Marc Zeitlin

I realize that - I was just saying that if it is OK to use tapered rollers in the wheel axle, then it should be OK to use them in your assembly as well, since they're no more likely to get exposed to dirt/grime there than at the axle. It's not a question of the "style" of casting, it's a question of the geometry of the steering axis vs. the wheel axle position. On a motorcycle (or car, for that matter), if you look at the steering axis (or the kingpin angle on the car), it's ALWAYS tilted backward, meaning that the TOP of the rotation axis if further to the rear than the bottom of the rotation axis. Also, where that axis hits the ground MUST be in front of the center of the contact patch of the tire. Since you have removed the "trail" in the casting (you've put the axle directly in line with the steering axis), if you put a negative rake angle on the system, to mimic the angle of the stock gear strut, you will end up with the steering axis hitting the ground BEHIND the center of the tire contact patch. This is unstable. It's like trying to drive your car backwards at 90 mph - any tiny perturbation, and you're off in the weeds with a hard-over wheel, especially since you won't have "hard" steering connection - just the gas-springs. If you put trail back in, that will help by getting the contact patch behind the point where the steering axis hits the ground, but you still don't want the steering axis angle to be negative (bottom further aft than the top). For stability reasons, the steering axis should be tilted back at the top, as described above for motorcycles, cars, and steerable nose gear on aircraft. Yes, Rhino has the ability to do volume calcs, and knowing the density, you get weight.

1) The wheel itself already used tapered roller bearings, and they take at least as much of a beating as the rotating bearings you're looking at. Dirt and Grime is not an issue if you use sealed bearings. Why would this be a special case? 2) A backward rake angle will destabilize the system. You need rake and trail (See any motorcycle fork geometry) for stability. You used to have a "L.E./COZY" style casting to hold the nose gear - you've gone away from that , so you've lost any trail you might have had. Take a look at any straight gear nosewheel - they're angled forward (wheel steering axis contacts the ground AHEAD of the tire contact patch - NOT behind it). Why did you change from what probably could have been a stable system to an unstable one? 3) Springs do not provide "damping" - damping is an ABSORBTION of energy, not a transmission of it from one form (kinetic) to another (potential). There is no damping mechanism in the Wilhelmson gear (I know, I have one). 4) You have no "shimmy damper" that I can see. With a very small diameter rod providing the steering force from the top to the bottom of the strut, you have a system that is very torsionally compliant. Without shimmy dampening right at the wheel, with high compliance (low stiffness), and especially with an unstable geometry, you will be susceptible to vicious shimmy in this system. 5) Assuming that you're using a 3D solid modeler, you should be able to calculate the weight of the system easily, by assigning materials to each part and asking for material properties. Are you not using a modeler that can calculate weights automatically? 6) In your original design, you were using a steering tiller - what happened to that idea that made you switch to using the rudder pedals? 7) What will happen if you partially extend the gear 1/2 way (to move the plane around with no weight in it on the ground), the wheel goes off to one side, and then you attempt to fully extend the gear without straightening the wheel? Will the engagement mechanism engage with a 90 degree wheel angle? Can you rotate the wheel 90 degrees sideways when the engagement mechanism isn't engaged to facilitate ground movement? 8) With the gear fully extended and the engagement mechanism engaged, you state that you have a 43 degree rotation angle capability. This will increase the turning radius from the stock 98" to about 135 - 140". Not a real big deal, but not as tight as before. You understand that you are giving up some ground handling capability here, right? 9) The dual gas springs (or equivalent) that you plan to use between the rudder pedals and the steering mechanism introduce another point of compliance into an already very compliant system. This greatly complicates the vibration modes of the system, as well as substantially lowering the vibration frequencies of the system (and increasing the possible amplitude of vibration). 10) I don't see any modifications to react the huge side loads that will be developed when steering at high speeds - since it's not possible for the system to ONLY castor (the gas springs are always engaged), there is the possibility that with the wheel off center, or in a crabbed landing, there will be LARGE forces that will be reacted into the gear strut and the NG-30's. How will your system deal with these loads, that a standard system never sees? There might have been more, but I can't recall off the top of my head. Thanks for the opportunity to review your design - it's very interesting.

I will agree with this statement. Having the ring substantially stiffens the structure, and having some cross-members as you sketched would substantially stiffen the conical mount design. Since the back side of engines with conical mounts is not substantially different than engines with Dynafocal mounts, what you've proposed would be possible. However, all that said, no amount of cross members will turn an engine mount of this type from an indeterminate structure to a determinate one - unless you were willing to use ball/pin joints at every tube intersection. There's "CORRECT" and then there's "ADEQUATE". While I will certainly agree with you that this design is not optimal from a load paths standpoint, without a stress analysis, it's not possible to claim that it won't work. It IS possible to claim that it MAY have bending moment issues, but that's something that can be addressed by looking at the population of non-cross-braced conical mounts and obtaining the failure rate. If the imposed bending moments, even under vibration loads under "G" loading, are lower than the fatigue limit of the 4130 steel, then it doesn't matter that the mount would be structurally more efficient with different load paths (and the claim of structural efficiency is dependent upon the increased thickness of the existing tubing weighing more than the extra tubes for bracing would have weighed). It would work - that's all that anyone will care about. Absolutely true. But 23 years of experience without failures would make it ADEQUATE, if not optimal. I have no idea if that's the case, but it might be. It's one of the things to consider. A seperate set, yes, but intimately tied in. If the engine attached to this mount put out 20 HP and weighed 40 lb., I don't think this discussion would have occurred. Conversely, if it put out 2000 HP and weighed 1500 lb., it would be fairly obvious that the mount was inadequate. We're somewhere in the middle, and that's why you had the gut feeling that there MIGHT be a problem. And there might - as I said before, I'd feel better about it with some more bracing on the top, but IF O-320's have been mounted to this design for 20-40 years, and have worked, then my gut feeling was misplaced. Here are some pics of mounts used for lower power engines that have similar configurations: http://www.greatplainsas.com/scengmounts.html Well, you bring up some good points, but using certified GA aircraft, most of which were designed 60 years ago by folks who may or may not have done any analysis or much testing, and which, by and large, have failure rates of systems and components that would NEVER be tolerated in a consumer or automobile marketplace as an example of something to emulate is not one of them. Always.

Folks: Well, we finally have a schedule for the COZY forum at OSH. Even though it's not up on the Airventure web page yet, I've been told that we will be giving the forum on Friday, July 28th, 2006 at 1:00 PM. I don't know which forum tent yet. We will have three presenters for the Forum: 1) Me (Marc Zeitlin) - general COZY/canard info - condensation of 2004's talk with some extras from the suggestion box - I'll address some common modifications 2) Daryl Lueck - - not sure as to Daryl's topic yet, or he may have a supporting role for the other talks 3) Chris Essylstyn - will discuss his heavily modified, 12" stretched, O-540 powered, retractable landing gear COZY MKIV. This should be VERY interesting! As an added attraction, Nat Puffer has informed me that he's hoping to make it to OSH, the dinner, and the forum this year, although he will have to drive to get there :-). Daryl will be scheduling the COZY dinner - I'm hoping that it will be Thursday evening, as I have to depart for the east coast on Friday afternoon after the forum, but it'll be a great time no matter what or when - Daryl and Kim always do a great job.

If you join the COZY mailing list at: http://www.cozybuilders.org/mail_list/ you'll get a database of 586 builders/flyers/wannabees, with all contact information. You can easily search for people near you. Builders will most certainly be interested in helping hands.

Yes, I read it. There's nothing new in it, except the student's lack of understanding of the fundamentals of lift. Actually, I guess that's not new - most people don't understand how Newton, Bernoulli, and/or Coanda play into the concept of lift. Mostly they don't understand that Coanda has nothing to do with lift production on a normal wing, and that Bernoulli and Newton are two different ways of saying EXACTLY the same thing. At any rate, from the standpoint of the canard downwash effecting the AOA of the main wing, this is also no surprise, and is taken into account in ALL aircraft design, whether canard or conventional - the downwash from the front wing can ALWAYS affect the AOA of the rear wing. While the paper is a nice science experiment, the student understands and admits that because the Reynolds number of the small model is very small, the results can not necessarily be extended to full size canard aircraft, at least with respect to the magnitude or angles at which the effect occurred. How's that?

I had heard once that it was to reduce the moment coefficient for the main wing. Having reflex near the T.E. of a wing raises the moment coefficient, and that would entail needing more canard authority (which might have an effect on deep stall susceptibility). Removing the reflex (flattening the bottom) would lower the Cm. I have no idea how accurate that account was.

And has already, for those interested in checking. See: http://www.canardzone.com/forum/showpost.php?p=10634&postcount=5

Ummm, there WAS a reason for it. But if they told me what it was so that I could tell you, they'd have to kill both of us :-).

It's the thing that Burt's in, because he hasn't put hydraulic controls on any of his homebuilt designs, nor the vast majority (if any - can't think of one off the top of my head, but there might be an instance here or there) of Scaled's designs, no matter how large. And because he's not interested in adding weight to reduce reliability or control harmony. Don't feel so bad that you're in it with him.

Absolutely correct, but usually, folks are trying to make their aircraft safer, or perform better, or do something that other aircraft can't do. Not just make them more dangerous and heavier. Since you don't address the questions that have been asked, or explain why you believe that your ideas will make the aircraft better, the only thing we can assume is that there are no reasons, and you're doing it just to be different.

Yes, that's right. Scaled hired me because they want the safety and comfort of those that can only think inside the box, and Richard, who was instrumental in bringing the Berkut to market, also is a staid, inside-the-boxer. There is the remote possibility that we have a slight clue about this issue, no?

I wonder if you can point me to one aircraft under 12K lb. MGW that uses hydraulic controls for flight surfaces - especially non-redundant ones. Having a redundant pump does not a redundant system make. So you're going to use boosted controls on an aircraft that weighs less than 2000 lb? You are aware, of course, that there are MINIMUM force requirements for control surfaces for controllability reasons, right? Given that canards already have pitch forces than many feel are too light, you'd need negative boost for pitch. Uh huh. Which trim tabs would those be? Pitch? Nope. Yaw? Nope. Roll? Nope. Maybe I've forgotten an axis? Hmmm. So you're going to increase the empty weight of the aircraft by about 10% so that you can reduce reliability, complicate the systems, screw with the handling characteristics, and eliminate trim systems that don't exist? You know, of course, that even aircraft with hydraulically boosted controls have trim systems..... Which autopilot for small GA aircraft works with hydraulic controls? Or are you just trolling with this post to see what reaction you'll get out of the engineers in the group?

Sure - I've done it many times (at least on a COZY - you don't state what type of aircraft or model you're talking about). You take the spring hook off of the rudder eyelet, and then fish around inside the tube until the OTHER end falls off the eyelet inside the end of the tube. To reinstall, reverse the process. I generally found that it took about 5 minutes to get the hook back on at the bottom of the tube, and then another couple of minutes fighting with the needle nose pliers to get the hook over the eyelet on the rudder. I can't imagine them falling off by themselves, given how hard it is for me to get them either on or off.

"Does Not" and "Can Not" are two different things. Greg has explicitly said that his aircraft COULD go a LOT faster, but he limits the speed because the Vne of the COZY is 220 mph IAS, so he doesn't go faster than that. And yet Jack Morrison's E-Racer has gone ~300 mph, as have some Berkuts, and Klaus's V.E. is in the 250 mph range. Drag does rise, and the aircraft are not designed for those speeds, but they can achieve them with a lot of power and judicious drag reduction. Greg's jet could EASILY beat those #'s, if he was willing to exceed Vne. As Richard Riley has pointed out, the Critical Mach # of the canard is somewhere between .55 and .75, and the main wing may not be far different. This gives an upper limit for TAS. Without the drag polars, we can estimate a Cd of 0.005 for each wing at low AOA's and get an approximate speed/power curve. It's pretty obvious (since it's been done) that 300 mph is achievable with enough power, and "enough" is defines as somewhere in the 300 HP range. Let's not forget that that's a tiny 1 seater. A very efficient one, no doubt, but tiny. It IS possible, and it's been done. I've said so many times, and others have proved it. The question is not CAN it be done, but how close to the edge are we when doing it?

You must have missed the multitudinous discussions of Mach Tuck on both the mailing list(s) and the fora. The Roncz canard is theorized to have a critical Mach # of about 0.55 - 0.75. At high altitudes, with a turbocharged engine, canards can approach that. That was my reference - to go a LOT faster in one of our aircraft, the canard airfoil would have to be changed in order to avoid this phenomena. Read the thread: http://canardaviationforum.dmt.net/showthread.php?t=2085 Especially Richard Riley's last sentence in his last post.

Well, you'll have to explain it to me, because I'm not sure what you "got" - there was no joke.... A lot closer to the latter than the former. It's not like folks with the long nose L.E.'s are seeing any speed decreases.....

Klaus is trying to get the last few knots out of an airplane that has already had everything even remotely easy done to it (assuming he's doing to the L.E. what he's already done to his V.E.). Karollina's point about canard thickness being a substantial portion of the drag of the aircraft is just wrong. Go look at drag polars of various airfoils appropriate for canards, find the minimum Cd, and then think about how many pounds of drag dropping that Cd will eliminate. Then consider that a reason to change the airfoil of the canard would be to eliminate the propensity for Mach Tuck at the high speeds that Klaus is trying to achieve. There may very well be airfoils out there that would have better Cd's at the Cl's used for the canard, but the thickness is not the major issue. That's what I was directing my comments to, along with the comment that fuselage lift could replace wing lift. Since wings are far more efficient (Cl/Cd) lifting surfaces than fuselages, that's the last thing you'd want to do. Plus, thin airfoils have terrible stall characteristics - not optimal for a canard aircraft. Take a look at: http://www.mh-aerotools.de/company/paper_3/yaka.html You can see that the minimum Cd for all the airfoils listed (with smooth surfaces) is about 0.005, in the range of Cl's we're talking about. You're not going to substantially reduce that (at low Cl's) by changing airfoils. A look at Appendix IV of "Theory of Wing Sections" shows that almost all minimum Cd's are in the range of 0.004 to 0.006, with the lower "drag buckets" coming only in a very small range of Cl's (with thin sections). You might change OTHER characteristics by going to a thinner airfoil, but not the minimum drag coefficient, at least not substantially. This implies that changing canard airfoils will not have a large effect on top speed (from a drag standpoint). I was referring to Klaus's aircraft. His O-200 puts out substantially more than 100 HP. Everything can be improved. The things I listed are the main ones, and will get you many knots relatively simply. The things you list are second order effects - optimizing them MIGHT get you a few knots, but it's hard to say.

Since canards already have "suitable" TO/Landing performance, the addition in weight and complexity of moving the whole canard airfoil is completely unwarranted. It's on thing to take on the complexity in a 6-8 seat bizjet - it's quite another in a 4 seat (or 2 seat) GA aircraft. It's why you don't see multi-surface fowler flaps on GA aircraft, either. If there were any truth to Karolina's statements about drag, then it might be worth thinking about. However, her suppositions about drag and airfoil shapes, with respect to L.E.'s and COZY's are incorrect. Witness the many canard aircraft (Berkut, supercharged E-Racer, souped up V.E.) that can fly well into the high 200 - 275 Kt. range with the stock airfoils. It's not the thickness of the airfoils that's creating any issues here. I'm not sure what your first sentence means. Personally, I wouldn't even think of adding flaps, and the concomitant complexity of canard rotation/sliding to my aircraft - the reduction in reliability alone would make it a non-starter. I can land and stop in less than 2000 ft, normally - 3000 when at gross weight and forward CG. There are almost no airports that I want to go to that I can't get in and out of. If I wanted to go faster, I'd concentrate on intersection drag of the gear, cooling drag, and I'd turbocharge my engine. Leave the airfoils out of it.

I'll bet someone could. To meet the FAR landing speed requirements, the Starship required flaps on the main wing. Flaps substantially change the moment coefficient of the wing, and to counteract the extra nose down moment of the main wing, the canard needed to be further forward. The swinging forward of the canard was coupled to the flaps on the main wing - when the flaps were deployed, the canard swung forward, moving the center of lift of the canard forward. This kept the lift/moment balance of the aircraft correct. It's not "speed" related - it's flaps related. Now, you only deploy flaps at relatively low speed, but no matter what the speed, if the flaps weren't deployed, the canard doesn't move. The mechanism is extremely complex. The patent # is 4,641,800. Interestingly enough, while poking through the patents, a derivative patent for a SLIDING canard is presented by a "John A. Lockheed", in 1989, that uses a COZY III as the basis for the figures. The figures are reasonably explicit as to the workings. Whether this aircraft was ever built, I have no clue - I've never heard of it or seen it. The patent # is 4,848,700.

You write down whatever you like. Nothing. The FAA couldn't care less how much time you spent building your plane. Since the FAA doesn't care, you can write down anything you like, and claim as much or as little time for whatever you do. Sure, and write down the times you spent eating lunch while staring at the plans, too. The only thing the FAA cares about is whether it was built for "education and recreation", that the "majority" of the aircraft was built by the builder(s), and if you want the "Repairman's Certificate", that you can show that you were one of the builders and are intimately familiar with the aircraft. Feel free to log anything you like.

Would thinking about it change anything? The aircraft is designed with a maximum recommended propeller diameter. Assuming that you do not put a propeller larger than that on the aircraft, and that you fly within the recommended CG range, the only way that you could touch the prop tips to the ground during a landing is by performing a hard carrier landing while at a very high AOA. I have only heard of a one COZY MKIV flyer that has ground 1/4" or so off his prop tips in a hard, high AOA landing. I'll leave speculation regarding how he managed to do this to the peanut gallery. So, the answer is, no, we don't think about it, because there's no need to. Unless you screw up pretty badly in multiple dimensions, you won't touch the prop.

According to Kai Christensen, the owner/builder, it is NOT based on a COZY, but a L.E. derivative.

Marc Dunno, I'm not a seamstress, but if I were to guess, I'd say 6-10, depending upon how you laid it out. You can get 60" wide Sunbrella for about $20/yd, and you could do the whole fuselage with 6 yards of that. You'd need about $10 worth of snap buckles and straps, so that's what - $130 or so (plus a spool of heavy duty exterior thread). Making the pattern is pretty trivial. Carl Denk describes how he made his cover (not sunbrella) in the COZY archives. Go to a show, find someone with a cover, and look at it. Not rocket science.

The 2004 presentation was filmed by Mark Beduhn, who is an amateur filmmaker. You should do no such thing. I have to say, the lack of respect for copyright protection on these fora are, IMNSHO, deplorable. The DVD was filmed and is owned by Mark Beduhn. If someone wants a copy, they should contact Mark and see if he has any more copies to sell. As of 8/4/2004, he was selling them for $15 including S&H. As far as suggestions for the presentation, I asked: "Please reply directly to me via email at the address below - have at it." Posting suggestions here is not the optimal way to get your opinions heard, and your suggestions done.