A stronger RV nosegear

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Thought I'd go public with my current homebuilt airplane project. Open it up for comment, advice and ridicule.

It was two years ago I decided that I can't afford two airplanes. Even ignoring the money question (a pretty big thing to ignore) I haven't the time to fly or maintain another. Can barely do the one I've got.

Like everybody else on here I've got a whole stable of planes I'd like to build. The backcountry STOL with a car engine. The electric motorglider. The scale warbird. The aerobatic 103. None of those are ever gonna make it off the digital napkin.

But of course I need a project. When I started the RV, I wanted a cross country plane that could also do aerobatics. These days I want an aerobatic plane that can also do cross country. So it still covers 95% of what I want to do in an airplane. But, maybe there is a way to push it up to say 98%?

Having finished my phase one in the middle of a pandemic, I haven't yet done any real cross country flights. But, small plane cross country isn't really compatible with working for a living and needing to get home on a schedule. Too weather dependent. Plane is capable enough I could visit relatives in California with a 4 hour flight instead of an 11 hour drive, would just need to limit myself to summer so as to be reasonable confident of good weather.

Aerobatics wise it will be a long time before the airplane is the limit instead of my skill. It would be nice if aerobatic weight/CG limit allowed me to share aerobatics with someone, but I don't actually know anyone who wants to do any with me. Not worth getting an extra parachute for that "maybe someday, somebody" mission even if the plane could handle it. (right now I could do it, but the passenger would need to weigh 120lbs or less with chute).

One thing I had wanted to do with this plane, but think it may not be a good idea as it stands is do back country fly in camping. I love going into the wild by truck/bike/kayak/foot. Would be nice to add plane to that.

Plenty of people do take RVs into the back country. The nose wheel however does have a reputation of being not quite robust enough. Back around 2000 or so when I had to decide between tailwheel and tricycle gear Vans still marketed the nose gear as being better for STOL work as you can get to a higher angle of attack and land shorter.

Shortly after I started my fuselage Vans came out with the RV-8A tricycle version and it was about 15 lbs lighter and easier to build than the conventional gear model. The gear can bolt right to the main spar instead of needing "gear tower" structures up front to support the legs. That 15 lbs and easier to build was enough to push me to tricycle. I had also been flying a biplane for several years at that point and had gotten my fill of S-turning.

Over the next few years I got married and my building slowed way down. The nose gear model RVs started to get a reputation for flipping over on grass or soft fields. In 2005 the NTSB took the unusual step of looking into a possible problem with a homebuilt design:

Text of Study
Photos and Data

In 2007 Vans came out with a service bulletin on it:

SB 07-11-09 - Van's Aircraft Total Performance RV Kit Planes

Basically they made a thicker leg with a shorter fork to increase the clearance of the nut before it would drag into the ground. They also set a maximum static nose wheel load of 375 lbs and say to always fly with the fairing in place.

This fix does seem to have reduced the incidence of these problems, but not eliminated them. I do have the new leg and fork on my plane. But I'm still not comfortable taking my plane off pavement. It can take off and land plenty short, but I don't know if it can handle the rough/soft field part.

So, my goal: take my existing RV-8A and modify it so it can handle some back country airstrips. Not trying to get it onto the proverbial Alaskan gravel bar, just into and out of the smoother/longer Utah and Idaho back country airstrips. I don't expect to make it into a supercub or anything.

Perfection isn't possible. Hitting a big enough rock/ditch/badger hole will flip/wreck anything. If I can make it so it can handle the rougher fields as well as a tailwheel RV that will be good enough.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Step one is to start with a survey of what others have done. I'm not the only one to want a stronger nosegear, enough RVs around that it is well trodden ground.


The one with the largest installed base is the "anti-splat" mod. The problem with this one is it isn't actually to make the gear stronger. Per an older discussion by the designer the brace is only supposed to come into play after the nose gear has yielded. The intent isn't to prevent the failure but change the failure mode so it is less likely to flip over.

If it does work it allows the leg to fail, but prevents the prop strike, vertical/rudder damage, canopy breakage and possibly serious injury that result from an overturn. That is a good goal, but unfortunately we don't know if it even achieves that. Some have flipped even with the mod, there aren't good enough stats on what fraction of planes have it installed to tell if it actually reduces the chance of flipping on nose gear failure. So, not what I'm aiming for.


With the RV-10, Vans redesigned the nose gear. Instead of using a Whitman style spring gear, it uses a hinged "stiff" leg and a stack of rubber washers on the engine mount for the spring. This allow the gear to only deflect up, and not back allowing the leg/fork to hit the ground and dig in. The RV-14A uses the same gear design, in 2019 they came up with a kit to apply this new design to the RV-7A and 9A. New Engine Mount and Nose Gear Option for RV-7A and RV-9A Finish Kits - Van's Aircraft Total Performance RV Kit Planes

However, there aren't enough -8As for them to bother with this mod for the -8A. Most who want the tandem seat go for tailwheel. I could do something similar but would need to design and scratch build it myself. The rubber washer stack would go right where I ran all the wires through the firewall so that would also need to be moved. A small job compared to the rest of it.


A few builders have put a bigger wheel on the nose, going up from the standard "Lamb" tires to the 500x5 size used on the mains. Usually also going up to 380-150-5 on the main wheels as well. http://mesawood.info/myrv7a/building/langing_gear_mods_main.htm This is a 23% bigger radius, I know how much better my truck handled rough roads going from 28" to 31" tires (10% different). The bigger tire means a longer fork with putting a larger moment into the leg. That 7A linked stiffened the leg with fiberglass, another 8A builder got a custom thicker leg from Langair.

For the really big change, an Australian builder put on a telescoping nose gear:




I can't find any pictures of what the support for that leg looks like.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
To begin with, I can't make a "better" (or more accurately "differently optimized") gear if I don't understand the one I have now. Since it is a kit part, the plans don't have full dimensions. It just has an image of it and instructions for installing the part. I made initial calculations based on measuring the drawing in the plans, but the loads needed for deflections seen in my landing videos were way too high.

So, with last year's condition inspection, I carefully measured the nosegear leg. And to get measured deflection with a known load, I took pictures with the weight on gear, and with a load pulling down the tail until the nose wheel is off the ground. By comparing the pictures I can measure the deflection. For the known load I took the nosegear weight from my W&B and subtracted the weight of the cowling (20.5 lbs). The cowling is pretty much right over the nose gear, so this should be close.

MERGEDr.jpg
(with six inch scale bar, marks at 1/4 inch)

Measuring the deflection from that image, and using the linked spreadsheet the predicted deflection from known load matches to within 0.01" / 0.07 degrees. Better than expected. I can then use that to compare strain (just using My/EI) to a 220kpsi max from Neal Willford's spreadsheet.

I put in the loads from FAR 23.499 based on Van's maximum 375 lb nosegear static weight (I can't actually load my that heavy, 350 lbs is the highest I can get if I stick by the maximum 50 lb front baggage limit, and I don't like the way it flies with the CG that far forward). Based on this calculation it easily has over 50% safety factor for all of the FAR 23.499 loads. This agrees with the NTSB "The FEA shows that the nose gear strut has sufficient strength to perform its
intended function."

So, if they are strong enough, how are they breaking?

Some are clearly pilot error. Land too fast and force it down nose gear first will break just about anything. But many of these accidents are happening on rollout or fast taxi. What I think may be happening is that the leg is nearly undamped, when it deflects back after hitting a bump, it bounces forward, then back again almost the same distance. If you hit one bump, the leg deflects back, bounces forward and then you hit another bump as it is going back again could push it enough that the leg end/nut digs in. Once it stops rolling and digs in it’s going to fail if you are going fast enough. Fast taxi is easy enough to avoid, but every landing will have a time when you are going too slow for the elevator to hold the nose up but still fast enough to bend the gear.

This would explain why many of these accidents happen on the runway or taxiway that the same plane and pilot has landed/taxied dozens of times without problems. In previous passes they weren't going the right speed to hit the bumps at the "resonant" frequency. This theory is at least partially backed up by this video.
Investigation: AO-2017-001 - Landing accident involving Van's RV-6A, VH-TJM, Starke Field ALA, Queensland, on 2 January 2017

That pilot did land nose first after the first bounce, and single stepping though the video you can see that not only does the nosegear touch down first on the second hit, it also hits with the gear already near the rear of its swing as it vibrates from the first bounce.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
So, my plan so far: Read a lot of books, try to learn enough engineering to actually design the thing. (Been working on this part for years now.) All my thoughts so far will require at least a little bit of welding, so I want to get a ox-acetylene torch. (always looking for an excuse to buy new tools) Finally got paid for some consulting work done last year, so just got the torch today. Next, simultaneous to the design phase use some left over 4130 tube I have and make a lot of scrap learning to stick that together. Maybe get some sheet like in Scapper's videos. Design/build a folding bicycle or velomobile to make some welded parts that actually get used.

Hopefully by the time all that gets done I'll have decided on a design and can get to building. Expect building will go fairly fast at that point, but wouldn't be surprised if that is years away yet. Keep the RV on pavement until then. (Maybe go to Ibex or somewhere easy like that).

How much am I willing to spend? I figure anything I design will be heavier and draggier than the standard gear. I'll allow up to the 15 lbs I saved going with tricycle gear, 5 kts off the top speed and $2000 or so. As is usual in homebuilt aircraft accounting, tool cost isn't included and my time is free (or else I'd be over budget already). Ideally make it bolt-on, so I can take it back off if it doesn't work.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Thought I'd start posting my design thoughts. First, the minimal effort: just use it as is. Only go to fields where there is a recent report of smooth conditions. Load camping gear, etc. so as to be toward the aft end of the envelope to keep weight off the nose. Hold the nose wheel off as long as possible when landing, and full aft stick when taxiing. If too fast go around, don't force it down. (rules out "go around not possible" strips, but safer not to go to those anyway.)

This is what Van's would recommend. Pretty much doing all this stuff already (except loading aft, prefer to keep it in the middle). Do want to try to see if I can do a bit better for my personal mission.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Idea two: Maximal effort.

Since the problem is the nosewheel, why not get rid of it and go tailwheel. The RV-8 is different from other RVs in that the main gear is not attached to the engine mount, but instead to "gear towers" support structure about half way between the firewall and the wing spar. Mine is an 8A, so I don't have that structure (giving me more leg room). Could go with a redesigned engine mount that could mount gear legs, but they would be either too far forward or very long and swept back.

Another thought is to attach the mains to the wings. Would take building new wings to handle the point loads, but the fuselage spar carry through should be fine. Could make it a tapered wing instead of the hershey bar. That would move some of the load inboard, which means you could increase span and area for a lower stall and better climb. Keep the same MAC.

Taper3view.png
(white is current, light gray is tapered wing and new gear) Per Shrenk's approximation the root bending moment is the same as the original. Kept the rear spar in the same place. Run the ailerons out to the tips to try to recover some of the roll rate you lose to the higher span. Some basic performance estimates 9% lower stall, 8% better glide and 4% better climb. Could design up to the 1600 lb aerobatic weight instead of the 1550 I have now. Looks great too. Straight leading edge inspired by my Hayate obsession.

Downsides: As sketched above using the same flap hinge (so the flap motor and arm stays the same) leads to only about 3 inches of clearance at full deflection. Not compatible with rough strips. Also the landing gear are coming out of the middle of the fuel tanks. Moved them outboard a bit to make up for some of that but still a problem. Could move the tanks aft of the spar, but that makes my aft CG problem worse. Longer span and more area but same tail will lead to lower stability, possibly worse spin recovery.

Would be a lot of work. Could probably get a RV-8 standard kit built before I could design and build this. Sell my current plane when it comes time to buy the engine and avionics. RVs sell for enough over the kit prices I would probably do a bit better than breaking even. Could maybe even put a constant speed prop on it.

But, you know what? I like the nosewheel. All RVs have great in-flight visibility, mine also has great visibility on the ground. I can see that rock/badger hole before I hit it. The insurance companies (who probably have even better data than the NTSB on accident rates) charge more to insure a tailwheel RV than a nosewheel. So the tailwheel planes get into trouble more often than the "weak" nosewheels. As I said before, I've had my fill of S-turning.

Also, it appears from reading on-line that some pilots try to use a tailwheel to somehow make up for a self perceived lack of masculinity. Fortunately, I don't have that problem.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Idea three: Lower the wheel.

Using the 380-150-5 tires on the mains raises up the plane about one inch. Could build a new fork to lower the nosewheel by that same amount to increase the amount of deflection needed before the leg hits. This is essentially what Vans' service bulletin did, but even more so. Maybe make the so it surrounds the leg, if it hits it acts as a wide skid instead of a sharp edged screw/nut that digs in easily.

lower.png
(white is existing, red is proposed)

Compared to the existing design, loads (from the same spreadsheet as above) work out a few percent lower in the leg. As the fork covers the leg end and nut, it will be a bit harder to install (need to rotate the fork with the nut, then hold the nut while rotating the fork back.

Pros: Pretty simple mod. Higher obstacle clearance. More deflection needed before the leg digs in.

Cons: Not really any stronger. The lower angle on the leg/fork connection does seem to reduce loads about 8%, but otherwise it will fail at the same load. If the ground is soft enough to sink the "skid" in, still have the "pole vault" effect flipping you over.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Idea four: bigger wheel

Just like with my 4WD truck, I like the idea of going with bigger tires all round. Up the mains to 380-150-5, and then up the nose to 5.00-5. However the longer fork needed puts a bigger bending load on the end of the leg. The way others have done this is by using a RV-10 nosewheel fork, then machining it thinner with a reducing bushing to fit on a 7A or 9A leg. The RV nosegear forks are made with pretty thick aluminum, and then the bushing for the thicker leg pushes it even farther out. If I make a custom fork, using steel instead of aluminum I should be able to have the wheel axle closer to the original position than using a 10 fork. Should be closer to the design load.

Just looking up 1.25" ID bushings (it is a flanged bushing, not a ball or roller bearing) it looks like they have a 1.75" OD flange and 3/16" wall. Shuold be able to mount it in a 1.75 0.0625 wall tube. Take the 6.75" tire radius and 1 inch clearance and half the 1.75 tube to get a 8.625 distance to the leg centerline. Should be closer than the -10 fork.

Running the longer for radius thru my same spreadsheet again, it reduces the load on the root of the leg, but increases at the front bend. Still looks like I have better than a 50% safety factor.

Since I'd be making a custom fork, go ahead and use the same "rounded skid" design as idea three. biggerWheel.png
Just looking at that image I might be able to shorten the fork more, if I allow the the same growth factor percent as the standard fork.

Pros: The bigger wheel gives me more clearance, as well as reducing the forces you you get from any given impact. Extra rubber means more damping when the leg is bouncing.

Cons: Bigger wheel means more weight. The extra length of the fork means larger bending forces on the leg, possibly making it fail earlier.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Idea five: Trailing link

Make a "stiff" custom leg, then put a trailing link flexible fork on that leg. Might be possible to make it lighter than the standard gear. That 1 inch thick solid steel rod is very heavy, if it was replaced by a hollow tube (higher diameter to get the stiffness) the weight saved might make up for the extra in the fork. The design could be set up for more travel. The standard design only has about 1.5" of travel with the FAR 23.499 loads, more would be nice for those ruts/rocks on an unpaved field. Since the leg and fork are both being redesigned, could design it for the bigger wheel from the outset.

trailingLink.png
White is standard, red is new. Drawn with 4" of travel. Solid line is at standard weight, dashed lines show unloaded and maximum load positions.

Could go with a hinge point aft of the leg with a spring in front of the leg for a thin profile, or with a hinge point in front of the leg with springs behind. As I've studied the possible layouts for this I found it looks like two mountain bike shock absorbers on either side of the wheel has the desired load and travel range. Would be wide, but should fit under a standard Van's fairing.

This is the one I've been mostly working on. Initial thought was to use a streamlined tube, possibly with internal reinforcement as the leg, spent quite some time trying to predict the strength of of a streamlined tube in bending. There are tables for round tube, but not for streamlined. Would probably need to just build and test. But with the cost of streamlined tube that is painful.

For the leg it looks like a 1.5 inch tube with 0.125 wall should be strong enough, you could put a 1.25 inch tube inside that to fit into the existing socket on the engine mount. That 1.25 inch tube would need to be quite thick to be strong enough, might want to run another set of tubes to the firewall bolt point to stiffen the upper end of the leg (should still fit in the cowling, but adds more stuff blocking the cooling air exit).

On the fork I thought I had the hinge in back, spring in front layout all worked out in my head. But when I went to draw it up for this post I found that the nose was going to drop down to take away much of the ground clearance the design was intended to give me. The above image shows how the hinge in front, mountain bike shocks in back would work. Shocks would mount to the two small red circles. As drawn there is just barely room for the bigger tube and a wheel fairing at full deflection.

Pros: Much more travel, can use the bigger wheel. Can put in a spring/shock absorber that has damping so not just relying on the tire. Leg failure mode would change to just absorbing energy as it bends, not releasing that energy afterwards to flip the plane over.

Cons: The second most complex design (after going tapered wings/tailwheel). As drawn above adds about 3 inches to the base height, even using the larger tire on the mains will still change the ground angle about 1.6 degrees. Makes a nosewheel first landing slightly more likely.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
Was out at the airport yesterday replacing a leaky fuel fitting. Because of all the hoses, etc. under the engine it is one of those "turn it 1/16 of a turn, flip the wrench over and do the next 1/16" places. So I had a lot of time to look at the engine mount where the nosegear attaches. Lots of stuff going through the spot where a pivoting leg's spring/shock absorber would go. Three oil and one fuel hose, the throttle cable and the primer system. But thinking about it more, if you hinge it at the point where the front of the current socket is (see the image in post #3 to see what I mean) and then put a tension spring/shock absorber at the back of the socket it would have plenty of clearance. Couldn't use a compression spring, as the bottom of it would be blocking the exit air path.

This would mean more angular deflection for a given vertical deflection than one that pivots at the firewall, but that just lifts the pivot point up and out of the way of obstacles. The design of an appropriate tension spring with some kind of damping is also something that will take some thinking. Might just be able to use bungees. Not sure how they do in the engine compartment (high temps, lots of oil and occasional gas contact).

Also took a "fish scale" and pulled the plane around the hangar with it. The rolling resistance on a smooth hangar floor at about a 1200lb weight (empty but with 30 gal of gas) took a 50 lb pull. So a hard surface rolling friction coefficient of around 0.04. Use 0.8 for the tire spin up, then use maybe 0.1 for a soft surface. Starting to get some numbers to design by.
 

bifft

Well-Known Member
Joined
Apr 17, 2011
Messages
385
Location
Utah
From the discussion thread, I decided that a good goal (or at least a limit it makes no sense to exceed) for how strong the nosegear should be is to make it able to handle as big a rock as the main gear. So I need to figure out what the mains can take. Just like the nose gear, the main gear is a kit part where the plans don't specify the exact dimensions. I can get to the root and end easy enough, taking the fairing off to measure the whole thing requires removing the wheel.

Made a spreadsheet based on the one for the nosegear to estimate the deflection. Now need to measure actual deflection under known conditions. My wing jack got stolen, doing the annual I was building a stack of cinderblocks and using my car jack on it.

So, I replaced my stolen wing jack. (A cheap jack bought on-line with a custom base and supports made from scrap angle). While at the airport yesterday jacked the plane up and measured the deflection. Sure was a lot easier than piling up cinder blocks.

main_merged.jpg

From the W&B, that is with a load of 466.8 lbs on that wheel. From the pictures I measure 1.89 inches vertical and 1.55 horizontal deflection. About three times what my spreadsheet estimates. So looks like I need to take the wheel off. That's enough bother it may wait until this year's annual (when it has to come off anyway to lube the bearings).

Have also started (slowly) working on what I'm calling "very finite element analysis" software where using the same math as the deflection spreadsheet to calculate dynamic behavior such as hitting a rock at some specified speed. Will need to get the spreadsheet to accurately predict deflection first. (I figure if the deflections are accurate, the loads/stresses should be pretty close).
 
Last edited:
Top