Trim Problem. Exponential control response.

The Russian stunter I bought last year at the World Champs seems to suffer from a rather exponential control response. The control response gets quicker with greater control deflection making it somewhat more difficult that desired to get consistent round maneuver sizes. The squares are great with a clean sharp corner and straight lines and flat bottoms. At the trim session at the Golden State Stunt Champs I heard Ted Fancher remark WOW! when I did my inside squares. I think the exponentiality might actually be helping in the squares.

Any suggestions on how to trim out this exponentiality and hopefully not hurt the squares too much would be greatly appreciated. I've had a lot of help from many top flyers out west in getting other aspects of the models trim cleaned up but this problem still exists. In most other respects the model is flying very well now.

Hopefully a trim cure can be found that won't require a design change like building new components. For the future I'd like comments on what design features one would use to make control response more linear also.

I'm sure some of you trim experts out there can offer up some ideas I can try when the weather warms up. Thanks in advance

Jim Rhoades

Jim
Assuming that the bellcrank and horns are of standard configuration, and you don't have any fancy exponential mechanizm in the plane, I'd have to think that you are feeling some aerodynamic effect.

The Yatsenko control surfaces are of a "cove and glove" construction that may not be easy to seal with tape. Might slip a strip of felt or something fuzzy like that of the proper thickness into the flap gap and elevator gap. Just enough to block the air flow but not enough to stiffen up the movement. Make it a temporary installation, Just in case.

Frank and Randy,

Thanks for the response. The model is not a Yatchenko although of the same basic design school. They look a lot alike. It was built by Eugevny Yakolev with great workmanship. The model is conventional built up wood construction and has conventional nylon hinges, not like the Yatchenko setup. The hinge gap was pretty wide, about 3/32", to allow for pulling the pin for flap removal. The characteristic was present when I got the model and the first change I made was to seal the hinges along with much less tip weight. I had to narrow the line spacing a bunch as the turn was much quicker. The C/G was pretty far forward to start with. The C/G has been moved back and the handle spacing narrowed but the problem still existed. The flap to elevator ratio has been increased a couple steps and brought back one of these steps. The exponential feel is still there.

Maybe the answer is just a narrower handle spacing. I'll try that.

The controls are conventional with the only exception being the belcrank which has the arms swept inboard slightly. Brett Buck says it's just about the same as he uses. Just enough to give a 90 deg leadout angle at neutral.

The numbers, (moments, areas etc.) are very similar to my Norm Whittle designed Sultan with the exception of the higher aspect ratio tail surfaces with much less stabilizer area. The stab is slightly narrower than the elevator. Everything is on the thrust line of the engine but I don't think that has any bearing on this problem.

Any more Ideas?

Jim Rhoades

Jim, do yo know what the size of the tail is, in terms of % of the wing? A small tail(<18% or so of the wing) can give the effect you describe. In a tight turn the wash from the wing reduces the effectiveness of the tail and makes the plane less stable. It can overturn in harder maneuvers.

Phil,

The tail surface is about 23% of the wing on this model so that is definitely larger than your quoted 18%. Slightly more of this area is in the elevator than the stab as I noted before. Most modern US stunters nowadays have stabs bigger than the elevator and lower aspect ratio tails. It has a nice thick wing and airfoiled tail surfaces about 5/8" thick.

The problem is in the round maneuvers not the squares or hard corners. The squares are great with tight clean corners. The model is really flying quite well and is easy to fly well in most all other respects. If I could trim out the exponentiality it would be a really great flying model.

I was hoping to get out and fly it today but It's been raining a bit overnight and things are qretty soggy out. If the sun came out now and it warmed up a few degrees I still might make it but that is not too likely to happen as it's noon here already and the clouds are looking pretty solid. Maybe another day, soon I hope.

Jim Rhoades

edited to explain problem.

Hay, Jim...Since this thing is an inline...how are the controls setup? Are the pushrods 90 degrees to the control horns? Are the control horns 90 degree types, or are they 'corrected geometry' types, i.e., inclined? What are the TE's of the flaps and elevators like? And finally, what is the cross section at both the hingelines? One thing I'd try is taping a piece of 1/8" square to both top and bottom of the elevator TE, to thicken the TE, assuming it's a thin TE now. Steve

Steve Helmick
Renton, Washington
USA

Steve,

The controls are set up with corrected geometry as you say. The flap T E is 1/8" square and the elevator T E is just over 1/16" square. The cross section at the hinges is square wing and stab T E and round L E on the flap and elevator with 3/32" gap now sealed.

It might be interesting to try thickening up the T E as you suggested.

Jim Rhoades

23% should be no problem.

Sounds like the control loads aren't high enough. Under partial control- i.e. the round maneuvers- you can't feel what it is doing.

This can be done two ways, but they aren't two differnt ways of doing the same thing. they are differnt.
- move the pushrod in slightly at the bellcrank(shorter 3rd arm)
- add some nose weight and see what happens after everything else is retrimmed.

Jim,
Frank is absolutely right. With a conventional bellcrank/control horn arrangement, control response is greastest about neutral and decreases as you get farther from neutral.

Ergo the problem must be from some other cause. Maybe the controls are simply too "quick" and this gives the impression of "exponential" response. Sealing the hingelines, as Frank suggested, (if possible) might make a big difference. Adding small increments of nose weight might also help, and also of course a smaller handle spacing. "Stretchy" lines might also give this "feel". (Leadouts or handle leads). Initial stretch would tend to make the controls feel less responsive until the stretch was accommodated by load.

Randy,

I've tried narrowing the handle spacing but only after making other changes that had increased the sensitivity such as sealing the hinge lines and moving the C/G back. Maybe I need to try narrowing it farther but the model was comfortable to fly so I left it where it was. The model was a bit nose heavy to start with, the C/G was at about 18% as I recall.

I use a Kaz Minato handle which has cables but is set up so there is little if any give or stretch. I haven't had the problem with other models using this handle. I like it a lot. It is light and fully adjustable. I Have a kit for a Fancher hard point handle I might put together and try but I really don't feel this is where the problem lies. I need to put it together anyway so this could be a good excuse to do it.

I'll probably bring the model with me to VSC and fly it after the contest.

Jim Rhoades

I think you got a lot of good advice, and it's about what I would try. Certainly you can drastically change it on other models by running the CG aft and narrowing the spacing to compensate, along with sealing the hinge lines.

I would also certainly look into reducing the elevator travel relative to the flap, particularly with the hinge lines sealed. That's one of the effects of the smaller elevator fractions (35% on my airplane) and the low aspect ratio that we usually run around here. Go easy because a little change will make a big difference in how the airplane flies.

If none of that works, building a new stab along the lines of an Impact dimensions is something to consider. That's the nice thing about a take-apart, you can take it apart.


I also caution you that the airplane flew pretty well when I flew it and maybe "fixing", it along the lines I suggested, will turn out to be worse than it was. The guys who built it knew what they were doing. I think you are doing roughly what I would, but at some point you may find you are better off just going back to where it was.


>I use a Kaz Minato handle which has cables but is set up so
>there is little if any give or stretch. I haven't had the
>problem with other models using this handle. I like it a lot.
> It is light and fully adjustable. I Have a kit for a Fancher
>hard point handle I might put together and try but I really
>don't feel this is where the problem lies.

Might be surprised at that. I certainly was - I thought the whole thing was downright silly - "how can a 1/16" cable possibly 'give' when I have 65 feet of .015 on there?" - but I was completely wrong. I didn't even *try* the prototype handle Ted gave me for a year or so, until Ted bugged me about it. I was using a Baron-style handle but I would expect the same effect would happen with the Kaz handle. By they way, I don't think the source of the problem is the cable stretching lengthwise - it's the bending of the cable after it comes out of the handle.

Brett

Brett,

I have tried reducing the elevator travel in relation to the flaps in two steps and felt that the second move, near 1-1 was too much and I didn't like the feel so I backed off to only one step which is less elevator travel than when you flew it. I think that feels best but still has some exponentiality feeling.

So far I haven't fixed anything that I cant change back easily. I definitely think it is flying better than when you flew it. If nothing else getting rid of an extra full 1 oz of tip weight certainly cleaned up the hinging problem.

I'll try a hard point handle. It certainly can't hurt and is a fairly easy fix. I have to build the handle kit though.

Building new tail surfaces is a possibility but one I'd rather avoid as the carbon fiber re-enforced spar has to be very close fitting to the fitting in the fuselage. I suspect that the fuselage fitting was laid up around the spar rather than fitting the spar to it. Matching the colors would be a problem too.

Thanks all for the suggestions. It is still a work in progress and has been a fun experience.

Jim Rhoades

Jim,
Fattening the TE was a thought I had too.
Brett's advise, all of it, it certainly good, and the use of the hard point handle is excellent.
I couldn't believe the difference in my big low aspect pipe ship when I first used this handle.
Chris...

No help with your problem Jim but here's a happy snap I have of you at the UK Nats last year. Nice to have that brief meeting with you on that cool day.

HH

[photo not recovered: 22196.jpg]

Attachment #1, (jpg file)

Jim,
I'll look forward to seeing you and the model at VSC this year.
Take care of yourself!

Jim,

Your comment about the aspect ratio of the tail caught my attention. Did you work out the actual number?

As I'm sure you're aware, aspect ratio has a great effect on the amount of lift a surface of a given size produces with a given amount of angle of attack change (or, in the case of a stab/elevator) the amount of deflection input by the pilot.

Higher aspect ratios produce lift much more rapidly in terms of A of A than do low aspect ratios. That's why sailplanes have long, high aspect ratio wings and very finely controlled amounts of elevator input to avoid overstressing the wing by too rapid an increase in A of A.

Same thing is true of a high aspect ratio tail. A given amount of deflection will result in a more rapid increase in lift (and the accompanying more rapid pitch change it produces in the airplane to which it is attached). The lift is also produced with less drag (another important reason for high AR on sailplanes).

These facts argue in support of Brett's suggestion about more flap per unit of elevator movement. In this respect this would be a recognition that you don't need as much elevator deflection to get the appropriate amount of pitch change for the lift produced by the wing.

The very quick corners your ship flies are probably at least partially the result of this phenomenom. In rounds, however, it is possible that the refined control inputs required to lock the rate of pitch change in for nice round, concentric maneuvers just aren't possible. Small elevator inputs could be resulting in too great a change in lift produced thus causing the rate of pitch change to wander.

One test that might be of interest would be to double over a piece of very light cardboard and tape it to the leading edge of the stab over its complete span so that the leading edge is a half inch or so further forward than at present. This'll do a couple of things, enlarge the area of the tail itself as well as lower the aspect ratio.

Doint this will result in a tail which needs to have the elevators deflected further for the desired pitch change and will produce more drag when doing so. This will simultaneously slow the rate of change for a given handle deflection as well as make the airplane feel more stable during the pitch change (the increased induced drag from the less efficient tail will move the neutral point aft during pitching manevuers ... which makes the airplane feel a bit more nose heavy.

By the way, the current trend to lower aspect ratio tails belies the success some had in the past with high aspect ratios. The best example from a competitive standpoint was Les McDonald and the Stilettoes. I don't know what the AR was but it was noticably higher than most of his contemporaries. Noone can say they didn't fly just dandy.

Also, Wild Bill's (and Bob Baron's) long winged airplanes with very high aspect ratio tails were very competitive. In particular, the Humbug had an AR of somewhere around 10 as I recall. Also, as I recall, Bob flew these ships with a very slow control set utilizing a "lot" of handle deflection for corners which is good evidence of the phenomenom we're talking about -- i.e. very rapid lift change with small deflections.

At any rate, Jim, it might be fun to try that trick with the cardboard to see how the control response changes. All it takes is some sticky tape. No knives or saws required.

Ted