What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Strong Crosswind, not a great outcome

scsmith

Well Known Member
Well, its confession time. After 400+ hrs in the RV-8, and over 500 hrs of total tailwheel time, I just groundlooped the RV-8. Fortunately nothing damaged but my ego and confidence. But...

Usually when something doesn't turn out so well, I can play back in my head and think through what I did, what I should have done differently, etc. Learn from it.

In this case, I'm not sure I could have done anything differently.

Before the story, you probably need to know I have a rocket-style steering link, which I really like. It has a nice amount of spring give but is very responsive at low speeds.

In the pattern, the ATIS reported 16 kts at 260 degrees. Runway heading 190 degrees, paved, 75 ft wide. I'll do the math for you, that's 70 degrees off runway heading. The wind turned out to be a bit stronger than 16 kts though.

On final, I transitioned from coordinated flight to a stabilized slip. (I've always used a stabilized slip on final -- you just use your feet to keep the airplane aligned with the runway, and the stick to control the lateral drift. ) Over the numbers, wow it took a LOT of slip, but speed good, right on centerline, aligned with centerline. As I floated a bit, some of the slip naturally comes out since the wind is not as strong right near the ground. Still a lot of slip. One main wheel down, then the other. Still good alignment, right on centerline. Here is one thing I **might** have done differently: I did not make much effort to hold the tail up high. It was taking a lot of rudder pressure (left rudder) to keep aligned, and I just let the tail drop slowly as I slowed down. I felt the tail wheel hit, still good alignment, still on centerline, rolling out....still taking a lot of left rudder, but all good. Stick well right of center and a bit aft of center. **another thing I might have done differently: I could have had the stick full back to really firmly plant the tail wheel. Not sure it would have mattered**

Rolling out nicely when IN AN INSTANT, I was sideways, because the swiveling tail wheel had come unlocked and just swiveled. When I processed that and made the intentional command to go full hard left rudder and left brake, my foot was already there, way ahead of my brain. All I could do was just hold that and wait, pretty much a passenger at this point. The airplane slid in a gradual arc slowly diverging more in heading and scrubbing speed until it left the pavement, then it pretty much finished the groundloop. All the corners of the airplane stayed level thank goodness. I had to get out to turn the airplane more to get the tailwheel out of the soft stuff and up onto the prepared gravel alongside the runway so I could proceed to the apron with my tail between my legs. While I was out, I assessed the actual wind. It was at least 18-20 kts (being an old sailor I'm a pretty good judge of wind speed) at pretty close to perpendicular to the runway-at least 70 degrees, maybe a bit more.

Thinking back through it, I am not sure if the tailwheel unlocked because of more sideload than the latch pin could hold (might be worn, will inspect tomorrow) or if it unlocked because I reached full left rudder while steering on the roll out and the cam-action unlocked it at full rudder. I do remember that I was steering pretty smoothly, no weaving back and forth, and I was not being buffetted by gusts. It all seemed pretty smooth to the instant the tailwheel swiveled. Given that my foot was already full left rudder at the instant I thought about it, I'm kinda thinking it got to full travel from steering on roll-out.
If that's the case, it suggests that during roll-out, one must guard against getting to full rudder travel, and instead feed in more left brake with partial rudder.
One reason I am skeptical that it unlocked because of full rudder command is that the spring in the steering link would have been stretched to its stop and so even if the rudder did get to full travel, the tail wheel would not have.

It is also possible, I suppose, that the tailwheel didn't actually swivel, but rather broke traction and skidded because of insufficient downforce. I really kind of doubt that. Although I didn't have the stick hard back, it was certainly most of the way back, which is kind of a reflex to me as soon as the tail wheel touches. It might have been instructive to walk the runway and look at the skid marks (if there were any). Presence or absence of skid mark for the tailwheel would have answered this question definitively. Unfortunately, its a tower controlled airport, with big jets and stuff, so I can't imagine being allowed to go look at the runway after I taxied to the apron.

So, at this point, I'm kind of feeling like I did the best I could and not clear I could have done better. But ...

Had I held the tail off much longer with strong forward stick pressure, I would have had perhaps better rudder control, but perhaps not ENOUGH rudder control with the whole fin sticking up in that wind. I might have ended up with the rudder on the stop and still turning right. Now what?

With the tail wheel on the ground, it should be much more willing to go straight. Had I sucked the stick back hard, maybe the added downforce would have helped the tail wheel track the airplane straight....unless the swivel unlocked.

I think the tail wheel swivel unlocked. I think at that instant, I was doomed.

thoughts?
 
Last edited:
Well, its confession time. After 400+ hrs in the RV-8, and over 500 hrs of total tailwheel time, I just groundlooped the RV-8. Fortunately nothing damaged but my ego and confidence. But...


thoughts?

My thoughts----------glad you and the plane are still in flyable condition.
 
Great description of what took place.

I'm too inexperienced to comment any further other than to say, I'm glad you and your plane are both Okay.
 
Thanks for the kind thoughts.

One bit of humor to add...while I was sliding and diverging toward the edge of the runway, but not to the edge yet, I heard the guy in the tower say, "wow, look at that" or something to that effect, and I heard in the background the alarm sound that sent a fire truck out to me. Afterward, I wondered what made him key the mic to say that, rather than just say it. Probably second nature to him to key a mic when he talks.

I'm hoping Paul Dye and a few others that I think have described landing in stronger cross winds than this will share some insights.
 
Steve, on my 8 I cannot apply immediate differential braking with my heals on the floor and to raise them to apply the brake I had to release rudder input for a split second. I now land with my heals off the floor in x-wind conditions. Differential braking combined with rudder input is very effective.

I had the same issue when I flew the Airbus.
 
Steve, on my 8 I cannot apply immediate differential braking with my heals on the floor and to raise them to apply the brake I had to release rudder input for a split second. I now land with my heals off the floor in x-wind conditions. Differential braking combined with rudder input is very effective.

I had the same issue when I flew the Airbus.

Yes for sure. I promise if my airplane is still on the runway, my heals are off the floor and my toes are on the brakes. I have nice pedal extensions below the stock pedals that allow rudder use without brakes if desired, and with brakes if desired.
 
Two cents

I'm so glad this event ended without damage. You can simply add this to your piloting experience.

Based on your description of the sequence of events, holding the stick full back as soon as the tailwheel reached the ground is required!
The only other thing I'll offer is not to delay the application of brakes as the rudder reaches the full rudder stop.

I have passionately argued for builders not to modify the brakes on their RV-8 to prevent dragging. Pads are cheap ($40) to reline. It is very important that the brakes be "Right There" when needed in a situation like this.

Returning from Sun-N-Fun several years ago, I landed the Doll in a gusting 30K direct crosswind. The tires were chirping repeatedly during the rollout with heavy braking, but the aircraft stayed on the centerline. With crosswinds that strong, you cannot allow the aircraft to get too far into a swerve or the crosswind will start the win the battle for control.
 
Sounds to me like you did almost everything right. I've landed my 8 in 30kt direct crosswind and it's a workout alright. In this one, Monday morning qurterbacking, I would say it's extremely important to get the stick FULL back to get maximum down force on the tail wheel. Even at 15-20 knots it exerts quite a bit of downward pressure and will help keep the little wheel from sliding.

Glad you didn't bruise any more than your ego and thanks for reminding us all how quickly thongs can go south.
 
Chains

I have loose chains on my 8 and a friend has the Rocket link. He has told me about his breaking lose in a strong crosswind, but the loose chains on mine allow full deflection without that happening. I'd suggest changing back to springs and chains and running them loose. Seems to work much better and not as twitchy. Glad nothing was hurt!
 
Steve don't beat yourself up. It is easy for the peanut gallery to Monday morning quarterback but we weren't there. It sounds like when the tailwheel broke free it happened so fast that a shot of throttle would have been too late and may have made the situation worse? I'm happy you and the airplane came away unscathed. I think you are on the right track inspecting the workings of the locking mechanism.
I may go do some cross wind practice. I'm getting old and rusty my reflexes are slowing down and am likely over confident as well. Thanks for posting and making me reflect and realize it.
 
With the tail wheel on the ground, it should be much more willing to go straight. Had I sucked the stick back hard, maybe the added downforce would have helped the tail wheel track the airplane straight....unless the swivel unlocked.

I think the tail wheel swivel unlocked. I think at that instant, I was doomed.

thoughts?

Steve,

Congratulations on not damaging the airplane or yourself; quite an accomplishment I'd say!

If the rudder was really deflected enough during the rollout to break the tailwheel loose then the tire must have been skidding big time? Seems like you might have felt this happening?

More likely perhaps that the tail wheel broke loose at a lesser angle, due to burrs either on the latching pin or on the inside radius of the control arm? I wrestled with my tailwheel not locking for some time before I realized the problem was not the pin but the control arm. Filed off some burrs near the corners of the notch that takes the pin and everything was good.
 
Whewwwww......bet the heart got going.
My initial thinking is that the ability of the rudder was simply exceeded. All airplanes want to weathervane. Once it starts followed by braking, the mass of the tail wins over braking and contributes to the groundloop.
Remember, you're already slightly cocked , the tail weighs more than the area between it and the pivot point, gear/engine, and a 'jackknife' begins.
As many groundloops I have experience or listened too, I can't recall any that were stopped once 'all **** broke lose'.
Humble pie.....is a scary aviation moment without scratching your plane.
I don't order it.....but eating it over my 47 years of flying everything made me a better pilot. Have thanked the lord a few times my 747 is a tricycle.
You did just fine....
Tip....if things get bad with choices nil...consider a judgement call , see if the tower will allow landing across the runway rolling out onto the taxiway. You pretty much have experience the aircraft limit. Anything more and the risk of damage increases rapidly.


H Retting
Old and no longer bold
 
I don't know if it was a factor or not, but from your description it sounds like it could be.....

I have been cautioning builders for many years (ever since there was a rocket link on the RV-7 prototype) that they need to confirm that when the rudder is hard over at the left and right stop, that it isn't rotating the arm on the tail wheel beyond the disengage point.
I believe a lot of the ground loops that have occurred over the years were caused by this. Chains and springs have enough slack that this isn't a problem.

This is one of the reasons I don't like the rocket style links, and why the RV-7 prototype was switched back to springs.

It is also possible for the tail wheel to break out under load if the arm and/or lock pin has gotten worn.

Glad to hear it turned out to be nothing more than a learning experience.
 
I watched a groundloop in a non-RV where the pilot kept the tail up and with brakes and power, attempted to control the direction. It was well off the runway and trying to go sideways when one wingtip touched and then it went over on its back.

Simultaneous brake and power did it for him.

I'm glad you're both ok and upright.

Dave
 
I don't know if it was a factor or not, but from your description it sounds like it could be.....

--snip--

It is also possible for the tail wheel to break out under load if the arm and/or lock pin has gotten worn.

Glad to hear it turned out to be nothing more than a learning experience.

Fearing this, I case hardened both the notch area and pin using Cherry Red hardening compound. I inspect both regularly and in over 200 hours of use I cannot yet detect any wear. And I am flying a Rocket which puts more stress on both pieces.

YMMV, but I believe this does indeed help.
 
Yes, I experienced tail wheel break out on landing, and the subsequent ground loop in my T-18. I simply got too close to the left edge of the runway and gave a big quick jab of right rudder, more than I'd ever given with this tail wheel, but just a brief moment. It went around just as you described. It was the first flight after I had disassembled, cleaned and greased the tail wheel post & lock pin assembly. This is a Van's style rod spring with a popular aftermarket tailwheel system. The chains are 'kinda tight'.
"Last thing worked on, First thing to break"
I got away with that back when I had the Maule tail wheel assembly and leaf spring on it. Go figure.
I'm wanting to mill the cam slot a bit wider to delay the release to full swivel, on recommendation of one of the Thorp gurus.
I did manage some pretty gruesome April cross wind landings in the RV-8, one @ Sedona pushed me over to the leeward side of the runway, but it stayed in control. Returning home, a gas stop @ Red Bluff had something like 19-Gust 24 70-80 degrees and I stayed in the center, but as we slowed to under 20 mph, the tail wanted to weather vane bad, trying to push us to the up wind side. Taxiing to the pump was difficult with the wind at our back. The RV-8 chains are "kinda loose"
I feel your pain. We don't get much cross wind experience in the Pacific Northwet, so trips to California, Arizona, Texas etc give me the Willies.
Aerobatics for me, are great. I just don't like strong crosswind landings. It's something for me to work on, like my 'off heading' slow rolls.
 
I have a different theory on this I would like to throw out for discussion.....

In a xwind, as the the aircraft slows, forward stick used to hold the tail up. As speed further decreases, more rudder will be required to hold alinement. Now when the tail finally drops, the rudder and more importantly, tail wheel (still locked) will be very displaced. When it touches down, it will swerve the aircraft until it unlocks.

This is somewhat opposite to what I think Steve is saying. To avoid this, one of two techniques could be used:

1) get the tail wheel on the deck sooner, rather than holding it off, while the rudder has more authority and therefore it and the tai wheel would be straighter.
2) If holding the tai wheel off as long as possible, neutralize the rudder just as the tail wheel touches down, so that the tail wheel is straight when it touches. Once on the deck, a straight tail wheel will hold the aircraft straight more effectively.

Is this (a severely cocked tail wheel at touchdown) a more likely explanation of what happened here and in other ground loops.

Isn't a less cocked tail wheel at higher rudder deflections what is going on in a "loose chains" rigged airplane??
 
Throttle

thoughts?

In my -8 I've learned to be ready on the throttle during roll-out. If a sudden gust hits and you suddenly find yourself out of rudder and opposite brake, a quick stab of throttle will give you a lot of rudder authority and straighten things right out. The one time I really needed this, it worked like a charm (in 580 RV-8 hours).

Skylor
 
I think Gary's thought is correct.

I have the link on my Rocket rather than the chains because I like the look. The problem with it is that you can't put the tail down with any rudder deflection or it can skid the tailwheel. The more rudder, the worse it is. I tend to center the rudder for an instant as the tailwheel comes down.

My link broke last year so I flew without any tailwheel steering for a couple of weeks. Really liked how that felt so I'm guessing that is how a set of loose chains feel. They allow the tailwheel to be straight with a large rudder input.

Russ
 
....confirm that when the rudder is hard over at the left and right stop, that it isn't rotating the arm on the tail wheel beyond the disengage point.

+1 for sure. My original tailwheel setup (with cables and springs, not a single-sided link) had the pin very near disengagement at full rudder deflection, and in fact it did go full swivel unexpectedly while landing in a good crosswind at OSH.

Today, a close look will find that the tailwheel steering attach points on the rudder horn have been moved inboard about 1/2" on each side. It allows full rudder deflection with less tailwheel deflection.

Side note; the ball bearing JD Air tailwheel I'm flying now seems to have a very firm lock pin. Can't say that about all tailwheels.
 
Thanks for the report Steve and glad it worked out ok with no damage to airplane.

The discussion about the tail wheel unlocking sure is interesting, have not had it happen but know its been close.

I am a total neophyte with 140 hours tail wheel time, most in the 8, so the envelope is adjusted accordingly. I try to not fly on gusty windy days. But somedays it just blows up and there you are, you have to land. The most I've survived is about 18 knots at 45 degrees to runway.

The approach and landing is all about technique. We try to figure out what works and stick with it. You've got a lot more experience in the 8 than I, so I pay attention.

Sounds like you did everything just right and in the end lost it anyhow, how come? Probably skidding tail wheel or it unlocked.

An unlocked tail wheel is a challenge. A few years ago I was getting a RV tail wheel check out in an RV-6. The IP signed me off but I though I did terrible. Later I found out why, the darn tail wheel steering was disconnected the entire check out. It was busted. No wonder I had such trouble keeping the thing aligned on take off and landing. Felt like I was wearing out the brakes.

So tail wheel steering is important.

The only thing about your flight that I do different and won't change is the approach. I don't slip but fly a crab and drop a wing and kick it straight to land. That's pure technique, carry over from other airplanes where a slip was out of the question.

Again, thanks for the report.
 
+1 for sure. My original tailwheel setup (with cables and springs, not a single-sided link) had the pin very near disengagement at full rudder deflection, and in fact it did go full swivel unexpectedly while landing in a good crosswind at OSH.

Today, a close look will find that the tailwheel steering attach points on the rudder horn have been moved inboard about 1/2" on each side. It allows full rudder deflection with less tailwheel deflection.

Side note; the ball bearing JD Air tailwheel I'm flying now seems to have a very firm lock pin. Can't say that about all tailwheels.

Dan, I think you are wrong on this one. The same linear displacement (the rudder pulling on the steering arm) will produce more angular rotation of the tail wheel with a shorter lever arm.

Think of a derailleur on a bike. A bigger rear gear is a lower gear ratio producing less wheel rotation for a given pedal travel.

NOPE. Re-read. You are right. Moving the attach point in on THE RUDDER HORN will produce less angular deflection of the tail wheel.

I think I will take a look at this on my steering link equipped 6.

Thanks Steve for prompting this thoughtful discussion.
 
Last edited:
I don't know if it was a factor or not, but from your description it sounds like it could be.....

I have been cautioning builders for many years (ever since there was a rocket link on the RV-7 prototype) that they need to confirm that when the rudder is hard over at the left and right stop, that it isn't rotating the arm on the tail wheel beyond the disengage point.
I believe a lot of the ground loops that have occurred over the years were caused by this. Chains and springs have enough slack that this isn't a problem.

This is one of the reasons I don't like the rocket style links, and why the RV-7 prototype was switched back to springs.

It is also possible for the tail wheel to break out under load if the arm and/or lock pin has gotten worn.

These are exactly the two things I am thinking about. Will inspect today.
A third factor that I will check today is that with full rudder deflection, and pushing pretty hard on it, I wonder if the pedal might touch the firewall with less than full braking. I'm sure I would have checked that when I built it, but maybe not including the flexing of the system under hard pushing.

For those that have brought up differential braking as part of the control process, the answer is 'yes, of course'. Honestly, after 30+ years of tailwheel flying, I don't think I can describe exactly what I do with rudder and brake -- its pretty subconscious. I think I start mixing in some brake at about half rudder, and add more brake pressure as I add more rudder beyond that. You are basically doing whatever is needed to fix any heading error. The startling thing about this event is how instantaneous it was -- everything was fine, rolling out nicely, I'm sure I had some mix of brake and rudder. And then it wasnt.

It was kind of a spooky thing, but I remember as it was sliding how quiet it was -- I could not hear any tire noise at all. I will be really furious if I check and find that with a hard push on the pedal, there isn't any braking available because the pedal is touching the firewall!
 
Last edited:
I have a different theory on this I would like to throw out for discussion.....

In a xwind, as the the aircraft slows, forward stick used to hold the tail up. As speed further decreases, more rudder will be required to hold alinement. Now when the tail finally drops, the rudder and more importantly, tail wheel (still locked) will be very displaced. When it touches down, it will swerve the aircraft until it unlocks.

This is somewhat opposite to what I think Steve is saying. To avoid this, one of two techniques could be used:

1) get the tail wheel on the deck sooner, rather than holding it off, while the rudder has more authority and therefore it and the tai wheel would be straighter.
2) If holding the tai wheel off as long as possible, neutralize the rudder just as the tail wheel touches down, so that the tail wheel is straight when it touches. Once on the deck, a straight tail wheel will hold the aircraft straight more effectively.

Is this (a severely cocked tail wheel at touchdown) a more likely explanation of what happened here and in other ground loops.

Isn't a less cocked tail wheel at higher rudder deflections what is going on in a "loose chains" rigged airplane??

Gary, I've experienced this behavior with tight chains, but less so with the rocket link. The link has enough stretch with an internal spring that when the wheel comes down, it straightens out some to minimize swerve. Loose chains would no doubt be better in this respect, but I don't like the lag time. I might think about going back to my Tail Lynx system.
I agree with you that doing this earlier in the roll-out also minimizes it.

Note also that the swerve, if any, is away from the wind, since that's what you are doing with the rudder.
In this particular case, I specifically noticed that there was no swerve when the wheel touched. An interesting thought though - what if it was turned so much that when it came down, it just started skidding, and then never regained traction. Hmmm. Again, I wish I could take a look at the marks on the runway. But now, in addition to it being a towered airport, it is also 300 miles away.
 
Last edited:
A question for Scott and Dan

One question for Scott and Dan, about setting up the wheel steering so that the cam is not close to releasing the pin to unlock the wheel at full rudder.
If it isn't, how would you ever get the wheel to unlock when you want it to?

Mine is set so that at full rudder, plus a bit of spring deflection in the link, it releases.
During roll out, as I mentioned, at full rudder, the link spring is deflected the other way, letting the wheel straighten some, so it should not unlock.

A bit of wear on the arm or pin, though, and it might. I will inspect today.
 
I think you only want it to unlock when you stand on one brake and drive the tw to a very large angle, like >70 deg.

For the victim, if you were holding everything in with a large rudder deflection and TW input during the rollout and the TW suddenly disengaged, you were doomed. One would have to be some mix of Chuck Yeager, Bob Hoover and Superman to react to that. Once the rotation started and the yaw inertia got wound up you were nothing more than a very interested passenger. There was nothing you could have done. There but for the grace of a well rigged TW go all of us.

Check your TW as suggested (I'm gonna do that too) and be very happy that you aren't doing an engine teardown or worse. I think you were damned lucky. I wouldn't worry about your pride. Better to have that dented than your wallet!
 
I think you only want it to unlock when you stand on one brake and drive the tw to a very large angle, like >70 deg.

But it can never get to a large angle if it doesn't unlock. If you go full rudder, and stretch the springs, and it doesn't move far enough to cam out the pin, that's it. Its locked. Its not going anywhere. (assuming its not worn enough to release with force)

its not supposed to release unless the cam action retracts the pin, and the cam action can't retract the pin unless the wheel can turn to that angle, and it can't turn to that angle unless the rudder deflection plus spring extension allows it to.
 
But it can never get to a large angle if it doesn't unlock. If you go full rudder, and stretch the springs, and it doesn't move far enough to cam out the pin, that's it. Its locked. Its not going anywhere. (assuming its not worn enough to release with force)

its not supposed to release unless the cam action retracts the pin, and the cam action can't retract the pin unless the wheel can turn to that angle, and it can't turn to that angle unless the rudder deflection plus spring extension allows it to.

No, you don't need a large rudder deflection. What you need is a large difference between the angle of the arm connected to the rudder and the wheel pivot (kingpost?). When you stomp on a brake, or you push the tail with your hand when pulling the airplane out of the hangar, it puts a sideload on the tire, which rotates the kingpost. If the kingpost angle gets beyond a certain limit the pin pops out. In fact I think that is the only scenario in which you want the pin to disengage. YOu don't want to be able to disengage it with rudder inputs - for the reasons discussed above. centered. I have never had mine disengage due to rudder inputs. That would not be good.
 
+1 for sure. My original tailwheel setup (with cables and springs, not a single-sided link) had the pin very near disengagement at full rudder deflection, and in fact it did go full swivel unexpectedly while landing in a good crosswind at OSH.

Today, a close look will find that the tailwheel steering attach points on the rudder horn have been moved inboard about 1/2" on each side. It allows full rudder deflection with less tailwheel deflection.

Side note; the ball bearing JD Air tailwheel I'm flying now seems to have a very firm lock pin. Can't say that about all tailwheels.

One question for Scott and Dan, about setting up the wheel steering so that the cam is not close to releasing the pin to unlock the wheel at full rudder.
If it isn't, how would you ever get the wheel to unlock when you want it to?

Mine is set so that at full rudder, plus a bit of spring deflection in the link, it releases.
During roll out, as I mentioned, at full rudder, the link spring is deflected the other way, letting the wheel straighten some, so it should not unlock.

A bit of wear on the arm or pin, though, and it might. I will inspect today.

Steve. I pondered this question also, specifically today as I sat astride the jacked up tail wheel, drill and 3/8" wrenches in hand. After doing some yanking on the tail wheel and rudder (gently ) I was able to convince myself that when turning an arc with full rudder, I think the internal link spring will be tension, limiting the tail wheel rotation. Stomping a brake should tighten the turn and put the spring in compression, allowing enough rotation to cam out the lock pin.

At least that is my theory I have moved the link mount point on the rudder post in about .6". Flying Thursday and will see if the theory is correct
 
Last edited:
Great job keeping the airplane from damage Steve - assuming you have cleaned up the seat already...... ;)

Been up at Stead, so just getting caught up here - I think you have plenty of interesting discussion already. What I can tell you from 1900 hours in the -8 is that when the crosswind gets that high, I generally try to land diagonally on the runway - that buys a little margin. I have had single-side link steering on all my RV's pretty much from the start, and have never had any issues with them versus chains - and I fly many other airplane's with chains, rarely noticing any difference.

With the number of hours you have on the plane, if you still have the original steering arm, I am going to join the chorus voting for a worn out notch - time to replace it. I have replaced them a couple of times (you can flip them if you still have two horns and add some life sometimes), and finally went to JDair arms - they seem to be harder, and wear out less (I have yet to wear one out).

Not much more to add - glad you and the airplane are fine. One of these days, I need to go out and ground loop one of the RV's, just to get it out of the way (if you know what I mean....).

Paul
 
Thanks for sharing

So awsome that you would share that here with us. I have made a few crosswinds that didnt turn out pretty and that was in C172's.

The tailwind thing was explained to me in 7eca training and he emphasised that the stick go Fully back as the tail dropped. Perhaps that was the thing missing, but hey you are safe, the plane is in one piece and you have shared it with others to make us all think! Thanks
 
No, you don't need a large rudder deflection. What you need is a large difference between the angle of the arm connected to the rudder and the wheel pivot (kingpost?). When you stomp on a brake, or you push the tail with your hand when pulling the airplane out of the hangar, it puts a sideload on the tire, which rotates the kingpost. If the kingpost angle gets beyond a certain limit the pin pops out. In fact I think that is the only scenario in which you want the pin to disengage. YOu don't want to be able to disengage it with rudder inputs - for the reasons discussed above. centered. I have never had mine disengage due to rudder inputs. That would not be good.

This is only true if there is enough slack in the linkage (whether chains or rocket link or whatever) to ALLOW the wheel to turn far enough for the cam action to release the pin.

If the link limits the motion relative to the rudder, then the wheel can not move unless the rudder moves, except for the spring or slop in the linkage.

If you don't believe this, then try this -- artificially limit the rudder travel to some smaller deflection, like say 15 degrees. Now try to release the tail wheel by pushing on it or anything else. Aint gunna happen. The linkage will prevent the wheel from turning far enough to release.

What you want is for full rudder PLUS deflection of the spring in the linkage from pushing the wheel farther than the rudder, to release the pin. If you get to full rudder AND full deflection of the spring, and it has not released, then it will never release. The wheel will not be allowed to turn far enough to release if it is restrained by the linkage.
 
Last edited:
Steve,
Nice description on what happened and your analysis. There is no greater teacher than experience, and I'll bet you'll be a little more ready next time - although you did exactly what what necessary. I'll agree with what Danny, Widget, and others said about getting the stick all the way back - on every landing... I'll also reinforce the idea that rudder (max) is king, followed by differential braking to maintain centerline. Doing this, you won't even know if your steering link pin, or spring is weak - or even broken! I landed my RV8 a few months ago in a strong crosswind and didn't notice that something was amiss until I was taxiing to my hangar. When I jumped out of the plane and checked my tailwheel, I found the rod end bearing on the link had snapped, and the steering link was dragging on the ground. Didn't really notice it on landing because I used nearly full rudder and differential braking, in which case the tailwheel would be beyond the release point. Taxiing was noticeably different though.
 
Found the problem

So, I just cleaned, inspected, and studied the tail wheel linkage.

No noticeable wear. JD will be happy to know that the newer, replacement steering arm is holding up very well. (the original one was a bit soft and showed some wear in a short time. JD sent me a replacement, and its good).

There is an asymmetry in the rocket-link style steering link, because the hole in the rudder horn is aft of the rudder hinge line, by MORE than the hole in the wheel steering arm is aft of the wheel pivot. This causes a different wheel rotation for the same rudder deflection in the two directions. If you have a hard time visualizing this, it is the same as the situation with an aileron bell crank geometry that gives differential aileron deflection.

(note this doesn't happen with chains because for each turn direction, the steering arm is pulled by the chain on that side, while the other chain goes slack)

So, this is what I have:
In one direction, the rudder gets to the stop, and the wheel has not released. Then, if you push on the wheel in the direction to turn it farther, stretching the spring in the link, the wheel releases with about 1/3 to 1/2 of the available stretch. (be nice if it were a little more).
In the other direction, the wheel releases before the rudder gets to the stop. Not good. and guess which direction that occurs.....yup, left.

The suggestion by Dan H to desensitize the tail wheel some by moving the attachment of the link inboard at the rudder horn is a good one for chains, but it won't work for the rocket link, because the "dog leg" bend in the rocket link hits the rudder. So, to accomplish the same thing, what is needed is a longer steering arm on the tail wheel.

If a longer steering arm is not available, I guess I will go back to chains.

Ultimately, I'm pretty sure this explains what happened. Steering to stay straight on the runway caused enough rudder deflection to release the tail wheel, and it didn't even take full rudder.:mad:

Good lesson for everyone to learn from.
 
Last edited:
One question for Scott and Dan, about setting up the wheel steering so that the cam is not close to releasing the pin to unlock the wheel at full rudder.
If it isn't, how would you ever get the wheel to unlock when you want it to? Mine is set so that at full rudder, plus a bit of spring deflection in the link, it releases.

You have it right. My setup just requires more stretching of the spring to unlock.

Note, I'm not using a link. I use two cables (tougher than chains) and pull springs. The link may not have enough spring travel. If so, don't move the forward attach point. Find stiffer springs to insert into the link.
 
Last edited:
At least that is my theory I have moved the link mount point on the rudder post in about .6". Flying Thursday and will see if the theory is correct

Be sure the link isn't hitting the rudder when you turn in the direction opposite the side that the link is on. E.i. if your link is on the left, be sure the link arm doesn't hit the rudder with full right rudder. Mine did when I moved the hole only 0.4" on the horn.
 
X-wind landing technique: steer for the ditch!

Hey Steve:

Tailwheels are not very trustworthy. That's why the tricycle gear is so popular.

"Steer for the ditch" is an old Beech 18 saying - if the ship starts drifting towards the ditch, turn the control wheel towards the ditch to use adverse yaw to keep the ship on the runway. Beech 18s have the most awful brakes, but very good ailerons (almost 40deg down deflection). That deflection is there to help keep the plane straight in a crosswind - the same applies to your -8 in a wheel landing. I only had my 18 off the runway once in ten years! Lotsa luck involved there...

You mention rudder use and brake use...but no aileron, or flap. If you used full flap, that is not the best plan with a tailwheel ship. If you do like to use full flap in a X-wind, get rid of it when the wheels touch - but it is best to use either none, or maybe 1/2, and dump it when the wheels touch.

Ailerons are very effective on the RV series, and really help with X-winds. Flaps dilute this effectiveness - as they do on any ship. Also when you dump the flaps you will feel the tailwheel settle on the runway immediately.

Landing at an angle into the wind is another trick - if the runway is wide enough. Landing with less energy is important as your 'runway length' is very short. Try the Stearman technique to develop this skill.

You are one of the lucky ones (no damage). Get back out there and try a few techniques that were used to get Stearmans on the ground and on the runway in similar circumstances. Remember these were big box kites with weak brakes, but plenty of rudder and aileron. The 2-wheel technique taught on those ships (land on the upwind main wheel and tailwheel) will work fine on your 8 - but it takes practice to learn and more practice to keep the skill. This type of landing gets the ship on the ground with less energy than a wheel landing, so it can be used in the diagonal landing technique easily.

But - if there is any doubt about the X-wind component, check if you can keep the plane aligned with the runway centerline while still in the air. This will also tell you the bank angle required, the rudder pressure required, and if you need to land diagonally.

In any case - steer for the ditch!

Carry on!
Mark
 
Tailwheel

Regarding the dual tailwheel steering springs there are two types in common use. The plain type have a history of coming off, especially if you are unfortunate to have a serious tailwheel shimmy. Pitts and others use what most call a boat tiller spring, attached at both ends with links made from wire. They are pretty reliable but the links do wear and need to be replaced, usually every 3-500 hours. I had this type springs with the Aviation Products 4" tailwheel on my Pitts. An even better setup is a locking tailwheel. It really helps the landing but is miserable to taxi when unlocked and free to swivel.
I combined the best of all options with a single spring link to a tailwheel that does not swivel. It turns with the rudder but does not swivel. A cockpit control unlocks a pin to make it full swivel for tight turns while taxiing, almost all taxiing can be done with the pin engaged and normal steering.
 
steering for the ditch

Thanks Mark. Interesting reasoning.

I keep the stick all the way into the wind, but I never thought of it as adverse yaw to help steer, I always just think of it as being sure the upwind wheel stays on the ground. Mostly from gliders, but also from Skywagons that have less aileron power than would be desired. I found myself with the yoke on the stop and the upwind wheel still trying to come off the ground more than once in the C-180. Flaps seem to make it worse, just as you say.

I had one "notch" less than normal full flap, which means 20 degrees rather than the normal 40 degrees.

In the Citabria, it was religion for me to land "two point" meaning upwind main wheel and tail wheel, just as you describe. But the -8 seems to like to wheel land much more, so that's my habit now. That said, I got the upwind wheel down, then the other main, then the tail wheel, and all was well. And all stayed well for a few seconds after that.

Might have helped to have FULL back stick rather than partial back stick -- a habit I let lapse when I started wheel landing. I will renew that habit. But ultimately, when the tail wheel swiveled, the departure was so sudden, and so abrupt that no one could have reacted sufficiently. As I said in the OP, my foot was way ahead of my brain. It was too late. All that was accomplished with full rudder and full brake was to delay the departure until much of the energy was scrubbed off. But I was still basically a passenger at that point.

Thanks to everyone for comparing notes. A useful refreshing of various ideas and techniques, healthy for everyone.
 
I combined the best of all options with a single spring link to a tailwheel that does not swivel. It turns with the rudder but does not swivel. A cockpit control unlocks a pin to make it full swivel for tight turns while taxiing, almost all taxiing can be done with the pin engaged and normal steering.


Your set-up sounds kind of like the P-51 Mustang. I always thought that was ingenious and the best of all worlds. If the stick is aft of neutral, the tail wheel is locked, straight ahead. If the stick is forward of neutral, the tail wheel is free swiveling.

I would love to have a set-up like that.
 
Tailwheel

Your set-up sounds kind of like the P-51 Mustang. I always thought that was ingenious and the best of all worlds. If the stick is aft of neutral, the tail wheel is locked, straight ahead. If the stick is forward of neutral, the tail wheel is free swiveling.

I would love to have a set-up like that.
Yes like the P51 but I chose not to link the release to the elevator controls. The Aviation Products worked so well on the Pitts that I never changed that. Just took it apart and inspected it every couple hundred hours. Never had it unlock in about 700 hours with a lot of landings and a lot of crosswinds.
The Russian Sukhois have a locking tailwheel same as the Twin Beech, no steering with it locked.
 
Yes like the P51 but I chose not to link the release to the elevator controls. The Aviation Products worked so well on the Pitts that I never changed that. Just took it apart and inspected it every couple hundred hours. Never had it unlock in about 700 hours with a lot of landings and a lot of crosswinds.
The Russian Sukhois have a locking tailwheel same as the Twin Beech, no steering with it locked.

Looking at the Aviation Products website, all I see are steerable and swiveling tailwheels, just like standard Vans set up. The main variable is the type of tail spring. I don't see any with controlled locking pins. Can you post a model number or link?
 
Carl;
Yes there are few T-18s, but they have a tight little community. I was going to sell mine after I got the RV-8 going, but it turns out my mentor at work and other friends ( and now new friends) have them too.
No, it wasn't me at Bremerton. I don't know who, and if the ship was unharmed, we'll probably never find out. :eek:
It takes a kind of bravery to admit a ground loop online. The discussion it brought up could prevent some ground loops in the future.
 
Last edited:
One question for Scott and Dan, about setting up the wheel steering so that the cam is not close to releasing the pin to unlock the wheel at full rudder.
If it isn't, how would you ever get the wheel to unlock when you want it to?

Only the old timers in the RV community will remember that the original RV tail wheel assembly was just a simple yoke with no full swivel capability. It worked great but it was a pain to push a tail dragger RV backwards into a hangar.
Someone began selling an after market full swivel tail wheel, and because it was so popular, something similar was incorporated as standard into all the kits.
As designed, the only time the tail wheel is intended to unlock is with someone on the ground pushing the tail and pivoting the tail wheel well beyond the deflection angle that is possible by pushing on a rudder pedal.
This can happen when using the standard chains and springs because there is typically some slack in the chains as well as the compression of the spring that has to be over come. This all results in some level of rudder movement before you get to the point that further movement induces a steering force on the tail wheel.
So with chains the rudder can never turn the tail wheel far enough to hit the cam-out point, but the chain slack and spring compression allow you to manually deflect the tail wheel far enough to hit the cam out point.

BTW, a lot of people say they don't like the chains because of the slack they can feel while taxiing. The slack is actually the biggest benefit of the chains... it provides for a dual rate steering. When moving at high speed on the runway with the tail wheel still on the ground, it allows for a small amount of rudder movement (often times all you are using) without inducing any steer on the tail wheel. This makes the steering at high speed much less sensitive, but then still brings the wheel steering into play when you need it. The rocket style links all tend to be very sensitive at high speed, along with the potential problem of causing the tail wheel to break into full swivel mode with full rudder input.
 
Pretty much the same thing happened to me on my RV 4 last year with 25+ knots 40 to 60 degs off the left wing.

With all of the severe up and down drafts and with the X wind component I was at times with full right rudder, left wing down and side slipping. Kept myself line up down the center line, left wheel down and held it, right wheel down (still full R rudder) followed shortly by the tail wheel while congratulating myself.

Wrong, as soon as the tail wheel was down I was heading 30+ degrees off the run way still with full right rudder and feeding in brake (rubber is still on the runway). I thought for sure I was going to drag a wing tip or hit a light. No audience except my GF in the back. She thought all was well until the F bomb came out as we were now heading to the fence. Luckily no damage.

Back to the hanger and analyzing what happened and what I now realize that once you have near to full rudder deflection the tail wheel will come unlocked and be full castering until you release it enough and it comes back towards center and let it lock. I have the stock springs, tail wheel etc. Everything passed inspection but I ordered new parts and replaced them. I realize it is the design of the locking mechanism and if you're using full rudder before touchdown and you're not able to relax your rudder it will remain in the free swiveling mode.

On the ground lift your tail and prop it up, push in your rudder until the tail wheel unlocks and now its clear on how it works. It won't lock again until you relax your rudder back to a certain point from full deflection.

Another small contributing factor is that the RV 4 has a smaller tail and rudder. Once your tail is on the ground the fuselage does blanket the rudder further reducing effectiveness.

Tim
 
This is only true if there is enough slack in the linkage (whether chains or rocket link or whatever) to ALLOW the wheel to turn far enough for the cam action to release the pin.

If the link limits the motion relative to the rudder, then the wheel can not move unless the rudder moves, except for the spring or slop in the linkage.

If you don't believe this, then try this -- artificially limit the rudder travel to some smaller deflection, like say 15 degrees. Now try to release the tail wheel by pushing on it or anything else. Aint gunna happen. The linkage will prevent the wheel from turning far enough to release.

What you want is for full rudder PLUS deflection of the spring in the linkage from pushing the wheel farther than the rudder, to release the pin. If you get to full rudder AND full deflection of the spring, and it has not released, then it will never release. The wheel will not be allowed to turn far enough to release if it is restrained by the linkage.

ok this is not how I understood it to work. But before I say something dumb (again?) I will go look at it and verify :rolleyes: On the airplane I currently fly I have a Lang TW, which works great. I also have a -4 project but that TW has not currently gone more than about 0.1 kts. I certainly want to understand this before I fly the airplane. Thanks for posting it.
 
Only the old timers in the RV community will remember that the original RV tail wheel assembly was just a simple yoke with no full swivel capability. It worked great but it was a pain to push a tail dragger RV backwards into a hangar.

[snipped]

I remember; I've owned two -4s with that setup; still flying one of them.
And....

[heresy warning]

I don't use chains at all. Started that way in a Swift a long time ago, and when I bought the 1st -4, couldn't stand the 'push-pain'. Everybody says "What if you lose your brakes?" Well, I'll probably ground loop. On the other hand, how many times have people had an issue with a tailwheel, or chains/links, or breakaway link, *with good brakes*, and not groundlooped? (Evidence: the subject of this thread.)

The only times I've departed the runway unplanned was when I was a very green student in a Luscombe (heel brakes), and in the same Luscombe after an IA worked on the tailwheel and "Kinda got it working" (with the unstated meaning that it locked in one direction, but not the other).

Everybody has a list, conscious or unconscious, of what they can tolerate, and what works for them. For me, never having tailwheel steering (for over 20 years) has meant that I'm always ready with the brakes.

I did own a T-18 for a while, before the Swift, that had a hard link to the tailwheel. It can work, but you need to know it's there & be ready with brakes & to center the rudder when the tailwheel touches.

Worth what you paid; etc etc.

Charlie
 
Tailwheel

Looking at the Aviation Products website, all I see are steerable and swiveling tailwheels, just like standard Vans set up. The main variable is the type of tail spring. I don't see any with controlled locking pins. Can you post a model number or link?
Sorry if I wasn't clear. Pitts was standard full swivel. Steerable/locking was on two other airplanes and were homemade except for the actual generic wheel/tire.
 
Dan, I think you are wrong on this one. The same linear displacement (the rudder pulling on the steering arm) will produce more angular rotation of the tail wheel with a shorter lever arm.

Think of a derailleur on a bike. A bigger rear gear is a lower gear ratio producing less wheel rotation for a given pedal travel.

NOPE. Re-read. You are right. Moving the attach point in on THE RUDDER HORN will produce less angular deflection of the tail wheel.

I think I will take a look at this on my steering link equipped 6.

Thanks Steve for prompting this thoughtful discussion.

First time I have seen an actual "post loop". Good thing all turned out ok full circle ended up back on the path. :eek:
 
Back
Top