18 replies to this topic

[Aris26]

As I was searching the forum, I saw many diffusers and I want to start a conversation on that. To be clear, I do not pretend the expert, I have only done one Msc module in Aerodynamics, but the lecturer was focus especially on diffusers because they are very important. The use of diffuser is to allow the low pressure vortex that is created on the back of the vehicle to suck the air underneath increasing it's speed and lower it's pressure. I attach some pictures from an analysis that we did on Star ccm+ to show you what I mean.  Diffuser.jpg   106.17KB   14 downloads Vx220 has already a diffuser and the only that you can do is to add the endplated to form the air. To put a diffuser that is paraller to the ground and is very lenghtly is completely stupid. It destroys the meaning of diffuser which is to allow the wake behind the vehicle to suck the air from underneath. Except from the loss of downforce you increase the drag also, because our purpose is to reduce the two small vortices which a diffuser also contributes by sending the flow upwards. I made that post to help guys to improve their car peformance not to pretend the smart, unfortunately many tuners make parts without any special knowledge just to sell them.

[mbes2]

Hi,

This has already been done i.e http://www.vx220.org...they-important/

There are some more like this.

[Aris26]

Ok, if it is not necessary maybe it is better to be deleted my topic

[Mike (Cliffie)]

Please don't delete and please always add your views.

[mbes2]

Above is true, as its good to get new views and ideas. I've read this subject to death and back... still don't understand

[2-20]

But is there a problem with the diffuser angle being above 10 degres ?

[sford]

From my understanding of reading a few books, once the angle of diffuser increases beyond a certain amount, the void created by the diffuser may also need additions vanes to stabilise turbulent air. This is all subjective though as I think ride height and other under body aero comes into it when you are talking about this level of modification. Simply adding a diffuser isn't going to make the car handle in a massively different way. I do applaud your attention though. Spending time working out models, applying software and your knowledge is very much appreciated. I have some computing capability at my dosposal but dont have the knowledge to set up/perform the analysis. There is a section in the premium members area that may be of interest!

Edited by sford, 28 April 2013 - 09:23 PM.

[Aris26]

After the 9.64 degrees it loose it's effectiveness to suck the air. It is still beneficial but not as it could be. The drag seems to reduced more above 9.64 degrees but the downforce also. There are vortices that are created on the diffuser and as you increase the step from the flat underneath to the diffuser area, the vortices increase and u make cause flow separation. Sford I agree on what u wrote, it all matter on the ground clearance. Reducing the height of the vx220 not only u increase the downforce by the reduced pressure underneath but you make also the diffuser works better. I would be interested to participate in the premium section discussion. I payed the membership on June but never worked, I contact some coordinators but never took a reply...

Edited by Aris26, 29 April 2013 - 12:55 AM.

[fezzasus]

PM any of the site admin, they should be able to help. Ask for their email address and you can forward your paypal receipt as proof.

[siztenboots]

I will chase up the problem with the subs , meantime have you seen the white papers from Southampton. They did some great work on this. I've had a quick look at your account and its something the higher level admins need to look at , so I will raise a to fix request.

[FLD]

There's also some work done on diffuser / wing combo's. Apparently big wings help steeper diffuser angles keep clean air by adjusting the flow over the back of the car. I cant explain it better than that I'm afraid as it bakes my head!

[siztenboots]

maybe the steep diffuser could stall , so having a more aggressive top wing puts more energy under the wing.

[Nev]

Getting the car really low is the critical thing in my experience. Unfortuntely this causes serious handling problems (summarised below). However, these problems aside, I did some experiementing with Nipper with the following config:

1/ Front ride height down to 8cm, rear was about 9cm.
2/ Big wing.
3/ 2" projecting splitter (which was 1 cm deep, hence the clearance to tarmac was 7cm)
4/ No rear floorpan
5/ No difusser (not even the OEM end/back small floorpan)

With this combo, the aero effect was seriously noticable, not by just a little/marginal bit, but a massive amount more down force. Once you got over about 90 leptons the steering got a fair bit heavier. After about 120 leptons the car was being crushed onto the tarmac. The effect became very noticable when slowing down to corner and the steering became so light I thought I might understeer badly. Where previously the car became very light at the front end at high speed (such that you could use 2 fingers to steer), I now found that you needed both hands on the wheel. Perhaps a diffuser would help with that setup still further, but in isolation I can't see it making a dramatic effect TBH.

This info was gained by actual experimentation. My subjective views on this are as unbiased as I can be; in other words this setup really worked. If you are serious about downforce, this is one way to do it, however you'd need to address the big problems of:

1/ next to impossible clearance for road driving.
2/ very little suspension travel.
3/ wishbones out of their intended horizonal plane.
4/ steering arms not horizonal and consequential bump steer.

All of those 4 problems are addressable, but at some expense. I did try getting new machine hubs made to raise the hub centres by 4cm, however the company could not make them in a reasonable time frame. I am still considering getting them made else where.

HTH.

Edited by Nev, 29 April 2013 - 08:28 AM.

[FLD]

Could be, the way it was in my head was that the diffuser angle had a function of inlet pressure and outlet pressure. The wing reduces air balooning round the back of the car so reduces the outlet pressure. Obviously this is all in my head and things can get very wrong in there!

[Winstar]

As I was searching the forum, I saw many diffusers and I want to start a conversation on that. To be clear, I do not pretend the expert, I have only done one Msc module in Aerodynamics, but the lecturer was focus especially on diffusers because they are very important.

The use of diffuser is to allow the low pressure vortex that is created on the back of the vehicle to suck the air underneath increasing it's speed and lower it's pressure. I attach some pictures from an analysis that we did on Star ccm+ to show you what I mean.  Diffuser.jpg   106.17KB   14 downloads

Vx220 has already a diffuser and the only that you can do is to add the endplated to form the air. To put a diffuser that is paraller to the ground and is very lenghtly is completely stupid. It destroys the meaning of diffuser which is to allow the wake behind the vehicle to suck the air from underneath. Except from the loss of downforce you increase the drag also, because our purpose is to reduce the two small vortices which a diffuser also contributes by sending the flow upwards.

I made that post to help guys to improve their car peformance not to pretend the smart, unfortunately many tuners make parts without any special knowledge just to sell them.

There is lots of bollox on the internet about aerodynamics and how it works on every car will be diferent hence why F1 spend so much time on it, even then someone goes and smashes a front wing with no disernable loss in performance.

Your explination of the diffuser is only partially right in that the aim is to diffuse to air from below the car into the depression at the back of the car the influence of the angle is that to achive this you need to keep the flow attached to both the diffuser and the ground and this angle is dependent on the ride hight. Bellow you can see the effect of dropping the ride hight to 50mm and keeping the standard diffuser angle.



Additionally even if you keep the flow attached then there are other effects that come into play

From the OEM and short style diffuser results below you can see that there is only a small are of low pressure, if you look a the stream lines it can be seen that this is because the high depression in the center of the car causes the most of the flow to be pulled down the center of the diffuser. Additionally you can see that the air flow under the body upto the rear wheel archesis pretty uniform but after that there is flow leakage from the rear wheel arches.





So to stop this you need to have multiple vanes in the diffuser that extend as far to the ground as possible, these are 40mm from ground, and extend the full length of the undertray


However you can get almost the same downforce by having just the outer vanes extending the full length of the undertrayt to stop the leakage from the wheel arches


You could probably fine tune the diffuser a bit more to get a few more % increase but cutting off the leakage from the wheel arches is what make the difference.

As for your Star model I suggest using the trim-cell mesher and wake refinement for that type of problem rather than the polymesh, only exception is if your using porous regions to represent the radiators like in my models.

[Aris26]

I do not understant what we say different? I said the same thing above that the only improvement that it can be done is to add the endplate, vx220 has already a diffuser. Take a look at the conditions and tell me your opinion

Star ccm+ Condition Settings

Surface Mesh:

Surface Remesher

Volume Mesh:

Polyhedral Mesher / Prism Layer Mesher

Base Size:

0.005

No Vol. Control

Growth Rate: 1.3

Air Velocity:

40 m/s

Inlet Pressure:

63.579 Pa

Outlet Pressure:

0 Pa

Inlet Temperature:

344 K

Outlet Temperature

344 K

Turbulence Model

K-Epsilon

The problem with the arces, we found it also at our models and when it was a group discussion the professor told us that it is not leakage to the arces but it sucks the flow from outside and it needs sideskirts. If you see your pictures the flow underneath is increasing from the second half of the two sides of the car, forming like triangle.

In the second picture that u uploaded, there are turbulences inside the diffuser, are u sure that u run the model correct? I have not seen that again, how they are created inside the diffuser at a speed of around 45m/s. The length of the diffuser should be as long it is needed to connect the underflow with the wake behind the vehicle. If you make it longer than this, I cannot see how u will take advantage of the wake effects?

Edited by Aris26, 29 April 2013 - 04:30 PM.

[Winstar]

I do not understant what we say different? I said the same thing above that the only improvement that it can be done is to add the endplate, vx220 has already a diffuser. Take a look at the conditions and tell me your opinion

Star ccm+ Condition Settings

Surface Mesh:

Surface Remesher

Volume Mesh:

Polyhedral Mesher / Prism Layer Mesher

Base Size:

0.005

No Vol. Control

Growth Rate: 1.3

Air Velocity:

40 m/s

Inlet Pressure:

63.579 Pa

Outlet Pressure:

0 Pa

Inlet Temperature:

344 K

Outlet Temperature

344 K

Turbulence Model

K-Epsilon

The problem with the arces, we found it also at our models and when it was a group discussion the professor told us that it is not leakage to the arces but it sucks the flow from outside and it needs sideskirts. If you see your pictures the flow underneath is increasing from the second half of the two sides of the car, forming like triangle.

In the second picture that u uploaded, there are turbulences inside the diffuser, are u sure that u run the model correct? I have not seen that again, how they are created inside the diffuser at a speed of around 45m/s. The length of the diffuser should be as long it is needed to connect the underflow with the wake behind the vehicle. If you make it longer than this, I cannot see how u will take advantage of the wake effects?

Which version are you running? for exernal flow I'd always try to use the trim cell mesh with prisim layers then either the wake refinement model in the more recent (versions V6.04 onwards IIRC) or volume refinement in the wake your vector plot would suggest you don't have any refinemt there. Also where on earth are you running your car 344k is 71 Deg C, although it doesn't mater if your using constant density. Which setting for the K-e model?

The point I was trying to make that you missed out that the it's critical not to have a diffuser too steep for the application as if your flow seperates then all the diffusion will be lost, this is what is shown in the second image I posted. You should be able to quickly recreate this on you're model by taking the diffuer angle up to somthing like 20 Degrees.

yes the leakage I was refering to was flow being sucked in from the wheel arches the last 2 models have end plate baffels that run inside the wheels and down length of the diffuser.

I don't do automotive CFD any more (it doen't pay enough lol) my VX model is just somthing I play with when I have spare time.

[Aris26]

Could be, the way it was in my head was that the diffuser angle had a function of inlet pressure and outlet pressure. The wing reduces air balooning round the back of the car so reduces the outlet pressure. Obviously this is all in my head and things can get very wrong in there!

FLD if u want take a look at SAE 2008-01-0327 and SAE 980030, u will see with numbers what u say. The benefit of the wing except of driving the flow upstream itself, it increase the wake behind the vehicle and it make more effective the diffuser also. The incline of the diffuser is important to take the most possible advantage of the wake but to do not be so steep to cause vortices at the top of the underneath flow, reducing it's speed. For that they say that the optimum angle is 9.64 degrees. Except the incline, the length of the diffuser and the endplates are also important. There is flow that is sucked underneath from the side body flow behind the wheels. So the most important end plates are the two at each side of the diffuser to reduce that effect. Regarding the lenght it has to connect the underflow with the wake, if it is longer u do not only reduce the potential benefits from the increased speed underneath but at the part that are 'outside' of the car it is created lift because the flow underneath has increased pressure compared with the almost ambient pressure that the wake has. In conclusions u destroy everything if the flow from the diffuser is not connected to the wake.

[Aris26]

I do not understant what we say different? I said the same thing above that the only improvement that it can be done is to add the endplate, vx220 has already a diffuser. Take a look at the conditions and tell me your opinion

Star ccm+ Condition Settings

Surface Mesh:

Surface Remesher

Volume Mesh:

Polyhedral Mesher / Prism Layer Mesher

Base Size:

0.005

No Vol. Control

Growth Rate: 1.3

Air Velocity:

40 m/s

Inlet Pressure:

63.579 Pa

Outlet Pressure:

0 Pa

Inlet Temperature:

344 K

Outlet Temperature

344 K

Turbulence Model

K-Epsilon

The problem with the arces, we found it also at our models and when it was a group discussion the professor told us that it is not leakage to the arces but it sucks the flow from outside and it needs sideskirts. If you see your pictures the flow underneath is increasing from the second half of the two sides of the car, forming like triangle.

In the second picture that u uploaded, there are turbulences inside the diffuser, are u sure that u run the model correct? I have not seen that again, how they are created inside the diffuser at a speed of around 45m/s. The length of the diffuser should be as long it is needed to connect the underflow with the wake behind the vehicle. If you make it longer than this, I cannot see how u will take advantage of the wake effects?

Which version are you running? for exernal flow I'd always try to use the trim cell mesh with prisim layers then either the wake refinement model in the more recent (versions V6.04 onwards IIRC) or volume refinement in the wake your vector plot would suggest you don't have any refinemt there. Also where on earth are you running your car 344k is 71 Deg C, although it doesn't mater if your using constant density. Which setting for the K-e model?

The point I was trying to make that you missed out that the it's critical not to have a diffuser too steep for the application as if your flow seperates then all the diffusion will be lost, this is what is shown in the second image I posted. You should be able to quickly recreate this on you're model by taking the diffuer angle up to somthing like 20 Degrees.

yes the leakage I was refering to was flow being sucked in from the wheel arches the last 2 models have end plate baffels that run inside the wheels and down length of the diffuser.

I don't do automotive CFD any more (it doen't pay enough lol) my VX model is just somthing I play with when I have spare time.

Also where on earth are you running your car 344k is 71 Deg C, although it doesn't mater if your using constant density.

He suggested to keep the K-ε stock settings as it is the most correct.  Untitled.jpg   107.46KB   3 downloads

We used a hell of a lot time to refine the mesh and at the end the professor came and put his final personal touch. Take a look at the picture and tell me your opinion.

We run the models up to a diffuser angle of 35 degrees to see the effects. It was never caused flow separation, only vortices to the upper part. Until 10.9 degrees the downforce slightly increased and stabilized but after that it reduced massively. In the pictures that u upload it shows a large vortice at the diffuser. This is not possible, I have read honestly more than 25-30 SAE papers and I have never seen a so large vortice, only tiny's that are on the upper part of the diffuser, if u take a look on the SAE that I posted, u will notice of what I tell u. If u have placed that diffuser on the car, cut the last part of it that is outside the car's body, it destroy's ur downforce completely. What I write is from freindly interest to a potential track competitor , I do not want to be advantaged on u by the aerodynamics

Edited by Aris26, 29 April 2013 - 05:23 PM.