Topic: 30% Douglas Skyraider - HAS FLOWN

Ali.

Yes................the stab centre section is 2 hr epoxied over it's full chord (about 18") to 1/2" stab seats on the underside. These seats are locked in between birch ply formers at either end, as well as being  bonded full length to a cyparis stringer on the underside edge. The 'lug' is also epoxied, and bolted to the centre 1/4" birch former with an M6 st/st bolt, washers & a nyloc (best check it to make sure it won't come loose ! )...............add to this the fact that the vertical fin is an integral part of the fuz, for it to go anywhere, it'd have to rip it's way through ALOT of structure.

The joiner is 1" diameter, so almost the same as on the SisT 190 WING..............don't think you need to worry.

Phil

...............under crackers.................tail wheel assembly almost complete, I've asked for it to be shipped ASAP so I can get it installed before moving onto the wings in the next 2/3 weeks............

saw this at the club night last night, when |Phil and andrew kindly came to give us a bit of a chat and brought along the Skyraider and B26...
'Kinell!! we are talking serious workmanship there

A splendid night and thanks again Phill, I reckon we were all fairly gobsmacked by it by the end. Fantastic sight to behold in the flesh, as it were

Where did all those pies go??

Cheers Graham...................hope the rest of the club found it interesting.................

.....as for the 'pie & peas'........it was a 1st for me being a 'southern softie' originally, my 1st 'pie & peas' supper.................and most enjoyable it was too !.................

Phil

Fin post is next...................this is made in exactly the same way of on the B26..............hard 1/2" balsa, capped either edge with 1/16" ply.............very light, but immensely rigid. With the mid rib cut and correctly positioned off the post, the 1/16" ply core to the rudder was used to get the correct 'sweep back' to the post. (the core will eventually be heavily fretted out, NOT solid  )



Whilst being careful to maintain a warp free fin, the remaining structure was added.................1/4" lite ply ribs and false LE. 1/4" sq Cyparis spar, with 1/16" balsa webbing......................you can also see 2 of the 3 rudder hinge hangers from 1/14" ply.



One complete side of the rear fuz/fin was then sheeted over in 1/8" balsa................this had to be done in smaller sections due to the compound curvature of the structure in this area.



The other side was left open to allow internal inspection my my inspector.

Phil thats one Class build
Very impressive 

Shane

Next major task is to get all of the servos installed into the back end...................but on a model such as this, you can't 'guess' what you'll actually need, a little more 'science' and thought has to go into the decisions.

So...............how do you calculate how much torque (kg/cm) is actually required for each control surface ?.

1st, we need to know a few basic figures.......these being the 'area' (sq inch's) of the surface, it's average 'chord' (inches) and also the estimated maximum speed of the flying model (mph). These figures are then put through the following formula, using the following graph to work out the 'Torque Coefficient'.

Required Torque = Torque Coefficient (divided by 1,000,000) x area x chord x speed (squared)



The Torque Coefficient is work out using the maximum required deflection of the control surface. I went for 30 degrees on elevator, and 45 degrees of rudder, with a maximum airspeed of 100 mph.

Elev = 78 / 1,000,000 = 7.8 power 5

Rudder = 98 / 1,000,000 = 9.8 power 5
 
Taking into account a mechanical advantage on the elevators of 1.5 to cater for the 45 degree movement of the servo, and only 30 on the surface (45/30 = 1.5), the elevators require 472 oz/inch of torque, converting to around 33kg/cm.

So, 2 x 15kg JR 8511 digitals on each elevator 1/2, plus the use of servo tabs will be ample torque, especially being that a full 30 degree's deflection should never be needed at 100 mph !
 
The rudder on the other hand, with a surface area of 361 sq/inch requires a MASSIVE 3272 oz/inch of torque, converting to 233 kg/cm    (no mechanical advantage between surface and servo) to achieve it's full deflection. However, being that max rudder deflection should never be required in the air (please don't tell me you plan on doing full power, knife edge passes Ali  ), only on the ground, I've cut the airspeed number in 1/2 to 50mph............., this gives a new requirement of a much more reasonable 58 kg/cm..............so, fitting 3 x 15kg digi's, again with a servo tab, (and possibly a lead static balance in the LE should be a workable solution.

I have passed all this data onto my inspector, and I await an answer before I go ahead and start cutting hatches and installing servo mounts.

Regarding hinges..................as mentioned, the rudder is hung of 3 substantial 1/4" ply hangers, so I'll be using M3 high tensile cap head bolts in brass tube bushes as the pivots (as per the C47 & b26 rudder hinges)................but for the elevators, I'm considering using these MONSTER 'hinge point'. From a German suppier, they are moulded in 30% glass filled nylon (not just 'plastic' like Robart hinge points), but more importantly are 1/4" in diameter, and just over 100mm long................5 of 6 of these could be used per elevator half.

i think you had better pass on Alis details too :ww

amazing figures..

Cracking build Grommit

Great stuff

Ian.

Do you have to take into account using surfaces heading into wind.  ie the overall speed is increased.  Would you not be better using 4 of those servos?  I know it seems extreme but loosing control of a model this size i would imagine makes the bum cheeks tweek.  Give Mr Nicols a ring he'll know the feeling 

Do you have to take into account using surfaces heading into wind.  ie the overall speed is increased. 

AAARRRGGGHHH!!

NO IT ISN'T!

PDR

depends if the runways moving.

depends if the runways moving.

GO TO YOUR ROOM



PDR

Regarding hinges..................as mentioned, the rudder is hung of 3 substantial 1/4" ply hangers, so I'll be using M3 high tensile cap head bolts in brass tube bushes as the pivots (as per the C47 & b26 rudder hinges)................but for the elevators, I'm considering using these MONSTER 'hinge point'. From a German suppier, they are moulded in 30% glass filled nylon (not just 'plastic' like Robart hinge points), but more importantly are 1/4" in diameter, and just over 100mm long................5 of 6 of these could be used per elevator half.

Phil,

Have you had a look at these monsters,



You can get them from Darryl at Sierra Precision, here http://www.sierraprecision.com/prod14.html


Mark.


Edit.....   if someone can tell me why thi imge wont display, you'd make me really happy

Sure Mark 

Your link points at the webpage not the image.
http://www.rcmf.co.uk/gallery/displayimage.php?album=lastup&cat=12&pos=0

Here is the image path.
http://www.rcmf.co.uk/gallery/albums/userpics/11549/Hinges.jpg

So...

  I'd forgotten all about those !...................me thinks they are a MUCH better alternative to the glass filled ones..............thanks for the reminder mark.

AAARRRGGGHHH!!

NO IT ISN'T!

PDR

If you have a measured distance on the ground and you time how long it takes for a plane to fly between the two you can obviously work out its effective ground speed ie 50mph.  If the path of the plane was directly into a 10mph wind then its overall airspeed has to be 60mph.  If i'm wrong i don't understand at all.  Just to be clear i understood and agreed with your point of view AKA the moving runway.

If you have a measured distance on the ground and you time how long it takes for a plane to fly between the two you can obviously work out its effective ground speed ie 50mph.  If the path of the plane was directly into a 10mph wind then its overall airspeed has to be 60mph.  If i'm wrong i don't understand at all.  Just to be clear i understood and agreed with your point of view AKA the moving runway.

We're discussing control surface loadings. If the model is flying at 50mph at a given throttle setting in still air the control surface loadings will be "x". If it does the same at the same throttle setting in a 10mph wind the control surface loadings will STILL be "x"; the airspeed will still be 50mph but the groundspeed will drop by 10mph.

At any given power setting the model will achieve the same airspeed regardless of wind strength - if you look through the various equations which describe aircraft performance you'll find there isn't even a symbol to denote "wind speed". Wind is only of interest to aeronautics in so far as it affects navigation and for the special case of wind-shear. In all other respects it is completely irrelevant.

I believe there might be a point Phil's missed here, but I'll deal with that by PM.

PDR

Steady Pete......steady.....

Count to 10 slowly..........there, that's better isn't it.......

It really makes you wonder sometimes....doesn't it? 

Andy

We're discussing control surface loadings. If the model is flying at 50mph at a given throttle setting in still air the control surface loadings will be "x". If it does the same at the same throttle setting in a 10mph wind the control surface loadings will STILL be "x"; the airspeed will still be 50mph but the groundspeed will drop by 10mph.

At any given power setting the model will achieve the same airspeed regardless of wind strength - if you look through the various equations which describe aircraft performance you'll find there isn't even a symbol to denote "wind speed". Wind is only of interest to aeronautics in so far as it affects navigation and for the special case of wind-shear. In all other respects it is completely irrelevant.

I believe there might be a point Phil's missed here, but I'll deal with that by PM.

PDR

Thats where the confusion stemmed for me.  Phil didn't specify whether he was using air speed or ground speed.  My mistake

you know, if he tries DSing it, the ground speed could suddenly become the ground PLUS! air :ww

Having had a reply from my inspector, his reply was as follows.............

As to servo' power, I advocate 15-20 kg.cm/sqft for ailerons, elevator and rudder. Twice that for flaps. Your elevator servo's are OK (30 deg's at 100mph would tear the tail off!!). For rudder, with 2.5 sqft area, two of your 15kg.cm' servo's would be more than enough for the above reason and also that part of the area is balance tab which counteracts the load.

I would not add counterweights; the theory is correct but our model structures are too flexible and unwanted oscillations occur which destroy the gearboxes of even the most robust of servo's (my Beau' and JT's Magister suffered from this and we eventually removed the mass balancing altogether).



So.............we're happy there.

BUT.............following a PM from our resident 'PDR' earlier on, I have totally neglected to consider the effect of 'prop wash' (or it's speed, and the forces it exerts) over the tail control surfaces.

The proposed prop for the 400cc radial is a 39 x 25 carbon, or 'wind shovel' as pete has appropriately named it. This, it will spin up to a maximum rpm of 3,800, so as you can imagine, it's ganna get a wee tad drafty back there 

I have calculated that a total of 63% of the rudders area is within the prop wash, and 54% of the elevators.......................so, once my brain is working properly again in the morning, I'll re-do the torque calculations using the prop pitch airspeed figure, rather than my 'best guess' of 100 mph (airspeed  )

Back soon...............

Phil

You'll have to rethink the ganged Futaba 148's then Phil

Stunning as usual, sure is an experience watching the build progress.

Ian.

Great build Phil  Interesting to see a bit of hard figures in the analyse of the dynamic loads for a change (usually a lot of gutfeel & guesstimate ) From where does that graph & formula originate, full scale?

  Johan

Great build Phil  Interesting to see a bit of hard figures in the analyse of the dynamic loads for a change (usually a lot of gutfeel & guesstimate ) From where does that graph & formula originate, full scale?
 

It looks like someone has done a "generic" hinge moment calculation. I have a slight concern that it should really be accompanied by a set of assumptions and the limits on which it is based, but then I went off and did some sums based on the data Phil published here and it comes so close to that curve as to make no difference.

PDR

It looks like someone has done a "generic" hinge moment calculation. I have a slight concern that it should really be accompanied by a set of assumptions and the limits on which it is based, but then I went off and did some sums based on the data Phil published here and it comes so close to that curve as to make no difference.

Pete, Phil,

How about an article in the Masterclass section...... 'An idiots guide to calculating servo torque'

We would all have a chance to get it right then......I hope I'm not asking too much, but I'm sure it would be well appreciated.....

I think, I could have eventually got there, based on the airspeed, angles and area of control surface etc, but I too, would have missed the effect of the propwash.......

Cheers

Andy

The 'article' was given to me by my chum John 'B24' Deacon...............and apparently came from Model Airplane news many years ago.................thats all I know. He tell's me he's been using it for years on his models, as well as those he's an over 20kg inspector for, and has yet to see a control surface/servo related problem.

Pete kindly PM'd me again this morning after doing the 'speed' calcs for me...............propwash speed off the 39 x 25 at 3,800 is 95 mph...............dropping to maybe 75/80 10ft rearward, over the fin & stab.
It also shows that the flat out, level speed of the model should be around the 80/90 mph range, so for a 30% SPAD, thats about right, the fullsize was a bit of a lumbering old beast at the best of time.

As mentioned earlier, trying to achieve a full 45 degrees of movement at full speed will most likely induce a catastrophic structural failure, so it's not even worth considering. Dropping the airspeed to 50mph dropped the requirement to 58kg/cm.....remember that'll still give a full 45 degree's of deflection.

The chances of this being needed are slim................a lot of movement may be needed on the ground to taxi etc, but this will be at low speeds, with engine at a high (ish) idle, plus we have a massive steerable tail wheel doing a lot of the work as well. The raider has a massive, flat fuselage side, so rudder input to fly Ali's favourite 'farnborough pass' shouldn't be huge either, even though it'll be at high airspeed. All of this says to me that getting as close as possible to the 50/60kg mark is plenty.

Now.................do I do that with multiple digitals.................VERY expensive, as well as being heavier on wire, and involving a more complex wiring loom, and the use of 'matchbox's' to get all servos moving identically, or do I consider using just 1 of these.............................

Tonegawa Seiko SP-050

A precision electro-mechanical servo motor for Radio Control applications with requirements exceeding those of recreational hobby models...the "Muscle" servo for RPV,UAV, & UVS projects.Industrial and special effects robotics...ruggedly designed and built for hostile environments and performance-critical applications.



SPECIFICATION

Operating Voltage :- Vmin-max: dedicated 4.8-12.0 VDC, not shared w/receiver

Recommended operating voltage :- Vin: 6.0-8.4 VDC

Starting torque ;- Ms 65.33 Kg/cm (907 oz/in) @ 6.0 VDC

                             91.50 Kg/cm (1271 oz/in) @ 8.4 VDC

Speed :- 0.29 Sec/60 deg (no load) at 6v - 0.21 sec/60 deg (no load) at 8.4V

Weight :- 280 grs (10 oz)

Size :- Length :- 100mm (3.94") including mounting flange

                        75mm (2.95") body only

           Width :-  44mm (1.73")

           Height:-   93mm (3.66") w/output arm installed

                         79mm (3.10") body only

The flap's are definitely going to require at least 1 of these per surface, as their area is bigger than the rudder, and full deployment will be needed, all be it at relatively low speeds...............the tail wheel steering is also in need, as having destroyed an 18kg sail winch servo on the C47 in under 25 flights, it looks like an ideal candidate to replace that, and be used in the raider.

If you have a look at Weidner Modellbau...............they spec them as well on pretty much every one of their monster size aerobatic models...............and knowing the germans, they don't do things by 1/2's, and don't use 'rubbish'.................they appear to come pretty well recommended...............

http://www.weimo.net/html/h2.htm

They are simple to wire up, the signal being taken from the Rx as normal, but the power from a separate high voltage power system. This can be either 6v or 8.4v if more torque is required, and can be run on heavier duty wire, via a high current battery backer to give some fail safe redundancy. I'm going to need such a syatem on the flaps regardless, so adding in 1 more PS-050 isn't too much of an issue.

 8)

 Those should do the trick..... 8)

For ease of use, I suggest you look at this

www.weatronic.com

Mine is on order and will be with me at the end of the month. (for the Pitts)

Does everything you need for a large model including the equivalent of a matchbox on every channel, and also gets over the 20kg twin RX rule. Plus it has so many other features, have a look.

Might solve some radio problems for this and your other large models.

Steve

rather than my 'best guess' of 100 mph (airspeed  )
Phil

Cheers for that Phil

Now.................do I do that with multiple digitals.................VERY expensive, as well as being heavier on wire, and involving a more complex wiring loom, and the use of 'matchbox's' to get all servos moving identically, or do I consider using just 1 of these.............................

Only problem i can see is putting all your eggs in one basket.  Less likely to have a servo failure and its much simpler which is good but if the one servo were to fail you would have total loss of that surface

For ease of use, I suggest you look at this

www.weatronic.com

Mine is on order and will be with me at the end of the month. (for the Pitts)

Does everything you need for a large model including the equivalent of a matchbox on every channel, and also gets over the 20kg twin RX rule. Plus it has so many other features, have a look.

Might solve some radio problems for this and your other large models.

Steve

Impressive...............498 Euros (don't really need Gyro & GPS) isn't bad at all when you consider the cost of 2 PCM Rx's and something like a Powerbox/RX Buddy....it's VERY reasonable. One unit with just power and servos plugged in would be good....................the more connections that can be done away with the better.

Which system have you done with Steve ??

Impressive...............498 Euros (don't really need Gyro & GPS) isn't bad at all when you consider the cost of 2 PCM Rx's and something like a Powerbox/RX Buddy....it's VERY reasonable. One unit with just power and servos plugged in would be good....................the more connections that can be done away with the better.

Which system have you done with Steve ??

The 10-20R - 10 Channel, 20 servo, no Gyro, no GPS, don't need it for the areobatic models I fly.

Looking forward to it, i'll let you know how I get on.

Well worth a look at though.

Steve

Pete, Phil,

How about an article in the Masterclass section...... 'An idiots guide to calculating servo torque'

We would all have a chance to get it right then......I hope I'm not asking too much, but I'm sure it would be well appreciated.....

I think, I could have eventually got there, based on the airspeed, angles and area of control surface etc, but I too, would have missed the effect of the propwash.......

Hmmm...unfortunately it isn't quite that simple. Hinge moment calcs aren't a simple matter of pumping numbers into a formula; you have to either know or be able to take a good stab at the characteristics of the local pressure field. The curve Phil has will work reaosnably well for a small range of sizes, sections and speeds, with the assumption that there's no abrupt turbulent break from the hinge line and a few other things. I'd suggest using that curve and then de-rating the answer (ie multiply the answer by about 150% to get a torque requirement). This is for two reasons. Firstly it's to allow for the approximations in the method, and secondly because the torque quoted for a model servo is usually the STALL torque - if subjected to this load the servo won't blow back, but it won't move the surface either.

PDR

The 10-20R - 10 Channel, 20 servo, no Gyro, no GPS, don't need it for the aerobatic models I fly.

Looking forward to it, I'll let you know how I get on.

Well worth a look at though.

Steve

My feeling exactly...................didn't really see the need for a Gyro or the GPS in 'our' type of models. After reading all the blurb, the 10-20R appears to be an excellent unit...............I'd be most interested to hear what you think once you've got it up and running as it also looks like an ideal candidate for the B26 as it'd make for a much lighter set up as only 2 battery packs are required compared to the Rx Buddy's 4, and I'm sure this new unit weighs a lot less than 2 Rx's and a buddy added together........................please keep me posted.

Going back to the servo situation, the more I think about it, the better the Tonegawa set up looks.

Yes, 1 servo per surface means that 'if' that servo were to fail, then that surface is lost. However, being of 'industrial' origin, my experience of such components (eg:- SMC pneumatic systems & air rams etc....) has shown far superior reliability, and ease of operation over our 'model' products. If these servos are manufactured to UAV and 'robotics' uses, then I'd imagine those environments are going to be pretty harsh compared to a 10 minute flight in a model aeroplane. We'd be using 2 x JR digis to there maximum performance, a single Tonegawa wouldn't be pushed so hard.

The problems I see associated with adding multiple servos to a surface are these.............

1) If 1 servo fails, does the 2nd have the power to drag the dead servo around as well as continue to operate the surface ?
2) It's heavier, due to the 2nd (and even 3rd) set of wiring (heavy duty, silicone coated wiring which is V heavy), associated servo mounts & structure, multiple electrical connections, multiple control horns & linkages.
3) It takes much longer to built, set up and get operating reliably
4) The industrial route is actually cheaper....................the Tonegawa servos are around £140 each..........the JR 8511 are over £80 each...................plus 1 Matchbox will be required per surface to tune the multiple servos into each other............so another £40+ per surface.

Further discussion with Ali and my inspector needed on this one before any decisions are made.................

Back to the build......................

With the completed stab installed on it's tube, I thought I'd give it a bit of a shove about to make sure all was well, and there was no excessive flexing and horrible crunching noises..................ohh dear  :-[

Pushing backward on the LE caused a small gap (0.5 mm approx) to open up at the LE of the joint between outer and centre section. Due to the fact the joiner tube, and the securing lug are both rearward of 50% root chord.........I guess I should have seen this one coming. Anyway............easy fix, a second set of 1/4" ply retaining lugs were set into the stab and centre section to hold the LE closed.................



The wiring was also installed for 2 servos in each stab 1/2, and a possibility of a further 2 for the rudder. The wire used is a heavy duty. silicone coated wire more suited to the high current demands of these big digital or industrial servos.

Once that was done, it was a simple matter of completing the sheeting of the fin, fitting and shaping the LE all the way from the fuz break up to the tip block, and adding and shaping the balsa tip block with it's 1/32" ply core.....................



A final fettle of the 2 fuz sections is next before they can both be glassed in prep for a short time in storage whilst the wings are started. I'll be fitting the elevator servos, mounts & wiring to the stabs this week, followed by glassing them. The cowl will also be completed with a high density pink foam nose ring, again before glassing. By the time this is done, I should have the engine from Ali so I can then crack on with the engine, tanks, throttle & choke servo installations....................

Phil,

Seperate power leads to servos - a REAL pet hate of mine! You only need one large poitive lead and one large negative lead. The bigger leads are both more robust and lighter for a given current rating because it lakes less insulation to cover one big wire than a dozen little ones.

So how about running one pair of LARGE wires from the radio bay area to the tail to provide power "busbars". At the back end fit a socket block with servo connectors (like a receiver socket block). Then just run a set of signal wires (one per function or one per servo depending on whether you're worried about this wire failing) in a seperate bundle to be connected to the individual servo connectors at the socket block. I'm sure someone like Ashley would be able to get you the servo socket block. If you wanted to you could take spurs off this busbar to other sockets in other places to feed (say) the canopy opening servo etc.

I presume you've decided against a full triple-bus electrical system as we discussed a while back...

PDR

Pete.

Using the modern 'Powerbox' or 'Weatronic' type power & Rx system, these are designed with multiple outputs per channel, so the manufacturer intends you running say 2, 3 or 4 separate servos, with their leads from the surface, right through to the 'box'..............so the power busbar is contained within the box, rather than at various points around the model.

I have a few mates who have used this busbar method in big models before, but both have now ripped it out and replaced it with one of the modern systems........................

I have know idea why this change has happened, as I partially agree that the busbar system is a perfectly usable and reliable system, it's just not the way in which the current breed of gear is designed to be run.

Cheers.

Phil

As previously decided, a pair of JR 8511's is to be housed in the underside of each stab half.............not sticking out in the breeze as on many 'Mega' models..............but fully enclosed under flush fitting hatches.

1st job is to get the servos mounted, so this was done with these rather neat little machined aluminium brackets that I now have available from stock (these fit JR & Hi-Tec, I have Futaba ones on the way as the hole spacings are different)



The ply mount plates were than packed to the correct height & angle to sit inside the 4 servo wells, and then securely bonded to the 1/4 lite ply false TE, the 1/4 cyparis stringer between spar & TE, as well as the top skin as well................this gives a foot print of around 4" x 1.5", and is also attached to 2 main structural members, so transferring servo loads well into the stab structure.



Servo leads were then extanded (cut and lengthened, rather than plugs & sockets) before being installed under 2mm lite ply hatches. The hatches are secured with M2 x 6mm countersunk screws into 1/4 lite ply corner blocks, with snake inner in the holes.................nice & tidy.



So as all 4 servo operate in the same direction, all 4 output shafts point in the same direction, but to enable all 4 sets of linkages to attach to the elevator horns in exactly the same location on both sides, the 2 servos in the port stab needed to be offset slightly compared to the 2 in the stbd stab................once glassed, painted & detailed however, the positions of the servos shouldn't be noticeable.

Talking of horns.................I now have the CNC milled 1/16" Carbon horns needed for the build..............28 of them in total.



Each linkage required 2 horns to allow me to sandwich the ball link (rod end) in the middle to avoid any flexing of the horn. Each surface, 7 in all, has 2 servos, so 14 servos, 2 horns per servo..................28 horns. Yes I could have produced them by hand, but the time, and number of bandsaw blades needed to cut the stuff simply isn't worth it, so these were custom cut by Weidnor Modellbau in Germany, and arrived in less than a week 

Next job is rudder and elevators...............this I expect to take quite a bit of time as they are massive surfaces with a large number of hinges to line up to obtain free movement, as well as the multiple control horns to line up etc.................so, before I set to and get them underway, being it was a nice day, it had it's 1st outside excursion all in 1 piece...........

I name this ship...............





Back soon.................

Phil

WOW!