[KVRC ~ish] CCE LMP01

[RicME85]

Thanks Richard!
Will work on those when I get home in two weeks.

[machin]

Hi Ric,

I took the liberty of looking through your previous designs and your results and I see that you already had a lot of those features on your previous cars....

I think your round 3 car looked pretty good... I'm wondering if it might make a better starting point, if for no other reason than you already have a set of aero coefficients with which to work with...



First thing would be to tidy up the side pod/crash structure....

The next step would be to graft on the diffuser exit from your round 4 car:



Both those should tidy up the car, and reduce drag a little, but don't expect miracles.

The round 3 car had quite a front bias to its downforce, which essentially means the front end is producing a lot more drag for additional downforce that the car can't utilise. I would be inclined to lose one element from your upper front wing: this should improve your balance and overall efficiency.

Again, we can't expect miracles with these changes, but I think that leaves you with a good basis to start with. I took the liberty of photoshopping what the car might look like (a picture paints a thousand words and all that).



There's an interesting article in the latest Racecar Engineering in which they made an F1 car model and subjected some changes on it to see what effect they had: trimming a small amount OFF the diffuser width manipulated a vortex sufficiently to increase downforce by 10% AND reduce drag: i think the moral here is that once you've got the basics sorted there are going to be no easy wins: to match the front runners requires some pretty in-depth CFD analysis.

Perhaps you could post some CFD images from under the Diffuser of your Round 4 car and we can try and figure out what's going on. Start with a surface pressure plot...?

[LVDH]

Hi,

I was wondering about something. Most cars have this big gap in the side pods. Mine has always been totally closed. I understand that the air entering from the sides to the underbody and then into the diffuser section is what creates a lot of the downforce on these cars. So with the design you are proposing here it looks like a nice amount of air which does not go into the cooling side pod should be directed, and with a nice rotational component to the underbody. From just looking at it it should be more air compared to a design like mine. Yet, whatever I do with open side pods it does not work (for me).

Am I doing something wrong or is there a different reason for these open sections?

[machin]

I always thought the purpose of the open sides was to encourage air through the front end aerodynamics... rather than any attempt at supplying the diffuser, but to be honest one thing I've learnt about aerodynamics is that it is bloody complex!

My thinking was this: air will always take the easiest route around an object (or in scientific-speak: "It follows the pressure gradient"): to ensure that the air goes through the front-end aero (rather than around the sides of the car), you want an unrestricted flow between the wheels and out towards the back of the car, opening up the sides like this (in my thinking) increases the ease with which air can flow through the front end... Essentialy, it has more ways to "get out" once it has been between the front wheels: either over the sidepod, or down the sides....

However, I am intrigued by your finding that it "doesn't work". In terms of aero coefficients, how does it affect the results? Reduction in Cl.A? Increase in Cd.A and/or drastic change in COP (if so, which way; forwards or backwards)?

[CAEdevice]

Very itneresting thread Machin (and very interesting question LVDH).

My answer, at the moment, is not very interesting: I choose the open sidepods shape at the beginning of the development and I could not find the time to test any alternative. It seems the air goes through the front diffuser more easily, but I could not make the air turn and feed the floor at all (untill now... but I'm still working about it).

[machin]

Thanks Matteo. I think the fact that your car has the separate wheel pods (and Variante's to begin with) shows that the concept can be made to work (even if it is not feeding your diffuser like you mighty hope)...

Matteo, I'm interested to know: are you running multiple iterations of your car between rounds and are you finding big improvements in coefficients for small changes in geometry? (like my example above from RaceCar Engineering where they found that REDUCING the width of their car's diffuser by 25mm introduced a vortex which cleaned up the flow inside the diffuser and gave them 10% more downforce AND less drag...)...?

[machin]

So I continued to look at the Round 3 Brook Motorsport car and think of other ways that it could be improved: (this is easy in photoshop; not quite so easy in Google Sketch!);

• Continued to smooth off all the edges; i.e. increase the radius of any fillets, particularly on the front wheel fairings
• inclined the front suspension and side crash structure fairings in the pursuit of extra downforce
• Joined the wheel fairings as LVDH suggested.
• increased length of wing mirror fairings to allow a more tear-drop shape

The question then though is given the above changes, what sort of improvement can be expected?

The car started with the following coefficients:-

Cl.A = 3.35m^2
Cd.A = 1.33m^2
COP = 1.395m

My first main change was to reduce front downforce to get rid of the excess drag that the front end was creating by losing the two element upper wing and replacing it with a single element upper wing (the lower wing assembly remains unchanged). The aim here is to reduce the front downforce by about 30% , with overall downforce dropping by about 15% (Cl.A reduction of about 0.51m^2 overall)

Assuming that the upper front wing wasn't too efficient (say 4:1 efficiency) this loss of front downforce would reduce drag by about 0.1275m^2 (0.51 / 4 = 0.1275).

That leaves us with:-

Cl.A = 2.83m^2
Cd.A = 1.20m^2
COP = 1.65m

Smoothing off all the corners and tidying up the side crash structure area and opening out the diffuser will help further, but its not going to be a huge difference, conservatively I'd hope maybe downforce might increase by a further 10% and drag might come down by 15-20%:

Cl.A = 3.1m^2
Cd.A = 1.0m^2
COP = 1.65m

That's a reasonable improvement... and would result in a lap time reduction of about 2 seconds per lap at Magny Cours... but its not going to win a round.... Unfortunately the conclusion that I am coming to is that to match the front runners the only path to take is to study the CFD results and perform an iterative development process (i.e. make a small change and see if it leads to an improvement, repeat as necessary) looking at:-

• Diffuser profile
• Diffuser throat position relative to wheels
• Diffuser strake quantity and span-wise position
• Wing element angles
• Wing element slot Gaps
• Wing element shape

[CAEdevice]

Thanks Matteo. I think the fact that your car has the separate wheel pods (and Variante's to begin with) shows that the concept can be made to work (even if it is not feeding your diffuser like you mighty hope)...

Matteo, I'm interested to know: are you running multiple iterations of your car between rounds and are you finding big improvements in coefficients for small changes in geometry? (like my example above from RaceCar Engineering where they found that REDUCING the width of their car's diffuser by 25mm introduced a vortex which cleaned up the flow inside the diffuser and gave them 10% more downforce AND less drag...)...?

I think you will find interesting details about that in my car for the 5th race.
At the moment I can say that reducing the exit angle of the diffuser reduces drag with a small loss of df.

About the rounded edges: I tried something, but the solver/mesher has great difficulties with a more complicated model, so it stops working most of the times.

[machin]

At the moment I can say that reducing the exit angle of the diffuser reduces drag with a small loss of df.

But I suspect (like the RCE example), that this solution is very specific to your car; if everyone made the same type of change I think we would see some improving, and some getting worse because the flow through the diffuser depends hugely on the airflow around the rest of the car...

...I suspect the same could be said of almost every part of the car; specific car = specific solutions... So applying general aerodynamic principles will only get you so far (balanced Cl.A = 3, and Cd.A=1 I think would be reasonable), but much more than that and you need to look at specific flow details of your own car, because the interaction of all the various parts is very difficult to predict (without CFD)...

...would you agree?

[CAEdevice]

At the moment I can say that reducing the exit angle of the diffuser reduces drag with a small loss of df.

But I suspect (like the RCE example), that this solution is very specific to your car; if everyone made the same type of change I think we would see some improving, and some getting worse because the flow through the diffuser depends hugely on the airflow around the rest of the car...

...I suspect the same could be said of almost every part of the car; specific car = specific solutions... So applying general aerodynamic principles will only get you so far (balanced Cl.A = 3, and Cd.A=1 I think would be reasonable), but much more than that and you need to look at specific flow details of your own car, because the interaction of all the various parts is very difficult to predict (without CFD)...

...would you agree?

Yes, the reaction of the airflow to the diffuser expansion depends on the vorticity of the flow, that is influenced by a lot of parameters (honestly speaking, I can't control all of them, so part of the optimization is obtained varying the parameters and guessing about the effects...).

Ps: Please, feel free to discuss my car

[RicME85]

Wow, excellent work Richard!
When I get home I might re-produce both versions of the car you have designed to see what numbers each gets. Would be interesting and would help the hypothesising in this thread which in turn should offer some insights to other competitors and any possible future competitors.

[machin]

Ps: Please, feel free to discuss my car

I'm not sure I'd be able to tell you anything!..... In fact... I (and no doubt a lot of the other competitors) would be interested to hear how you use the CFD results to improve your car. I presume you follow some sort of step-by-step process as I've laid out below? Could you provide some insight into the questions I've highlighted in red?

1. Compare your results to your competitors and the last version of your own car.
2. Work out where to concentrate your effort.... (Try and be realistic, and don't bite off more than you can chew; don't try and change too much in one go, e.g. "Improve front end downforce", "improve diffuser performance", "reduce drag whilst maintaining downforce" are probably realistic goals, "increase downforce, reduce drag and improve COP" is probably asking a bit much)
3. Study the CFD results.What analysis do you find most useful? Surface pressures? Streamlines? Total pressure plots? What are you looking for when you look at each analysis (I.e. What indicates an area that needs attention/could be improved)
4. Based on what you saw on your CFD analysis implement changes to try and improve the car in these areas. Can you provide some guidance of some examples of what you do to solve the problems you found in step 3? Like manipulating surfaces to avoid stalling, controlling vortices, etc?

I think that would be great Matteo....

[machin]

Wow, excellent work Richard!
When I get home I might re-produce both versions of the car you have designed to see what numbers each gets. Would be interesting and would help the hypothesising in this thread which in turn should offer some insights to other competitors and any possible future competitors.

That would be cool! I'd have a go myself... but my Google Sketch skills are a bit lacking!

[CAEdevice]

Ps: Please, feel free to discuss my car

I'm not sure I'd be able to tell you anything!..... In fact... I (and no doubt a lot of the other competitors) would be interested to hear how you use the CFD results to improve your car. I presume you follow some sort of step-by-step process as I've laid out below? Could you provide some insight into the questions I've highlighted in red?

1. Compare your results to your competitors and the last version of your own car.
2. Work out where to concentrate your effort.... (Try and be realistic, and don't bite off more than you can chew; don't try and change too much in one go, e.g. "Improve front end downforce", "improve diffuser performance", "reduce drag whilst maintaining downforce" are probably realistic goals, "increase downforce, reduce drag and improve COP" is probably asking a bit much)
3. Study the CFD results.What analysis do you find most useful? Surface pressures? Streamlines? Total pressure plots? What are you looking for when you look at each analysis (I.e. What indicates an area that needs attention/could be improved)
4. Based on what you saw on your CFD analysis implement changes to try and improve the car in these areas. Can you provide some guidance of some examples of what you do to solve the problems you found in step 3? Like manipulating surfaces to avoid stalling, controlling vortices, etc?

I think that would be great Matteo....

I would add a "0" step (I check it with more urgence than the final classification): the correspondence between my local solution (OCFD with OpenFoam 2.1 Windows compiled, without wings refinement: it is the best compromise between reliability and results accuracy) and the official one. Consider that I've lost a podium (1st race) for a wrong setup of the local solver.

About the questions in red:

- Post processing: I use total pressure (-1000 +1000 Pa) with Paraview
- Areas to improve: I changed the diffuser after the 2nd race (it is the most important part of the car, it makes the difference between Df/Drag ratio lower or higher than 3). I'm using Naca inspired airfoils (optimized with local simulations). At the moment I'm trying to reduce the lift generated by the wheel covers and by the front suspension covers, without any results to be honest

[machin]

I wanted to put some basic ideas down for interpreting the CFD results from vehicles that will hopefully help newcomers to the KVRC championship. I don’t in any way pretend to be an expert in CFD analysis; a lot of the established competitors in the KVRC championship clearly know a lot more than me, but this should get you started:

Once the basic car design has been achieved there is really no alternative to CFD analysis (or full scale windtunnel testing!) in order to progress the performance of the vehicle from “also-ran” to “competing for wins”.

PRESSURE DISTRIBUTION

When an object moves through the air it displaces that air around it. The air attempts to follow the shape of the object as it passes by. In so doing the velocity of the air changes in both direction and magnitude relative to the object. It is this relative change in velocity direction and magnitude which generates pressure on the surface of the vehicle. Using CFD it is possible to generate a surface pressure plot of your vehicle as it passes through the air (red is generally high pressure, blue is low pressure) :-



The summation of all of the pressures on the body of the vehicle lead to the three main coefficients we are interested in: Downforce, Drag and Centre of Pressure (COP):

• The generation of downforce is the summation of the pressures on the top and bottom facing surfaces of the car. High pressure above and low pressure below = downforce. Low pressure above and high pressure below = lift.
• The generation of drag is the summation of the pressures on the front and rear facing surfaces of the car. High pressure in front and low pressure behind = drag.
• The relative distribution of the areas of high and low pressure on the surfaces of our vehicle determine the COP.

It is clear to see from the image above that the front and rear wings of our vehicle generate significant downforce (high pressure, red, when viewed from above, and low pressure, blue, when viewed from below)… But also high drag (high pressure, red, when viewed from the front, and low pressure, blue, when viewed from behind). Unfortunately this is something that is unavoidable, although we should try and design wings which have a high “efficiency”: as much downforce as possible for as little drag as possible.

Sidepods, canopies, and wheel fairings can also produce significant amounts of drag and lift if they are not designed well. In this case you want to look out for areas of low pressure (blue) on top of these objects (indicating they are generating lift), and red in front or blue behind (indicating they are generating drag).

A tear-drop shape is one of the best shapes for reducing drag (refer to previous post with Cd Values for a range of different shapes), and should be employed where possible on all fairings. However low pressure is generated on our car when the airflow follows a convex surface as can be seen on the curved top of this canopy:-



Whereas this flat-topped wheel fairing produces a lot less lift, although there is still some low pressure where the air flows onto the top:


We should attempt to design these objects so that the air is directed (“gently coaxed” is probably a better description!) around the sides of the fairing, rather than up and down the back of them.

STREAMLINE ANALYSIS

The next analysis we might want to do is to look at stream lines, particularly around our wing and diffuser elements. We can see on our NACA aerofoil data (see previous pages of this topic) that at some angle of attack the downforce generated by our wing suddenly drops (although drag continues to rise). This is the “stall point” and results in an inefficient wing (high drag for low downforce). The stream-line analysis of the underside (low pressure side) will indicate whether the air flow “stays attached” to the wing surface, or separates from it (the wing is said to have “stalled”). If flow becomes detached then the angle of attack of the wing element should be reduced incrementally until attached flow is achieved (this may take several iterations to achieve). This will give you more downforce for less drag.

(aircraft wing shown, -imagine the image upside down for a racecar wing)

TOTAL PRESURE PLOT

The final analysis I want to talk about is the Total Pressure Plot which is generally shown as a slice through our model showing the total pressure (Both static and dynamic elements, and is better thought of as the “total energy” in the flow) of the air as it passes our vehicle. Our downforce generating elements (wings and diffuser) should be fed with air which has the highest possible total pressure if they are to generate efficient downforce. The rear wing in the image below is in good high energy flow (red/orange).



The image below shows a transverse Total pressure slice through a diffuser showing a large area of low energy flow (dark blue patch) generated by the wake from the tyre contact patch: this low energy flow is reducing the “active area” of the diffuser.



There are several solutions to this problem:

• introduce bodywork which stops the formation of the low energy wake in the first place.
• Introduce bodywork which directs the low energy wake out of the diffuser,
• to purposely direct a vortex into the diffuser to contain and control the low energy wake.

Vortices are generated when air from a high pressure region spills/migrates into an area of low pressure whilst moving in a direction perpendicular to the migration. Wings, barge boards or dive planes (all without end-plates) can generate strong vortices.

(Another Aeroplane wing)


The strategic placement of these devices upstream of the low energy wake can produce a vortex which interacts with the low energy wake area: the aim being to stop it spreading throughout our diffuser. Trial and error is required here to find the best placement for the vortex generators.