Accumulated drag coefficient for an automobile

[ryan298]

Hello everyone,

I have a question regarding the accumulated drag curve shown in the figure below for the CFD simulation of external flow over an automobile. (It is written Vehicle Cd in the figure)

It can be seen in several regions that the accumulated drag curve is dropping (ie., having a negative slope, for instance, as in the region between 2500 and 3500 for the x axis). Now this should mean that in those regions, the drag is negative locally, as only this would bring a reduction in the overall drag. But how is this possible,? How can drag be negative locally in the said region for instance?

Correct me if I'm misinterpreting the curve.

PS: 1) The figure I've posted here is taken from a source which I've referenced below

"Levin, Ridgal, Aerodynamic analysis of drag reduction devices on the underbody for SAAB 9-3 by using CFD, Master's thesis, Chalmers University of technology, 2011"

2) It is kind of urgent, so if someone could help me out ASAP, it would be great

 

[AlephZero]

Start by understanding a much simpler geometry that shows the same effect.
http://www.grc.nasa.gov/WWW/k-12/airplane/dragsphere.html

 

[cjl]

Aleph - I don't see how your link has anything whatsoever to do with the question. The question is asking about a single flow condition in which portions of the geometry appear to have a negative contribution to overall drag, not about how drag can vary based on differing flow conditions (which is what your link is discussing).

 

[AlephZero]

OK, my eyesight wasn't good enough to read the labels on the graph. I took it as a plot of drag coefficient against speed.

To be honest, I don't have any idea what plotting "Cd against chord length" is supposed to mean.

Anyway, it looks like the OP got some sort of answer from a different forum.

The total drag is a combination of pressure drag and viscous drag. The pressure drag can be negative or positive depending on the local pressure relative to free stream.

... which doesn't seem to shed much more light on it, IMO.

Google can't find paper referenced by the OP, so there's no help from going back to the original source.

 

[dean barry]

Normally, keeping the shape of something the same, but changing the size ( that looks like what the graph is describing) leads to an different (specific) Cd value.
Pick the length of your car according to the x axis, read the Cd value from the y axis, then the drag force (f) in Newtons can be calculated from the usual :
f = v ² * Cd
where v is the velocity in m/s

 

[256bits]

The study,
http://publications.lib.chalmers.se/records/fulltext/143210.pdf

Appendix A, Fig 74, Page i ( Page 81 of 88 of the PDF )

 

[256bits]

The graph is somewhat difficulty to interpret as there is only one point that is the real value of Cd, and that is the endpoint of about 2.9 at the 5500 mark for the whole vehicle. Cd is for the whole shape and not really for parts of it. One could not for instance cut the car at the 3500 mark and expect a Cd of 0.14 from the graph, as that would not be taking into account an end effect.

I suppose it is a representation of the contribution to the drag, or to the final value of Cd, of the parts of the car as we progress down its length. Perhaps the graph ordinate labeled as Cd makes it unintuitive. Accumulated, to me, represents something that is added up, and I don't think that is what the graph is all about. Progressive Cd might be more apt for the graph title.

 

[Ranger Mike]

the nose of the car hits that wall of air ( think of a window curtain that weights 14.7 pounds per square inch). There is a huge amount of drag..it starts to drop off at 1500 mark due to following the surface of the hood (bonnet) until the path of the air hits the windshield ( windscreen) and goes up again) after it flows to the roof top. This longest surface privides high speed low pressure ( relative to all the other readings) it goes negative until it hits the end of the vehicle and jumps up to maximum observed reading.
if you note, the readings drop after a period of following a surface that is " straight" and fairly lengthy..hood, windshield , roof. any time the air flow can follow a surface and the surface does not change over 7 degrees, the air can remain attached. When you have transition areas that disrupt this flow you have turbulance and high drag...like at the end for the roof.

 

[Damo ET]

I'd have to agree with Ranger Mike. Although I am not sure how you could generate a graph like that, it appears what it is showing is that the flat portions of body work are reducing drag from the previous non flat portions. Straightening airflow directly countering non aerodynamic body shapes in-front, which cause restriction.Damo

 

[autodoctor911]

From what I see in that graph, all portions of the graph are positive coordinates, and have some drag. The only thing I see that is negative is the slope of the graph where the drag on the sections where the body panels are parallel to the wind. This would only mean there is less drag on that portion of the car than the part further to the front of the car.

I think the left side of the graph should be CdA or D for drag, not Cd.

If it were the Cd it should drop down way lower where it does. The Cd where the roof is horizontal, and the surface drag is the main force, should be very low. there is a lot of Area there though.

 

[ryan298]

Thank you everyone!

 

[cjl]

From what I see in that graph, all portions of the graph are positive coordinates, and have some drag. The only thing I see that is negative is the slope of the graph where the drag on the sections where the body panels are parallel to the wind. This would only mean there is less drag on that portion of the car than the part further to the front of the car.

I think the left side of the graph should be CdA or D for drag, not Cd.

If it were the Cd it should drop down way lower where it does. The Cd where the roof is horizontal, and the surface drag is the main force, should be very low. there is a lot of Area there though.

It's the accumulated drag, meaning the drag from the front of the car to that coordinate. That means that a negative slope actually does mean a negative drag contribution from that portion of the car. Also, calling it the Cd is correct.

 

[autodoctor911]

So, which is it? accumulated drag or Cd? I don't think anyone would say that Cd is the same as accumulated drag.

Accumulated drag would be a quantity of units, most likely Newtons, or LB/Force.

Cd is a coefficient.

I don't think it makes any sense as accumulated drag, for the same reason the OP finds it implausible. There is no way you would have a negative drag for portions of the body, especially areas that are parallel to the direction the measurement is being taken. as any pressure would have no effect in the direction being measured.

Also, a Coefficient of drag can not be negative as far as I know. if it could you could have perpetual motion.

I guess you could have negative drag in a localized area, but it would have to have a surface where pressure coming off another surface is pushing forward on a surface that is not parallel to flow.

I don't know, it is hard to determine what it is supposed to represent. To me it would make a lot more sense if it were labeled Distribution of Drag.

 

[cjl]

Accumulated drag coefficient, if you must be specific, and as I already explained above, it makes sense for that portion of the car to be contributing negatively to the overall drag (the overall drag coefficient for the car can never be negative, but the local contribution of one part can). The overall drag coefficient is equal to the accumulated drag coefficient at a coordinate equal to the length of the car, so for this car, the Cd is approximately 0.3.

 

[autodoctor911]

OK, it just seems weird to me to display something like that.

How would they calculate it. Is it the measured drag at each point divided by the cross sectional area at that point?

So, in those portions that are contributing negatively to the overall drag, there must be some kind of pressure on a surface that creates a force towards the front of the car.

I just can't imagine what surface in the center of the car is angled such that it can create a forward force.

 

[cjl]

I would imagine you would get a plot like that from a CFD program, though I suppose you could put a model in a wind tunnel with a bunch of pressure taps to get the surface pressure distribution (I'm not sure how you would get the distribution of viscous drag from a wind tunnel though).

As for what kind of surface can create a forward force? As I said, I would expect the local velocity at the top of the windshield to be substantially higher than freestream, which means the local pressure is well below freestream. Since the windshield is angled forwards, a low pressure region at the top of the windshield will result in a force pointing forwards and upwards.

 

[autodoctor911]

ah, yes, I see now. I also looked at the other figure showing Cp plot. It confirms what you are saying.

I knew that you got lift from the low pressure on the top of the car. I didn't even think about the fact that it is pulling the car forward in front of the upward curves, and back behind where the curve comes back down.