Statistical investigation on flight test results.

I'm sure there are other factors as well that can not be reproduced in a video game to compare to the original aircraft's climb rate...example: panel fit's , rivet exposure drag etc....it also seems to me that the true rudder performance can not be achieved in a video game as well...keep at it though! :)

I'm sure there are other factors as well that can not be reproduced in a video game to compare to the original aircraft's climb rate...example: panel fit's , rivet exposure drag etc....it also seems to me that the true rudder performance can not be achieved in a video game as well...keep at it though! :)

Rivet exposure drag did help the aircraft get better speed, but not to a scale people might expect.

I read somewhere that for instance the Spitfire was tested with flush rivets and regular rivets and the difference in terms of max achievable airspeed was around 5 mph, so around 8 km/h.

Rivet exposure drag did help the aircraft get better speed, but not to a scale people might expect.

 

I read somewhere that for instance the Spitfire was tested with flush rivets and regular rivets and the difference in terms of max achievable airspeed was around 5 mph, so around 8 km/h.

Oh OK that's very interesting Chuck, thanks

Ok, why is this called a statistical investigation? It sounds more like calculus to me. I reached differential equations in college and decided that was enough, and I have forgotten most of what I learned.:book:

Rivet exposure drag did help the aircraft get better speed, but not to a scale people might expect.

 

I read somewhere that for instance the Spitfire was tested with flush rivets and regular rivets and the difference in terms of max achievable airspeed was around 5 mph, so around 8 km/h.

I believe the difference was much bigger - 22mph as tested on Spitfire K5054. It has been tested to determine, which areas were having significant influence on the overall drag and which areas they could get away with cheaper regular rivets in the production series of the type.

Not that it matters, and sorry for OT.

Great thread. No matter what flight models the ED will come up with, there will be still people complaining. :D Perhaps an official set of performance graphs / dog houses would be beneficial.

3. There must be found a very special reasons to explain why 109K4 ROC (in real flight, not in calculations!) must be 22.5 m/s, so far from the trendlines.

May be changes in airframe are wider than it looks like? May be drag is greater for some reason, wing bulbs indeed? Later war German aircraft are known to had building flaws but we here have a "perfect" model, may have something to do with calculated vs real charts mismatch? Just a guess.

Whatever, this is like a parallel Universe, people complaining about their Luftwaffe rides are overpowered :lol: :lol: :lol:.

S!

After reading of this, I do not find many common points with what you say qualify as realist...

http://kurfurst.org/

Again... huh?

lol...just a theory I have :)

with what you say qualify as realist...

With what WHO says?

Now Jojo is a maths man. You gotta be in this line of activity. I could, if I forced myself, to follow the math, but since I don't, I'm gonna back off this thread pretty soonish. But not before I say this: if you want to retort what's said in the OP, this is not the way to do it. So please, pretty please, present YOUR calculations next and I'll see if I can be arsed to follow them :D

OK, carry on gents & gentettes... now where the hell did I put me coat?!?

You lost the point: we are speaking about the theoretical estimation that somebody wants to see as a true reference, so we would better to search for the reason what is wrong with this estimation.

S!

Yo-Yo: I don’t think I quite follow your logic in this thread: On the one hand we are to dismiss the calculated results from Messerschmitt and Focke-Wulf because they were calculated and not based on flight test.

However, you yourself now do a calculation to shows trends that coincide with your viewpoint that the climb rates in DCS are now correctly modeled. Is that not a double standard?

In addition, I don’t think you can make the sort of simplifications you do in your method of calculation:

In post #1 you claim: ”The second simplification is that we intentionally neglect exhaust thrust and radiator drag. First of all, at climb they have opposite signes and approximately equal”.

Where did you come up with that? Where is the proof that it is so?

I’m not at all convinced you can make such simplifications and in fact I think I think the very reason you get the results you do proves the point. But if we are to treat this scientifically, If you claim that you can neglect these since they cancel out then the onus is on you not me to prove it is so.

I summed up my arguments about what I think the climb rate should be for both the K4 and the Dora and my method of calculation here so no need to reiterate that:

http://forums.eagle.ru/showpost.php?p=2398986&postcount=3

Regarding the trend chart you posted in post #8, I in fact compiled a similar figure showing the climb rate for the Me109 with different power to weight ratios (P/W) about a year ago which I have attached below. Granted, this compilation is for 2 Km altitude but translating this to the sea level then the climb rate target for the K4 at 1.8 ata should IMHO be around 23.3 m/s not the 25-26 m/s which seems to be the currently targeted climb rate in DCS.

So to sum up: Sorry, but I’m more inclined to trust the Messerchmitt and Focke-Wulf calculations rather than your calculation as outlined in post #1 above. Especially since they agree so well with my own calculations. :)

Pilum,

Seeing yo-yo's graph I think it's pretty clear why he calls his argument statistical. Compared to the trend line established by the other 109s the 109K-4 data point is an influential point. Including it in the regression would make dVy/dN for the 109 much lower.

I'm not clear on what role mass is playing in the graph, or if it is play a role at all (as yo-yo says he is assuming 3400kg for the Spit and 109?).

Pilum,

 

Seeing yo-yo's graph I think it's pretty clear why he calls his argument statistical. Compared to the trend line established by the other 109s the 109K-4 data point is an influential point. Including it in the regression would make dVy/dN for the 109 much lower.

 

I'm not clear on what role mass is playing in the graph, or if it is play a role at all (as yo-yo says he is assuming 3400kg for the Spit and 109?).

Well I think it is a weak statistical argument: If you look at the trend line for the 109 (thin brown line) this is drawn based on a scatter of data points at 1300 hp and then just one point at 1450 hp. Now if the point at 1450 hp was just a teeney-weeney bit higher or lower then the result at 1800 hp (i.e. the K4 DB605D at 1.8 ata) would be impacted significantly. So any error on the part 1450 hp value has a huge impact on the results.

Hardly a sound statistical foundation surely? OTOH, if you read off the Me109 climb rate at 1800 hp in Yo-Yo's figure in post #8 you get 23.5 m/s which I can only agree with ;)

Look more closely. There are two points for 1450hp. Also, some of the lower points are for a little more than 1300hp, and others a little less than 1300hp. So there are three different bhp ratings.

But using the thin brown line in your own figure in post #8 this shows a climb rate of 23.5 m/s at 1800 hp which agrees with my calculation Yo-Yo :)