wing clamping plates - dimensions

[JimCrawford]

Hi All,

I would like to know the dimensions of the wing clamping plates that are used on the Nipper fleet. Mine (GAVTC - Slingsby Mk3) are 19mm wide but I've seen wider ones, although I cannot remember where or when. They could be up to ~30mm wide before they foul the canopy frame.
I'd appreciate the information if anybody happens to have a ruler handy when they next attend their aircraft. Only the width is needed, the length is, of course, the spar chord.

[ben]

Hi Jim

I have a factory one in front of me. It is .75" or 19mm wide. I can send you a used one to copy if you wish but I think Paul Grellier has them in stock. I am sure you are aware they are chamfered to "square off" the spar angle.

Regards

Ben Faulkner
G-NIPS

[JimCrawford]

Thanks Ben,

I would guess that 19mm is the stock size. I'm not trying to make one, I'd just get one from Paul, but this whole business of the bolt inspection or replacement prompted me, just out of curiosity, to do some sums. As of now I haven't managed to reconcile the numbers and, as a time served physicist, that irritates me. Either the numbers or the sums are wrong. I'm in the process of tidying up my notes to send to Francis and I'll copy them onto the forum but right now I feel as if I've kicked over a can of worms and I'm looking for a bigger can to put them in.
I don't mind going down in flames if I've misunderstood the way the assembly works or made some fundamental error in my calculations, but it is worth the risk to understand.

Jim
GAVTC

[Tipsy Flyer]

IMG_2262.jpg (31.86 KiB) Viewed 11660 times

Hi Jim,
My contribution !

Regards
Glen

[G-ARBG]

Hi Jim,

I acquired these clamps amongst a lot of spares from Raymond Cuypers some years ago but never felt inclined to use them having never seen anything other than the standard 19mm type used on a Nipper. I do not recognise the part number 990003 and the word 'specials' after JALB is unusual for an aircraft part ? However I thought you may be interested to see them.

regards
David GARBG

[G-ARBG]

Jim,

Dimensions of clamps shown in my photo 35mm wide x 12mm depth.
R.
David GARBG

[JimCrawford]

Thanks for the numbers - very interesting. I hope to post the sums this evening after I've managed to get no.1 grand- daughter to bed. A rather hyper young lady.

Jim

[JimCrawford]

This note describes the investigation I made into the wing bolt installation for the Tipsy Nipper aircraft, my curiosity having been sparked by the recent requirement to have the bolts tested or replaced by purpose made studs. I hope it fits into the user group format and apologise if parts get rather like Fortran, which shows my age!

Head above parapet, tin hat on!

Imperial units are mostly used throughout as this is the format used in the majority of references.
The strength characteristics of AN and A25 bolts are regarded as equivalent, [ref 1]. I understand, but have no confirmation, that the stud material is of a higher tensile strength than that of the bolts. I have used the LAA value of 6500 lbs for the ultimate tensile strength of an AN5 bolt [ref 2], although 6710 lbs is given elsewhere [ref 3].

My aircraft operating documents show weights of 660lbs semi-aerobatic and 725lbs normal category and a maximum G loading of 5 at aerobatic weight. My permit to fly shows weights of 685lbs and 750lbs respectively and aerobatic manouvers limited to 4G loading, although there is a mysterious limit of 2G on steep turns.
The Nipper was originally available as a certified aircraft so I assume it was designed to BCAR section K which defines semi-aerobatic and normal category limits as 4.4 and 3.8 respectively. The negative loads are not relevant to this analysis.

To avoid presenting a messy spreadsheet with all the permutations, I’ve used the section K limits at the higher weights and of these the 685lbs / 4.4G is the worse case.

1. First the easy stuff:

Aircraft max weight 685lbs
Wing weight (bit of a guess) 80lbs
G loading 4.4

Weight supported by bolts; 4.4*(685-80) = 2662.0 lbs
(we don’t have to work out the bending relief along the spar, because we are only interested in the bolts)

Factored for tailplane download; 2662.0*1.05 = 2795.1lbs [ref 4]

Assuming symettric loading, the load on each bolt; 2795.1*0.5 = 1397.6lbs
Proof tensile strength; 6500/1.5 = 4333.3lbs
I’ve not made any adjustment for the shear at the spar/fuselage junction as the maximum shear in a terminal velocity dive will be half the aircraft weight, 342.5lbs, and isn’t enough to make a significant difference on the tension/shear interaction curves.

2. So far so good.

The tension in each bolt will be increased by the preload generated by the torque used on installation. Using the standard formula [ref 5],

F= T/(K.d)
Where
F is the tension (lbs)
T is the torque (lb.in)
K is the fudge (sorry nut friction) factor.
d is the bolt diameter, 0.312 in.

setting the torque to the middle of the ops manual range – 120lb.in
and the K factor to 0.13 – middle of range [ref 6]

Then the bolt preload is; 120/(0.13*0.312) = 2958.6lbs
And the total bolt load; 4356.1lbs

So the 4356lb tension in the bolt at section K semi-aerobatic load of 4.4G and aircraft weight of 685lbs is comparable with the proof load of 4333lb.

There are obvious issues with the wide range of assembly torque allowed and the rather uncertainty of the K factor which, in various references, can vary from 0.1 to 0.2 but there is another parameter to check.

3. The plot thickens;

The bolt transfers the lift and preload tensions via the spar clamp. This has an area of 3.8sqin
so the pressure at the clamp/spar interface is 4356/3.8 = 1146.3 lb/sqin The only reference I’ve found for the proof load on sitka spruce across the grain is 600lb/sqin [ref 7] rather than 740lbs/sqin ultimate which is more usually quoted..So the load which is satisfactory for the bolt is nearly twice that which would cause a crush failure in the top surface of the spar. These are the numbers I’ve been unable to reconcile. Either I’ve misunderstood the assembly, incorrectly applied the numbers or used incorrect parameters. It would be interesting to know, since all Nippers must have had their bolts removed for the recent AD, how many exibited signs of the clamp digging into the spar.

To achieve a clamp pressure of 600psi I had to make two changes;
1. Increase the width of the clamp by 50%. This could be achieved by fitting a 30mm wide plate, say 5mm thick, under the standard clamp. No wider otherwise the plate could foul the canopy frame at the forward edge.
2. Set the preload torque to 80lbs.

This does not include any allowance for the variation in K, which could typically be +-20%.
It does however, by lowering the preload to 1972.4lbs, give a reserve factor of 4333/(1397.6+1972.4) = 1.3 for the bolt.
The 35mm wide clamps described by Dave GARBG would give a reserve factor of 1.17 for the spar.

Make of this what you will.

Jim

Ref 1 Royal Aircraft Establishment
Technical note Structures 248 September 1958
Comparative Strength tests of tension bolts with UNF and BSG threads

Ref 2 LAA note TL 1.16 issue 6 page 25

Ref 3 Metallic Materials and Elements for Aerospace Vehicle Structures.
MIL-HDBK-5E, Department of Defense, June 1987.
TABLE 8.1.5(b l ). Ultimate Tensile Strength of Threaded Steel Fasteners

Ref 4 LAA note TL1.15 issue 2 page 2

Ref 5 Fastener Technical Reference Guide

Ref 6 Fastener Technical Reference Guide

Ref 7 Technical Report FPL-GTR-190
Chapter 5
David E Kretschmann
Mechanical Properties of Wood


A couple of postscript items to ponder;

Whatever the width of the clamp there is a sharp disconuity of crush force at the edge. Is there any scope for using a sacrificial ply plate under the clamp such that the edge ‘damage’ is contained within the plate and the forces transferred more smoothly into the spar.

The assembly drawing calls up two full size plain nuts and appears to show a split pin drilled through the top (outer) nut. The stud is supplied with a full nut and a castellated nut. The ops manual emphasises fitting the split pin. Using a split pin allows only 30 degree increments in nut rotation and will result in uncertainty in the torque which has been shown to be quite critical. I would prefer to use a half nut locknut with a mild grade of thread lock combined with a tell tale paint marker.


Jim

[JimCrawford]

An update on the story so far:
This issue has bounced backwards and forwards between me and John Tempest at the LAA, although I've had nothing since my last email on the 10th of January.
The situation as it stands now is that I corrected my calculations with the result that the situation is eased somewhat, but still significantly out of kilter re the crush load on the spar. I'm not sure how I can present this on the forum, maybe I can turn the spreadsheet and the graphs into jpg screenshots and load them as pictures.
In words the correction is to allow crush relief as the wing load increases. To illustrate this imagine the aircraft on the ground, this is the baseline data point. Now fly the aircraft and apply a 3G load. This load is reacted against the head of the bolt which will stretch by ~ 2 thou. so the spar expands elastically by the same amount and, in doing so, relaxes some of the preload crush.
As the flying load increases the load on the bolt increases and it's reserve factor decreases, but the crush on the spar surface decreases and the reserve factor increases.
Extrapolating the calculations to the ridiculous, the bolt is good to about 9G and the spar lifts off the top longeron at about 20G (!) Note this is only relevant to the attachment bolt joint, not any part of the rest of the airframe. I expect that many other interesting effects would be seen well before then, probably the engine departing the airframe, the wings clapping tips or the tailplane collapsing.
I'm investigating whether the crush limit can be taken as that of the ply capping which is likely to be higher than that of the spruce booms.
I'll attempt to post a graphical illustration of the spreadsheet results and meanwhile hope to get some reply from the LAA.

Best Wishes

Jim