I know the team owners wanted to keep the 3 highest paid employees and drivers out of the budget cap - which, of course, meant it wasn’t and isn’t a budget cap at all but they have screwed themselves IMO. Once the good drivers suss that they can basically ask for as much as they like all that potential profit will go the way of the driver with the best negotiating position. Already experienced engineers are being screwed on salary or replaced by know-nowts because they are inside the budget limit, so we have a situation now where th most important members of the team are screwed and the ones that matter less are cashing in. The beginning of a long downward spiral for F1 probably. If I were Gene Haas I’d cash in at the top of the market for team value.
I agree.
The budget cap has resulted in fewer experienced F1 employees in teams with them being replaced by a few new graduates to save money over a period of time. Not surprisingly the team’s capability has reduced on average across the grid. Meanwhile the team’s value has increased, drivers are paid more and the team owners and 3 most highly paid people are paid even more.
Even the new graduates salary has dropped lower than in the city and at some consultancies so F1 is not attracting the best graduates now.
On the subject of Andretti whilst their exclusion is due to the greed of the people currently fucking F1 it is true that they would probably struggle hugely, probably more than Haas. When I was helping out Lola I went to the US often to support CART customers and the very best of them had about 10 to 20% of what is necessary to build a F1 car and, IMO, pretty weak understanding of the conceptual requirements of a racing car as opposed to data on setup stuff. If they manage to get some good people and build a car and test it whilst not in F1 they will learn what they really need - I am pretty sure they don’t yet - and if they use the testing cleverly they can learn fast since the actual teams aren’t allowed to do much testing and the 2 most important parameters are missing or bodged in the simulators so running a car could be a really big advantage to them over existing teams and lift 2 fingers at them. I bet they won’t though.
I would mostly agree with this.
Being outside the rules should have been by used by Andretti to do what you have suggested (build a representative car to test and gain all that wonderful experience as well) to get the data that they need to really understand what they need to develop. Once they have that data from a high quality wind tunnel and the track, then they can ensure that the simulator gives them what they need.
The biggest problems are that you can never get real car aero data from a wind tunnel because of aero elasticity. It gives an approximate quasi static bit of data but it is wrong at every speed since relative movements are rarely of negligible effect and things like floor and front wing deflection are big and, if modelled at all, only valid for one car speed. The only question is how wrong…
The second problem is friction coefficient, I have no idea how much better the thermal models are than they used to be but unless you can accurately calculate the temperature at every stage of the grip limited phase for the circuit, and have an accurate plot of temperature against friction coefficient the two most important grip parameters are not accurately known.
Which is why limiting circuit testing is such a catastrophe for young engineers and drivers with a lot to learn IMO.
In the end you need to use the tools to provide you with valid directions. There will never be absolute accuracy of course, but the tools I have led the development of, have provided very effective directions to work in and have helped win many championships.
In the end, you need both track testing (to provide data and experience), along with simulations and test rigs. But as the rules are limiting track testing more and more, the race car has to be instrumented more and more with all the associated limitations.
Yes, porpoising was a classic demonstartion of the inadequacy of simulation for certain aspects of performance leading the less experienced teams (ie pretty well all of them) going into the season 2 years ago completrely “blind” and shocked to get a massive problem none of the test rigs or programmes could predict.
I had experiernce of what causes it how to solve it from 40 years ago and it was by no means only the easily fixed and much publicised at the time and since Peter Wright theory.
It was absolutely no surprise to me that most teams had a problem and took a long time to solve it, if they ever did.
That is as much a problem of no track testing as the absence of data for adequate simulation.
If you watch front wing movement and tyre sidewall dynamic deflection on an actual car being driven round an actual track with bumps, warp and kerbs and you know how much difference they make to a quasi static model it is scary
How much of this is just the limits of what hardware can actually model in a reasonable timeframe? Versus a failure of the teams to actually use the resources they have?
I know the likes of Boeing do some insane levels of modelling, but then they’re a rather bigger enterprise making things that fall out of the sky if they fuck up.
Aeroplanes are aerodynamically trivial compared to an open wheeled car running along the ground so much easier to model.
About 50% of the aero in an F1 car is reducing the damage done by the wheel wakes, which 'planes don’t have at all, and getting the downstream aero effect of things upstream to have least effect on the crucial downstream aero parts.
Having a stable rear grip is critical to driver confidence and the front wing, its end plate and particularly the front tyre wake are all trying to screw it (with great success usually it is very, very difficult to prevent/limit)
The wind tunnel data comes from a fairly rigid model which may or may not been made to some calculated deflected shape which, even if correct, which is unlikely, is only valid for one speed and ride height, the elastic deflection of the floor and front wing in particularly is big enough to have dominant effects on balance and overall downforce.
Typically the model is tested over a range of ride heights, yaw and steering angle all of which have a big effect on rear stability but its aero shape is only valid for one car speed and height - as soon as a car is going faster the aero shape of the underside is different even at the same ride height than that tested.
As i wrote earlier it is a question of how wrong? and IME it can be very wrong - porpoising isn’t the only surprise I have seen, rate of change of steering lock when correcting oversteer had a big enough effect to cause the car to spin on one car I dealt with, easy to see the problem, not so easy to measure or fix…
The modelling I’d seen was airflow with the landing gear out, but still wasn’t requiring modelling to incorporate the ground, so maybe halfway between those two scenarios I’m guessing.
I realise I also don’t know what they were doing that modelling for exactly - if it’s just to stress check some of the landing gear components, that feels appreciably different to car handling.
I used that example because when I saw it it was being touted as the most complex fluid dynamics modelling that had been done to date, in terms of calculations done I think.
As to tyre modelling, there was a wide variation across the grid.
At BAR Honda we developed a thermo mechanical model in 2003 when it was clear that Michelin were pulling out. We worked with Michelin and another company to embed thermocouples in the tyre build and so were ale to get validation data for the surface temp through a multi array, temp just below the tread and temp on the inside of the tyre with another mutli sensor array. The tyres only lasted a few laps but gave us invaluable data. That model was developed over the years through rewrites etc and in 2011 was used as a multi rib model (good for ribbed tyres but also good for slick tyres as it was a simple finite element approach that worked extremely well). With individual ribs modelled in different threads of a fast processor, thereby enabling use real time for tack data and for the simulator.
There were many other things that we did as Loic Serra (Vehicle Performance Director at Merc and now of to Ferrari) was in charge and as an Michelin engineer he knew a lot about tyres and tyre modelling and how to get round the legal limitations on what you could do with tyres.
Williams were the worst that I came across.
Sauber had a very Michelin like tyre model (developed y 2 ex Michelin modelling engineers) but it could only be used off line (accurate but processor and accurate material property intensive) and so they developed a more lightweight model for use on the simulator and in lap time simulation.
NASCAR etc really struggled in this area and so a few years ago were using tuned Pacejka 6.2 data sets, which I never liked.
I didn’t see one that could calculate the growth in tyre temperature as ther car braked so was trustworthy at turn in, nor one that predicted the whole tyre contact patch area temperature and look up Mu from a tyre/temp/friction table, or predict if the overheated or got a balance front to rear.
We all know nothing makes grip like having optimum tyre temperature on all 4 wheels and lack of that, the most important single grip parameter, and knowledge of downforce, the second most important, major non transferable fudge factors were needed to get a match.
But I retired 13 years ago and the latest I saw was Toyota which was Michelin engineer driven. Even if the temperature could be reliably predicted the temp/Mu curve seemed to be pretty well unknown.
Simulations have been brilliant for targeting where to look for performance but widely expected to do things they don’t have the data to achieve - or didn’t anyway.
The models were more advanced than that when I left. And yes it modelled turn in well.
The big mistake people made was to use look up tables for Mu (generic data and you need specific), whereas we modelled the physics of the contact patch and the compound grip and the tarmac surface properties.
Liam Lawson to Mercedes for 2025?
If it isn’t a consideration, it damn well should be and Mercedes should be doing whatever they can to get him there!
(Leaving aside all the Red Bull machinations to keep drivers in their harem…)
