Understanding the numbers on the data sheets

[Carl Beck]

What opens a .rar file?

 

I know, I'm clueless.

 

John

 

It's an "improved" ZIP file compression ... actually it's the native format of the WinRAR archive.

 

On Windows it's opened with the later versions of WinZIP

 

on the Mac OS-X, it's opened with current versions of STUFFIT

 

I had to "up-date" my Stuffit version...

 

FWIW,

Carl B.

[74_5.0L_Z]

I have made extensive improvements to the excel file, and I have replaced the file above with the new version. It now includes an interactive graph. By that I mean that you can select the data that appears on the graph through the use of some dropdown menus. In order for this feature to work properly, you must set your Excel Macro security to medium and allow the application to run the macros.

 

If you don't know the current level of your Excel Macro security, set it to medium as follows:

 

From Excel

Select Tools -> Macros -> security and select medium.

 

Do this before trying to run the new spreadsheet.

 

Again, I will gladly accept any feedback for improving this tool.

[Michael]

Q is the dynamic pressure, 0.5*density*V^2. If the testing were at 80 mph (=117.3 ft/s), at "standard" conditions Q comes out to 16.3 Lbf/ft^2, in the usual engineering units.

 

Cross-sectional area times Q times the drag (or lift, etc.) coefficient gives the drag (or lift, etc.) force.

 

I really regret that I didn't participate in your wind tunnel test....

[74_5.0L_Z]

Thanks Michael for pointing out what should have been obvious.

 

I was thrown by the definition given in the initial posting of the data:

"Q down pressure on entire car measured near center of roof ". I didn't realize that Q was just the denominator of all the Cd,CL, ... equations divided by the area. I was using it all along without using knowing it.

[74_5.0L_Z]

Michael

 

I ran the numbers for a few test cases. Like you said :

 

Q is the dynamic pressure, 0.5*density*V^2. If the testing were at 80 mph (=117.3 ft/s), at "standard" conditions Q comes out to 16.3 Lbf/ft^2, in the usual engineering units.

 

However, the snapshot that is posted here is for an extrapolated 120mph, and each test configuration listed has a different value of Q(psf). By that I mean test configuration 1 average has a Q(psf)=16.294, test configuration 2 average has a Q(psf)=16.246, test configuration 18 average has a Q(psf)=15.865, etc... If the Q(psf) were as you stated then it would be constant for each configuration because the dynamic pressure is only dependant upon velocity and air density.

 

Thanks for trying though. If you or anyone else are able to discern the physical relationship of Q(psf), let me know and I will incorporate it into the spreadsheet.

 

I should also point out that I was not part of the team that did all the work at the windtunnel. I only endeavored to put this spreadsheet together so that the data would be a little easier to visualize for everyone in the HybridZ community. By doing this, I feel like I am contributing to the aerodynamics effort(I too wish that I could have been there).

 

Thanks,

Dan McGrath

[Michael]

In any actual wind tunnel test there will be variations from run to run; repeatability is never perfect, especially if the schedule is tight and the fan rpm control system is ... well, not the best that money can buy. So in every run, the wind speed in the test section will be slightly different, despite the desire for the same nominal conditions. Sometimes it will be 80.7 mph, other times 78.9 mph, and so forth. But each time, the value is recorded, converted to dynamic pressure, and saved for future reference. Then, the physical forces for each run are normalized by the corresponding dynamic pressure for that run, to be reduced into coefficient form.

 

If you want the actual lift and drag at 120 mph instead of at 80 mph, multiply the numbers for 80 mph by (12/8)^2. At 120 mph Q = 36.7 lbf/ft^2.

 

In the end, variations during the tunnel runs are pretty small. We're trying to improve CD from 0.45 to 0.35 (for example); a 2% error here or there isn't going to make that much of a difference.

 

Do we have any means of doing a group teleconference, where we could all go over the highlights of the data? I'm not trying to imply that somehow magically I have all of the answers, but it would be easier to explain some of the lingo verbally, so that we're all on the same page.

[74_5.0L_Z]

Michael,

 

Variations in the velocity between one run and the next makes sense for the variation in Q(psf). Further, it may be that the spreadsheet whose snapshot I am using as a reference always reports Q(psf) for the velocity of the original test and not for the extrapolated speed.

 

Have you looked at the excel spreadsheet that I put together?

 

Thanks

[Michael]

Looking at your data set, I think that the most important info is contained in three columns: CD, CL_front and CL_rear.

 

The number of runs is very impressive for a one 8-hour shift, and compares favorably with what our engineers achieve for airplane-type testing in our tunnels.

 

I would caution that every wind tunnel is a little bit different, so comparing numbers from your test with published numbers or measurements in other facilities is always a bit dicey. That said, the baseline numbers for lift and drag coefficients sound about right; drag is a bit high, but it would have been lower (I think) with a moving ground, or a least a boundary layer sucker upstream of the test section. There may also be some "blockage" effects that can not be entirely eliminated, despite what certainly appears to be commendable good sense on the part of the facility operators.

 

I am however surprised that the various approaches at drag reduction did not yield more than they did. But not being part of the crew I can only speculate whether lift reduction, as opposed to drag reduction, was the primary objective. An anecdotal comparison - so, let's not get too worried about it - is that of the Hot Rod 2nd-generation Camaro - which evidently dropped something like 40% in Cd, with the various modifications tried during testing in A2.

 

It still seems to me that the culprit is the Z's sharp hood lip. The G-nose doesn't help (I'm not surprised), but what "should" help is radical surgery to the hood, dropping the radiator upper crossmember by some 2" and sloping down the hood lip. This is unfortunately difficult to implement and is only practical for a V8 car with lots of engine setback (and probably firewall setback).

[Mikelly]

Our "goal" was to test parts that are so popular among our Zcar Members and with the various vendors... Do those "cool" looking parts "work?" That was the goal for the first 20-22 tests. Once we rolled Roddy's car in, it was more an excersise on "what should work on the track".

 

When we go back, I'd welcome your participation, Michael.

 

Mike

[JMortensen]

I am however surprised that the various approaches at drag reduction did not yield more than they did. But not being part of the crew I can only speculate whether lift reduction, as opposed to drag reduction, was the primary objective. An anecdotal comparison - so, let's not get too worried about it - is that of the Hot Rod 2nd-generation Camaro - which evidently dropped something like 40% in Cd, with the various modifications tried during testing in A2.

 

It still seems to me that the culprit is the Z's sharp hood lip. The G-nose doesn't help (I'm not surprised), but what "should" help is radical surgery to the hood, dropping the radiator upper crossmember by some 2" and sloping down the hood lip. This is unfortunately difficult to implement and is only practical for a V8 car with lots of engine setback (and probably firewall setback).

I saw the same thing, but my impression of the problem was that the Z's used were all too high off the ground. Reading the Hot Rod article the first thing they did was make the airdam go right down to the floor, but the Z airdams look to be 5-6 inches off the floor.

 

The 70's Camaro used in the Hot Rod article has a laid back front end like you describe too, but they were able to get the drag down to .201. If you have the magazine handy, and you look on page 114 at the bottom left corner, they were able to get a stock front end with very sharp hood and headlight areas areas down to .337 by using a huge airdam and grill plug, and then by copying the tail design from the first Camaro got the total drag down to .292.

 

My gut tells me that a deeper, better airdam could get our Z's down to the low 3's or high 2's.

[Mikelly]

I agree, but again, I reference my post above... most members don't have coil overs and sectioned struts. We will be going at it from a whole 'nuther perspective for round 2!

 

I'd think that closing off the grill, dropping the car low to the ground with maybe 2 inches or less clearance for the air damn, and using all the other tricks we learned would get us deep into the .3Xs I don't see a Zcar getting into the .2Xs though. However, I'd love to be pleasently surprised!

 

Mike

[OlderThanMe]

Also the Gnose was tested in the standard form is like funneling air under the car.

 

Funnel, just like stock but with more shelf to grab air and push it under...

 

An ok scraper...should be less gap under the "bumper" edge. Just stretch the airdam upwards to seal off the gap.

 

OTM

[TitaniumZ]

One of the reasons the cars in the tunnel photos looked rather "unlow" was that the tires were resting on scale pads attached to data recorders beneath the floor.These pads were about 1 to 1.5 inches above tunnel floor.

The point remains though;low is better.It's up to the the individual car owner to decide what amount of ground clearance is practical for them.

[Mayolives]

Well put, Titanium. Had we tested a "slammed low" car, I doubt the results would have been very useful to most members. Yes, low is better, but most of us need to drive our cars on streets with obstacles that would be hard on a 2" high front nose. Event track guys with trailers need to be able to load and unload with out dragging the front end. We really need an easily attached and removable front air dam/splitter. Perhaps Titanium can design one. If so, I'm in!

[Michael]

Did the scales really protrude 1-1.5" above the tunnel floor, thus elevating the car by that amount? If so, then this is a pretty significant flaw in the A2 tunnel. Scales should ALWAYS be flush with the floor! One solution is to make a plywood template with cutouts for the scales, and a ramp several feet in front of the car, thus elevating the flow back up to "curb height" relative to the car.

 

How much more expensive - and how much busier - is the larger tunnel?

[Mikelly]

Yes The scales did indeed protrude...We had to use blocks to roll the cars up onto the scales.

 

How much busier and more expensive??? $1000 per hour, and the likes of "all the NASCAR elite" use it. To add insult to injury, that tunnel requires you bring your own engineers to collect the data...

 

I don't know that the "scales" protruded that far, but they were at a level we had to roll the car up and onto them... I asked Bob about the "impact" of the rise and he said it had no real effect... I also don't think these cars were "that" high in the air, but I will double check my data.

 

Mike

[OlderThanMe]

So we could just lower a car an extra 1.5" for the test right?

Someone could fabricate a custom spring replacement to adjust ride height easily. It would basically replace a spring and threaded collar on a coilover setup with a threaded rod on either side of the strut to adjust the ride height.

[Mikelly]

I went back and looked at all the pictures and some video clips. The scales were flush with the floor... What we were dealing with is the "stops" that help the wheels from rolling back and forth... I"ll post pics later, but the scale is DEAD LEVEL with the floor, so throw out that theory...

 

The cars were tall, because they were at stock suspension with larger than stock wheel/tire combos, which in my mind valedates the numbers even more. 90% of guys driving these cars are running around in stock configuration with plus'ed up wheels/tires.

 

Mike

[jt1]

My memory of the wheel pads is that the center of the pad is flush with the floor, with a "ramp" on each end that is about 1.5" high. We used the blocks to roll the car over the ramp, then down.

 

Part of our original plan was to let the air out of the tires to check lower ride height, but as the sidewall bulged out the tire would bear on the floor around the scale pad and affect the results, so Bob nixed that right off the bat.

 

Checking ride height is gonna be expensive, because if we use a coilover car you have to unstrap the car, roll it off the floor, jack it up, take off the tires, make adjustments, put it back on the pads, ......., etc. All that eats up a lot of tunnel time, but I bet the results would be worthwhile.

 

John

[Mayolives]

Michael

How much more expensive - and how much busier - is the larger tunnel?

 

I don't remember what the cost was to rent the other tunnel but I do remember the operator saying it was booked for the next 14 months. That place is a revolving door with one NASCAR team after another. One of Todd Bodine's trucks was being unloaded on the afternoon we were there. I noticed that it had several bulges in the sheet metal work, towards the rear of the cab on each side, that had been taped in place for testing. The operator said these teams were very secretive about what they were doing and usually arrived at the tunnel with unmarked cars and haulers. They normally bring there own engineers and do they own thing with out the help of tunnel personnel.