eta: super efficient 3 wheeled car (project thread)

[diesel_john]

http://www.pagendarm.de/trapp/progra...les/NACA4.html
this site has an enteractive program which changes the shape of the airfoils. I moved the sliders until i got a shape that i could fit a large rectangle inside of. then downloaded the x&y coordinates and multiple both columns by the same number to the size of a truck in feet without going over legal road width. don't know the effects of my assumption on drag. I do see where the flow is going to stay against the surface a lot longer at slower speed. what do you think about the above site. the attachment in post # 46 would be the top view with the rectangle vehicle inside. along the bottom i would have a splitter all the way around. the sides would come straight up a couple feet with this profile then just rotate the curve over the vehicle like a jump rope to form the top. the biggest drag may be the joint effect at the ground, and lots of wetted area but if I am staring with a Cd of 0.02 or such wild number how could it not be extremely low, until the engine and driver overheat of course.

[Peter7307]

Originally Posted by roflwaffle

Motorcycle engines have crap BSFC compared to larger car engines due to their power to weight ratios. No reverse gear isn't a big deal since we can always just use an electric motor for that, but it still may weigh more than a reverse gear in a manual trans.

An extra gear in the transmission weight very little since the shafts , casing , lubricant shifting rails and so on are already there for the other gears.

BSFC is a stand alone measurement and is not related to weight.

High or low BSFC readings reflect the fuel consumed for an engine bolted to an engine brake (dynomometer).
This is normally measured in pounds of fuel burned to produce one horsepower for one hour.

Pete.

[diesel_john]

What's a good BFSC number for a car? I think JD has tractor that got
18.65 Hp.-hr./Gallon is that good? I think a gallon of diesel weights around 7.428. so 18.65/7.428= 2.51 hp.-hr./lb. oh, I guess that was at the wheels.

[Fuzzy]

Originally Posted by GenKreton

The equation of similitude I derived does only work under the assumption that ratios in the dimensioning cannot change.

When I first sat down with it I copied (w^2/w_m^2)^2 by accident (was doing multiple things at the same time.) I think this is what lead me to consider S/S_m.

Originally Posted by GenKreton

If you halve the length, you halve everything including radial dimensions. The reason for this assumption is if length and width, for example, change disproportionately then the coefficient of drag also changes, as you stated, and it introduces complexities. I'm not even sure Cdo/Cdo_m is valid without sitting down with some paper to check it.

I worked it into a spreadsheet using an NACA 2412, some generic GA-like wing-plan numbers, and some Cd numbers from XFOIL (a 2d aerodynamic performance estimation code), and it seemed to correlate pretty well.

Originally Posted by GenKreton

The world of secondary flow and eddies makes their modeling relationship nonlinear rendering that a looser approximation at best, or wrong at worst. The goal of the drag modeling equation was eliminating the need for full scale testing, and you are correct, if we modify it by changing geometry disproportionately then we do need to make an actual scale model for testing.

I'm not (or no longer am) familiar with "secondary flows"... is this relating to the inherent difficulties of predicting 3-dimensional flows (of which bodies of rotation {like a trike body} are)? I think codes like XFOIL would help bring the numbers predicted by the "Drag Scaling Equation" closer to "the ball-park" without the need of windtunnel/model testing.

Originally Posted by GenKreton

A quick example would be if you had an airplane airfoil that is 3x larger than you want for your car you do something like the following:

(1/3 scale factor) * (23.77/7.382 average density at sea level divided by 35,000 ft, scaled) * (60/600 grossly estimated traveling speeds) * (drag on airplane airfoil) = drag added to your car

Also if you only scale, preserving geometric relationships, then the coefficient of drag will not change on most of the sizes we would be discussing. Since you are considering scaling down size AND speed it is especially true. As you decrease speed the secondary layer flow decreases - the area where the fluid sheers off a surface, slowing down. This explanation is probably more than necessary I now realize

Yes, the boundary layer is affected by "scaling down size AND speed"; but I feel that the "Drag scaling equation" doesn't properly account for the laminar/turbulent transition location which is generally determined by Re, maximum body thickness, and the length where maximum body thickness occurs. Although, your assumption that this 'transition location (and thus drag) is negligible considering the Re range being considered' is probably sufficient for 'napkin' estimates.

[Fuzzy]

Originally Posted by diesel_john

http://www.pagendarm.de/trapp/progra...les/NACA4.html
this site has an interactive program which changes the shape of the airfoils.

<meh> the NACA 4-digit series equations can be found floating around the internet (or Theory of Wing Sections).

Originally Posted by diesel_john

I moved the sliders until i got a shape that i could fit a large rectangle inside of. then downloaded the x&y coordinates and multiple both columns by the same number to the size of a truck in feet without going over legal road width.

John, I'd be interested to see what you decided on. Could you PM the coordinates to me?

Originally Posted by diesel_john

don't know the effects of my assumption on drag. I do see where the flow is going to stay against the surface a lot longer at slower speed. what do you think about the above site. the attachment in post # 46 would be the top view with the rectangle vehicle inside. along the bottom i would have a splitter all the way around. the sides would come straight up a couple feet with this profile then just rotate the curve over the vehicle like a jump rope to form the top. the biggest drag may be the joint effect at the ground, and lots of wetted area but if I am staring with a Cd of 0.02 or such wild number how could it not be extremely low, until the engine and driver overheat of course.

This sounds similar to the Nissan Goodwood 'Gravity Racer' (linkage)or whatever from a couple years ago. It is important to send most of the flow around the car instead of over, since that will limit how much lift you have to compensate for... allowing a more stable vehicle with lower drag.

[Big Dave]

As one sees ideas being developed one must remember they have all been tried before. Same with three-wheelers. Three wheel vehicles with two wheels up front and one trailing have been around since the earliest days of automobiles. Their history has been on-again, off-again, and mostly off.

This is for a very good reason, not some deep dark conspiracy. Two-forward three wheelers are treacherous in a corner. They tend to swap ends really easily. One tire does not have the lateral traction of two. Worse yet, the zero camber of a single rear tire gives it no mechanical advantage over the tendency of a vehicle to “throw” its rear end around in even moderate-g turns.

I look at this design and I see a second mistake – engine in the rear. This exacerbates the tendency the throw the rear end around as now the weight of the engine is added to that force.

One of the very few (sorta) successful three wheelers was the British Morgan, built during the Great Depression The car had an air-cooled 45 degree V-twin mounted exposed in front axle. The car looked for all the world like a somebody in a three-wheeled MGA had t-boned a Harley-Davidson.

http://rides.webshots.com/photo/2588...98837763rCjmXx
http://rides.webshots.com/photo/2561...98837763WxRpEu
http://rides.webshots.com/photo/2137...98837763ELPtQe

That V-twin engine actually drove the rear wheel through a drive shaft. The Morgan was actually fairly successful because the engine was forward of the yaw axis and tended to resist the three-wheelers tendency to ground loop.

I don’t mean to discourage the idea of a two-forward three wheeler. It is an efficient idea. But I cannot imagine anything more discouraging than having to sit in your wrecked three wheeler whilst some fireman uses an emergency grinder to cut through your tube frame to extract you. Damage control always starts with avoiding a wreck.

If it were me, I would approach this somewhat differently. I would start with a VW TDI engine/transaxle and a vehicle about 60% the width of a Golf/Jetta. All the drive train and suspension issues have been worked out over decades of development. Front wheel drive is old hat in this day and age. My passenger compartment would be a tandem seating arrangement with a fairly upright seating position. An upright seating position gives you a very comfortable ride and you have to be able to see over the engine compartment. At the back I don’t use one tire but rather two skinny tires set at about a 2 degree camber. The rear wheels have a much narrower track than the front, allowing a very nice aero shape. With this setup, you get all the aerodynamic excellence of the three wheeler without driving a ground loop waiting to happen.

[Peter7307]

Originally Posted by Big Dave

One of the very few (sorta) successful three wheelers was the British Morgan, built during the Great Depression The car had an air-cooled 45 degree V-twin mounted exposed in front axle.

That V-twin engine actually drove the rear wheel through a drive shaft. The Morgan was actually fairly successful because the engine was forward of the yaw axis and tended to resist the three-wheelers tendency to ground loop.

The three wheeled Morgan was built well before and after the great depression of the 1930's and the main reason for the commercial success was related to taxation and low fuel consumption.
Having 3 wheels they were classified as a motorcycle and avoided the higher tax on vehicles having four and more wheels.
They also seated comfortably two and three at a squeeze and had reasonable weather protection.

Three wheeled Morgans are known as TriMogs by their owners.

Cheers. Pete.

[Big Dave]

No contemporaneous three wheelers had the Morgan's resistance to ground looping. Nice idea: Don't kill your customers. They might come back for another one.

With the exception of the Morgan - with its engine mounted forward of the yaw center - all two in front three wheelers tended to swap ends. If this was an easy problem to overcome, it would have done so decades ago.

[Christopher Jordan]

Originally Posted by Big Dave

At the back I don’t use one tire but rather two skinny tires set at about a 2 degree camber. The rear wheels have a much narrower track than the front, allowing a very nice aero shape. With this setup, you get all the aerodynamic excellence of the three wheeler without driving a ground loop waiting to happen.

LOL! A.K.A.: BMW 4 wheel Isetta with 2 wide in front and 2 close in back! Europe has several very much like the 3-wheel Morgan- about 5 companies, I think. This 3-wheel usage went back to the 1800s adult tricycles; but never got anywhere; until recently- (1980s-90s-now)- with underpowered human powered recumbents and velomobiles mainly. Even now a VW or Peugot or something shows up very shortly; then disappears.

Then something like Canadas T-Rex pops up, shocks everyone with spectacular burn outs, high top speeds, then goes only to the high end market. Usually seems to get a "can't be done - spins around - flips over." description. Brazil races similar but longer ones. I don't know for sure where I saw them- reverse trikes websight, probably.

Those mid '60 Harley/Corvair/VW choppers had weight problems, and soured the US attitude for years. They are 2 back wheel configurations,but put a 'thumbs down' on all trikes.

I would gladly make my Ford/Kia a 1 back wheel trike- weights in the front, but the back is tippy as all get-out as it is now; with the stock height-weight-tires-solid axle. Tear-drop with a whale-tail fin......

And that would bring the thread back to eta: a side by side 2 person trike! But the Twike (made in Germany) side-by-side 3-wheel EVtrike is already sold in the US!

countersTrike

[DAN]

front struts are about the worst. old pinto and mustang twos had a wishbone front end that come out in one piece. they were often used in things like cobra copies. good luck finding one. just about anything is better than struts. they are the cheapest thing they can make.