Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by Victor Bravo, Sep 16, 2019.
I emailed her asking if the presentation was available anywhere.
I don't mean to sound sarcastic but do you have two cars and a tree. Block the metal with one car and push with other. If the metal let's go it's just a scratch.
As a matter of fact, it is. Jump to 21:30 for the landing gear portion.
I would probably make that 15-30" radius bend with my backhoe bucket. The bucket has about that radius, so just clamp one end to the boom stick and the other end to the bucket teeth and then rotate the bucket. Done.
More thanks T. I owe you a drink if we ever meet.
I can't believe she just hard bolted through the legs for the fuselage attach!
I've just been reading through the composite gear leg threads here. I know Billski prefers steel, but I think you can get a composite leg to work satisfactorily. I laid out my nose leg with 8 plies of uni glass near the OML. All cloth was too soft, but the outer unis give it nice flex- about the same deflection as the Ti leg . The strains look fine for a 4G load.
Why, that's mighty kind of you sir! (especially since I'm the ESA YouTube channel manager, and it took me about fifteen seconds of "work" to get this for you.)
But hey, since you're buying... Oban 14, neat, please.
I won't make an uneducated comment about the wisdom of that, but I will say that it's working for her airplane, and for enough time to have shown issues, I would think.
The main gear legs on the RAF VariEZE and LongEZ were both composite. They originally had some "creep" issues in warm climates, but I believe the "cure" was to paint them with a UV barrier, then black, put them in a basic fixture and put them out in the sun for a post-cure heat-up. Then paint them white. Someone more knowledgeable can confirm or deny that.
This weekend Steve Smith sat down with Robbie Grove and chatted about how to do gear like this in carbon fiber. Bottom line, it's doable, but it doesn't like sharp-ish corners where bending loads the outside fibers in compression and the inside in tension. The main issue is that the combination of forces creates a separating force that pulls the inside fibers inward and pushes the outside fibers outward. To mitigate this effect, you need annular fibers around the bundle to help resist delamination. That's not a deal killer, but it does complicate things. The solution that suggests itself is a couple or three (or maybe more) plies of carbon sleeve around the unidirectional fibers to help resist delamination.
I think glass is better for these legs, but you pay your money and you take your choice. As I mentioned above, I like having some type of through thickness reinforcement for gear legs even if there are lots of them that don't have it. I'd prefer a 3D woven part, but that's big$$. Tufting or stitching are also viable, especially if you're dealing with an RTM preform.
I plan to have a straight, flat, horizontal center section, and large radius "quarter-circle" arch bends starting at the side of the fuselage and going out/down to a small flat section at the bottom on each side to bolt the axle onto. The airframe attch system is clamps that are co-located with the wing strut attach, riveted to the corner of the fuselage/longerons. Rubber pads cushion the steel fitting, prevent galvanic corrosion,a nd allow a small amount of compliance. The stock design gear differs only in that it has small radius bends yielding short straight "gear leg" sections as you had posted.
Why not just use the Grove radius blocks? I'd be worried about the rubber creeping or setting up a resonance.
You need to do a sanity stress check on whether the longerons are up to the load. I had to locally beef up my 1x.065 4130 longerons. There are pictures online of people (Pitts? Great Lakes?) who didn't beef them up and had them crack. If I can find the links, I'll post them. Decathlons have beefy local reinforcements.
I do not know what level of cold work will make cracks grow in 7075-T6, so I would get all conservative and work to use as large a radius as I could fit, which means make the straight sections at the leg mounts as short as will work. The positive camber (unloaded) shown on Grove's stuff is usually preferred so as to have the tires straight up when the loaded airplane is rolling along the ground. You probably already know that, but I wanted that out there for others to see.
Good luck and report back so we can learn from you.
Well, you can all learn a delightful lesson from me right now, about not having one's head up one's empennage, resulting in not being able to see a major design flaw ahead of time.
I had a talk tonight with a highly experienced aerostructures guy. I sent him the sketch I posted here. He pointed out that if I bent the "leaf spring" into a rolled curve, then it wouldn't behave like a leaf spring anymore, because the curved shape would not "flap" up and down evenly like a straight section would.The curve would stiffen itself in a manner of speaking, forcing the majority of the load and flexure into a much smaller section of the "spring", which would possibly overload (or permanently deform or crack) the material that otherwise would have acted like a diving board and lasted a long time.
So although the large radius curve would probably solve the fabrication stress problem, it would immediately create an operational stress problem no less disastrous.
So my clever solution to this problem was not going to work, and I don't need anybody to calculate any more heavy mathematics to help design the landing gear this way.
OR, in the words of Gilda Radner back in the funny days of Saturday Night Live.....
Uh.......Uh....... there are chapters in classic stress books (such as Roark, ch. 8) that address the curved beam you have.
Uh................Uh.................and then there are the many arched gear legs in service.
Uh.......Uh......You try to taper the width (or thickness) of the leg to get a nice constant working along the leg in order the get the lightest leg. But you usually end up with a stress concentration at the inboard radius. I've done hand calcs (using the classic stress books) and made many FEA runs on straight, 'small' radius and large radius gear legs. Please have your advisor cite some references, because otherwise I missed something big time.
With a tapered width leg, you don't get the same deformed shape and stress distribution you get with a constant width 'diving' board.
uh......uh......Billski, we need you. Please report for expert engineer duty as soon as you wake up!
I concur with flyboy 2160. Arch gear legs in any form are workable. Seriously, what is the difference between two small curved areas and one big one on each side?
As for your advisor, I do not think I would pay any more attention to his judgement on structures...
Going further, arched landing gear do work. Now they may be a bit more effort to optimize, but they do work just fine. The familiar straight sections with small radii bends are generally lighter at the same energy absorption, and so for bent and then heat treated metal they are generally preferred. If I were building with materials and processes that preclude small radii bends, I would build a continuous arch as pictured. Prior to yesterday, that meant laminated Graphlite rod gear legs, where we had to keep the curvature down (radii large), but I can now see a reason to design in aluminum the same way.
See George Sychrovsky's site: http://www.curedcomposites.com/gear.html
I am going to go out on a limb here. Victor Bravo, contact me via private message...
The 1/4 bend gear will work just fine. Your friend is trying to tell you that the ideal taper will not be straight, but a paper pattern, a bandsaw and belt sander will be fine. Some of these theoretical types can get carried away sometimes. (Whistles innocently).
You can also get the equivalent of Graphflite uni carbon rods in uni glass if you need more flexibility.
Well, I am not taking any delight over this...
What it sounds like is this "highly experienced aerostructures guy" was saying was that your gear will be less than optimal. He obviously did not understand you and most likely knows little about what works in Wittman landing gear systems and their derivatives. Making the entire leg operate at high stress and allowing it to pivot about the rocker plates will reduce the amount of weight you must lift aloft to do the job.
There are a number of methods for optimizing this spring. Some folks working in constant thickness springs will tell you that you simply do a straight taper from the tread patch to the mount. Even if all of the load entering the spring were vertical and the leg stayed put (assumption of small deflections), this is not optimal and understrength. Include the fact that you splay the gear in landing which increases the arm between tire and mount, that there is drag force from tire spin up that both bends the leg aft and puts significant twist into the leg, and then there is some shear load from the tread contact and calculate the total stress state in the critical locations at each cross section. Yeah, it is sophisticated about maximizing the springiness you can get for a certain airplane and amount of material.
When I did my steel legs, I wrote a program in Excel using what is in the old Part 23 as summarized in Pazmany's landing gear book, then included all of the mechanics of materials issues with the leg, put in a gear leg shape profile and tapered shape, calculated forces and stresses in leg at frequent intervals along it, and let it iterate until it gets the deflected shape at the peak load during the landing cycle. Then I tailored the width of the cross section at each station, and checked if I had enough travel to suck up my max landing without exceeding my max g's and without hitting other parts of the airframe with the wheel, tires, brakes or legs. I ended up with a 3 g landing gear that has straight legs when sitting on the ground, slightly curved legs when the gear is hanging, and if I do thump it into the ground at a max FAR Part 23 vertical speed, everything will be OK except my bruised ego - you know it will make the best five light plane thumper landings video someplace.
My approach to constant thickness springs is a little different. 1st I decide on the undeflected and deflected shapes. The deflected shape is created from the undeflected one by giving the entire spring a certain strain. That is usually yield. So deflected is the hardest you can smack it down and have it not bend. From that, I can determine the thickness. Now, I calculate the moment along the whole spring for the deflected state. From that, I can get the width.
After all that, it needs to be checked for all load cases. Side, braking, spin up etc.
Repeat until you have the minimum spring that will take your calculated loads. The maths is not particularly hard, but it is tedious.
Billski, I get enough delight from your comments here for both of us LOL
RE Forming 1.00-thick 7075.
As a materials engineer there is likely only one-way to form relatively tight/safe bend in 7075. Here is how I would specify a BR 5.00 [estimated from of my charts for thicker plate]...
Per AMS-H-6088 [was MIL-H-6088] or AMS2770 [wrought aluminum HT processes]...
Solution heat treat 7075-O [or -T6] 1.0-thick strip-of plate @860F-to-930F [time as specified for thickness].
Shock-quench-cool the SHT plate to room temperature by 'quickly' immersing/holding it in cold-water-glycol tank [until it's at least 140F].
Remove the quenched stock from the tank and dry it off. Apply forming lubricant as needed.
Approximately 10-to 30-minutes after quenching/lubing form all bends while in the 'as-quenched' ['AQ' or 'W'] unstable temper [not quite dead-soft].
After forming, then begin 'age heat treatment' to attain the -T62 temper** 240F-to-260F for ~24-Hrs.
for 7075 I greatly recommend -T73 or -T76 temper over -T62 temper for grossly improved stress corrosion cracking and exfoliation corrosion resistance... especially in this 'thick plate'. These tempers would require different 'aging HT' protocols.
Separate names with a comma.