iRacing Dallara DW12 Aerodynamic Research

I tested the car at Talladega Speedway in order to test at the highest speed possible. Telemetry was used to determine dynamic ride heights.

(On this oval track, Diffuser settings are limited by iRacing to Sidewalls and Strakes both OFF.  On medium length speedways, iRacing allows Sidewalls ON, Strakes OFF and on some shorter tracks they allow Sidewalls and Strakes both ON.)

Front and rear wings were set at their minimum settings.  The results of many laps found that the optimum “least drag=most speed” was produced with “dynamic” ride heights of 1.000” front and 1.350” rear when traveling 233-234 mph on a  straight section of the track. This was counter to “rules of thumb” promoted by others indicating that a “nose up” attitude produced less downforce and drag.  

At speeds of 233-234 mph, the downforce with these ride heights produced having the lowest wing settings allowed (-4.5/-10.5) was:  Front: 420#; Rear 1200# as determined by spring deflection using 1:1 motion ratio. (Note: An exact motion ratio of ride height change to spring deflection is not available from public information so actual downforce figures should be considered to be approximate. It is assumed in this discussion that the ratio is different in the front vs the rear so downforce figures for the front are doubled.) Front: 840#; Rear 1200# as determined by estimated spring deflection using “estimated” motion ratio.

Initial “Static” ride height would be depending on the springs chosen.  For example: (2) 2800# front springs would compress 0.075” inches each (420x2=840# front downforce) so static ride height in the example above is 1.075” static front ride height;  (2) 1500# rear springs would compress 0.40” (1200# rear downforce) so static ride height in the example above would be 1.750”.

Spring rate choice affects handling, depending on track bumps and the transition from banked turns and less banked straights. (For example, the car at Talladega became very difficult to drive with rear springs stronger than 1500# as the transition from banked turn to straightaway creates a loose condition.) In general, from an strictly aerodynamic standpoint on ovals, the stiffer the better as the ideal "aerodynamic attitude" is maintained.

The next test was to learn how lowering the ride height would affect drag.  The car was lowered enough that it almost bottomed on the right side in the corner banking but the 0.35” front to rear “dynamic” rake was maintained on the straights.  Dynamic ride heights were 0.700” front and 1.050” rear.  Speed  and lap time dropped off approximately 2%.  

At speeds of 229-230 mph, the downforce with these ride heights produced while still having the lowest wing settings allowed (-4.5/-10.5) was:  Front 1000#; Rear 1500#.  Conclusion, lowering ride height will increase downforce AND drag. 

Another test was made that produced a dynamic ride height of 0.500” front and 0.850” rear. Here the car bottomed out in the corners, but still a downforce calculation was made on the straights:  Front 1600#; Rear 1600#.  Conclusion, lowering dynamic front ride height to 0.500” dramatically increases front downforce. The iRacing front to rear downforce calculator does not show this.  (This has important implications for the DW12 on road course, however there, the third spring must be taken into account.)

The Diffuser Wicker  created a very slight increase in drag and 25# rear downforce.  It is recommended that in most cases—set this diffuser at ¾” Sealed as its use will allow a very slight decrease in rear wing.

Each “click” of the front wing increased front downforce by approximatley 20#, e.g. 6 clicks produced a change of 120#.  Each click of the rear wing increased rear downforce by approximately  6.5#, e.g.  18 clicks produced a change of 120#.  The ratio of #/click front to rear was not linear. The higher the wing settings, the more the rear wing changes per click in proportion to the front. So on low wing settings, the ratio might be 4 clicks rear for each one click front, but at higher wing settings, the ratio is closer to 2 or 3 clicks rear for each click front.  (These ratios are affected by wing wicker settings.)

On tracks where turning “grip” is relatively more important than maximum straight line speed, downforce is improved by lowering the car, front and rear to the lowest possible dynamic ride height with minimum bottoming.  Maximum use of the diffuser or “wing under the car” also produced the best results—maximum wicker and maximum sidewall/strakes available.  This usually requires stiffer rear springs and often results in front and rear dynamic ride heights being equal.  (It is suspected that the “faster with nose up” rule of thumb is related to reducing rear ride height to lower the car—this does not increase top speed, but does increase downforce and speed through the corner, and therefore may increase speeds on the straights.)

Keep in mind that there are many, many variables that affect handling and speed. Finding the perfect setup for best lap times still takes experimental trial and error testing.  The knowledge above only serves to help us "zero in" faster on the best result.


Donald Wayne Strout 10/28/2015

What does the 3rd Spring Do?

On the IndyCar in iRacing, there is a "third spring" in the front and rear.  Pictures are worth a lot of words so:

Pic of Front Suspension

Pic of Rear Suspension

Here is a link to article that provides some interesting data. Pics are from that article.


Article
http://www.lolachampcar.com/3rdSpring.html

Bottom line is that the "3rd Spring" essentially provides resistance to pitch and squat relatively independent of the four "corner" springs.  Note the rods connected to the suspension rockers that go to third rocker supported by the "3rd Spring" assembly.  If the car rolls to one side in a turn, with one suspension rod going up while the other moves down, then the third rocker simply rotates and the "3rd Spring" does not move.  If the car dives while braking or is lowered by aero downforce, than both suspension rods move closing the "gap" and then compressing the "3rd Spring".

Keep in mind that IndyCar aero downforce can nearly double the wheel loading at very high speed, so with the "3rd Spring" it is no longer necessary to carry that aero load with the "corner" springs, allowing the car to handle with more compliance in turns.

The "gap" is the distance that the rocker will travel before engaging the "3rd Spring".  Choosing that setting is an important part of chassis setup.

The actual "3rd Spring" may be a form of dense foam bump-stop progressive rate spring rather than the coil spring in the photos, but the principle of action is almost the same. 

Caster Angle-The Most Misunderstood Front End Setting

Caster Angle is an important setting to achieve desired handling characteristics.

Race cars almost universally employ positive caster.  The steering axis is "leaned back" at the top, which places the tire contact patch behind the point where the steering axis intersects the road.

Positive Caster create Five (5) effects:

1)  Steering Effort and Straight-Line Stability.   Because the contact patch is behind the steering axis's intersection with the road, increasing positive caster increases the force necessary to drag the tire sideways. This occurs on both front tires and is greatest at beginning of a turn when the wheels are starting out in the straight ahead position.


2) Camber Change.  With positive caster, the negative camber of the outside tire is increased while the negative camber of the inside tire is decreased, or the positive camber of the inside tire is increased. This is usually beneficial as it compensates for chassis roll and generally increases the grip of the front tires.

3) Cross Weight.  Because the steering axis in inclined, in a turn, the outside tire is lifted away from the road, and the inside tire is pushed down toward the road.  When the inside front tire is pushed down, it also increases the weight of the outside rear tire, essentially creating a negative change to cross weight. The more steering input, and the more positive the caster, the more effect.  This essentially makes the car have less understeer or more oversteer as steering input is added.


4) Auto Steering.  The same principle described regarding Cross Weight change also can be used to make the car turn more easily on oval tracks. On oval tracks, the inside tire is usually set to have less positive caster which actually causes the car to turn left with little or no steering effort.

5) Effect of Steering Corrections. The same principle regarding Cross Weight when turning produces the exact opposite effect when making a steering correction to an excessive oversteering situation. Turning the steering the opposite direction of the normal turn will push the outside tire down, placing more weight also on the inside rear. The more positive caster, the more exaggerated and rapid this change in handling occurs--sometimes large positive caster settings results in the driver over-correcting.

So, on a race car, caster essentially is a tool to change the handling characteristics of the car during turning. More positive caster makes the car looser in mid corner and more stable in a straight line.

Many, many years ago, my Father taught me about caster by using a pencil and a pin. He took the pin and stuck it into the side of the pencil at a 90 degree angle.  Then he leaned the pencil back and rotated it. The pencil was the steering axis (king pin in the old days) and the pin was the wheel axle.  Rotating the pencil counterclockwise with the pin on the right illustrated the outside tire in a left turn--the pin went up. Rotating the pencil counterclockwise with the pin on the left illustrated the inside tire in a left turn--the pin went down. My first lesson was 56 years ago when I was 8, but I have used the concept ever since.

Driver Feedback in IRACING

IRACING is a wonderful simulation. But it is not exactly the same as real racing.

First, you have no feedback of lateral or longitudinal G Forces to your inner ear.  This "seat feel" or the ability to use G Force feedback to the inner ear is what most real life "fast" drivers have and develop to a high level.

When you are racing in a simulator, your primary feedback is through your eyes and to some extent through your hearing.  Let's say you get VISUAL and AUDIO feedback, just like real life.

But, even with a motion seat or motion cockpit that moves around, that "seat feel" is just not there. (More on this in a separate article as there are some refinements that can be made to a motion cockpit that can get close to real seat feel, but many motion cockpits provide so much "noise" that the important G force feedback is lost.)

In real life, I would place the relative importance of feedback as: 

Seat Feel (G Forces)

Visual

Audio

So on the typical simulator, you are missing the most important feedback.

The human mind is pretty good at adapting. So, very quickly, one learns to use visual feedback for steering input. (It helps to use settings in the .ini file for DriverRotateHead in the range of 0.75 to 1.  IRACING ARTICLE With practice ,this provides a very good visual indication of yaw, allowing an accurate indication of understeering/oversteering and enables the driver to "catch" a slide reasonably well.) Then one uses the sound of the engine as feedback for throttle input. 

IRACING allows you to adjust the various audio output to increase tire sound, while decreasing wind noise. On some cars with street tires, the tire squeel is a decent feedback for cornering force. But on cars with slicks, all you hear for tire noise is sort of a whoosh.

There is a great deal of controversy regarding Force Feedback of the steering.  I find that you can tell when the front tires loose traction, but generally, I already know that from visual feedback regarding the car's direction compared to my steering input.  Some drivers claim to be able to get more out of it, but I remain skeptical.

The most important input of all is braking. Correct braking technique is the key to driving fast on a road course.  And here you have very little feedback. 

Your first set of pedals are likely to be the potentiometer type where movement determines brake force. Push the pedal more distance and get greater force. IRACING allows you to change the relationship a bit to make it less linear, but there is no force feedback. Some drivers adapt to this quite well and seem to sense the position of their foot. Others find that they need brake force and muscle memory to tell them how much braking input they are introducing. So, the next step for these folks (me included) is to a load cell brake where the force you push is measured rather than the position. 

Load cell brakes tend to be a bit too abrupt, so there are all kinds of mechanical tricks applied to provide "feel" by using springs, rubber bumpers etc. in the quest to provide a pedal that has BOTH position and force feedback.

Then, there are those that believe that the only way to simulate a hydraulic braking system is to use a hydraulic braking system, where the force is measured with a pressure transducer. Still, the "feel" boils down again generally to the use of elastomer bumpers or discs that are compressed by the hydraulics. 

One unappreciated feedback provided by IRACING is the "UI" or user interface box with a little red bar that rises and falls in proportion to braking input.  The default setting for this is quite small and hardly noticeable when driving. But, they provide methods to double the size of the box, with Control Page Up or a 200% setting in graphic "Options".  

With the command Alt K, the box can be moved with a mouse to the location of the driver's choice. A side benefit is the box also indicates the transmission gear and provides a green bar for throttle.

So, by enlarging the box and positioning it strategically, it serves almost like a HUD or heads up display and the driver can see the red braking bar rise and fall in his peripheral vision.

All of a sudden, trailbraking and general brake input modulation feedback is not only possible--it is convenient and easy!  So is the use of "maintenance" throttle while braking or "rolling thru" a corner.

So, once you enlarge and reposition the UI box, you have good visual feedback for steering, good audio feedback for throttle, and good feedback for braking modulation.

Driving fast is all about using, brakes, throttle and steering in the correct combination with precision timing. (As well as having the car setup to optimize those inputs.) The driver is the "brain" that collects the information (feedback) and with a developed physical coordination makes the appropriate inputs. Improve the quality and speed of feedback to the driver and he will be more precise; he will be better able to judge required changes to the setup; and with practice he will be fast. 

On most ovals, the use of the brakes is less important than in road racing, but throttle modulation is often the critical determinant of speed. The enlargement and repositioning of the UI box's green throttle bar provides excellent feedback for that slight amount of throttle release necessary during corner entry as well as on corner exit with a loose set-up. 

One last word about feedback---your input and output electronic devices will all suffer to some degree from 'lag" or "delay" between what you do and what the sim sees as well as what you see as feedback.  These cannot be eliminated, but they can be minimized.

Mock Racing Article on Lag

The first "lag" is from the ping or milliseconds it takes for the internet to send a round trip signal from you to the sim servers and back--ranges from 50 to 150 milliseconds. The second lag is from the display.  A really good one has about 10-15 milliseconds of input lag; a LED TV might have 40-50 milliseconds. Then there is the "refresh rate" that varies from 60hz (16 milliseconds) to 144hz (7 milliseconds.) The third is the frequency that the sim "reads" your input--typically 60 times a second or 16 millisecond intervals. The fourth is the rendering of the GPU--a combination of Frames per Second or FPS as well as "buffering" or "stabilization" or "synching" that creates output lag. Finally the steering wheel typically has input/output lag or latency that varies from a very low 5 milliseconds to up to 30 milliseconds.  

Wow, you can see that the driver might actually be responding to visual cues or FFB cues with a lag or delay of 100 to 200 milliseconds. A car moving at 120 mph travels 18 feet in 100 milliseconds!  Almost like driving looking out the rear view mirror. Reducing lag will make the driver more competitive.

I am convinced that over time, many drivers adapts to this "alternative reality" by both optimizing their hardware and adapting to the hardware and system so that this lag no matter affects their "normal" lap.