Pilotable ferry acceleration problem

[L_Rob]

I hope that this question is not completely inappropriate here. I’m converting one of my AI ferries to make it pilotable. I’ve managed to get most of the dynamics fairly realistic (roll on turning, pitching frequency, etc) but I can’t get the throttle response right. Up to 17-18 knots its fine, but above that speed the bow comes up a few degrees and the boat accelerates rapidly --- as if it’s trying to fly, or is approaching the step (like a float plane).
I started with a aircraft.cfg and .air file from a similar model, but it too had a similar handling problem.

I’ve spent a lot of time temporarily changing everything I could think of that might have an effect in the .cfg file (better, but still not right) and I’ve been editing the .air file with AirEditor after finding an extensive document on the subject recently posted (I forget where) explaining the various features. I’m currently using an .air file based on a Piper Cub from FSX to keep things simple. Among other things I’ve tried increasing the Cd of the landing gear in the .air file, changing the size, position, lift coef, etc. of the wing, changing the induced and parasitic Cds, changing the float positions, cg location, on and on, in the .cfg file, no luck so far.

Any suggestions *greatly* appreciated!

 

[Roy Holmes]

This page and the calculator it contains may give you some clues.
http://illustrations.marin.ntnu.no/hydrodynamics/resistance/planing/index.html

I tried the same thing a few years ago with a speed boat. It would plane as explained on the referenced page and the nose would, at some critical speed, pitch up and the boat would back flip. I think I got it up to about 60 kts then lost interest. I do remember that the two aft float type contact points need to be well back for any kind of pitch stability. Put the wing area to zero.

Roy

 

[L_Rob]

Thanks so much for this! I'll see if I can use the reference to make sure the ferry does not plane!

After playing around more with the .air file, I did discover that I can dramatically change the throttle response by fiddling with table 511- "Propeller efficiency vs. advance ratio and blade pitch". I may be able to fine tune quite a bit more, but at least the boat will now hold a relatively constant and appropriate speed. The 511 tables for various aircraft I checked show a large variation, so I think I'll have to review the theory a bit. However I still have a hunch that there is another fundamental issue.

 

[L_Rob]

I thought I might briefly report on progress, or maybe lack of it. Above about 17 to 18 knots the speed still oscillates over a period of a minute or two, but the change in speed is down from maybe 12 knots differential to 5 or 6. If you are not watching the speed on the GPS you would not notice it. A couple of other models I checked had similar problems. I have still not figured out what’s the fundamental cause of the behavior, so I think I’m just treating the symptoms. Having tried all sorts of things, the following seemed to help: I significantly edited table 473 in the .air file (pitch moment coef vs. alpha) so that there is no pitch up or down moment when the boat is on its lines, but that the pitch (down) moment coef increases very rapidly and to a high value by the time the boat is bow up by 5 degrees, pushing the bow down. (The max pitch up seems to be about 3 degrees for this model). The total pitch moment is linked to the lift coef, so I played with that too.

In addition, adding a thrustanglespitchheading line to the general engine data in the .cfg file and setting it so in effect the propshaft is angled ~25 degrees down helped a modest amount. The effect is linked to the engine location, so I experimented with that. Note that there is a typo in the SDK; there is a missing ‘s’. And for some reason, using a constant speed prop did help slightly. But I’m still hoping that someone will be able to fill me in on what causes this behavior in the first place.

Larry

 

[robystar]

Hi,

The speed oscillation is most likely due to some contact points that have an xyz placement that raises them above the water when the vessel pitches up and down.
Have you tried simply lowering the contact point(s) for the wheel(s)?

 

[L_Rob]

Hi,

The speed oscillation is most likely due to some contact points that have an xyz placement that raises them above the water when the vessel pitches up and down.
Have you tried simply lowering the contact point(s) for the wheel(s)?

Thanks for this! That just might be it. It fits with what I observe. I spent a lot of time adusting them up and down / every which way without getting the desired result, but that was before I changed the underwater surfaces. I'll try again with adjusting contact points and see what happens. If this works I'm gonna feel awfully dumb.

When I lowered the contact points before it simply pushed the boat further out of the water... as I remember.

Larry

[edited to add]

After some experimenting, I'm getting close. I've confirmed my early memory. In P3D at least, (presumabely FSX too) raising or lowering the contact points raises or lowers the visual model, but did not change the speed fluctuation.

However by trial and error I found that adjusting the "Static Compression”, and the Max/Static Compression Ratio and getting it just right the boat will now hold constant speed through the part of the speed range I tested. I'm going to quit while ahead for the moment, but I think this may be the answer. Right now I have different values of both variables for the two front contact points vs. the rear, and its sensitive. I'm leaving the Damping Ratio/Water Rudder Effectiveness unchanged for the moment; it's set way overdamped. I also found that I could effectively adjust the amount the bow plunges by changing the Max/Static Compression Ratio on the front floats. I had played with MOI values, but this is seems to address the issue directly. I had tried all of this earlier, but did not hit on what looks like the right combination.

More testing to follow. But thanks *so much* Robystar for putting me onto what may be the answer!

Larry

 

[robystar]

Hi,

Lowering your wheel(s) should not lower the position of the vessel. The float points would do that.

 

[L_Rob]

Hi,

Lowering your wheel(s) should not lower the position of the vessel. The float points would do that.

Hi robystar,

I’m not quite sure I follow you.

I was referring (or thought I was referring) to contact points coded as floats (4) as in:

Class: Integer defining the type of contact point: 0 = None, 1 = Wheel, 2 = Scrape, 3 = Skid, 4 = Float, 5 = Water Rudder Per the SDK.

No wheels on the boat.

What am I missing?

Thanks for the help,
Larry

 

[robystar]

Hi,

" What am I missing? "
The wheels (class 1)! At least one of them for steering and propulsion, to be placed slightly lower than the floats (That is how I would do it for a pilotable one).

 

[L_Rob]

Hi,

" What am I missing? "
The wheels (class 1)! At least one of them for steering and propulsion, to be placed slightly lower than the floats (That is how I would do it for a pilotable one).

Thanks robystar,

I have been using one centerline 'class 5' contact point, and lowering it below the four 'class 4' floats. Several of the freeware boats I checked did it this way. However, I find that a payware boat has done it the way you suggest.

Interestingly enough, the wheels were given a 0 degree steer angle while the front two floats had this set at ~20 degrees. (Steerable floats then?)

This gives me something to experiment with! Can't wait to try it out.


Thanks again! I owe you!

Larry

 

[robystar]

Hi,

I guess they just did not bother to adapt the steering angle of the floats because it is of no importance?

 

[L_Rob]

Hi,

I guess they just did not bother to adapt the steering angle of the floats because it is of no importance?

At least in P3D, you can steer with the floats only, and a couple of the freeware boats I looked at did it that way. I am also able to steer with a water rudder (class 5). It does not seem to make a difference as long as you set the appropriate variables.

I did discover that if I code a new set of contact points as wheels along with the floats, without which the boat sinks, it seems to have no effect at all. I eventually made them ridiculously long, gigantic wheels, etc and it had no effect. Who knows.

However I finally have found what may be a solution to the problem of rapid acceleration beyond a certain speed.

I started looking at Table 512 in the .air file: where the input is the normalized advance ratio of the prop and the output is the prop power-in coef. for different betas. Unless I screwed up my calculation, the x axis scale is ~10 times bigger than it needs to be for a boat. I've made quite a bit of progress playing with these curves. I found a nice little article on prop theory on the MIT website which helped a lot. Normally the power in coef goes down with increasing advance ratio, but that's sort of the opposite of what I want in this case. At least this gives me something to play with that really has an effect.

Larry

 

[robystar]

Hi,

Good luck because that goes over my head!
Do remember though that FSX emphasizes more on the cfg and not so much anymore on the .air file (some things that formerly were governed by the .air file are no longer but are now adaptable in the aircraft.cfg).
I am glad that I managed at least to get aircraft take off as if launched and land as if they were trapped by the arresting cables on a carrier. And that involved mostly adapting the .cfg and finetuning through the .air file.
Keep us informed though.

Roby

PS: I have not the faintest if it is different in P3D compared to FSX.