A legal but flexible T-Tray Splitter: The ‘See-Saw’ solution

For over a decade, the FIA have tried to reduce front wing performance by increasing its ride height. Moving the wing clear of the track for less “ground effect”, reduces the wings efficiency and handicaps downforce. When the major aero rules changes came in for 2009, the loss of the central spoon section and the smaller allowable working surfaces for the front wing, made getting downforce from it even harder.
Through out this period, teams have sought to gain front wing performance, largely by trying to make the wing closer to the ground. Either via flex or by altering the attitude of the car (i.e. rake). As has been explained before in this blog, the issue with making the front wing lower by raking the car is that the T-Tray splitter gets in the way. Teams have sought to make the splitter flexible to allow it move up and allow for a lower front wing.
to combat this the FIA have a deflection test to ensure the splitters are not flexing and that front wing ride height is maintained. In response to accusations about several teams splitters, at Monza last year the FIA doubled the test to 5mm of movement for a 2000 Newton (~200kg) load. Yet in 2011 we still see cars with a nose-down raked attitude and wings nearly scraping the ground. How can a splitter meet the FIA deflection and still flex on track? I have a theory for a splitter construction, that actually exploits the method of the FIA test to provide the splitter greater stiffness during the test.

Typically teams run splitters mounted to the underside of the monocoque. The splitter is often made from metal to act as ballast, with additional carbon fibre bodywork to form the aero surfaces. Beneath the splitter runs the Skid block (plank). Made to FIA dimensions the plank features holes for measuring wear.
The splitter is bolted securely to the underside of the tub by bolts and in some cases with a small strut at the leading to aid installation stiffness. Disregarding the strut, the splitter is effectively installed in a cantilever arrangement. The protruding section of splitter will need to bend upward when grounding on track or on the FIA test rig.

With a typical splitter, wear will occur only at the leading edge of the plank

With a car in a raked attitude, when on track the splitter will exhibit a classic wear pattern, the tip of the splitter will wear away in a wedge shape roughly equivalent to the rake angle of the car. During normal running, for cars with high rake angles its likely no other wear may take place along the length of the plank. If the car runs a front ride height that’s too low, the splitter will wear away leading to exclusion at post race scrutineering.

See-Saw Solution

A pivot half way along the splitter creates a 'see saw' effect

Rather than run a cantilever mounted splitter, my theory would be to run the splitter mounted on a pivot. Taking the length of the removable section of splitter, the pivot woudl need to be half way along its length. Which would be roughly inline with the heel of the monocoque. Not having any significant mount at the rear of the removable section of splitter would allow the splitter to pivot like a ‘see saw’.

With the 'see saw' splitter, grounding on track will bend the plank & create two wear spots.

Now when the splitter grounds on track, the leading edge will tilt up and the trailing edge tilt down. This ‘See-Saw’ effect, will allow a slightly lower front ride height as the splitter will be deflecting upwards. To achieve this the plank will need to flex, as the front section of plank must now be a minimum of 1m long, far longer than the splitter. The drooping trailing edge of the splitter will now make the plank contact the ground, leading to a distinctive wear pattern. Now having plank wear in two placed, beneath the splitters leading edge and the trailing edge. This will also have the benefit of spreading the wear over a larger area of plank and reducing the likely hood that the front inspection hole will be excessively worn. The fulcrum point need not be the overtly obvious pivot I have drawn and the entire exterior of the splitter could be covered in bodywork, which will have enough strength to keep the splitter in place when stationery, but deform enough to allow the splitter to ‘see-saw’. But in this guise the splitter will not have the strength to meet the 200kg load from the FIA test.  so how will it pass the test?

The current format of the FIA test, actually aids the pivoted splitter.

The FIA test is carried out on the multi functional rig that is used for the other regulatory checks. The car is driven up onto the rig and then steel pins protruding up from the rig, locate in corresponding holes in the plank. The sections of floor under the wheel are dropped away and the cars ~580kg (640Kg less driver) weight sits on its belly (the plank\reference plane floor).
Then a hydraulic strut with load and displacement sensors extends upwards beneath the front splitter. The 2000n load is applied and the deflection measured.

With a typical splitter the FIA load bends the splitter like a cantilever

For a cantilever splitter, the test tries to bend the splitter upwards straining on the bolts at its tail end.

For a 'see saw' splitter, the weight of the car is on one side of the fulcrum, making it harder to deflect the other end upwards

Where as for the ‘see-saw’ splitter the test tries rock the splitter, effectively trying to bend the splitter like beam about its fulcrum. But the cars weight is sitting on the tail end of the splitter, preventing the splitter tilting upwards. As long as the splitters beam strength is enough to meet the test, then it will pass. Being a long metal structure, it should not be hard to make the splitter strong enough.
So as the FIA tests the cars weight sat down on the splitter, it actually aids the splitters ability to beat the test. If the test were to apply the load to the splitter, when the car is supported on its own wheels and not its floor, then the car would surely fail the test.

The biggest flaw is this theory is the wording of article 3.17.5 which describes the test and the mounting of the splitter. But typically the FIA rules are both vague and overly specific at the same time. The regulation states that mounting between the “front of the bodywork on the reference plane” and the “survival cell” (Monocoque) must be not be capable of deflection. The definition of “front of the bodywork” might mean its leading edge, but might not incorporate stays further back along the car. Equally the design of the fulcrum need not be the pivot I drew, but a simpler solid but flexible part, that is not suspected to deflect.
As with all borderline legal parts, this would need to be carefully assessed against the wording of the rules. But where’s there’s ambiguity, there’s a chance to exploit.

The legal interpretation of the regulations not withstanding, this is a feasible solution.  The biggest risk to running it, is if the FIA change the test process without notice.  This could catch the team out, although normal FIA process is to warn the team and ask for the design to be altered and pass the test at the next event.  Thus unlikely to cause an exclusion or  ban.


Footnote: A team have asked the FIA for clarification on the use of this splitter construction with a view to using it themselves.  Charlie Whiting has made it clear it would not be and added that the deflection may now be altered to ensure the rules and test are not being exploited.

60 thoughts on “A legal but flexible T-Tray Splitter: The ‘See-Saw’ solution

  1. Why do you not work for a team??? Genius idea and brilliantly worked out.

    One complaint: when posting pics of cars, can you link the individual pic rather than the gallery? When on a phone it’s hard to make out which picture relates to your comment

  2. Interesting article, great concept if you could get away with it.
    Question on it, aren’t movable aero devices not allowed at all (Sans the DRS?) Or has this since changed?
    Also, how would like go in regards to stablilty say mid corner if you ran over the kerbing and it changed pitch upwards? Is it enough to un-settle the car at all?

    Once again, its a great concept, thinking outside the square and trying to attack the grey areas.

    • Moveable aerodynamic devices are banned, as is deflection of bodywork. But the FIA have set a precedent by saying if it passes the test, then its legal!

      In my personal view this solution would be illegal.

      • I’m sure they take this into account, but 200kg seems like a small amount for a car generating thousands of pounds of downforce. Does the 5mm of deflection at 200kg on a static rig take into account 900+ kg force at maximum turning speed? I.e. does 5mm at 200kg equate to some deflection along a curve at maximum cornering downforce?

    • Since all parts deflect under load, the question is just how much deflection that is legal, and how this deflection is measured? That will always leave some room for interpretation.

  3. movable aerodynamic device?

    If your explanation of the compliance of the splitter and wearable plank is correct, then the “floor structure” is flexible.

    The fact that the flexible floor passes the “test” seems irrelevant or just plain badly tested.

    • They can say that “flexible aero devices are illegal” but as there are no perfectly stiff materials suitable to make a race car out of – not to my knowledge anyway – so there will always be some amount of flex. To allow for this constraint of the physical world the tests have to allow for some flex. Add to that ambiguity of rules and regs and the teams find ways to exploit it.

      I think this is one of the problems with the current regs, they only encourage “illegal” innovation.

      • I would say there is a difference between the flexing of a bit of bodywork from a material that can never be 100% rigid, and the inclusion of a pivot to induce flexing in something that is supposed to be rigid,

        I guess it will come down to the FIA deciding which rule has precedence – The letter of the law saying no movable aerodynamic parts, or as long as it passes the test it’s legal. Strictly, I think that the pivot, if it is the real deal, uses the mechanics of the test to hide a real movable part.

        If it’s decided it is illegal, I hope they don’t punish Vettel though, even a McLaren Fanboy like me can appreciate the class act he has been this season.

      • @Steve Hopkins – I think we in agreement here. I’m not saying that I think that the pivot is necessarily legal. F1 has to allow for some flexibility given physics. They don’t have to have super ambiguous regs or tests that allow for such a solution to exist, but they do.

        So, given all that, and the strictness of the regs for more obvious design features, the only place left to innovate is by skirting the rules, exploiting the test procedures, etc.

        If it is a moveable device, and if it really is something against the rules, I would want it removed as it is unfair. At the same time, if the “illegal” or grey area is the only area where innovation is allowed, I say that teams should continue to pursue such innovation so that maybe the rules will get changed to allow engineers to work in the open.

  4. You’ve not said it explicitly, but I presume you are hinting that uneven wear patterns seen on the plank under Webber’s car at Monza might be explained by this sort of arrangement?

  5. An great “work around” the regulations Craig. My only thought is wouldn’t the arrangement you describe be acting like a second class lever (a bit like a wheelbarrow) from the back of the splitter to the rear of the car? Would you end up with the full mass of the car acting against the downward force on the back of the splitter? But presumably with a neutral weight distribution (or not far off under this year’s regulations) it would be fairly easy to ensure a minimum 200kg force acts on the back of the splitter from the car…?

  6. Brilliant. The more you write the more interesting it gets. I cannot imagine why you still are freelancer and a team has not got you in their ranks as tech director

    • The real question there being. Does Craig actually WANT to be a tech director? Its quite something different designing/overseeing design of the systems for a certain team and explaining that to the public as a journalist.

      I would not be suprised if some teams might occasionally want to make an offer not to publish to Craig, but I am very glad its not come to that.

  7. I never knew the FIA deflection tests were performed with the car resting on its belly and not on its wheels. I’m a bit surprised the FIA never realised this could create an opening to be exploited.

    • Resting on it’s wheels would make the amount of deflection more difficult to test as the team could simply say that their suspension allows for the car to move up more easily than others, etc…

      The only true solution is to mount the car to the ceiling, suspended above the test rig. With the suspension fully unloaded you could then push up on the splitter and see how much it moves.

  8. I have some doubts whether the described see-saw action would actually be possible.

    If you look at the photo of the damaged RB car, there appears to be a metal pin that connects the front part of the splitter to the monocoque. The pin, I assume, would prevent vertical movement of the splitter’s front end. Also, the plank would have to be not only mighty flexible but also easily stretchable to accommodate the vertical movement of the splitter’s back end.

    Sorry if there’s an obvious explanation of the above, I just don’t see how it’d move while being fixed like that.

    • A slank pin easily bends during compression loads, so it doesn’t stop the deflection upward, only downward. The plank also wouldn’t need to bend much, and I don’t see why it would have to stretch at all?

      • This is my proposed theory, I havent mentioned it being the Red Bull. I dont know how they make their splitter work?

        Although I should point out their strut (pin) at the leading edge is not always fitted, but even then doesn’t look that substantial.

  9. Another amazing piece of analysis Scarb’s, thanks!

    Someone above has mentioned the strut or pin which seems to fix the front of the splitter to the bodywork, how do you see this affecting the proposed ‘see-saw’ action?

    I know you can’t reveal your sources with this kind of material, but do you think this design is specific to one particularly dominant team over the last couple of years? Also, do you think others, say a team from Woking or one from Maranello for example might be looking at a similar solution?

  10. To make the splitter flex using either method, the splitter must rely on the leading edge of the plank making contact with the ground.

    1) Why wouldn’t the nose of the plank wear beyond the set limit?

    2) Why do we never see any plank smoke from this wear process?


  11. Even if you assume that the trailing edge of the splitter is not attached to the chassis (a “see saw” requirement?), the blank is one piece. There also is two blank mounting points located very close to the trailing edge of the splitter (leading edge of the chassis?) So now we have a piece of rock hard plywood screwed down right OVER the point of where the splitter must move downward (“see saw” effect).

    It is not going to happen without a lot of force being applied to leading edge of the plank, and thus, very high wear.

      • A lot less wear? What do you assume is causing the splitter to flex when the car is on the track? Where does the force come from to provide the flex?

      • I dont think bending a 1m section of laminated beechwood, is that tough a test. There’s no specifiction for the skid block material only a specific gravity. With a very thin carbon fibre aerodynamic skin, it woudl be relatively easy to apply enough force to bend the splitter upwards. Wear will occur at the leading edge, but would be reduced by the Ti skid blocks (allowed to be insterted into the plank) and also by the wear at the rear of the splitter.

      • @ Brian, the force to bend the splitter with this system would be much less than the normal force required on other cars. That’s why it’s effective, it allows the splitter to “flex” with less load on it and therefore less wear, which means they can a lower front end.

  12. I assume we are talking about a “see saw” system. With your proposal you need to move the plank (and rear edge of the splitter?) below the bottom of chassis floor plain. Is this not correct?
    The plank is mounted to the chassis at this point, so you are talking about bowing or bending the blank a very short distance from the mounting points. This will take a lot more force than a 1m lever.

    Is this an incorrect interpretation?

    • The splitter is about 50cm long. The front section of plank must be a minimum of 100cm, so it could be longer. There’s no specification for where the planks fasteners are to be mounted Reg 3.13.2.
      So there’s a lot of freedom to engineer the plank and its mountings, to allow the small amount of flex necessary to work this solution.

  13. Is there anything that prohibits the FIA from placing a strain gauge, to be logged in the ECU, in an appropriate location so as to monitor deflection… real time? I would think that would be the ultimate deflection test, there would be no way getting around it, and just plugging the car into a data logger would tell you immediately if the car passed or failed.

      • I guess the FIA is what happens when you put a bunch of true racing/engineering nerds in a closet with a bunch of cash, a good cad tech, and a few CNC machines. Speaking from my own experience, probably yours as well, engineers don’t always work on the logic principal, much less the common sense principal.

        Thanks for making my weekend. Ciao!

  14. I assume we are talking about RB being the team currently most likely to be benefiting from your flex proposal. That said, viewing the latest photos of the RB underside as it hung from the hook in Monza, you can see the two mounting assemblies right in the location where your proposed system needs to flex. I would think you would need 3-5 mm of movement downward at the rear of the “see saw” to get the rake benefit you are after.

    Also viewing the same photo, it is clear that there is a support rod very close to the leading edge of the splitter. Now this support can be made to bend, but it is going to require more force from contact with the ground and thus more plank wear.

    The visual clues we have from RB just don’t support the use of a see saw system by RB.

    Great theory and discussion. Did scarbsf1 develop it on his own?

    • I havent made any reference to Red Bull with this theory.

      I also made the point, the typical splitter can be regarded as a cantilever, if we ignore the support strut (rod). So I’ve not included a support strut in my proposal.

  15. Pingback: Pit lane exit warning lights for drivers in Korea | F1 Fanatic – The Formula 1 Blog

    • Thats a good point, I guess that would be down to the detail design of the compliance in the mountings, with the pivot, plank and surrounding bodywork providign the small amount of stiness\damping needed to keep the splitter still.

  16. “the design of the fulcrum need not be the pivot I drew, but a simpler solid but flexible part, that is not suspected to deflect”

    This sounds like the most logical next step for the teams after the hinges at the rearmost aspect of the splitter mounting were banned in 2007. The deflection test does seem to have it’s limitations given that it’s an area the teams have been trying to exploit since the early 2000s. Is there a test to stop the teams from hiding a 2006-7 style sprung mount of sorts to create the alleged mass damper effects of this period?

  17. Looking at the redbull t-tray it seems to be connected to the chassis by a cable. Could it be that this cable it used in conjunction with the seesaw affect to control it? For example connected in some manor to the brake or accelerator peddals…

  18. Scarbs, excellent theory and since some have questioned the sizes and loads, thought I’d put a few numbers to it to check out how it might work…

    Some basic dimensions from the regs, hope they’re correct, these are my assumptions

    1) The ‘plank’ starts 330mm behind the front axle CL and finishes at the rear axle CL

    2) The wheelbase is approx 2883mm

    3) The splitter extends approx 500mm from the front edge of the floor

    4) The static load on the front wheels is 291kg [rear 342kg, though this isn’t relevant to the calc]

    In order to make the system invisible to the scutineers it must not move the car at all during the deflection test.

    If there is a fulcrum involved somewhere in the splitter assembly, then the ‘see-saw’ will act as a lever and tend to help the ram to lift the car, rather like the action of a pry-bar. For example if the see-saw has equal length members either side of the fulcrum it will double the lifting effect, or 200kg x 2 =400 kg at the test load.

    Question – how much load would be required to actually lift the car by loading at the front of the splitter?

    Taking moments about the rear of the plank, remembering that this is also the axle line..

    291 x 2883 = (load at front of splitter to lift car) x (2883 – 330)


    Load at front of splitter to lift car = 291 x 2883 / (2883-330) = 329kg.

    Clearly this is less than the 400kg calculated above using the example of an equal lever therefore a short lever would simply lift the car because it would not have enough weight on the fulcrum to hold it down. In order to be invisible to the scrutineers, ie not lift the car, the fulcrum must be situated some distance back from the front of the splitter in order to be closer to the car centre of gravity to ‘use’ a greater proportion of the weight. The required position can be established again by simple moments about the rear of the plank.

    291kg x 2883mm = 400kg x distance forward from rear of plank


    Distance of fulcrum forward from rear of plank = 291 x 2883 / 400 = 2097mm

    or 786mm back from front axle line, which is 456mm back from splitter front edge, also about where the splitter meets the main body of the tub?

    The entire mechanism would therefore be operating over a length of approx 456 x 2 = 912mm

    Co-incidentally the regulations say that the forward piece of the plank cannot be any less than 1000mm long front to back, which ties in rather well with the geometry required to disguise the effect and the hardware during a load test. A see-saw with a forward section of 450mm and rearward of 550mm would be about right I think to allow for the lack of a driver during the testing. The stiffness of the plank itself would not inhibit the mechanism because it’s a seperate piece that moves with the see-saw beam. All that is then required is that the mounting for the rear of the see-saw and plank assembly to the tub system has low vertical stiffness in order to allow enough flex to avoid loading the nose of the plank when it touches the track. As you rightly point out, all this could be made part of the tub and essentially invisible to a casual inspection. The struts that some teams use at the front of the splitter are very thin and look as if they are intended to buckle under end load as a result of splitter deflection in use.

    Looking at the underside of Mark Webbers car, purely as an example you understand, accepting that the web photos are grainy, I could convince myself that there is a seperate front section to the plank of about 1 metre long…

    It’s entirely feasible, and it’s surely cheating isn’t it?

  19. Could the splitter be connected to the front wing via the the pin/cable which appears to connect it to the monocoque, through the nose and into the front wing? There have been pictures of cables in the wing before and Vettel’s seemed to be hanging on after his crash in Suzuka practice.

    The front of the splitter could then be pivoted upwards as load is places on the front wing. Although this would surely be a moveable aerodynamic device!

  20. From what I understand, you claim that the plank has to touch the ground with more than 200kg pressure. If the plank scratches on the ground to trigger the see-saw mechanism, wouldn’t that consume a lot of power?

  21. Pingback: UPDATE: “See-saw” Splitter, FIA issue a Technical Directive « Scarbsf1's Blog

  22. Pingback: Acerto de Hamilton base para recuperação de Jenson | Papaya Orange

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s