Underbody panel testing
 

I found this in Piechna's book. Sardou ([1]) tested underbody panel configurations on the same car in different wind tunnels, both and without a moving floor.

http://ecomodder.com/forum/attachmen...7&d=1269633271

The chart below shows how adding different underbody panels influences aerodynamic drag and lift, but also lift for the front and rear axles.

http://ecomodder.com/forum/attachmen...8&d=1269633271

Worth noting are two things:

1. Adding panel #1 down the center increases drag,
2. Adding the center panels (#'s 3, 2, and 1) decreases lift in the rear.

[1] Sardou M., The Sensitivity of Wind-Tunnel Data to a High-Speed Moving Ground for Different Types of Road Vehicles, SAE 880246, 1988.

diffuser
 

kinda wish they had re-evaluated diffusers to see if there was agreement there.
Looks like a good book.Thanks!

tunnel
 

so from that chart,, pan everything,front and rear but leave the middle of the car unpanned ,,except the middle sides? . wonder if you just pan front to rear but leave the whole middle alone? very interesting. also shows how much pans can drop cd,, thanks for the great info.

I bet the middle one increased drag because with all the panels in place the air entering the engine compartment through the grille didn't have enough room to escape so the whole engine compartment became a "parachute"

We STILL have accept that this info from the tests could be completely different with a different car!

Brian

That sounds like a good explaination.
Nothing a grille block can't partially fix:)

The amount of drag increase from panel 1 is negligible compared to the reduction in rear lift. Though the increase in front lift does support the underhood parachute hypothesis.

I think it's amazing that the panels reduced Cd from 0.30 to 0.27.

I will emulate Sardou in starting my underbody paneling at the rear of the car.

Gazing at my own Insight, it would seem that panel #5 is the only area not covered, though Honda's design may differ slightly.

I also looked carefully at the rear area. By integrating the cargo well floor with the the bumper cover, even the #5 area is somewhat covered. It does look to me like there is adequate clearance on the torsion beam rear axel to use one large coroplast sheet to cover up everything remaining open, but JMO. I seriously doubt that this improvement will be measurable.;)

the drag resuls are pretty much as expected. while details may vary a full belly pan seems to decrease drag and lift.

what i find strange is the increase in lift caused by the front and rear pannel, and than the spectacular decrease caused by the center pannels.

i wish they'd tested them separately, because the numbers seem to sugest while they only cause a minor decrease in drag they cause a decrease in lift large enough to cancle out the front and rear increase... but one cannot conclude if they do this on their own or together with the first pannels.

Jack the car up, off its rear wheels, to find out what the suspension looks like near maximum deflection. Options include a flexible panel spanning the axle, or an underbody panel mounted to the axle.

As to whether the performance improvement would be measurable, well, how good is your measuring technique? ;) Also, a boat-tailed Insight is more sensitive to airflow at the aft underbody than most cars.

On the subject of diffusers, MetroMPG posted a promo video a while back that accompanied the release of the Insight. It was not written for a technical audience, but the voiceover guy mentioned that "the coefficient of drag was reduced by keeping frontal area to a minimum"*, and the (7° by my measurement) slope behind the rear axle "reduces turbulence underneath the car".


*Not as bunkum as it sounds. Keeping height to a minimum improves fineness ratio and Cd, not just CdA.

I ran this experiment slightly differently, but the bottom line is; there is plenty of clearance. I jacked up one side until the tire just cleared, the theory being that this represents the limit of weight transfer, and took the clearance measurement. There is 1.5-2 inches of clearance remaining, so a straight forward large panel covering the entire rear works.

[QUOTE]As to whether the performance improvement would be measurable, well, how good is your measuring technique? ;) [/UNQUOTE]

I've been doing some "testing of testing," but I remain convienced that small incremental improvements will be below the measurement noise on a FE twichy car like the Insight. An approach which tests several "known" improvements at one time has its own shortcomings, but the combined improvements will probably be measurable.

[QUOTE]On the subject of diffusers, MetroMPG posted a promo video a while back that accompanied the release of the Insight. It was not written for a technical audience, but the voiceover guy mentioned that "the coefficient of drag was reduced by keeping frontal area to a minimum"*, and the (7° by my measurement) slope behind the rear axle "reduces turbulence underneath the car".[/UNQUOTE]

I pretty much agree that using a rear belly pan to smooth up the rear and extend the effective area is going to make the diffuser area more effective, particularly if you install a boattail. The lower spring perches and the lower shock mounts will protrude, but even these could be covered by small bulges.

Some of the effects of increased drag may also be due to Venturi effect, maybe?

Assuming the parachute effect with panel 1 in place is correct, i.e., panel 1 prevents air from properly escaping the engine compartment, how about:

Use aluminum window screen mesh, held taught on a frame, as panel 1.

That way, the mesh lets cooling air escape from engine compartment, preventing parachute effect, but the flat, taught surface would smooth airflow from front to back. The alu mesh would also withstand exhaust pipe heat, and being an excellent heat conductor would dissipate hot spots.

I was considering a piece with louvers cut in it for that reason. that would also actually create a bit of suction to help remove the air from the tunnel.

Looking again a the drawings, I'm not sure that a parachute effect is the issue. There should be plenty of room for air to escape from the gap left aft of #4 for the front suspension.

Okay. I have a 2003 Golf 2.0l gasser. I have been wanting to try this for 9 years now. The car has raised 2 boys to 8 and 6 now and is in a condition that my wife will allow it. My question is what if these panels were to be dimpled? you know, like a golf ball? Remember when the "Myth beaterz" (not sure if I can say the real name on here) show fellas did that one with the ford Tarus and the modeling clay? It had like 900 xtra pounds on it and got the best MPG? If one were to do such a mod and have the entire underside of a car done this way what would happen? and another question, what of SALT? I hate SALT! they put it on the roads here in Ohio during winter and it eats cars. What would a person do to stop that aside from constantly removing and washing? For that matter, what would happen if I "dented" my whole VW Golf Car t?

I also figure the Parachute issue would be fixed by allowing the trapped air to escape out the hood meets windshield area. Suck air in at the nose, up in, over the motro, and out the wipers area. Nuttin a swazall can't create. Would most likely have to add closeable hood louvers to allow heat to escape when stopped or ideling.

PoorSmartGuy -

I think the dimples apply to spinning objects. In terms of parachute, I think you want to go for a "fish scale" overlap wherever possible.

CarloSW2