Beware old and new aero data

[JulianEdgar]

Can I put forward an idea?

A preamble: I love historical context. There is nothing in any area of car aerodynamics of the past that I am not interested in. Show me old aero cars and data - and I am always interested. That's a major reason I went and saw the Tropfenwagen in two museums in Germany, and admire the Tatras, Citroens and Chrysler Airflow so much.

But you need to acknowledge that old car aero data comes with some very major shortcomings. Primarily, they are in the measuring and testing methodologies that were then used.

Take the wonderful Tatra T87. It had a drag coefficient, when recently measured full-size in the highly-respected Volkswagen wind tunnel, of 0.36. Nothing special, eh? However, models of the car were being measured at the time of the car’s development at CD = 0.24!

Or what about the ground-breaking pre-WWII research vehicles, created by the German FKFS institution under the leadership of Wunibald Kamm? The models were dubbed the K1, K2 and K3. One reference states that K1 had a CD of 0.23 (and a model measured 0.196), K2 apparently had a CD of 0.23 – but K3 had the much higher CD of 0.37! So what happened to K3? – after all, it looked much the same as the two earlier cars. Well, K3 was tested full-size in the Volkswagen wind tunnel in 1979… In fact, it’s very likely that K1 and K2 had drag coefficients that were similar to K3. Good at the time - but absolutely nothing to try to take lessons from today.

So old info - based on measurements of models, or done in wind tunnels without moving floors or moving wheels - is all highly suspect.

And what about NACA data, or that great old textbook, Theory of Wing Sections by Abbott and Doenhoff? It's a book I have (I think I have all the major aero textbooks) but it's one that has as little relevance to the shape of current cars as data on low drag wheel covers on a WWI aircraft has. It's all interesting - but tech papers on wheels on low drag cars with smooth undercovers are in a quite different world to a WWI aircraft! (Not to mention the different galaxy of aerofoil profiles measured in free air!)

I read somewhere 'what is the time cut-off for scholarship'? The answer to that question is very easy: when the current evidence suggests that our understanding has changed. Patterns of airflow don't change, but they way that they are measured, modeled and predicted has changed massively over time.

I've recently been reading a 1960s textbook. It's very well regarded - in fact an absolute classic on engine tuning. But the material on tuning air/fuel ratios and ignition timing is so simplistic as to be useless for any current engine. In fact, to be honest, it sounds like they're trying to tune a lawnmower. And yet the author worked on, and helped develop, F1 engines....

If you don't keep up with the current research evidence, you're doomed (I think anyway) to falling into the trap of "it's always been like this" - when, maybe, it isn't any more.

[j-c-c]

" Patterns of airflow don't change, but they way that they are measured, modeled and predicted has changed massively over time."

And that sums up best what will always be the one thing certain, I think.

[aerohead]

Just read this thread.I'm away from my library and do not have access to my materials.
While I don't disagree that there is evidence that there are discrepancies between early data compared to more contemporary data obtained for the same vehicle/model,there do exist modern data which compares favorably with early reporting.
I don't believe that broad stroke generalities best serve this argument,but rather, maintain that any specific vehicle should be treated on a case-specific basis.
The Tatra was incapable of 'low drag' due to its aftbody.
Of thefive 'K-cars' built by Fachsenfeld at FKFS,only the Mercedes-Benz'experimental-chassied,diesel-powered,overdrived,1941-42 K-5,personal personal car of Kamm's was rated as low as Cd 0.23.It's the only 'drop-nosed' K-car.It's the only K-car with full wheel skirts.K-5 maybe the only K-car with Kamm's patented cooling system (which is what he is actually famous for,as the
'Kamm' -back is actually the property of Fachsenfeld,and the reason Hitler wanted Fachsenfeld at FKFS).K-5 had the longest aftbody of all the K-cars,the secret to its low drag.
Airfoil shapes would be relegated to sub-critical Reynolds number solar racers,and NACA (NASA) quanta would be valid.
The really 'low-drag' shapes which could be applied to passenger cars would be based upon Paul Jaray's,1922, streamline half-body,which modern wind tunnels and CFD re-verify,with each generation of university investigation.
In his second edition of Aerodynamics of Road Vehicles,Hucho wrote that any difference in drag,associated with wheel rotation during testing indicated such an insignificant difference in results that it wasn't pursued.General Motors Research Labs had come to the same conclusion from it's testing at Cal Tech in 1953.It would have been considered an absurd proposition in truck and trailer testing,due to wheel environment turbulence.Only open-wheel race cars were deemed worthy of the extra trouble/cost.
Today's F-1,Indycar,and NASCAR' billion-dollar racing programs mandate that 'ALL' areas of aerodynamics be investigated.
One curiosity always hiding in the room is that,given rolling-road wind tunnels,and million-dollar public relations budgets,and the ability to develop 'porous' low-drag wheelcovers,when tasked with setting land speed records,all auto manufacturers choose the totally sealed,convex disc,MOON wheel covers,the lineal descendant of WW-I era aviation; and can be seen on all extant Boeing 727s.

[freebeard]

K-5 maybe the only K-car with Kamm's patented cooling system

[ears perk up] So what was that like?

Originally Posted by j-c-c

" Patterns of airflow don't change, but they way that they are measured, modeled and predicted has changed massively over time."

And that sums up best what will always be the one thing certain, I think.

That's why I like Computation Fluid Dynamics. Software has changed massively over time. Having a moving floor is possible in meatspace, having an undulating/textured moving floor is possible in CFD.

It pairs well with Generative Adversarial AI.

Two Minute Papers: This AI Does Nothing In Games…And Still Wins!
hardware.slashdot.org: AI Drives Innovators To Build Entirely New Semiconductors

Increasingly powerful computers arguing with each other about the ultimate design for a use case. It should be fun to watch.

[JulianEdgar]

Originally Posted by aerohead

Just read this thread.I'm away from my library and do not have access to my materials.
While I don't disagree that there is evidence that there are discrepancies between early data compared to more contemporary data obtained for the same vehicle/model,there do exist modern data which compares favorably with early reporting.

I did not say that all new data disagrees with old, I said that old data should be treated with suspicion when the current evidence suggests that our understanding has changed.

Originally Posted by aerohead

I don't believe that broad stroke generalities best serve this argument,but rather, maintain that any specific vehicle should be treated on a case-specific basis.
The Tatra was incapable of 'low drag' due to its aftbody.
Of thefive 'K-cars' built by Fachsenfeld at FKFS,only the Mercedes-Benz'experimental-chassied,diesel-powered,overdrived,1941-42 K-5,personal personal car of Kamm's was rated as low as Cd 0.23.It's the only 'drop-nosed' K-car.It's the only K-car with full wheel skirts.K-5 maybe the only K-car with Kamm's patented cooling system (which is what he is actually famous for,as the
'Kamm' -back is actually the property of Fachsenfeld,and the reason Hitler wanted Fachsenfeld at FKFS).K-5 had the longest aftbody of all the K-cars,the secret to its low drag.

I am quoting data from Stromlinienautos in Deutschland (Kieselbach), P 84 for K1 - K4 and Aerodynamics of Road Vehicles (Hucho) (P 21) for K5.

As is typical with cars from the 1920s and 1930s, when tested in a modern wind tunnel, they all recorded much higher drag figures.

For example, you mention K5 as having a long afterbody as 'secret for its low drag'. It had a measured Cd in a modern wind tunnel of 0.37! So much for its secret. As I already mentioned, K3 was also measured in a modern wind tunnel at 0.37.

They were fascinating for the time, but to bring up these old cars as if they have anything to teach us today is just absurd.

Originally Posted by aerohead

Airfoil shapes would be relegated to sub-critical Reynolds number solar racers,and NACA (NASA) quanta would be valid.

Yes, I agree - so to use data relating to NACA aerofoils for shaping normal road cars is treading on very thin ice indeed. One obvious discrepancy is the thickness of the boundary layer on the rear parts of real road cars versus an aerofoil.

Originally Posted by aerohead

The really 'low-drag' shapes which could be applied to passenger cars would be based upon Paul Jaray's,1922, streamline half-body,which modern wind tunnels and CFD re-verify,with each generation of university investigation.

Yes Jaray was a very good aerodynamicist, and refined his craft working on the aero of Zeppelin airships. But his pure shapes are a long way from practical road cars.

Originally Posted by aerohead

In his second edition of Aerodynamics of Road Vehicles,Hucho wrote that any difference in drag,associated with wheel rotation during testing indicated such an insignificant difference in results that it wasn't pursued.General Motors Research Labs had come to the same conclusion from it's testing at Cal Tech in 1953.It would have been considered an absurd proposition in truck and trailer testing,due to wheel environment turbulence.Only open-wheel race cars were deemed worthy of the extra trouble/cost.

These views are now completely outdated for road cars. Any aerodynamics textbook published over the last few decades covers the major increase in drag associated with testing vehicles with rotating wheels. It's why all car manufacturers now use 5-belt wind tunnels with rotating wheels.

Originally Posted by aerohead

Today's F-1,Indycar,and NASCAR' billion-dollar racing programs mandate that 'ALL' areas of aerodynamics be investigated.
One curiosity always hiding in the room is that,given rolling-road wind tunnels,and million-dollar public relations budgets,and the ability to develop 'porous' low-drag wheelcovers,when tasked with setting land speed records,all auto manufacturers choose the totally sealed,convex disc,MOON wheel covers,the lineal descendant of WW-I era aviation; and can be seen on all extant Boeing 727s.

What have Boeing 727s, Land Speed Records, F-1, NASCAR and Indycar got to do with road car wheel design for lowest drag? The most recent tech research shows that, for road cars, fully enclosed wheels are not always best for low drag. I've already cited those papers here in another thread. It's an excellent example of why not keeping up with the literature means falling into the trap of "it's always been like this" - when, maybe, it isn't any more.

[aerohead]

Originally Posted by JulianEdgar

I did not say that all new data disagrees with old, I said that old data should be treated with suspicion when the current evidence suggests that our understanding has changed.



I am quoting data from Stromlinienautos in Deutschland (Kieselbach), P 84 for K1 - K4 and Aerodynamics of Road Vehicles (Hucho) (P 21) for K5.

As is typical with cars from the 1920s and 1930s, when tested in a modern wind tunnel, they all recorded much higher drag figures.

For example, you mention K5 as having a long afterbody as 'secret for its low drag'. It had a measured Cd in a modern wind tunnel of 0.37! So much for its secret. As I already mentioned, K3 was also measured in a modern wind tunnel at 0.37.

They were fascinating for the time, but to bring up these old cars as if they have anything to teach us today is just absurd.



Yes, I agree - so to use data relating to NACA aerofoils for shaping normal road cars is treading on very thin ice indeed. One obvious discrepancy is the thickness of the boundary layer on the rear parts of real road cars versus an aerofoil.



Yes Jaray was a very good aerodynamicist, and refined his craft working on the aero of Zeppelin airships. But his pure shapes are a long way from practical road cars.



These views are now completely outdated for road cars. Any aerodynamics textbook published over the last few decades covers the major increase in drag associated with testing vehicles with rotating wheels. It's why all car manufacturers now use 5-belt wind tunnels with rotating wheels.



What have Boeing 727s, Land Speed Records, F-1, NASCAR and Indycar got to do with road car wheel design for lowest drag? The most recent tech research shows that, for road cars, fully enclosed wheels are not always best for low drag. I've already cited those papers here in another thread. It's an excellent example of why not keeping up with the literature means falling into the trap of "it's always been like this" - when, maybe, it isn't any more.

*The suspicion issue has been reported on for decades.
*Since,as of Hucho's earlier writing,and specifically to good agreement of drag coefficients for specific cars tested in modern tunnels,compared to original literature,only a fair assessment would be derived from those very same shapes being tested with rotating wheels and the variability published.

*Kieselbach may have it wrong.It may have been
Ludvigsen who reported on the confusion of K-car nomenclature.
*The Landsberg (sp?) Castle car tested by VW is not the K-5.It's Cd 0.37,in light of its mutilated belly pan,would not be germane to the K-5.
*The Schl'o'rwagen,at full-scale,if I remember correctly,returned a Cd very close to that of early reporting at Gottingen,compared to the VW tunnel,with static wheel testing.
* Hucho has corroborated Hoerner's numbers.
* Horner has corroborated NASA's (NACA) numbers.
*Which all support Jaray's Zeppelin numbers.
*Jaray's small car returned lower drag in more modern testing than originally reported by Klemperer.
*Modern testing of 'typical cars' of old is not germane to modern testing of 'specials'.We need specificity.
*As to the K-5's 'secret,' the entire premise of much of the FKFS research dealt with body length vs drag.'Verhungungsverhaltnisse'.They corroborated Walter Lay's research,who's research corroborated Jaray's,along with Elliott Reid at Stanford,and others.
*I would want to tread very lightly when implying that the older research has nothing to offer us.I smell confirmation-bias,and that has no place in science.
*Aircraft,by design,are separation free at 'flight' conditions,and are all ruled by surface friction drag,of shock-wave drag depending on Mach#.It's really not germane to road vehicles,of which drag is fundamentally a function of flow separation.
*'Practical' is not a scientific metric.It cannot be quantified.It's very subjective and only an arbitrary invention of the human mind.
*If truly low drag is the topic,then body length is the arbiter.It's not negotiable.Physics as it is.
*And again,FKFS gave us the template for active aerodynamics in the mid-1930s which would address the 'practical' length argument.
*If outdated,can you explain why contemporary record vehicles continue to borrow from near-century old technology?
*I have your book.Your sources haven't proven anything,nor made a case against full-coverage,convex wheelcovers as an all-emcompassing universal absolute.What your messengers have published is highly contextual and dismisses an environment in which even lower drag can be attained if automakers weren't enslaved to the Paris dressmaker.
*Ignoring brake cooling,if porous wheels have superior low drag characteristics compared to full-coverage,convex discs,why don't we see them when land speed records,or fuel economy records are set?

[JulianEdgar]

Sorry, but I honestly can't make any sense of a lot of what you have written. So I'll take up just the one point.

If you wish, you can dismiss technical papers that don't agree with your premise that full wheel covers are always best for drag. You can even drag in odd statements about Paris dressmakers. I, however, like going on the evidence of peer-reviewed technical papers.

The current technical literature unambiguously shows that full wheel covers are not always best for low drag in road cars.

I don't like seeing people being given incorrect advice. It wastes their time, energy and enthusiasm.

[aerohead]

Originally Posted by JulianEdgar

Sorry, but I honestly can't make any sense of a lot of what you have written. So I'll take up just the one point.

If you wish, you can dismiss technical papers that don't agree with your premise that full wheel covers are always best for drag. You can even drag in odd statements about Paris dressmakers. I, however, like going on the evidence of peer-reviewed technical papers.

The current technical literature unambiguously shows that full wheel covers are not always best for low drag in road cars.

I don't like seeing people being given incorrect advice. It wastes their time, energy and enthusiasm.

The conclusions drawn in the peer-reviewed technical papers are contextual.I believe you touch on that in your book.I don't advocate driving with MOON covers,in light of brake cooling issues.I've done that and devolved to a porous cover which accommodates cooling air to the calipers and drums.
And on a case-specific-basis I'm okay with a messenger reporting on research of a specific vehicle, in light of all that vehicles design particulars.But that's as far as I'll go.
Automakers are beholden to the stylist.Aerodynamicists may be given a little latitude to best optimize a wheel cover as long as it's in keeping with the spirit of the stylists vision.It's true of any feature on a car.Some committee may accept the data from the wind tunnel,and then base a new car design solely on aesthetics.Volkwagen has a chance to sell a Cd 0.31 Golf/Rabbit in 1975,but couldn't because of contract constraints with Georgetto Giugiaro.We had to wait decades before we could get a Cd 0.31 VW,and thousands died in battle in the meantime,over the difference in oil to power those high-drag VWs. There's more at stake in aerodynamics than academic arguments.
I didn't give any advice that I know of without caveats.

[JulianEdgar]

Originally Posted by aerohead

The conclusions drawn in the peer-reviewed technical papers are contextual.I believe you touch on that in your book.

Well, of course. But 'contextual' is the whole issue here: the context is that in some cases, full wheel covers increase drag. Isn't that the point?

Originally Posted by aerohead

I didn't give any advice that I know of without caveats.

I am sorry, but I have seen lots of advice here - including from you - that is stated without any caveats whatsoever. Here's an example that was being used in a signature:

"At 55 mph, a 10% drag reduction translates to a 5% increase in fuel economy. At 70mph,a 10% drag reduction translates into a 6% increase -Phil Knox (Aerohead), Aerodynamics Seminar #2"

No caveats there.

(Nothing on what proportion of total drag is made up by CD vs rolling resistance - which is car-dependent? Nothing on the change in engine efficiency - BSFC if you like - as the rpm changes? For a discussion of this sort of thing, see Barnard, Road Vehicle Aerodynamics Pages 54-55, and Stone, Motor Vehicle Fuel Economy, Page 136.)

[freebeard]

[time for some popcorn. It's gonna be epic]