[KVRC] Khamsin Virtual Racecar Challenge 2016

[cdsavage]

Shouldn't be too much longer - there was one entry that had some problems with simulation. Once that's done we'll just release the full set of numbers with the lap times.

[CAEdevice]

Hi, I have a question about the 10mm rules applied to the rear wing section.

The required thickness is referred to the whole section or the the two closed profiles separately? I mean: the second (usually upper) profile has to be 10mm thick for most of his length?

[cdsavage]

For the entire span of each element, there should be a point along the chord where the thickness is at least 10mm. There's some leeway given for complex 3d-shaped wing profiles where it might be difficult to maintain thickness in the curved areas.

[CAEdevice]

For the entire span of each element, there should be a point along the chord where the thickness is at least 10mm. There's some leeway given for complex 3d-shaped wing profiles where it might be difficult to maintain thickness in the curved areas.

Thank you!

[RicME85]

Just been reading about Microsoft's latest developments for Windows 10 and it got me wondering if it would help Julien with OpenFoam and OCCFD...
https://blogs.windows.com/buildingapps/ ... n-windows/

[machin]

Hi, a quick update from my side as the provider of Virtual Stopwatch:

After the the test round we noticed a few things with the results that we decided needed to be tweaked in Virtual Stopwatch concerning the intake, exhaust and cooling pressures, as per below:

• We noticed that the exhaust routing on a few cars looked like it might be quite tight. To account for this we've reduced the target exhaust pressure to 4.5Pa.m^2. This accounting for the fact that some of the original pressure allowance of 9.8 Pa.m^2 is (theoretically) now being used to overcome higher pressure losses in the pipework system within the car. In reality all cars met this figure anyway (we think all cars exhausts look realistic), so this change just stops people going more extreme.
• On the intake side there was one entry with inlets that looked like they would require a long tortuous route to the engine which would give more power loss than was originally being calculated. We didn't want to change the target pressure (that would unfairly penalise other cars), so we've tweaked the intake pressure power de-rate curve slightly; in essence there is now a bigger drop in performance if you don't meet the 0Pa.m^2 target.
• On the cooling differential side we noticed an annomally which is a function of the way the boundary conditions are applied.... In real terms a 0 pressure differential along a duct system results in no flow... however, in our case the calculation of a 0 pressure differential isn't "fair" since we've forced a specific flow rate boundary condition onto the system (in essence on some cars the inlet needs to "suck" (that's an intake pressure integral of <0Pa.m^2) air into the inlet in order to achieve the pre-determined flow... in reality the pressure wouldn't drop that low you would actually get a lower flow rate.... anyway, long story short: the power de-rate curve has been modified on the cooling side so I think everyone will now see less power reduction. The power reduction is still significant though so you should still be aiming for the target of 20Pa.m^2.

The updated version of Virtual Stopwatch is available on the usual link:-

http://www.competition-car-engineering. ... KVtest.htm

[FYI: I have not seen anyone's detailed geometries: I've only seen the official pictures posted on the website]

[CAEdevice]

On the cooling differential side we noticed an annomally which is a function of the way the boundary conditions are applied.... In real terms a 0 pressure differential along a duct system results in no flow... however, in our case the calculation of a 0 pressure differential isn't "fair" since we've forced a specific flow rate boundary condition onto the system (in essence on some cars the inlet needs to "suck" (that's an intake pressure integral of <0Pa.m^2) air into the inlet in order to achieve the pre-determined flow... in reality the pressure wouldn't drop that low you would actually get a lower flow rate.... anyway, long story short: the power de-rate curve has been modified on the cooling side so I think everyone will now see less power reduction. The power reduction is still significant though so you should still be aiming for the target of 20Pa.m^2.

Hi Machin, thank you for the update.

About the point quoted above: can you explain wich is the law that corrects the engine performance starting from the cooling pressure differential?

I was thinking about it and I am convinced that a good compromise (waiting for the porous CFD model that would fix anything) would be to apply the reduction only to half of the total power:

- Previous model: Power = Power.max * (reduction coefficient)
- My proposal: Power = 0.5*Power.max + 0.5*Power.max* (reduction coeffcient)

We would also avoid zero power cars.

[machin]

That's pretty much what we decided: 0Pa.m^2 = 40% power. 20Pa.m^2 =100% power, with a square law inbetween and minimum power of 40%. That should be enough of a drop in lap time to make cooling flow important.

Out of interest: do we have anybody intending to submit a car who didn't take part in the test race? The introductory class didn't have much of an uptake... I'm interested to hear what would make more people enter....?

[CAEdevice]

I think that we should also consider the geometric quality of the submitted cars (I mean CAD quality, not CFD quality). I am surprised to see that partecipants without any cad skill did not choose the intro class.

I think I will release a basic car obtained with extrusions and rounds only: I hope it could help new partecipants with no CAD skills. Maybe the official intro car has a too advanced geometry to be edited with SketchUp.

Finally: what about r71 team?

[Alonso Fan]

Out of interest: do we have anybody intending to submit a car who didn't take part in the test race?

Yes

[RicME85]

I think I will release a basic car obtained with extrusions and rounds only: I hope it could help new partecipants with no CAD skills. Maybe the official intro car has a too advanced geometry to be edited with SketchUp.

This. As I mentioned way back, making the extrusion of the sidepod to meet the floor whilst keeping the curves was impossible in Sketchup...hence you doing it for me and posting the file here.
There are far too many faces in the supplied bodywork to be able to just grab and extrude a surface.
I have said it previously and the idea got shot down but it might be an idea to look at it again going by a) the lack of intro class entries and b) the lack of ability of some of the entrants....more supplied parts. A selection of wings etc that produce different levels of drag and downforce that can be swapped in seconds, same goes for floor/diffuser and the rear section of the rear wheel pod.
IMO the intro class should be about learning how aerodynamics work. If the intro class came with a body like it does but say 5 versions of each of the currently missing parts that all have varying levels of performance then the beginner entrant can spend time learning how different AoA works or how different wing profiles work.

I liked Richard's entry using the intro class, when the results come out I might consider going down the same route for this season as despite having a good base, I havent really got the time to improve the quality of the model as it will be a case of starting from scratch and rebuilding the geometry but with finer tolerances.

[CAEdevice]

Anyway, Chris' model is great, I have been tempted to use it as a base. Maybe some partecipants should dedicate more time to learn some basic cad skills.

On monday (or tuesday) I will release simplified version of the car that I used for the preseason race (but my 2015 car, syill legal except for the sidepods, is already available on GrabCAD, with more than 4500N of df in race conditions: just type KVRC or CAEdevice to download the step or the stl).

[RicME85]

Learning aero is difficult enough without having to learn CAD.
I have tried FreeCAD, Solidworks, 3DS Max, Blender and god knows what else. I have watched hours of tutorial videos and cannot get my head around any of these programs which is why I continue to use Sketchup and will for the foreseeable future.
Not everyone has time on their hands (Im struggling with time and I dont work) to learn a whole new package. If the uptake on the intro class is low...do something about it, make it more approachable. Maybe something like my 'box of bits' idea is the way to do it, maybe its something else. Exposure is definitely something that needs working on. I dont think the KVRC website works very well, it looks a jumbled mess and doesnt display thinks properly. I think it could also do with some pictures of what the intro class looks like and maybe a video.

There used to be some excellent Sketchup modelling tutorial videos posted on this forum from back in the Formula Sketch days, things like how to design a sidepod and how to create an F1 style front wing from extruding the profile and manipulating it to follow two separate paths. I have tried to find them but not luck, I will however try again,

[CAEdevice]

To learn CAD enough to model a decent car is not that difficult. Who is not ready to learn the basics of modeling is not probably ready to study things like vortices or boundary layers. Not all the partecipants are the new Adrian Newey, a genius despite the CAD skills.

I guess that a car modeled with only extrusions and rounds can reach 4000N of df and obtain points.

On the other side: consider your case. During last year you were able to improve very much you car from both CAD and CFD point of view. I can imagine you had to invest some energy in it, but nothing comes without an effort.

[RicME85]

You say that but I have tried on numerous occasions over the last 4 years to get my head around Solidworks etc but to no avail.

My suggestions for the intro class is about easing people into KVRC gently, give them a taste of what the full version of it is like by holding their hand at every step of intro. Get newbies hooked on getting a car setup so they get the bug to improve the car themselves by tweaking supplied parts or designing their own from scratch.

As for last year, yes I had to invest a lot of energy and time, I would hate to think of how someone working 40 plus hour weeks and bringing u a family or someone studying full time would get on trying to put the same time and effort as I did last year seeing as my day to day life consists of getting up at 10am then sitting at my computer either messing around with Sketchup or gaming then going to bed in the evening, no work, no studying.

Anyway, it was only a suggestion. A small competition like KVRC will only get by and improve with ideas from those involved, shutting them down and dismissing the abilities of others wont help it.