What's new
Van's Air Force

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

Aerospace Welder / Engineer recommended I not weld the W818 aileron rod.

jj_jetmech

Well Known Member
I read all the threads on this topic and decided to have my W818's welded. I'm now on my second set of rods after the first welder did a horrible job. I could'nt stand the sight of them so I ordered 4 inserts and more tubing.

This time a friend took them to a welder here in so-cal who's company makes parts primarilly for the Defense Depart and NASA. This was to be a favor as they do not weld for walkins etc..

He called me with his friendly recommendation which was in short, don't weld these unless you plan to heat treat them after. Stating that he had completed a Rockwell hardness test on the inserts and a quick calculation indicated that welding would cause a loss of 70% of the threaded inserts strength...

This all makes sense I just didn't give it much thought prior to this discussion.. It's on the plans as an option, many are in service and I haven't heard of any failures.

He indicated he was familiar with RV's and if it was his he would rivet them. He then said he would gladly weld them if that's what I wanted...

I'm now full circle and getting out the rivet gun following his recommendation. I can't help but think maybe he was being over cautious if there is such a thing in this case.

Any thoughts, or comments? Is this worth further discussion?
 
Last edited:
Any thoughts, or comments? Is this worth further discussion?

You'd have to calculate the strength needed in the part and the strength of the welded part, then compare the two for a worthwhile engineering discussion.

From a practical perspective, there are thousands of welded rods out there with no failures that I'm aware of, so field experience seems to indicate that there is no issue with welding the threaded inserts to the rods.
 
From a practical perspective, there are thousands of welded rods out there with no failures that I'm aware of, so field experience seems to indicate that there is no issue with welding the threaded inserts to the rods.

Not sure about the assumption above, but regardless, I agree with the welder.
 
I had mine welded by a friend who builds NASCAR cars for Hendricks Motor Sports. I recently gave those rods a really close inspection for no reason othere than they have been in service for seven years and 650 hours.

Mine are painted white so any crack should be easy to see. So far, they look as good as the day I had them welded. That doesn't mean they won't failin the future, just that I'm happy with them.

The reason I had them welded was that I couldn't do a nice job riveting them.
 
Now that its been brought up I'm picturing the threaded end snapping off. Probably take side force though as opposed to just popping off under pulling load.
 
Old subject. The search function will pull up good previous discussions.

Sticking strictly to the assertion in the first post, a loss of 70% of the threaded insert's strength is unlikely. AN490's are made from 4130, heat-treated to Rockwell 25C to 30C, which would put tensile strength in the 120,000 to 135,000 psi range. For comparison, normalized 4130 is about 97,000 psi, and annealed is down around 81,000 psi. What the nice man probably meant to say was that the welding process would reduce the AN490 to about 70% of it's original strength (90,000/135,000). This is based on the very reasonable assumption that the welded part would end up close to the normalized condition.

MIL HDBK-5J says the single shear strength of an 1/8" AD rivet is 389 lbs. Ignoring corrections for bearing strength, and treating the insert joint as four single shears (two rivets), you have 1556 lbs capacity.

In comparison, assume 50,000 psi yield for an annealed condition after welding. The root diameter of the 1/4-28 thread is about 0.2 inch. So, 0.1^2 x 3.14 x 50,000 = 1570 lbs load capacity.

Snapping off? Normalized or annealed conditions are more ductile, less brittle. The welded fitting is more likely to bend or elongate rather than snap, although that plastic behavior will come with the reduced ultimate strength.
 
Last edited:
Stranger than fiction!

Kyle wrote:
From a practical perspective, there are thousands of welded rods out there with no failures that I'm aware of, so field experience seems to indicate that there is no issue with welding the threaded inserts to the rods.

I am told that this method (fleet experience shows the part does NOT break) is a path for a certified aircraft to prove airworthiness of some assemblies. An example that we are all flying: the HS elevator hinge points just outboard of the center pivot bearing do not have the strength to pass a Part 23 inspection. But, none have failed, so they would pass due to fleet flying experiences.

Eyeball engineering says these brackets should have 4 rivets ea, same as the outer brackets. Again, I know of no failures, so the ol' eyeball is wrong in this case.

I would suggest the same process applies to the aileron pushrods (welded): none have broken, so they are 'good to go'.

Carry on!
Mark
 
Is this a full circumference weld or just a series of rosette welds in the same locations where the rivet holes are drilled???
 
welding

If the weld is not done properly it creates a crack initiation point at the edge because of the sharp corner. The control forces are light on an RV, I don't see the benefit of welding aileron rods compared to the risk of welding.
 
Nice calculations Dan! I dont think we are going to pull these apart with the stick force these aircraft see.
 
Welds

Slightly larger diameter (5/8 & 3/4" rube) have been in use on the various Pitts Aircraft since the late 60's. Certified Pitts models include S1S, S1T, S2A, S2B and S2C.
The end of the tubing should be "fish mouthed" at a 30 degree angle. All Pitts factory aircraft are TIG welded and they DO NOT normalize the welds.
I have never heard of a rod end failure on a Pitts.
Cutting the tube end straight and welding around the end is asking for trouble in my opinion.
The rod ends are almost always cadmium plated, the cad should be wire brushed off, then polish the weld area with some fine emery. The small rods on the RV make it really difficult to make the weld look nice, the larger rods used on the Pitts are easier.
The cad plating is also very toxic when burned which makes it very important to remove it before welding.
 
I've met DanH one time and from that discussion, from seeing his plane, and from reading his many posts of technical nature here in these forums, I am convinced he is a Mechanical Engineer posing as a Truck Dealer.

Thanks for your many contributions to these forums, Dan. :)
 
Last edited:
Further Detail

Dan, thanks for putting some numbers to this and the MIL-HDBK-5J ref.

I should note that my conversation with this engineer was in great detail, he was clear in his reasoning for this recommendation as follows.

Once annealed by the welding process the mechanical properties and structural integrity of this part will have changed forever unless heat treated.

Although the effected parts would initially be comparable in strength to rivets after being annealed or normalized the concern is for the altered part over time.

Basically leave the metals as intended and put the 2 rivets in, this far exceeds the design requirements without compromising the heat treated parts.

The concern is not pulling apart it is the idea that the threaded portion of the AN490 could fail.

What I have learned, initially I thought welding was the way to go, turns out its not. Plus by not welding the inside of the tubing it can be fully protected by epoxy primer and not burned off by welding.

Ive renamed these the Analeron Control Rods!
 
Slightly larger diameter (5/8 & 3/4" rube) have been in use on the various Pitts Aircraft since the late 60's. Certified Pitts models include S1S, S1T, S2A, S2B and S2C.
The end of the tubing should be "fish mouthed" at a 30 degree angle. All Pitts factory aircraft are TIG welded and they DO NOT normalize the welds.
I have never heard of a rod end failure on a Pitts.
Cutting the tube end straight and welding around the end is asking for trouble in my opinion.
The rod ends are almost always cadmium plated, the cad should be wire brushed off, then polish the weld area with some fine emery. The small rods on the RV make it really difficult to make the weld look nice, the larger rods used on the Pitts are easier.
The cad plating is also very toxic when burned which makes it very important to remove it before welding.

You bring up some good points..

It just dawned on me that my unwelded Analeron rods and AN490's are coming back to me today. I've removed the Cad Plating from the 490's and now I'm riveting them... I'm not buying 4 more that would be 12, I'm cleaning Van's out on these things... I could brush cad them.. I guess..
 
Last edited:
From Sportplane Construction Techniques by Tony Bingelis:

control-rod-assembly.jpeg
 
Are we arguing theory or everyday reality here?

I can acknowledge (because I know nothing about welding) that welding this joint may technically weaken one or both parts being welded. However, is it weakening it enough to make any difference in the real world? Since Van's drawings show a straight cut tube with a circumference weld, shouldn't we assume that Van's engineers have decided this is strong enough for the purposes intended of it?
 
Welding

Dan, thanks for putting some numbers to this and the MIL-HDBK-5J ref.

I should note that my conversation with this engineer was in great detail, he was clear in his reasoning for this recommendation as follows.

Once annealed by the welding process the mechanical properties and structural integrity of this part will have changed forever unless heat treated.

Although the effected parts would initially be comparable in strength to rivets after being annealed or normalized the concern is for the altered part over time.

Basically leave the metals as intended and put the 2 rivets in, this far exceeds the design requirements without compromising the heat treated parts.

The concern is not pulling apart it is the idea that the threaded portion of the AN490 could fail.

What I have learned, initially I thought welding was the way to go, turns out its not. Plus by not welding the inside of the tubing it can be fully protected by epoxy primer and not burned off by welding.

Ive renamed these the Analeron Control Rods!

Welding does not anneal a steel part. If done with oxy/acy torch and cooled gradually, the welded part is something close to the original normalized condition of the steel tube. The same process can be used with TIG, after completing the weld, heart it up cherry red with a troch and cool very gradually by slowly backing the torch away.
 
Follow the drawings

Are we arguing theory or everyday reality here?

I can acknowledge (because I know nothing about welding) that welding this joint may technically weaken one or both parts being welded. However, is it weakening it enough to make any difference in the real world? Since Van's drawings show a straight cut tube with a circumference weld, shouldn't we assume that Van's engineers have decided this is strong enough for the purposes intended of it?

Exactly what I was going to say.

The default for builders is to follow the drawings because that is how the designer wanted it built.

The drawings clearly show a welded joint option...:)
 
Weld

Regarding post number 15, third example form the top, this is an extremely rare case where I disagree with Tony.
The Pitts factory drawings for the S1S show in detail how the fishmouth is cut on the end of the tubes. The simple straight weld around the end of the tube as shown in Tony's drawing can create stresses during welding that lead to failure. The fishmouth method virtually eliminates that possibility.
Regarding protecting the inside of the tube, if the tube is sealed tight it will probably be good for 100 years. If you want internal protection, after welding drill a #40 hole near one end, inject some tube oil, rotate the tube around, drain the excess oil and put a drive screw in the hole. Drive screws are in the Spruce catalog.
 
Once annealed by the welding process...

The welding process does not anneal.

Basically leave the metals as intended and put the 2 rivets in, this far exceeds the design requirements without compromising the heat treated parts.

One of two perfectly reasonable choices.

The concern is not pulling apart it is the idea that the threaded portion of the AN490 could fail.

Which is why I chose the thread root minor diameter for the example. Not that it makes any difference. The pushrod itself will fail in compression at about 535 lbs.
 
Last edited:
I welded and treated mine internally for rust prevention. I am going to take a guess here and say that the aileron tubes do not get anywhere near as much abuse as the little flap pushrods do and nobody seems to worry about them.
 
Gee. Too often I see deviations from Vans plans and instructions here that I just wonder about. Rarely, especially in the firewall-aft part of the build, have I seen a deviation or alteration that is a genuine improvement.

Stick to the plans, you'll be happier. Vans and Company knew what they were doing, almost always.
 
Dan, I didn't say it WOULD snap off. I just said I could picture it snapping off. Especially with me welding. Now, about that compression failure. Do I need to sleeve it with some water pipe?
 
Gee. Too often I see deviations from Vans plans and instructions here that I just wonder about. Rarely, especially in the firewall-aft part of the build, have I seen a deviation or alteration that is a genuine improvement.

Stick to the plans, you'll be happier. Vans and Company knew what they were doing, almost always.

I seem to recall the older plans, or at least instuctions,actually called out welding as an acceptable alternative.
I would not disagree with your statement in general however. Your right on.
 
Gee. Too often I see deviations from Vans plans and instructions here that I just wonder about. Rarely, especially in the firewall-aft part of the build, have I seen a deviation or alteration that is a genuine improvement.

Stick to the plans, you'll be happier. Vans and Company knew what they were doing, almost always.

Neither option discussed is a deviation or alteration from the plans. Welding these rods certainly creates more debate than riveting. Clearly both options exceed the design requirements making either suitable.

So the plans call out a -12 rivet, that looks a little long 1.8722145 D :D I know that's another thread..
 
I seem to recall the older plans, or at least instuctions, actually called out welding as an acceptable alternative.
I would not disagree with your statement in general however. Your right on.

Not even the older plans, the much newer RV-7 plans specifically added a note "Optional Weld in Place of Rivets"

It's not an old RV4/RV6 thing...:)
 
Not even the older plans, the much newer RV-7 plans specifically added a note "Optional Weld in Place of Rivets"

It's not an old RV4/RV6 thing...:)

I think you mean not "only" the older plans......

I am enjoying the banter in the thread BTW.... on my 3rd bag of popcorn already :rolleyes:
Part of the reason for that is because for some people, these 4 rivets have been some of the toughest to install in building the entire kit.

I have seen some that looked pretty bad. In those cases, having had them professionally welded would have probably been a much better choice.
 
Last edited:
So the plans call out a -12 rivet, that looks a little long 1.8722145 D :D I know that's another thread..

Seems a little long but that is what is needed.
The reason being is a lot of rivet material gets used up when the shank of the rivet swells within the hollow area of the tube.
There in lays the challenge I mentioned in my previous post... the rivets start out being a bit on the long side.... along with forming the shop head on a curved surface.... It is a prime situation for having the rivet clinch over to the side.
 
Last edited:
Seems a little long but that is what is needed.
The reason being is a lot of rivet material gets used up when the shank of the rivet swells within the hollow area of the tube.
There in lays the challenge I mentioned in my previous post... the rivets start out being a bit on the long side.... along with forming the shop head on a curved surface.... It is a prime situation for having the rivet clinch over to the side.

I was kidding,,based on other posts! But thank you... I riveted the rods this afternoon with -12's.. Clamped them in my soft jaw vise, got out the 2X and and a bar drilled to accept a squeezer set, placed a machined head die in the bar to produce a rounded shop head... Should have done this weeks ago... Now what to do with the welded set of W-818's? Maybe I'll put them on the classified page FS::rolleyes: actually I plan to destructively test them...
 
Shrink fit, cross drill, dowel.... button weld 4 places. move on.
 
Last edited:
weld

Done a little checking on this site and found that max push or plull on this part was 256 pounds so when I made the part with gas welding I built a fulcrum out of wood with a 10 ft arm and placed 60 # of lifting weights on the arm to test to twice the pull.
Bob
 
I suspect Van's picked the rivet technique over welding as I believe no where in any vans built plane is the builder required to do any welding. Making it a welded joint adds to the tool list.

with that said, rivets can get loose. A loose rivet/bolt shear strength is greatly reduced.

I work in defense manufacturing industry, do a lot of welding myself, work with some very smart welding engineers. I personally would trust a properly welded joint over a riveted joint any day of the week.

I would also trust a nicely rivetted joint over a farmer with a mig welder he just bought at lowes and has no clue what he is doing, welds look like a bird flew over after a healthy meal.
 
I would also trust a nicely rivetted joint over a farmer with a mig welder he just bought at lowes and has no clue what he is doing, welds look like a bird flew over after a healthy meal.

This is not a good application for a MIG weld........ Paul made the same slip in his visit with Vetterman Jan 2015 KP.......:eek:
 
I was kidding,,based on other posts! But thank you... I riveted the rods this afternoon with -12's.. Clamped them in my soft jaw vise, got out the 2X and and a bar drilled to accept a squeezer set, placed a machined head die in the bar to produce a rounded shop head... Should have done this weeks ago... Now what to do with the welded set of W-818's? Maybe I'll put them on the classified page FS::rolleyes: actually I plan to destructively test them...


Mail them to me and I'll test them on a certified calibrated machine. One static, one fatigue...
 
Testing

Mail them to me and I'll test them on a certified calibrated machine. One static, one fatigue...

PM me your address I will definitely send them to you.. We should build a riveted W-818 and test it as well...

Both will exceed our design needs as proven in the field but it would be cool to have a defined winner....

This could be like the Super Bowl of Analeron Control Rods... :D
 
We should build a riveted W-818 and test it as well...
Both will exceed our design needs as proven in the field but it would be cool to have a defined winner....

Cool! Go for it.

TIG or OA welds? If TIG, I'll bet a beer on welded.
 
I had mine done by a welder friend. He does perfect work and I really doubt there will ever be any problem with the strength of these rods. Much nicer than trying to get a good rivet head with those long rivets.
IMG_8362-M.jpg


My only concerns with this would be corrosion protection, since the CAD plating got baked off during the welding. The tubes have been repainted, but the threaded rod ends have the potential to get corroded with rust.
 
That is the wrong way to weld it. Sets up stresses that could lead to cracks/failure. Should have two 30 degree V cuts on each tube end.

If this was a suspension member in an off-road race truck, then your opinion might be valid.
In the RV aileron pushrod application, welded as shown is more than adequate.

The majority of the prototypes at Van's Aircraft have been flying for many years and many thousands of hours with aileron push-rods welded as depicted in the plans (and shown in these photos).
 
The majority of the prototypes at Van's Aircraft have been flying for many years and many thousands of hours with aileron push-rods welded as depicted in the plans (and shown in these photos).

And if you dont know what empirical data is, well, what Scott posted here is as good a definition as anything Webster has to say.
 
Welding

Tell that to Van's. A circumference weld as shown in the pic above is the way it's shown in the drawings.

The majority of metals used in EAB construction have a grain. There are numerous references to grain and welding protocol. These include The Techniques of Aircraft Building, published by Acro Sport Inc, as well as numerous EAA Publications. The consensus of all this is that welding across the grain should be avoided to the extent possible.
AC43.13 does not specifically address welded rod ends. However the principals involved are the same as for welded steel tube repairs. Welded steel tube repairs are addressed in AC43.13 Pages 4-62 thru 4-78. In each example on these pages, the end of the tube is either a fish mouth(two 30 degree V cuts) or a scarf (30 degree straight cut). The purpose of this is to avoid, to the degree possible, welding against the grain.
Just one more reminder that several Aviat Pitts models are type certificated aircraft. None of the Vans Aircraft are type certificated. The testing and engineering that the Pitts aircraft were subjected to far exceed any similar process for any Vans model. I have to sincerely question any one who would choose and promote their own "empirically proven" method, which goes against decades of established protocol, on thousands of aircraft, many of them type certificated.
Regarding "Aerospace Welder" this is a rather meaningless generic term. There are many different categories of "aerospace welder" most of them having little or nothing to do with welding on EAB aircraft. If you refer me to someone who has every Aerospace Welding certification known to exist, now we are having a meaningful conversation. Aerospace Welder, as a stand alone statement, means nothing.
 
Welded mine

I've been involved in three RV builds to this point. Welded these every time. I believe it is necessary to have a "vent" hole in the tube while it's being TIG welded. I took the opportunity to enlarge that vent hole to #30, then add a few tablespoons of Poly Fiber TUBE SEAL LINE OIL (ACS sells it). Plug the hole with a solid blind rivet, then I painted all non threaded surfaces with epoxy.

Question, what is the recommendation for the CAD coating, mentioned earlier, on the threads evaporating off after welding?

Thanks!
 
Welding

I've been involved in three RV builds to this point. Welded these every time. I believe it is necessary to have a "vent" hole in the tube while it's being TIG welded. I took the opportunity to enlarge that vent hole to #30, then add a few tablespoons of Poly Fiber TUBE SEAL LINE OIL (ACS sells it). Plug the hole with a solid blind rivet, then I painted all non threaded surfaces with epoxy.

Question, what is the recommendation for the CAD coating, mentioned earlier, on the threads evaporating off after welding?

Thanks!
I wire brush the rod end to remove the cad and then polish with fine emery or wet or dry paper. Wash with lacquer thinner or acetone.
In most case the vent hole is a good idea, but I have not found them necessary on a small tube. Leave a tiny un welded area on the final weld, let it cool, then weld up the small gap. If you use a vent hole, the drive screws do a better job of sealing than a blind rivet. Put a dab of tank sealant or JB weld on the screw before driving it. The drive screws were standard protocol for the old tubing structures for the military.
Certified aircraft do not generally protect the inside of tubing structures unless the aircraft is configured as seaplane or amphibian.
I VERY LIGHTLY sandblast the thread after welding
I try to estimate the amount of thread that will be used and put a light coat of primer on that portion, then mask it and prime and paint the rest. If you need additional "clean thread" run a castle nut backwards over the thread.
 
Illustrations to go with Jim's explanation. Empirically not required here, but it is considered good practice among tube and fabric pros. Required? I dunno, but it can't hurt.



 
Back
Top