Aerodynamic vs. vehicle stabilty

[SPARKY]

well said   

[bvillercr]

The thread went sideways when Blue said that he was talking to a SCTA official about providing aerodynamic proof of stability over a certain speed for certain vehicles.  So you want people to ignore this kind of statement?

[Cajun Kid]

I think B'ville is correct,,, the thread went negative after that statement.

[blackslax]

well said   

Thanks for the reach around sparky.  I guess this means that you will be needing those leathers I offered?

[bvillercr]

well said   

Thanks for the reach around sparky.  I guess this means that you will be needing those leathers I offered?

Don't get carried awar here, you'll have to start a new topic on the subject.

[blackslax]

I think B'ville is correct,,, the thread went negative after that statement.

I agree with what you are saying, but people could pass over the statement or voice that you disagree is one thing.  To loose all sense of decorum and start chest beating is uncalled for.  We could have 5 pages of "your a jerk", "no, you are", "no, you are" and that would be about the same.  

It always amazes me how a person's chest can grow at a far greater rate than their intellect when they have a tiny little computer screen to hide behind.

Or to put it in the words of my brother...."the main thing, is to keep the main thing the main thing;" and I think a discussion of aero vs. stability is the main thing.

[bvillercr]

Read Blues first sentence on the post.  He wanted it to go in this direction.  If he was interested in keeping the thread on tract he could post more often.  No chest puffing that I could see.  I don't know who is hiding, I and many others will be at speed week, where we will be running and proving our theories.

[Cajun Kid]

Go FAST and Be Safe  B'ville,,, we are all pullling for you to set yet another record.

Keep us posted and take a bunch of pics (in car too)

Charles

[Blue]

Slings and arrows, aim here.  Just got back into town, leaving tonight for Bonneville so everyone can beat me up in person.  Sorry if the absence was construed as a lack of participation.

Actually, the 675 mph lane change was attempted by Craig Breedlove and resulted in an upset.  I was the schmuck who had to figure out why it happened and fix it.  Whether we want to do a rapid lane change or not we sometimes have to.  As far as can cars and bikes go very fast without being aero stable?  Sure, the question was really: how fast?

Some of the people here are the best in the world at this sport and viewed my comments here as somehow questioning their already proven competence.  Sorry, no.  I'm one of those engineers who actually believes that when theory doesn't match reality we need to fix the theory.  I was actually looking for information and got a lot of it and I thank every single angry racer who told me to stuff it.  You all increased my understanding of multiple steering geometries in multiple sports in just a few posts.

After speed week (I'll have the white shirt on with my logo) and after everyone has all of the salt washed off, we can come back and post more.  I look forward to learning a lot.

[Blue]

Most LSR designers use the idea of a "Cp" as a plot of the lateral area of their vehicles and presume that if the CG is forward of the 50% point of this plot, then the vehicle is stable in yaw.  This is simply not true.

First, most symetrical aerodynamic surfaces rotate around the "quarter chord point" or only 25% of the length.  This is called the "yaw neutral point".  Very few LSR vehicles have their CG forward of this.

Help me out here. . .  I haven't had my tea today like Tony. Just where are these two points? Assume a 100" wheelbase. The CP is at 75" from the front spindle(s). Now if the CG is 50% forward of that point it resides at (75" x .50 = 37.5"). So the CG is at 37" from the front of the car? But you say that is yaw unstable?  Why? Is it too far forward?

Now where is this quarter-chord point? You say it is 25% from the CP. Same exampe, if the CP is at 75" the yaw neutral point is (75 x .25 = 19') in front of the CP. On a 100" car that yaw neutral point is at 56" which is a rearward bias! So where do you want the CG to be on a 100" car? 56" is too close to yaw neutral but 38" is unstable? All the sedans I know have the CG in the forward 50% of wheel base in the 45-48% range. What am I missing?

Edit: Since I see Willie is here I made "spindle" both singular and plural in the interest of diversity.   

OK, sorry to not address this specifically.  I really messed up by starting this thread on the front end of month of long hours and multiple road trips.

First, we need to uncouple the aero from the wheelbase.  Lots of cars have lots of bodywork well aft of the rear axle, so tails can be far from the CG while the CG is close to the rear axle for traction.

The fundamental of the 1/4 chord point relates to any streamlined aerodynamic object of smooth contour, i.e. an airfoil without separation.  Cars, bikes, and especially streamliners are very complicated aerodynamic objects.  To actually get the yaw neutral point, we have to look at aft vertical tails, forward vertical surfaces, round edges vs. square (round in front and sharp in back moves the CP aft), where the separation is, etc. and how all of this changes with yaw angle.  Look me up on the salt.  As far as whether my knowledge is opinion or BS, I'll bring a few reference books and a box of yarn tufts and tape.  Between the books and some yarn in the breeze anyone can figure this stuff out.

[ack]

I hope this provokes a lively discussion, this is a subject that I believe everyone in LSR can understand and benefit from with the end result of faster records and safer vehicles.

First, I mean no disrespect to Costella or any builder or contender in LSR.  My concern and comments on the previous thread come from my knowledge of vehicle mechanical-aerodynamic stability and control.  My concern is that speeds have now exceeded the area of mechanical stability and entered the area where aerodynamic stability dominates and the knowledge base of LSR is not yet wide or deep enough for safety.

I spoke to a senior SCTA board member last year about the need for the higher speed vehicles to prove their stability with analysis before going too fast, and we had a good discussion about the costs and where the break point could be for that analysis.  I am still searching for a reasonable CFD cost solution for everyone and have not found it yet.  However, my experience in S&C (stability and control) concerns me about some of the designs I see vs. the speeds people are seeking.

Most LSR designers use the idea of a "Cp" as a plot of the lateral area of their vehicles and presume that if the CG is forward of the 50% point of this plot, then the vehicle is stable in yaw.  This is simply not true.

First, most symetrical aerodynamic surfaces rotate around the "quarter chord point" or only 25% of the length.  This is called the "yaw neutral point".  Very few LSR vehicles have their CG forward of this.  Even so, at relatively low airspeeds (below 200 mph) the dynamic pressure is low enough that mechanical stability can override the aerodynamics.  Above 300 mph, the opposite is true and any vehicle that is not solely stable aerodynamically will not be recoverable if it loses traction.  Downforce can increase the mechanical advantage, but it is a bad trade since downforce usually leads to pitch instability.

Tails or other large vertical surfaces mounted far aft are used in some designs and can radically improve the overall vehicle's yaw neutral point.  However, blunt tails (like chute tubes) can reduce their effect.  Some of the vehicles currently seeking 400 mph are nearly neutral in stability due to their aft CG and high degree of aft separation.  There are solutions and a few in the 400 mph club have done a very good job of addressing this issue.  Some haven't, and that scares me.

At least a first-order, algebra-based stability calculation should be required of any motorcycle going over 200 and any car going over 300.  As speeds increase, the mechanical stability is going down exponentially with speed (dependent on surface condition, traction, and tire dynamics) and up linearly with downforce.  Countering this, aerodynamic instability increases with the square of speed.  At some speed the two lines cross and things can go bad very quickly.  Since most motorcycles do not have downforce, this equation leads to the need for positive yaw stability at the starting line.  Worse, downforce-based stability is at the mercy of driver skill;  and I like to be kinder to my drivers.

The REAL danger is that this "negative stability" speed may have already been achieved without external upset and then the vehicle makes another similar run and encounters an upset due to surface or wind conditions and suffers an uncontrolled departure; i.e. SPIN.  Think about all of those guys who have gone fast in roadsters or stock body cars and then spun at less speed.  Their driving skill may have saved them in the past, this does not mean it will forever.  At any combination of speed, surface, and wind condition it is the LSR vehicle's job to go straight, not to demand an ever-increasing level of dynamic driver input.

In aviation, we call the ability to handle instability the "velvet glove": a VERY complimentary term for the pilot.  And a not-so-complimentary one for the engineer who made it necessary.  As an engineer I don't like being the butt of jokes, so I make the things that I design stable and controllable.  My pilots appreciate this and bitch about other engineers instead.

All of this relates to yaw stability and spins.  Pitch and roll stability is another subject entirely and much more complex.

Here is my take on CFD for what it is worth:

Swift Engineering has just finished a detailed CFD study of the ACK Attack to try to determine why the bike becomes unstable at speed with the rear doors attached. This was done on their brand new Cray Supercomputer and CFD software which was installed in July.  http://www.swiftengineering.com/

The bike originally was designed with the rear of the body cut off and open with an open area of about 8” X 24”.  After setting the record in 06 and running as fast as 349 mph the bike handled very well and showed no signs of instability.

In 07 after adding doors to the rear we crashed at the Bub event when the bike became unstable and began to oscillate side to side at about 300 mph. In 07 we blamed the crash on track conditions.

In 08 at the Shootout we spent the first three days trying to figure out why the bike began to experience a side to side phugoid oscillation at about 330 mph which increased in frequency as the vehicle accelerated. After many tweaks on the bike I finally told Rocky to open the doors when the oscillation began.  When he did the bike became absolutely stable. We set the record with the doors off in 08.

We commissioned Swift to do the study and try to determine why the bike became unstable with the doors on. The first phase of the study looked at the bike at 350 mph and 6500 ft density altitude at 0, 2, 4 and 8 degree of yaw. The study showed that the bike was unstable at these speeds and tended to wind cock away from a cross wind. They said the rider could correct for this up to a point at which the bike would be uncontrollable.

The second phase studied the bike with the tail off my friend Ken Mort who helped with the design and has many years experience with the 130 ft wind tunnel at NASA Ames bet me a beer that they would find it is unstable with the doors off and that is exactly what they found. There was a slight increase in stability but nothing that would correlate with what the empirical data told us.

 The other interesting thing they ask was what was keeping us from going faster with the doors off. They calculated we needed 180 HP with the doors off and 163 HP with the doors on to go 350 mph.  The reality is we used all of at lease 800 HP with the doors off to go 360 and the bike was just not going to go any more than a few miles per faster. While these calculations might hold true for a moving body in free air they don’t work for ground vehicles.

The folks at Swift are very smart with practical knowledge and I appreciate the work they did on this project.  While the CFD is well developed and understood for the Champ and Indy cars which they work with at 200 MPH and aircraft at higher speeds as my friend Ken Mort says there is really no good understanding of the dynamics of ground vehicles running at half the speed of sound especially a motorcycle. We did receive a lot of interesting data from the study such as boundary layer thickness, pressure gradients, drag in different configurations and other data but the basic question about the change in vehicle stability with the door on was not resolved.

I would not trust CFD data to reliably predict the stability of a ground vehicle at 300 MPH + and knowing the cost of CFD; to require such studies at this point I belive would be counter productive and detrimental to the sport.   

One last comment as the bike has been displayed at a number of venues including the SEMA shows and a number of people who claim aerodynamic knowledgeable and expertise have come by looked at the vehicle and they know exactly what’s wrong with it and explain how I have designed it wrong. I always answer “how fast has the motorcycle you designed gone”? 

[Blue]

Here is my take on CFD for what it is worth:

It's worth quite a bit IM<HO.

I make my living in aero mods specifically focusing on areas where CFD is weak and inaccurate.  As CFD scientists like to believe, CFD can solve everything and anything.  This is true IF and only if the vehicle in question has no separation or stagnation.  We struggle every day to make aircraft unbelievably simple and totally streamlined.  Complex vehicles like LSR can NEVER be fully streamlined.  In these areas, wind tunnel analysis breaks down due to lack of accurate road-to-vehicle interaction and CFD breaks down because separation is an integral part of LSR aero, not an afterthought to be squashed like a bug.

The chute tube cover on Ack replaced disorganized separation with unstable vortexes on both sides.  Much like the experience we have all had driving down the freeway behind a tractor-trailer, a semi-stable vortex builds on one side of the vehicle and grows until it sheds and imparts a yaw moment.  Then the other side grows one and the cycle repeats.  This phenomena is easily understood by anyone following a truck, yet CFD and wind tunnels have a very hard time duplicating it accurately let alone predicting it.

Correct design of aft doors can cure both problems of drag and stability, but it's not so simple to contour the doors and tail correctly to prevent both separation and shedding vortexes.

[Rex Schimmer]

I think that "Blue" has made a very interesting observation of the reasons for the instability problems with the Ack Attack after the addition of the chute covers and also I think that his statements regarding CFD and LSR vehicles are something that we all can agee with to some point, remembering that the JCB streamliner was totally designed using CFD and it was certainly successful. Eric "Blue" is a big believer in tuft testing, I would ask Ack if you have ever done tuft testing with the Ack Attack before the tail modification or after or both??

Rex

[A2WindTunnel]

Prime example of this topic.  Did anyone at speedweeks find out any info that led to this crash from the driver/team?

Link to story: http://www.sltrib.com/sltrib/home/50114112-76/car-speed-kirk-salt.html.csp

[jl222]

  I drove the long course after the drivers meeting and I could feel a lot of bumping inside of truck from bumps in track, in other words it was not as smooth as in other years. Not big bumbs very small just not graded completely
out.  I thought at the time that cars might have a hard time hooking up, especialy unsprung cars. Dave Mcdonald said their Firebird wanted to dance around more than usual and I heard they spun later in the week. It will be interesting to see how many spins and crashes this meet had compared to past meets.

              JL222