Jump to content

Replacing Sullivan Snakes


Recommended Posts

Yes, but if the ends were anchored properly and the tubes were allowed to bend a little (to accommodate the expansion), then the thermal expansion would be cancelled out as they should be when using snakes (because the 2 tubes expand equally).

Edited by Gary Manuel
Link to comment
Share on other sites

Dale, you say this model is a favorite. Has it done this since new? How is the airframe stored, drying it in front of the fire, and a bit of hairdryer will shrink a damp airframe. And wood re-sizes a lot more than plastic, due to damp.

Those snakes are from a long established and respected supplier. We have been using them for decades. My hack has them, longer than yours, same temperature range, no re trim issues, I even have 3001 servos in it. 
Those snakes do not move that much over that temperature range. A fair test, put some spare snake in the fridge, measure length. Warm it in a 50° C oven, remeasure. I reckon, unless using fancy instruments, you will find no difference. It would not be iro  8 mm.

An answer, if you have an airframe changing length is fit a closed loop (pull pull) system

Link to comment
Share on other sites

13 minutes ago, Don Fry said:

Dale, you say this model is a favorite. Has it done this since new? How is the airframe stored, drying it in front of the fire, and a bit of hairdryer will shrink a damp airframe. And wood re-sizes a lot more than plastic, due to damp.

Those snakes are from a long established and respected supplier. We have been using them for decades. My hack has them, longer than yours, same temperature range, no re trim issues, I even have 3001 servos in it. 
Those snakes do not move that much over that temperature range. A fair test, put some spare snake in the fridge, measure length. Warm it in a 50° C oven, remeasure. I reckon, unless using fancy instruments, you will find no difference. It would not be iro  8 mm.

An answer, if you have an airframe changing length is fit a closed loop (pull pull) system

A properly fitted snake should also cope with a shrinking airframe. It's all about anchoring the ends, keeping the length outside the anchor to a minimum and allowing the length between the anchors to flex a little.

 

Note that the curve between the anchors is necessary to accommodate expansion but has the unwanted side effect of allowing a little slop due to the difference between internal diameter of the outer, and external diameter of the inner tubes as they bend around the curve. This is why I don't like snakes much, but they do have their place.

 

Edited by Gary Manuel
Inner / outer corrected
Link to comment
Share on other sites

The apparent length change of the snake outers suggested by the visible movement of the unpainted blue section at the servo end is more likely actually the wooden airframe changing length. I suspect that the wood will shrink at the higher temperature which will actually add to the very small expansion of the snake itself.

As previously stated, properly securing the snake outers at both ends will minimise this effect to really only the short sections of snake between the securing points and the servos and control horns.

I have looked at my Airsail Chipmunk plan and it shows the elevator and rudder both with external controls, so I suspect you have built yours with internal horns as a modification. There could well be something moving in there. Perhaps you could at least cut an opening in the fuselage underside beneath the tailplane to see what is happening there. If there are bent wire joiners they could be changing with temperature.

Brian.

Link to comment
Share on other sites

Sullivan Golden Rods do grow and contract under changes of temperature. 

Many moons ago I built a PFM Team Special. Golden rods used on all control runs including ailerons with a central single central servo. On a warm day the ailerons would droop and on a cold day they would reflex and I could never trim the rudder and elevator.

One of the great advantages of golden rods is  the low friction.

Obviously all the comments regarding anchoring the outers correctly are all relevant.

My answer was to remove the fittings from the ends of the inners and remove them from the model. Thread some steel braided bowden cable inner down the inside of the golden rod inner. Solder a clevis adapter on each end, leaving about 3mm of float so that the golden rod can expand and contract along the steel cable. Thread the rod inner with its new steel core back into the outer and screw the clevises back on.

So now you have the temperature stability of the steel cable with the free running of the golden rod.

Shimples.

Edited by Wookman
Link to comment
Share on other sites

On 20/06/2021 at 12:20, Shaun Walsh said:

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.

 

I made it anywhere between 1mm and 2mm for nylon, but it seems figures for nylon can differ over a large range depending on nylon type. Glass filled nylon is a different animal entirely, much lower expansion, I found numbers that indicate only about 0.3mm on the same length with the same temperature change. I don't know if Sullivan snakes have a sprinkling of glass in them?

 

Some points of note, personal experience -

 

If you use large control horns, the effect is of course minimised... Caveat is, need to make sure the unsupported length of inner can cope with the extra movement.


Most of my install are on a gentle curve with the outer securely fixed / glued at both ends, supported in between at two or three formers (or similar) - not glued. That makes (if all is well with the world) the outer expand as much as the inner. Again serving to minimise any temperature effects, as the only important expansion is that which affects the ends of inner that stick out from the outer.

 

Most of my installs are probably around a half meter length of snake.

 

I have never noticed more than a few clicks change going between low single digit temperatures up to mid twenties. I use the things on most builds - always on elevator, sometimes also on rudder although the large movement of a rudder tends to favour a closed loop, I think. Your mileage may vary, etc.

Link to comment
Share on other sites

I had the same issue with a 'V' tailed 120" electric thermal soarer I built a few years ago, which had a straight control run from the servos to the control horns. In the winter the trim change was large enough to cause an ever increasing dive if left uncorrected. I completely resolved the problem by replacing the snake inners with with carbon fibre.

It has put me off using snakes on long control runs, although I'd still use them on smaller models.Eglider.thumb.jpg.95f22ae721ec3355c4cdf5c6366dcf63.jpg

Link to comment
Share on other sites

On 21/06/2021 at 22:41, Don Fry said:

Dale, you say this model is a favorite. Has it done this since new? How is the airframe stored, drying it in front of the fire, and a bit of hairdryer will shrink a damp airframe. And wood re-sizes a lot more than plastic, due to damp.

 

 

Don, yes it has done it since new, but as the model is only about 6 months completed, this is the first winter it's had. (Was completed and first flown in summer, and noticed the rudder would shift around a bit then. Airframe is stored in my mancave, which will reach 0° in winter, and perhaps mid 20's in summer. The fuse is fibreglassed, so i expect it would be more ridged than the plastic snake. 

On 21/06/2021 at 22:41, Don Fry said:

Those snakes do not move that much over that temperature range. A fair test, put some spare snake in the fridge, measure length. Warm it in a 50° C oven, remeasure. I reckon, unless using fancy instruments, you will find no difference. It would not be iro  8 mm.

 

And to this, i must disagree. I found 4 samples, including the offcut off one of the very snakes in this model. I got them to 1°, measured, heated to about 40°, measured again, and cooled again to about 1° to confirm. Now this isn't precision measuring with lab equipment, but i reliably found a change of 1% in all cases.

By pure chance, one of the samples happened to be 796mm long when hot, 799mm long when hot, a growth of 3mm, completely echoing what i found in the model a few days ago. 

 

So i am conclusively convinced that these have a huge change in length with regard to the temperatures i fly in (when considered against the total travel when in use), so i will have to put up with it or change it. Wookman has the idea, I'll try that next time the model is in for maintenance.

 

Thanks for the discussion gents, it's been interesting.

 

 

Edited by Dale Bradly
Link to comment
Share on other sites

I did my own experiments with some spare snakes and there is noticeable expansion. What I find odd is the extreme movement at the control surfaces. I have never noticed any change in trim on my own Airsail Chipmunk which I built with snakes in 2003 and have flown regularly since. 
 

I can only think that a combination of the loose snake outer and your control geometry is to blame. Until it was raised, I’d never really considered the self compensation of a bowing outer fixed at the end points although I’ve always ensured good adhesion on my builds using them. 
 

Simple test: When the airframe is warm, push the end of the outer back into the former to its original location - does the control surface return to its original position?

Link to comment
Share on other sites

34 minutes ago, Martin Harris - Moderator said:

Simple test: When the airframe is warm, push the end of the outer back into the former to its original location - does the control surface return to its original position?

Nope. And it wouldn't as this has no bearing on the amount of actual pushrod fixed between the two horns.

Link to comment
Share on other sites

Dale.

If I understand your latest post correctly, you are planning on lining the nylon inners with steel ones. You have said that you are less interested in the science and more interested in better materials to use as the inner snake. To this end, your question has been answered. This will improve the situation but not fully resolve it. Nylon has a coefficient of linear expansion of about eight times that of steel. This will result in your 25mm of control surface movement reducing by a factor of eight down to around 4mm. This is a great improvement which you may be happy with, but it is not the cure.

 

As has been mentioned previously, the cure is to properly anchor the ends of the snake outers and allow the length between the anchors to flex a little.

 

Edited by Gary Manuel
Link to comment
Share on other sites

The only expansion affecting the control surface in the fixed ends with some bowing permitted scenario will be from the unsupported length of the inner which should be as short as possible anyway if adopting good practice. I admit I've failed to appreciate the importance of allowing some bowing in the past but it's an integral part of the Bowden cable system when you think about it.

 

Did you actually try pushing the outer back towards the former?  The distance between the two ends of the inner is fixed.  Pushing the outer backwards (i.e.to what should be the fixed point) will force the inner to take a longer route so the distance between the horn and the servo arm will change, returning the control surface to its previous position. The similar expansion/contraction of the inner and outer during temperature changes simply allows the change in length to be accommodated by the longer/shorter "cable" route - not the case if the outer is free to move relative to the airframe when the expansion will move the control surface.

 

Allowing one end of the outer to float will allow unwanted movement (the reason I've always taken pains to fix them) which can lose precision or in some cases, allow catastrophic flutter to develop.

  • Like 1
Link to comment
Share on other sites

Catastrophic flutter, yes I have heard that happen on a high speed swept back wing model, not mine.

 

I launched it for him and it flew great, fast and seemed in perfect control.

 

Then in a 30 degree ish dive we heard this " strange noise" which was elevator flutter.

 

Loss of control and it carried on in the dive, flew into the deck.

 

One right off.

 

He also crashed his other model that day, a largish biplane.

 

I recon this crash was due to a discharged Rx battery as the model was difficult to start, it took ages to get going.

 

I did suggest that by that point the Rx battery was getting a bit low.

 

A bad day back in the late 80's.

 

Flutter can and will happen if allowed, I know, I heard it and saw/watched the result.

Edited by Rich Griff
Link to comment
Share on other sites

40 minutes ago, Martin Harris - Moderator said:

The only expansion affecting the control surface in the fixed ends with some bowing permitted scenario will be from the unsupported length of the inner which should be as short as possible anyway if adopting good practice.

 

Call it 25mm poking out at each end, 50mm total.

 

If we can assume an expansion of 3mm for 20degC over 1 meter, then, applying that same expansion to the 50mm of exposed inner gives a change of just 0.15mm.

 

Link to comment
Share on other sites

43 minutes ago, Nigel R said:

 

Call it 25mm poking out at each end, 50mm total.

 

If we can assume an expansion of 3mm for 20degC over 1 meter, then, applying that same expansion to the 50mm of exposed inner gives a change of just 0.15mm.

 

Or to put it another way. That's an improvement by a factor of 20, which is actually better than switching from Nylon to Steel.

 

Note that this "improvement factor" of snake cables is not related to the material used. It could be made of a material which doubles in length with a change of 1 degree and the improvement would still be 20:1. This is because it's directly related to the total length of cable compared to the unsupported cable (1000/50 = 20). This is the fundamental design principle of snakes / Bowden cables / bicycle brake cables etc.  But it only works if the ends of the outers are anchored solidly and the central part is allowed to flex to accommodate both inner and outer tubes as they expand and contract at equal rates.

Edited by Gary Manuel
  • Like 1
Link to comment
Share on other sites

A complete guess, a figure plucked straight from the air, 5000 - 20000 models have them installed in these last decades of either plastic or nylon, SLEC or Sullivan snakes

I think we may have come across problems caused by excessive thermal expansion of said snakes.

What we have come across, causing issues, is the outer not being fixed and moving, and servos being asked to do too much work in their inner arc.

Snakes are a go to surface connection method all these years.

Edited by Denis Watkins
  • Like 2
Link to comment
Share on other sites

Here's another thing. Without the curve between anchor points, the outer snake will have nowhere to go when it expands or contracts. It will either break one or both anchor points, or it will deform and probably break some part of the airframe. This is inevitable and the very strong expansion / contraction forces are at work continuously, including over winter and summer storage.

Link to comment
Share on other sites

IIRC this was a bit of an issue a while back, so much so that I checked the instructions in a set of Gold ;n Rods. My memory was that the instructions stated that the outers should be anchored at their ends and installed in a gentle curve. When I checked my memory was incorrect, there was no mention of installing in a gentle curve. However having used these for over 25 years, installed in such a manner, in a number of models I don't recall it ever being an issue.

The other thing is that the OP is in NZ and might conceivably encounter temperatures somewhat closer to the tested 40 degrees- something that is quite unlikely in flight or outdoors in this country.

Link to comment
Share on other sites

Dale, I absolutely believe your problem is caused by thermal expansion of the snake inner.

Why do I say that?

Last week during the hot weather my new Viscount was left in the sun upside down after being assembled while I flew other models.    It was then placed in the pits ready to fly.  Control checks prior to taxiing out showed the elevator  movement which should have been 50% up and down was about 80% up 20% down  The rudder was also out of alignment.

The flight was cancelled and advice sought from the several very experienced modellers present.  After investigations and tests the consensus was that the elevator and rudder snake inners had expanded in the heat.  Most of us had used snakes for years on various models and this was the first any of us had seen such an extreme occurrence.  The model was placed in the shade and left to cool off.  As the model cooled off over a couple of hours the  the control surfaces gradually returned to near normal positions.

The snakes are just under 1m long, properly fitted and secured.  The underside of the fuselage which was in full sun is dark blue in colour and the dark colour extends the whole length of the fuselage.  Control surfaces have quite a large cord (scaled from the full size aircraft) with a relatively small angular movement required to give the required control.  This probably amplified the effect of the expansion in the snake core.

 

What to do about it?  It was too large an error to deal with using trims, so short of regularly checking and resetting the transmitter settings on hot days a more permanent answer was required.  This looked like involving replacing the snake core with a different material.  Carbon rod was considered but would probably have been too stiff to follow the curvature of the fuselage -  the reason for using snakes in the first place.

 

The answer came in the form of Bowden cables sourced from Modelfixings, together with solder adapters for the ends.  The cable and its sheath fitted easily inside the snake outer, with the sheath being cut to the length of the snake outer.  It involved an afternoon of keyhole surgery to fit these into the model and the results seemed promising.

 

The model was assembled at the patch again yesterday and left in the sun for several hours to warm up.  There was no movement of elevator or rudder from their centre positions.  Control checks were normal and the model was flown without incident or requiring any change of trim from earlier flights.

 

Will I use this method again?  Probably, if long snakes are the best solution for the model in question.  Better to eliminate a possible problem rather than have to fix it later.

Will I take extra care during pre-flight control checks - absolutely.  I doubt that the model would have survived had I flown without spotting the problem.

IMG_3314.JPG

Link to comment
Share on other sites

31 minutes ago, Martin & Sue HARDY said:

Dale, I absolutely believe your problem is caused by thermal expansion of the snake inner.

Why do I say that?

Last week during the hot weather my new Viscount was left in the sun upside down after being assembled while I flew other models.    It was then placed in the pits ready to fly.  Control checks prior to taxiing out showed the elevator  movement which should have been 50% up and down was about 80% up 20% down  The rudder was also out of alignment.

The flight was cancelled and advice sought from the several very experienced modellers present.  After investigations and tests the consensus was that the elevator and rudder snake inners had expanded in the heat.  Most of us had used snakes for years on various models and this was the first any of us had seen such an extreme occurrence.  The model was placed in the shade and left to cool off.  As the model cooled off over a couple of hours the  the control surfaces gradually returned to near normal positions.

The snakes are just under 1m long, properly fitted and secured.  The underside of the fuselage which was in full sun is dark blue in colour and the dark colour extends the whole length of the fuselage.  Control surfaces have quite a large cord (scaled from the full size aircraft) with a relatively small angular movement required to give the required control.  This probably amplified the effect of the expansion in the snake core.

 

What to do about it?  It was too large an error to deal with using trims, so short of regularly checking and resetting the transmitter settings on hot days a more permanent answer was required.  This looked like involving replacing the snake core with a different material.  Carbon rod was considered but would probably have been too stiff to follow the curvature of the fuselage -  the reason for using snakes in the first place.

 

The answer came in the form of Bowden cables sourced from Modelfixings, together with solder adapters for the ends.  The cable and its sheath fitted easily inside the snake outer, with the sheath being cut to the length of the snake outer.  It involved an afternoon of keyhole surgery to fit these into the model and the results seemed promising.

 

The model was assembled at the patch again yesterday and left in the sun for several hours to warm up.  There was no movement of elevator or rudder from their centre positions.  Control checks were normal and the model was flown without incident or requiring any change of trim from earlier flights.

 

Will I use this method again?  Probably, if long snakes are the best solution for the model in question.  Better to eliminate a possible problem rather than have to fix it later.

Will I take extra care during pre-flight control checks - absolutely.  I doubt that the model would have survived had I flown without spotting the problem.

IMG_3314.JPG

If your snake inner & outer are of the same material & the outers are absolutely secured against movement at both ends then I'm afraid the explanation doesn't make sense. 

  • Like 1
Link to comment
Share on other sites

After checking my Airsail plans, I would suggest that the snakes may be over supported through the fuselage at every former if run through tight holes in them. It's 18 years since I built mine and I don't recall the exact details of getting them in place but it hasn't given me any problems.  Maybe there's insufficient flexibility in Dale's outer to allow any bowing if the outer was passed through tight fitting holes or secured at the intermediate formers?  In the former case, expansion over the length of the outer may well have caused the failure of the glue bond at the forward former if the snake was unable to bow.

 

It's a good lesson in the correct installation of snakes!

  • Like 1
Link to comment
Share on other sites

When fitting 'snakes' or bowden cables to a model I always ensure the ends are roughened and securely epoxied at either end. They rely on the relative lengths of 'inner' and 'outer' being maintained, they will not work if one end of the outer is free to slide in and out because of temperature changes, or any other reason for that matter. I suggest Dale applies a liberal quantity of epoxy to secure the servo ends of the snake outers to the model, to stop them moving. If you replace the snakes without securing the outers properly you will still have the same problem regardless of what material they are made from.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...