Feedback Thread - F-14 Tomcat patch Jan 27th 2022

[IronMike]

7 minutes ago, Spurts said:

Track analysis.  Both launched from ~0.8M 36,000ft at 1.05M closing target at 34,000ft from 40nm range

                         AIM-120C5                   AIM-54

Max Speed      3.59M@8s                    3.45M@30s

Loft                   22deg                           13.6deg

Max Alt             60,414ft                         48,570ft

Loft type           0.0-0.4G until               0.5G pull down

                           18.75nm range             18.8nm

Active                8.13nm                           10.2nm

                           51,779ft                         45,052ft

                            2.32M                           2.42M

                           -25.6 pitch                     -10.6 pitch

Terminal            1.8M  7.1G  -67.8pitch      1.35M  3.8G   -40.4 pitch

                          1:16 ToF                            1:17 ToF

AI drops chaff immediately on launch with Phoenix, not with AMRAAM.

AMRAAM pulls nose to 26.2deg in 2s from launch, Phoenix pulls nose to 13.5deg in 16s from launch.  Sounds like guidance is making the difference here.

 

 

Thank you. Which AIM-54 was that?

[cgstever91]

could this have waited until the other issues were sorted i would assume that the performance of the aim54 would have been seen in testing. i just dont understand why it was pushed in the state its in i can understand that it takes time to fix bugs or to swap to a different api but this is just bad.

[IronMike]

2 minutes ago, Gypsy 1-1 said:

Just curious - having looked at your CFD whitepaper from back in the day, how did you guys get the non-0 and high AOA region of flight or how did you calculate it? What mesh and software did you use? What is standing in the way of switching the 54 to the new API still? Is it not made available by ED yet? 

This is part of switching it to the new API.

2 minutes ago, cgstever91 said:

could this have waited until the other issues were sorted i would assume that the performance of the aim54 would have been seen in testing. i just dont understand why it was pushed in the state its in i can understand that it takes time to fix bugs or to swap to a different api but this is just bad.

Unfortunately no. It is an important step that had to be taken and was needed to continue further from this point. As we said, it is actually still overperforming in pure real life shot replications a tad, and only now, and with your help as well, can we see how guidance affects it. It was grossly over-performing previously and that had to be fixed, too.

[cgstever91]

is there currently a timeline for a fix to make the 54 usable or are we going to have to wait for the guidance integration with the new api. 

 

[dundun92]

4 minutes ago, cgstever91 said:

54 usable

It is usable though, Your very own tacview shows a 40nm shot at 35kft with an AIM-54C can intercept a target that maneuvers at missile active. How is that unusable?

[IronMike]

10 minutes ago, cgstever91 said:

is there currently a timeline for a fix to make the 54 usable or are we going to have to wait for the guidance integration with the new api. 

 

The timeline is asap. However, it is usable, alas at a reduced range for now. Let's please not exaggerate things entirely. The guidance suffers still the same issues as it did before, only that now in conjuction with the more closely matching performance, it produces results that disappoint set expectations, in parts due to under-performing, and understandably so, but in parts also due to expectations that stem from being used to an over-performance that mitigated these issues so far (but not in a good way at all). We will try to fix it asap and match both sides again.

Edited January 27, 2022 by IronMike

[NeedzWD40]

Nothing scientific to report here, just some oddities in performance between the three variants (AIM-54AMk47, AIM-54AMk60, and AIM-54Mk47):

It seems like the AIM-54CMk47 has some really strange changes to its guidance routine that results in a lot more weirdness than the AIM-54A series. For example, the AIM-54CMk47 climbs to an markedly higher altitude, but seems to drag more on descent. It also seems to simply quit tracking in the terminal phase, often flying right in front of the target at longer ranges. When testing launches beyond 70nmi, the missile almost always seems to get confused at goes stupid in the final seconds of flight, whether TWS or STT launches. Under that range, the missile seems to have virtually no issues and tracks very well in either mode. By contrast, the AIM-54AMk47 is able to accurately guide itself and hit targets further out and with greater accuracy, TWS or STT. I was able to destroy a bomber target in STT from 90nmi with the A model with room to spare, a feat that the C model seemingly went stupid over. The Mk60 had virtually no problems with kinematic performance at this range and greater, the only limitation being the AWG-9's ability to track a target.

Beyond that, within about 50nmi the AIM-54C seems to remain the superior variant over either A model, especially against maneuvering targets.

[IronMike]

3 minutes ago, NeedzWD40 said:

Nothing scientific to report here, just some oddities in performance between the three variants (AIM-54AMk47, AIM-54AMk60, and AIM-54Mk47):

It seems like the AIM-54CMk47 has some really strange changes to its guidance routine that results in a lot more weirdness than the AIM-54A series. For example, the AIM-54CMk47 climbs to an markedly higher altitude, but seems to drag more on descent. It also seems to simply quit tracking in the terminal phase, often flying right in front of the target at longer ranges. When testing launches beyond 70nmi, the missile almost always seems to get confused at goes stupid in the final seconds of flight, whether TWS or STT launches. Under that range, the missile seems to have virtually no issues and tracks very well in either mode. By contrast, the AIM-54AMk47 is able to accurately guide itself and hit targets further out and with greater accuracy, TWS or STT. I was able to destroy a bomber target in STT from 90nmi with the A model with room to spare, a feat that the C model seemingly went stupid over. The Mk60 had virtually no problems with kinematic performance at this range and greater, the only limitation being the AWG-9's ability to track a target.

Beyond that, within about 50nmi the AIM-54C seems to remain the superior variant over either A model, especially against maneuvering targets.

Thank you! Could you post us some tacviews for that? Those would help a ton! Thanks again!

[NeedzWD40]

2 minutes ago, IronMike said:

Could you post us some tacviews for that? Those would help a ton!

Can do. FWIW test conditions were in the Marianas with the H-6 and J-11 as targets; I will try and correlate those conditions on the Caucasus.

[IronMike]

1 minute ago, NeedzWD40 said:

Can do. FWIW test conditions were in the Marianas with the H-6 and J-11 as targets; I will try and correlate those conditions on the Caucasus.

Thank you ever so kindly!

[cgstever91]

9 minutes ago, IronMike said:

The timeline is asap. However, it is usable, alas at a reduced range for now. Let's please not exaggerate things entirely. The guidance suffers still the same issues as it did before, only that now in conjuction with the more closely matching performance, it produces results that disappoint set expectations, in parts due to under-performing, and understandably so, but in parts also due to expectations that stem from being used to an over-performance that mitigated these issues so far (but not in a good way at all). We will try to fix it asap and match both sides again.

 

thank you 

[IronMike]

3 minutes ago, cgstever91 said:

thank you 

My pleasure. And thank you for your feedback. Rest assured it is being both heard and taken seriously. No one wants to get the aim54 tweaking and fixing behind us as quickly as possible as we do, I can promise you that.

Edited January 27, 2022 by IronMike

[LF_Luftmanu]

Thank you @HB for a more realistic performing -54 down low. If the guidance issues are fixed, then we can enjoy a more accurate missile than before.

 

I agree with @NeedzWD40 on his analysis. Sadly, I can't provide any tracks now.

 

Kind regards.

Edited January 27, 2022 by Ala12Rv-LuftManu

[cgstever91]

1 hour ago, IronMike said:

Thank you. Which AIM-54 was that?

that was a c

[NeedzWD40]

38 minutes ago, IronMike said:

Thank you ever so kindly!

A second set of tests in more controlled conditions has these results:

The AIM-54C was fired in TWS at ~90nmi, approximately when the "HOT TRIGGER" light came on. The missile climbed to an altitude of almost 105,000ft at a pitch angle of ~45deg. At the top of its climb, the speed had dropped down to ~900KTAS. Missile made no attempt to acquire the target or follow presumed AWG-9 commands; it seemingly gave up on the descent as speed dropped down to under 700KTAS. The second test was done in STT and launch was performed at the same distance, approximately 90nmi. Behavior was identical to the TWS shot.

The AIM-54AMk47 was launched in conditions approximating the AIM-54C's, except that the pitch angle seemingly topped out at 22deg. Altitude reached was similar to the AIM-54C, however energy retention was slightly greater and the missile made a couple of attempts to acquire the target. STT behavior was similar, with the exception that maximum altitude reached was ~110,000ft. The missile did not have the kinematics to hit given the conditions, despite seemingly attempting to acquire the target.

The AIM-54AMk60 was launched in the same conditions as the previous two; however, maximum altitude attained was ~98,000ft with nearly twice the speed: ~1900KTAS. The missile acquired the target twice, then hit with a speed of approximately 1500KTAS. An STT launch resulted in very similar results as well as a hit.

When altering the launch range to ~70nmi, the AIM-54C topped out at ~84,000ft, with a speed of 1450KTAS. The maximum pitch angle attained was 25deg. The missile acquired the target twice and hit. A second altering of the range to ~77nmi resulted in the missile missing, with a maximum height of ~95,000ft but a maximum speed of ~1200KTAS. The missile acquired the target but did not have the kinematics to make a hit.

I've tried to keep test conditions as ideal as possible but naturally there will be some variances in launch altitude, speed, etc. with each track.

Overall, it would seem that the Mk60 variation may be performing too well as it seemingly holds twice the energy as the Mk47 powered variations. However, the AIM-54C clearly has some guidance trouble that isn't affecting either A model, since ~70nmi seems to be a magic barrier as to whether it tracks and guides properly or performs wasteful energy-losing maneuvers. If so desired, I can make some additional tracks of the AMk47 model.

I've attached the Tacview files of these tests that will hopefully give some insight into what is going on.

F-14 AIM54 Tests.zip

[Brody]

Mach 1.0, Alt 23,900'

Shot at 26 and 28 nm, TWS Auto, both trashed

Copy-pasta'd .miz but swapped to f15C with AMRAAM, shot as close as I could to same parameters

Result is AMRAAM in this case slightly out-performing, and has more energy in the terminal phase.

Not a very scientific experiment, but take a look...
 

Tacview-20220127-152132-DCS.zip.acmi Tacview-20220127-150213-DCS-F-14A_IA_PG_BVR.zip.acmi

[ChockP51]

1 hour ago, Gypsy 1-1 said:

Just curious - having looked at your CFD whitepaper from back in the day, how did you guys get the non-0 and high AOA region of flight or how did you calculate it? What mesh and software did you use? What is standing in the way of switching the 54 to the new API still? Is it not made available by ED yet? 

Judging from the mesh preview and post process figure, I believe they use ANSYS fluent. The angle of attack can be adjusted by changing velocity component at the inlet boundary or rotating the mesh but that’s not the usual way we do it.

 

[dundun92]

44 minutes ago, NeedzWD40 said:

Overall, it would seem that the Mk60 variation may be performing too well as it seemingly holds twice the energy as the Mk47 powered variations.

So I looked at the tacview, and what this looks like is that the Mk47s loft profile is broken for some reason. The Mk60 is following a nice, smooth ballistic loft as seen here, without any of the old snap-down behavior (though this would have to be tested more to make sure its not just this specific shot profile)

Meanwhile, the Mk47s are using an odd loft profile where they climb at a seemingly fixed pitch until stall speed, bleeding a ton of speed

EDIT: i looked at another tacview and its not that its using a fixed pitch angle, its just overlofting and staying in the loft too long. Its like the loft is tuned perfect for the Mk60, but on the Mk47 it gets too slow too fast for that loft to work

Edited January 28, 2022 by dundun92

[dundun92]

2 minutes ago, BrodyZ_ch said:

when you outrun your phoenix... any suggestions for 20mm timeline? the harriers want to know

20mm Timeline.acmi 326.62 kB · 0 downloads

Yeah thats the broken loft profile I mentioned above. If you look at the tacview it stalls out at the top of the loft

HB definitely needs to look into this, as this pretty much makes lofted Mk47s useless

[ChockP51]

2 hours ago, IronMike said:

There is a couple things to consider: 1) the CFD data is not everything. Plume effects are not included, so it is likely to be taken as conservative a touch. 2) You need to adjust CFD data, etc to achieve known real life results within a set environment as DCS, and conditions will alter your results, when deviated from those real life shots. 3) The FM is now pretty close and introduced with the new weapons lua FM, which yes is part of the new API. 4) The guidance part is still on the old API, and thus interferes with the performance results.

This is a rather simple run down of things, not everything is or will be this straight forward and it requires a lot of tweaking and trial and error to get it just right. But currently it is not the FM that is off as much as the guidance causing the missile to make turns that are working against it, which sets off the FM. The shots in the whitepaper have been done without loft, in a straight line. Try that, and you will see it matches very closely already (while in fact even over-performing still a tad in these situations). The guidance issues then introduce problems that cause it to under-perform in certain situations, which we are currently dealing with and have to solve together with ED. As Cobra mentioned, it is a multi-step process, and we kindly ask you all to have a bit patience. While it might not seem like it at first hand, this is already a huge step forward and in the right direction, and we will not stop until we reach the point that satisfies us in performance matching real life shots most and above all. Thank you.

 

Interesting didn’t know that the plume effect wasn’t include. The plume normally increase the pressure at the base and thus greatly reduced drag coefficient when the motor is firing. However, these drop in base pressure depend on plume strength and it could be the other way round as well

Edited January 28, 2022 by ChockP51
Wrong pic

[commiccannon]

Did a whole bunch of testing today posting tacviews just incase it helps

15,000 25 NM target 8000 A model 0.9.acmi 35,000 40 NM target 25,000t A model 1.2.acmi 25,000 35 NM target CO-Alt A model 1.1.acmi 25,000 35 NM target CO-Alt C model 1.1.acmi Tacview-20220128-110450-DCS.zip.acmi Tacview-20220128-105529-DCS.zip.acmi Tacview-20220128-104848-DCS.zip.acmi Tacview-20220128-104048-DCS.zip.acmi Tacview-20220128-103633-DCS.zip.acmi Tacview-20220128-103223-DCS.zip.acmi Tacview-20220128-102753-DCS-TomcatTesting.zip.acmi Tacview-20220128-100752-DCS-TomcatTesting.zip.acmi Tacview-20220122-075158-DCS.zip.acmi Tacview-20220122-072101-DCS-F-5E Mission - AWACS.zip.acmi

[IronMike]

Thank you!

[Callsign JoNay]

I did a fun before and after comparison on the MK60. Top half of the screen is before, bottom half is after.

 

Pre-patch missile achieves a full Mach faster top speed over the post-patch missile, but their speed/energy was similar in the terminal phase. 

EDIT: Missile shot is at 5:08 by the way. I included the flight profile leading up to the missile shot to show any inconsistencies before the Fox-3, but feel free to skip.

Edited January 28, 2022 by Callsign JoNay

[nighthawk2174]

3 hours ago, IronMike said:

There is a couple things to consider: 1) the CFD data is not everything. Plume effects are not included, so it is likely to be taken as conservative a touch. 2) You need to adjust CFD data, etc to achieve known real life results within a set environment as DCS, and conditions will alter your results, when deviated from those real life shots. 3) The FM is now pretty close and introduced with the new weapons lua FM, which yes is part of the new API. 4) The guidance part is still on the old API, and thus interferes with the performance results.

This is a rather simple run down of things, not everything is or will be this straight forward and it requires a lot of tweaking and trial and error to get it just right. But currently it is not the FM that is off as much as the guidance causing the missile to make turns that are working against it, which sets off the FM. The shots in the whitepaper have been done without loft, in a straight line. Try that, and you will see it matches very closely already (while in fact even over-performing still a tad in these situations). The guidance issues then introduce problems that cause it to under-perform in certain situations, which we are currently dealing with and have to solve together with ED. As Cobra mentioned, it is a multi-step process, and we kindly ask you all to have a bit patience. While it might not seem like it at first hand, this is already a huge step forward and in the right direction, and we will not stop until we reach the point that satisfies us in performance matching real life shots most and above all. Thank you.

 

Is motor drag reduction simulated now?

[turkeydriver]

Please note the missile guidance is an ED limitation right now- its not limited based on what HB can provide- just what the sim can provide currently.