RealAir Turbine Duke V2 Engine Tweak

[light_blue_yonder 15]

The Turbine Duke has been a long favorite of mine, but the turbine lag is becoming a bit bothersome after wading into X-Plane recently.

Since RealAir is no longer in operation, there is probably not going to be a fix for the programming based on old FSX limitations.

I'm thinking of using this summer quarantine time to get into coding. Here's my question:

How easy is it to alter the engine values for the aircraft? I tried to use the new fuel flow table in P3D V4, but it just messes up the max %NG value and the max torque slightly.

How feasible do you think it would be to code a custom model for the engine to bypass the flawed P3D system? I'm assuming it's basically impossible for someone with very little coding experience and next to no understanding of the P3D system. But as long as I can be pointed to a place to start, I can start messing around with no real end goal.

EDIT: Altering the fuel_flow_gain to anything other than 0.051 results in the max %NG increasing from the default 102.0 to 102.5 and the max torque to increase from the default 1300lb-ft to 1320lb-ft. This leads me to believe that there is a custom coded engine model that relies on the value being 0.051.

EDIT 2: I managed to restore the default max torque by redefining the max torque to a lower value in the aircraft.cfg, but the %NG problem remains. This doesn't make sense to me, since the fuel_flow_gain is defined based on N1, why does changing it alter the N1 value. If N1 is the independent variable in manipulating fuel flow, why does changing fuel flow make it act like an independent variable that changes N1?

Edited May 21, 2020 by light_blue_yonder

[jabloomf1230 2,445]

Consult the P3d SDK. Fuel flow can be broken into stages by modifying an entry in:

[turbineenginedata]

in aircraft.cfg. It doesn’t fix the problem to my satisfaction, but with a bit of experimentation you can make the response less annoying.

fuel_flow_gain.n

Fuel flow gain constant table.

This can be used in lieu of the single constant (above) in order to make the fuel flow gain a function of N1. The table will load sequentially (.0, .1, .2, ...) up to a maximum of 5 entries. Each entries first value is the N1 input, the second is the scalar at that N1. The constant will be linearly interpolated in between data points.

fuel_flow_gain.0 = 00.0, 0.011
fuel_flow_gain.1 = 25.0, 0.011
fuel_flow_gain.2 = 60.0, 0.05

[light_blue_yonder 15]

Yes, I did try that. Here's the problem I found with that:

If I increase it by enough to have a 12s spool time, then the max %NG and max torque will be higher than default. Not by much, but since I don't understand the P3D system well, I don't know what else may be messed up.

In other words, it's exactly as you stated. Doesn't fix the problem much, but a little bit better as a compromise. What I'm trying to do is essentially trying to reverse-engineer RealAir's code, which is next to impossible for someone like me without some guidance.

[jabloomf1230 2,445]

It’s tough for a user to figure out, let alone modify the flight characteristics of a complex aircraft designed for P3d, such as anything designed by either PMDG or A2A. The RealAir TD falls into that complex category also.

First, there are the performance settings in aircraft.cfg. Then there is the AIR file which can be modified easily with the tool AirWrench. There are discussions on fsdeveloper.com regarding modifying the AIR file for turbine airplanes. Lastly, there is code added by the developer of the aircraft. It’s a witch’s brew of settings.

[DavidP 1,408]

15 hours ago, light_blue_yonder said:

the turbine lag is becoming a bit bothersome

I have no experience flying real turbo props but understand lag is an issue.

Do you have such experience or are you basing the issue when comparing to XPlane?

Just curious..

[Bert Pieke 7,584]

Just to add to that.. The original Author, Rob Young was no slouch in this department..

By all means fiddle with it, to your heart's delight.. but I am not holding out hope for a breakthrough..

[fppilot 4,352]

I am perfectly satisfied with the current characteristic as it is closer to reality than subtracting that characteristic.  I have logged over 500 hours in the TDv2 and it is my go to aircraft for longer GA flights at higher altitudes and speed.  I have an operations sheet developed for the TDv2 that might help you out.  But it is an aircraft you must get ahead of and stay ahead of.  Sounds like a contradiction, but it is not. The greatest learning curve is slowing, descending, and with the approach.  All require managing speed by managing and mastering controls during those phases.  Turbines are that way in the real world.  

[light_blue_yonder 15]

1 hour ago, Bert Pieke said:

Just to add to that.. The original Author, Rob Young was no slouch in this department..

By all means fiddle with it, to your heart's delight.. but I am not holding out hope for a breakthrough..

Yes, I do understand that, and he is orders of magnitudes more knowledgeable in this area. However, he developed this software for P3D back when some of the newer V4 SDK tools are not available to him. That's what I'm trying to see if I can do.

59 minutes ago, fppilot said:

I am perfectly satisfied with the current characteristic as it is closer to reality than subtracting that characteristic.  I have logged over 500 hours in the TDv2 and it is my go to aircraft for longer GA flights at higher altitudes and speed.  I have an operations sheet developed for the TDv2 that might help you out.  But it is an aircraft you must get ahead of and stay ahead of.  Sounds like a contradiction, but it is not. The greatest learning curve is slowing, descending, and with the approach.  All require managing speed by managing and mastering controls during those phases.  Turbines are that way in the real world.  

I don't have a problem with the aircraft as it is, and I'm perfectly able to fly it as it is. (Perfectly able to fly it, not able to fly it perfectly. 😆) I'm just trying to mess with it a bit just for the sake of it making it a tad bit better for not good reason other than I have time and RealAir is no longer in business.

I'm actually curious if anyone has looked into RealAir's coding and knows of their system somewhat, since I have no clue where to start other than the aircraft.cfg and that isn't doing me much good.

Edited May 21, 2020 by light_blue_yonder

[robert young 2,897]

I'm interested in the concept of dividing the fuel flow params into three parts and I did not know that was possible, but I suspect the result will be very similar to using one param. As I recall (it was quite a few years ago now), you can very easily ramp up the turbine response but unfortunately the core coding of the FSX/P3d prop turbine is that it results in gross over speed of the propellor and out of control N1 particulary on start up and initial take off acceleration (if I recall correctly). So the problem is not getting the engine to respond quicker, but to stop getting overspeeds.

I spent literally months and hundreds of hours adjusting the relationship between several params in order to get as quick a response as possible without overspeeds. If it wasn't for the start up routine, which is extremely sensitive to prop and engine overspeeds, the overall response times could have been shorter. Once you have taken off and speed is around 120 knots or greater, you can abuse the throttle levers and not much will happen, but on the ground shoving them forward fast will lead to an overspeed. Sadly the original Turbine Engine tweaker way back at FSX release time forgot to add sufficient numbers of params to control all of this.

The key params in the aircraft cfg that combine to influence the response time and subsequent overspeeds if not tuned carefully are as follows:

Propeller Diameter

Propeller MOI (inertia)

Prop_tc (time constant)

minimum on ground beta

beta max and min

fuel flow gain

There are a few more but the above are the important ones.

These params all interact and you have to find the sweet spot between over-sluggish response and over speeding prop especially on the ground. One big influence is the throttle condition lever which by FSX default starts the aircraft in high position. You need to get this lever back to idle condition or the overspeed on start up will be more likely and taxiing can get out of hand.

That all said, the delay in response is not a disaster, just a little slower than ideal.

Taxi tip. I use the beta and reverse range to do most taxiing.I tuned the throttle beta/reversers so you can gently pull them a little backwards of idle to give a little bit of reverse. That contains the forward movement much better than braking. BTW one thing I regret is the wheel brake settings. Unfortunately the differential_braking param appears broken in both FSX and P3d. So you have to adjust the general or toe brake scaling to clamp down on the oversensitive directional control with typical rudder/brake pedals. You need to get the scale well down (0.5 or 0.6) to stop the twitchy steering but then you will have less braking power, which is not too much a problem given the powerful beta-reverse throttle control.

Hope this helps!

[Seagreen 5]

This reminds me of an incident in the comedy "Frasier" where, along with his brother. he returns from dining out and when asked how it was, replys, "perfect but with one tiny flaw we can pick at all night" to which the inquisitor retorts " Oh!. Just how you like it".

I remember there was a real and personal reason why RealAir had to stop but given Robert Young's astonishingly cultured, informative and reasoned reply to the initial query, I wonder if he "can be got at" and if enough people could persuade, cajole, bully, force, blackmail, stamp our feet, scream, petition or whatever it would take to make RealAir get back together and create a private jet along the lines of a Lear or Challenger etc. I know Gulfstream have banned any attempts at modelling their products so that is out.

Look at the quality of Sean's Diamond!

Can you imagine that attention to detail and love these guys put into their work, the clarity of the guages and the unsettling realism of their sounds of, say, the "Legacy", put into a jet? There isn't a company to touch them.

I think God in His heaven is wondering the same thing and would go as far as to say should they not, they will have to answer to Him! Perfection is their calling and duty!

I am getting on and hope to see it in my lifetime.

[Sabre57 46]

The turbine duke can be a finicky beast.  But when treated kindly with slow deliberate engine control management it really is a fun beast.

[dbw1 444]

I have a couple thousand rw hours on the Pt6-34 and to be honest I am quite happy with the characteristics that Real Air created.