Main Spar Design

[Radge]

I've just about finished my latest little project and as I seem to have developed a system of having 2 or 3 on the go at the same time(glue's drying on one so work on the other) I'm deciding what I fancy doing next and although I'm experimenting with as many different techniques as possible, I'm really reluctant to stray from spruce top and bottom spars with a shear web, it really does seem to make tapering quite an easy affair, however, I've noticed a lot of older plans and even the latest Corby Starlet plan use the notched single spar method which I would be tempted to alter to top, bottom and web.So to aid my decision on the selection of my next balsa bash could I beg on your experiences and ask you to indicate the pro's, cons, strengths and weaknesses of these methods and , or any other suggestions please.

Thank you all kindly

Radge

[kc]

The notched spar seems to work as some of the great designers used it ( Boddington etc ) but it seems likely to weaken the wing especially if not cut acurately. I have avoided it and always used spruce spars top and bottom and webs on all recent models. Stick with what you know works!

[Mark Powell 2]

Spruce top and bottom, balsa web, vertically grained. Every time. A notched single spar breaks every rule known, regardless of who designed it. .

If you are fussy (and I have done this just now for the first time), taper the spruce spars from root to tip. Ideally the taper should be 'horizontal' (ie: looked at from above it tapers). Most of the work of a spar is done by the outer surface. Use 3/8 X 1/8 laid flat, tapering to (say) 3/16 x 1/8 at the tip rather than a constant 1/4 square. Less fussy, use constant 3/8 x 1/8.

If you are really nuts, taper it as a 'hollow' parabola, but that is going to extremes.

Even with dihedral, build the spar first, including, if there is one, the flat centre section, as one piece. Then build the wings around it. Better and simpler than joining the spar afterwards. you can use accurate scarfe joints easily, for a start. Don't let the cat jump onto the half of the spar that is initially floating in the air.

[Erfolg]

For small models the notched spar probably works, at least when done reasonably well.

I guess it can make assembly easier, a sort of egg box type packing, quite strong, in many instances.

However as the stress levels go up, your spar type structure comes into its own.

If optimising bending and shear stresses are the objective and ultimate handling performance is your objective, then tapering is worth while. However, if dodgey landings (tip), easy build and you are not chasing fractions of a gram and the ultimate, aileron roll rate, and you are not Gorilla launching a glider, why bother.

Stick with what you are happy with.

Edited By Erfolg on 01/09/2012 16:19:56

[Bob Cotsford]

The notched spar (egg-box) construction is fine on small spans, though if the rib slots are not a snug fit the wing will end up like a banana. It makes for a quick and easy build though and if the leading edge is sheeted it can stop or at least limit the banana effect. A compromise is to use the slotted spar as the web and add a thin flat spar top and bottom. It's best used on thick wing sections either way.

[Radge]

Thank you KC, Mark Erfolg and Bob. I was right in asking this question as you have all responded in a marvellous, informative and explanitive manner. This, as I thought as much but through inexperience was unsure, has confirmed that I shall stick with spruce, top, bottom and web. Now then Mark, please elucidate on the cat situation. Lol. Thank you all again.

Bless

Radge

[Simon Chaddock]

Not quite sure why Mark suggests tapering a spar 'looking from above' and keeping the thickness constant.

The most efficient use of material would be to taper the thickness and keep the width constant as it puts the most material at the greatest depth.

Like this full size Mosquito front spar.

Note the top and bottom spars locally thickens to pick up the engine mounts. The complete spar is faced both sides with ply as a shear web.

I would entirely agree that such structural detail is not really required on models as they can achieve more than sufficient strength to weight with much simpler structures.

[Mark Powell 2]

Simon,

Because I read that the two spars should be as far apart as possible (which makes sense), so using 3/8 x 1/8 rather than 1/4 square is good, and tapering accurately my way, rather than ending with 3/8 x 1/16 at the tip  and 3/8 x 3/32 in the middle is a lot easier. Steel girders always have the outside laid flat.

Edited By Mark Powell 2 on 02/09/2012 17:12:10

Edited By Mark Powell 2 on 02/09/2012 17:15:51

[Mark Powell 2]

Simon,

Further, you say 'sufficient strength'. I crash models quite often. People have commented both on my not brilliant piloting (I have somewhat 'monocular' vision, thought it does not seem to bother me anywhere else) and how my models resist crashes. Yet they weigh no more than anyone else's.

BTW - cutting lightening holes everywhere, as Chris Golds does in his otherwise excellent plans, saves amazing little weight. I even weighed these 'holes' on my 7 lb EDF CG Swift. Three quarters of an ounce

[Erfolg]

As has been said that the principle stresses are concentrated in outer fibres, according to elastic bending theory. Then again there are a lot of provisos, such as the beam is considered to be made of lamella, that are free to move, relative to each other, the material is isotopic (that is identical properties throughout, which wood certainly is not) and a few more, which do not immediately come to mind.

Yet in principle, the further apart the better.

Of course our friends Timoshenko and Euler all had thoughts on beams.

[kc]

Mark suggested replacing 1/4 sq with 3/8 by 1/8.  The cross sectional area of 3/8 by 1/8 is less than that of 1/4 square surely? 1/2 inch by 1/8 would be the same area as 1/4 square.

But actually 1/4 sq is much better than 1/2 by 1/8 because there is more area to glue the web onto the edge of the 1/4 sq ( assuming web is glued on the outer edge rather than sandwiched between spars. Between spars needs much greater accuracy in construction and therefore likely to be weaker )  

Wings are not just subjected to loads like a beam but have to cope with cartwheels, hanger damage etc   So I suggest 1/4 sq spruce for .40 size models and 3/16 sq spruce for smaller lighter models. Forget tapering!

Edited By kc on 02/09/2012 19:34:54

[Peter Miller]

Just a thought i=on tapered spars.

I have in the past laminated the spars from 1/4" X 1/8" The layers at the root out to 1/3, two layers out to 2/3 and a single layer from then out. You get the strength where the loads are for the same weight.

The change in lamination must be tapered and you can vary the wood sizes to suit the size of model

[Mark Powell 2]

kc,

The 3/8 x 1/8 laid flat is more structurally efficient, tapered or not. My reduction in cross sectional area is deliberate. Weight saved already, even without tapering.

Glueing spar webs on the side relies much more on the glue. it also uses, in the case of two 1/4 square spars, 1/2 inch more wood. Heavier.

'I recommend....forget tapering'.

We are looking for 'better', not just 'adequate'. Cartwheels - they are made to fly, not to crash   Heavier planes crash harder than light ones, that is why small planes are more crash resistant than big ones. Heavy extremities, such as tips, produce more strain on central parts, such as wing joints.

Peter,

That is a good way too and easier than trying to get a straight taper with one piece. Though I have never done it. Yet.

None of this matters on 'day to day' planes (probably why they are day to day), but on 'difficult' subjects like EDFs, weight saving, and large gliders, structural efficiency, can make a big difference. if carried out everywhere.

Also weight saving at the extremities increases manoeverabilty.

Edited By Mark Powell 2 on 03/09/2012 04:55:32

Edited By Mark Powell 2 on 03/09/2012 04:55:56

Edited By Mark Powell 2 on 03/09/2012 05:03:26

[Mark Powell 2]

It is not just spars.....

Centre wing joint. If, as is often done, there is a 1/8 ply (say) wing joiner, discard it. Instead, put a 1/16 one at the front of the spar and another at the back. Make the end of the front one two ribs out, and the rear one four ribs out. This avoids a suddent stress concentration. Note the rear is longer than the front. In a crash the wing tries to move forwards, but is secured at the centre. More stress on the rear of the spar than the front. Double up the rear joiner for a couple of inches. Use no other than Araldite 'Precision' 24 hour epoxy for the spar/ply joint. All other epoxies go like toffee after a year or so and some, Devcon especially, are brittle right from the start. Always taper the spar, less weight at the tips, less inertia, 'distributed' strength, more manoevrability. Use thinner ribs as you move towards the tips, 1/8, 3/32, 1/16. Taper the TE width.

I have done all this on an 80 inch high aspect ratio (that's efficient too) 'hotliner'. It is the fastest thing on the field. stays up for more than an hour on a good day, with only 6 or 7 minutes of motor, rolls faster than the eye can see, and you can pull full up elevator after a 180 mph (measured) vertical dive without anything breaking, and it is ten years old. The wings are an open structure, with a D-box LE, tapered spruce spars on the surface, web between the spars not glued to the sides, etc., covered with clear yellow Profilm. The flaperon joint is sealed.

Edited By Mark Powell 2 on 03/09/2012 06:06:41

[Erfolg]

I have no issues with a little theory being put into the practise of model building. As KC has also indicated there are a lot of issues which affect the design.

Engineers use relationships such as the Universal Bending Formula for calcs to check a design, having recognised a number of facts. They will have noted that model aircraft wings, particularly large gliders bend in a Parabola, where as the UBF assumes the beam bends in a constant radius.

Engineers will also note that the shear web actually does something, it keeps the extreme fibres at a constant distance apart. So they will be loath to remove sections, without a compensation, or checking if the compression member could fail in crippling.

All the bits and pieces we attach to the spar affect how it bends, the load is often also partially torsional.

Engineers will have noted that the beam is not a UDL (uniformly distributed Load) Nor a point load except where struts are concerned. The load varies from tip to root, again for a number of possible reasons. As we know the twisting force also varies with speed. The wing structure will also be liable to oscillations at some critical speed. All of this points to the need for a lot of judgement, rather than first reaching for the calculator.

Most will satisfy themselves that the root section is OK, they will ensure that the spars are as near the surface as is practically possible.

If they are really looking for the best performance they would almost certainly look at other materials such as Carbon Fibre for spars etc.

Yet most practical engineers will design what they think is adequate, a few calcs on the critical areas, and then trust in good practical/theoretical practises. Often coming to the conclusion, for many purposes spars top and bottom, no tapering, with a shear web is as good as anything, for a model.

Certainly no notches, looking askance at breaks in shear webs etc.

If you want to go to town on Theory consider both Timosheko and Euler. both devised methods for treating beams, other than the simple