Tailgating a lorry/semi (modelled in CFD)
 

I conducted a little experiment this evening to see the aerodynamic effect of driving close to a large vehicle and where a good position to slipstream is.

The red/orange zone is where the air is at lower velocity and the idea would be to get the car in that zone so that the car is entering air at approx 20mph instead of 60mph (less resistance).

http://i4.photobucket.com/albums/y14...comparison.jpg

Looks like around 1 car length is the place to be at 60mph, however painful death is a risk and the grim reaper has to occupy your passenger seat.
http://i4.photobucket.com/albums/y14...experiment.jpg

It's a great way to wind up wearing the semi's rear bumper as a new facial implant. It'll look real stylish in a coffin...

Seriously, those trucks can stop a lot faster than you would think they can, and your reaction time is generally a lot more than 1 car length at 60 MPH.

Saving fuel is one thing, but doing it in an unsafe manner is entirely another.

-soD

That's an excellent representation of how close you need to draft to make any significant improvements in fuel mileage. Indeed, if anyone was contemplating this as an actual method to practice, there is next to zero reaction time for braking or obstacle avoidance.

decapitation
 

Jane Mansfield lost her head this way.If you've grown tired of your cranium I recommend it.

Yeah I probably should have added a disclaimer on there somewhere. I was just experimenting since I have the modeling software and the skills to use it. You can see the yellow section goes further back and would still allow some fuel saving while mitigating the risk of painful death.

Deja vu over again, deja vu over again...http://www.youtube.com/watch?v=lttgT1XZVvE

...or go to 8:45 mark.

That's an urban legend...it was either her wig came off, or a portion of her scalp, not her whole head Jayne Mansfield - Wikipedia, the free encyclopedia

Head, wig, scalp, doesn't matter if your dead.

Notice in the second simulation the drag appears worse than in the first one.

Can you give us a % of Cd improvement at 1, 2 and 3 seconds behind the truck at 45, 60 and 75 mph? :)

Viio, could you do another round of simulations, with the truck's wheels removed? There is always a portion of air flowing under the truck which would slightly alter the wake's shape.

(Not that I'm going to be reading any truck's tags with my nose, just curious about the advantages of staying at 2-3 car lengths.)

Contrary to the color codes in the simulation, you can use drafting to save fuel at longer distances than one car length... at approx. 30-40 ft. distance, I save around 0.5 l/100km with my setup which equals to about 10% fuel savings. Closer in, you may save more, but this cannot be sustained for longer periods of time safely.

Safety will always be an issue. Much depends on how you drive. Burger in one hand, soft drink in the other, cruise control on, foot on the dashboard: yes, you will have a hard time to brake in time. If your eyes, hands and feet are were they belong, your safe driving envelope will grow.

so long,

tinduck

I like your style :)

Do you mean with just repeated with the wheels removed, or do you want to see the model with zero airflow underneath?

CFD work
 

Thank you for doing and sharing. I'd love to hear more about the software you are using and any tips.

If you are game, it might be interesting to repeat the experiment except model from top view. :D

Rock on!

Just remove the wheels and keep the height the same.

Here's the experiment with a top down view. I've changed to flow trajectories instead of the colour map so it's easier to see.

It's 60mph again, and anything that is red is 10mph or less.

http://i4.photobucket.com/albums/y14...omparison2.jpg

I'll do the wheels-off view next.

Wait! Wait! My 2 dimensional mind just realized that this is in 3D, so the wheels should be on to make it more realistic. Unless the wheels of the model are much wider (the whole width of the truck, for example) than they should.

Well I've done a wheels off before reading your post, so we've got one anyway. As you might expect in 3d, the car just piles into high velocity wind anyway, and the wake of the truck is higher up.

http://i4.photobucket.com/albums/y14...cknowheels.jpg

Nice simulation pics. Could you post the scales with any new images you post? Its really difficult to know anything without a scale.

Also, is the gradient showing air speed or pressure?

It's velocity, with red being the slowest. I'll post scales from now on. I'm glad the images will be useful enough to warrant more info!

We already know that driving closer the truck saves fuel from mythbusters and other tests done earlier in the books like hucho... What I would like to know what are the savings at proper safety distance like @ 62 MPH 100km/h 3 seconds rule --> 273 feet or 83,4 meters

Also modify the truck lenght so that it is 83,4 meters/4 --> 20,85 meters

Then do simulations also to 62,25, 41,7 meters safety distances. According to Huchos book the drag should decrase with similar steps or how was the example in the book. Someone who has that book can tell the basic laws valid to this drafting.

Lastly make a truck train with 4 trucks with 41,7 meter safety distance and keep the same distance to the last truck with the car 83,4 meters. You should see some nice results this way.

How long does it take to make one simulation like that?

Here is waht I wrote to TDIClub forum on the similar topic:

There are benefits already at that distance about 3 second and more. Savings will come related to the vehicle length in front of you and how large cross section does it have. Also your own car Cd effects and the cross section you have in your car. If I would drive behind small car there would be no gains at those distances. Check book aerodynamic of road vehicles by hucho for hard data. If you like you can also find mythbusters test on the same topic Drafting Semi uploaded by tremorfalcon. 100 tf gave 11% better fuel consumtion.

In europe max length for truck + 2 trailer combination is 25.25 meters. On mythbuster that was only truck with one trailer about 17 meters. So with 8 meters longer combination you can drive that 8 meters more behind and still see the same savings. So that 100 ft or 30 meters would be with 25.25 meter long truck 44.5 meters or 146 feet. and still get 11% fuel savings in that same test truck. If your car smaller you get more savings. So about 3 second behind you can see about 1-6% fuel savigs. If you are driving behind a long truck "convoy" savings will be even greater because the distace needed is then about one convoy not one truck lengt and you can see that 11% at 3 seconds behind the convoy.

Holy crap Vekke, you take this stuff seriously :) This was just a mock-up model but you make me want to do a precise one now. It'll take a couple of hours - thankfully I have tomorrow off work so I'll produce something then.

Only thing is, I can run the simulation and tell you the air velocity, but I'm not sure how that will translate into fuel savings. I've got no education on aerodynamics as yet, just a bit of modelling skill and plenty of enthusiasm.

This is my work at the moment so more I know easier it is to sell and also design new products.

That accuracy for the truck is more than enough what you have at the moment. Just make it little bit longer or then use the safety distance what you get if you multiply the trucks current length by 4. It close enough to three second rule. at the moment it looks like its about 16 meters long? Also at the front there is only one front axle at the front between those two you have at the moment, but even that should not effect the big picture.

The catch here is that you should be able to get nice results by first doing three simulations and after that make a simulation about "truck train" and you should be able to compare results easily to wind speed. and also cd when you have the data from huchos book.

This is in SolidWorks, no? Perhaps I'll have to talk to some product design students to see if we can't figure out a Cd feature or something.

For reference: 60mph/100kph = 88 feet/sec and 27m/s.

So, driving a relatively safe 2 car lengths back, the car should be positioned about 175 feet or 54m back from the rear of the semi. 80m would be about 3 seconds.

If the airflow coloring is different at the car than it is going over the semi we should see some improvement, right?

Okay, so here is another model. The colour code is included in my screenshot, windspeed of 25m/s applied (56mph which is the UK speed limit for a truck of this size).

Truck is 15.1m long, 5.1m from top to bottom. The car is 3.7m long. The distance from the rear of the truck to the front of the car is 54.0m as requested.

Seems to me that there's no slipstreaming/drafting effect at this distance. Clicking the picture will open up a 1920x1080 version.

http://i4.photobucket.com/albums/y14...lo54mapart.jpg

It's slower air than the blue stuff outside the semi's wake. So instead of hitting 56 mph wind, you're getting 42. I'd say 14 mph less air drag is successful!

Oh I missed that, yes I suppose so. Blue air is at full speed, the first green area is about 5 metres per second slower (or 11mph).

my conclusion so far is that you can safely draft semi's and see significant improvements in mpg's

seriously though, if you cannot avoid hitting a slowing/braking semi from 100 or 150ft back, then you have no business driving on the highway in the first place.
IMHO

Yes if the coloring is different than the blue on top of the truck your car will face that 6 m/s slower air and use less fuel. Now It would be also cool to know for how long/ far behind that green 6m/s slower air will go. I have estimated the effect with that 3 second distance to 1-6% savings. If you drive 5m/s slower 18 km/h or 11.1 mph slower 55 MPH -->50MPH your fuel consumtion is about 13% less according aero rolling resistance calculator.

From myhtbusters test they found out the following results in their tests: Wer can now put those distances to that simulation and get following colors
100 ft=30m=11% greenlight and green dark so about 17.5 m/s air speed--> 7.5 m/s saved 11%-->1,46%/(m/s)
50ft=15m=20% yellow air mostly 13.4 m/s
20ft=6m=27% orange air mostly 7,8 m/s
10ft=3m=39% red colored air mostly so 5m/s

If you calculate the difference mean that 6 m/s would mean about 8,8% fuel saved at that distance. For better perpective you need also the upper picture because the truck is narrower and the green is not so long in width section.

At 60mph, 88 feet of distance leaves you with 1 second to react. You are dead if that semi rolls over a 2x4 in the road. On a drag strip, you are looking for that light to drop and the best anyone can do is 0.4 seconds to start. You cannot maintain that level of awareness for 5 minutes let alone a typical commute or road trip.

Personally I will not follow any closer than 3 seconds, and I have always managed to stop when I need to; sometimes I still have had to make a complete stop unexpectedly.

Hey, great looking stuff!

Far enough back to see the truck drivers mirrors on both sides at all times, then, yes there is some benefit which makes "drafting" worthwhile. But not any closer.

The second benefit is that other traffic will want to flow around the unit of "you" and the truck as the truck is seen as being the obstacle. The occasional other driver will insert himself, but not often enough to worry over. IOW, let the truck be the front door and be aware he'll slow on grades. . so "you" as the back door will be managing the overtaking & passing traffic on four-laners to the left lane. Done well this will speed them all around faster and make for fewer driver inputs to each vehicle in re traffic concerns.

So that 14-mph reduction is earned without causing the truck driver to feel there's a problem riding his tail on the one hand, and any skill application to move others around more quickly benefits both on the other. This would be win-win, versus the pure-D bull**** of close drafting where there is no benefit to either party worthy of the name.

.

Sven7, was that a video from your car?

Those proportions look a little off. On the Continent trucks are usually 18m long and 4m tall. I know that UK doesn't have the 4m height limit, but I thought most of them were no more than 4-4.5m tall. I could be wrong, though.
Reducing the height to 4m would reduce the wake's length and make it more symmetric (top-down vs side view).

But even with those proportions in the view of the car 54m back it looks so far behind. Maybe I'm just too used to Polish drivers driving bumper to bumper at highway speeds - not for economy, but because they think it's faster. Maybe I'm wrong about that, too:rolleyes:

Yeah my proportions are a little squiffy but close enough to learn from. I could model again with a 4m truck. I know Vekke is keen to see a road train so I'm going to model that too.

Here is one using a more accurate model of a European HGV, and the proportions are fixed - 18m long, 4.5m high.

http://i4.photobucket.com/albums/y14...mx45m25m-s.jpg

Next is a render to see just how far the truck's wake goes, and then the road train.

No :thumbup:

Okay, so this is a learning process for me, and I think I've made a bit of a boob. :confused:

This time I've set the colour gradients so that it's clearer to see where the air is back at 25m/s, and the wake is much smaller than original expected.

Just after the 40metre indicator on the ruler you can see that the air is back to full speed (where a normal car height would be). That makes me think that the wake is no longer effective 35 metres behind the truck.

Any thoughts?

http://i4.photobucket.com/albums/y14...aketo25m-s.jpg

The problem i have with drafting semi's, at least on the Garden State Parkway in New Jersey, is that semi drivers rarely ever follow speed limits, and rarely ever do they drive with constant load. They accelerate hard up upgrades and let off on down grades. They themselves drive very inefficiently, which in turn forces you, the drafter, to drive inefficient. Thus, the potential gains are not as great.

Draft semi's who cruise at a constant speed on flat highway works amazing, though.

Yeah, I liked the earlier wakes more. I'm not going to draft any your new trucks, viio:p

No really, what happened? Did you change any of the parameters, or only the coloring? In the previous models it appeared that the 19m/s part of the wake extended quite far, here even the 22m/s vectors end much earlier.

I dunno, thats how I drive too and I've hit 100% above EPA driving that way. It surently isn't very inefficient. DWL is not as good as P&G and pulsing up a hill to glide down it is much more efficient as long as you have a manual trans or can keep your torque converter locked.

I understand that. I think what i meant to say is that it's difficult to gauge when a trucker is going to speed up or slow down on small rolling terrain when you are trying to draft 10-15 feet away from it. In these situations, i find myself using the gas and brake more often in order to maintain that distance, more so than i would when driving normally without drafting.

There have been times when a trucker has slowed down considerably and then just floors it and accelerates away from me. For me to have any chance of staying in the wake, I also have to floor it and try to keep up. Then they'll get up past 60mph and i'll have to play catch-up to try to get back in the wake.

Basically what I'm saying is, it really depends on how the trucker is driving whether or not you are going to get a decent amount of benefit out of it, at least around here.