Does your car s front end go all light and jittery at speed? Here are a couple of popular solutions to the problem of unwanted front-end lift

Transcription

1 AIRDAMS AND SPLITTERS Does your car s front end go all light and jittery at speed? Here are a couple of popular solutions to the problem of unwanted front-end lift Words, illustrations and photos: Simon McBeath. Gulf Aston Martin Vantage photo courtesy of Aston Martin Racing 26

2 Massive splitter adorns GTE Pro Gulf AMR Vantage 27

3 Hillclimb Mini with deep airdam It s a matter of academic historical debate as to whether lift-reducing devices first appeared on the front or rear of competition cars. It may be that they started appearing at both ends at more or less the same time, depending on whether the car in question generated more lift at one end or the other. Either way, it all generally kicked off in the 1960s, when little chin spoilers at the front might have been seen along with small rear spoilers, as designers attempted to reduce or cancel the aerodynamic lift to which increasing speeds and sleeker styling were both contributing. It isn t actually clear whether anyone realised back then that adding devices that reduced lift or created genuine downforce, even at the expense of creating additional drag, was the way to faster laps. It s entirely possible they were simply trying to make their cars less unstable at speed! How times change Nowadays those chin spoilers have evolved into a number of other variously simple or complex devices including airdams, splitters and dive planes (also known by some as canards). And in this issue we re going to look at the effects of airdams and splitters using computational fluid dynamics (CFD) software, courtesy of ANSYS UK. In previous issues we have used a simple CAD model of a generic saloon car (based on the centreline profile of a Mitsubishi Evo VII found on the web) to study how rear wings and rear spoilers can generate sufficient downforce to cancel the natural lift generated by the car s intrinsic shape. But in order to get a better idea of how front end, close-to-the-ground devices such as airdams and splitters work, the CAD model this time needs to be just a bit more realistic. In short, it needs an underside featuring some of the lumps, bumps and cavities that usually blight a production car. To that end, an engine bay has been created into which a lump passing a crude resemblance to a transverse engine and transmission is placed. And as if it didn t know which end it was driven from, the model also has a rear axle unit (with industrial-size differential ) and associated cavity created. Then at the front, two typical apertures have been cut in the front panel to feed air to a block representing the radiator. This should provide a more realistic basis on which to try out some airdam and splitter combinations [Fig1]. Airdams Received wisdom is not always an infallible source of guidance (especially if it has been received from an internet forum ), but airdams are supposed to help the aerodynamics of a car in various ways. First, they reduce the amount of air that passes under a car and which then interacts (collides if you prefer) with the aforementioned lumps and bumps. Reducing the airflow under here can, it is said, lead to a reduction in drag, despite the extra increment of frontal area that the airdam itself adds. Secondly, the airdam reduces the car-to-ground gap, which accelerates the air s velocity and locally reduces the pressure under that part of the car. Thirdly and less obviously until you see the output from a CFD simulation there are also losses in the airflow s energy behind the airdam, and this also sees a modest drop in the pressure here, analogous to the modest pressure reduction in the car s wake at the rear. 28

4 So a couple of different sized airdams, one 50mm deep, the other 100mm deep, were attached to the baseline model with its rough underside, and each model in turn was added to our virtual wind tunnel to go through the CFD mill. The simulated force results on the baseline and airdam models are shown in Table 1. Table 1 the effects of airdams on the aerodynamic forces of our generic saloon model at 100mph Drag, N (CD) Lift, N (CL) Baseline 1088 (0.390) 1675 (0.600) Baseline + 50mm airdam 1086 (0.389) 1452 (0.520) Baseline + 100mm airdam 1109 (0.397) 1292 (0.463) Divide Newtons by 9.81 to get kilograms. Multiply kilograms by 2.2 to get pounds. Fig. 1: The CAD model of our generic saloon/sedan, now with rough underside Fig. 2: Upper surface pressure distributions at 100mph Fig. 3: Lower surface pressure distributions at 100mph Both airdams had negligible effect on drag, despite adding to frontal area, yet both reduced lift reasonably significantly, and that s really all that matters! Looking first at what these results represent, drag is hopefully self-explanatory, and corresponds to the force that resists the passage of the car through the air. Lift is the vertical force that the car s passage through the air generates, and in this case it is positive, that is upwards, therefore offloading the tyres and reducing grip. Also given in brackets are the possibly more familiar coefficient of drag and the coefficient of lift, calculated from the force results. Don t be misled into thinking these are forces and coefficients to apply to a real car; the CAD model is far too simplistic and hence the CFD simulations are only illustrative. But what we can do is to use the simulations to get a feel for the effects and approximate extent of changes, and in that respect it helps to re-tabulate the results showing the changes to the coefficients relative to the baseline, as in Table 2. Table 2 the effects of airdams shown as Δ (Greek letter delta) values, or the difference between each configuration and the baseline ΔCD ΔCL Baseline + 50mm airdam Baseline + 100mm airdam Delta values enable quicker and easier judgements to be made on the effectiveness of a change, and in this case we can see that the 50mm airdam gave us 80 counts of lift reduction, or downforce if you prefer (where a count is a change in a coefficient of 0.001) for, effectively, no change in drag. The 100mm airdam was pretty efficient, too, in that it gave 137 counts of downforce for just seven counts of additional drag. Overall the car still generated positive lift, but the airdams had beneficial results. So these airdams didn t quite fit the aforementioned preconception that airdams reduce drag, and that might be because our model is still very simplistic, and it might be that the airdams were quite small compared with the quite large ground clearance (170mm) of the base model. But we can say that both airdams had pretty well negligible effect on drag, despite the fact that they added to the car s frontal area, yet both reduced lift reasonably significantly, and that s really all that matters! Let s have a look at some graphics from the CFD simulations to try explaining how the airdams achieved the results they did. First, check out figures 2 and 3, which show the distribution of local pressures on the surfaces of our car model. Red and orange are areas of raised 30

5 Fig. 4: Pressure distributions along a centreline slice of the car at 100mph Fig. 5: Centreline pressure distributions with 50mm airdam Fig. 6: Centreline pressure distributions with 100mm airdam pressure, blue and green are areas of reduced pressure. Looking at the red on the front of the car, it s easy to see where a lot of the drag comes from. But it s perhaps not so obvious where the lift comes from in these views: figure 4 may help here. This shows the pressures on a slice down the car s centreline, and the scale has been altered to show more clearly where the pressure was raised above ambient (zero on this plot) and where it was reduced. It s now more obvious that the low pressure over the roof and the front of the bonnet would have been creating a lifting force, even though the pressure under the car appeared to be slightly reduced below ambient. But notice, too, how the pressure in the front of the engine bay was increased; this would also be contributing an increment of lift. Compare this now with figure 5, which is the 50mm airdam case. First impressions show the general pressure distributions to have been similar. But look more closely at the front end and you can see that there was less positive pressure in the engine bay, and more negative pressure under the front end of the car. And there was also evidence that the airdam had persuaded more air to pass over the car look at the little suction peak of blue on the rear lip of the boot in figure 5; it is more pronounced than in figure 4, implying increased flow over here. So the airdam subtly altered flow velocities and hence the pressures on the whole car, but its greatest effect was, of course, at the front end. Figure 6 shows the same view of the 100mm airdam simulation, and it s essentially a case of more of the same here. Splitters Moving on to try simulating how splitters function, and here we modelled one on a level just below the main floor level of the car, which in this instance meant attaching it to the bottom of the 50mm deep airdam. In real world applications the splitter height may be limited by technical regulations in sanctioned motorsport, or it may be unrestricted, but in any case the car would likely have less ground clearance than our model, so don t interpret what we ve done here as being ideal; it s purely illustrative. The splitter we evaluated here extends forwards only by 75mm, and Table 3 shows the results alongside those of the baseline case and the 50mm airdam only case, while Table 4 shows the delta values. Table 3 the effects of a 75mm splitter Drag, N (CD) Lift, N (CL) Baseline 1088 (0.390) 1675 (0.600) Baseline + 50mm airdam 1086 (0.389) 1452 (0.520) 50mm airdam + 75mm splitter 1149 (0.412) 1270 (0.455) Table 4 the delta values from a 75mm splitter, relative to the baseline ΔCD ΔCL Baseline + 50mm airdam mm airdam + 75mm splitter Hillclimb Audi 80 with airdam-cum-short splitter So this 75mm forwards-protruding splitter attached to the 50mm airdam produced 145 counts of lift reduction for just 22 counts of extra drag relative to the baseline car: in percentage terms this lift reduction corresponds to 24.2% less lift than the baseline case. This translates very 33

6 Fig. 7: Centreline pressure distributions with 75mm splitter attached to the 50mm airdam Fig. 9: Centreline velocity distributions and flow vectors with a thicker splitter Fig. 8: Centreline velocity distributions and flow vectors with the 75mm splitter roughly to 24.2% increased grip at 100mph compared with running no airdam and splitter who wouldn t want some of that? Thus in these examples, the 50mm airdam with 75mm splitter performed slightly better than the 100mm deep airdam alone. Let s look at how the splitter creates downforce. Using the magic of CFD once again to illustrate the pressure distributions, we can see there are two clear mechanisms here; raised pressure on the splitter s upper surface, thanks to the air velocity slowing right down; and reduced pressure on the splitter s lower surface thanks to the air flowing rapidly underneath it. Figure 7 showing the pressure distribution along the car s symmetry plane illustrates this nicely, although you can see that the pressure reduction under this splitter was not very pronounced. This is probably down to two things; firstly, the ground clearance is quite high; and secondly it s a rather thin device that isn t shaped to deal with the direction from which the airflow actually approaches. Figure 8 illustrates this in close-up, using flow vectors to indicate the local airflow directions: it s abundantly clear that the air s approach direction as it encounters the leading edge of the splitter is not horizontal as might be supposed, but markedly downwards. As a result, the air is unable to make the sharp turn around the leading edge and is actually separating from the underside of the splitter s leading edge. British GT Aston Martin Vantage with splitter shaped with knowledge of how the airflow actually encounters the device So the air speed immediately under the splitter is not high, as it should be, but instead is as low as it is behind the airdam. Thus, the local pressure under this splitter is not as low as it could have been, even at this relatively high ground clearance. This is why a splitter should have, at the very least, a generous rounded leading edge, as in figure 9, which shows an improvement in the flow under the splitter using double the thickness and a bigger leading edge radius, or perhaps better still, a leading edge that is turned up into the airflow and shaped more appropriately. You may well have seen such things on topline competition cars such as the Aston Martin Vantage GT3, for example. At this point many of you will be wondering why we haven t looked at larger splitters, and why we haven t extended those splitters back under the front compartment. That is something we shall look at in a future issue as a separate mini-project, because when panelling under the engine bay in order to generate more downforce-inducing splitter area, there are also implications, as will perhaps be obvious, for ensuring there is adequate extraction of the flow allowed in through those front apertures for cooling purposes Meantime, even the humble airdam can be of worthwhile benefit, and a modest-sized splitter that simply protrudes forwards from an airdam can also offer a significant increment of downforce. Many thanks to Ansys for the use of the CFD software. This is why a splitter should have, at the very least, a generous rounded leading edge, as in Fig.9 35