3 Piece Box Spar Wing Modification For Single Seater SSDR

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

buckie555

Member
Joined
Oct 21, 2019
Messages
10
I'm in the process of building a wood and fabric single seater of an established design that currently has a 1 piece cantilevered wing of box spar contruction. I am keen to modify this to be a 3 piece construction with a fixed centre section and two outboard sections, each approximately 2.5m long.

I realise that I will need to do a full structural analysis of my proposed modifications, however for now I would just like to get a rough ballpark idea of how heavy these modifications are likely to be. I am building to quite tight weight requirements and want to check that it's in the realms of being feasible for my airframe weight budget.

I have looked at a few different established 3 piece wooden spar designs such as the Chilton DW1 and the Falconar F11 however they have slightly different spar construction to mine. Namely solid plank spar and split caps.

My box spar is constructed of full, constant width, caps top and bottom that taper in thickness (height) and faced on either side with a ply shear web.
To give a sense of scale the spar is approximately 75mm constant width and where I envisage the joins, the spar cap thicknesses are approximately 35mm and 30mm top and bottom respectively with a spar cross section height of 150mm.

From what I've seen externally, I believe the Colibri MB2 3 piece wing was a similar construction but I could be mistaken.

My initial thoughts are for a fairly standard arrangement of vertical straps made of 4130 or 2124, reinforcement blocks, reinforcement ply doublers, etc for both the main box spar and, to a lesser extent, the rear drag spar.

Has anyone got drawings of the MB2 wing join arrangement or for that matter any other wooden aircraft of similar spar construction to above. I have an idea of how to proceed but I would like to see a proven example before I start doing structural calcs and FEA etc.

Any guidance, insight, gotchas or pointers to resources would be very much appreciated.

Thanks, Neil
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,933
Location
KBJC
I would like to see a proven example before I start doing structural calcs
Take a look at KR-2s. More images available from the google.
1623120839484.jpeg
I suspect your caps will need to grow in depth near the joint to accommodate the bolts (whether a single row per cap, or a double row as seen above). Your weight penalty will be the extra wood, the plates, and hardware. Perhaps also whatever scheme you cook up for the skin closure as well. Not a terrible price to pay if you don’t have space to fabricate a single, long spar.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,853
Location
Rocky Mountains
Thatone
Good example.
Keep in mind when doing the calculations for the wood/bolt bearing load that you can't use the total area of contact like we tend to do with thin ridged metal bolted joints. Long skinny bolts tend to bend and the wood crushes/deforms more closer to the metal strap.
Hardwood doublers may be worth the weight in this area?
Phenolic inserts for the bolt holes is another option that has been used by others.
 

fly2kads

Well-Known Member
HBA Supporter
Joined
Jan 2, 2010
Messages
1,808
Location
Justin, TX
The Taylor JT-1 Monoplane and the Turner T-40 are two other examples that I know of that use a similar type of joiner between the center section and the outer panels. The Mono uses a box spar, and the steel straps have two rows of bolts and lightening holes. The T-40 has a laminated plank spar, and the steel straps have a single row of larger bolts. The T-40 fittings would be a little faster to make. Neither one uses the vertical section of the steel fitting pictured above.
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
Those examples were exactly what I was looking for, thank you. That KR-2s solution is the first one I’ve seen where the upper and lower tie bars are joined together. I like the simplicity and the assuredness of alignment between the two it provides. I take your point Hot Wings regarding the area. I can see that the relatively soft spruce within the straps sandwich would provide for lower bolt preload than would otherwise be the case, so we can’t bank on the shear on the strap faces taking the load but instead the shear within the array of pins (bolts). My initial thought was that the pins elongating their bores and deforming the caps could be the limiting factor. I see there’s a trade-off between bolt size and number to balance their weight, minimum distance to the cap edges and gaining enough cross-sectional area to take the shear and not deform the bores. I like the idea of inserts. Perhaps ash cylinders or maybe even aluminium.
 

cvairwerks

Well-Known Member
Joined
May 12, 2010
Messages
284
Location
North Texas
Ignore the mess in the warehouse, but here's a couple of photos of the spar joint for my Fairchild AT-21. The one photo is of the inboard, lower joint on the left side forward spar, while the second is the outboard forward spar. I don't know if the outboard is an upper or lower section. The joint for the rear spars are identical, but smaller. The forward straps are about 27 inches long, and the spacing at the joint bolt is about 8 inches. The forward spars have an upper and lower set at each joint and the rear spars likewise. There are some steel tubing splices that go with the bolts between the wing sections, but there are boxed and buried for now.
Forward.jpgOutboard.jpg
 

cvairwerks

Well-Known Member
Joined
May 12, 2010
Messages
284
Location
North Texas
Tim, same style, with the exception that there is a built up "I" section that bridges the front bolts and the back bolts, that resides in the gap between the spar ends. As to the bushings, same style, but I'm not sure if they are magnesium or aluminum.... Lots of corrosion on them and the bolts will not press out. Somewhere, I've got the drawings for the PT bushings and I'm hoping the AT drawing is on the microfilm. Making them is going to be a job for a cnc lathe with a bar feeder or an old style screw machine!
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
OK so I've made some progress on this. I've done a spar stress analysis based on the Schrenk approximation. Verifed my calcs against a known good example and am confident that the spar loads are reasonable. From there I've done a stress analysis of proposed wing attach fittings. The dimensions of the proposed WAFs look to be in the same ballpark as for a couple of similar existing designs. For each scheme I've determined the ultimate axial stress in the strap fitting, the ultimate shear stress in the strap mounting bolts, the ultimate shear stress in the strap joining bolts and the ultimate strap bolt bearing load. One solution is 4130 straps, AN bolts and aluminium bushes to reduce the bolt bearing stress. That works out to approximately 10lbs for the set. My calcs assume that all the bolt loads are taken purely in shear and so the frictional effect of bolt preload is neglected. So then I thought, well if I'm going to use aluminium bushes around the strap mounting bolts why not just use aluminium bolts instead. Picture dowels machined from 7075 with external threads on both ends. The threads begin outboard of their intersection with the straps so they are a snug fit within the wooden cap strip and joiner straps. They're secured both ends with washers, bolts and an appropriate locking mechanism. If I also make the straps from 7075 or 2024 then the weight comes down to about 6lbs.

With this scheme the axial stress in the straps is < 23,000 psi, ult shear stress in the aluminium dowels is < 23,000 psi, and the ult strap bolt bearing stress is as per allowed in ANC-18 (i.e 4700psi limiting factor of upper spar cap in compression x L/D factor x ply reinforcement factor, etc). I used 23,000psi based on the quoted fatigue strength for the alu. The joining bolts would still be AN. My thoughts are:

Am I being overly conservative / not conservative enough using the value of 23,000psi?
Will those modest stresses allow me to sleep comfortably or should I just stick with the more conventional 4130 straps and AN bolts?
Is there some other glaring deficiency in this scheme that I've overlooked?

4lbs weight saving may not seem like much but it is significant in an aircraft with an empty weight of only 450lbs.

Any thoughts, comments, criticisms would be very much welcome. It's been an interesting learning experience so far.
 
Last edited:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,049
Location
Saline Michigan
In my not so humble opinion, not conservative enough. While we use FOS Of 1.5 in metal structures, folks generally use considerably higher FOS of in fasteners.

Any yielding in a bolt is a failed joint - preload is lost, joint goes sloppy and then fatigue is rapid. A wise designer will protect bolts and rivets from yield at ultimate airplane g's using appropriate FOS.

Shear yield is 0.577 times tensile yield, 75000*0.577 = 43000 psi.

I am traveling and can not get at my texts on this to check, but you should do so. Maybe someone will chime in...

Several reasons for higher FOS:
  • Rivets and bolts are loaded with stress concentrations, so they fatigue more readily than other stuff;
  • They have to be preloaded to prevent fatigue, and that uses up some of their available strength in all directions,
  • Failure modes for bolts and rivets tend to be severe and hard to detect ahead, so you want to distance yourself from these failures.
In addition, our AN bolts and almost all commercial bolts and nuts have rolled threads, which are stronger and more fatigue resistant than cut threads, which is what you most likely would use in a custom fastener.

Please do some checking into airplane structure texts on bolted joint design and typical FOS.

Billski
 

AdrianS

Well-Known Member
Joined
Jul 5, 2014
Messages
674
Location
Australia
What diameter are the pins?

Could you use gun-drilled 4130? That should be lighter than a solid bolt.
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
Billski,

I take your points regarding the fasteners and I certainly want there to be a considerable safety margin.
In my example that equates to:

43000/23000 = 1.87 Safety Factor @ ultimate (compression failure at upper surface of spar assuming ultimate compressive strength of spruce @ 4700psi
43000/23000 = 2.5 Safety Factor @ limit load (elastic fibre yield at upper surface of spar assuming proportional limit compressive strength of spruce @ 3530psi

I thought those would be sufficient but perhaps not.

The 4130 and AN bolt option looks roughly in line with a couple of other proven examples that I looked at. I suppose I am particularly concerned with how much to de-rate the calcs for the aluminium's pins succeptibility to fatigue compared to the aforementioned.

AdrianS,

That's an option. Out of 7075 the pins would be 10mm dia x 5 off per strap pair.
If using AN bolts it's AN3 x 12 off or AN4 x 8 off per strap pair.

The aluminium option is an interesting one but perhaps this is one of those areas where I err on the side of caution, accept the weight and go with the steel.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,853
Location
Rocky Mountains
how much to de-rate the calcs for the aluminium's pins succeptibility to fatigue compared to the aforementioned.
<< >>
accept the weight and go with the steel.
I'm scared of fatigue failure. To me it's the same kind of fear I have of land mines - you just can't tell if the field has been fully cleared. Unless there is a real need just don't go there.
<< >>
My particular interest is in part 103s. Every time I do this trade study I end up with steel straps and lots of little bolts in the wood...........and accept the weigh penalty (if any).

A lot of my thoughts on this are just semi-irrational fear based.....................which is a poor substitute for science. ;)
edit: Not intended to be a political observation.
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
I'm scared of fatigue failure. To me it's the same kind of fear I have of land mines - you just can't tell if the field has been fully cleared. Unless there is a real need just don't go there.
<< >>
My particular interest is in part 103s. Every time I do this trade study I end up with steel straps and lots of little bolts in the wood...........and accept the weigh penalty (if any).

A lot of my thoughts on this are just semi-irrational fear based.....................which is a poor substitute for science. ;)
edit: Not intended to be a political observation.
I hear you and it scares me too! What started off as a "I wonder why I can't do this" has led me down an interesting rat hole but it's probably best to revert to a more conventional solution.
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
Ok so Billski's comments prompted me to re-visit my spreadsheet with the approximate values for a KR2 main spar as a sanity check.

Note that these are based on some simplifications, not Schrenk, not using MOI, etc and apologies for any errors or omissions.
Some values are approximate but should be close enough for a sanity check?

Assuming that we want the WAF to be at least as strong as the spar itself.
Assume limiting factor will be compression failure in upper spar cap.
Upper spar cap 2-5/32" x 2".
Assume spar web approx 0.375" wide.
Centre to centre spacing 5-3/16"
Assuming cap will fail in compression at 4700psi.
Cap will therefore carry 2-5/32 x 2 x 4700 = 20,269 lbs at failure.
Load acts on a moment arm of 5-3/16 / 2 = 2.59" to neutral axis
Max bending moment = 20,269 x 2.59 = 52,497 inlbs
Assuming semi eliptical load distribution, total moment outboard of the wing attach fittings is approx 0.58 x max bending moment = 30,357 inlbs
Acting over same moment arm gives 30,357 / 2.59 = 11,720 lbs compression in top cap
WAF strap is 8" x 1.5" x 1/8" 4130.
The top cap uses 2 pairs of two straps. Each pair is joined by a single AN 3/8" bolt.
Each pair of straps sandwiches the spar cap and is secured by AN3 x 8 off bolts.
Including the ply shear webs the length/dia ratio of the strap bolts is approx 14. From ANC-18 giving a wood crush strength de-rating of 0.44.
Therefore bolt bearing crush strength = 4700 x 0.44 = 2,068 psi.
Ultimate axial stress in each strap fitting = (11720/2) / ((1.5-0.375)*0.125) = 41670 psi (90,000) => SF = 2.16
Ultimate shear stress in wing join bolts = (11720/2) / ((pi()*((0.375/2)^2))) = 53297 psi (75,000) => SF = 1.41
Ultimate shear stress in strap bolts = (11720/8) / ((pi()*((0.1875/2)^2))) = 53297 psi (75,000) => SF = 1.41
Ultimate strap bolt bearing load = 4700 * 0.44 * (0.1875 * (2.156 + 0.375)) * 8 = 10359 lbs which is almost 11,720 => almost all the load can be taken as shear

Assumes Shear stress for an AN bolt 75,000 psi.
(note: the /2's are due to the load being shared by the two straps)

Unless I've messed up my calcs, which is entirely possible :) , the safety factors for this known proven application are what guided me rightly or wrongly towards my choices for the 7075 option albeit with the uncertainty regarding the fatigue values.

I would have thought that the above actually overstates the axial load somewhat as it assumes an equal max load across the depth of the upper cap when actually it's only at max load at the top surface of the cap and tapers off towards the neutral point. That and the fact that the bolts are not in the plane of the upper surface but are closer to the neutral axis and therefore also see a lower load.

My understanding from ANC-18, is that the joint should be sized so that there is sufficient strength to take the loads purely in shear and any contribution from the strap shear friction due to bolt preload being an additional safety margin. I understand the neccessity of ensuring a non-slip joint in a metal / metal application to prevent fretting and fatigue but don't have practical experience of how easy that is to achieve with a metal / ply / spruce / ply / metal joint.
Maybe it's much easier to achieve than I envisage it? Don't want to damage the wood by crushing!

I'd be grateful if anyone can point out any mistakes or gross error assumptions that I've made.

Thanks,

Neil
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,049
Location
Saline Michigan
Hmm. I was not talking about the wood to strap calcs, only that your aluminum bolts joining the center section to the outer panel were not conservative enough. FOS was a little above 2, and if memory serves, you need 4 or more. And shear yield is 0.577 ti

All of the rest of the calcs - was not even addressing them.

Billski
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
Hmm. I was not talking about the wood to strap calcs, only that your aluminum bolts joining the center section to the outer panel were not conservative enough. FOS was a little above 2, and if memory serves, you need 4 or more. And shear yield is 0.577 ti

All of the rest of the calcs - was not even addressing them.

Billski
Ah maybe there’s been a misunderstanding then. I probably should have made it clearer in my earlier post. For the aluminium based design, I was only proposing that the straps and the strap mounting bolts (that mount the strap to the wooden spar cap) be aluminium, the joining bolts that bolt the centre and outboard straps together were always going to be AN bolts.
 

Heliano

Well-Known Member
Joined
Dec 24, 2015
Messages
140
Location
Campinas, SP, Brazil
One quick reminder (may you are aware of that): be careful when considering using AN bolts with aluminum bushes (sleeves): galvanic corrosion. AN bolts are cadmium protected, but even so I would avoid such arrangement. Compression failure stress of 4700PSI for stika spruce is realistic, but I would use an extra FOS for wood properties vary quite a lot. One key aspect of this type of joint is wood bearing strength. One can not analyze a single bolt separately, because there is a combined effect of the multiple bolts, and this effect depends on bolt separation and arrangement - if they are inline or in a zig-zag pattern, for example. If yow are willing to be extra conservative, build a partial spar and do a destructive test (in that case my condolences - it is a lot of work).
 

buckie555

Member
Joined
Oct 21, 2019
Messages
10
I'm scared of fatigue failure. To me it's the same kind of fear I have of land mines - you just can't tell if the field has been fully cleared. Unless there is a real need just don't go there.
<< >>
My particular interest is in part 103s. Every time I do this trade study I end up with steel straps and lots of little bolts in the wood...........and accept the weigh penalty (if any).

A lot of my thoughts on this are just semi-irrational fear based.....................which is a poor substitute for science. ;)
edit: Not intended to be a political observation.
In your designs have you used any hardwood or alu bushes around the strap mounting bolts or did you just go for bare AN bolts through the capstrips?
 
Top