flow attachment in rapidly expanding diffuser

Here are our CFD links and discussions about aerodynamics, suspension, driver safety and tyres. Please stick to F1 on this forum.
bigpat
19
Joined: 29 Mar 2012, 01:50

Re: flow attachment in rapidly expanding diffuser

Post

Blaze1 wrote:
bigpat wrote:Yes you are correct that the IMSA cars were front grip limited, but this was a good thing because of the frightening corner speeds they could achieve, understeer is always preferred!!!! The Jaguar XJR-14 with the outrigger bi plane front wing, was the best attempt back then to achieve a better front balance. If this was combined with the front floor/sidepod treatment used by LMP's in the Toyota GT-One and latter LMP's then the downforce and efficiency would have been truly astounding....

As for flat floor vs diffuser area, the flat floor was forced upon designers in the rules, and generally they will look to a bigger diffuser to drive it, if it can be kept under control, sensitivity wise. Given free reign, and they wouldn't have them, rather running fully contoured tunnels like the ground effect cars, and IndyCars back in the day.

In order to control boundary layer, and to keep the air energised, we used to fit small vortex generators about 30mm back from where the diffuser commenced on F3000's. Our diffusers were quite aggressive, with the first section actually having a reflex curve away from the flat floor, rather than a gentle curve. Using dye to visualise the airflow, we stopped separation for a further 250-300mm along the diffuser. On a high speed track ( Phillip Island, AUS), we were able to back the twin element rear wing off 1 hole, which was worth at least 2-3 km/h at the top end.....
Pat, your comment about the front floor design in prototype cars got me thinking and I was just about to write how the following drawing of the Toyota GT-One's front diffuser perhaps supports your remarks:
http://www.mulsannescorner.com/ToyotaGT-One-TM1.jpg
There is no flat section ahead of the diffuser in the picture, however:

http://www.mulsannescorner.com/BenzCLR2.jpg
In the above picture of the 1999 Mercedes-Benz CLR LM front splitter/diffuser there is a flat area of floor before the front diffuser starts. Mulsanne Mike states: "Note that the nose diffuser is relatively conventional, not as agressive as the raised nose Toyota GT-One. The diffuser appears to be planar, transistioning from the flat splitter to the flat diffuser kick across an edge."

By less aggressive, I take that to mean the nose area of the CLR isn't raised, which would have given the diffuser greater span and subsequently volume. This makes me wonder why Mercedes went for such a design (flat section ahead of diffuser).
The cars I used to play with.....
That's a nice car. I remember watching Alonso thrashing a similar one around Monza on Eurosport many years ago. :)

BTW, what other changes could be made to the diffuser (I assumed the car was run in a spec series) and is 'one hole' the minimum config change?
I think by stating not as aggressive, I believe it meant in downforce potential. As Le Mans is a low downforce track, many assume simple, and tried solutions. The typical nose diffuser would likely produce less drag than the GT-One concept, but more pitch sensitive. More on this later...

The GT-ONE front end concept was a game changer, but it must be remembered that it was originally designed to suit loopholes in the rules after watching what Porsche and Mercedes produced the year before. Originally 5 production cars had to produced for homologation. This dictated wide footwells for 'passengers'. Toyota saw that only 1 production car actually HAD to be produced, to demostrate raod rules compliance, and decided that it didn't need to worry about the narrow footwell that the raised/slim nosebox it produced. ( the smartest thing they did, being to satisfy the luggage room requirement by claiming the fuel tank area was the luggage trunk!)This was missed by others, most notably Mercedes who not present at Le Mans in 98, concentrating on the FIA championship.

A couple reasons the Merc concept didn't work, and ultimately gave it wings.....

As seen in your pics, the Merc diffuser spills into the front wheel wells. Traditionally louvres on the top help relieve high pressure air within, and extract the air from the diffuser with it.... but with a drag penalty. To compensate, the louvres are blanked off, for low downforce configurations. The 98 CLK GT 1 ran with louvres. Remember the wheel arches by nature of their shape try to produce lift. So now with them blanked we have less load on the front axle. Of note is that Mercedes had a concept that served them well enough the year before to be the class of the field in the FIA series, and understandably felt they comfortable that they understood the concept well enough...

Second, the diffuser would not be a straight angled ramp, there would be contour to it, (possibly small reflex curve at the base ) though not evident from that picture angle. Having a flat front leading edge to the nose it good for downforce potential, but once it grounds over bumps, it stalls the feed to the diffuser. The Porsche 911 GT-1 was the same, and suffered the same fate, flipping over a crest like the Merc. This requires stiff springs in heave to keep the nose off the deck.

Finally Mercedes exploited long front overhangs for more area from which to produce downforce. This meant the front edge of the nose moved up and down more than the wheels did, increasing pitch sensitivity. In the tens of thousands of km's of testing conducted by Mercedes, nearly all was on high speed, smooth tracks/ ovals. Come Le Mans test weekend, with the bumpy nature of the circuit, the cars were required to be sprung heavily to keep nose off the deck for it wouldn't stall. The cars could be seen on footage having a high frequency 'nodding' tendency, bouncing on the tyres sidewalls due to stiff springing to tie the nose down.

After the test weekend, Mercedes were in a panic, with their relative lack of pace. It is rumoured they turned to Williams F1 team, and Adrian Newey in particular to come up with a quick fix. The dive planes fitted to the noses after Webber's flips were one of the fixes for race weekend. The fact that Mercedes claimed they increased front downforce by up to 25% shows just how little downforce the front end had, when designed to run a t Le Mans.

In the end, the nose lifting over humps, and an angle of attack of anything over 4 degrees, resulted in the large flat underbodies catching the air, and flipping the cars.

The Toyota GT-One, like Porsche 956/962's had a radiused entry to the front floor, with the choke point back away from the leading edge, which would help to reduce pitch sensitivity.

As for the F3000's we ran, yeah they were great to work on. It wasn't a spec series, as we had 95 Euro cars, fitted with Formula Nippon parts, and custom parts too. There was a 93 Reynard with 94 floor, sidepods and gearbox, 97 front wing etc. we called it 'Frankenstein'. It was really cool to be able to try different things. we played with:
Monoshocks
Third dampers /springs
Suspension geometry
Different wing configurations, front wing turning vanes
Winglets, bargeboards, diffuser strakes, in house carbon parts etc.
Spark cuts, downshift blippers etc.

Great times, we learned so much, and were able to be creative. And funnily enough, most of it wasn't expensive to do, just time. Sadly, spec series cars stifle this creativity.

Hope this didn't bore you!

OO7
OO7
171
Joined: 06 Apr 2010, 17:49

Re: flow attachment in rapidly expanding diffuser

Post

Cheers Pat.

I remember reading comments about the Toyota luggage/fuel tank loophole, but I wasn't sure if they were true or not. I thought someone was just exaggerating to demonstrate how smart engineers circumvent problems. :)

There were an number of issues that appeared to contribute to the Mercedes CLR back flips as you mentioned. The Mulsanne's Corner article states that soft rear springs may have exacerbated the porpoising, but explained that such reports may have been purely anecdotal. From video footage when compared to other cars such as the Toyota GT-One and BMW LMR, The Mercedes CLR certainly did appear visually to be softly sprung at the rear. Unlike the other cars, it even appeared to run a slightly lower rear ride height than the front through some of the turns and over some of the curbs such as the Dunlop Chicane section.

BTW your post was far from boring. It is always refreshing to learn a different perspective and or new details on various topics, so I appreciate your responses and the info you provided on your time working with F3000 cars, that must have been a lot of fun! :D

Cheers

bigpat
19
Joined: 29 Mar 2012, 01:50

Re: flow attachment in rapidly expanding diffuser

Post

Yes they were fun! Taught me lots of lessons. Opened a number of doors for me, in many different categories. I still have fun working on all sorts of cars, old and new, and always learning something!

Thanks for the kind words too! :D

OO7
OO7
171
Joined: 06 Apr 2010, 17:49

Re: flow attachment in rapidly expanding diffuser

Post

I just watch the following video and the first thing that came to mind was 'Giant diffusor'.
http://youtubeonrepeat.com/watch/?v=Ieg ... =70&to=140