Pat Sy­monds on the value of aero­dy­nam­ics

F1 Racing (UK) - - CONTENTS - @F1rac­ing _mag face­book.com/ f1rac­ing­mag PAT SY­MONDS

One of the sur­pris­ing as­pects of this year has been the as­ton­ish­ing lack of per­for­mance from Wil­liams and Mclaren – and, at the be­gin­ning of the sea­son, Force In­dia. Peo­ple of­ten won­der how, with the in­cred­i­ble ar­ray of tech­nol­ogy avail­able to the teams, per­for­mance drop-offs like this are pos­si­ble. They’re not just con­fined to the lesser-funded teams; in 2015 Red Bull pro­duced a car that didn’t be­gin to show the team’s nor­mal po­ten­tial un­til mid-sea­son.

So what can be the rea­son? The first thing to re­mem­ber is that a For­mula 1 car needs to be thought of in sys­tem en­gi­neer­ing terms. Per­for­mance isn’t a func­tion of any sin­gle as­pect of those in­ter-re­lated sys­tems. Wise choices at the de­sign stage, to­gether with a full un­der­stand­ing of the in­flu­ence of cer­tain pa­ram­e­ters on lap time and straight line speed at dif­fer­ent cir­cuits, pay div­i­dends through­out the sea­son.

While all de­sign vari­ables in­flu­ence each other, there are cer­tainly some that have a greater con­nec­tion to over­all per­for­mance than oth­ers. The three most in­flu­en­tial over a range of cir­cuits are power, grip and aero­dy­namic ef­fi­ciency. The first two, to some ex­tent, can be com­pared across cars. There are only four types of power unit and, with the ex­cep­tion of Honda’s, they are used in dif­fer­ent chas­sis. This can rule out gross dif­fer­ences in per­for­mance in this area. There will al­ways be some dif­fer­ences as­so­ci­ated with the in­stal­la­tion of the power unit in the car, but these are very much sec­ond-or­der in­flu­encers.

Sim­i­larly, all the cars are us­ing the same tyres, which gives a de­gree of per­for­mance nor­mal­i­sa­tion – es­pe­cially on a sin­gle qual­i­fy­ing lap on new tyres. While it’s true that the cur­rent tyres are very sen­si­tive to op­er­at­ing con­di­tions, when an­a­lysed sta­tis­ti­cally over a num­ber of events, a pat­tern should rapidly emerge.

This leaves aero­dy­nam­ics as the most likely cause of dis­par­ity be­tween ex­pected and ac­tual re­sults. But surely the so­phis­ti­ca­tion of the tools these days should put paid to that type of con­cern? Oh, that it were so.

The aero­dy­nam­ics of an open-wheel rac­ing car are im­mensely com­plex. Much of the flow is tur­bu­lent and this makes it more dif­fi­cult to pre­dict us­ing Com­pu­ta­tional Fluid Dy­nam­ics (CFD), and even a wind­tun­nel, since there are many as­pects of the flow that are dif­fi­cult to model.

Both CFD (in the form com­monly used by teams) and wind­tun­nel test­ing op­er­ate in a time-av­er­aged do­main. What this means is they ef­fec­tively mea­sure forces or con­di­tions which will be vary­ing over a pe­riod of time, and then take an av­er­age re­sult over this time pe­riod. In iso­la­tion this can give a very poor pic­ture of ac­tual per­for­mance.

Let’s take a sim­ple case where the down­force is vary­ing in a sim­ple, rel­a­tively slowly re­peat­ing pat­tern, like a sine wave with a vari­a­tion that goes both 10 per cent above the av­er­age and 10 per cent be­low. The aero­dy­nam­i­cist will re­port the av­er­age num­ber but the driver will only be able to ex­ploit some­thing around the min­i­mum down­force, which is maybe eight or nine per cent be­low the av­er­age.

Now con­sider a car that has nom­i­nally the same down­force, in other words the av­er­age fig­ure is the same, but now the vari­a­tion is plus or mi­nus 15 per cent from that av­er­age.

The driver once again will drive to some­where just above the min­i­mum, say 13 or 14 per cent, which will give him con­sid­er­ably less down­force than on the pre­vi­ous car, even though the re­ported num­ber is the same. In fact it may be worse than this be­cause the sec­ond car, hav­ing more vari­a­tion, will un­doubtably give the driver a feel­ing of se­vere in­sta­bil­ity, dis­cour­ag­ing him from find­ing the limit.

In ad­di­tion to these sort of prob­lems, it is com­mon to hear teams re­port­ing that they have lost the cor­re­la­tion be­tween their wind­tun­nel and the car on the track. Most peo­ple as­sume this means that the aero­dy­namic mea­sure­ments made in the wind­tun­nel are dif­fer­ent to those mea­sured on the car. There is an in­fer­nal tri­an­gle con­sist­ing of CFD re­sults, wind­tun­nel re­sults and car re­sults, and none of them give a true an­swer. One might think that the car must be the best mea­sure­ment but un­for­tu­nately a rac­ing car isn’t a sci­en­tific de­vice, and by in­stru­ment stan­dards the mea­sure­ments are some­what crude.

On top of all that, the aero­dy­namic forces ex­pe­ri­enced by the wheel, be they the lift of the ro­tat­ing wheel or the down­force pro­duced by the brake duct winglets, are dif­fi­cult to

mea­sure be­cause they aren’t pass­ing through the sprung mass.

So what is good cor­re­la­tion? The an­swer lies not just in be­ing able to mea­sure num­bers that when fed into a sim­u­la­tion will give a rea­son­able pre­dic­tion of lap time, but more im­por­tantly a sit­u­a­tion whereby trends can be fol­lowed with con­fi­dence. If the flow field pre­dicted by CFD or even that ex­pe­ri­enced in the wind­tun­nel is some­what dif­fer­ent to that seen on the car, then it be­comes ex­tremely dif­fi­cult to pre­dict the ef­fect of changes.

When Wil­liams had a prob­lem af­ter fit­ting a new rear wing at Sil­ver­stone this year the prob­lem didn’t lie in the rear wing it­self, but in the flow de­tach­ing from the floor when the DRS was used and then not re-at­tach­ing when the DRS closed again. Not only was this not de­tected in the wind­tun­nel, it also wasn’t de­tected dur­ing prac­tice the pre­vi­ous day.

This is a mea­sure of just how crit­i­cal the aero­dy­namic per­for­mance can be to the con­di­tions un­der which it per­forms. When I was at Wil­liams we even used to clean the lead­ing edge of the rear wing in pit­stops be­cause the dead flies it col­lected could ad­versely af­fect its per­for­mance.

Good cor­re­la­tion ex­ists when any trend seen on the car is pre­dicted by the ex­per­i­men­tal tech­niques used to de­velop the car. Eas­ily said but not eas­ily achieved, and the more com­plex and crit­i­cal the flow fields are that de­liver per­for­mance, the more dif­fi­cult that nir­vana of cor­re­la­tion be­comes.

WHEN I WAS AT WIL­LIAMS WE EVEN USED TO CLEAN THE LEAD­ING EDGE OF THE REAR WING IN PIT­STOPS BE­CAUSE THE DEAD FLIES IT COL­LECTED COULD AD­VERSELY AF­FECT ITS PER­FOR­MANCE

Mclaren and Wil­liams have stug­gled this sea­son, but the rea­son is not nec­es­sar­ily as sim­ple as a lack of fund­ing

Flow visu­al­iza­tion paint is one way of try­ing to mea­sure air flow on the car it­self

Good cor­re­la­tion be­tween wind­tun­nel and the car on track means that aero­dy­namic trends can be fol­lowed with con­fi­dence

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.