Aerodynamic vs. vehicle stabilty

[hotrod]

More oil tests

The first photo attached here shows the flow separation and then re-attachment on the hood scoop then the formation of a rotor behind the airscoop as the flow again separates.

The second shows the flow separation line on the rear pillar as the air tries to curve over onto the low pressure area behind the rear window.

As an extension of Blue's comments on end plate locations. On the WRX body that separation line on the rear quarter defined by the oil would be an excellent place to mount a side plate (end plate) for a spoiler.

Larry

[saltfever]

3. Bvllercr's video: A little tough to do detailed aero from the dust, here goes.  First, there is an unstable shedding vortex off of the front tire.  This is responsible for the perfectly spaced dust clouds going down the side of the car.  This is an important thing to minimize because it disturbs the flow on the side of the quarter panel and since the left and right vortexes will seldom be synchronized, it induces a yaw instability. 

I wouldn't jump to conclusions about that front tire. The motor is putting out 1,000+HP (per side). Look at the exhaust zoomies! You have tremendous pulses of gas disturbing the flow along the side. 

[saltfever]

4. Chute position:  I see a lot of chutes that eject straight back into dirty air and this is not good.  The chute is ejected into air that is recirculating and this delays the chute opening.  In some cases, the bag drags on the ground before it opens.  A lot of teams have theirs mounted at an up angle, and this is better, but the angle often is below the angle of the spoiler, so the chute is still buried in dirty air.  Either arrangement will work fine when we're straight, let's worry more about how it works when "oh-shoot-we're-spinning-I-pulled-the-handle-and-why-hasn't-the-darned-thing-opened-yet???"  A Honda red hat holder narrated his spin video for me and said that chute opening seems fast during a clean run and awfully slow when spinning.    Mount it vertically. Yes, I said shoot the chute straight up. As soon as the drogue clears the aft lip it's in clean air and it will deploy in a fraction of the time that it takes in dirty air.  Use a d-bag too, if you're spinning a d-bag will help prevent the chute from fouling on the car.  Bob Stroud and I talked about this at length and the more vertical the better.

A very important observation and I don't wish to minimize or discredit any of the quote. However, chute position is CRITICAL! Be very, very, very, careful about how you approach your chute attachment point. The chute must be pulling through the CG. If the initial yank is above or below the CG the load will unweight (or lift) the appropriate end of the car. That will lead to very bad things. Also, consider what happens in a partial spin as Blue mentions. If you are yawing 45 to 90 degrees and the chute deploys with a yank not in-line with the CG what do you think will happen? What might have resulted in a flat spin might throw you into a pencil roll! I understand Blue is discussing orientation and not attachment point. My point is to approach this carefully! How will Blue's proposed orientation affect initial opening? Where will the chute be (what altitude) will it be upon initial yank? How will that pull on the CG? This is a very serious subject.

[1212FBGS]

This means we need course reports, like seeing the size of the roostertail behind Speed Demon triple at the 4 mile marker.  That was some soft salt. 

apparently you've never seen Burkeland, Vesco turbinator run, or the salt spray sucked up by a 400 or even a 300mph vehicle before
Kent

[Interested Observer]

Blue:

Ref: Your reply #215:   "It only takes 20 to 40 mph for most of the separation and reversed flow effects to show up on the yarn."


While the flow regime indicated by tufting at 20 mph and 300+ mph may be similar for highly streamlined bodies, one has to wonder about the degree to which this holds for more irregular or production-based vehicles.

It is not hard for the semi-layman to suspect that for these shapes there may be several different regimes due to successive flow separations as velocity increases.  That is, tufting results at 20 mph (which might almost be considered potential flow) may not be representative of flow at considerably higher velocities.  Indeed, this might even cause one to look askance at low speed wind tunnel results.  Is there any validity to this viewpoint?

It would undoubtedly be illuminating for many LSR devotees if you could be prevailed upon to elaborate a bit on this admittedly complex question.

[Cajun Kid]

I.O. 

Darn, just when I was starting to understand, there ya go using them big fancy words again !!  LOL   

"Illuminating"  "askance"  "devotees"  "prevailed upon"

Charles

[Blue]

3. Bvllercr's video: A little tough to do detailed aero from the dust, here goes.  First, there is an unstable shedding vortex off of the front tire.  This is responsible for the perfectly spaced dust clouds going down the side of the car.  This is an important thing to minimize because it disturbs the flow on the side of the quarter panel and since the left and right vortexes will seldom be synchronized, it induces a yaw instability.

I wouldn't jump to conclusions about that front tire. The motor is putting out 1,000+HP (per side). Look at the exhaust zoomies! You have tremendous pulses of gas disturbing the flow along the side.  

Actually, I took that into account when I made that statement.  If we look at the intermittent draw of dust from the underside of the car to the side of the door sill it looks like this is independent and well aft of the exhaust pulses.  Also the spacing between the dust plumes does not vary, if they were caused by the exhaust we would see exhaust timing in the spacing.  That's a big leap of faith until we look at video and pictures of other cars without zoomies and see the same effect.  It can also be seen on Thrust SSC, and that was going a lot faster without zoomies.  It's something I've observed for years, this video just gave a good example.  The back of the tires are highly separated areas on any car, it is simply a question of how the separation manifests itself.

Completely separate from this thread is whether the exhaust can be arranged to help or hurt the local flow.  Yes to both.

[Blue]

4. Chute position:  I see a lot of chutes that eject straight back into dirty air and this is not good.  The chute is ejected into air that is recirculating and this delays the chute opening.  In some cases, the bag drags on the ground before it opens.  A lot of teams have theirs mounted at an up angle, and this is better, but the angle often is below the angle of the spoiler, so the chute is still buried in dirty air.  Either arrangement will work fine when we're straight, let's worry more about how it works when "oh-shoot-we're-spinning-I-pulled-the-handle-and-why-hasn't-the-darned-thing-opened-yet???"  A Honda red hat holder narrated his spin video for me and said that chute opening seems fast during a clean run and awfully slow when spinning.    Mount it vertically. Yes, I said shoot the chute straight up. As soon as the drogue clears the aft lip it's in clean air and it will deploy in a fraction of the time that it takes in dirty air.  Use a d-bag too, if you're spinning a d-bag will help prevent the chute from fouling on the car.  Bob Stroud and I talked about this at length and the more vertical the better.

A very important observation and I don't wish to minimize or discredit any of the quote. However, chute position is CRITICAL! Be very, very, very, careful about how you approach your chute attachment point. The chute must be pulling through the CG. If the initial yank is above or below the CG the load will unweight (or lift) the appropriate end of the car. That will lead to very bad things. Also, consider what happens in a partial spin as Blue mentions. If you are yawing 45 to 90 degrees and the chute deploys with a yank not in-line with the CG what do you think will happen? What might have resulted in a flat spin might throw you into a pencil roll! I understand Blue is discussing orientation and not attachment point. My point is to approach this carefully! How will Blue's proposed orientation affect initial opening? Where will the chute be (what altitude) will it be upon initial yank? How will that pull on the CG? This is a very serious subject.

There is almost no connection between ejection angle and pull point when a d-bag is used.  As a rule of thumb, the chute will fly 1/2 to 1/4 of its inflated diameter above the salt.  We need to draw a line between the center of the chute at this altitude and the CG of the car and make sure that the pull point is on this line.  If it is higher, we risk a wheelie, if lower the car can lift in back and stuff the air dam.

Once the pull point is set, the chute ejection geometry should be solely based on getting the chute into clean air as soon as it leaves the pack.  Shoot it straight up;  The pull point, inflated diameter, and riser length determine the load on the car, not the angle of ejection.

Even if the chute is significantly above the pull point at extension, the D-bag will drag into trail as the chute pulls out long before inflation ("long" being fractions of a second).  If a d-bag is not used, the chute will never get above the pull line since the local air flow against the drogue will determine inflation.  Although this may sound extreme, think about it:  If we shoot a drogue into the air stream at 20 to 50 mph (spring vs. pnuematic chute ejection) and the air is going by at 200 to 300 mph (50 to 400 times the dynamic pressure), which way is the drogue going to go?  The way the air wants it to go and nowhere else.

The point is to get it into that high velocity air and out of the dirty, slow, reversed air at the back of the car.  Once it gets there, then it will pull though the pull point;  which as saltfever points out should be straight through the CG, not higher, not lower.

[Blue]

This means we need course reports, like seeing the size of the roostertail behind Speed Demon triple at the 4 mile marker.  That was some soft salt.  

apparently you've never seen Burkeland, Vesco turbinator run, or the salt spray sucked up by a 400 or even a 300mph vehicle before
Kent

Actually I have many times.  Apparently you've never driven or groomed a course and actually walked on the salt at the 4 to 6 mile areas where the underground stream cuts the course at a 60 degree angle.  It's quite an experience in May driving through standing water at the end of the road, standing on salt at the 3 mile as hard as a rock, then watching your feet sink at the 6 mile.  The salt is always different at every quarter mile, every week, of every season and everyone's understanding of the salt would benefit from participating in grooming a course.  Which I did, under the guidance of the best; for which I am very grateful.

While not an "old salt", I've spent much of my life on playas and salt beds.  Over the years I've been lucky enough to learn from people who have more words for their condition than eskimos do for snow.  The salt from the 3.7 mile to 4 was soft this year and from 4.2 to 5 it was REALLY soft.  SCTA worked late and re-groomed these sections during speed week and we should all be grateful for that.  Any observant person looking at several previous high speed runs can watch the spray vs. position and normalize this data vs. how hard the drivers were staying in the throttle and report from this how soft or hard the salt is at each mile.

It's rocket science, not rocket art.  And I met plenty of rocket scientists at Bonneville.
 
LSR runs at Bonneville because it's flat and available.  In an ideal world where Bill Gates is a land speed racer, we would have 10 miles of pavement.  Not a variable, plastic, frangible surface that deposits on everything and corrodes metal.

[Blue]

Blue:

Ref: Your reply #215:   "It only takes 20 to 40 mph for most of the separation and reversed flow effects to show up on the yarn."


While the flow regime indicated by tufting at 20 mph and 300+ mph may be similar for highly streamlined bodies, one has to wonder about the degree to which this holds for more irregular or production-based vehicles.

It is not hard for the semi-layman to suspect that for these shapes there may be several different regimes due to successive flow separations as velocity increases.  That is, tufting results at 20 mph (which might almost be considered potential flow) may not be representative of flow at considerably higher velocities.  Indeed, this might even cause one to look askance at low speed wind tunnel results.  Is there any validity to this viewpoint?

It would undoubtedly be illuminating for many LSR devotees if you could be prevailed upon to elaborate a bit on this admittedly complex question.

"potential flow"?  Are we talking about the difference between inviscid and viscous solutions for widely varying Rn or the differences in development of separation phenomena into unsteady, partially stable, and/or stable recirculation at low vs. high Rn? <VBEG>

OK, all kidding aside IO brings up a good point:  the character of separation differs with Reynolds number, i.e. speed.  The point of the previous post was two fold:  #1, the existence of separation and the majority of the areas of reversed flow seen at very low speeds will translate well to high speeds.  If they didn't, sub-scale wind tunnels would be useless.  #2 TEST!  There is absolutely nothing I can tell anyone about their car that is better than the data they can get for themselves.

To a point, separation on not-so-streamlined-production-based-vehicles will actually get better, not worse, with speed.  This is due to the thicker boundary layers associated with higher Rn, and the difference is mostly irrelevant for the shapes common in LSR.  At lower speeds, we may see organized or semi-stable recirculation where higher speeds will see disorganized flow;  the point is that we will see the separation even at low speeds.  So once we are going fast enough to straighten out the tufts we're going fast enough for reasonable data.  When people have gotten that far and actually tested, then we can revisit how recirculation and separation vary with speed.  One step at a time.

[Blue]

More oil tests

The first photo attached here shows the flow separation and then re-attachment on the hood scoop then the formation of a rotor behind the airscoop as the flow again separates.

The second shows the flow separation line on the rear pillar as the air tries to curve over onto the low pressure area behind the rear window.

As an extension of Blue's comments on end plate locations. On the WRX body that separation line on the rear quarter defined by the oil would be an excellent place to mount a side plate (end plate) for a spoiler.

Larry

Outstanding data and an excellent observation.  Kent and John, don't listen to me, do what Larry is doing.

[saltfever]

 

There is absolutely no connection between ejection angle and pull point.

That is not the point I am concerned about.

As a rule of thumb, the chute will fly 1/2 to 1/4 of its inflated diameter above the salt.  We need to draw a line between the center of the chute at this altitude and the CG of the car and make sure that the pull point is on this line.  If it is higher, we risk a wheelie, if lower the car can lift in back and stuff the air dam.

I agree. I am not concerned about the pull point of a fully opened, functioning chute. The pull point through the CG is fairly well understood by most everybody. Also, if you extend the CG centerline straight back you will see it is invariably in dirty air. Which, as you mentioned above, is where most of the chute will generally be flying.

Once the pull point is set, the chute ejection geometry should be solely based on getting the chute into clean air as soon as it leaves the pack.  Shoot it straight up;  The pull point, inflated diameter, and riser length determine the load on the car, not the angle of ejection.

Are you sure? That is my concern. I have seen no data on this but would be very interested in learning about it. I understand what you are saying. I agree that clean air is most desirable for predictable, successful, chute performance. But I am concerned about the altitude of the chute on initial opening shock. Not the flight altitude of an inflated chute. As I interpret your thesis you would shoot the chute (no pun intended) above the roof of the car (or higher) into clean air. That is fine but it will hang there until impact? What altitude will it be at impact? Sure it will settle down to the flight height you mention after deployment. But that is not the issue. LSR chutes hit with 4-6Gs and that initial impulse will be way above the CG at an extremely critical time. Any hit like that will upset the car. And if you are in a spin it will roll the car! What you are proposing sounds nice but where is the testing data? Who has done this? Using my layman math it doesn't compute. 240MPH is 352ft/sec. That means it takes less than 1 second to get full deployment. My car is 5.5 ft high. If I loft the chute 6-8ft high (a couple of feet above the car into clean air), where will it be in relation to my CG when it hits? Assuming a ballistic trajectory of about 1 second, it is not going to be even close to my CG. If my math is wrong correct me . . . I am not embarrassed, I want to understand and be safe.

(Edit) Blue, my apologies because I see you were editing post No. 232 when I was composing this reply to that post. Your edit adroitly anticipated some of my concerns. However, please consider. The clean air above the car is stationary and laminar. We are launching a 240mph package into it. The chute package is moving, not the air! It will undoubtedly have some flight characteristics that could "hang" it in the air. At any rate I doubt it will loose altitude at 32/sec/sec. and yet it must get to my CG height in less than 1 second for deployment. Who is going to be your test pilot.  

[Blue]

OK, I understand how this would seem like bad math, the fact is that it is a mass vs. moments and this went through my head so fast I didn't even consider it.  Let's look at the two cases with extreme ejection (pneumatic): D-bag and no D-bag.

D-bag: When we shoot straight up, the d-bag is going 40 to 50 mph and hits 200 to 300 mph air as soon as it clears the spoiler lip.  The air now takes over and the d-bag goes in trail at about 40 G's of acceleration in the flow axis (5 lb chute vs. ~200 psf dynamic pressure vs. ~1 ft2 of drag area for the bag).  This 40 G's totally dominates the bag's velocity of ejection.  The bag and car separate to the limit of the line with the line pulling the bag in line with the pull point and the CG.  The chute deploys and hopefully the car hasn't gone too far around by this point.

No D-bag:  When we shoot straight up, the drogue hits clean air and the same thing happens except the chute will inflate as soon as about half of it's length clears the spoiler.  It's at this point that we see the need for a d-bag.  Without one, the chute inflates and basically stops in mid air while the car continues (5 lb chute vs. 200 psf dynamic pressure vs. 40 ft2 of drag area for the chute equals 800 G and the chute stops right where it inflates!).  The car moves forward from the point of chute opening until the riser is tight and then the car accelerates the chute.  If the car has any angle in the CG-pull point-chute line, the car will rotate.   

Read that carefully.

In both cases, the chute pretty much opens at the nominal height: centered just above the spoiler lip.  The point of shooting it up was to contrast this with shooting aft where the air is more than disturbed, Bvlllercr's video shows the flow to actually be reversed and sucking back into the bumper.  Lots of cars have this issue and dragging the bag and the chute on the ground is the result.  Shooting up at speed does not mean that it will shoot as far up as it will at a standstill.

As a matter of who has done this, drag boats now eject their chutes up and outboard at extreme angles so that at least one of the two is trailing when they flip-spin.  The yank straightens them out and prevents, not causes a roll.  We need to learn from other motorsports, just be careful what we learn and how it applies to what we are doing.

The air behind even the worst streamlined car gets less and less disturbed the farther back we go.  Behind one to 3 car lengths, it's disturbed, but not reversed like it is at the bumper.  Clearly, if we run a short riser and open the chute in really bad air right behind the car the chute will squid up, down and sideways and yank us all over the place like a funny car.  If we run too long of a riser we risk going around a ways in the spin before the chute inflates and pulls us straight.  Ejecting the chute straight up gets it to inflate at its nominal flying height in the shortest possible time without a mortar or rockets, which are used on fighters for ejection seats and spin recovery chutes.

[SPARKY]

Great stuff guys  thanks 

[jl222]

  No trouble dragging chute on ground before opening on the 222 camaro, we use a funny car type chute [ short lines] it opens instantly when applied. In 08 Bvillercr got real sideways and Glen annouced '' we have a spin''
but Troy hit both chutes and did some steering and it never went clear around. Chutes are angled up 33 deg.
 We once had a guy look us up at El Mirage because he was inpressed with how fast and well our chutes opened

            JL222