BOAT TAIL Drag Reduction Estimates

[aerohead]

This is just a quicky for those contemplating the addition of a tail and what it might do.I can't do a graphic today,will do numbers.
Data is from Hoerner's AERODYNAMIC DRAG,Fig.3.23,p.35,1951,which is referenced to Kosin & Lehman,'Aerodynamic Characteristics of Fuselage Bodies,'Deutsche Luftfahrtforschung,1942,p.I 241.

What I will do,is have you imagine your car already' with' a full boat tail based on the Aerodynamic Streamlining Template,which with proper ground clearance extends to about 80% when it 'strikes the 'line'.

Note: The aft-body of the example fuselage is responsible for a 567% drag reduction when added to the point of max camber.

From zero-to-20% chopped-off,the drag rises 3.95%
30%-chop = 12.5% drag increase
40%-chop = 29.6% drag increase
50%- chop = 59.6% drag increase
60%- chop = 99.85% drag increase
70%- chop = 257% drag increase
80%- chop = 330% drag increase
90%- chop = 426% drag increase


These values are for 'profile-drag' only( covering form and skin-friction drag ) ,so your car will still have components of cooling system drag,interference drag,induced drag.
You can see the significance the rear of the car plays.
With wings,the aft-body is responsible for a 1,000% drag reduction.
Most modern cars will achieve Cd0.13 with a full tail.With under-car wheel fairings the Cd can go even lower.

If you can do a breakdown of your existing drag components,you can use the table to estimate the profile drag as a function of boat tail length, holding all the other parameters as a constant to arrive at the theoretical Cd for your mods.
I've already done a parametric comparison to Walter E.Lays wind tunnel investigations and there is very close agreement between his work and the fuselage data.
Jaray,Kamm,and Lay,all did pumpkin seeds with forms close, if not in perfect agreement( Lay) with the Aerodynamic Streamlining Template.
The pressure recovery profile appears to be universal for streamline bodies of revolution regardless of thickness ratio.
It's my opinion that the 'phantom tail' concept is losing air.When you make a very close investigation of the research you find context which must be taken into account when making general claims about the performance of a tail.
Gotta go,clocks about run out!

[jime57]

Originally Posted by aerohead

Note: The aft-body of the example fuselage is responsible for a 567% drag reduction when added to the point of max camber.

From zero-to-20% chopped-off,the drag rises 3.95%
30%-chop = 12.5% drag increase
40%-chop = 29.6% drag increase
50%- chop = 59.6% drag increase
60%- chop = 99.85% drag increase
70%- chop = 257% drag increase
80%- chop = 330% drag increase
90%- chop = 426% drag increase

So,if I interpret this correctly, the best one can hope far in about 3.95% worse than the ideal template, with a boattail that is probably impossible long and low. RobertSmall's proposal looks to be about a 30-40% chop, if one degrades it a bit for the upward slop of the bottom, so he is looking at something like 20 % drag increase from the perfect template. The stock Insight looks to be something like a 70-80% drag increase from the template. Am I in the ball park?

[3-Wheeler]

I'm just hoping for a decrease in drag similar to Metro's.

Jim.

[RobertSmalls]

Originally Posted by jimepting

RobertSmall's proposal looks to be about a 30-40% chop, if one degrades it a bit for the upward slop of the bottom, so he is looking at something like 20 % drag increase from the perfect template. The stock Insight looks to be something like a 70-80% drag increase from the template.

I can't study this thread tonight, but I had to at least crunch these numbers.

If the stock Insight's drag is 75% greater than that of a full length streamlined car, then the latter has a Cd of 0.14 or so. Apply a 20% penalty to this, and you have 0.17. These numbers use a baseline Cd of 0.25, which is a stock Insight with mirrors, grille openings, and stock underbody coverage.

[aerohead]

Originally Posted by jimepting

So,if I interpret this correctly, the best one can hope far in about 3.95% worse than the ideal template, with a boattail that is probably impossible long and low. RobertSmall's proposal looks to be about a 30-40% chop, if one degrades it a bit for the upward slop of the bottom, so he is looking at something like 20 % drag increase from the perfect template. The stock Insight looks to be something like a 70-80% drag increase from the template. Am I in the ball park?

I ran the profile out with some minimum constants for cooling system drag ( 0.011 ),'features drag'( gaps,cut lines,door handles,antenna,wheel openings,etc.) ( 0.007 ),and then subtracted out 0.015 for a complete wheel fairing set using data from the Sunraycer.
It appears that if we do everything 'right',we could get a car down to a conservative Cd 0.126,at 80 % of the template length.
The last bit of drag would be insignificant,residing within the 'phantom' tail.
Running the tail out to 100 % would only get the Cd down to 0.123.

[aerohead]

Originally Posted by RobertSmalls

I can't study this thread tonight, but I had to at least crunch these numbers.

If the stock Insight's drag is 75% greater than that of a full length streamlined car, then the latter has a Cd of 0.14 or so. Apply a 20% penalty to this, and you have 0.17. These numbers use a baseline Cd of 0.25, which is a stock Insight with mirrors, grille openings, and stock underbody coverage.

I'm seeing the stock Insight at 30 % aft-body.
If I use Cd 0.25 and add the Sunraycer's original wheel fairings,I'm seeing Cd 0.241 at stock length.
If I run the tail out to 50 %,and with cooling system mods( ductwork) we might be looking at Cd 0.163.
With a one-wheel range-extender trailer with gap-fillers taking it out to 80 % template,Cd 0.119.

[RobertSmalls]

What do you make of my grille block, in terms of its contribution to Cd: Insight grille block

I will probably also block most of the are to the driver's side of the license plate, leaving <10in² of cooling air inlet low in the bumper, and ducted to the front wheel wells.

I see in the other thread you translated the above percentages into Cd figures:

zero------ Cd 0.488
10 %----- Cd 0.375
20 %----- Cd 0.299
30 %----- Cd 0.240
40 %----- Cd 0.194
50 %----- Cd 0.162
60 %----- Cd 0.135
70 %----- Cd 0.125
80 %----- Cd 0.118
90 %----- Cd 0.118
100 %---- Cd 0.115

It is interesting to compare your table to this one:

Specifically the half-sphere (approximating a 0% chop) and streamlined half-body (which has Cd=0.09 with zero ground clearance and no features).

Like the EV1 and the UFE-III, the Insight's rear glass doesn't exactly follow the template. I believe a 35-40% chop is a better approximation of the Insight's shape. That would place it around a Cd of 0.21, i.e. a 0.04 Cd reduction from aeromods other than the tail, and the potential for 0.10 Cd reduction from a complete boat-tail.

My tail extension will be cantilevered, not towed, and its underside will depart from that of the template to allow good clearance angles. This departure occupies 17% of the tail extension's area, in rear projection, which could not result in a drag penalty of more than 0.017 Cd.

It looks like I'm on track to build a car with a Cd of 0.13-0.15! I will be impressed if I achieve a figure in that range.

[jime57]

Got my Hucho!

I found interesting his comment that there had been no agreement between CFD simulations and the results achieved by real cars ;-)

[aerohead]

Originally Posted by RobertSmalls

What do you make of my grille block, in terms of its contribution to Cd: Insight grille block

I will probably also block most of the are to the driver's side of the license plate, leaving <10in² of cooling air inlet low in the bumper, and ducted to the front wheel wells.

I see in the other thread you translated the above percentages into Cd figures:

zero------ Cd 0.488
10 %----- Cd 0.375
20 %----- Cd 0.299
30 %----- Cd 0.240
40 %----- Cd 0.194
50 %----- Cd 0.162
60 %----- Cd 0.135
70 %----- Cd 0.125
80 %----- Cd 0.118
90 %----- Cd 0.118
100 %---- Cd 0.115

It is interesting to compare your table to this one:

Specifically the half-sphere (approximating a 0% chop) and streamlined half-body (which has Cd=0.09 with zero ground clearance and no features).

Like the EV1 and the UFE-III, the Insight's rear glass doesn't exactly follow the template. I believe a 35-40% chop is a better approximation of the Insight's shape. That would place it around a Cd of 0.21, i.e. a 0.04 Cd reduction from aeromods other than the tail, and the potential for 0.10 Cd reduction from a complete boat-tail.

My tail extension will be cantilevered, not towed, and its underside will depart from that of the template to allow good clearance angles. This departure occupies 17% of the tail extension's area, in rear projection, which could not result in a drag penalty of more than 0.017 Cd.

It looks like I'm on track to build a car with a Cd of 0.13-0.15! I will be impressed if I achieve a figure in that range.

Robert,if the grille block isn't cooking anything under the hood then I guess you're good to go.
Since the hybrids have expensive MOSFET and IGFET power transistors requiring cooling,I'm at a loss for recommendations with these advanced powertrains.
With respect to the drag reduction contribution from the block, crude logic might suggest that if you've cut airflow by 50%,then maybe that load has been halved.Just a SWAG.
Don Sherman got 0.4 mpg @ 70 mph with complete block to Pinto.2.0 mpg when airdam was added,then after cutting some block away for minimum cooling,gain came in at 1.4 mpg @ 70 mph.
With respect to the tail,I see no reason why you couldn't achieve your 0.13-0.15 numbers.

[cfg83]

aerohead -

I started with this image :



Are the red vertical lines the correct chop-off points? :



Orrrrr, the original black vertical hatched lines?

CarloSW2