Replacing Sullivan Snakes

[Dale Bradly]

I have a problem. 

A favourite model i though was having unusual trim changes. Conclusive proof today, when it was about 1°, and a heap of both rudder and elevator trim was needed. last time i flew the plane it was about 25°, and probably warmer inside the fuse. The blue outer, yellow inner snakes on the elevator are changing length significantly, as in about 25mm of travel from hot to cold. This is clearly way too much, we aren't talking a click or 2 to correct for. 

Does anyone have any great suggestions on how to replace, or moreso what to replace the inner with? I'm thinking some alternative product that can be slid up inside the outer in place of the existing yellow inner?

 

[Don Fry]

I’m not a fan of snakes, but not because they don’t work OK.

That is a lot of change, and far more than I would expect. Is the outer cable properly supported down its length, with attachment supports every 20-25 cm?

[Andy Stephenson]

The trim change with temperature associated with plastic snakes prompted me to replace them with piano wire in a patternship where trim is all important. Of course the ideal would be to make any full length pushrod from the same material as the fuselage so balsa ones would be ideal. If there any sharp bends in the outer then piano wire would not be a solution. The only other option is to use Bowden cable if there is not too much unsupported length at either end.

[Shaun Walsh]

1 hour ago, Andy Stephenson said:

The trim change with temperature associated with plastic snakes prompted me to replace them with piano wire in a patternship where trim is all important. Of course the ideal would be to make any full length pushrod from the same material as the fuselage so balsa ones would be ideal. If there any sharp bends in the outer then piano wire would not be a solution. The only other option is to use Bowden cable if there is not too much unsupported length at either end.

At first I thought that the expansion of plastic snake inners wouldn't make any difference to model trim, then I looked up the coefficient of thermal expansion of polyethylene and calculated that a 1 metre length of snake would change length by 3mm for a 20C temperature change which would probably mess up elevator trim quite a bit. It's amazing what you learn!

[Don Fry]

Shaun, maths is not my strong point, but I think your 3mm should be 0.3mm. I’m on slightly stronger ground on plastics, and I don’t think the snakes are common or garden polythene. Polythene will give a figure with a three in it, but the more likely candidates have a expansion coefficient half of that.

 

As an aside, I have a biggish hack, acquired as is. It has a pair of Sullivan snakes, red/yellow. They are 95 cm long. It operates in temperatures from 5° C to 30° C. ( my limits for temperatures). It does not need trim changes beyond the odd click here and there. 
I still wonder how the snakes are installed in the aircraft with a reported 25 mm of travel.

Edited June 20, 2021 by Don Fry
Clarity

[Piers Bowlan]

25mm is a heck of a lot of expansion with temperature, even if the model was two meters in length! Surely something else is going on? Perhaps one of the ends of the snake has become unglued and is sliding in and out somewhat? Just a shot in the dark!

[Jonathan W]

I would agree that Sullivan snakes are normally very stable and also have a smooth action, requiring only small adjustments for temperature. Maybe something else is awry.

[Shaun Walsh]

28 minutes ago, Don Fry said:

Shaun, maths is not my strong point, but I think your 3mm should be 0.3mm. I’m on slightly stronger ground on plastics, and I don’t think the snakes are common or garden polythene. Polythene will give a figure with a three in it, but the more likely candidates have a expansion coefficient half of that.

 

As an aside, I have a biggish hack, acquired as is. It has a pair of Sullivan snakes, red/yellow. They are 95 cm long. It operates in temperatures from 5° C to 30° C. ( my limits for temperatures). It does not need trim changes beyond the odd click here and there. 
I still wonder how the snakes are installed in the aircraft with a reported 25 mm of travel.

Coefficient of of thermal expansion 150ppm/C for polyethylene x1000 for 1 M length x20C for temperature difference gives 3mm, have I misplaced  decimal point somewhere?

Good point about other polymers being used, any idea what they might be? I chose HDPE as a worst probable case.

Edited June 20, 2021 by Shaun Walsh

[Don Fry]

I think you are decimal point out. For no other reason that I don’t notice plastic sheet expand that much for that temperature range, and my fagpacket  gets .3mm  / m for 20° C

I would quess it’s a high density variant, HDPE, but it is not plastic carrier bag in a different shape. There are hundreds of polyethylene variants
By the way. This is interesting, but I am by trade a chemist, not an engineer. But my logic system says if the inner expands, and the outer expands, and the temperature increase is the same for both, and both have the same, or similar material;

Increase in length will be the same for both. Why get a trim change change? 

 

Now my machine does not change trim to any extent, despite having rodsa meter long.  I think I might notice 3 mm, I would doubt the trim could handle that. 
What does give large variants is poor or failed installation. Ie missing or failed anchors on the outer. 

[Martin Harris - Moderator]

Assuming a wooden airframe which will expand by a fraction of the (small) expansion of the snake (0.3mm not 3mm as I assume your estimate meant?) there will be a change in length from servo to horn - the endpoints of the outer being fixed will stop/limit bowing losses.  However, as a coarse flyer, I haven't ever noticed temperature induced trim changes although I might have re-trimmed marginally on a few occasions without thinking about the reason behind it.

[Don Fry]

Martin, cheers. So you are telling me, your aircraft also, trundles along without issues.

But, if the outer, and inner are anchored, would not the wood be irrelevant. And the wood also moves under the flying stresses. Quite a lot. 

Now, point one, it’s been a hard busy day. Boat holiday preparations.

Two, I've cooked dinner.

Three, lovely wine.

 

But small stuff. I don’t like snakes, but this brand, at least, work. They do not move by 25 mm, 3mm or whatever. Reality statement, given the vocal nature of Yank modelers, can’t we trust them that the silence of vilification on the brand, is praise indeed. 
 

For I think the 4th time of asking. How has Dale anchored, and placed intermediate anchors on the outer. Because, 25 mm of difference has nothing to do with expansion. It’s about failing to follow the instructions, or the installation has failed, ie, a glue joint has failed. Or something else. But snakes don’t alter by 25 mm. That an inch in old money. That’s longer than the thread on the end of these things what you screw the clevis on.

[john stones 1 - Moderator]

Servos behaving at centers ?

 

[Don Fry]

Could be, logical solution.

[Martin Harris - Moderator]

I've certainly had no problems with Sullivan snakes and have used them for many models.  You'll get no arguments that they haven't ever grown by 25mm!

 

As I understand it, the point of anchoring the outer is to stabilise the movement which would allow the outer to bow under compression or straighten under tension which would effectively change the linkage distance from the servo arm to the horn.  That is completely unacceptable and would lead to lost movement, unpredictable centring and possible flutter.  I can't see that it would have any effect on any change in length through expansion which would be the same whether the outer was anchored or not.

 

 

[Dale Bradly]

Sorry I was unclear,  the 25mm of travel is (a guesstimate) of the control surface deflection between hot and cold,  not the growth of the snake itself  which will be significantly less.  Yes the outer is well secured, and the snake operates in a more or less linear direction,  so I absolutely fail to see how the outer's security can impact the inner length. 

I'll get some more info later,  but I'm less interested in the science behind it  and more about suitable materials that could replace the inner component. 

[Graham Davies 3]

Evening Dale,

 

This is quite interesting and I appreciate your problem. However, I think you are better to nail down the root cause of the problem before changing material as I suspect you may find the problem still exists.

 

Have you tried a little control experiment? I appreciate this wont be easy, but set the fuselage such that the control surfaces are aligned to a ruler and then heating or cooling the area with the snake. If you see ANY variation at the trailing edge, you know that there is an expansion/ contraction problem. If you don't, your smoking gun is elsewhere.

 

The reason I suggest this is because as others have mentioned, I find sullivan snakes reliable and have never had more than a click or so of trim change throughout the season.

 

I have had far bigger trim issues with dubious servo centring.

 

Graham

[Dale Bradly]

Thanks Graham, I intend on doing exactly that over the next few days,  but I have already proven it, but am going to obtain further proof. 

Servo centering I firmly believe is not the issue, but again will prove that.  

[Martin Harris - Moderator]

It strikes me that if you're seeing an inch of control surface movement for that tiny difference in expansion, the linkage must be set up with very poor mechanical advantage.  How many degrees of arc does the servo arm move through from one extreme to the other?  I'd aim for around 45 degrees either side of neutral depending on radio and servo specs.  The more servo movement for a given deflection, the less effect expansion will have on the control surface with the benefit of more holding power and accurate positioning.

[Dale Bradly]

Right, some more info, and the conclusion that temperature only is to blame.

 

The model is an Airsail Chipmunk, snakes as per the plan.

The snakes are about 800mm long, and there are 7 bulkheads they pass through, so well supported. I do recall gluing them in, and certainly can't move the outers now.

Futaba S3001 servos on elevator/rudder. 

Unfortunately the surface ends are internal, so not visible and surgery will be required to get in there.

 

Servos at Midpoint, i.e. where i had them setup/subtrimmed during build in summer.

 

"Cold" photos. Whole plane about 1.5°C. (i.e. a typical Sunday morning at my club in winter)

 

 

View from aft. (fuse is inverted).

Elevator has approx 12mm of "fwd stick" applied. A boat load of left rudder on too.

 

 

After a couple of hours inside in front of the fire and waving the wife's hair dryer around it, estimate the model is 25-30°C.

during this time, care was taken not to bump the model, knock surfaces, and the radio has NOT been powered up, so the servos have not moved. (removing the centering issue)

 

 

I now have about 6mm of up elevator, and some right rudder. The only thing that has changed is the temperature of the model.

 

 

 

The servos when "hot". Note at the end of the blue outer where it comes through the bulkhead, a ring of dark blue (no overspray) is visible where the outer has grown by what i estimate to be 1mm. As I type this, the model is cooling down to room temp (currently 19°C) and this oversprayless ring is disappearing back into the bulkhead. One assumes the inner and outer are made of the same plastic type. 

 

Now you could say that clearly the blue outer isn't glued well at this particular bulkhead, but that isn't the answer as it shows that the inner will be growing at the same rate, and you can't secure that in place.

 

I measured the apparent growth in the pushrods, and it appears to be about 4mm at these temperatures over this distance. So this 4mm of pushrod change has resulted in 18mm of elevator change at the trailing edge of the elevator. As mentioned the rudder i didn't measure, but the photos above suggest my guess of 25mm wasn't far out. 

To answer Martin Harris' question, elevator is about 40° either side of centre  to give me about 18mm of throw either side of neutral, and rudder is about 45°/100% to give 30mm throw. (Manual states 16 and 30mm respectively). The elevator servo horn needs to be changed, as i have this at 70% travel after the last flight as it had too much elevator, however I haven't changed the mechanical yet to keep this the same while i figure this out. But all the adjustments in the world won't make up for 4mm of pushrod length change.

 

Not happy, but without major surgery, not much i can do other than carefully checking prior to each flight, retrimming as needed and accepting the fact that it'll forever be a moving target.

 

Edited June 21, 2021 by Dale Bradly

[john stones 1 - Moderator]

Models shrinking.

Answered your own question Dale, put up with it or commit to surgery.

[Gary Manuel]

Your problem is 2-fold.

1. The snakes are dead straight, which is not allowing them to curve as the temperature changes and cancel out any relative expansion. The yellow inner is effectively acting as a straight pushrod.

2. Your outer snakes are not anchored properly at the ends. They are sliding in the bulkhead. May be similar at the other end.

 

Cold.

 

Hot.

 

 

Edited June 21, 2021 by Gary Manuel

[Shaun Walsh]

6 minutes ago, Gary Manuel said:

Your problem is 2-fold.

1. The snakes are dead straight, which is not allowing them to curve as the temperature changes and cancel out any relative expansion. The yellow inner is effectively acting as a straight pushrod.

2. Your outer snakes are not anchored properly at the ends. They are sliding in the bulkhead. May be similar at the other end.

 

Cold.

 

Hot.

 

 

 

Carbon fibre has a very low coefficient of thermal expansion but if you replace the inner snake with CF tube it probably wouldn't be flexible enough to accommodate the bends in the outer tube without binding.

[GrumpyGnome]

4 minutes ago, Gary Manuel said:

Your problem is 2-fold.

1. The snakes are dead straight, which is not allowing them to curve as the temperature changes and cancel out any relative expansion. The yellow inner is effectively acting as a straight pushrod.

2. Your outer snakes are not anchored properly at the ends. They are sliding in the bulkhead. May be similar at the other end.

 

 

 

Ummmm, I may be a bit thick but I don't se how either of those 2 things would help....

 

1. irrespective of the straightness or otherwise of the outer, the inner will still expand and as the hinges on the control surfaces will be loose, this has to be the 'path of least resistance' resulting in a control surface movement

2. If the run is straight, this has no effect surely. If curved, then 1. above applies, no?

 

I think it's a fact of life, made worse if the pushrod joins the control horn at it's innermost position...

 

GG

[Dale Bradly]

Agree with your point 2 Gary, but this is a symptom of the expansion, not the cause creating the deflection.

[PatMc]

I'm not sure about Gary's first point. I've been re-sizing & overlaying the same two images in semi opaque layers & came to the same conclusion as Gary's second point.

Dale, I think the unseen tail ends of the snake outers must be firmly anchored (or at least better anchored than the visible front ends).   

Edited June 21, 2021 by PatMc