investigating Thrust to weight ratio: confused

[bkthunder]

Nice work Ironhand! 

Did you test at M 0.8 just to add a data point?

 

 

Just to clarify, the point I'm making is that the MiG-29 is the only aircraft - among those tested - that can't maintain speed in a climb (from sea level to 6k feet) given the same T/W ratio (all aircraft use static thrust to calculate), and given the same starting speed, 3g pull and 80 degrees climb. 

In your spreadsheet you can see that up to 2000m (about 6500ft) the T/W ratio is >1  and I maintain that at the low speed I tested at, differences in drag between the different aircraft are negligible, and if anything the MiG-29 should be less draggy than the F-18.

 

So, unless the F-18, F-16 and F-15 have a much better T/W ratio at low speed, I don't see why they all have better climb performance than the MiG-29.

I guess now we should do the same test at M 0.8 with the other airplanes, or @Yo-Yo could shine some light on the matter (maybe U.S. engines are better at M 0.5?)

 

[Ironhand]

I chose M 0.8 because, according to the chart, that would allow me both the altitude and time to record any real changes in speed. Low speed and consequently low altitude make that more difficult. That plus the fact that M 0.8 was about as fast as I could go without introducing trans-sonic and supersonic variables into the situation (assuming DCS models them).

 

I don't know the reason for the differences among the aircraft. Nor would we know if something was amiss unless we had thrust-speed-altitude charts for the other aircraft. Static tests are just that. They won't demonstrate performance differences among engines at various altitudes and airspeeds. We can make assumption but that's all they are--assumptions..

 

Concerning your flight in the spreadsheet, I actually wasn't going to include it but I thought it interesting. At the 1000 m mark, you were still pulling 1.6 Gs and had an AoA of 9°. So drag, ect would have been greater than later in the flight. But since I have no way of quantifying the various drag components, I went with the 1.1. But effectively it would have been less than that. I included it because it was similar to my 1.1 data point. Somewhere after that point, you start losing airspeed. Whether it's exactly at 1.1 or something a bit less would take creating more frequent data points. I might do that at some point...or I may not bother.

Edited October 3, 2021 by Ironhand

[Cmptohocah]

Any chance you guys can verify the sustained turn rates I have posted earlier?

 

This can verify correct thrust-to-drag ratio in one go. I don't have access to DCS at the moment

[bkthunder]

10 hours ago, Cmptohocah said:

Any chance you guys can verify the sustained turn rates I have posted earlier?

 

This can verify correct thrust-to-drag ratio in one go. I don't have access to DCS at the moment

17 hours ago, Ironhand said:

I chose M 0.8 because, according to the chart, that would allow me both the altitude and time to record any real changes in speed. Low speed and consequently low altitude make that more difficult. That plus the fact that M 0.8 was about as fast as I could go without introducing trans-sonic and supersonic variables into the situation (assuming DCS models them).

 

 

 

At least from my side, I am trying to eliminate drag from the variables as much as possible, hence why I tested at low speed. 

 

Mach 0.8 doesn't allow you to clearly test for thrust:

 

- you are adding speed (=drag) so differences between airframes and their drag have a stronger influence on the outcome

- the initial speed makes it so that the first part of the climb (up to ~6k feet where engine thrust is unaffected by air density) happens due to stored energy rather than pure thrust out the back. 

 

17 hours ago, Ironhand said:

Concerning your flight in the spreadsheet, I actually wasn't going to include it but I thought it interesting. At the 1000 m mark, you were still pulling 1.6 Gs and had an AoA of 9°. So drag, ect would have been greater than later in the flight. But since I have no way of quantifying the various drag components, I went with the 1.1. But effectively it would have been less than that. I included it because it was similar to my 1.1 data point. Somewhere after that point, you start losing airspeed. Whether it's exactly at 1.1 or something a bit less would take creating more frequent data points. I might do that at some point...or I may not bother.

 

 

I maintained a constant 3g pull in all tests for all airplanes. If you check the track/tacview you can see that in all cases I maintain 3 g up to about 4000ft, then it immediately drops down to 0.5-0.9g during the climb. 

I.E. I was not pulling more g or for longer in the MiG-29 than in the F-18, F-16 etc. 

[Exorcet]

52 minutes ago, bkthunder said:

 

At least from my side, I am trying to eliminate drag from the variables as much as possible, hence why I tested at low speed. 

If you want to take drag out, try level acceleration instead of a climb.

 

Engines off (in FC3 this doesn't impact flight controls or displays) and decelerate at constant altitude, you get the drag force from F=ma (some error from windmilling engines).

 

Then accelerate at full power, F=ma gets you the combined thrust and drag force, but you can just subtract drag since you know it from the deceleration test.

[Cmptohocah]

6 minutes ago, Exorcet said:

If you want to take drag out, try level acceleration instead of a climb.

 

Engines off (in FC3 this doesn't impact flight controls or displays) and decelerate at constant altitude, you get the drag force from F=ma (some error from windmilling engines).

 

Then accelerate at full power, F=ma gets you the combined thrust and drag force, but you can just subtract drag since you know it from the deceleration test.

I was having in my mind something similar:
Mainly testing the produced trust at sea-level by measuring acceleration from 0 to 100km/h for example. This would give a good measurement of static thrust without too much wind resistance. Only thing that needs to be determined before that is the dynamic friction force produced by the wheels, which can also be measured by rolling the aircraft to a some amount of ground speed and then measuring the time/distance it takes to come to a complete stop with the engines spooled down. Then it would be just a matter of F=ma with a fairly low error which would be produced by drag.
Sadly I can't test this yet my self.

[Ironhand]

16 hours ago, bkthunder said:

…- the initial speed makes it so that the first part of the climb (up to ~6k feet where engine thrust is unaffected by air density) happens due to stored energy rather than pure thrust out the back…

…

I maintained a constant 3g pull in all tests for all airplanes. If you check the track/tacview you can see that in all cases I maintain 3 g up to about 4000ft, then it immediately drops down to 0.5-0.9g during the climb. 

I.E. I was not pulling more g or for longer in the MiG-29 than in the F-18, F-16 etc. 

 

Well…at this point I am thoroughly confused. The chart we’ve been referencing indicates that thrust decreases as altitude increases starting from sea level. As altitude increases, air density decreases. Yet somehow before, at, or after 6k feet (not sure which), air density no longer has an effect on thrust. So I’m confused. I’m obviously not understanding something.

 

The other part of that paragraph confuses me as well. Your speed is your speed. And, especially if you are losing speed prior to the climb as I was, how is the subsequent increase in speed due to stored energy rather than thrust?

 

As for the last bit, how you tested the 3 aircraft has no bearing on what I said about your MiG-29 flight at the 1000 m data point. At that point, you still had 1.6 Gs on the airframe and an AoA of 9. So it’s not a clean data point. That’s all I’m saying. In hindsight, I probably shouldn’t have included it.

 

At no point have I been comparing the -29 aircraft to anything but the chart. When it says I should accelerate, I do. When it says I shouldn’t, I don’t. Does it deviate from the chart a little? Probably, though, maybe not.

 

@Cmptohocah,

 

I grabbed a few minutes this morning and made a sustained G turn at 1000 m. In full burner I got 5.3 at 600 km/hr. I used the chart in the pdf that includes the thrust-speed-altitude chart we’ve been using. Clean aircraft and 1500 kg fuel. The sim and chart matched.

 

 

Edited October 5, 2021 by Ironhand

[Cmptohocah]

6 hours ago, Ironhand said:

@Cmptohocah,

 

 

I grabbed a few minutes this morning and made a sustained G turn at 1000 m. In full burner I got 5.3 at 600 km/hr. I used the chart in the pdf that includes the thrust-speed-altitude chart we’ve been using. Clean aircraft and 1500 kg fuel. The sim and chart matched.

 

Great stuff. Thanks for verifying that. It means that both thrust produced by the engines and the drag experienced should be correct - at least for the 1000m altitude, that is. Any chance you could try the same thing out at 5km and 8km alt, if you find the time?

[Ironhand]

4 hours ago, Cmptohocah said:

Any chance you could try the same thing out at 5km and 8km alt, if you find the time?

Sure. Though it may not be for a few days.

[bkthunder]

On 10/4/2021 at 12:20 PM, Cmptohocah said:

I was having in my mind something similar:
Mainly testing the produced trust at sea-level by measuring acceleration from 0 to 100km/h for example. This would give a good measurement of static thrust without too much wind resistance. Only thing that needs to be determined before that is the dynamic friction force produced by the wheels, which can also be measured by rolling the aircraft to a some amount of ground speed and then measuring the time/distance it takes to come to a complete stop with the engines spooled down. Then it would be just a matter of F=ma with a fairly low error which would be produced by drag.
Sadly I can't test this yet my self.

 

So I *think* I tested this correctly. 

 

Mig-29A

total Mass: 13770Kg

Flaps up in all tests to reduce drag.

 

first I tested the ground friction as you suggested. 

 

Initial speed: 35 m/s

final speed after 10 seconds: 31.3 m/s

= -0.38 m/s2 deceleration

13770 x -0.38 = -533.58 kgf

 

 

Acceleration test - measurement started when in full AB and no brakes. 

Initial speed: 39.8 m/s

final speed after 10 seconds: 125.2 m/s

= 8.54 m/s2 acceleration

13770 x 8.54 = 11991.57 kgf + 533.58 = 12525.15 kgf

 

 

So:

 

- 12500 kgf vs 16600 kgf (static thrust)

- 12500 kgf vs 16000 kgf according to the chart below

 

 

 

Conclusion: 

 

The MiG-26A lacks 4000 kgf of thrust. 

Can someone confirm my calculations are correct? 

 

 

Thanks

 

[Cmptohocah]

2 hours ago, bkthunder said:

 

So I *think* I tested this correctly. 

 

Mig-29A

total Mass: 13770Kg

Flaps up in all tests to reduce drag.

 

first I tested the ground friction as you suggested. 

 

Initial speed: 35 m/s

final speed after 10 seconds: 31.3 m/s

= -0.38 m/s2 deceleration

13770 x -0.38 = -533.58 kgf

 

 

Acceleration test - measurement started when in full AB and no brakes. 

Initial speed: 39.8 m/s

final speed after 10 seconds: 125.2 m/s

= 8.54 m/s2 acceleration

13770 x 8.54 = 11991.57 kgf + 533.58 = 12525.15 kgf

 

 

So:

 

- 12500 kgf vs 16600 kgf (static thrust)

- 12500 kgf vs 16000 kgf according to the chart below

 

...

 

Conclusion: 

 

The MiG-26A lacks 4000 kgf of thrust. 

Can someone confirm my calculations are correct? 

 

 

Thanks

 

Your mathematics is spot on.
 

In my humble opinion the following things need to be revised in order to bring the errors to a minimum:
- dynamic wheel friction test: make sure you start measuring the deceleration once the engines are completely shut down and spooled down in order to not have any engine thrust effect the friction force calculation. Perhaps you can wait until the aircraft comes to a full stop in order to catch more of deceleration. Yet again, maybe 10 seconds is enough. I wouldn't know.
- ground acceleration test: perhaps you can measure the acceleration from 0km/h to 100km/h in order to avoid drag having big impact on the measurement. Make sure you start with the wheels on and AB on, so you have as much thrust as possible when accelerating from 0km/h. Due to wheel brakes not being able to hold the airplane at full AB perhaps the measurement should start at some other value than 0? Perhaps 20km/h?

Would be great if you could upload a Tacview of the tests, so I can have a look at it.

Great work BTW.

Edited October 6, 2021 by Cmptohocah

[bkthunder]

I tested pretty much as you describe, waiting to be in full AB and brakes released, and waiting for the engines to be spooled down as much as possible (but it's hard 'cause the runway is not long enough). 

 

Either way, I have sampled speeds at different points and at different intervals, there are of course some small differences, but in general I can confirm the numbers I posted before. 

I attach here the track so you can check for yourself. 

 

P.S. I also tested with F-18 and F-16,  and strangely enough, while they both have a higher thrust than the MiG-29, they also fall way short of their declared static thrust by several thousand lbs!

Tacview-20211006-104628-DCS.zip.acmi Tacview-20211006-112533-DCS.zip.acmi Tacview-20211006-123714-DCS-Climb TW ratio test.zip.acmi

[Cmptohocah]

59 minutes ago, bkthunder said:

I tested pretty much as you describe, waiting to be in full AB and brakes released, and waiting for the engines to be spooled down as much as possible (but it's hard 'cause the runway is not long enough). 

 

Either way, I have sampled speeds at different points and at different intervals, there are of course some small differences, but in general I can confirm the numbers I posted before. 

I attach here the track so you can check for yourself. 

 

P.S. I also tested with F-18 and F-16,  and strangely enough, while they both have a higher thrust than the MiG-29, they also fall way short of their declared static thrust by several thousand lbs!

Tacview-20211006-104628-DCS.zip.acmi 34.22 kB · 0 downloads Tacview-20211006-112533-DCS.zip.acmi 19.5 kB · 0 downloads Tacview-20211006-123714-DCS-Climb TW ratio test.zip.acmi 23.53 kB · 0 downloads

Great stuff! Thank you!
I will have a look latter today or tomorrow.

[Ironhand]

4 hours ago, bkthunder said:

Initial speed: 35 m/s

final speed after 10 seconds: 31.3 m/s

= -0.38 m/s2 deceleration

13770 x -0.38 = -533.58 kgf

 

Good work.

 

That first part is a bit off though because of a math error. 0.37, not 0.38 which works in favor of your argument.

 

Quote

final speed after 10 seconds: 125.2 m/s

 

How did you account for the extra drag after your tires blew, though? 125.3 m/s is about 450 km/hr and the tires shredded somewhere in the high 300s.

 

Quote

P.S. I also tested with F-18 and F-16,  and strangely enough, while they both have a higher thrust than the MiG-29, they also fall way short of their declared static thrust by several thousand lbs!

 

To my mind, that seems to indicate an error in our approach here. Wish I was in a position to use the sim. These investigations are fun.

Edited October 6, 2021 by Ironhand

[GGTharos]

Installed thrust can easily be 20% less than declared static thrust.

[Cmptohocah]

I have checked the 104628 TacView file and here are the findings for MiG-29A:
m  - mass of the aircraft [kg]

F_q - dynamic friction force of the landing gear [N]

F_t - thrust [N]

a - acceleration [m/s*s]

t - time [s]

v - speed [m/s]

 

m = 13,770kg (assumed correct as provided by @bkthunder - can't verify this)

F_q = m * a_q
a_q = (v_q - v_q0) / t_q = - 0.37
F_q = 13,770kg * -0.37 m/s*s = -5094.9N

 

Thrust:
F_t +  F_q = m * a =>  F_t = m * a -  F_q

a = (v - v₀) / ( t - t₀)
t₀ = 157.25s
t = 159.31s
v = 23.33 m/s [84 km/h]

v₀ = 4.33 m/s [15.6 km/h]

a = 9.223 m/s*s

F_t = 13,770 kg * 9.223 m/s*s + 5094.9N = 132,232.71 N = 13,479.38 Kgf
 

Edited October 6, 2021 by Cmptohocah
Added formula for acceleration.

[bkthunder]

Confirm the mass is 13770 Kg, this is straight form the mission editor loadout page (you can take the track file posted in the OP and check).

 

Curious how you got acceleration of 9.223 m/s2, if I understand correctly, you measured measured over 2 seconds as opposed to 10 seconds in my test? 

 

@Ironhand didn't realize the tires were blown!

 

 

Ok, anyway do we agree according to our calculations, with some error and variance, we are far from the 16000 Kgf of thrust the plane should have, according to the chart posted earlier? 

Are we missing something?

 

 

 

[Cmptohocah]

2 hours ago, bkthunder said:

Confirm the mass is 13770 Kg, this is straight form the mission editor loadout page (you can take the track file posted in the OP and check).

I don't have access to DCS at the moment. That's why I asked for TacView
 

2 hours ago, bkthunder said:

 

Curious how you got acceleration of 9.223 m/s2, if I understand correctly, you measured measured over 2 seconds as opposed to 10 seconds in my test? 

That's correct. I took only 2.06s of acceleration since I tried to keep the air resistance at a minimum, ie. ground-speed of less than 100km/h. I recorded the data at 15.6km/h since TacView shows throttles at 100% at that moment.
Acceleration was calculated as:
a = (v - v₀) / ( t - t₀) = (23.33 - 4.33) / (159.31 - 157.25) = 19 / 2.06 = 9.223 m/s*s

 

 

[Cmptohocah]

For anyone wanting to automate the process, here is JavaScript code:

const kphToMs = kph => kph * 0.277778;
const acc = (v0, v, t0, t) => (v - v0) / (t - t0);

// TEST PARAMETERS
const m = 13770;
const g = 9.81;

// Dynamic wheel friction
const tq0 = 0;
const tq = 10;
const vq = 35;
const vq0 = 31.3;

const aq = acc(vq0, vq, tq0, tq);

const Fq = m * -aq;

// Acceleration parameters
const t = 159.31;
const t0 = 157.25;
const v = kphToMs(84);       // 84   km/h
const v0 = kphToMs(15.6);    // 15.6 km/h


const a = acc(v0, v, t0, t);

const F = m * a - Fq;

console.log(`Calculated thrust for aricraft mass of ${m}kg, is ${Math.round(F)}N or around ${Math.round(F/g)}kgf.`);

 

[Cmptohocah]

Here's a chart of longitudinal acceleration vs time/speed:

It seems that thrust-to-weight ratio for mass of 13.7 tons, is somewhere around 1.0-1.1 (on average) up to 100km/h (red line).

Edited October 7, 2021 by Cmptohocah

[Cmptohocah]

21 hours ago, bkthunder said:

P.S. I also tested with F-18 and F-16,  and strangely enough, while they both have a higher thrust than the MiG-29, they also fall way short of their declared static thrust by several thousand lbs!

Tacview-20211006-104628-DCS.zip.acmi 34.22 kB · 4 downloads Tacview-20211006-112533-DCS.zip.acmi 19.5 kB · 3 downloads Tacview-20211006-123714-DCS-Climb TW ratio test.zip.acmi 23.53 kB · 3 downloads

What were the masses of the F-18/F-16?

[Ironhand]

Just to add some data:

 

I come up with 13,541.66 kgf when tested on the runway. Very similar to Cmptohocah's numbers.

 

Acceleration:

Mass: 13588.4 (Removed the amount of fuel used to the point of time start which is why it's not 13770.)

Initial Velocity: 13.4 k/h (100% thrust--full afterburner)

Final velocity: 186.4 k/h

Time Difference: 5.16 sec

Acceleration: 9.313 m/s2

Force: 126,550N or 12804.51 kgf

 

Friction/Drag:

Mass: 13537.9 (Again fuel used subtracted from 13770 kg.)

Initial Velocity: 190.6 k/h (Engines off and spooled down before time start.)

Final velocity: 180.7 k/h

Time Difference: 5.15 sec

Acceleration: -.534 m/s2

Force: -7,229N or -737.15 kg/f

 

Total Thrust: 13,541.66 kgf

 

@Cmptohocah

 

Two sustained turn tests:

8000m, 700 IAS yielded 4.5G

5000m, 800 IAS yielded 6.6 G

 

EDIT: In the Tacview files for these two flights, the a/c has a total mass of 13,000 kg when fuel reaches 2,078 remaining. So the Tacview files only show the time span around that point. /EDIT ENDS

 

Combined with my previous 1000m flight, it looks like the installed thrust is modeled as less than the static thrust but, based on my limited G testing, is in line with real world performance.

 

Tacview files attached (hopefully the right ones).

 

Tacview-MiG-29 Sustained G 5000m-800IAS.zip.acmi Tacview-20211007-094623-DCS-MiG-29 Sustained G 8000m-700IAS-1.zip.acmi MiG-29 Thrust Testing Roll 2A.zip.acmi

Edited October 7, 2021 by Ironhand

[Ironhand]

13 hours ago, Cmptohocah said:

For anyone wanting to automate the process, here is JavaScript code:

const kphToMs = kph => kph * 0.277778;
const acc = (v0, v, t0, t) => (v - v0) / (t - t0);

// TEST PARAMETERS
const m = 13770;
const g = 9.81;

// Dynamic wheel friction
const tq0 = 0;
const tq = 10;
const vq = 35;
const vq0 = 31.3;

const aq = acc(vq0, vq, tq0, tq);

const Fq = m * -aq;

// Acceleration parameters
const t = 159.31;
const t0 = 157.25;
const v = kphToMs(84);       // 84   km/h
const v0 = kphToMs(15.6);    // 15.6 km/h


const a = acc(v0, v, t0, t);

const F = m * a - Fq;

console.log(`Calculated thrust for aricraft mass of ${m}kg, is ${Math.round(F)}N or around ${Math.round(F/g)}kgf.`);

 

Ummm…my ignorance is showing. How do I set this code so that I can use it? I tried copy/ pasting it between <script></script> tags in an html document but that didn’t work. Never tried to set something like this up before.

[Cmptohocah]

31 minutes ago, Ironhand said:

Ummm…my ignorance is showing. How do I set this code so that I can use it? I tried copy/ pasting it between <script></script> tags in an html document but that didn’t work. Never tried to set something like this up before.

It was my bad for not explaining better.

This should still work between the script tags, check out your browser's console. Depending on what browser you use you should be able to right-click and click on "inspect". Also you could alternatively use an online service like "jsbin" and click "run" button - output should be in the console.

[Cmptohocah]

6 hours ago, Ironhand said:

Combined with my previous 1000m flight, it looks like the installed thrust is modeled as less than the static thrust but, based on my limited G testing, is in line with real world performance.

It could also be that drag was adjusted to match the produced thrust and hence have same performance as the chart - at the moment I have no proof of this.

The only thing we do know is that, in DCS, MiG-29A produces somewhere around 13.5t of static thrust at max power (full AB). Now it's a question of finding out what kind of static thrust the real engines provide when on a Fulcrum A.

From that chart that ED shared, it seems that they should produce a combined thrust of 16t, but I am not 100% sure as I would need more information related to this particular chart.

Edited October 7, 2021 by Cmptohocah