What's new
Van's Air Force

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

Suspension travel is the key to RV-7A nose gear collapse

Column buckling?

<snip>
Simply put so that I could understand the issue without a math degree it was explained to me that the steel used in the nose gear leg when put into a press and having rearward force applied showed an interesting characteristic. At around 1,500-1,700lbs of force the integrity of the metal changed and the gear leg effectively became a noodle. With this force applied they could move it around at will as if it had no resistance to offer. As the force was reduced the strength returned and the gear leg acted as intended.
<snip>

Bryan,

Your explanation sounds like column buckling. It seems quite possible that the slope of a particular bump could put the gear leg largely into compression, and the bending momemt Raiz has described is just what's needed to upset the balance of the column. Once a column starts to go, that's pretty much all she's wrote.

Good analysis, Raiz.
 
this looks like a lightly damped system

Great work Raiz - so your model says that a 2" bump that get run over at 10kts with no baking utilizes 60% of the yield strength of the gear leg. That is a surprising result, and makes me wonder about what happens if you hit it faster or slower? Is 60% as a bad as it gets, or is there a bump size / bump shape / speed combination that puts the gear leg into failure without the nut ever touching the ground?

Another intersting thing that I noticed was on your graph - the fuselages second bounce is very nearly the same amplitude as the first. I guess that means there is very little damping in the system. I assume if you let the model run longer the fuselage continues to bounce merrily along?

I wonder if your model can be modified to include the wooden gear leg stiffener that some builders use? I think that would provide some damping. It would be a very interesting analysis.
 
Last edited:
Making the model even more useful

A number of flip-overs have been on grass. I speculate without any supporting facts that putting the nose wheel into a hole is a common fail/flip scenario.

The model would be very useful if it could also look at that. That is where there is suspicion of an encounter between the nut and the ground. Checking a series of depths would also be useful. I suspect there is a chart that could be drawn - an envelope - that could define how big a hole and how fast you hit it which would be near or beyond failure.


Thanks for your good work.
 
i'm not an engineer, so you can explain this to me if its a bit off base, but isn't this just a bunch of theory? what if there is a variable that you do not know about and don't have plugged into your model? You will never know without running actual ground tests with recording equipment, right?

More or less, yes. That's what I mean when I say it's unvalidated. However, that's not to say its not useful, as modeling and experimentation normally go hand in hand when you are trying to solve a difficult technical problem. Tests are expensive, so it is common to build a model, validate it with a few experiments and then explore the limits quickly and cheaply using the model.

A good flying analogy would be navigation by dead reckoning using just a compass and clock above a broken overcast. Provided you don't make any mistakes in your planning or flying and provided the winds are as forecast, you will arrive at the intended destination. It is however at risk of a single error somewhere, so it's much more reliable if you can identify a few landmarks through the gaps in the clouds. That is what I mean by validation and it can be by instrumenting an aircraft or by checking the results against specific cases.

Raiz
 
Tire Spring Rate

Raiz,
I am curious how you modelled the spring rate of the tire. I have a copy of Pazmany's book "Landing Gear Design For Light Aircraft" in which he reproduces load/deflection curves for various sizes of Type III tires. The source is not credited, but it was one of the aircraft tire manufacturers. There are four graphs which summarize their test results:
1. Static load/deflection - low rate loading - isothermal compression.
2. Impact load/deflection - rapid rate loading - adiabatic compression
3. Impact energy/deflection - rapid rate loading - adiabatic compression
4. Contact area/deflection
Graph 3 seems most appropriate to this investigation.

Kent Byerley's videos show a substantial fore/aft movement as well as rearward rotation of the vertical portion of the nose gear leg as the tire rolls over the grass surface. Does your model reflect this?

Can you describe your model methodology in a little more detail? I don't think this has been covered. Is it an FEA model, a spreadsheet model or...?
Thanks.
 
When my gear leg flexed back it happened a couple of inches under the intersection fairing just below the cowl. The gear leg fairing broke at that location and the back of the nosewheel fairing almost hit the bottom of the cowl. As quickly as it happened it snapped back into postion. This happened while I was parking and my father in law and my friend were standing a few feet away watching from the front of the plane. From the pilots seat I didn't even experience a dip in the attitude of the plane. I had no idea that it had happened until shutdown and being told. Looking back now at Kents videos there is a definite flexing a few inches under the cowl in the are where mine gave up when he was rolling across grass.
 
Great work Raiz - so your model says that a 2" bump that get run over at 10kts with no baking utilizes 60% of the yield strength of the gear leg. That is a surprising result, and makes me wonder about what happens if you hit it faster or slower? Is 60% as a bad as it gets, or is there a bump size / bump shape / speed combination that puts the gear leg into failure without the nut ever touching the ground?

Another intersting thing that I noticed was on your graph - the fuselages second bounce is very nearly the same amplitude as the first. I guess that means there is very little damping in the system. I assume if you let the model run longer the fuselage continues to bounce merrily along?

I wonder if your model can be modified to include the wooden gear leg stiffener that some builders use? I think that would provide some damping. It would be a very interesting analysis.

At very low speed, the airplane rides up over the bump but at higher speeds there is insufficient time for the fuselage to move out of the way and the stresses are higher. Bigger bumps always cause a higher stress and I've put a failure example in below. Bump profile also has an impact, especially steepness, but it is mostly bump height that matters, which is why I introduced this thread by saying that suspension travel is the key.

So here is a failure scenario. 10 knots, 3.5" high bump:
graph6.jpg


And here is another way to break it: 10 knots, 2.5" bump but moderate (0.15g) braking:
graph7.jpg


Notice that, in both examples, the nut stays clear of the ground until the leg gives way. These are both 1800lb AUW and 375lb nose load.

In case it is not clear (because of the particular examples above) you can also fail it on the first bump, especially if the bump is relatively steep. However, the reason the second impact is often worse than the first is because the first bump acts like a launch ramp.

There certainly is very little damping in the model (and in the airplane too) but when you look at the damping forces generated by that amount of wood and compare them to the other forces at work, it is clear that it is not going to have a significant effect over the period of a single bounce. What it possibly does help with is the back and forth shimmy-like motion but the model is steering me away from the idea that that jittery nose leg motion is linked to the failures.
 
Raiz,
I am curious how you modelled the spring rate of the tire. I have a copy of Pazmany's book "Landing Gear Design For Light Aircraft" in which he reproduces load/deflection curves for various sizes of Type III tires. The source is not credited, but it was one of the aircraft tire manufacturers. There are four graphs which summarize their test results:
1. Static load/deflection - low rate loading - isothermal compression.
2. Impact load/deflection - rapid rate loading - adiabatic compression
3. Impact energy/deflection - rapid rate loading - adiabatic compression
4. Contact area/deflection
Graph 3 seems most appropriate to this investigation.

Kent Byerley's videos show a substantial fore/aft movement as well as rearward rotation of the vertical portion of the nose gear leg as the tire rolls over the grass surface. Does your model reflect this?

Can you describe your model methodology in a little more detail? I don't think this has been covered. Is it an FEA model, a spreadsheet model or...?
Thanks.

Sorry terrye but I interpret Pazmany's graphs differently. Graph 3 is simply the area above the curve from graph 2, so they are effectively the same. Pazmany explains this in his book. Nevertheless, to answer you question, I actually used the constant pressure method, because that is what the FAA used in their RV nose leg study. However, I tend to agree with you that 3 (or 2) would be a better choice. The constant pressure method will result in the tire bottoming earlier but it doesn't make much difference, because the tire bottoms out before the leg fails, whichever of these models you use.

Kent's video is a classic and my model predicts a similar behaviour. However, unless Kent wants to run his plane over a big bump, I'm not going to claim equivalence. ;) only joking Kent. It's not shown on the graph but the model is showing horizontal displacements at the axle about 50% higher than the vertical displacements and a rotation of around 4 degrees over a 2" bump. But, I think Kent's video is showing movement due to the rolling resistance of softer ground as much as due to vertical bumps.

The model is a bit of a hybrid to deal with the different aspects of the problem. The model of the actual leg does use a finite element approach but it is implemented in excel. The simulation (that produces the recent graphs) is a numerical solution of a limited set of the motion equations (pitch and horizontal motion). I've modelled the tire using the approach used by the FAA in their RV nose gear study except for the interaction with steps, which used the Pacejka tandem cam method. It also includes rolling resistance and the wheel rotation speed shanges that occur with changes in tire deflection. Did I answer your question, or am I on the wrong track?

Raiz
 
A number of flip-overs have been on grass. I speculate without any supporting facts that putting the nose wheel into a hole is a common fail/flip scenario.

The model would be very useful if it could also look at that. That is where there is suspicion of an encounter between the nut and the ground. Checking a series of depths would also be useful. I suspect there is a chart that could be drawn - an envelope - that could define how big a hole and how fast you hit it which would be near or beyond failure.


Thanks for your good work.

hevansrv7a,

You are right, potholes (as we call them in the UK) are worse than mounds, especially if the falling nose meets the ground coming up on the far side of the hole. Here is a 2" deep hole, 3m wide (the mirror image of the 2" bump). It is almost enough to fail the leg without any braking.
graph8.jpg


Here is another example of a 1 1/2" deep flat-bottomed pothole with moderate braking. This is what I think may have happened to G-CDRM at Croft Farm, which is on YouTube here: http://www.youtube.com/watch?v=NfaCGc16jQ0

graph9.jpg


Notice that, in the video, the leg failure appears to occur on the second dip, just as the model predicts. I would be very interested in any further examples that help to validate the model in this way.

The nut still stays clear of the ground in these examples. The only time I can get it close to the ground is when the bump is high and steep.

Of course, to suffer from this problem whereby the falling nose hits the rising ground, the speed and width of the hole must match up, which must be relatively rare. On the other hand, it may explain why most landings are accident free.

Raiz
 
Pazmany graphs

Sorry terrye but I interpret Pazmany's graphs differently. Graph 3 is simply the area above the curve from graph 2,

You're right, finger trouble, I meant Graph 2 is the most appropriate for this investigation. Unfortunately Pazmany does not include the 11x4.00-5 tire that we are using. I tried to find some data on it, but no luck. I found that Lamb is now Cheng-Shin and there didn't seem to be an English language button on their website.

I assume you are using the 5.00-5 Type III 6PR data?
 
You're right, finger trouble, I meant Graph 2 is the most appropriate for this investigation. Unfortunately Pazmany does not include the 11x4.00-5 tire that we are using. I tried to find some data on it, but no luck. I found that Lamb is now Cheng-Shin and there didn't seem to be an English language button on their website.

I assume you are using the 5.00-5 Type III 6PR data?

terrye, I was also unable to find good data on the 11x4.00-5 tire but the method used in the FAA study can be applied to any tire size. I therefore modeled the 11x4.00-5 tire (not the 5.00-5/6 tire) using this FAA method. However, it is a very simple method in that it does not model the adiabatic behaviour of the gas. I will look to take this into account by refining the FAA method in a future iteration of the overall model. My guess is that this will show that the tire is less inclined to bottom out but that the bump required to fail the leg is not significantly affected.

Thanks for the input BTW. It can be difficult to spot these things without critical review of this nature.

Raiz
 
Strategies

This thread is full of excellent data. I am not an engineer, so I am a little lost with what to do with all of it. Assuming the data is 100% accurate, what could a builder do to improve the situation. I have seen one thread concerning installing a RV 10 leg. What other options might be considered?

A kevlar/carbon fiber wrap on the leg?

An additional component to the leg like a rod or dowel?

What would be the downside to those changes or any others?

Just looking for knowledge.

Thanks
 
This thread is full of excellent data. I am not an engineer, so I am a little lost with what to do with all of it. Assuming the data is 100% accurate, what could a builder do to improve the situation. I have seen one thread concerning installing a RV 10 leg. What other options might be considered?

A kevlar/carbon fiber wrap on the leg?

An additional component to the leg like a rod or dowel?

What would be the downside to those changes or any others?

Just looking for knowledge.

Thanks

Scott, that's a good question, but not easy to answer. It's difficult to install an RV-10 nose leg on the 2 seat airplanes as it is quite a different design - it uses a rubber spring at the top. The nose leg is made from 200ksi steel, so to add significant strength would need a significant amount of material very well attached - which increases the weight on the nosewheel (which increases the chance of failure). A carbon wrap may be able to add some useful strength, but how much to add and where (same amount all the way up, more at the bottom - I would have no idea) then where would those loads then go? Would the engine mount then start to break?

How about a completely carbon/glass nose leg (with a streamlined profile) with the strength and stiffness tailored to give acceptable bending characteristics? Is it possible - no idea! But bound to be more expensive than the factory leg.

I think the take away is that there are circumstances where the standard nose leg can become overloaded when a C172 (or similar) might have coped - such as potholes or bumpy grass strips, especially when there is a lot of braking going on, or the static load on the nose wheel is high (say >300lb). One solution is to modify the airplane, the other is to modify your flying habits - such as make sure the load on the nose wheel is light if you use rough runways and don't brake hard. I've operated my 6A from a bumpy 2000' strip for the last 3 years without any nosewheel problems.

Pete
 
column failure and column modification

Raiz,

On a quick run through, I did not notice discussion on Euler/column failure and wondering if that is also taken into account on the graphs provided. Also, are you using a constant area tube or variation in dia and wall thickness? (Maybe Richard can provide some indication of variation)

Given that you have most of this in an excel sheet,.. any chance you can generate a graph of how much difference in diameter/wall is required to move you to a point that "std" holes and bumps are no longer a threat. Kind of a moment of inertia to bump size relationship. Guessing some significant improvement could be made with minor increase in weight, and fairing would result in negligible impact on speed. Might be interesting to see what "minor changes" might yield before we move to fixes by use of unobtainium. (then we can start discussing gear mount and firewall cracking)

I'll add my thanks for the work and explanations.
 
Think9a,

ok, the easy questions first: Yes buckling is taken into account and I have modelled a non-constant cross section, solid leg based on measurements of an actual RV-7A leg (new type) made by Mike Cencula (thanks Mike).

Now the tricky one: Adding strength is almost certainly the WRONG direction to go. Instead we need more suspension travel. If we were to add strength by making the leg a larger diameter and hollow, we would reduce the suspension travel i.e. it would be less flexible.

Now I am really struggling to find the right way to explain this, so let me try two different ways. First the practical explanation: Assume you have 4" of suspension travel before the leg breaks (as we do) and imagine running over a 4" bump fast enough that there isn't time for the fuselage to lift out of the way (as is the case). The leg is deflected to its maxium of 4" i.e. it is about to break. The load on the leg will be about 2100 lb and, since the static load on the leg is 375 lb, the fuselage will get an upwards shove equivalent to 5.6 g (2100/375). Now assume we make the leg stronger - say 3000 lb - but the maximum deflection drops to 3". Obviously, a 4" bump will break the leg but, in doing so, it will have exerted a 3000 lb load on the fuselage, which equates to 8 g. Consider the opposite approach, a leg with 6" of travel and a maximum load 2100 lb. At the top of the 4" bump, it is only loaded to 2/3 of its maximum (1400 lb) and the acceleration would be just 3.5g.

Now the engineering explanation: At the most basic level, the role of the suspension is to store energy. In the case of a landing aircraft, it must store the kinetic energy of the sinking aircraft (half mass times vertical velocity squared). In the case of a bump, it is equal to the potential energy of the aircraft when lifted by the height of the bump (mass times bump height times gravitational acceleration). Note that both of these are proportional to the mass of the aircraft. The suspension is essentially a spring and the energy stored in a spring is equal to half (peak) force times (peak) deflection. To accomodate the kinetic or potential energy, the designer of the aircraft can choose a stiff suspension (high force and low deflection) or a soft one (low force, high deflection). In practice, he picks a compromise between the high accelerations caused by a stiff suspension and the addition space and complexity associated with high deflection. The RV-7A leg has a moderate energy storage capacity (i.e. less than that required for a certificated aircraft) and a fairly stiff gear and, as a result, a low suspension travel.

In my view, the RV-7A leg already has enough strength and I wouldn't want the problems of beefing-up everything attached to the leg, which would be necessary if the leg were stiffer. I therefore believe we need a gear design that can accomodate more travel.

More travel either needs (i) a leg material with lower ratio of elastic modulus/strength; or (ii) a mechanism (e.g. RV-10 or C172); or (iii) a different shape of leg (i.e. flat instead of round).

One possibility for a material with low modulus/strength would be a glass fibre composite rod but that would be relatively expensive and complex to attach. The "RV10-like" pivoted leg and rubber spring is probably the front running mechanism provided it can be fited within the confines of the cowl. In terms of flat legs, I didn't anticipate this when I built the model, so it will take me a while to incorporate it.

More travel also brings with it other challenges - prop clearance and shimmy, for example, so they should not be forgotten.

Raiz
 
At the most basic level, the role of the suspension is to store energy.

Raiz

You can't store energy forever. So shortly after the landing gear "stores" the energy, ( I would say: "absorbs" the energy) the energy must be dissipated somehow.

A good example is the landing gear on US Navy Carrier aircraft. They have relatively long throw, and lots of hydraulic components to absorb the energy.

A spring steel "strut" can not "absorb" any energy, and must return it in the form of springing back, or "bounce".

The energy has gotta go somewhere. It either throws the plane back into the air, or it heats up some hydraulic fluid. Or, ............

There ain't no free lunch. ;)
 
You can't store energy forever. So shortly after the landing gear "stores" the energy, ( I would say: "absorbs" the energy) the energy must be dissipated somehow.

A good example is the landing gear on US Navy Carrier aircraft. They have relatively long throw, and lots of hydraulic components to absorb the energy.

A spring steel "strut" can not "absorb" any energy, and must return it in the form of springing back, or "bounce".

The energy has gotta go somewhere. It either throws the plane back into the air, or it heats up some hydraulic fluid. Or, ............

There ain't no free lunch. ;)

I agree. In the case of the RV-7A nose gear, it bounces.

Raiz
 
Question

So, Raiz-

If a relatively large "bump" causes the nose gear to fail in your model, why aren't the nose gear failures in the mode of nose gear "collapse", where the nose ends up on the ground with the nose gear failed upward and forward into the bottom of the cowl and into the prop?

Does your model predict the failure of the gear back under the plane, with the following "pogo" manuver that results in the plane flipping over on its back?

Trying to follow you engineers along on your disucssion, but not really keeping up!:eek:
 
what about tie down holes and "steps" from shifting slabs?

Raiz,

got it on increaseing strength,.. did not pick up the total input back into the plane.

We have some ramps from concrete slabs where they have moved in relation to each and there are some beginning to be sizeable steps. Also what about the step function when you hit a recess for the tie down, not a real deep hole,.. but fairly sharp edges. (some where I go look like they were planning on tying down a commercial jet).
 
RV-6A better than RV-7A?

I lack the engineering skills to critique the discussion above, but I think I can offer evidence that there is more to it. All of the above may well be true, but it's not likely the whole story.

The RV-6A was not noted for flip-overs. It still is not if this forum is a good survey. If the 8A has the problem, I'm not aware of that, either. Thus it is reasonable to infer there is something different about the 7A and 9A which makes them more vulnerable to the problem. The recent SB for which most of us changed the forks did not change any of that but did shorten the down portion of the leg and, I think, made the entire thing a little stiffer (it feels that way to me).

Since the nose tire is the same size, we can rule that out. For the 7A and 9A, the gear legs are longer and the airplanes when not flying sit more nose-down.

When a small wheel hits something tall enough (a hole or a step) that it cannot roll over then bad stuff begins to happen. Ever try to roll your rolling luggage out of the elevator by pushing instead of pulling? The problem here is geometry. Tire inflation can play a big role, too, since a higher inflation keeps more of the size of the tire and a bigger circle rolls over a bigger step.

Since the 8A is a whole different shape (longer, etc), I'll ignore it for now. The 7A and 9A center of mass must be in a position to encourage the pole-vault effect. We know this because it happens.

Imagine that you had a Van's side-by-side with a very low tail and your nose wheel stuck but the gear did not totally fail. The center of mass would not tend to rise but might even drop and lever the wheel out. Now imagine the 7A as it is with the nose down slightly and the center of mass above the line of the top of the nose leg. A force sufficient to stop the nose wheel will force the center of mass to begin an arc upward and forward. This is easier to imagine if you picture a 7A with a taller main gear and a shorter nose gear; now you are half way to a flip over without even trying. If the nose leg collapses and the airplane nose comes down to the ground, then it may flip over it.

I'm not suggesting making the nose gear longer. I do wish the main gear and the nose gear were shorter, like the 6A. I wanted to change over, but the diameters of the legs are wrong. It would take all new parts. I would gladly give up some x-wind performance for better safety on grass and rough fields.

Hi hevansrv7a, your post got me thinking and I ran some RV-6A cases on the model. On an equal nose wheel weight basis, the RV-6A results are similar to the RV-7A. This surprised me, because I also had the impression that the RV-6A has fewer problems.

However, if the model is right in predicting that it is lack of suspension travel that is important, we should expect similar results, as these two aircraft share the same leg.

Therefore, in understanding whether the model is correct or not, it becomes important to know for sure whether the accident rate is different for the different models. Do you (or does anybody) have any definitive data on this?

Raiz
 
So, Raiz-

If a relatively large "bump" causes the nose gear to fail in your model, why aren't the nose gear failures in the mode of nose gear "collapse", where the nose ends up on the ground with the nose gear failed upward and forward into the bottom of the cowl and into the prop?

Does your model predict the failure of the gear back under the plane, with the following "pogo" manuver that results in the plane flipping over on its back?

Trying to follow you engineers along on your disucssion, but not really keeping up!:eek:

PCHunt, this must be a misunderstanding, because the model does predict the leg will fold under in all cases when the aircraft is moving forwards. It predicts the point of initial yielding will be either just above the castor bearing or at the thinnest part of the leg roughly above the axle. In either case, the castor nut will touch the ground soon after the initial failure and the nose-over follows on from there. I haven't modeled the actual pogo movement, as the game's over once the leg yields.

Raiz
 
... I ran some RV-6A cases on the model. On an equal nose wheel weight basis, the RV-6A results are similar to the RV-7A. ...

Raiz, Is that a valid assumption? Because of the different main leg geometry (7A sits much higher) I believe the max static load that most 6As fly at is significantly lower than the typical 7A load - I don't have any data at the moment but will try to get some.

Pete
 
About the 6A

Hi hevansrv7a, your post got me thinking and I ran some RV-6A cases on the model. On an equal nose wheel weight basis, the RV-6A results are similar to the RV-7A. This surprised me, because I also had the impression that the RV-6A has fewer problems.

However, if the model is right in predicting that it is lack of suspension travel that is important, we should expect similar results, as these two aircraft share the same leg.

Therefore, in understanding whether the model is correct or not, it becomes important to know for sure whether the accident rate is different for the different models. Do you (or does anybody) have any definitive data on this?

Raiz
Raiz, I am becoming a big fan of yours, even though it took me a while. Kudos! I hope you stick with this project a while longer as we are all learning a lot about a subject that we thought we had already talked to death.

A minor correction - the 6A legs are shorter. I know the diameter of the root of the mains is smaller. I don't have the data on the nose leg. I am pretty sure it is shorter, though. I once called Van's seeking to use 6A legs on mine and was assured it could not be done for the reason of how they fit into the airplane.

Since they are shorter, one wonders what their suspension travel and strength are and how that bears on the earlier discussion about the feasibility of making the legs stronger on the 7/9 A models.

I don't have real data on the 6A rate of failure or flip over, but when the 7A was a new model, there were literally thousands of 6/6A's and the problem was not being discussed yet. My strong inference is that the 6A is much more immune for some yet to be defined reason. I still can't recall hearing about a 6A failure but I'm sure someone can help with that. I also don't recall any 8A failures and can't come up with any reason that does not involve the much different geometry of the airplane vs. the legs. Maybe I'm just using selective data-recall?

I think that it's a much bigger deal to note that when you analyzed the hole scenario as compared to the bump scenario the problem looked significantly worse. It's equally a big deal that the nut catching the ground was only a problem in the simulation AFTER the failure. That implies that the expensive and inconvenient "fix" that most of us did was perhaps time and money not well spent.

If the SB that had us install different nose wheel forks does eventually prove to have beneficially affected the problem, I will suspect it has more to do with the change in the bending forces at the point where the leg goes vertical (down) from slanted (down and forward). The vertical arm is shorter by a significant percentage to compensate for the higher top of the fork (the plane sits at the original angle to the ground. Subjectively, it feels stiffer to me when I taxi than the original did. My non-engineer visualization of this suggests that the forces on the upper leg my thus come from a different angle, too.
 
Numbers vs. Pcts

The FAA study of this issue covered 23 aircraft.

4 - RV6As
9 - RV7As
4 - RV8As
6 - RV9As

The pictures included in the report do not seem to indicate failure near the lower part of the gear.

Reading the remarks about each accident is highly instructive.

See http://www.ntsb.gov/publictn/2006/RV_Photos.pdf

Even today there are 2.5 times a many 6's as 7's. (see Van's Hobbs meter) It's a reasonable guess that the average 6 has been flying longer than the average 7. Unfortunately, we don't know how many are the A model in any series. Given that, the percentage of these events for the 6A would seem to be considerably lower than for the 7 or 9 A's Thanks for the very helpful data.
 
Solution?

I admit that I did not study every post in this thread, but is there a solution being recommended here? Is it more complicated than switching to a larger nose wheel?

As for technique - no one believes that a tricycle gear airplane can be landed on the nose wheel, but sometimes we goof and at the very least land on all three wheels at the same time. I'd like to know that my airplane can survive the occasional botched landing (within reason of course). That's called a forgiving design. Any solution needs to be retrofitable to existing RV7A's or we risk severely devaluing the current fleet. Adding a spring (a la the RV10) is not an option in my opinion since that would probably require a new engine mount as already noted.

So what's the recommendation here?
Stronger steel?
Different material/metal?
Larger nose wheel?
Shorter main legs?
All of the above?

Thanks,

Tom
 
Figures 19-23 at http://www.ntsb.gov/publictn/2006/RV_Photos.pdf show in a simple way that aircraft with tall landing gear are more prone to nose over than aircraft with shorter gear following a "tripping" event. This is somewhat analogous to an SUV being more likely to roll over than a sports car. A casual reading of the data says that 6As are less likely to tip over, and center of gravity height difference provides a possible explanation. All designs being a compromise, I will not be rushing to shorten my landing gear. On the other hand, I won't be installing larger tires either.
 
I admit that I did not study every post in this thread, but is there a solution being recommended here? Is it more complicated than switching to a larger nose wheel?

As for technique - no one believes that a tricycle gear airplane can be landed on the nose wheel, but sometimes we goof and at the very least land on all three wheels at the same time. I'd like to know that my airplane can survive the occasional botched landing (within reason of course). That's called a forgiving design. Any solution needs to be retrofitable to existing RV7A's or we risk severely devaluing the current fleet. Adding a spring (a la the RV10) is not an option in my opinion since that would probably require a new engine mount as already noted.

So what's the recommendation here?
Stronger steel?
Different material/metal?
Larger nose wheel?
Shorter main legs?
All of the above?

Thanks,

Tom

Tom, this topic is covered frequently in these forums, please read the first post in this thread. I stand behind what I wrote then, and since this discussion almost three years ago, there have been very few tip overs on-airport. I know many replaced their nose wheel axle/bearing setups after this (and other similar threads of the time). Many reported noticeably different taxi behaviors.

Somewhere there is a video of a first landing in a -10, during which the nose wheel skidded the plane to a stop after a perfect landing on pavement. The plane then taxied back with only a huge flat spot on that tire to show for it. This should make believers out of anyone as to the problems with a non-rigid axle.
 
Raiz, Is that a valid assumption? Because of the different main leg geometry (7A sits much higher) I believe the max static load that most 6As fly at is significantly lower than the typical 7A load - I don't have any data at the moment but will try to get some.

Pete

Pete, I have that impression too, i.e. that RV-6As typically operate at lower nose load but I also don't have any good data, so anything you can dig up would be good.

This diagram illustrates your point about the geometries. Each color represents a different plane Green=9, Magenta=8, Black=7, Red=6.
RVs.gif

The short horizontal bars represent the allowable CG range and the fact that the RV6 bar is further back than the RV7 bar (relative to the rear wheel) indicates that, all other things being equal, the load on the nose of the RV6 will be lower than on the 7 (but practice may differ). The lower MAUW of the 6 and the lower CG position will also help. Notice the 8 with the wheel further forward. This also reduces the nose load but the higher CG will tend to count against that under braking. I plan to post a comparison of all 4 aircraft over the same bump but I have a bit more work to do on the model before that is possible.

Raiz
 
Raiz, I am becoming a big fan of yours, even though it took me a while. Kudos! I hope you stick with this project a while longer as we are all learning a lot about a subject that we thought we had already talked to death.

A minor correction - the 6A legs are shorter. I know the diameter of the root of the mains is smaller. I don't have the data on the nose leg. I am pretty sure it is shorter, though. I once called Van's seeking to use 6A legs on mine and was assured it could not be done for the reason of how they fit into the airplane.

Since they are shorter, one wonders what their suspension travel and strength are and how that bears on the earlier discussion about the feasibility of making the legs stronger on the 7/9 A models.

I think you are absolutely right about the rears being shorter on the 6 but the nose leg is the same part number U-603-3 on the 6A, 7A and 9A. The 8A has a different leg U-809-2 but I've no idea what the difference is. Here is the source: http://www.vansaircraft.com/pdf/sb07-11-9.pdf. I've just posted a diagram that shows the side view of all three aircraft.

Raiz
 
This all makes sense to me (sort of anyway, not being an engineer.) It would seem to me that this is another good argument for building light, especially FWF. I would love to have a CS prop, but that would take up more of this precious gear travel dimension. This validates some of my building choices, and no, I have never had aft-CG problems.

The 9A gear is the same, and in fact my whole FWF, including the gear leg, came from a 6A that had a hangar fall on it. The axle (circa 1994) in the nose gear was IMHO FAR superior to what comes with today's A's.

Bob Kelly
 
This all makes sense to me (sort of anyway, not being an engineer.) It would seem to me that this is another good argument for building light, especially FWF. I would love to have a CS prop, but that would take up more of this precious gear travel dimension. This validates some of my building choices, and no, I have never had aft-CG problems.

I've got a C/S prop, the old heavy starter motor, and heavier six pac instruments. But I can still walk back to the tail, push down and raise the nose without too much effort. Or two people standing on the steps at the same time will throw the tail to the ground. It's because I distributed other weighty parts behind the mains and CG. But don't get the wrong idea........... because my specs work out great CG wise for fore & aft........yet there is not a signicant amount of extra weight on the nose, when the aircraft is rolling on the mains.

L.Adamson --- RV6A
 
RVator references to nose gear

I did a little checking into references in the RVator newsletter about nose wheel/leg since I remembered a nose wheel test rig article. Here's what I found:

1. Fifth Issue 2007. P9 setting of nose wheel rotating torque in Matco nosewheels. (I think this torque could be reduced with a proper spacer and pin as already discussed in this forum). P19 Service letter November 9, 2007 shows gross weight and weight on nosewheel for various models. P25 Mandatory Service Bulletin 07-11-09 Upgrade of nose gear leg and fork.
2. Sixth Issue 2004. P8 Nose Gear Design by Van. Discusses the loading components, failure modes, contact of fairing/swivel nut, low tire pressure and binding of tire on fairing. Closes with a request for users to share their experiences with Van's Aircraft.
3. Fifth Issue 1998 (and this is the one I was really looking for). P3 Improving the Breed - Van's Develops a New Nose Gear Leg by Van. This discusses the early issues of RV-6A nose gear legs, both fatigue failures and overload failures. The interesting things about this article are i) Harmon Lange's "rotating runway" test rig for cyclic bump loading and ii) Ken Krueger instrumented the nose gear of the factory airplane and recorded the stress levels on asphalt and turf runways. This data was then used to refine or calibrate the "rotating runway" test rig. This resulted in a new nose gear design U-603-2. It also emphasized that the nose gear leg must act as both a spring to absorb irregularities as well as a strut to accurately locate the nose wheel.

So, while the current discussion is interesting, I am wondering:
1. Is Raiz' model "bump" more severe than will actually be encountered in service?
2. Is there a plan to do rig testing or aircraft instrumentation to verify the model?
3. What changes to the leg can be made to improve it's strut function without detracting from its spring function?
 
Comparison of 6A, 7A, 8A and 9A

Ok, so in an earlier post, I promised to look at a comparison of the 6A, 7A, 8A and 9A. I've assumed the RV8A leg is the same as the others (but I don't know this for sure). I used a 10 ft long, 1.5" deep pothole and 10 knots as the reference case.

For the first comparison, I set the AUW to 1650 lb and the nose load to 325 lb (worst case that they can all cope with and still be within Van's recommendations). This is intended to highlight any differences due to geometry and gave the following "% strength remaining" results:

Model ____ No braking ____ 0.15g braking
6A ________ 50% __________ 29%
7A ________ 49% __________ 25%
8A ________ 49% __________ 26%
9A ________ 49% __________ 26%

Although the 6A is better, the differences are very small compared to the effect of braking.

Next, I considered the worst recommended loading case for each model. For the 6A that is 1650/375 (AUW/nose weight). For the 7A and 8A 1800/375 and for the 9A 1750/325. This gives:

Model ____ No braking ____ 0.15g braking
6A ________ 42% __________ 23%
7A ________ 40% __________ 17%
8A ________ 40% __________ 17%
9A ________ 49% __________ 25%

In this case, the 6A and 9A come out slightly ahead (less likely to fail) compared to the 7A and 8A due to the lower all up weight or nose load but the differences between models are still small compared to the influence of braking.

Finally, I looked at nose laod on a 7A at 1800 AUW:

Nose load ____ No braking ____ 0.15g braking
__ 275 ________ 56% __________ 26%
__ 325 ________ 49% __________ 20%
__ 375 ________ 40% __________ 17%

In summary, the model predicts that braking is the most important factor, followed by nose load, then all up weight and (only) then the differences between models. Only if there were systematic differences in braking or nose load would I expect to see differences between models. For example, if braking on the roll-out were less common in the 9A (because of the lower landing speed) we might expect fewer failures. Alternatively, if the 7A tends to be nose heavy compared to the 6A, we might expect more 7A failures. I think both of these examples have some grounding in reality but I don't have any definitive information. However, overall, the model suggests the likelihood of failure should be similar for all models with maybe the 6A and the 9A somewhat ahead of the 7A and 8A. I know that's not in line with the general perception that the 7A and 9A are generally worst than the 6A and probably the 8A as well.

The next step will be a look at the accident statistics.

Raiz
 
You're doing some fantastic analysis work on this Raiz, really - but until you can do some testing to prove the models, or find test data that has already been taken, you can't really make the leap to a re-design with any sort of confidence. I don't trust unverified models, because I have too often seen them fail due to unexpected variables that hadn't been taken in to account. But what you have done is very valuable - have you thought of contacting the engineering guys at Van's to see if any of it fits with what they have already done, and the testing they have performed.

I am not being critical, just suggesting the logical next "engineering" step. Analyze, test, then design - then test the results to see if the model works in the real world.

Paul
 
Paul, I agree entirely. I have no intention of proposing a change to the fleet without proper validation but I do intend to extract whatever useful information I can from the analysis. I'd be happy to support Vans or anyone else with the facilities and expertise to develop a better gear.

I have tried to engage with Vans. I sent them a summary a month or two before the first post on this forum but only radio silence from then on. If anyone can engineer a connection for me, that would be great.

Raiz
 
Nose leg failure statistics

The following results come from the NTSB database and include all accidents in the period 1/1/2000 to 12/31/2009. Actually, there is one exception (N448GM), which is in the FAA study but apparently not in the NTSB database. I have only counted those nose leg failure accidents where there was no previous problem (such as engine failure or a stall) preceding the nose leg failure incident.

Model ___________ 6A __ 7A __ 8A __ 9A
# nose leg fails ___ 24 ___ 8 ___ 3 ___ 5
total # accidents _ 78 __ 20 ___ 9 ___ 13

To compare the different models in terms of their relative likelihood of experiencing leg failure, we need to know how many of each model were flown in this 10 year period. By analysing the FAA register, I got these figures for the number of aircraft-years of each type with an airworthiness certificate (as an approximation of the # flying):

Model ____________ 6A ___ 7A __ 8A ___ 9A
# aircraft-years ___ 6770 _ 1404 _ 905 _ 1266

100 aircraft-years could mean 10 aircraft for the full 10 year period or 20 aircraft for an average of 5 years etc. It provides a way of dealing with the fact that 6As have been around longer than 7As.

If we assume that all models are flown the same amount per year on average and exposed to the same types of strip and piloting technique, we can compare the relative likelihood of a nose gear failure in failures per 1000 aircraft-years:

Model _________________________ 6A __ 7A __ 8A __ 9A
# leg fails per 1000 aircraft-years __ 3.5 _ 5.7 __ 3.3 __ 3.9

Thus, on the face of it, the 6A, 8A and 9A are very similar but the 7A is apparently more susceptible to leg failure. A word of warning though. A couple of accidents either way would change the picture significantly.

So, does this invalidate my model? At the overall level of lack of suspension travel being the issue, I think the model remains intact. However, it apparently lacks the ability to predict the greater likelihood of failure in the 7A unless the 7A generally has a heavier nose load than the others. If anyone has access to general information on nose loading for the different models, I'd love to know about it.

Raiz
 
Last edited:
Nose gear failures much more likely in the UK

Now here's one to get you thinking. The equivalent anaysis of leg failure rates for the UK Vans fleet shows they are much more likely to suffer nose leg failure than the US fleet:

Model ___________________________ 6A __ 7A __ 8A __ 9A
US # failures per 1000 aircraft-years _ 3.5 __ 5.7 _ 3.3 __ 3.9
UK # failures per 1000 aircraft-years __ 28 __ 44 __ N/A __31

The N/A is because there have been no nose leg failures in the UK on 8As (there are only 2 flying).

The rate of nose leg failures is about 8 times higher (across all models) in the UK than the US. Why? If it were piloting technique, we would expect to see a higher accident rate for the fleet overall but this is not the case. The # accidents per 1000 aircraft-years for all 6, 7, 8 and 9 (TD and nose-gear) combined for all accidents except nose leg failures is 6.6 for the UK and 8.0 for the US.

The only thing that comes to mind is that a much larger proportion of UK RVs are operated from grass strips.

Raiz
 
Now here's one to get you thinking. The equivalent anaysis of leg failure rates for the UK Vans fleet shows they are much more likely to suffer nose leg failure than the US fleet:

Model ___________________________ 6A __ 7A __ 8A __ 9A
US # failures per 1000 aircraft-years _ 3.5 __ 5.7 _ 3.3 __ 3.9
UK # failures per 1000 aircraft-years __ 28 __ 44 __ N/A __31

The N/A is because there have been no nose leg failures in the UK on 8As (there are only 2 flying).

The rate of nose leg failures is about 8 times higher (across all models) in the UK than the US. Why? If it were piloting technique, we would expect to see a higher accident rate for the fleet overall but this is not the case. The # accidents per 1000 aircraft-years for all 6, 7, 8 and 9 (TD and nose-gear) combined for all accidents except nose leg failures is 6.6 for the UK and 8.0 for the US.

The only thing that comes to mind is that a much larger proportion of UK RVs are operated from grass strips.

Raiz

Raiz, I had a -9a the second in the UK I think which I operated off grass in N Yorkshire. I sold it because I saw an accident waiting to happen when on soft grass.

The point I wanted to add, after having not seen it in the previous 80 odd posts is that soft grass compounds the problem that your analysis highlights. As the weight of the plane settles on the grass it moves the mains back a little. The result being full up elevator is unable to hold the nose off very long after touch down. Since the mains are now further aft than normal the loading on the nose wheel is further increased which only exacerbates the situation which you have analysed.

I suspect it is not an insignificant increase though.

With its shorter main gear legs the -6a is less susceptible to this.
 
Soft ground

Steve I think you make a very good point. The soft ground not only moves the rears back but it also creates a braking effect on the mains, which also tends to load up the nose gear even more.

Soft ground also plays a role on the nose wheel loads directly, because it increases the rearward component of the load on the nose leg, encouraging it to fold under.

Ray
 
Maybe a stupid question BUT what are the efects after landing and the flaps are quickly removed and the stick in an aft pos, does that not lift the nose weight better than all the way on the rough runway with flaps down.

Thsis a a great concern to me having made my choice on a RV7A and now this reading.

Here in SA some of the strips are very much called a bush strips.
 
Agent Cooper,

I'm not sure it would make that much difference - once you're on the ground the balance of forces is now about the main wheels, rather than the CofG. Also the ground attitude is at a low angle of attack, so the lift produced will be low, and the speed is dropping. I find that landing on bumpy strips the inertia effects predominate - the aeroplane pitches and I bounce over bumps without the elevator authority to do much about it, especially with an aft cofg. I try to minimize the nose wheel load and lower it to the ground smoothly before I run out of elevator authority.

Pete
 
Maybe a stupid question BUT what are the efects after landing and the flaps are quickly removed and the stick in an aft pos, does that not lift the nose weight better than all the way on the rough runway with flaps down.

Thsis a a great concern to me having made my choice on a RV7A and now this reading.

Here in SA some of the strips are very much called a bush strips.


I mostly operate out of a short grass strip and always immediately raise the flaps at touchdown. Although I have not done accurate testing, my seat of the pants impression is that raising the flaps allows the nose wheel to to kept off the ground to a slower speed, allows more control when lowering the nose to the ground and helps keep the nose in the air while applying brakes.

Opinions vary but I find average Australian grass and gravel strips OK in my 9A. I would keep away from soft, badly potholed and strips with short/sharp undulations.
Read all the suggestions on correct pilot technique and modifications and incorporate those you think are sensible. One extra modification I did that is rarely discussed is to alter the nose fairing cone to increase the fairing clearance fwd of the tire.

Fin
9A
 
I just saw this for the first time today. I have a question about the suspension value for the nose tire.

Does that imply that the tire compresses until the bottom rim almost touches the ground? If so, where does the tire go? In other words, does the tire bulge outward possibly contacting the wheel pant? If so, what does that do to the scenario?

I also saw three psi values: 25, 30, 35. I assume that is the nose tire pressure. Interesting that you seem to get less suspension as the psi goes up. I recently upped my nose tire pressure to 40 psi thinking that was the way to go. True or not on a 6A?
 
Last edited:
Interesting......

Ron, that is an interesting thought!

I let the air out of my front tire just to check that it would not catch on the fairing, if I would ever encounter a "flat" during flight or landing. But..... what you are saying is something else!..... if the tire is flattened by the load of the gear, it will not just be "flat" but it will bulge out!... a lot! And then.... it will catch on the fairing for sure! (unless you have more than the additional clearance that I already have, I guess)

I will be following this discussion!

Regards, Tonny.
 
Another RV Nosewheel

I am curious why the RV-12 has not been considered in this discussion. Is it due to the fact that the nose wheel tire is a 500 X 5" just like the mains, or the fact that it has less weight to support. As far as I know, there have been no RV-12 nose gear failures or roll overs. True though that there are less than 50 flying. Also, many of the ones that are flying don't have wheel pants yet. There is a youtube video showing a landing on a pretty rough island sod strip. Can't see the nose wheel, but you can tell from the cabin oscillations that it ain't the smoothest!!
Hope the below link works.
Tom

http://www.youtube.com/watch?v=nOKvdcZknGE&playnext=1&videos=1CnbaeuaY00&feature=mfu_in_order
 
Back
Top