Need more miles on a tank to average out the change in mpg. Even multiple fill ups with adding miles then divide. I see people who top off a tank at a station after a 30 mile drive and extrapolate that mpg to the moon. Use multiple tanks or you’ll get inaccurate results.
very nice advancement:) i suggest you look up the audi A2 concept diesel : That car had rediculous mileage using more / other methods you could apply too.
12:20 Funny you mention the aerodynamics of that Mercedes, it's the EQS which is their first all-electric model. To help maximize the range they designed it to have the best coefficient of drag of any production car on the market at 0.20, so your callout was spot on
Well yeah for a full electric vehicle that is already twice as heavy compared to a ice version the aerodynamic is important to maximize the range of the city vehicles otherwise it will end up like a brick but then comes the BEV pick-ups like : Rivian R1T, GMC Hummer EV that somehow do not care too much about efficiency more oriented for the utility & practicality on road and off the road + other adventure capabilities aspects..
@@SMGJohn EQS is not cheap at all.. inside is quite luxurious and the price starts above 95k-100k.. the driving for what purpose is intended (definitely is not a track car) mostly designed to drive smooth and quiet from busy airports to cities hotels and it does feels good for that so yeah.. not sure what you were pointing ..
By sealing the underfloor, you reduced drag from turbulence but also created downforce (as you said it felt more planted) and thus more rolling resistance.
There will not be enough downforce to significantly change the rolling resistance. Also, RR is a very small portion of overall drag, so it's not relevant anyways. @@dinanbimmertv1864
You should do coast-down tests. It is a very accurate and repeatable indication of drag. Certainly will be more accurate than depending on the vagaries of a fuel pump and the fueling/venting system on a car. Heck, the reason the gas pump clicks off on a modern car is more about the EPA-mandated evaporative control system on the tank and less about the fuel level in the tank. I bet you'd always get different fuel amounts pumped in just as a matter of course, even with all else held equal.
I'd like to suggest an even better method of coast down testing - Julian Edgar did this video on "throttle stop" coadt down testing : ua-cam.com/video/tVAokIdaXm0/v-deo.html
@@O-cDxA i'd like to double monkey boy's comment, the throttle-stop speed testing of drag method discussed by Julian Edgar in the video linked seems like a good idea and relatively simple to do.
@@artokiiskinen1058 Well on a modern car, that will overflow the venting system and can lead to undesirable effects. And I don't trust the accuracy for this type of calculation either way.
by increasing the length of the tail you increased surface area which means the air is attached to the body for a longer period of time. You would need to probably find a balance between the first tail and the second one to find your optimal length to curve ratio; i.e. find how much you can curve the tail before the air separates and that is a good as you can make it
Your experiments are done at a non controlled environment. This implies that changes in the wind speed , different temperatures, non exact driving or speeds can affect and alter the final results, being what we could say unexpected results. Very good on your part to try this out! It's the first time I see one of your videos and was fantastic! Keep it up :)
@@Itr-tv7kt Car bodies are shaped like a crude airfoil. Air underneath travels slower than air over the top, creating lift via a pressure differential. The less air underneath the body, the less lift. Negative lift is known as "downforce". Does his mod produce NET downforce? Almost certainly not. But "less lift" is mathematically equivalent to "more downforce", and will be experienced by the driver as more traction and roadfeel.
You can still make the underside more aerodynamic just a matter of what you’re comfortable with. Some Miata race cars use ABS plastic sheets underneath to really smooth things out underneath. Off road race truck teams with bigger budgets take that into consideration and throw a giant metal skid underneath the cab. Also 15 degree departure angle for the bottom if I remember correctly with diffusers
It was just where and how to attach things that would have been a PITA. Plus a flat sheet wouldn't have done it, it would have needed to be 3 dimensional. Subaru didn't make it easy....
@@thinkflight Considering you're using such tall skirts, you could probably utilize these to your benefit of making a flat bottom out of sheet material. Just connect the front effect, skirts, and tail, with a cutout for the wheels, should be far simpler than attaching to the frame itself. The most you might have to do is find central anchor points, but considering how far offset from the body you'll be, just layer up some duct tape and make a support with some dowels epoxy, entirely removable solution since the base of it is taped down, and you can screw the sheet into the support.
Try using (if your willing) to get a external tank similar to what Mythbusters used for their test and weight before and after to get a more exact number of fuel usage
You should try using a OBD monitor to measure real-time fuel consumption. In order to maintain a consistent weight you would still have to fill up before each test, but it would eliminate any noise in the data caused an inconstant fuel station pump.
Truck driver here, the whole industry has been very focused on aerodynamic improvements for years now. Lower reaching front bumpers, cab side fairings that reach further back towards the trailer bulkhead, reducing the air gap, trailer side skirts (even on flatbeds which are terribly not aero efficient!), “flow below” devices mounted fore and aft of trailer axles, trailer tails (not cost effective tho, and prone to damage/failure because they have to fold), and fairings at trailer rear edges to smooth airflow over door hinges and marker lamps… there’s tons of stuff going on. Every little thing is like a 1-2% improvement over baseline on their own, with some hitting around 5%. Check out the Freightliner Supertruck concept vehicle.
A Bluetooth OBD reader (~$10 Amazon) and an app like torque would likely be much more accurate for fuel economy numbers. Great videos, hope you continue this series!
Great video, and 17% is an amazing result for a car thats already quite aerodynamic stock. I have a Mercedes with air supension that lowers about an inch at 75mph, ive also lowered the car another inch. Now the car is more efficient at 80mph than it is at 60mph when the supension rises. It makes a huge difference overall, even the topspeed is higher.
if you dont trust the gas station, you can always fill up some petrol cans and get a very exact amount inside your car. loved this 2 videos! here in germany we have so many rules, you will never find out something like this by your own :(
thought about this too, but since he's not driving it until it runs out of gas, he's measuring how much he puts in until the pump stops pumping, so a jerry can, even if you knew exactly how much was in it, wouldn't really inform how much gas you had used in a test.
@@maxymoo2764 because of this some test it like this: they fill it up till they can see it. It's of course not good to fill it up that much but you can compare it better
I would love to see this become a regular series! You could also look into tire rolling resistance, and as side mirror regulations are changed, the effect of replacing them with cameras. I second the idea of finding a long downhill slope and doing a rolling comparison. This way you could compare different cars, even compare gas and electric cars! Could be an annual event :-)
When Donut Media made an attempt to improve fuel economy someone mentioned video mirrors and somebody claimed they cost $4,000 per set. I asked for a source and he named semi mirrors.
@@drippingwax lmao. Talk about price gouging big corporations. A backup camera is only like 250 but starting from your own cameras and running it through a laptop and feeding the footage to small displays would be the cheap solution. Could be as low as $250, small cameras are dirt cheap
@Willham I was talking about aftermarket backup cameras. And of all the dash camera videos I’ve seen online, I really haven’t been disappointed in the video quality. They are usually at least 720p. I can’t speak on price since I haven’t shopped for them but I’d imagine there is a fair bit of competition just by the amount of truckers on the road alone
Cars and heavy goods vehicles and buses are all shipping with camera mirrors right now. Honda E car has them, Volvo and Mercedes have been using them on trucks and buses too.
@@gravemind6536Yes, because manufacturers jump at the option of replacing a reliable and cheap component with an unreliable expensive one, especially if the customer demands it because it improves efficiency and increases the perceived greenness of the product.
I would advise you too look at the viewing angles for the lamps to ensure the vehicle is still compliant. I dough it is! Further to this I know the rear overhang would be illegal in Australia as it can only be 60% of wheel base. Fuel consumption test are difficult to quantify unless you can weigh the fuel tank before and after test and ensure the comparison tests are done under the exact same conditions.
You could always buy some Flow-vis (short for Flow Visualisation) is a paint-like substance used for aerodynamic testing during practice sessions in Formula One. A high-contrast luminous colour, it is applied to an area of the car - e.g. one side of the front wing. (Or make some at home!!)
Keeping your RPMs below 1800 will help a bunch. My 2022 has manual shift mode. When I get above 50 mph, I shift to manual and put it in 7th gear. BAM!!! 42+MPG with a bone stock Impreza wagon. Even @ 70 mph
A automatic will often run at higher RPM and lower load at lower efficiency than needed if you let it for air pollution reasons, especially on diesels. that, or it just gets stuck in higher gears too long
When i flew model rockets a little short kinda flat nose cone was faster than a long curved pointed nosecone . It had less surface area. So on your trunkated put a shorter point so the flow has to break around the trunk edge and flow to the point. Instead of drag off the edge.
Just for those interested, Echoes are in mountainous canyons and very large buildings with flat walls to bounce accurately off of and hit you some time later, just like a “Delay” effect for guitars and vocals. Your garage is reverberant which is a million echos blended smoothly and is what a cathedral sounds like. Not echo. And definitely not “echoey”. : )
I would also look at your turbulence under the tail, and have a look with tufts. It may be wise to add some vertical diffuser elements to the tail underside
I don't know how much of a diffuser he could add without it scraping. Not many people modify their cars like this, but when they do, they usually compromise the bottom for usability.
Some other comments have pointed out, but those skirts are also creating downforce, a way for you to reduce your rolling resistance and induced drag could be to raise the tail a little but and vent a little bit of air behind the front wheels, this would reduce your underbody downforce reducing induced drag too. This vent could be something like a skirt with a small protusion into the air stream, it sounds weird that adding frontal area to the car can reduce the drag, but its one of those counter intuitives examples to drag more there to save else where
have you standardized vehicle weight, and tire pressure? Tire pressure can have a massive effect on efficiency. also added weight from more material at the back axle could be to blame. I would suggest adding lightness and trying to see how far you can slope the read down before you hit turbulence, as a shorter tail would also reduce weight, and improve axle balance.
The original: It scooped up from the bottom more. The top part didn't rake down so sharply and the sides 'boat tailed' in sharper converging at a point just below the taillight level. Even, fully enclosed transparent wing mirror fairings? Maybe? Love these uploads btw :)
Super interesting. I did some tests on a 3 tonne truck a few yars ago with an inflatable tail - inflated by the stagnation of the air at a duct on the top of the body. I didn't have very good metrology but thought that I had reduced the drag by about 5%. The inflatable idea was that when it was stopped the tail would deflate and have bungee cord in it that would collapse it down to a short length for parking. I had to rent the truck for the testing so I ran out of time and energy but I think it would work well. People often put stuff on the front of vehicles to aid aerodynamics - we have a thing called NoseCone in Australia but it is really at the wrong end IMHO. Avoiding the separateion of the boundary layer on the rear is waaaay more important than making the front rounded...
Coilovers, lighter wheels, and oddly a tune could help too. Looking at race cars that go for efficiency in fuel economy is cool. I also understand you aren't trying to buy the house in parts. Shoot I'd donate if you did this with something smaller. As a car enthusiast what you are doing is pretty cool.
The commentary toward is probably more spot-on that you would expect. I'm the guy that typically doesn't refuel until the reserve light comes on, and I get to know my cars fairly well as I drive them and pay attention to things like fuel dispensed. There are some gas stations I simply avoid because they'll consistently dispense almost 10% more fuel than others in the area. In cars where I know when I'm almost completely empty, just how much should be filled, going over a certain amount is pretty obvious. I had one car with a 12 gallon tank that would typically take 12-12.5 when the computers range was in the single-digit miles remaining. On gas station would routinely dispense 13.5. Another gas station never dispensed more than 12.5. Unless you have some calibratable way to measure the fuel, I would recommend either only using one pump at one station, or doing multiple runs using different stations and averaging all the results together in hopes of reducing the significance of that particular mode of variance.
Great simple experiments, well summarized. To monitor realtime data, and collect data on vehicle performance; you could invest in an On-Board Diagnostic (OBD) dongle. This would allow realtime monitoring and collection of data on a mobile device (via wifi, or bluetooth). Data parameters available varies a bit by make, model and year, but often can view realtime fuel consumption, or estimate quantity remaining, as well as odometer readings. Logging would allow export to a spreadsheet for some fun maths. BTW: did you do a calibration check on tire pressure? This an often overlooked item and can make ~5% efficiency difference. Varies with temperature and seasons, so is worth checking once a month, or prior to efficiency runs.
I hope these vids start seeing more traffic man, such good content, and you do enough work to justify it! Love your content, which i found thru Daniel's channel. Love all of your ekrano-esque creations btw, such a cool phenomena that seemingly NO ONE is taking commercial advantage of, yet, you and Daniel are making a whole series of interesting ground effect vehicles in a relatively short time frame JUST for the appreciation of physics and to 'feel' that effect! Theres so much intersting phenomena to behold, and all the micro details of each design can make or break the entire character of the flight... so cool! Keep up the awesome work buddy! Cheers from HTX
I think one of the reasons why the tail *seems* less efficient (but might not be) is the combined effect of the underbody bypass and the tail. The previous tail might have reacted better to the previous body kit, but this one might benefit more from the smoother airflow coming from the new system
Great job. 17% is huge. I would guess that you'd need to do several runs with each mod, in two directions, and tuft test each run to get a clear idea of what is actually happening.
You could use a pressure sensor to measure your air pressure in the tail section. Then pressure drag is equal to this multiplied by the area of your truncated cross section.
Aerocivic (ex)owner here, as well as owner of one of the highest fuel economy Insights on the road (100mpg+). Happy to see you continuing to experiment.
Add to front a little splitter below the airdam. 30mm further front is enough for most of the benefits in separation. (based on Nascar 2019 CFD results etc) I would also cut the rear bumper and only make cone for the remaining lowpressure zone. (or remove bumper and use good rationalizing on the cone starting height.)
Haven't you considered an _inflatable_ tailcone? One made from [clear] plastic film, able to be inflated at low pressure (maybe even by scooping air in motion?) and deflated while parked.
I thought of that a long time ago for semi trailers but didn't try it. Once on the highway, the flat back of the trailer would deploy a kind of sock that would fill up with air that slopes down to a point. When you slow down, it would automatically deflate and roll itself up out of the way. That could save a lot of fuel for trucking companies.
He says, adding a rear wheel skirt would require modifying the car so he doesn’t do it. Then he proceeds to make all kinds of modifications to the body of the car. LOL
Seeing the rear of the new boat-tail made me think of the british WW2 stern design. Apparently they realized having a flat stern instead of a slow taper was more efficient, because the turbulence swirled around and pushed the boat forward? I may be wrong about that but it's worth looking at.
That reminds me of the kammback, which found a truncated tail was more efficient like used on the Prius and Insight. Then consider the bullet shape called the boat-tail, which is essentially a kammback too, that mimics the same design for stability in flight.
Just wanted to share my mileage tests by saying when filling up the gas tank the only way I could think of to keep consistent was filling up neck so I could not put another drop in. So that worked for me, but some cars might have problems with smog systems. I was not testing pollution effects so my mileage worked just fine. I was able to get 27.6 mpg in my F150 but if doing a full tank to empty don't drink to much, it's a long drive.
I had a 2014 subaru impreza sedan cvt that i used for my 1hr commute that got a consistent 36.9 mpg, i now have a 2010 honda fit manual trans that gets 38.1 and am in the process of k24 swapping a 2011 honda crz in hopes to achieve 40 plus mpg with some tuning, this video was great to watch!!
Legal or not, it could also be that the cops who followed you decided that they just didn't want to know; "No, nope, to close to end of shift/next break, etc...."
you can check mileage with the onboard computer. drive say 80mph on a level road in a high gear. Use the same gear to compare. do several runs and note the windspeed and air temperature each time for every setup. If you get better mileage in worse or equal conditions then the better efficiency is confirmed. Onboard fuel metering has about 0.5% accuracy and can be offset if the tire size is not nominal but the tire size doesn’t affect the test obviously if all tests are done on the same tires and pressure (check pressure each time). The fuel metering can be adjusted through obd if offset but you don’t need that either since all tests are using the same calibration Note the computer mpg average over say 0.5miles. Each run on the same road preferably
Im guessing that the variance in the results from the gas pump come from the mechanism of the stop valve in the nozzle. Since its purely mechanical, it might not always fill to the same point each time. Ive noticed on my car that sometimes the amount it fills will be off by a half gallon or so when i fill the tank from the same mark. Awesome video though!
It is a back pressure pop valve. Using the same pump and nozzle is good, but the car, temp, air pressure, humidity/dewpoint is what changes along with the saturation of the vapor trap cannister. If overfilled on visit 2, then visit three may cut off as much as two liters early because the vapor cannister has become wet from last overflow.
Yo yo yo!!! - Please find a secluded hill and try out rolling in neutral. If you can’t rely on your pedal or gas pump input then just see how far your car rolls (with you in it of course) from top to bottom!!!! Your results should overhaul anything gas mileage might tell you since you know for a fact one is more air efficient
Filling in the rear of the tail surely hurt you a bit. The plate cover is a large surface for the low pressure wake to act upon. Remember that it's pressure region's acting on a moving body itself that contribute to drag! Having the tail hollow the first time around meant that realistically only the edges of the foam were subject to wake forces, with the gradient increasing to probably around atmospheric when you reach the body. With such a better optimized curve on the tail, you had way higher velocities at the bluff cutoff, so every square cm had an even more severe pressure drop to contend with vs the former. If you cant get a point (which you probably can't achieve efficiently anyways) find a way to leave her hollow!
Well, the old design created a large cavity, and there is no way to tell what the internal cavity pressure was without some further analysis. Sometimes it can result in higher pressure when you create desirable recirculation inside the cavity, but in this case I suspect recirculation was minimal. So it was probably transmitting the wake recovery pressure, or a lower pressure, up against the hatchback and bumper inside the cavity. Which means not much of a reduction in base drag. The new closed boattail shape is probably better.
I could imagine that the new tail causes more drag, because the back area, where the numberplate sits is substantialy bigger. Without directly changing its size again, you could try to angle the surface at maybe 45 degree, effectively creating more of a tip, instead of a big rear stub. Do you have something like discord for better exchange of ideas?
I have an FA24, ambient temperature makes a huge difference. In the summer when I am driving in 70-75F in the morning and 85F in the afternoon, I see ~30-31mpg. Now in the fall, morning temps around 45F and afternoon temps around 60F, I am seeing 35-36MPG. Same route, no traffic, trying to average 35mph or 50mph on the backroads mileage is always about the same over 3-4 tanks, leaving me to believe it is ambient temp, or if the additives package has changed in the gas. On another note, I had a pickup truck that I put a downward wedge cap on to do a fluid dynamics paper on (emulating a couple that were already done), that alone improved efficiency about 10-15% on a regular basis.
I think some of the efficiency losses may be because of side winds. Perhaps a shorter tail section would reduce drag in a cross wind. Also, may as well leave the back portion of the tail section off so that there's a clear view of the existing license plate holder. I think such a design is called a "partial cam back." (Not my words) it may be the side wind and driveability solution as it would be shorter and (perhaps) still have most of the efficiency gains. Very interested in this as a series. Kinda hoping for a more final version, perhaps made from fibre glass laid over an existing test piece. Also, may as well test the pointy front. I hear it's kinda bad, but have no numbers to tell exactly how bad.
Drop weight Electric fan conversion light weight pulley conversion, hot air intake instead of cold air intake. 5psi over tire pressure. Tune the ecu to run at 15.5afr instead of stock 14.7 stoic also automatic transmission oil cooler will cause the transmission oil to slip less and improve efficiency of the torque converter lockup coupler. I’ve done all of this and it all works I’m a mechanic 15 years
You may know your stuff with aerodynamics, but establishing a baseline fuel efficiency needs to have minimum of 10 runs under similar condition, wind speed and direct plays a major role in fuel efficiency. My SUV gets 32mpg, but winds weather head, tail or from the side can throw (increase or decrease) off fuel efficiency by up to 15%. After each change or theoretical improvement, need to make equal number of runs and then average them after removing anomalous results. Called the scientific method. I have logged 10k miles of fuel efficiency in my own car, using data logging, engine performance, and weather condition. So many well intentioned people come to inaccurate results, because they can't or don't consider all factors.
Before the back addition, the curves helped with the airflow. You could try angling it more towards the bottom, so only the area with the license plate is flat and nothing else, and even that could be slightly angled (something like 15 degrees or less).
This is a really cool series! Maybe one day if this makes enough you could try to find a cheap and already teardrop-ish car and try to make it insanely light and efficient. Also the Mercedes was an EQS, the new electric one that I remember seeing in videos s year or two ago, but it seems they haven't gotten very popular, so that's a cool car spot!
When I am passed by a semi on the freeway in my RV I feel pressure first against it pushing me away and then towards it sucking me in behind. Those admirers of the car that stay next to it will have an affect on air pressure in front.
@@JoeOvercoat My car is older so I only use ethanol-free, and while it does cost more, I get about 2-3 mpg using it. For consistency, he should definitely use it when testing at different times of the year.
Removing the skirting introduced more drag, than the ever-so-slight increase in efficiency of the new tail could hope to compensate for. The ground effect between the chassis of the car and the road, is ~30% of the total drag of a modern car! You were definitely on the right track, when you tried to address turbulence underneath the car. Another thing to address, (I'm sure you thought of this) is tire rolling resistance. Tire pressures make a big difference. When a car becomes this efficient, tire inflation pressure becomes a greater percentage of overall drag, in that it greatly influences rolling resistance. Ambient temperature also makes a difference. A cold tire on a cold road on a cold day has much more resistance, than the same tire at the same inflation pressure, on a hot summer day.
I might suggest using an OBD reader to get your air-fuel ratio while doing these tests, it's possible your car had richened it out for one reason or another causing more fuel consumption. Cars actively manage all of the conditions that make the car run and could be doing one thing or another to make catalytic converters heat up or cool down or a bunch of other things. I think averages from a set of tests using one variation for another would be a better indicator. Love the videos, 18% is incredible.
When flying, after polishing our small aircraft (115 Horsepower) We can expect around a 5-knot increase in it's cruise speed at 25 Liters per hour. This represents around a 4% increase in cruising speed. It would be interesting to see what would happen if you smoothed the rear material and polished the car. Also, perhaps putting the fuel units in L/100km as well as MPG might help appeal to a wider audience.
Equally interesting is the dimples on a golfball that create lift by essentially creating an air pocket around the ball. At least that’s how I think it was described as how that works or something similar. I think I recall long ago someone recreating a dimpled effect on their car which smoothed the airflow because of the “pillow” of air on the body.
The splitter is giving you the most gains. The boat tail could be reshaped a bit get the flow into the dead space, also refining the splitter will help. One thing you might to try is divert some the air from the top of the boat tail to the bottom and clean up some of the dead air.
I don't care about the fuel efficiency part But those skirts HELPED me a LOT because i race, and how simple your skirts are make me excited to try them on my racer
Fantastic! I also did DIY aerodynamic modification to my car (Kei car) and improved fuel efficient a lot. I'm so glad I found this video! currently I'm on raodtrip in the whole mainland Japan with my aerodynamic car lol
Why not go full Ricer mode, add the stickers and the sponsors and don’t forget add colors. Red means fast and Green for efficiency , don’t ask me how the color works just Ricer science.
Laminar flow past the leading edge of the vehicle is oddly enough, a source of additional drag. You aren’t trying to generate lift, you are trying to limit air contact with the body. Instead of making the car resemble a wing, try to make it look like an upside down speed boat.
What i think what this video is when clicking on the video: A guy upgrading a car with high quality materials What it actually was: a guy upgrading his car with cardboard, styrofoam and duc(k) tape You did this very good!
I know I am late to the party, but - do you know with certainty that when the pump at the gas station cuts off, it is always accurate ie. what about foaming, or more precisely, the splashing that trips the trigger? It only needs a single splash to cover the 'trip hole' in the mechanism, and you could be quite a way out... Just a thought. (I used to work on such mechanisms, and they are far from precision mechanisms. I reckon you would have to do, maybe, three runs and average them out (-:)~
Take a look at experimental aviation for fuel flow and totalizers. Plenty to choose from and very accurate. Thanks very much for your efforts in this endeavor, very interesting. Efficiency IS in the management of the air. I added disks to close off the rear dual wheels on my tow rig (second gen Dodge 3500, manual gearbox, Cummins powered RWD.) I log every tank (have sense purchase), noting if towing Ect, and I can clearly see a slight gain, .37 mpg over 40 fills. Sounds like mouse nuts, right?. AT $5 plus a gallon, for now, every little bit helps.
i agree with your last comment about the pump not being as accurate when shutting off, you can get around that by using a fuel cell instead and measure by weight
One of the things with your low tail vs high tail comparison to think about is the amount of air taken in through the grille. You had a big drop in efficiency when the splitter and skirts were gone in part because the engine bay was probably less efficiently removing air, causing it to spill into the wheel well. That's messy enough aerodynamically that many cars that are new have active air dams hidden in the grille for improved efficiency. The angle of attack on the extended rear diffuser can speed up the underbody airflow and get more of that messy, hot engine bay air out.
A simpler test that will eliminate the most errors is a down hill coasting test. Others have mentioned it. And it can be done in other forms. Dead stop from exactly the same place on the same hill under the same weather conditions and simply monitor the top speed and coasting distance. Do this 10 times with the regular car to establish average deviation. If your aero works well the top speed will increase as well as total coasting distance. I bet the car in it's stock form will fluctuate +-3% or more so an increase of 17% without knowing the average deviation is more likely 10-14% in actual improvement. That's still very good. So keep going. Make sure all things are kept equal. Tire pressure, weight in the vehicle, weather especially wind conditions. If the wind or traffic turbulence fluctuates the air +- 3% that's not helping get good data. stay away from dirt it will very the rolling resistance, stay away from testing in traffic and avoid stops and starts in the test, that puts the engine in the worst efficiency rpm range and can easily be effected by the driver, also avoid turns as they induce turning drag and are also effected by the driver. Go straight down a hill and only measure coasting top speed and distance and a more clear picture should come from that. Good work keep it up.
Great testing. You queried why you got the results you did. Please restore the car back to baseline and then record 20 consecutive tanks. That will help you understand variability. You will also see that reporting efficiency gains to 0.1% precision are ambitious.
(if he's married) Props to his wife for letting him play around with his engineering ideas. As embarrassing as this might be for "what will the neighbors think" types, it's actually pretty cool and how a lot of tech/science advances. Apple was created in a garage. This dude is in heaven working on this stuff.
For testing I would recommend measuring instant mpg, not trip mpg, and at as high of speed as possible to make the differences more apparent. If you can do 90+mph on that lake bed that would be much better. Calculating based off a trip introduces so many variables, not least of which is the driver. You might drive more cautiously with the tail, thus improving MPG.
@@aapjehoan and static wind is easily measured and you can somewhat normalize the data by taking data points going both directions. It's much easier to calculate a potential variance knowing local windspeed than trying to guess/calculate driving style impact.
Your new tail is less efficient because ot the vortices created at the edges were the upper surface and the side surface meets. This is beacause your tapering angle is larger at the top surface than at the sides, creating two vortices similar to airplane tip vortices and these cause the excess drag. You can even see these vortices even at low speed in your smoke test at 11:42. Other input is that seapation is usually just a bubble, the airflow closes up after the bubble so it is not necesserily as big of effect as one might think. (Taillight covers are not so crutial.) So even if you have some separtation but if you manage to taper the tail of the car more your wake region can actually be smaller. As for the wheel arches you do not have to fully close them you can merely make the gaps smaller.
I think it's very likely you fill up to a different point every time. You'd probably need a more accurate sensor in your fuel tank, or a transparent hose that makes the fill status very clear. I've found for my 50l fuel tank, that filling from empty till the click, then letting it really settle and filling again till the click gives me am extra 3l of fuel. I've been tracking my fuel consumption for aound 50000km, 2500l of Diesel, and sometimes for no reason my sonsuption goes up by like 5%, even though I commute the same long route every day.
To get beyond 18%, here is a suggestion. Some air will flow underneath the car even you put an air skirt in the front. Starting from the rear bumper, you can taper the underside of your tail addition. The angle of the taper is gentle. See 2000 Honda Insight or Chevy EV1.
You could use the bottom flat sections of a couple of truck bed liners to make under trays. Use rivthreads to attach them to the bottom. You also could use abs sheets. Even though most of the bottom of my '19 prius has trays, I used airtabs to jump the exhaust gaps and suspension gaps which has resulted in a 3mpg gain. Still a work in progress.
In order to fill your tank to a consistent level partially fill at the pump and then use a dip-stick and top up by hand from a can to the same level. As an alternative to the dipstick get a clear plastic tube; insert it into the part filled tank; seal the end with your thumb; partially withdraw tube now filled with fuel; have the tube describe an S shape rotated 90°; take thumb off end; now the fuel in the visible portion of the tube will show the level in the tank. This should allow consistent filling of the fuel tank.
a 0.3 gallon deviation on 4 gallons sounds like a lot, but it could be fuel temp related (filling up in the morning vs evening affects fuel volume as the tanks in the ground heat and cool leading to slightly different densities. it's why its done by weight in aviation.) or it could be down to aeration of the fuel as it's delivered. the "click" happens when the air tube is obstruced, and having a nozel that produces slightly more foam while dispensing will trigger the stop early. It's why Mythbusters always used a fuel cell, you can just put it on a scale to get the fuel weight
Forget weird aero that can potentially kill you and DO simple things: REDUCE WEIGHT. REMOVE THE ROOF BARS (that is killing your aerodinamics) GET lighter ryms, Clean your Trunk of stuff you don't use. Those are key and will not kill you.
While you eliminate one variable by using the same gas pump you still have the issue of the automatic shutoff. It will never be consistent. You have a couple of options. 1. Drain the tank completely. Not sure if there is a drain plug on the tank or not. Once it is drained put a measured amount of gas in the tank. Then run the car until the fuel is exhausted. Then put gas in the tank, start the car so until the fuel lines are primed then drain the tank again and refill with the same measured amount of gas. Repeat drive with the aerodynamic changes. Another option is to put a small tank and pump in the trunk and move the fuel lines. Either way you need to know precisely how much fuel you put in and then run until empty.
12:36 hmmmm... at first I thought the cabin side-skirts were like a single line (or two) of thick bristles (like a thin broom). Maybe that would be a good practical solution for an actual daily car, as when you drive over an obstacle (eg a high speed hump), the bristle would give way, but under normal high speed driving, they would act as a solid surface/wall, forcing the air to go around.
Head to squarespace.com/thinkflight to save 10% off your first purchase of a website or domain using code thinkfight
the mercedes was an eqs
nobody cares about advertises. you can see the most replayed part was where the advertise ends.
Need more miles on a tank to average out the change in mpg. Even multiple fill ups with adding miles then divide. I see people who top off a tank at a station after a 30 mile drive and extrapolate that mpg to the moon. Use multiple tanks or you’ll get inaccurate results.
very nice advancement:) i suggest you look up the audi A2 concept diesel : That car had rediculous mileage using more / other methods you could apply too.
ua-cam.com/video/BTPnQneEtVs/v-deo.html&ab_channel=JayEmmonCars
12:20 Funny you mention the aerodynamics of that Mercedes, it's the EQS which is their first all-electric model. To help maximize the range they designed it to have the best coefficient of drag of any production car on the market at 0.20, so your callout was spot on
Well yeah for a full electric vehicle that is already twice as heavy compared to a ice version the aerodynamic is important to maximize the range of the city vehicles otherwise it will end up like a brick but then comes the BEV pick-ups like : Rivian R1T, GMC Hummer EV that somehow do not care too much about efficiency more oriented for the utility & practicality on road and off the road + other adventure capabilities aspects..
@@poplaurentiu4148
Twice as heavy, twice as cheap to drive.
@@SMGJohn EQS is not cheap at all.. inside is quite luxurious and the price starts above 95k-100k.. the driving for what purpose is intended (definitely is not a track car) mostly designed to drive smooth and quiet from busy airports to cities hotels and it does feels good for that so yeah.. not sure what you were pointing ..
@@poplaurentiu4148
Wow, you drive for free? Interesting.
@@SMGJohn I'm not really sure i fully understand what you try to say regarding to EQS..
By sealing the underfloor, you reduced drag from turbulence but also created downforce (as you said it felt more planted) and thus more rolling resistance.
So a flat Bottom would be more beneficial.
But. Is that bad in california? It's mostly mountain driveing...??????
Not to mention when it rains it pours.. Idk
More rolling resistance, more power demands from the motor to overcome the resistance, more fuel
There will not be enough downforce to significantly change the rolling resistance. Also, RR is a very small portion of overall drag, so it's not relevant anyways. @@dinanbimmertv1864
You should do coast-down tests. It is a very accurate and repeatable indication of drag. Certainly will be more accurate than depending on the vagaries of a fuel pump and the fueling/venting system on a car. Heck, the reason the gas pump clicks off on a modern car is more about the EPA-mandated evaporative control system on the tank and less about the fuel level in the tank. I bet you'd always get different fuel amounts pumped in just as a matter of course, even with all else held equal.
I'd like to suggest an even better method of coast down testing - Julian Edgar did this video on "throttle stop" coadt down testing : ua-cam.com/video/tVAokIdaXm0/v-deo.html
Either of those two methods will be more accurate than a partial fillup.
or a fill up until the clicks stop. trickle until the tank basically overflows. should be pretty accurate too.
@@O-cDxA i'd like to double monkey boy's comment, the throttle-stop speed testing of drag method discussed by Julian Edgar in the video linked seems like a good idea and relatively simple to do.
@@artokiiskinen1058 Well on a modern car, that will overflow the venting system and can lead to undesirable effects. And I don't trust the accuracy for this type of calculation either way.
by increasing the length of the tail you increased surface area which means the air is attached to the body for a longer period of time. You would need to probably find a balance between the first tail and the second one to find your optimal length to curve ratio; i.e. find how much you can curve the tail before the air separates and that is a good as you can make it
Agreed. Look at nascar. Very important to minimize front drag from vents etc.
Maybe vortex generators could work?
Was just going to say that. More surface area = more viscous friction.
If I remember my aero class correctly 11 percent is the sharpest curve before detachment of flow
@@TheObserver567 racecars of any kind, even nascar are not good examples of aerodynamics. they want downforce for handling. pretty much the opposite.
People love these type videos. 18% improvement is almost amazing. It’s significant for certain.
Almost a fifth of tank getting saved.
Almost amazing? The things I would do for 18% improvement… without engine or tune mods
Its even more impressive that he starts with pretty efficient car whatsoever.
are you guys ever drove a car?🤔🤭
@@oxygenium92 I'm pretty sure Subarus aren't crazy efficient
Your experiments are done at a non controlled environment. This implies that changes in the wind speed , different temperatures, non exact driving or speeds can affect and alter the final results, being what we could say unexpected results. Very good on your part to try this out! It's the first time I see one of your videos and was fantastic! Keep it up :)
I love how you tried to make the car more fuel efficient and accidentally created a dirt cheap mod for more downforce on your car at the same time.
Now I can try ricing without being permanently riced out
His mods actually look better than like 90 percent of the ricer mods i see lmao
How did you come to that conclusion btw?
i dont think it's creating downforce.. its reducing drag which helps with MPG.. if you're adding downforce it would add drag = it will worsen the mpg
@@Itr-tv7kt Car bodies are shaped like a crude airfoil. Air underneath travels slower than air over the top, creating lift via a pressure differential. The less air underneath the body, the less lift. Negative lift is known as "downforce".
Does his mod produce NET downforce? Almost certainly not. But "less lift" is mathematically equivalent to "more downforce", and will be experienced by the driver as more traction and roadfeel.
You can still make the underside more aerodynamic just a matter of what you’re comfortable with. Some Miata race cars use ABS plastic sheets underneath to really smooth things out underneath. Off road race truck teams with bigger budgets take that into consideration and throw a giant metal skid underneath the cab. Also 15 degree departure angle for the bottom if I remember correctly with diffusers
I always read on the area of 12 degrees...
It was just where and how to attach things that would have been a PITA. Plus a flat sheet wouldn't have done it, it would have needed to be 3 dimensional. Subaru didn't make it easy....
@@thinkflight
Maybe build a box frame along both sides of the underside of the car so you can mount a flat sheet lower than the vehicle components.
@@thinkflight Considering you're using such tall skirts, you could probably utilize these to your benefit of making a flat bottom out of sheet material. Just connect the front effect, skirts, and tail, with a cutout for the wheels, should be far simpler than attaching to the frame itself. The most you might have to do is find central anchor points, but considering how far offset from the body you'll be, just layer up some duct tape and make a support with some dowels epoxy, entirely removable solution since the base of it is taped down, and you can screw the sheet into the support.
Try using (if your willing) to get a external tank similar to what Mythbusters used for their test and weight before and after to get a more exact number of fuel usage
i think without a large private property that probably wouldn't be street legal but that would be pretty neat
@@dogbee in the USA nobody cares what you do to your car
You should try using a OBD monitor to measure real-time fuel consumption. In order to maintain a consistent weight you would still have to fill up before each test, but it would eliminate any noise in the data caused an inconstant fuel station pump.
That Mercedes is an EQS. +1 to coast-down tests instead of gas pump variance. Rad video as always.
the merc looks like a bar of soap, although the tech is cool
@@driftmad1839 It also drives like it's on ice: terrible
@@SkylineLofe yours for only a few hundred thousand dollars
@@driftmad1839 EQS's are 100-150k USD. So not a few hundred thousand.
Truck driver here, the whole industry has been very focused on aerodynamic improvements for years now. Lower reaching front bumpers, cab side fairings that reach further back towards the trailer bulkhead, reducing the air gap, trailer side skirts (even on flatbeds which are terribly not aero efficient!), “flow below” devices mounted fore and aft of trailer axles, trailer tails (not cost effective tho, and prone to damage/failure because they have to fold), and fairings at trailer rear edges to smooth airflow over door hinges and marker lamps… there’s tons of stuff going on. Every little thing is like a 1-2% improvement over baseline on their own, with some hitting around 5%. Check out the Freightliner Supertruck concept vehicle.
Its not a fair test,
Racing stripes will put at least +100hp to your engine...
Awesome and interesting results. Thanks for the video!
Imagine if the car was red!🤯🤯
@@jasmijnarielthe skirts were red, so it goes faster at any speed.
A Bluetooth OBD reader (~$10 Amazon) and an app like torque would likely be much more accurate for fuel economy numbers. Great videos, hope you continue this series!
Great video, and 17% is an amazing result for a car thats already quite aerodynamic stock. I have a Mercedes with air supension that lowers about an inch at 75mph, ive also lowered the car another inch. Now the car is more efficient at 80mph than it is at 60mph when the supension rises. It makes a huge difference overall, even the topspeed is higher.
I do the same thing with my bikes. Round tubes are terrible in the wind. Adding some foam and tape to the trailing edges reduces drag by 10x.
Keep in mind the air that enters the engine compartment through the grill has to exit somewhere
if you dont trust the gas station, you can always fill up some petrol cans and get a very exact amount inside your car.
loved this 2 videos! here in germany we have so many rules, you will never find out something like this by your own :(
thought about this too, but since he's not driving it until it runs out of gas, he's measuring how much he puts in until the pump stops pumping, so a jerry can, even if you knew exactly how much was in it, wouldn't really inform how much gas you had used in a test.
@@maxymoo2764 because of this some test it like this: they fill it up till they can see it. It's of course not good to fill it up that much but you can compare it better
You dont need to be accurate, you need to consistent. Using the same pump is good enough
@@TheJttv i gave him just a solution because he sayed he didnt trust the gas station.
Yup das wär ein klasse Weg ein Date mit dem Richter zu bekommen
I would love to see this become a regular series! You could also look into tire rolling resistance, and as side mirror regulations are changed, the effect of replacing them with cameras. I second the idea of finding a long downhill slope and doing a rolling comparison. This way you could compare different cars, even compare gas and electric cars! Could be an annual event :-)
When Donut Media made an attempt to improve fuel economy someone mentioned video mirrors and somebody claimed they cost $4,000 per set.
I asked for a source and he named semi mirrors.
@@drippingwax lmao. Talk about price gouging big corporations. A backup camera is only like 250 but starting from your own cameras and running it through a laptop and feeding the footage to small displays would be the cheap solution. Could be as low as $250, small cameras are dirt cheap
@Willham I was talking about aftermarket backup cameras. And of all the dash camera videos I’ve seen online, I really haven’t been disappointed in the video quality. They are usually at least 720p. I can’t speak on price since I haven’t shopped for them but I’d imagine there is a fair bit of competition just by the amount of truckers on the road alone
Cars and heavy goods vehicles and buses are all shipping with camera mirrors right now. Honda E car has them, Volvo and Mercedes have been using them on trucks and buses too.
@@gravemind6536Yes, because manufacturers jump at the option of replacing a reliable and cheap component with an unreliable expensive one, especially if the customer demands it because it improves efficiency and increases the perceived greenness of the product.
I would advise you too look at the viewing angles for the lamps to ensure the vehicle is still compliant. I dough it is!
Further to this I know the rear overhang would be illegal in Australia as it can only be 60% of wheel base.
Fuel consumption test are difficult to quantify unless you can weigh the fuel tank before and after test and ensure the comparison tests are done under the exact same conditions.
You could always buy some Flow-vis (short for Flow Visualisation) is a paint-like substance used for aerodynamic testing during practice sessions in Formula One. A high-contrast luminous colour, it is applied to an area of the car - e.g. one side of the front wing. (Or make some at home!!)
Flow vis paint is just Mineral oil and fluorescent dye.
Flow vis paint is a great idea!
@@GunganWorks yeah, I noticed a few videos right now on recipes for it.
Keeping your RPMs below 1800 will help a bunch. My 2022 has manual shift mode. When I get above 50 mph, I shift to manual and put it in 7th gear. BAM!!! 42+MPG with a bone stock Impreza wagon. Even @ 70 mph
A automatic will often run at higher RPM and lower load at lower efficiency than needed if you let it for air pollution reasons, especially on diesels.
that, or it just gets stuck in higher gears too long
When i flew model rockets a little short kinda flat nose cone was faster than a long curved pointed nosecone . It had less surface area. So on your trunkated put a shorter point so the flow has to break around the trunk edge and flow to the point. Instead of drag off the edge.
Just for those interested,
Echoes are in mountainous canyons and very large buildings with flat walls to bounce accurately off of and hit you some time later, just like a “Delay” effect for guitars and vocals.
Your garage is reverberant which is a million echos blended smoothly and is what a cathedral sounds like. Not echo. And definitely not “echoey”. : )
This I did not know!
I would also look at your turbulence under the tail, and have a look with tufts. It may be wise to add some vertical diffuser elements to the tail underside
I don't know how much of a diffuser he could add without it scraping. Not many people modify their cars like this, but when they do, they usually compromise the bottom for usability.
Some other comments have pointed out, but those skirts are also creating downforce, a way for you to reduce your rolling resistance and induced drag could be to raise the tail a little but and vent a little bit of air behind the front wheels, this would reduce your underbody downforce reducing induced drag too.
This vent could be something like a skirt with a small protusion into the air stream, it sounds weird that adding frontal area to the car can reduce the drag, but its one of those counter intuitives examples to drag more there to save else where
What kind of protrusion from the skirt, and what kind of angles for the skirt meeting the undertray do you reckon would work best?
have you standardized vehicle weight, and tire pressure? Tire pressure can have a massive effect on efficiency. also added weight from more material at the back axle could be to blame. I would suggest adding lightness and trying to see how far you can slope the read down before you hit turbulence, as a shorter tail would also reduce weight, and improve axle balance.
The weight of the tailnisnt enough to effect axel balance
@@MaYbYl8eR foam board is extremely lightweight.
I'd like to see this done on something super inefficient like a lifted silverado 😂
The original: It scooped up from the bottom more. The top part didn't rake down so sharply and the sides 'boat tailed' in sharper converging at a point just below the taillight level. Even, fully enclosed transparent wing mirror fairings? Maybe? Love these uploads btw :)
Super interesting. I did some tests on a 3 tonne truck a few yars ago with an inflatable tail - inflated by the stagnation of the air at a duct on the top of the body. I didn't have very good metrology but thought that I had reduced the drag by about 5%. The inflatable idea was that when it was stopped the tail would deflate and have bungee cord in it that would collapse it down to a short length for parking. I had to rent the truck for the testing so I ran out of time and energy but I think it would work well.
People often put stuff on the front of vehicles to aid aerodynamics - we have a thing called NoseCone in Australia but it is really at the wrong end IMHO. Avoiding the separateion of the boundary layer on the rear is waaaay more important than making the front rounded...
Absolutely LOVE this series! Please keep going! 🙏🏼 The more you refine, the better the data.
this guy took it to the limit.
you might enjoy it.
ua-cam.com/video/QhmKVe0u_oI/v-deo.html
Coilovers, lighter wheels, and oddly a tune could help too. Looking at race cars that go for efficiency in fuel economy is cool. I also understand you aren't trying to buy the house in parts. Shoot I'd donate if you did this with something smaller. As a car enthusiast what you are doing is pretty cool.
+1 for tune even on a N/A subaru
Another tip is to tape over all the panel lines and shut lines on the front of the car. People do this on racetracks to reduce drag a bit
The commentary toward is probably more spot-on that you would expect. I'm the guy that typically doesn't refuel until the reserve light comes on, and I get to know my cars fairly well as I drive them and pay attention to things like fuel dispensed. There are some gas stations I simply avoid because they'll consistently dispense almost 10% more fuel than others in the area. In cars where I know when I'm almost completely empty, just how much should be filled, going over a certain amount is pretty obvious. I had one car with a 12 gallon tank that would typically take 12-12.5 when the computers range was in the single-digit miles remaining. On gas station would routinely dispense 13.5. Another gas station never dispensed more than 12.5. Unless you have some calibratable way to measure the fuel, I would recommend either only using one pump at one station, or doing multiple runs using different stations and averaging all the results together in hopes of reducing the significance of that particular mode of variance.
Great simple experiments, well summarized.
To monitor realtime data, and collect data on vehicle performance; you could invest in an On-Board Diagnostic (OBD) dongle. This would allow realtime monitoring and collection of data on a mobile device (via wifi, or bluetooth). Data parameters available varies a bit by make, model and year, but often can view realtime fuel consumption, or estimate quantity remaining, as well as odometer readings. Logging would allow export to a spreadsheet for some fun maths.
BTW: did you do a calibration check on tire pressure? This an often overlooked item and can make ~5% efficiency difference. Varies with temperature and seasons, so is worth checking once a month, or prior to efficiency runs.
Agreed - snoop the ecu with and obd scanner!
Torque Light is free! :)
I hope these vids start seeing more traffic man, such good content, and you do enough work to justify it!
Love your content, which i found thru Daniel's channel. Love all of your ekrano-esque creations btw, such a cool phenomena that seemingly NO ONE is taking commercial advantage of, yet, you and Daniel are making a whole series of interesting ground effect vehicles in a relatively short time frame JUST for the appreciation of physics and to 'feel' that effect! Theres so much intersting phenomena to behold, and all the micro details of each design can make or break the entire character of the flight... so cool!
Keep up the awesome work buddy! Cheers from HTX
I think one of the reasons why the tail *seems* less efficient (but might not be) is the combined effect of the underbody bypass and the tail. The previous tail might have reacted better to the previous body kit, but this one might benefit more from the smoother airflow coming from the new system
Great job. 17% is huge.
I would guess that you'd need to do several runs with each mod, in two directions, and tuft test each run to get a clear idea of what is actually happening.
You could use a pressure sensor to measure your air pressure in the tail section. Then pressure drag is equal to this multiplied by the area of your truncated cross section.
Aerocivic (ex)owner here, as well as owner of one of the highest fuel economy Insights on the road (100mpg+). Happy to see you continuing to experiment.
Legend....
Add to front a little splitter below the airdam. 30mm further front is enough for most of the benefits in separation. (based on Nascar 2019 CFD results etc) I would also cut the rear bumper and only make cone for the remaining lowpressure zone. (or remove bumper and use good rationalizing on the cone starting height.)
12° is the magic angle for wind to leave a moving object. 😊
Haven't you considered an _inflatable_ tailcone?
One made from [clear] plastic film, able to be inflated at low pressure (maybe even by scooping air in motion?) and deflated while parked.
I thought of that a long time ago for semi trailers but didn't try it. Once on the highway, the flat back of the trailer would deploy a kind of sock that would fill up with air that slopes down to a point. When you slow down, it would automatically deflate and roll itself up out of the way. That could save a lot of fuel for trucking companies.
@@elliottdiedrich3068That's a very cool idea indeed. I would like to think that it would be worth pursuing. 👍
He says, adding a rear wheel skirt would require modifying the car so he doesn’t do it. Then he proceeds to make all kinds of modifications to the body of the car. LOL
Seeing the rear of the new boat-tail made me think of the british WW2 stern design. Apparently they realized having a flat stern instead of a slow taper was more efficient, because the turbulence swirled around and pushed the boat forward? I may be wrong about that but it's worth looking at.
That reminds me of the kammback, which found a truncated tail was more efficient like used on the Prius and Insight. Then consider the bullet shape called the boat-tail, which is essentially a kammback too, that mimics the same design for stability in flight.
12:25 Mercedes EQS. Their flagship EV and the production car with the lowest drag coefficient ever.
Just wanted to share my mileage tests by saying when filling up the gas tank the only way I could think of to keep consistent was filling up neck so I could not put another drop in. So that worked for me, but some cars might have problems with smog systems. I was not testing pollution effects so my mileage worked just fine. I was able to get 27.6 mpg in my F150 but if doing a full tank to empty don't drink to much, it's a long drive.
What year is your F150? Dad made me drive his 2006 for a while and at 55 MPH I maintained 19 MPG.
Nobody tailgated, though! :D
I had a 2014 subaru impreza sedan cvt that i used for my 1hr commute that got a consistent 36.9 mpg, i now have a 2010 honda fit manual trans that gets 38.1 and am in the process of k24 swapping a 2011 honda crz in hopes to achieve 40 plus mpg with some tuning, this video was great to watch!!
Legal or not, it could also be that the cops who followed you decided that they just didn't want to know; "No, nope, to close to end of shift/next break, etc...."
you can check mileage with the onboard computer. drive say 80mph on a level road in a high gear. Use the same gear to compare. do several runs and note the windspeed and air temperature each time for every setup. If you get better mileage in worse or equal conditions then the better efficiency is confirmed. Onboard fuel metering has about 0.5% accuracy and can be offset if the tire size is not nominal but the tire size doesn’t affect the test obviously if all tests are done on the same tires and pressure (check pressure each time). The fuel metering can be adjusted through obd if offset but you don’t need that either since all tests are using the same calibration
Note the computer mpg average over say 0.5miles. Each run on the same road preferably
Im guessing that the variance in the results from the gas pump come from the mechanism of the stop valve in the nozzle. Since its purely mechanical, it might not always fill to the same point each time. Ive noticed on my car that sometimes the amount it fills will be off by a half gallon or so when i fill the tank from the same mark.
Awesome video though!
It is a back pressure pop valve. Using the same pump and nozzle is good, but the car, temp, air pressure, humidity/dewpoint is what changes along with the saturation of the vapor trap cannister. If overfilled on visit 2, then visit three may cut off as much as two liters early because the vapor cannister has become wet from last overflow.
Yo yo yo!!! - Please find a secluded hill and try out rolling in neutral. If you can’t rely on your pedal or gas pump input then just see how far your car rolls (with you in it of course) from top to bottom!!!! Your results should overhaul anything gas mileage might tell you since you know for a fact one is more air efficient
Filling in the rear of the tail surely hurt you a bit. The plate cover is a large surface for the low pressure wake to act upon. Remember that it's pressure region's acting on a moving body itself that contribute to drag! Having the tail hollow the first time around meant that realistically only the edges of the foam were subject to wake forces, with the gradient increasing to probably around atmospheric when you reach the body. With such a better optimized curve on the tail, you had way higher velocities at the bluff cutoff, so every square cm had an even more severe pressure drop to contend with vs the former. If you cant get a point (which you probably can't achieve efficiently anyways) find a way to leave her hollow!
Well, the old design created a large cavity, and there is no way to tell what the internal cavity pressure was without some further analysis. Sometimes it can result in higher pressure when you create desirable recirculation inside the cavity, but in this case I suspect recirculation was minimal. So it was probably transmitting the wake recovery pressure, or a lower pressure, up against the hatchback and bumper inside the cavity. Which means not much of a reduction in base drag. The new closed boattail shape is probably better.
Great work, and brilliant imagination, to figure out such cool ways to make your car so much more aerodynamically efficent.
I could imagine that the new tail causes more drag, because the back area, where the numberplate sits is substantialy bigger. Without directly changing its size again, you could try to angle the surface at maybe 45 degree, effectively creating more of a tip, instead of a big rear stub. Do you have something like discord for better exchange of ideas?
I have an FA24, ambient temperature makes a huge difference. In the summer when I am driving in 70-75F in the morning and 85F in the afternoon, I see ~30-31mpg. Now in the fall, morning temps around 45F and afternoon temps around 60F, I am seeing 35-36MPG. Same route, no traffic, trying to average 35mph or 50mph on the backroads mileage is always about the same over 3-4 tanks, leaving me to believe it is ambient temp, or if the additives package has changed in the gas.
On another note, I had a pickup truck that I put a downward wedge cap on to do a fluid dynamics paper on (emulating a couple that were already done), that alone improved efficiency about 10-15% on a regular basis.
I think some of the efficiency losses may be because of side winds. Perhaps a shorter tail section would reduce drag in a cross wind. Also, may as well leave the back portion of the tail section off so that there's a clear view of the existing license plate holder. I think such a design is called a "partial cam back." (Not my words) it may be the side wind and driveability solution as it would be shorter and (perhaps) still have most of the efficiency gains.
Very interested in this as a series. Kinda hoping for a more final version, perhaps made from fibre glass laid over an existing test piece.
Also, may as well test the pointy front. I hear it's kinda bad, but have no numbers to tell exactly how bad.
Drop weight Electric fan conversion light weight pulley conversion, hot air intake instead of cold air intake. 5psi over tire pressure. Tune the ecu to run at 15.5afr instead of stock 14.7 stoic also automatic transmission oil cooler will cause the transmission oil to slip less and improve efficiency of the torque converter lockup coupler. I’ve done all of this and it all works I’m a mechanic 15 years
You may know your stuff with aerodynamics, but establishing a baseline fuel efficiency needs to have minimum of 10 runs under similar condition, wind speed and direct plays a major role in fuel efficiency. My SUV gets 32mpg, but winds weather head, tail or from the side can throw (increase or decrease) off fuel efficiency by up to 15%. After each change or theoretical improvement, need to make equal number of runs and then average them after removing anomalous results. Called the scientific method. I have logged 10k miles of fuel efficiency in my own car, using data logging, engine performance, and weather condition. So many well intentioned people come to inaccurate results, because they can't or don't consider all factors.
**Looking at a car with a 6 foot aluminized foam tail** - "Car looks incredibly plain now..."
Love these videos, it’d be great if you keep producing more. But don’t feel pressured to produce videos quickly and rapidly, take your time
Before the back addition, the curves helped with the airflow. You could try angling it more towards the bottom, so only the area with the license plate is flat and nothing else, and even that could be slightly angled (something like 15 degrees or less).
This is a really cool series! Maybe one day if this makes enough you could try to find a cheap and already teardrop-ish car and try to make it insanely light and efficient. Also the Mercedes was an EQS, the new electric one that I remember seeing in videos s year or two ago, but it seems they haven't gotten very popular, so that's a cool car spot!
When I am passed by a semi on the freeway in my RV I feel pressure first against it pushing me away and then towards it sucking me in behind. Those admirers of the car that stay next to it will have an affect on air pressure in front.
my idea for the gas pump issue is the different grades of fuel, ie the ethanol content during the climate shifts throughout the year.
“I’ll take zero ethanol, please.”
@@JoeOvercoat
My car is older so I only use ethanol-free, and while it does cost more, I get about 2-3 mpg using it. For consistency, he should definitely use it when testing at different times of the year.
Removing the skirting introduced more drag, than the ever-so-slight increase in efficiency of the new tail could hope to compensate for. The ground effect between the chassis of the car and the road, is ~30% of the total drag of a modern car! You were definitely on the right track, when you tried to address turbulence underneath the car. Another thing to address, (I'm sure you thought of this) is tire rolling resistance. Tire pressures make a big difference. When a car becomes this efficient, tire inflation pressure becomes a greater percentage of overall drag, in that it greatly influences rolling resistance. Ambient temperature also makes a difference. A cold tire on a cold road on a cold day has much more resistance, than the same tire at the same inflation pressure, on a hot summer day.
I might suggest using an OBD reader to get your air-fuel ratio while doing these tests, it's possible your car had richened it out for one reason or another causing more fuel consumption. Cars actively manage all of the conditions that make the car run and could be doing one thing or another to make catalytic converters heat up or cool down or a bunch of other things. I think averages from a set of tests using one variation for another would be a better indicator. Love the videos, 18% is incredible.
When flying, after polishing our small aircraft (115 Horsepower) We can expect around a 5-knot increase in it's cruise speed at 25 Liters per hour. This represents around a 4% increase in cruising speed. It would be interesting to see what would happen if you smoothed the rear material and polished the car. Also, perhaps putting the fuel units in L/100km as well as MPG might help appeal to a wider audience.
Equally interesting is the dimples on a golfball that create lift by essentially creating an air pocket around the ball. At least that’s how I think it was described as how that works or something similar. I think I recall long ago someone recreating a dimpled effect on their car which smoothed the airflow because of the “pillow” of air on the body.
The splitter is giving you the most gains. The boat tail could be reshaped a bit get the flow into the dead space, also refining the splitter will help. One thing you might to try is divert some the air from the top of the boat tail to the bottom and clean up some of the dead air.
I don't care about the fuel efficiency part
But those skirts HELPED me a LOT because i race, and how simple your skirts are make me excited to try them on my racer
Fantastic!
I also did DIY aerodynamic modification to my car (Kei car) and improved fuel efficient a lot.
I'm so glad I found this video!
currently I'm on raodtrip in the whole mainland Japan with my aerodynamic car lol
*Cybaru
a thing that you dont mention... slightly overinflate your wheels. it can reduce the rolling resistance quite a lot
Why not go full Ricer mode, add the stickers and the sponsors and don’t forget add colors. Red means fast and Green for efficiency , don’t ask me how the color works just Ricer science.
The gap at the back bottom could be creating a vacuum, good for downforce, bad for drag
Laminar flow past the leading edge of the vehicle is oddly enough, a source of additional drag.
You aren’t trying to generate lift, you are trying to limit air contact with the body.
Instead of making the car resemble a wing, try to make it look like an upside down speed boat.
What i think what this video is when clicking on the video: A guy upgrading a car with high quality materials
What it actually was: a guy upgrading his car with cardboard, styrofoam and duc(k) tape
You did this very good!
I know I am late to the party, but - do you know with certainty that when the pump at the gas station cuts off, it is always accurate ie. what about foaming, or more precisely, the splashing that trips the trigger? It only needs a single splash to cover the 'trip hole' in the mechanism, and you could be quite a way out... Just a thought. (I used to work on such mechanisms, and they are far from precision mechanisms. I reckon you would have to do, maybe, three runs and average them out (-:)~
Love the work (and results!) Keep on keeping on!
Take a look at experimental aviation for fuel flow and totalizers. Plenty to choose from and very accurate. Thanks very much for your efforts in this endeavor, very interesting. Efficiency IS in the management of the air. I added disks to close off the rear dual wheels on my tow rig (second gen Dodge 3500, manual gearbox, Cummins powered RWD.) I log every tank (have sense purchase), noting if towing Ect, and I can clearly see a slight gain, .37 mpg over 40 fills. Sounds like mouse nuts, right?. AT $5 plus a gallon, for now, every little bit helps.
i agree with your last comment about the pump not being as accurate when shutting off, you can get around that by using a fuel cell instead and measure by weight
One of the things with your low tail vs high tail comparison to think about is the amount of air taken in through the grille. You had a big drop in efficiency when the splitter and skirts were gone in part because the engine bay was probably less efficiently removing air, causing it to spill into the wheel well. That's messy enough aerodynamically that many cars that are new have active air dams hidden in the grille for improved efficiency. The angle of attack on the extended rear diffuser can speed up the underbody airflow and get more of that messy, hot engine bay air out.
A simpler test that will eliminate the most errors is a down hill coasting test. Others have mentioned it. And it can be done in other forms. Dead stop from exactly the same place on the same hill under the same weather conditions and simply monitor the top speed and coasting distance. Do this 10 times with the regular car to establish average deviation. If your aero works well the top speed will increase as well as total coasting distance. I bet the car in it's stock form will fluctuate +-3% or more so an increase of 17% without knowing the average deviation is more likely 10-14% in actual improvement. That's still very good. So keep going. Make sure all things are kept equal. Tire pressure, weight in the vehicle, weather especially wind conditions. If the wind or traffic turbulence fluctuates the air +- 3% that's not helping get good data. stay away from dirt it will very the rolling resistance, stay away from testing in traffic and avoid stops and starts in the test, that puts the engine in the worst efficiency rpm range and can easily be effected by the driver, also avoid turns as they induce turning drag and are also effected by the driver. Go straight down a hill and only measure coasting top speed and distance and a more clear picture should come from that. Good work keep it up.
Great testing. You queried why you got the results you did. Please restore the car back to baseline and then record 20 consecutive tanks. That will help you understand variability. You will also see that reporting efficiency gains to 0.1% precision are ambitious.
(if he's married) Props to his wife for letting him play around with his engineering ideas. As embarrassing as this might be for "what will the neighbors think" types, it's actually pretty cool and how a lot of tech/science advances. Apple was created in a garage. This dude is in heaven working on this stuff.
For testing I would recommend measuring instant mpg, not trip mpg, and at as high of speed as possible to make the differences more apparent. If you can do 90+mph on that lake bed that would be much better. Calculating based off a trip introduces so many variables, not least of which is the driver. You might drive more cautiously with the tail, thus improving MPG.
Yes and wind plays a major role
@@aapjehoan and static wind is easily measured and you can somewhat normalize the data by taking data points going both directions. It's much easier to calculate a potential variance knowing local windspeed than trying to guess/calculate driving style impact.
Do you account for the actual wind speed during the test. This is quite a variable when you want to go that precisely.
Rear defuser the turbulence created by the ground effect adds drag so you need to channel and speed up the air going out of the back
Your new tail is less efficient because ot the vortices created at the edges were the upper surface and the side surface meets. This is beacause your tapering angle is larger at the top surface than at the sides, creating two vortices similar to airplane tip vortices and these cause the excess drag. You can even see these vortices even at low speed in your smoke test at 11:42.
Other input is that seapation is usually just a bubble, the airflow closes up after the bubble so it is not necesserily as big of effect as one might think. (Taillight covers are not so crutial.) So even if you have some separtation but if you manage to taper the tail of the car more your wake region can actually be smaller.
As for the wheel arches you do not have to fully close them you can merely make the gaps smaller.
I think it's very likely you fill up to a different point every time. You'd probably need a more accurate sensor in your fuel tank, or a transparent hose that makes the fill status very clear. I've found for my 50l fuel tank, that filling from empty till the click, then letting it really settle and filling again till the click gives me am extra 3l of fuel.
I've been tracking my fuel consumption for aound 50000km, 2500l of Diesel, and sometimes for no reason my sonsuption goes up by like 5%, even though I commute the same long route every day.
To get beyond 18%, here is a suggestion. Some air will flow underneath the car even you put an air skirt in the front. Starting from the rear bumper, you can taper the underside of your tail addition. The angle of the taper is gentle. See 2000 Honda Insight or Chevy EV1.
I too really enjoyed this. Aero & efficiency are always on my mind.
You could use the bottom flat sections of a couple of truck bed liners to make under trays.
Use rivthreads to attach them to the bottom. You also could use abs sheets. Even though most of the bottom of my '19 prius has trays, I used airtabs to jump the exhaust gaps and suspension gaps which has resulted in a 3mpg gain.
Still a work in progress.
In order to fill your tank to a consistent level partially fill at the pump and then use a dip-stick and top up by hand from a can to the same level. As an alternative to the dipstick get a clear plastic tube; insert it into the part filled tank; seal the end with your thumb; partially withdraw tube now filled with fuel; have the tube describe an S shape rotated 90°; take thumb off end; now the fuel in the visible portion of the tube will show the level in the tank. This should allow consistent filling of the fuel tank.
a 0.3 gallon deviation on 4 gallons sounds like a lot, but it could be fuel temp related (filling up in the morning vs evening affects fuel volume as the tanks in the ground heat and cool leading to slightly different densities. it's why its done by weight in aviation.) or it could be down to aeration of the fuel as it's delivered. the "click" happens when the air tube is obstruced, and having a nozel that produces slightly more foam while dispensing will trigger the stop early.
It's why Mythbusters always used a fuel cell, you can just put it on a scale to get the fuel weight
Forget weird aero that can potentially kill you and DO simple things: REDUCE WEIGHT.
REMOVE THE ROOF BARS (that is killing your aerodinamics)
GET lighter ryms, Clean your Trunk of stuff you don't use.
Those are key and will not kill you.
Maybe add clear plastic tapered teardrop shaped covers on the mirrors aswell? See if that really gives you 5%
While you eliminate one variable by using the same gas pump you still have the issue of the automatic shutoff. It will never be consistent. You have a couple of options. 1. Drain the tank completely. Not sure if there is a drain plug on the tank or not. Once it is drained put a measured amount of gas in the tank. Then run the car until the fuel is exhausted. Then put gas in the tank, start the car so until the fuel lines are primed then drain the tank again and refill with the same measured amount of gas. Repeat drive with the aerodynamic changes. Another option is to put a small tank and pump in the trunk and move the fuel lines. Either way you need to know precisely how much fuel you put in and then run until empty.
12:36 hmmmm... at first I thought the cabin side-skirts were like a single line (or two) of thick bristles (like a thin broom). Maybe that would be a good practical solution for an actual daily car, as when you drive over an obstacle (eg a high speed hump), the bristle would give way, but under normal high speed driving, they would act as a solid surface/wall, forcing the air to go around.