Help with LARGE hood extractor (Porsche Boxster)

I'm working on installing a hood extractor on my Boxster and am wondering if anyone can give me some advice on these topics or any others.   I've also include photos below of the splitter (before I cut the opening) and the wing so you can get an idea of the rest of the aero. 

1) should the intake be bigger, smaller, the same size as teh exit?

2) do I need to add a piece of angle on the leading edge of the exit?  I've seen this done inconsistently

3) should I add more surfaces inside the extractor to direct air?

4) should I cover the exit with a grill?

Other Aero

I did this on my rx7.

Temps actually went up at first.  I used my patented ghetto viz paint and string tufts and recorded some laps showing the movement of the paint.

It showed that air was actually trying to go into the opening from the top.

I added a little gurney flap to the front edge and everything worked wonderfully!

I would say that the gurney flap depends on where the high pressure is at the front of the car.  In our case, a gurney flap was definitely needed!

I cut my opening slightly bigger than the stock opening.  If you make it too big, you can always add tape to help with aero and temps.

I don't know what having the opening bigger or smaller than the exit will do to airflow and pressures.

In reply to Box4VIR :

so, first question... are you planning to have an air-exchanger of some kind in this space (i.e. a radiator, or cooler of some kind)?

I don't know boxsters real well, but AAZCD's GRM $2k challenge build exposed me to the idea that they had their radiators fed by the bumper's oval openings.  Are they staying there?

Thanks for the feedback!  My Boxster is a non-S so it only has two radiators on the front corners (so yes the oval openings).  So I will not have a air-exchanger in the opening.  The S cars have a third radiator in this location, my engine temps have always stayed cool so I never added the third radiator

So, if not for a radiator, then why do it?  Won't it just add drag from forcing the air to change directions?

I would think if your looking for downforce, you would be better off with tunnels under the splitter and setting the splitter at the correct angle to make downforce (very similar to a diffuser)

My modification was about moving to a single center radiator. Box4VIR is doing this for aero, taking the air pressure from low in front of the car and using it for downforce. Picture his extractor with a nice big splitter joining it up front and a large opening in the center of the front bumper. I think the concept is good, but the design - positioning and angle - is based on the space available under the hood. What do you guys think can be done 'grassroots' to optimize the flow?

AAZCD said:

My modification was about moving to a single center radiator. Box4VIR is doing this for aero, taking the air pressure from low in front of the car and using it for downforce. Picture his extractor with a nice big splitter joining it up front. I think the concept is good, but the design - positioning and angle - is based on the space available under the hood. What do you guys think can be done 'grassroots' to optimize the flow?

If I were to do it again I would have move the brake fluid reservoir, I could have pushed the open back another foot with a much smoother transition.  

Box4VIR said:

If I were to do it again I would have move the brake fluid reservoir, I could have pushed the open back another foot with a much smoother transition.  

Yes, and that seems like it would be a better angle, but I'm wondering how you can get some good testing done and how to tune it. No wind tunnel or CFD available, but maybe with some strategically placed yarn and a Gopro you could get a better idea of where the air is moving. Maybe change the size of the front opening as a way to adjust and balance along with corrections to the rear wing..?

Box4VIR said:

4) should I cover the exit with a grill?

Do you have Armadillos in Virginia? In Ok and Tx, I can picture that design scooping them and tossing it at the windshield. Probably just a wide spaced grill up front similar to the radiator grills.

In reply to AAZCD :

Haha thankfully none here.  It also isn't driven on the road so hopefully I won't see to many animals.  If you want to see some scary videos though YouTube racetrack deer strike.  You really don't want to hit a deer at 130mph.  

Using a powerful backpack blower with the yarn can help prepping before driving around, saves time.

wvumtnbkr mentioned a flap at the front of the exit. You could try one on 1/2 of the side to compare with yarn. I'm curious as to the angle of exit being perhaps too extreme, forcing the exiting air straight up and disrupting the flow?

Speaking of deer, I was driving around the other night and hadn't driven more than 1 1/2 miles down my road when I had to stop completely in front of the 18th deer. While stopped a huge owl flew straight at my windshield. Took the hint, turned around and went home.

AAZCD said:

My modification was about moving to a single center radiator. Box4VIR is doing this for aero, taking the air pressure from low in front of the car and using it for downforce. Picture his extractor with a nice big splitter joining it up front and a large opening in the center of the front bumper. I think the concept is good, but the design - positioning and angle - is based on the space available under the hood. What do you guys think can be done 'grassroots' to optimize the flow?

I understand this is for some downforce generation.  I would suggest small tunnels in the underside of the splitter for more effective aero.

so... my questions are to help me ballpark a few things...

what ruleset are you working towards?

do you have track data, and if so... what's the slowest corner you have? what's the highest speed corner you have?

is the street driveable, or is it "track only"?  if the later, what kind of time do you have to get it on track before you race?

I'm guessing you probably won't need a 'wicker' / 'gurney' on the back edge of the front /upper 'element'.

offhand, my gut says your exit area is too big... and you don't need a grill over it.  especially since the flow isn't being slowed by going through a media (i.e. radiator) of some kind.  offhand, rule of thumb, is you can accelerate a flow ~60% from the initial area without choking things.  I'm going to guess that you minimum area will be somewhere in the turn to vertical, and you'll want to work to keep the area constant after that or err on the side of 10% larger than the minimum.

The tough thing is you kind of need to calculate/estimate these areas as.... what's the word... tangentials?  I.e. you can't take straight horizontal/vertical cross-sections of your tunnel but from a point 90deg 'above'/'below' from a line that passes through the "middle"

I might have to make a sketch later, because that's probably clear as mud.

edit: changed the ballpark area.  I still need to dig into my old coursebooks to check on this, but you want a maximum pressure change of ~0.528p0.

sleepyhead the buffalo said:

so... my questions are to help me ballpark a few things...

what ruleset are you working towards?

do you have track data, and if so... what's the slowest corner you have? what's the highest speed corner you have?

is the street driveable, or is it "track only"?  if the later, what kind of time do you have to get it on track before you race?

I'm guessing you probably won't need a 'wicker' / 'gurney' on the back edge of the front /upper 'element'.

offhand, my gut says your exit area is too big... and you don't need a grill over it.  especially since the flow isn't being slowed by going through a media (i.e. radiator) of some kind.  offhand, rule of thumb, is you can accelerate a flow ~50% from the initial area without choking things.  I'm going to guess that you minimum area will be somewhere in the turn to vertical, and you'll want to work to keep the area constant after that or err on the side of 10% larger than the minimum.

The tough thing is you kind of need to calculate/estimate these areas as.... what's the word... tangentials?  I.e. you can't take straight horizontal/vertical cross-sections of your tunnel but from a point 90deg 'above'/'below' from a line that passes through the "middle"

I might have to make a sketch later, because that's probably clear as mud.

Thanks for all the feedback!

- its only a track car (no track driving) I have been running NASA TT5 for a few years but I have a kid on the way so no more buying fresh tires and dedicating time to keep everything competitive.  I'm now just going to instruct and run in HPDE 4 so I'm not building it for a class anymore.  This modification isn't legal for TT5

- I do have data, I use an AIM Solo 2, I don't have that data handy but posted some old Harry's data to show you normal speeds on my home track (VIR)   This was pre-aero so corner speed is up   Looking at the data there's a mix of turns from 60mph to 100mph.  I wonder if aero can be effective at 60?  Any quick question, you said I probably don't need a gurney flap, do you think it would hurt to add one?

So, I suspect that your minimum area is around about what I've marked in the red box with the hash, and step 1 in this really is getting a measure what that area is.  Everything else backs out from there.

ideally, you really want 'straight' vertical side back from that point, iirc (especially for what I'm going to suggest next).   If you don't want to rebuild it, I understand... and you should definitely get a second opinion before doing so.  If you were putting a radiator here, it'd make sense to widen it, so the flow slows and has time to do the required heat conduction.   It, that's not really what this device is for.

Now, I could be wrong, and I'll see about getting stafford1500 to chime in when he's got a chance.

As for a gurney, I was wrong.  You do want one... but not like you're thinking.  What you want is a 'cowl flap':

It's a moveable 'door' aircraft use to make sure an engine cooling circuit stays 'active' even at slower landing/takeoff/climbing speeds.  Yours doesn't necessarily need to be moveable 'actively'... it certainly could, and would go towards making this effective at your slowest speed, and less drag at your highest speed.  But that's a complication that could be added later.

Ideal, imho, you'd want this 'cowl flap' to have sides that stay in the tunnel through its range, and the easiest thing to do would be to mount it to the front edge of your exit with a 'piano hinge' (red line in the image below)... one sourced for aircraft kit building.  Then you can figure out a way to put a 'middle horn that connects to an adjustable rod (or push/pull cable). 

The orange lines in the two images are an outline of how I think the tunnel would be shaped with 'straight back' sides, and the green outline is a rough guess of what your flap would look like.

Doing this with it only adjustable in the pits will make it easier to fab, let you focus on driving, and not tempted to fiddle with it on track.  Then you'll use some tufts and video, and data, to figure out (or verify) the right amount to have it open, depending on the track you're at (you should be near Dominion, CMP, and/or Summit... which can be slower than VIR).

Anyways, that's what I would be building if it was my car, and I didn't have a kid on-the-way/here.  I do have kids 'here', so I understand the decisions you're making.  Hopefully I'll get a chance to check it out at UTCC in a couple years.

In reply to sleepyhead the buffalo :

Wow great info!  I wish I asked you before I started this project. Here's my thoughts

1) I'm going to avoid rebuilding it but could add another set of side plates that are vertical. So air flow would only see the verticals.  
 

2) I love the cool idea! I could reasonably easily make it active.  This is a Boxster which stock has a rear spoiler that pops up at 70mph, so I already have an electronic system that "deploys aero". I could hook it up to that and have a low speed setting and a high speed setting.   Most racing organizations ban home made active aero but it should be fine in HPDE 4.  I could also use this system to change the rear wing angle I suppose by mounting the rear of the wing to actuators the stock actuators might work for this actually.  
 

3) hopefully even after the kid arrives I'll always be able to make UTCC / Hyperfest, if I can only do one event that would be it!   What car are you running?

edit: for season I cannot find an answer to this question.   How much downforce (In pounds) is a typical, efficient race wing producing at 130mph?

I just gave this some thought.  The premise of a good extractor is you create a longer path under the hood than over the hood to create an airfoil correct?  If I added a cowl wouldn't it increase the distance over the hood and defeat the airfoil?   Airplanes seem to use a cowl only to increase flow through heat exchangers so they aren't concerned with this.  Certainly a cowl will increase air flow but not in relation to the air going over the hood.  Am I thinking about this correctly? 

In reply to Box4VIR :

Length of the path is a rabbit hole you don't want to go down.  It's effectively a wrong way to think about lift/downforce generation.  The flow doesn't have to speed up because it takes a longer path, and they have to meet at the back at the same time... the actually doesn't happen, which is why the flow wraps up over from the bottom of a wing to the top at the tip (which is why you have endplates).

Lift/Downforce is about turning.  Calling this device an 'extractor' is a bit of a misnomer, for me anyways.  Extractors are attempting to get flow that has been slowed (for one reason or another, but principally for cooling) to speed back up and rejoin the free stream flow as close to the same speed as possible.  To do this in the subsonic regime, you need a low pressure at the exit, and an appropriately sized exit area.  Those two things change together with speed (but the inlet and area and the 'slowed area' can remain a fixed, and the exit area and strength of the exit low pressure will 'feed the system' from the back.)

You don't have a radiator, so you're purely attempting to turn the flow and take flow that's down low, and get it to go up over the car.  That turning results in downforce.  You probably want to speed the flow up into your minimum area (which I marked above) so it has excess energy to make it around the corner.  Also, this means that flow should hopefully overpower the freestream air coming over the 'front hood' element.

yes, this does cause the bumper/front-hood to look vaguely airfoil shaped... and the moveable cowl is effectively a 'flap' like on an light plane.  So, if we took a waterline from the center, it might look kinda like this:

a secondary benefit of speeding up the flow, and keeping it sped up through the exit, is that higher energy flow will go over the car... which should help it to stay attached to the back glass and the make downforce from the rear wing.  maybe.

Box4VIR said:

In reply to sleepyhead the buffalo :
3) hopefully even after the kid arrives I'll always be able to make UTCC / Hyperfest, if I can only do one event that would be it!   What car are you running?

edit: for season I cannot find an answer to this question.   How much downforce (In pounds) is a typical, efficient race wing producing at 130mph?

I probably won't be at UTCC until 2022; and I've got to figure out getting a TT license before then.  Hasn't been a priority in the past... and I've had a number of... interruptions in my ability to pursue HPDE/TT.

re:How much downforce does a race wing produce at 130mph?
Good question.  Depends on the wing's size (length and chord), the wing's section (i.e. shape), and the angle the flow 'meets' the wing at (AngleOfAttack, which is roughly the same as the angle on the car... notwithstanding the back glass of the car angling the flow away from 'level with the ground')

which wing do you have?  Or which wing are you looking at getting?

This is a good start for a nice wing and the numbers it makes. It's be nice if they showed other specs like width or end plate size (or distance from and angle of the car surface it's mounted to), but as a starting point this is really nice:

https://9livesracing.com/pages/cfd-testing

sleepyhead the buffalo said:

In reply to Box4VIR :

Length of the path is a rabbit hole you don't want to go down.  It's effectively a wrong way to think about lift/downforce generation.  The flow doesn't have to speed up because it takes a longer path, and they have to meet at the back at the same time... the actually doesn't happen, which is why the flow wraps up over from the bottom of a wing to the top at the tip (which is why you have endplates).

Lift/Downforce is about turning.  Calling this device an 'extractor' is a bit of a misnomer, for me anyways.  Extractors are attempting to get flow that has been slowed (for one reason or another, but principally for cooling) to speed back up and rejoin the free stream flow as close to the same speed as possible.  To do this in the subsonic regime, you need a low pressure at the exit, and an appropriately sized exit area.  Those two things change together with speed (but the inlet and area and the 'slowed area' can remain a fixed, and the exit area and strength of the exit low pressure will 'feed the system' from the back.)

You don't have a radiator, so you're purely attempting to turn the flow and take flow that's down low, and get it to go up over the car.  That turning results in downforce.  You probably want to speed the flow up into your minimum area (which I marked above) so it has excess energy to make it around the corner.  Also, this means that flow should hopefully overpower the freestream air coming over the 'front hood' element.

yes, this does cause the bumper/front-hood to look vaguely airfoil shaped... and the moveable cowl is effectively a 'flap' like on an light plane.  So, if we took a waterline from the center, it might look kinda like this:

 

a secondary benefit of speeding up the flow, and keeping it sped up through the exit, is that higher energy flow will go over the car... which should help it to stay attached to the back glass and the make downforce from the rear wing.  maybe.

This helps a lot!  It makes sense to me now.  The hood is going in to be wrapped today, once its back I'll fab up a flap thats adjustable (by hand)

While I've got some aero experts on the line (and while I'm hacking my car up) I thought I'd ask about another problem that has been annoying me.  My rear bumper seems to catch a lot of air going around the track and you can see it being pulled outward like a wind sock.  I saw a photo of a car that had the bumper cut even more and tucked in. I wouldn't add the aero foil (but I guess I could), but thought the "tuck" might help.  What are your thoughts?

Interesting - does the factory bumper hang like that or is there an inner fender liner and perhaps a cover panel for the bottom of the bumper?

I would imagine that car would have something like that factory.

If not, you could try making those types of elements, or maybe doing louvers on the bumper? (Like the louvers over the front wheels on the white car you show). If you make a diffuser, you could have that pocket of air sort of side feed the diffuser.

In reply to Robbie :

one challenge with trying to feed a diffuser with air from the wheel well, is that the wheel is turning opposite to the direction of travel of the car for ~50% of it's rotation.  So, that, and some other things, means the flow in the wheel well is 'highly confused'... and, (iirc) generally speaking you want smooth uniform flow in your diffuser to get the most out of it.  I think that's one of the reasons why there' usually a vertical wall separating the diffuser from as much of the wheel well as possible.

along the lines of the wheel stirring up the flow, is to remember that the wheel is kicking up flow off the back of the tire.  I can't recall if in IMSA the lower "cheese wedges" are rules mandated... or effective... or a combination of both.  But, frequently cars are using them and a bunch of angled elements to straight the flow off the back tire and extract some downforce from it on the way out.  But, you really need an outer face that's flush with the outerwall of the tire for them to be effective and not disturbed by the flow coming down the side of the car.

In reply to Box4VIR:

 Can you cut away the bumper?  Sure.
Will it reduce drag?  Maybe
I think I've come across one 2D study (or a simple 3D study) of a miata that did that, and found a slight increase in downforce.  But I don't recall the drag impact.

In a grey area like this... and considering my own inability to generate 3D simulations currently... my default suggestion is to try it out, assuming you have another bumper... and any easy setupt to swap between the two.  That the best way to figure out if there's benefit.  One session with, one session without.  Maybe with video and tufting.  If you do cut the bumper, do it above the wheel centerline at full droop, and add a flat inner panel between the fender and the body for the flow off the rear wheel to attach to and orient back/out of the car.

thanks for all the amazing advice guys!  I just got the hood back from the wrap shop and started reassembling everything.  I have another aero question.   The opening is about 2 feet long and I could probably crawl through the whole thing, so it should get a fair amount of airflow.   What if I bought a wing and cut it down to fit in the opening?  So the leading edge of the wing would be even with the bumper.  It would be 1/3 the width of the rear wing but should be very clean air comparatively.