with 1-2 kW fans that needs to be powered.
13V and 100Amps is only 1300Watts, not much for a high performance fan, a normal car has around 100Amps generator, so there is nothing left for the fuel and spark system.
We have fans mounted in the outlet duct, but these are only intended to be used with the car stationary. Once the car is moving on track the ducts will be far more effective than the fans are capable of keeping up with! - Tim
Is it better to have the heat exchangers stacked like this or is it better to have a air gap between them??
Why dont you just make the outlet narrower (in the transverse direction of the car) so that you clear the intake? I would imagine that it's OK as long as the outlet is the same area.
For sure this is a possibility which we thought about. In the end there was a target cross-sectional area we wanted for the outlet and the space back to the engine sets the depth of the outlet for us. That then defines the outlet width and not allowing the inlet tube to pass through the duct would have meant the duct would have been a long way off center. It's certainly a compromise, both in terms of adding complexity to the duct manufacture and potential separation at the duct outlet. This is simply our choice of compromises - Tim
how would something like this be implemented on a car with almost no space between the front bumper and intercooler?
Packaging is always a challenge, is was no different for us in this setup. The rule of thumb for the length of the inlet duct is for it to be at least equal to the height of the heat exchanger. So if your heat exchanger is 300mm high, you would aim for an inlet duct length of at least 300mm. For us this wasn't possible without the duct length getting ridiculous, as it is it already extends forward of the air dam. If moving the intercooler backwards is not an option, then you just have to make the inlet as long as you can get away with. Make the transitions as gradual as possible and avoid any sharp corners or steps - Tim
Water to air IC might be better.
Rather than vent out the bonnet, split the air and vent into the front wheel wells to pass air over the brakes.
In theory you can exhaust the ducting anywhere with relatively low pressure. For our case with a circuit car and especially when you're doing endurance racing like this car is designed for, the brakes need a good supply of cold air so feeding them with air that's already been warmed by the intercooler and radiator wouldn't be a good option - Tim
8:00 maybe I’m wrong but pressure isn’t density. + Remove one fan from the heat exchanger for extra efficiency because Racecar
The more gas particles the more air density. pressure meaning more air in the area = air density.
At the same air temperatures, you are in fact wrong.
There are different types of pressure, and the type he is referring to is static pressure (note this does not mean the flow is static, but you can think of it as internal pressure). The higher the static pressure, the more dense the fluid.
I'm with you on that one. The higher STATIC pressure doesn't mean that the air changes density. Normally the assumption for gasses and liquids flowing at low Mach numbers (low speed) is that they are incompressible and the simplified Bernoulli's equation assumes there is a constant density througout the flow. Not that this theoretical technicallity will change any of the performance of the setup.
The "paper" like material you used to make the mold how its called ?
This mockup was done with carboard and tape. We ended up making composite ducts and the mold for these was made out of foam - Tim
How will the fans effect flow? Will they act as blockage at higher speeds?
To a negligible extent.
Its important that they are suction fans, i mean behind radiator.
Some ppl are using blowing fans, in front of radiators - this might block the air flow.
@@jareknowak8712 After a quick google I came across this article: www.verus-engineering.com/blog/informative-8/post/radiator-fans-in-depth-explanation-and-information-32
In here it says that after 20-30mph airspeed and pressure alone are enough to outflow and electric fan's capabilities. To me that says they are only really necessary in road cars, and are sub optimal and an added complexity in a race car.
@@flyingfox09 thats true.
Above ~50kmh fans are not needed.
Thats why F1 cars do not use them.
Look at how big the fans are on dyno!
@@flyingfox09 Decent rad fans work above 20-30 mph, not sure where they got that number. But generally they're not going to do much above say 40-50 mph on most setups. If you have a thick radiator core with a high fin density (lots of aftermarket radiators fall into this category), the fans help out at higher speeds. I did testing to show they helped on my S13 with an SR20VET on track (stock Altima fans on high) with a non-ducted hood (about 5-10 deg F lower coolant temps). Once I ducted the hood they did lower coolant temps a tad, but not much, just a few deg F, so like 1-2 C.
@@jacobfannon4295 Interesting. I think hood ducting is probably a big factor. So you did testing with and without fans? I assume also on a track or at highway speed?
🙉😍
so when you have a volume of gas, and then you expand it, you end up with a lower pressure, think of it like how a ram air system works, you start with a bigger inlet, and compress it down to a smaller outlet, this increases the speed and density of the air, effectivly boosting the engine (like on a road bike). If you reverse this process and expand the air, your going to slow it down and reduce the pressure. So you should make your inlet as large as possible to catch as much air as possible, the rest comes down to packaging. (im not saying his setup isnt going to work, just his assumptions about fluid dynamics is wrong)
Here, I think you're getting confused between static and dynamic pressure. When the velocity of a fluid decreases, it's true that it's dynamic pressure also reduces. The dynamic pressure can be thought of as the kinetic energy in the flow. When the velocity of a fluid decreases, the static pressure increases. The reason this is true (at least in an idealized case) is due to conservation of energy. To put it casually, the total energy within an idealized flow is constant. So if the flow goes through an expansion, the velocity drops, this means the static pressure must increase for the energy in that flow to stay constant. Dynamic pressure is what you feel when you put your hand out the window with a car driving at high speed, you feel a high pressure, this is the kinetic energy of the air. It is the static pressure we are interested in manipulating with this style of ducting - Tim
Really bro, you had to use cardboard to make the ducting? I couldn't watch any further..
That's how prototyping goes .Once you find what works best then you make it out of permanent materials. I suppose those prototype cars out of clay or foam are silly as well? Please lol
You should have K24 swapped it instead. Far superior engine to the old.shitbox sr
Here's the thing - Everyone has their own opinion on what is the best engine in the world and that's fine. What you obviously don't understand however is the reasons for our decision. First of all a K24 would be absolutely useless to us since it would put us in a different class so it's not even an option. 'Use a K20 instead' I hear you say? Yup, could have done that and granted it's a great engine - Perhaps one of the best N/A 4 cylinder engines I've tuned. Problem is we didn't have a K20 and couldn't have sourced, built, installed and wired one in 4 weeks so again, simply not an option. On the other hand we did have a complete built SR20VE sitting in our 350Z.
Next let's address this 'shitbox' SR you speak of. While I wouldn't have used the term 'shitbox', I also don't have a lot of love for the run of the mill SR20DE/DET. The head flows poorly and the valve train is a disaster. This however is NOT an SR20DE/DET as it uses the VVL head from the P11 Primera. The P11 head flows exceptionally well (probably pretty similar to a K20) and uses Nissan's take on Honda's Vtec with high and low rpm cam lobes. It also uses a rock solid shaft-mounted rocker system fixing all the woes of the DE/DET heads. Ultimately there's not many 2.0 litre 4 cylinder engines I've tuned that can produce 600 whp on 24 psi and have a power band from 4500-9000 rpm - Hard to refer to that as a shitbox in my humble opinion but perhaps that's just me - Andre