Quantcast

Raptor Composite Aircraft

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Status
Not open for further replies.

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,776
Location
Memphis, TN
It can only be something different if participants are different. Im not saying it can’t be fixed, Im saying it’s not going to be fixed. As it is now, it’s heading for a box canyon. Anything different requires a chasm of the Pacific Ocean to be spanned. What is done after the finish line is crossed is up for grabs. I can’t see starting over with a new project after knowing what this one took. I imagine that there was hope of buyout, built in like a tech company being bought by google. Maybe Icon might want to expand its product line. Not a lot of power takeovers in GA much less homebuilts.
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,958
Location
Port Townsend WA
Generally successful projects start with a bare bones proof of concept and add luxuries to the production version, rather than start with everything you could ever want and later subtract from the production version.
I agree with that and suggested moldless proof of concept early in this thread.
He has a different way of doing things. Either way it would take more than one iteration.
 

berridos

Well-Known Member
Joined
Oct 10, 2009
Messages
1,019
Location
madrid
Some suspect that the plane showed dutch roll problems. If thats the case, the end of this saga would be an anticlimax, because for such an aero problem there is no solution at this stage. Two more hops confirming this and to the junkyard. Hope carbon at weight can be recyled somehow.
 

rbarnes

Well-Known Member
Joined
Aug 28, 2015
Messages
312
Location
Texas
Some suspect that the plane showed dutch roll problems. If thats the case, the end of this saga would be an anticlimax, because for such an aero problem there is no solution at this stage. Two more hops confirming this and to the junkyard. Hope carbon at weight can be recyled somehow.
About 6 vortilons along the leading edge seemed to do the trick on the scale model. Just can't figure out why PM only put one on the real plane. It's not an uncommon "band aid" to see on these style planes to help with the headaches of needing a swept wing.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,355
Location
Rocky Mountains
Could Give my opinion of why I think that is,
For some of us it is the squandered resources.

There are probably a dozen orphaned kit planes that could have been brought back to the market - in improved versions - and the first 50 kits given away and still have cash in the bank. There would be no new tech or advancement of the art but there might be a few more planes flying. Example:

How many new fast build Starlite kits could be produced and sold with current factory support for what has been spent on this project? Sure they aren't even in the same class, but there is only so much investment money available and we in the EAB market already at the end of that backwards flowing stream.
 

Turd Ferguson

Well-Known Member
Joined
Mar 13, 2008
Messages
5,330
Location
Upper midwest in a house
Some suspect that the plane showed dutch roll problems. If thats the case, the end of this saga would be an anticlimax, because for such an aero problem there is no solution at this stage. Two more hops confirming this and to the junkyard. Hope carbon at weight can be recyled somehow.
I find it interesting that 1) someone could identify dutch roll in a 4 sec ballooning flight, and; 2) that dutch roll is some kind of bad omen. Lot's of airplanes exhibit dutch roll to varying degrees (V-tail Bonanza, for one) as it's more desirable than some other forms of roll / yaw dynamic coupling. Certainly not something I would scrap the plane over.
 

jet guy

Well-Known Member
Joined
Aug 15, 2020
Messages
73
So much speculation and guessing and linearizing accelerations...

Now as to what the power and BSFC really is for the Raptor, well, you are all guessing...Once we admit to ourselves that they are guesses, when will we stop talking about it?
Well, there is guessing, and then there is educated guessing, sometimes called 'analysis'.

Many of the folks here seem interested in tinkering with ideas about aircraft design, and that also includes Raptor Peter. But let's do a thought experiment: you want to create an airplane that is as good as some known quantity that works well. So where do you start? Do you just wave your hands around and say 'that looks about right'?

Okay, so obviously it pays to do a little preliminary analysis, which is why no airplane is ever designed without undertaking such an exercise.

In this particular case we are interested in the power required by the engine in order to give us the required takeoff performance. That is a pretty basic starting point, and we would set as our objective an existing airplane like the SR22, which only takes about 1,000 ft to get airborne. Obviously we don't want an airplane that takes three times as much, so if our testing results show that, then something is very wrong.

Now if we look a little deeper we find that trying to figure out why a Cirrus performs the way it does is due entirely to some key information about its physical dimensions and aerodynamics. We know its dimensions and its performance, but Cirrus isn't going to give you the drag polar for the airplane, which you will need in order to take a close look at just HOW it achieves its performance.

But the good news is that all of those aerodynamic [and propulsion] characteristics are already baked into the published airplane performance, which we can access just by looking at the POH. It only takes a little bit of work with known physics analysis methods to extract the exact data.

Let's run an example. Our first order of business is always to compile thrust required and power required curves. But in order to do that we need to know a couple of key things that are known only to Cirrus, namely the zero lift drag coefficient, and the Oswald efficiency factor e.

How do we extract those?

Well, it turns out that there is indeed a well traveled road to do that. The zero lift drag coefficient CDo exactly equals the induced drag coefficient CDi at the airplane's best glide speed. This is true for each and every winged airplane.

Looking at the POH we see a best glide speed of 88 KIAS [at 3,400 lbm]. There is also a glide ratio given of about 9.6, but that includes propeller drag, which means the actual aerodynamic performance of the airplane will be a lot better than that. [We will find that our L/D ends up at 16 without the prop].

The importance of the CDo is that it always remains the same. Once you know this number, you can easily find the total drag coefficient CD at any speed, by simply then computing the lift induced drag, which is easy to do [and adding the two together].

A very useful relation for finding the CDo:

CDo = W^2 / q^2 / S^2 / pi / e /AR

This holds true only for the speed at minimum thrust required, which is the maximum L/D [glide ratio] speed. For brevity, I won't go into the derivation, but obviously it involves the lift equation and the Newtonian equations of motion for an aircraft in flight. All of that is pretty simple and a good derivation is provided in Anderson's Introduction to Flight.

We start the analysis by first finding our lift coefficient at this best glide speed using the lift equation, L = q * S * CL, and turning it around to solve for CL we get:

CL = 3,400 lbm / 145 ft^2 / 26.2 lbf/ft^2 = 0.895

[q is dynamic pressure.] We now find our CDo using the long equation above, plugging in a trial value for the efficiency number e, which we will need to iterate. Starting with 0.8 for that number we get:

CDo = 3400^2 / 26.2^2 / 145^2 / pi / 0.8 / 10.1 = 0.032

We know that our induced drag CDi must equal CDo at best glide speed, so we now find our lift induced drag coefficient:

CDi = CL^2 / pi / e /AR = 0.895^2 / 3.14 / 0.8 / 10.1 = 0.032

So it works! We have found our zero lift drag coefficient, and it exactly matches our lift induced drag coefficient at best glide speed. Each one is 0.032, so the total drag coefficient CD = 0.064.

BUT, we need to now double check that result at a different flight speed so that we can verify that our CDo does indeed remain the same. If we now go again to the POH we find that the airplane will cruise at 184 KTAS at 6,000 ft on 85 percent engine power continuous.

We find our drag coefficient by first finding our total drag, which will equal thrust in straight and level flight.

Drag = Thrust = power / v = 310 HP * 0.85 * 0.85 * 550 ft*lbf/s / 311 ft/s = 397 lbf

Where 550 is the number of ft*lbf/s in 1 hp, and 0.85 is both our percent engine power and our prop efficiency, and 311 ft/s is our speed of 184 KTAS.

Now that we know our total drag, we can find our overall drag coefficient:

CD = D / q / S = 397 lbf / 96 lbf/ft^2 / 145 ft^2 = 0.029

Well, we have a problem! Our total CD is now lower than just our CDo that we calculated, without even taking into account the drag due to lift, CDi.

So here is where we iterate the two unknown factors, the prop efficiency, and the Oswald number. If you use a spreadsheet you can simply iterate the two formulas until you get the right number. The Oswald number will decrease as the prop efficiency increases, and vice versa.

It turns out for this example that a prop efficiency of 0.9 will give an Oswald number of ~0.9 [0.89565]. Both of those are reasonable figures in line with known quantities: the C182 has an e of 0.84, the 35 Bonanza 0.82. The Oswald number is related to span efficiency from lifting line theory, so the higher aspect ratio will give a higher e, as we would expect in the SR22 with its aspect ratio of 10.1.

We note that the lift coefficient doesn't change, as we iterate the e and prop values.

CL = W / q / S = 3400 lbm / 96 lbf/ft^2 / 145 ft^2 = 0.245

We now find our lift induced drag coefficient:

CDi = CL^2 / pi / e / AR = 0.245^2 / pi / 10.1 = 0.002

Summing our CDo that we found at our best glide speed and which doesn't change, plus our CDi, we get:

CD = CDo + CDi = 0.028 + 0.002 = 0.03

Note: as we were iterating our Oswald number and our prop eff, we ended up with a new zero lift drag number, smaller than with our original assumption:

CDo = 3400^2 / 26.2^2 / 145^2 / pi / 0.89565 / 10.1 = 0.028

Note that we are using here our dynamic pressure q [26.2] for our best glide speed, 88 knots, since CDo is always computed at this speed only.

We are all set now, and have what amounts to the drag polar and can go ahead now and calculate our thrust required and power required, at every speed.

Glide ratio at 184 KTAS = L/D = CL / CD = 0.245 / 0.03 = 8.1

To find our maximum glide ratio, we go back and add our new, correct CDo [0.028] to our CDi at that best glide speed and get 0.056 [remember, CDi is the same as CDo at best glide speed].

We can also verify this:

CDi = CL^2 / pi / e / AR = 0.895^2 / pi / 0.89565 / 10.1 = 0.028

Exactly the same as our CDo, as it must be.

Our total CD = CDo + CDi = 0.028 + 0.028 = 0.056

Our max glide ratio is then:

L/Dmax = 0.895 / 0.056 = 16

Thrust required at any speed is: Tr = W / (L/D)

We can now go ahead and calculate the entire drag and thrust required curves at every speed, using nothing more than our CDo, which stays the same, and simply computing CL at each speed in order to find the CDi.

All you really need, aside from this analysis method is to accurately calculate atmospheric conditions, a good calculator here.

So are we merely 'guessing'? Well, yes our actual numbers for the two unknowns [Oswald number and prop eff] may not be exactly what the actual numbers are. That is only going to be established in flight test, just as all of the massive work that went into analyzing and predicting the aerodynamics and flight dynamics and structures in the design stage, will only be verified when the airplane is tested.

BUT, and this is a big one, the fact that our two values [e and prop eff] may not be exact is completely irrelevant. They work PERFECTLY in letting us PREDICT exactly the performance as it is in the POH [or as observed in flight test].

It doesn't really matter that these two variables may each be [very slightly] off, because one depends upon the other. You cannot get a bad result because both the two real numbers and our two extracted numbers give the exact same result. Both are baked in. A little more of one 'ingredient' and a little less of the other still gives the same result, because there are only two of them, and they vary inversely. Too much of one means less of the other, but the performance comes out exactly the same.

So we see that we can indeed perform very useful performance analysis by simply 'guessing' [iterating actually].

Additionally, one can carry this much farther by using more sophisticated methods to estimate propulsive efficiency, as well as the Oswald number. This is done all the time, in both industry and academia, and there are some really clever methods out there. It all depends on just how much accuracy you are after.

So back to the Raptor and engine power. In my earlier analyses of the takeoff performance, the exact same key numbers are baked in. If you go back to the equation used to find takeoff distance:

Slo = 1.44 * W^2 / g / rho / S / CLmax / {T - [D + Rr(W - L)]av}

We see that the entire THRUST term in parentheses on the right can now be solved when we possess the zero lift drag coefficient, plus the Oswald number. [Rr is the friction coefficient for rolling resistance.]

We can now accurately compute the SR22 takeoff distance using the above results, and we will get a number very close to the 1,028 feet in the POH.

A difference of even 5 percent is quite acceptable. Say 950 ft instead of 1,000. Or 285 hp instead of 300. That kind of difference is a lot better than nothing. And, as mentioned already, you can get very accurate results, even down to several significant digits, depending on how fancy you want to get with various intricate methods.

The simple reality is that there is nothing new and very little guesswork involved in performing very accurate aircraft performance analyses. This is always done when designing a clean sheet airplane and setting performance targets, as well as extracting those unknown numbers for a competitor aircraft.
 
Last edited:

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,958
Location
Port Townsend WA
With successful programs the second iteration builds on sound designs.
It's essentially almost identical to a Velocity except the wider width, the diesel engine and pressurization. So what would be the point of building a regular Velocity first? He hired a Velocity expert who presumably built the wings almost exactly as a Velocity anyway. He could have saved time and money by starting with a used Velocity, like Vans did on the RV-1. I doubt the sensible gradual evolving approach would have attracted deposits and the project would have ended early.
The most major apparent mistake was the complete disregard for logical weight estimation. Since he estimated the same 1800 empty weight as a Velocity even with the pressurization, so clearly no sensible weight goal was posted on the website at the beginning. The engine is the second big variable and we don't yet know how badly estimated the engine weight and performance is.
It could be they knew the weight was vastly more but simply neglected to change the web numbers. Nobody knows.
 

Scheny

Well-Known Member
Joined
Feb 26, 2019
Messages
232
Location
Vienna, Austria
Always in miles an hour because 170 mph sounds way cooler than 140 kts.
So it is not because americans can't distinguish miles and miles ;)
Thanks. Not sure what made me think it was electric, but the issue is the same with it engine driven. If he didn't program it to disconnect the belt pulley clutch at low RPM that will stall the engine when AC kicks on, it wouldn't surprise me at all if he didn't do it for WOT either.
I just manually flip the AC off on my car going up a mountain pass. I bet Peter figures a test pilot could do that on a prototype.
I drove one once, it accelerated so fast my face hit the windshield.
Diesel engines have twice the torque. Since I got a twin turbo Diesel (like the Audi engine, but 2.2l instead of 3l), I can go up mountain highways in the 6th gear. Aircondition can not be felt at all.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,214
Location
Saline Michigan
Some suspect that the plane showed dutch roll problems. If thats the case, the end of this saga would be an anticlimax, because for such an aero problem there is no solution at this stage. Two more hops confirming this and to the junkyard. Hope carbon at weight can be recyled somehow.
Dutch Roll is kind of a long period phenomenon to be sure about on a few seconds of ground effect flight...

Vortillons and winglet extensions below the wing have fixed Dutch Roll and increased stall resistance of the main wing on Long Ez and derivatives . I suspect that Dutch Roll can be "fixed" easily enough, as it has in all of the other ships in this basic configuration.

Billski
 

berridos

Well-Known Member
Joined
Oct 10, 2009
Messages
1,019
Location
madrid
How many fixes are tolerable for a supposedly top product? They all detract from performance. At the end it will resemble a Formula 1 car.
 
Status
Not open for further replies.
Top