Eagler's Nest
General Category => Off Topics and General Interest => Topic started by: Murray Randall on November 15, 2014, 06:02:54 PM
-
Here's pic of my Mod XL aileron installed with hinge pins installed in the wing. The pic shows that I did not install the 1/4 X 1/4 diagonals that act like drag/antidrug struts in the ailerons and maybe you can see that I modified the trailing edge piece. That trailing edge has quite a bit more glue area and is lighter than the plans design. Yeah its more work also. An aileron weighs 2.3 lb varnished. I was running a little trial test. There is 10 lbs of steel blocks on the aileron outboard tip which takes about 20 lbs force at the aileron actuator arm on the inboard end. And translates into a 8.4 lb stick force with my linkage. I wanted to get a feel of any hinge binding due to aileron deflections causing misalignment and get a look at aileron deflections and torsional stiffness. I might go back and add more weight. The wings varnished with all fittings come in at 23 lb. My question here is. Do any of you think that omitting the aileron diagonals is bad for your health? thanks Murray Randall
-
More than anything, I think you will have problems with the structure when you go to shrink the covering. I wouldn't intentionally delete any of the structure of the wing.....
But I am not an aeronautical engineer......
-
I would suspect one reason for having diagonals is to prevent the shape of the aileron from deforming enough to cause it to bind in the pocket. When deflected to full travel there is quite a bit of airload--is it enough to cause it to bow if the diagonals aren't present??
Wanna be a test pilot? ;)
-
Here's my thinking at the moment. Loading the aileron outboard tip is a harsh test condition. In service the aileron load is distributed both span and cord wise. An 8.4 lb load at the stick sounds high and that's one aileron. The opposite aileron is generating load also and feeding that load to the stick thru the cable system while one aileron is generating 8.4 stick load the opp aileron is contributing. Something quite a bit less than 17 lb total, but up there somewhere. I'd say the loading as I tested is worst case. Fabric shrinking will tend to scallop the trailing edge if carried to extremes, but if done uniformly will not twist the aileron. But fabric shrinking beyond wrinkle free is of questionable merit. Once the fabric weave is filled with the low compressibility "paint" the stiffness of the system increases by many factors. The process mechanism and results are analogous to applying resin to fiberglass cloth. My plan is to cover the aileron and run the little test again. Please reply with any suggestions, thank you. Murray
-
Putting all those diagonals in takes a big bunch of time though Sam
-
I haven't started my build yet (still need to finish the tailwind), But I plan to include everything in the plans. As I understand it, the wings were designed by a structural/aeronautical engineer. I don't know about the ailerons, but I'd guess they were also. I doubt there is anything in the plans that are not essential to the structural integrity of the aircraft (in the spirit of 'keep it light'). The challenge is to not add stuff that's not in the plans!
-
Looking at the picture...
I see a D tube at the front of the aileron which is responsible for carrying the torsional load as imposed by the test. The diagonals would do little to change or increase the load capability that I can see.
In structures remember that triangles are always a better and stronger structure than rectangles. By removing the diagonal braces you have left a rectangular structure which will be weak in jacking loads. Do you need diagonal braces in all bays of the aileron for this? IMHO having braces in one aileron bay would likely suffice. Not to mention any shear strength provided by the fabric covering.
It appears you have also removed a number of aileron ribs? This is probably the riskier change here as it provides less support for covering and the trailing edge. Not just potentially a scalloping problem when shrinking covering but a deflection problem from air pressure. And that could be a gateway to flutter?
The structures designed into an airplane have the benefit of having been tried and tested. They are of course not the only way to do it, just the way with known characteristics. Any change deviates from that and imparts some level of risk of potential failure. This risk will have to be managed in some way.
-
Oh would I love to see any analysis and testing of the Eagle wing. (Yes I have seen the 500 lb load test documentation.) I do not say that with any malice or doubt at all. No doubt about the integrity of the wing. I have done extensive finite element structural analysis of the Eagle XL and my modifications to the fuse and am totally satisfied that the XL fuse as designed is satisfactory. That fuse analysis was certainly not necessary. It just gave me confidence in my modifications. As a registered mechanical engineer and Eagle builder I would love to see the engineering on the wing. I would like to see any analysis of the fuse also but my own analysis of my fuse leaves me satisfied. The finite element programs that I use do not handle the beam and plate interactions of the Eagle wing to my satisfaction and I would simply like to see any work in that area done by others. But it is not necessary to treat an airplane design as a clean sheet of paper design. Components of Eagle are derived from existing designs with a satisfactory history. This is entirely satisfactory. It just a personal professional interest in the numbers, the technology that interests me. That's what happens when you mix shop rat genes with engineering genes. Murray
-
Proven safe design why skimp on safe Dirt is hard
-
Proven safe design why skimp on safe Dirt is hard
Responses like this are really not constructive... Nobody said anything about skimping on safe and there are many ways of building capable structures. If we never ask the questions we never learn and would never advance the state of the art. Much better to point out the weakness of a proposal or ways to reduce/manage the risk so we can all learn from it.
Would seem to me that if the D tube at the front of the aileron were stiff enough there would be little but extra mass to gain from diagonal members in every bay of the aileron. I picture a round of foam... 3 channels 120* apart... a layer of light carbon fabric wrapped around the foam tube... some carbon tow or uni run down the channels... another layer of light carbon fabric around the tube... vacuum bag... then remove foam. One very stiff and lightweight aileron front spar. Be even better if one of the carbon fabric laminations were laid up with a 45* bias.
Might be interesting to build one aileron per plans and another per improved concept. Then test each for torsion and bending loads to see how they perform.
-
Its good to see a response to the topic recognizing that there are no end of safe light methods of building Eagle components. My revision is tested with a proof load considerably in excess of operating loads. Isn't that sufficient to insure safe operation? I did neglect to mention that I put additional nose ribs in the aileron D nose in the first two inboard bays to increase the torsional stiffness. If you do not want to test or analyze please build to plans! I communicate off site with other builders making alterations. I would like to see these out of the box folks putting their ideas up on the site, but its not productive getting a generic not to the plans slam.
-
Murray,
Your torsion test was good but I would add another test for confidence. In pondering this, the aileron is supported at each end and thus subject to "droop" in the middle. The diagonals will integrate the LE and TE into a single plane which would be very stiff in the droop direction (think increasing hinge gap in the middle). So with the aileron hanging down from each hinge end how much will it droop with a weight applied in the middle? This is the one direction I could see it being weaker without the diagonal members.
As to your prior comments about finite element load analysis... I think that is in a way what all of us have been doing for years. We look at the structure and use our instincts and understanding to conceptually model loads and forces. We look for weaknesses like non-triangulated structure in particular loading planes for force directions. The benefit of the computational solution is the iterative capability to try lots more variations and thus not miss anything.
-
Murray, I like your process of approaching every part with the question of how it can be made better or lighter without compromising safety. I also like that you are willing to post your ideas. You also have calculations to back them up. It is experimentation like this that will improve any design.
My own experience in building the aileron was that there wasn't much glue area holding the diagonals. I first used Titebond III, but found that every time I bumped the aileron when moving it, a diagonal would pop loose. I redid the joints with T88 which seems mor tolerant of sub-optimal joints. No more diagonals poping off. I can't really think of a condition in flight that would stress the diagonals that way. Your test seems adequate to me.
-
I would agree with Doc 100%- on Murray’s work. As long as it is tested, has good theory of build behind it, and leans towards positive, it is all good.
We all know, how the Eagle, XL, and many many others come about… they start with an improvement upon a previous design. Maybe, a different product, maybe a different build way, generally a combination of both. And every once in a while, a brand NEW idea comes out. Though this is very rare and can be counted on one hand.
the real short list of totally NEW ideas- Jim Markse, Ken Rand (Rutan-follows the Rand way, with his design, one year later) as well as Glasair. and so it goes. I don’t know where I would put Pipstrel, but certainly they are doing builds that produce what no one else has for performance but not new, just perhaps better design. Again the list is short.
Now you toss out the list for “improvements” or new build ways, and the list suddenly gets HUGE!
Best of success to all and keep the minds open!
-
Tite Bond III quotes you a strength of 4,000 psi. While T88 quotes 7,000 psi. I like to think most woods are close to 10,000 psi. These are just napkin level values but I'm not surprised to see that Doc sees better results on those aileron diagonals using T88. Those numbers might suggest that it is maybe appropriate to use gusset/corner blocks here and there. It would be no big deal to do a simple comparison twix Tite B and T88. But why not just use T88. Of course I still say leave those diags out. Murray
-
I'm not sure it is really a difference in glue strength I was seeing. Because of the way the diagonals fit, there is no good way to use a gusset or corner block, so the glue area is very small and at least partially on end grain. The T88 fills gaps better and because of its viscosity, can create a bit of its own "gusset". I found the Titebond III convenient when the joints are nice and tight with lots of area. Also Titebond prevents one from trying to stretch the last bit of a batch of epoxy. Nothing worse than a joint with too little epoxy that is already starting to set up.
The more I think about it, the more I think the fabric will do the work of the diagonals.
-
Tightbond when applied correctly creates a bond stronger than the wood itself. If you glued two pieces of together with the appropriate clamping force and then tried to break the wood apart, you would find the wood split and not the glue joint. But this works best when the grains are glued in parallel. If you were to glue the end grains together, it would barely hold. This might be the case with the diagonals in Doc's ailerons. T-88 has no such requirement; it's just a powerful epoxy.
To Murray:
Material Properties of Sitka Spruce: (http://www.matweb.com/search/datasheet.aspx?matguid=1e56abdf98904f2ca53bff4bd1250cab&ckck=1)The compressive yield strength of properly dried Spruce is about 5,500 psi parallel to the grain, and a little over 300 psi perpendicular to the grain. I found similar numbers from numerous sources (US/UK/Canadian).
To put this into perspective, the maximum bending stress experienced by Lenard's wing when he loaded it with 1,000 lbs of bricks was, 3,100 psi compression and 2,325 psi tension on the front spar top and bottom respectively. The rear spar top/bottom was 1,950 psi compression/ 2,040 psi tension. Normal flying loads are much less. I ran these numbers several times and always get nearly the same results.
P.S. If there are any engineers reading this, I wouldn't mind a second look at these numbers. I can explain the formulas and assumptions.
-
Vince, I think you are exactly right about my diagonals. It was a case of poor joint configuration. I looked again at how I did it today. When adding the nose plywood, I extended it about an inch over the cap strips to create a gusset at the caps and diagonals. Now it is confidence inspiring. Even more so, now that it looks like the diagonals may be unnecessary. ; )
Now if I could just get back to construction after my usual spring stall/spin...
-
.
You're quoting some uncommonly low strength values from various sources, Vince. Thats me trying to be polite which is most unnatural for me. The easy way to determine the bending strength of a a piece of wood is to clamp lightly in bench vise measure up a distance, pull it to failure with a fish scale. I did that and got 11,780 psi one direction and 8,330 psi the other. There should be pic attached. Then weigh a little piece on the digital scale. Anyone can do this with a material they see in lumber yard. I'd say the accuracy of my crude set up is 10%. If my explanation below is not clear to anyone, please pipe up. My point, if there is one, is: test YOUR wood, test YOUR glued joints .
The applicable equations for the size sample you use are:
I = b times h cubed divided by 12 I = moment of inertia in lbs/sq in b = base dim in inches h = height dim in inches
S = M times c divided by I S = stress in lbs/sq inch M = bending moment = dist from vice up to spring scale times the spring scale reading in lbs c = half the distance from the compression side to the tension side of your piece
My sample of Stika Spruce from Aircraft Spruce was .5 by .262 failed with 1.125 lb in the weaker axis and 1.5 lb in the stronger axis, applied 6 in up. The fish scale doesn't show as such in the pic
-
Damn Sorry I said my test piece failed at 1.125 lb and 1.5 lb That should 11.25 lb and 15.0 lb Murray