Under the skin of F1 testing
Pre-season testing is fraught with tension for the Formula 1 teams. Here Gary Anderson explains how they prepare for and carry out the initial running of their new cars – fingers crossed, of course
Pre-season testing for the latest batch of Formula 1 cars is just around the corner. Since this is the second year of the current technical regulations, we can expect the car concepts to converge. The aerodynamic and tyre changes for 2017 were fairly dramatic and they brought roughly what was predicted in terms of lap-time improvement. This year, optimising that package will be the order of the day. During the season, there is always development and most teams will try for around one tenth of a second of improvement each race. That should add up to a 2s gain over the season. That, combined with the introduction of new chassis for 2018, with all the developments it wasn’t possible to add last year, means that at pre-season testing the cars should be 3s faster than they were last year. It’ll be very interesting to see which teams achieve that.
Most of the big teams have the infrastructure to manufacture components much more quickly than was possible a few years ago. This is what some have concentrated on, because once a drawing is printed or a machine is fired up to make a part, that component is already out of date. The later that everything comes together in the build schedule, the more time it allows for pre-build development.
Since testing now gets under way later in the year, teams have extra time for designing, optimising and manufacturing components. But the start of February will have been D-day for final assembly of the 2018 test package. Before this, hopefully the chassis will have passed all the FIA crash tests. These are high-pressure moments and you want to build everything to be as light and efficient as possible, while still meeting strength and stiffness requirements. One failure can set a team back dramatically.
Recovery from a failure will either mean a redesign and remanufacture of the chassis itself, which will take time, or the remanufacture of components by adding weight. Neither solution is easy or optimum, and it will mean that everything moves closer to when the cars must be shipped to Australia for the first race. Every component on the suspension side of things will also go through strength and durability tests. Setbacks here can also mean delays that will interfere with the build schedule.
The front and rear wings will also be tested by loading them up with hydraulic actuators. This is to simulate the FIA deflection tests to ensure these components are stiff enough to comply, and they will then be loaded up to potentially their maximum load to see if they will withstand that force. There will also be a safety factor of something like 20%. When you consider that a front wing at high speed will produce something in the region of 600-700kg of load, this is not easy. Combined with some aerodynamic stall characteristics, which will potentially introduce fatigue load, the problem is doubled.
Many teams, including Mercedes and Haas, have had problems in the past few years with front wings falling off. There is nothing that gets the driver’s attention quicker than bouncing over their own front wing, so you want to err on the side of safety with the first batch of components!
Many of these tests can be completed back in the workshop before testing begins, but there’s no substitute for circuit
mileage. There the car will get real use bouncing across kerbs, being subject to instantaneous changes of forces on all the suspension and steering components, and real-time engine and gearbox transient-load fluctuations.
The first running is what is normally called a ‘shakedown’. This is just to make sure that all the systems work and that the sensors are supplying the correct data. Normally, a new car will be built with a few extra sensors to help understand the water, oil, fuel and hydraulic systems better.
Sometimes shakedowns happen in parallel with the car launch. The presentation and the time that the marketing side of the company requires will eat into on-track running time and this can can be frustrating. But worse still is when a team has an independent launch somewhere with a bit more glitz and glamour. This can really impact on the car-build programme and normally the car ends up at the launch in a less-thanfinished state. Neither solution is very productive, but it’s important for marketing and the sponsors, so it’s something that the engineers have to live with.
Once a team is happy with the shakedown results, it does a fuel-pick-up test again to make sure it can run very low fuel levels. It will also do this again with the oil level to make sure there are no oil-pick-up problems that could cost the team an engine.
Getting the best performance and durability from a new car package is a lot easier than it was. Now the simulators are so refined, teams can test many different combinations of set-up before they hit the track. The test rigs mean that some of the reliability issues will be addressed before the car runs on the track, but there is no substitute for real-world mileage.
The first thing you want to get under control is simply being able to start the car and for it to leave the garage at the pre-set time. It’s amazing how often that doesn’t happen. Everyone needs to be ready at the same time, from the tyre man to the control-system engineer to the driver.
After that, it comes down to piling on the miles and making sure that you don’t stop testing until the end of the day. It’s
normally the little things that bite you – something getting too hot, or wires or pipes chafing on something, or bodywork fastenings coming undone.
The other thing a team will want to test is what happens if certain sensors fail. There will be plenty of default parameters within the control-system strategy that will hopefully identify a failed sensor. This could be clutch, gear or throttle position, oil pressure, or any number of temperature sensors.
Normally, the strategy is to use two or three inputs from other sensors to check if it’s an actual failure or a sensor problem. This all has to be done in milliseconds, otherwise there could be engine or gearbox failures. The systems also need to be proven as potential back-ups that can get a driver to the chequered flag if necessary.
Instead of heading into the first test with your fingers crossed that everything works as planned, you now have a fair idea of how the car functions with different set-up changes and what its performance profile looks like. But running on the track will always throw up questions.
As we often say, the driver feels the car through the seat of their pants, so the first thing is to make sure they are comfortable and can get those feelings without distortion.
If so, they will quickly know if the car is doing what the team has been telling them it should.
Lap time is always confirmation, but within a couple of reasonably quick laps they will know if the braking stability is there, if the car continues to load up progressively as steering lock is applied, and if traction is consistent off the corner. If not, they will also have a feeling of whether any negative characterstics are within set-up parameters, or there is something fundamentally wrong.
The experienced drivers have all been here before, and they’re as keen to get the feeling of those first few laps as anyone. They will know very quickly if it’s going to be a long, hard season.
No matter what the first feelings are, the teams will have a test schedule to try to get more from the package, understand it more, or see how the car responds to set-up changes. This will consist of a suspension-stiffness sweep around the set-up that their simulations have predicted as best. This will be something like +/-5% vertical-stiffness change, then a +/-5% roll-stiffness change. If either direction improves the performance, then they will pursue that with another step around the same magnitude.
Around the suspension-stiffness test, teams will also experiment with ride height. Red Bull ran quite a different ride height relative to most other teams last year, and I expect this to converge in 2018, but teams will still want to confirm that their development direction is correct.
When it comes to the aerodynamic package, teams are basically looking for confirmation of the loads predicted by the wind-tunnel and CFD. This is vitally important to allow a team to continue with its development direction.
To understand this, the teams will do a lot of aero runs.
These consist of running at a fixed speed in two directions to weed out variances caused by wind. Barcelona is pretty good for this since there are two almost parallel straights running in different directions.
Most teams will do this at the end of each timed run, and from that they can compare timed-lap aero data and driver feedback with the straightline aero data. Downforce is not vitally important on the straight but it is around the corners, so this is used to identify any transient or mid-corner losses of downforce induced with steering angle, yaw or roll.
They will also carry out a front-and-rear-wing sweep to confirm the changes that are predicted are there in reality.
When the front wings run close to the ground, they are prone to aerodynamic stall.
Confirmation of the characteristics of this stall is important to allow the teams to understand how to go about balancing
the car aerodynamically. Also vitally important is that the rear wing recovers instantly after using the DRS and the airflow reattaches to give the downforce back. If not, then the initial braking area can become a bit of a nightmare and they will be forced to close the DRS early, losing end-of-straight top speed.
A team will normally carry out these back-to-back chassis tests on what they would classify as the middle tyre compound. But, with Pirelli introducing a new range for 2018, some initial running will be needed to define which that tyre may be. This will be complicated by the fact that the Barcelona circuit has been resurfaced. The aim is for it to be similar to the old one, but a new surface takes time to bed in, and it will evolve a lot for the first few days in particular.
Gathering all this data on a soft or hard-compound tyre could very easily take you down the wrong set-up path, but you still need confirmation that the car can be balanced on these extremes of tyre compound, so the teams will normally run softer compounds in line with lower fuel some time during each day of the test.
When you get to the point of having what you believe is a reasonable set-up, you will have to start race runs to understand where the tyre degradation starts.
Normally, a team will try to simulate a race weekend with practice and set-up changes in the morning. Then they’ll do three qualifying runs before a two-stop race distance in the afternoon using two tyre compounds.
When you consider that all this running is being done in temperatures that are much cooler than the average will be at most of the races, car performance needs to be viewed with some scepticism. Hotter ambient and track temperatures usually lose downforce, but they also bring extra cooling requirements, which again cost downforce, so with a 20°C increase in temperatures you could have a downforce loss of around 5%. That doesn’t come back by fiddling with the set-up, but luckily it's similar for everyone.
The other thing that happens with increased ambient, but mainly increased track temperature, is that the aerodynamic stall characteristics of anything running near the ground alter. A full understanding of this is very important to allow a team to optimise ride heights when they get to hotter conditions.
During pre-season, it was always more important to me that the car balance could be tuned for each of the compounds, changes in temperature and different fuel loads, and whenever the driver went out on the track the balance was somewhere near what it was previously. If you can achieve this comfortably then you should be in a reasonable position when you get to the hotter climates.
Every team will have developments planned for the first race in Melbourne, but these will be in line with the initial car-research direction. Any change in that direction that comes as a result of initial testing is vitally important, since it could mean that already committed components are no longer required, or that a different path needs evaluating.
There is just under a month from the start of testing until the cars run in Australia, so there is not much time to recover if you have major problems. Within the design, there is normally room to recover from pursuing the wrong concept. Red Bull had to do it last year and managed to close the gap to Ferrari and Mercedes as the season progressed.
But it takes time and money to develop a new concept, so if you’re a big, well-funded team then recovery is possible.
But if you’re on a tight budget then it’s not so easy.