Spacewalker half scale 168" electric

[Andy Boylett]

http://www.modelfixings.co.uk have them.

Chris,
thanks for that, great shop.
Andy

[Mark Partington 2989]

Andy,

I agree with Phil about the use of circlips on control linkages.

I've never seen a circlip fitted on a control system linkage in full size aviation, (I've fitted them just about everywhere else on an aitrcraft though), the minimum that I've seen or fitted is a cotter pin with a washer and split pin, but more commonly it's a bolt and nut. Get yourself some proper bolts, the thread isn't very long, usually enough for a washer, nut and safety threads.


Mark

[Andy Boylett]

Hi Mark,
OK I hear what you are sayinng and I can easilly use a different system......
But, as well as that....
These fittings are proprietry machined clevises with a 3mm clevis pin that is retained in place by spring clip fitted into a circlip groove. They are made for 3mm sized control functions. I have looked to change the spring style retainer for a circlip as this will be secure compared thefitting supplied. I can understand full size being different as I was not aware of similar control systems to this on a full size (with a mere 16kg/cm torque). I am quite interested in understanding this and Phil said you have expertise in aircraft engineering. So rather than guess work or subjective views with regard to circlips, with your expertise in full-size linkages (I am guessing you have calculated this sort of stuff before) can you please help me calculate the loads that would be invloved and explain why a circlip would not suffice?
Data:
- I am using 31kg/cm torque servos with a 50mm arm.
- servo speed is 0.118 sec for 60 degree arc.
- Clevis pin is 3mm diameter and circlip groove is 2.8mm diameter.
- Calculated aileron servo required maximum torques at twice actual air speed is 16kg/cm.
- The clevis is high grade aluminium alloy, as is the clevis pin (the ally is harder than just normal ally).
- The servo arm is annodized ally (can't tell hardness without scratching and I don't wan tot do that).
If these fittings are hence not up to the job then I will be asking the manufacturer for a refund

Here are the specs for the circlip. As you will see the clevis groove strength is 35lbs and the circlip shear load stregnth 98lbs Tough little cookies....this probably explains why brushless motor outrunner cages can hang off the end of these.

ab2011-03-04 circlip spec 2.jpg (15.16 KiB) Viewed 12439 times

Regards
Andy

[Mark Partington 2989]

Hi Mark,
can you please help me calculate the loads that would be invloved and explain why a circlip would not suffice?

Andy,

It's nothing to do with calculations, it's all about redundancy, and for this redundancy a nut and bolt would be split-pinned or a nut or bolt would be wire-locked, whereas a cotter pin and split-pin would only be used on a trim tab, (the main control surface would be a nut and bolt with a split pin), or on 'secondary' control linkages, even a self locking nut would not be used on it's own nor would any other fastener that has no secondary 'retaining system'. I've had a thought about where circlips would be used, and concluded they are only used to stop things like bearings sliding out of position, holding shafts in place or as a secondary retainer for other fasteners.

As an example - the main bolt for a rudder would be a 1/4"nut and bolt, (possibly 5/16") with a split-pin, whereas the trim tab on the same rudder would be a 3/16" cotter pin with a collar and split-pin. both are fitted to an aircraft currently flying and capable of supersonic speeds.


Mark

Edit: I've just read the spec sheet for the circlip, and it's to fit a groove of .004" depth in a 1/8" dia pin, not a 3mm pin which is 0.2mm under 1/8". Never ever mix metric and imperial fasteners, they don't go together.

[Andy Boylett]

OK Mark, I understand that, but the issue here is whether an engineered circlip is capable of being one of those means of fixings.

We have not been discussing whether we should use one or two methods of fixing. If for CAA approved over 20kg builds you are suggesting that on all primary control surfaces there should be 2 methods then we all have a lot of work to do adding secondry fixings everywhere there is just is one...which is just about everywhere

So 2 seperate issues
- one or 2 fixings?....I will leave that one to you to answer.
- can an engineered circlip be a fixing?...my comments below.

Circlips
Firstly, one of the reasons a circlip is not used where a secondry fixing method is required as well is that it does not lend itself easilly to this. A nut can have a split pin through it etc, or be a castled nut for same....but circlips are usually just used on their own.

In my installation above I am not suggesting using a circlip in a high load load appliiaction, although as I have pointed out the prop is held on to the rest of the plane by one. The installation we are discussing is where a circlip holds in place a pin and the load on the pin is at 90 degrees to the direction of fitting of the pin ie the load does not pull the pin out of it's hole.

There is a major difference between the control links on our models compared to full size and one that was obvious early on my build. I was going to design purpose made actuator links that were all internal and would not spoil the outside line of the plane. I did nto do this because Phil did not like it and was concerned about any play possibly being an issue as the control movements are so small. And this is the difference. If I had put the controls internally then the moment arm would have been 10mm rather than the 50mm I am using now externally. The maximum calculated aileron servo torque at 120mph is 16kg/cm, so at 10mm this would be a load of 16kg on the control linkage and at 50mm the load is just 3.2kg. This is not a dissimiar load to that exerted on a toilet cistern handle ie its not very much.

Now, full size control linkages have very small control thows wrt the size of surface being moved and have very high linkage loadings. Do you have any examples of linkage loadings on a hydraulic actuator? These are massively different beasts to the ones we are using and if you calculate the forces involved in the linkages and the method of fixing you will understand why a seccondry fixing is required (I will explain further if needed).

I know that the circlip I propsed to use on this little link is absolutley fine and I can show the engineering behind that thinking if needed. I do not think suggested items should be dissmissed just becuase of subjective or circumstantial evidence. Technical reasons should be given and discsused so that we understand and do not work by 'black art'. hence why I started asking technical things.

Another point. Just to very clear. The 3mm circlip I suggested using is not a "C spring" and it is not a "C clip" (which is what is used in most small brushless motors).

So, is a 3mm engineered circlip suitable for holding in place a 3mm diameter clevis pin on a control link with a loading of 3.2kg? I think so.

[Andy Boylett]

Mark, re your last Edit....Come on.I am an engineer. I would NEVER EVER mix imperial and metric. What do you think I am????

I put a copy of the spec for a proper circlip in the above link from the only on-line one I could find quickly. These are imperial and designed for 1/8 inch shafts. I have ordered metric for a 3mm shaft, but you cannot download the manufacturers spec sheet on-line. The purpose was to show the non engioneers what loads a real circlip can actually take.

I have even reamed the hole in the servo arm so that it is a superb fit with the clevis pin.

Andy

[Andy Boylett]

Andy,
... I've had a thought about where circlips would be used, and concluded they are only used to stop things like bearings sliding out of position, holding shafts in place or as a secondary retainer for other fasteners.

Mark

Mark,
Engineers design for use of circlips where they are the optimum fastner for the appliaction. This can be in very high demand situations. One such example where circlips are used is to hold the gudgeon pins in place in many high powered bike engines - pretty tough one.
Where you have seen them on shafts this is most likely to take an end thrust load becuase they put lots of metal into shear loading which is really strong - like on my main motors.

From the way you were initially talking I assumed you were familiar with engineering fastener design and selection? As I said above, if one of the selection criteria for a fastner is that it must be backed up then a circlip cannot be used, hence why in the areas you have seen in aircraft linkages where the engineers specify 'dual', they are not used. As I do not beleive the LMA have specified that dual redundancy fastners be used this means a circlip of correct design is fine.

I gave you all of the data required above to be able to calculate the loadings on the circlip and hence show that it is ok. I will explain for you by showing you the sums why a circlip would not be used on a very high torque control linkage.
My application is to hold a shaft in place

ab2011-03-04 Circlip.jpg (14.09 KiB) Viewed 12370 times

The control linkage has a load of3.2kg or say 32N. If all of this was exerted at a precise angle of 90 degrees to the clevis pin then the there is zero load trying to push the pin out and hence zero load on the clevis pin.
Now lets assume there is an out of line of 1 degree on the linkage. This actually quite a lot but lets use this figure anyway. If we assume there is no friction between the clevis, clevis pin and the servo arm then the side load would be 32xsin1 = 0.56N or about 60grams!!
However, there is a bigger side load that can be induced due to the turning of the clevis pin as the servo arm moves. Now this is the one that an engineer working on a full size aeroplane would worry about. It is quite difficult to explain but I will use a simple example to describe it. Imagine that the clevis pin was threaded and the hole in the clevis arm was also threaded. Every time the servo arm rotates the clevis pin would also rotate on its thread and move sideways a little (this is what can happen if you use a threaded bolt!!). Now because the pin moves sideways it puts a load on whatever is stopping it move. If it moves away from where the circlip is, then circlip is pulled into the side of the clevis. Now as the roatation continues the circlip can try to ‘unwind’ out its groove. I have to point out that this requires extremely high loads, only found when used with hydraulic actuators.
In the example I have the side load is immediately overcome by the strength of the circlip and the shaft slides. If you want to work this out you look up the coefficient of friction between the materials being used and hey presto done.
In my example the actuator load is 32N and the side load is a small fraction of this. The circlip proposed can take over 150N, 5 times the total load.

[Mark Partington 2989]

Andy,

You're not listening to what I said.

The reason for redundancy on the fasteners is nothing to do with engineering design, but simply because without it the connection is not considered safe.

Ref you point about full size 'horns- for manual control they vary in 'length' between 25% and 50% of the distance from hinge to trailing edge at MAC, (rudder horns tend to be shorter than elevator as the pilots legs are more powerful than his arms), so not what you refer to as a 'short control throw', in fact probably longer than a lot of models. Even fully powered controls use a 'horn' as long as possible- e.g. surface length approx 40cm, 'horn' length 9cm (that's on a mach 2 fighter btw) or surface length approx 200 cm, 'horn' length of 35-40cm (500+ ton airliner), check out the elevator horn on a Spit or Hurri.

back later with more.....


Mark

[Andy Boylett]

Andy,

You're not listening to what I said.

The reason for redundancy on the fasteners is nothing to do with engineering design, but simply because without it the connection is not considered safe.

Ref you point about full size 'horns- for manual control they vary in 'length' between 25% and 50% of the distance from hinge to trailing edge at MAC, (rudder horns tend to be shorter than elevator as the pilots legs are more powerful than his arms), so not what you refer to as a 'short control throw', in fact probably longer than a lot of models. Even fully powered controls use a 'horn' as long as possible- e.g. surface length approx 40cm, 'horn' length 9cm (that's on a mach 2 fighter btw) or surface length approx 200 cm, 'horn' length of 35-40cm (500+ ton airliner), check out the elevator horn on a Spit or Hurri.

back later with more.....


Mark

Mark, I heard all of that but you are not listening and have I no idea what you mean about me not listening. I thought I laid out my replies clearly. I will try again....

You are not discussing the issue which is about using a circlip and not about using dual redundancy. They are 2 different things. So...

1. If you are saying that all over 20kg planes need to have dual redundancy on control fittings then tell the powers to be and if they agree we all have to fit them.

2. If they do NOT have to have dual redundancy then a single one will do and we can get back to the discussion about types and design as I was doing because Phil and then you said a circlip should not be used....so now is your chance to explain the technical basis for making that assertion?

I think a new thread should be started if you want to discuss point 1 further.... as I am sure many people will be very interested in it.

I am interested in 2. Do you have any technical knowledge or detail with regard to the design of circlip applications or why you said it should not be used?

Andy

[Andy Boylett]

Well I have to say that cutting wing tubes has to be done very carefully! Anyone need 4 very short wing tubes, 50mm diameter?

IMG_2790.JPG (31.23 KiB) Viewed 12267 times

[Phil Clark]

A few more questions Andy.....................

Your clevis pin is 3mm dia correct?

The circlip is 2.8mm internal diameter correct?

This gives a circlip groove of just 0.1mm deep correct?

You seem to know what the circlips are made from, and say the clevis pins are made from aluminium correct?

I would therefore assume the circlips are MUCH harder than the clevis pins and after a few installations/removals of the circlip; the grooves in the pins could become slightly damaged or worn? If wear or damage occurs, will the circlips not then become a little loose? I would suggest it definitely will.

I still have concerns over the integrity of a 3mm circlip & its ability to retain its original shape even when using the correct pliers (they are incredibly fragile from past experience). What is the likely outcome of a slightly damaged circlip running in a worn/damaged groove? (Not rocket science to figure that one out!!)

With only a 0.1mm deep circlip groove, surely it won't take that much wear before it's possible for a circlip to become misaligned with it’s groove or even simply 'fall off'. Though your model is electric powered, it will still be subjected to a considerable amount of vibration simply through airframe resonance......been there, done that with BIG electrics remember. Your model is timber & fabric so it is likely to resonate more than the rigid full composite structure I worked with.

Also, what are the shear load properties of the clevis pin when even small loads are applied to a 0.1mm wall? (Impossible to find out unless you know exactly what material the clevis pin is made from I would imagine?) The applications you have mentioned so far (brushless motor shafts) all have steel shafts which I would assume have a substantially higher shear load capacity than a 0.1mm deep groove in an unknown type of aluminium?

I am not an engineer, I am a model maker.....I am working from 25+ years of experience, 15 of which have been on projects between 10 and 80kg's in weight. Many of your calculations go beyond my knowledge I'm afraid, but I don't see what is wrong with pointing out concerns and working from 'Black Art' as you call it. The LMA is not an association frequented entirely by engineers, we come from all walks of life and the only way for the majority to succeed in their own projects is to look at other projects completed successfully by others and 'copy' their installations. If a particular method or component is flagged up as having caused issues elsewhere, then is it not just sensible practise to avoid it and look for another solution?

Phil

[Mike altham]

Hi guys

With working on full size and also the high g 3d models all the linkages use cotter pins or split pins.

These if correctly fitted have never failed. C clips I would never use.

Not in the same context but I have had large petrol cars and these had c clips on the ends of the wishbone steel pins and these had to be replaced every time I took it out as they popped off.

The best way that we use at work is have a high tensile hex or cap head bolt, add a washer under the head,
Put bolt through the clevis etc then another washer, a castlated locking nut, drill a hole through the bolt then split pin.

This will never come undone. Smallest we do though is 3/16" bolts not 3 mm

Dubro method is quite good as well.

Just a thought.

Cheers

Mike

[Andy Boylett]

Phil,
Interesting discussion.
There is nothing wrong with past experience, that is what we all base decisons on and the first thing engineers are taught....we keep on learning for ever and should keep on asking questions.

Circlips.
A circlip has a very small gap in its circumference, about 1mm in the case of the ones I am using and is made of thicker material than a 'C clip' so that if fits snuggly in its groove. The shaft you are therefore worrying about wearing would have to wear 2/3 of the way through it before the circlip could come off in a sideways direction. In the longitudinal direction the shaft would have to wear down to a diameter of less than 2.8mm along the whole lenght of the protruding shaft. This is just not going to happen. Circlips are also not as easy to bend or damage as you are making out (see below). Circlips are in fact very robust.

C Clips
I think that you may be confusing C clips with circlips. C clips are stamped out from thin material and are a cheap version of a circlip. They only cotact a shaft in 3 places and are fitted by twisting them sideways and then forcing them along the shaft. Sometimes they are called spring clips. It is C clips that are very suseptible to being bent when removed as they do not have circlip plier holes and people use pliersd and screwdrivers to remove them. It was very nasty little C clip that was supplied with the clevis we have been discussing. Because these these also touch down on only 3 places they are more eaisly displaced and become loose when slightly damaged. Many small brushless motors use C clips - I will show you some tommorrow.

You also write as though I am using circlips. I said right at the begining of this discussion "OK I hear what you are sayinng and I can easilly use a different system".

The rest of the the discussion has been to understand the basis on why you and Mark reject using a circlip because I so often see wrong conclusions being drawn. Mark's reason would appear to be that he sees dual redundancy fastners used when he is assembling full size planes and a circlip cannot be used like that. I would be interested to know if you have experince of a circlip being an issue due to wear because in my experience that would be unusual.

Mike, Yes, I always find C clips once removed are no good anymore and I have a supply of new ones for my motors. For my clevises I had already decided to drill a small hole through the end of them and then use a cotter or split pin (as you suggest ). These will be removed so infrequently that replecing them each time would be fine as well.
Cheers
Andy

[Andy Boylett]

I have now completed all of the aileron hinges. These have been quite a chalenge as the hinge point is some 40mm behind the back edge of the wing. This is beacuse the aileron has a completely rounded front edge that slots into the curve on the training edge of the wing.

I have made the hinges so that the ailerons can be removed. This is something I have always just liked doing so that I can remove the aileron in future if anything untoward happened . The hinges themselves are the Sierra Ginat Scale 10mm (3/8") ones. They are ally and are ribbed at one end for gluing in and the other end is left smooth.
The ribbed end is glued into cyparis blocks set into the aileron structure. The smooth end is a perfect tight sliding fit into 10mm thin walled ss tube. So, I have epoxied ss tubes into the wing side and left them protruding out because of the overhang....

IMG_2785.JPG (26.29 KiB) Viewed 12213 times

The gold part is the Sierra hinge that is annodised.
Here is one aileron being fitted..

IMG_2784.JPG (37.84 KiB) Viewed 12213 times

The end of the ss tubes is such that the hinges butt up nicely to them. To hold the hinges in place I have drilled and tapped the smooth end so that a 6mm nylon bolt can be inserted from the inside.

[Phil Clark]

1) Sorry Andy, I may not be a trained engineer, but I am FULLY aware of the differences between a 'C' clip and a Circlip

2) I obviously have hands like shovels as I find small CIRCLIPS very delicate & easily damaged

To answer your question regarding circlips showing signs of wear....yes I have. A slightly different application I agree, but on some retract spool valves I use, they use 2 circlips to retain the spool in the valve and act as end stops. Some cheap valves have used C clips, better quality ones circlips. I have had one instance where one of the better quality valves 'started' to fail (I spotted it before it became a problem) because the shallow grove the circlip was seated in started to show signs of wear around it's edge and the circlip had started to jump out of the groove. The circlip was obviously a hard spring steel of some sort, the spool of the valve was softer aluminium. The valve had not been tinkered with; the wear was due to vibration and general wear & tear. A retract valve is a relatively non critical piece of kit....we can cope if our wheels don't go up, or won't come down.........in your application??????

I see you are now fitting a split pin in addition to the Circlip......this is perfectly acceptable as this will serve 2 functions in helping to prevent the circlip coming off if it does becomes loose, and preventing the clevis shaft falling out in the event of a circlip failure.

Phil

[Phil Clark]

Aileron hinges..............

1) How are the bolts accessed?

2) I assume the st/st tubes are bonded into hadwood blocks?

2) What provision has been made inside the wing to prevent the thin edge of the st/st tube cutting into the underside of the bolt head?

[Andy Boylett]

Aileron hinges..............

1) How are the bolts accessed?

2) I assume the st/st tubes are bonded into hadwood blocks?

2) What provision has been made inside the wing to prevent the thin edge of the st/st tube cutting into the underside of the bolt head?

Hi Phil,

1. There is the largest access hatch I have ever made...

IMG_2782.JPG (35.16 KiB) Viewed 12202 times

You can see it runs all the way along and gives access to ailerons and servos.

2. Yes

3. I couldn't get my camera inside to show how I avoided that. The ss tubes do not go all the way through the wood blocks and finish short of the end of the holes. Hence the hylon bolts only touch the wood. I could actually use steel bolts, just nylon is lighter. One of the reasons I set the tubes in is that they are positioned to put the aileron in precisely the cirrect position. I did this by plugging the inner end of the ss tube with tissue paper, then greasing the hinges (so nowt sticks ), then epoxying the ss tubes and holes, then assembling the ss tubes onto the hinges, then sliding everything into place until it touched temporary wooden blocks I fitted to give correct clearance. Some time later I removed the aileron and then with a big pin punch and a 5lb hammer the tissue paper could be knocked out....and b****y stuck it was too . I can definately say the ss tubes are well glued in !

Cheers, Andy

[Andy Boylett]

1) Sorry Andy, I may not be a trained engineer, but I am FULLY aware of the differences between a 'C' clip and a Circlip

2) I obviously have hands like shovels as I find small CIRCLIPS very delicate & easily damaged

To answer your question regarding circlips showing signs of wear....yes I have. A slightly different application I agree, but on some retract spool valves I use, they use 2 circlips to retain the spool in the valve and act as end stops. Some cheap valves have used C clips, better quality ones circlips. I have had one instance where one of the better quality valves 'started' to fail (I spotted it before it became a problem) because the shallow grove the circlip was seated in started to show signs of wear around it's edge and the circlip had started to jump out of the groove. The circlip was obviously a hard spring steel of some sort, the spool of the valve was softer aluminium. The valve had not been tinkered with; the wear was due to vibration and general wear & tear. A retract valve is a relatively non critical piece of kit....we can cope if our wheels don't go up, or won't come down.........in your application??????

I see you are now fitting a split pin in addition to the Circlip......this is perfectly acceptable as this will serve 2 functions in helpingIt to prevent the circlip coming off if it does becomes loose, and preventing the clevis shaft falling out in the event of a circlip failure.

Phil

Phil,
I have seen those on retract spools as well (just sold a set of new ones, Dave now has them!). The valve puts its load 'full on' and 'full off' on the clip you talk about. It is also a considerably high load to be putting on a small circlip - we all know just how much force is needed on the end of the shaft, sometimes they are difficult to move . Not surprising it can wear then. The clevis pin application I am talking about is 90 degress different in that it does not take the load and has no inherent loading on it at all.

I was not proposing a split pin in addition to a circlip, instead of.

My other idea was to drill/tap the end of the pin and fit a 2mm bolt in it (threadlocaked of course ).

You also looked at and approved the Protech fittings in November (for me to use as hinges on the rudder and elevator). I showed you one made up. There are lots of photos near the start of this thread (a little way down page 1 there are photos). This is why I ordered 100 circlips.

[Phil Clark]

Phil,
I have seen those on retract spools as well (just sold a set of new ones, Dave now has them!). The valve puts its load 'full on' and 'full off' on the clip you talk about. It is also a considerably high load to be putting on a small circlip - we all know just how much force is needed on the end of the shaft, sometimes they are difficult to move . Not surprising it can wear then. The clevis pin application I am talking about is 90 degress different in that it does not take the load and has no inherent loading on it at all.

No it doesn't.

If a retract valve/servo is correctly set up, there should be NO load on the circlips at all. If the servo is pushing/pulling the spool against the circlip, it is straining with too much movement putting unecessary loads on the circlip. The servos travel should be adjusted so there is no straining/buzzing with the circlip 'just' touching the end faces of the valve body at either end of it's travel. This is how I set all my valves up but I have still had this one instance where a circlip seat became worn and the clip would have jumped off if I hadn't have spotted it. How it happend I do not know, all I know is it did happen, so the use of circlips for me on 'critical' components is to be avoided if at all possible.

I was not proposing a split pin in addition to a circlip, instead of.
My other idea was to drill/tap the end of the pin and fit a 2mm bolt in it (threadlocaked of course ).

A 2mm tapped hole through a 3mm 'pin'......that leaved just 0.5mm of 'meat' either side of the hole. Unless you can be sure that your tapped hole is spot on central, that sounds like very little material left to me. In steel, Ok at a push, but not in soft aluminium.

You also looked at and approved the Protech fittings in November (for me to use as hinges on the rudder and elevator). I showed you one made up. There are lots of photos near the start of this thread (a little way down page 1 there are photos). This is why I ordered 100 circlips.

I did.......but consider once fitted they should never be removed so the main caused of damage will be avoided and you have if I remember correctly 4 hinges per control surface, so for a critical failure to occur, all 4 circlips would have to fail.

[Andy Boylett]

Phil,
The rudder and elevator hinges have the nasty C clips on them. I was proposing to use propper circlips instead.

A circlip in a matching groove will not wear unless there is a load on it - maybe not set up correctly?

The idea of using a small bolt was the 3rd in line after circlip and cotter pin. It is actually not difficult to drill and tap straight when set up in the milling machine and indexing head because you can move the drill to centre acurately with the slide-ways. Anyway, as I said, third on the list

I would not like for one minute to suggest having an aileron not work is good, but I have 4 of them!!

After all this and since I have these fittings in pace already I think the simplist is to go with your original suggestion
"Andy, I'd still prefer not to see the C clips unless you can find some that are a VERY good & tight fit & you are willing to replace the C clip every time the linkage is disasembled."
The circlips will be a magnetude better than the C clips supplied, so should be fine then.
Regards, Andy