Spit mkIX level speed inaccuracies

[Coopes]

So i ran a test with a little help.

At sea level i hit 330 mph TAS which is only 336mph TAS on your chart

At Angels 22 could not hit 18 pounds.
Hit 18 pounds @20500ft with max speed of 389mph TAS

So only issue i found may be with the Full throttle height.

Last test was max climb from deck to 10,000 feet and was within 3 seconds of the actual time.

Biggest error was roughly 5% when up @ Angels 22

At 20500 ft FTH FS had error of 3.7%

btw i did not use the gauges but the actual data the sim is spitting out on the info bar.

I just found out that there is a known issue with the FTH that varies from map to map and temps that they are trying to track down. Overall its fine

[Dickie]

When did RAF switch to BAM100 Octane fuel? I wonder if maybe the tests that ED used were based on performance statistics using 87 octane as opposed to 100. However if RAF were almost exclusively using 100 Octane by the time they started rolling out merlin 66's then I dont see that being a factor

May 1940. But that only improves boost at lower altitudes anyway, below 12k ft

[Vanguard]

So i ran a test with a little help.

At sea level i hit 330 mph TAS which is only 336mph TAS on your chart

At Angels 22 could not hit 18 pounds.
Hit 18 pounds @20500ft with max speed of 389mph TAS

So only issue i found may be with the Full throttle height.

Last test was max climb from deck to 10,000 feet and was within 3 seconds of the actual time.

Biggest error was roughly 5% when up @ Angels 22

At 20500 ft FTH FS had error of 3.7%

btw i did not use the gauges but the actual data the sim is spitting out on the info bar.

I just found out that there is a known issue with the FTH that varies from map to map and temps that they are trying to track down. Overall its fine

If that is the case, then it looks like the gauges are inaccurate rather than the actual speed.

[Biggs]

Did another test tonight. This time taking a page from Coopes and doing it offline so that I could read the true altitude and speed from the external view.

However, Mesa made me a server with 'static weather' 15°C, (this is standard temperature according to International Standard Atmosphere) no wind, no cloud, and a QNH of 29.92.

I was air spawned at 18,000, full fuel full ammo. I flew trimmed out as best as I could and held the RPM down at 2500 initially to keep the Rad flaps closed as I built up speed, then went full RPM to achieve the max speed. I had Mesa pause the server when the rad flaps went down (air speed will not advance further when they are open), and read the results...

IAS 247kts @18,000ft = 371mph TAS | 19.5mph too slow
IAS 248kts @19,100ft = 378mph TAS | ~16 mph too slow (NOTE:This was it's FTH, above this, boost started to drop below +18.1)
IAS 243kts @20,000ft = 376mph TAS | 23mph too slow
IAS 235kts @22,000ft = 377mph TAS | 30mph too slow (27mph, if you're using 404mph as max)

@22,000ft the boost gauge should still read +18 but was on +16, however, the actual setting was +15.6 according to the external view readout.

So there you have it. The DCS spit doesn't not hold boost at the appropriate altitude and is anywhere between 20-30mph slower than it should be at seemingly every altitude.

Funny enough, that 376mph at 20K I achieved is practically identical to the top speed of the Spitfire mkV... On second thought, there's absolutely nothing funny about that.

Sorry fellas, Spit's broke as fuck.

[Thaine]

What conversion for IAS to TAS did you use, or was TAS a direct readout? Did you have any indications of ambient density, pressure or temperature?

[Onebad]

http://www.hochwarth.com/misc/AviationCalculator.htmlWe used this calculator to converse IAS into TAS using corresponding altitude each time. Air density and pressure was whatever DCS models with temperature set manually for the server we were testing on. Actually we made two sets of tests using two different air temperatures, both brought different top speed results aswell as different FTH's, each time the Spitfire was underperforming when compared to historical data.

And yeah, as Biggs said, the DCS Spitfire is on average about 5 mph faster than the Mk.V and much slower than the Mk.IX.

[Coopes]

Like i said i only found issues at the FTH. But when an insider let me know they know they have a bug with FTH and are looking into.
I tested in standard atmosphere at 15 degrees centigrade as was in Biggs chart. Different temperatures will give different results.
I made close to the speeds stated apart from the aforementioned.

If yours certain its broken we cant do shit about it here, go and post in the spitfire bugs section.

[Vanguard]

Was doing some flights with the Mustang up at 26,000 ft (supposed FTH with War Emergency) but was only pulling 61 inches with full speed ram, seems to not develop full power high enough, similar to the Spit.

[Bounder]

I was chatting to the guys on the DCS TS and was asked to have a look at the thread.

It looks like you are recording results from the outside view in single player (which shows TAS) with correct atmospheric conditions - this is good.

I think the major issue for me is you are comparing to just one speed test; as Coopes says I believe DCS uses a number (greater than one) of original test results and compares the DCS model to some sort of average. Every different individual RL aircraft will have variance in top speed and using just one data set is very limited, especially if DCS are using more. You really should compare against the RL data Yo-yo shared that was used to model the Spitfire. If that doesn't match up you have a case.

Stuff to think about when reporting results: as far as I know for speed tests DCS operates a tolerance of +/- 5% to real data so it's best to also include all speeds as a percentage (obviously include the raw data too). The data from the first post translates as follows:

@ 10k the IAS should be 337, I could only manage 305-307 IAS - a 32 MPH difference = - 8.9%
@ 18k the IAS should be 306.5, I could only manage just over 280 IAS - a 26 MPH difference = - 8.6%
@ 22K the IAS should be 300, I could only manage just over 270 IAS - a 30 MPH difference = - 10%

These results indicate the Spitfire is outside of tolerance limits but only against the one Spitfire test you selected. As said above, comparison needs to be against more RL data, preferably the data Yo-yo has posted.

Couple of other notes: results should always be reported in exact values e.g. @ 18k the IAS should be 306.5, I recorded 281 IAS maximum (definitely never "I could only manage just over 280 IAS"). I know this sounds harsh but you can't make accurate assessments on inaccurate data. All data should be video recorded and presented to show results and methodology. Don't use the in-game track file but instead use a video capture program e.g. Nvidia shadow play / Fraps etc etc

I'll try and add more as I think about it. Don't get me wrong, I think you're all doing a great job but the methodology must be scrutinised so that if/when you present your case it is strong and wont fail at the first hurdle.

Good luck!

[EDIT] just one more thing .... the radiator in the original test you are comparing against was noted as closed. Obviously we can't set it closed in DCS and this is a difficult variable to reproduce. It was said in one post that in your tests the aircraft was taken up to altitude at 2500 rpm so the radiator was closed and then the rpm was set to 3000rpms for the test. Need to be careful here as when you increase rpm the radiator is likely to open as more heat is generated at 3000rpm. You need to visually check the radiator is closed when at test rpm and manifold pressure. It might be worth running at 3000rpm and watching the radiator, see if it stays closed or opens and if it is opening does it close after the initial power increase? It's a very difficult thing to replicate in DCS but is worth thinking about.

It's difficult taking account of all of this when performing the test solo. Could somone host a small server with a mission running the correct temp / atmospheric pressure with outside views enabled. If so, one person could perform the test while a second watches the test in F2 view. That way the person viewing can watch and record TAS and check stuff like the radiator whilst the other person concentrates on flying?

[Biggs]

I was chatting to the guys on the DCS TS and was asked to have a look at the thread.

It looks like you are recording results from the outside view in single player (which shows TAS) with correct atmospheric conditions - this is good.

I think the major issue for me is you are comparing to just one speed test; as Coopes says I believe DCS uses a number (greater than one) of original test results and compares the DCS model to some sort of average. Every different individual RL aircraft will have variance in top speed and using just one data set is very limited, especially if DCS are using more. You really should compare against the RL data Yo-yo shared that was used to model the Spitfire. If that doesn't match up you have a case.

Stuff to think about when reporting results: as far as I know for speed tests DCS operates a tolerance of +/- 5% to real data so it's best to also include all speeds as a percentage (obviously include the raw data too). The data from the first post translates as follows:

@ 10k the IAS should be 337, I could only manage 305-307 IAS - a 32 MPH difference = - 8.9%
@ 18k the IAS should be 306.5, I could only manage just over 280 IAS - a 26 MPH difference = - 8.6%
@ 22K the IAS should be 300, I could only manage just over 270 IAS - a 30 MPH difference = - 10%

These results indicate the Spitfire is outside of tolerance limits but only against the one Spitfire test you selected. As said above, comparison needs to be against more RL data, preferably the data Yo-yo has posted.

Couple of other notes: results should always be reported in exact values e.g. @ 18k the IAS should be 306.5, I recorded 281 IAS maximum (definitely never "I could only manage just over 280 IAS"). I know this sounds harsh but you can't make accurate assessments on inaccurate data. All data should be video recorded and presented to show results and methodology. Don't use the in-game track file but instead use a video capture program e.g. Nvidia shadow play / Fraps etc etc

I'll try and add more as I think about it. Don't get me wrong, I think you're all doing a great job but the methodology must be scrutinised so that if/when you present your case it is strong and wont fail at the first hurdle.

Good luck!

[EDIT] just one more thing .... the radiator in the original test you are comparing against was noted as closed. Obviously we can't set it closed in DCS and this is a difficult variable to reproduce. It was said in one post that in your tests the aircraft was taken up to altitude at 2500 rpm so the radiator was closed and then the rpm was set to 3000rpms for the test. Need to be careful here as when you increase rpm the radiator is likely to open as more heat is generated at 3000rpm. You need to visually check the radiator is closed when at test rpm and manifold pressure. It might be worth running at 3000rpm and watching the radiator, see if it stays closed or opens and if it is opening does it close after the initial power increase? It's a very difficult thing to replicate in DCS but is worth thinking about.

It's difficult taking account of all of this when performing the test solo. Could somone host a small server with a mission running the correct temp / atmospheric pressure with outside views enabled. If so, one person could perform the test while a second watches the test in F2 view. That way the person viewing can watch and record TAS and check stuff like the radiator whilst the other person concentrates on flying?

Thanks for the input Bounder, but I think you should have a look at my second test. Post is on page 2, done May 4th

I was in a server Mesa created(read earlier post). All the atmosphere conditions were at nil, temp 15 degrees C, and he enabled external view and crucially important, he was able to pause the server. So when he saw my rads begin to open he would pause everything and we would take a reading. It was not a variable because I had already reached the fastest speed I could well before the rads began to open... when they did it merely meant that I could stop the test at that alt and could move on. There was no 'close to' in this measurement, everything was as accurate as a seasoned sim pilot could achieve.

My approach for this second test was not to try and match the speeds that BS.543 achieved, but to just try and come close to the WELL documented top speed for the LFmkIX @ 21-22000ft which is anywhere between 404 and 408 TAS... I would have been happy at 400 honestly.

I didn't even come close. @22000ft - IAS 235kts or 377mph TAS

nearly 30 mph off the reported top speed raises a few eyebrows.

If Yo-yo is interested I will try and find a way to record a track (Shadowplay doesnt seem to like DCS on my machine). but I had 2 other people helping me out with this test and they had eyes on what I was doing and can attest to my findings.