The Underbody (flat belly pan is suboptimal - explanation & solutions offered)

[aerohead]

Originally Posted by freebeard

You are thinking of the lurkers.

Anyone who'd read his comments deserved to 'hear' some counterfactual information.
Unless you'd studied fluid mechanics you might not even know the vocabulary engineers use,and the conditions/context which limits their acceptability for use.
We can't be casual about our use of precise language or embellish beyond utility.
Ockham's Razor!

[Pooft Lee]

After finding this and reading through, I wish he had stuck around. Formula SAE experience is better than what I have access to. But the follow up data didn't really happen, which is disappointing.

Back to the notorious audi photo, Mercedes is doing much of the same, so I think they're on to something.

While it doesn't look useful in the center, and may just be hiding attachment points, there's a clear pattern at the skirts, and they cared enough to extend the dimple from the pan to the side skirt. It looks like they're trying to generate vortices that fence underbody and outside airflow. A pretty cool alternative to low ground clearance. I've looked for CFD or other photos to confirm, but it would seem Mercedes likes to hang on to what they learn in their multi-million dollar testing facilities. Go figure. Thoughts?

[Pooft Lee]

The big challenge I would expect from a flat underbody would be keeping it consistent front to back. If air starts spilling from the high pressure you low it's going to mess with any nice airflow you have.

Does a diffuser help drive clean airflow even if it's messy in places along the pan? 3 degrees is the lowest angle I've really heard of but most of my reading on the topic was fron motorsport, not exactly what you guys are going for but there's a lot of room for improvement in a stock car. Increasing downforce should increase drag, but I think reducing lift is another story. Increased stability and lower drag should be doable in most applications

[aerohead]

Originally Posted by Pooft Lee

After finding this and reading through, I wish he had stuck around. Formula SAE experience is better than what I have access to. But the follow up data didn't really happen, which is disappointing.

Back to the notorious audi photo, Mercedes is doing much of the same, so I think they're on to something.

While it doesn't look useful in the center, and may just be hiding attachment points, there's a clear pattern at the skirts, and they cared enough to extend the dimple from the pan to the side skirt. It looks like they're trying to generate vortices that fence underbody and outside airflow. A pretty cool alternative to low ground clearance. I've looked for CFD or other photos to confirm, but it would seem Mercedes likes to hang on to what they learn in their multi-million dollar testing facilities. Go figure. Thoughts?

*Due to the length of commercially-produced passengers cars,they are guaranteed a supercritical Reynolds number.They don't need artificial roughness introduced.
*This allows for full flow attachment.
If you're seeing 'dimples',you may just be seeing recessed attachment points for servicing


There's no good reason to have any

*Any dimple would actually degrade performance of a smooth panel (Hoerner is a good source for this info).
*Adding any type of VG would betray a poor design to begin with.
*And a dimple is never chosen,other than in golf balls,as a VG.There are at least a half-dozen VG designs which would out perform the dimple.
*A specific vehicle,of which it's styling and major tooling had been locked in,if found to have trouble when assembled,could be 'tuned' in a wind tunnel,and a low cost gimmick could be added to mitigate design flaws discovered.But they would just be a Bandaid.

[Vman455]

Originally Posted by Pooft Lee

Increasing downforce should increase drag, but I think reducing lift is another story.

"Decreasing lift" and "increasing downforce" are different ways of referring to the same thing (coefficient of lift). Do you mean to say, "Increasing downforce with added devices such as wings should increase drag, but I think reducing lift without devices is another story"?

[Pooft Lee]

Originally Posted by aerohead

*Any dimple would actually degrade performance of a smooth panel (Hoerner is a good source for this info).
*Adding any type of VG would betray a poor design to begin with.
*And a dimple is never chosen,other than in golf balls,as a VG.There are at least a half-dozen VG designs which would out perform the dimple.
*A specific vehicle,of which it's styling and major tooling had been locked in,if found to have trouble when assembled,could be 'tuned' in a wind tunnel,and a low cost gimmick could be added to mitigate design flaws discovered.But they would just be a Bandaid.

Do you know a good place to read on the reynolds numbers? My knowledge with aero so far is pretty existential. There's a lot of guessing by feel and checking the results if you don't have CFD and don't know the math. I thought the boundary layer would thicken and continue to drag along a surface until it had to contend with a change in direction. Is the the turbulence you're talking about from the boundary layer slowing to a crawl and falling apart?

As far as the bandaid thing, a lot of engineering is bandaids but I see where you're going with elegant design. An afterthought shouldn't be praised as a breakthrough. I was referring to the Audi photos posted earlier with significant dimples 2-4" across that have a clear defined shape. I haven't posted enough here quite yet to link the photos, I believe they were page 1. But they seem concentrated around boundaries. It looks like the interest is around the edges of the floorpan, and the gap for the exhaust and transmission.

In an application where you can't be scraping every speed bump with nice ducting and diffusers, using vortex generators to fence airflow boundaries without impacting ground clearance is pretty elegant engineering. I'd really like to know if that's their real game with those large dimples on the undertray. On a golf ball, dimples trip up the boundary layer to keep it moving, if you created turbulent, faster airflow in the right places it should do a good job dividing different areas so you don't have unwanted spill over different pressure boundaries

[Pooft Lee]

Originally Posted by Vman455

"Decreasing lift" and "increasing downforce" are different ways of referring to the same thing (coefficient of lift). Do you mean to say, "Increasing downforce with added devices such as wings should increase drag, but I think reducing lift without devices is another story"?

I'm thinking lift from the underbody is caused by messy airflow getting stuck under the car, resulting in a decent amount of drag. If you clean it up with an undertray, and stop some of it from getting there with an airdam, you can eliminate lift and drag at the same time. I'm thinking reducing the air's influence that pushes up, instead of adding more that pushes down.

There are intersting examples of adding a spoiler, wing, diffuser, or both and actually improving the efficiency of the system. Ducktail spoilers done right can reduce the size of the wake behind the car while pushing the rear down, and spoilers and diffusers in tandem can "drive" each other. Whether or not average joe can make it work is a lot of luck and testing

But the thread is about the undertray, and I'm getting off topic

[freebeard]

This one?



It looks like the dimples recess the fastener and are tapered out to minimize their influence.

An airdam and spoiler can interact with each other (Porsche) and they're on opposite ends of the car. The airdam/diverter, wheelwells, bellypan and diffuser all act together. Airflow is segregated by fences, not dimples. "...the reynolds numbers" is actually Reynolds number[singular], a dimensionless quantity.

[Pooft Lee]

Originally Posted by freebeard

This one?

It looks like the dimples recess the fastener and are tapered out to minimize their influence.

An airdam and spoiler can interact with each other (Porsche) and they're on opposite ends of the car. The airdam/diverter, wheelwells, bellypan and diffuser all act together. Airflow is segregated by fences, not dimples. "...the reynolds numbers" is actually Reynolds number[singular], a dimensionless quantity.

Thats the one. I understand what you're saying, but can't see why audi would come up with such a dramatic way to hide a fastener if they didn't want it to serve a dual purpose. Back to the fence thing, canards in motorsport produce a vortex that helps keep airflow separated along the side of the car, I don't see why a similar idea can't be applied underneath. Are these just not shaped the right way? Would you want the teardrop pointed the other direction to produce a vortex? Here's Mercedes's similar application

[Pooft Lee]

They don't even bother with the fancy pockets for most of the attachment points

We're told it's "optimized" but don't get to hear why