Jussi

I calculated quickly from that first measurement and took only 9 first seconds just before that lump, I got 0.36 Cd average if using stock 1.93m2 Frontal area BUT I didn't notice that down hill, so it make that Cd worse.
But that result already shows that you don't have stock 993TT..

Are those Height readings in feets or in meters?

It is quite laborious to zoom that little picture and find out every speed point of every second, so that is has many accurate decimal..
I can get best results from Excel model table where are speed and time values as pairs.

Jean

Hi Phelix, a lot of great things can be said about the abundancy of information in your tests as well

I was referring to strictly rolling resistence of tires and hubs/bearings, not transmission losses as they don't apply to this test that is being done in neutral. Otherwise you would be 100% right.

Jussi

So reading from picture is little inaccurate but that average value tells more.
I can publish Jean's GT2 calculated Cd values from every second:
0,47
0,45
0,39
0,45
0,38
0,47
0,49
0,39
0,43
0,44
0,49
0,40
0,47
0,49
GT2 average:
0,44
so that result is reasonable.

Same for Jean's stock 993TT
0,32
0,33
0,35
0,36
0,35
0,41
0,29
0,38
0,31
0,38
0,30
0,34
0,32
0,33
and those make average as:
0,34
and that is also in specs


Toby's "half GT2" in little downhill test:
0,34
0,40
0,32
0,33
0,42
0,42
0,30
0,31
0,38
average:
0,36

So, if you want better results I need data in table (or AX) mode.

But still the most meaninful factor for these calculation is the test environment, you must have a very flat road (and no wind) and still better to do tests for both ways and use that average value. More speed is more accuracy, like in early days we did these rolling tests from 100 to 50 km/h and got inaccurate results, so over 100mph is much better, 200mph is very good
Slowing down time doesn't need to be more than ten seconds.

Jussi

Many of you have asked my "secret calculations".
So here they are:

As you see, Aero Drag is our/speed junkies' enemy

You can virtual test your variables: Car's weight, Frontal area, Cd, etc and see how they effect.

For example, If you have stock weight and shape 993TT and you want to go 220 mph, you must have at least 553 hp plus more to get positive acceleration and how much more you have extra hp over that needed value, that takes you quicker to there.

And what comes to that "Calculating Cd from rolling test" is solved here

(same equation as in first pic but turned to get Cd)
So when you do your tests and get results, please publish them here (or send to me.)
But bewere of speed tickets and don't blame me then

Have a great weekend

Felix

Isn't that loss what's calulated after a chassis dyno run when the rollers are allowed to coast down with the clutch depressed? As in the green line in the graph in the first post of https://rennlist.com/forums/993-turbo-forum/389143-dyno-tests.html

TB993tt

Well many people say how Rennlist saved them some money but they usually mean on some bargain parts.....

Thanks to Jean for raising the coast down test and Jussi for crunching the numbers I decided that in fact having the major surgery (sloped screen, sunroof and gutter delete) on my ~0.36 Cd car which would at best result in a ~0.34 Cd is just not worth it, especially for the $$$$$.
The car was due to go to Germany week after next but I have cancelled the job - thanks guys

We are shortly going to the airport runway so we can do some more of these tests but I am eager to try the 0-300kph using my current configuration. Knowing the Cd of ~0.36 is it possible to estimate a 0-300kph time using 580hp, 4WD, 1600kg - gives me an idea of what to expect - I am thinking 34s ?

Jean

Phelix,

I did not know that they depress the clutch for the coast down. When I was at Manthey (with you ) they just let the car drop RPMs while in gear. They are measuring losses of the drivetrain, gearbox, bearings etc.. In this case we are interested by the tire friction because drivetrain efficiency is calculated separately...At least that's how I see it ?

Felix

Hi Jean,

I always thought that at the end of the run the roller load is removed and the clutch is depressed and the rollers and drivetrain spin down to measure the frictional losses of the system. If the engine was braking the system the rollers they would spin down fairly quickly, no? And engine braking would enter the equation which isn't part of power measurement?

Jean

Felix

It is a bit confusing in fact. If they depress the clutch while coasting down they would be discounting gearbox losses, which are the most substantial losses I believe. Engine braking would not fool the measurements as they are measured by RPM anyway I would think, so it is irrespective of the speed at which the dyno decelerates?

The speed (the time it takes rather) at which the car decelrates while on the dyno (on coast down) is very similar to what you would get on the road, I just watched the video of my run on the Manthey MAHA and it does look like it is engine decelration..

I could be wrong I am just following my own logic, which might be no logic

Kevin

You depress the clutch or bump the trans into the neutral gate on the Maha.. You are measuring the geared losses with the clutch depressed. Unless you have a open diff like the 996TT. The limitd slip will lock up and rotate the gearbox cluster, along with your tires and CV's.

Jean

I have now done more accurate calculations of the drag losses. The previous calculation of CDA included friction and climbing losses. I now can isolate each of these components, so slope variance does not matter anymore as its effect can be pinpointed.. All of this is in relative terms accurate of course.

The total losses are rolling losses, climbing losses and drag losses. So one needs to be able to separate the three factors to be able to determine the closes way possible the CDxA alone , and subsequently the aerodynamic losses alone.

There is a set of quite unfriendly formulas to be used, my approach seems to be more detailed and complex than Jussi's (doesn't make it better or worse), and lets me separate clearly the three components by calculating the friction and slope losses rbased on the actual datalogs, rather than using accepted values found on the net, which is how used to do it earlier. Not much difference, however more accurate.

The results are as follows for my car at 195kph (circa 122mph):

Total real CDxA (almost impossible to separate CD from Area on a non stock car): 0.83

Total resistance due to slope: 11 hp (8% of total losses)
Total resistance due to rolling: 28hp (19% of total losses)
Total drag losses: 108 hp (73% of total losses)

The results are as follows for my car at 135kph (circa 85mph):

Total resistance due to slope: 8 hp (13% of total losses)
Total resistance due to rolling: 19hp (30% of total losses)
Total drag losses: 35 hp (57% of total losses)

Interesting to see how the percentages move with speed..

And here is the graph of the losses throughout the run;


I cannot cover it all or post spreadsheet screen shots as they are quite large, but I am happy to answer any questions or show more data.

Jean

Might be accurate what you are saying, but it is not what my video is showing or what they did with my car, I also would not understand how the losses are calculated if they leave accurate gearbox losses out of it by depressing the clutch. Maybe they do both?

Felix

But wouldn't gearbox losses be included even with the clutch depressed? Only the engine is "taken out of the loop".

Jean

Doh! Yes!

mickfluff

Last reply was 08.... no mention here by anyone of how much value will my car loose if I mod this or that... how much is my car worth? I need an agreed upon value insurance quote? how much will this or that car sell for?..... blah blah blah....This is an old school Rennlist thread that I would read back in the day when I had a Honda S2000 and thought owning a porsche wasn't possible. Where did this passion go? Imo the surge in pricing on these fine automobiles has done nothing but impacted the enthusiast world in a negative way. Sure some cars left in the hands of still true enthusiasts but not nearly as many.

Well it's 2017 and have owned multiple P cars.... currently have a modded 997 and an offer accepted this week on a 993 that goes to ppi this week end... not going to ask people what it's worth, if I got a fair deal or track it's ****ing value in my hands. I dug up this thread poking around for info as I plan to mod it, add aero, interior pieces, chop off fenders, add wheels, mess with suspension and more!

Carry on gents with your fine 993TT cars and hope to see a forum flooded again with threads like this....

Be well