News:


  • March 29, 2024, 04:52:45 AM

Login with username, password and session length

Author Topic: Titanium vs. Steel & Aluminum  (Read 895 times)

Offline Tim Wescott

  • 2016 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 12804
Titanium vs. Steel & Aluminum
« on: May 26, 2023, 01:53:10 PM »
With regards to Matt Colan's recent tragedy, I just make my own pushrod ends with an aluminum cap into which I've screwed a 4-40 socket head screw.  The screw is epoxied in and tightened, then the cap is epoxied onto the end of the rod.

See picture of Matt's failed part, and the sketch of mine.

I'm assuming that because these use steel screws that came from a reliable source, and because the construction supports the shank of the screw past the head that this isn't as susceptible to fatigue as the part that Matt used.

So -- should I worry?

AMA 64232

The problem with electric is that once you get the smoke generator and sound system installed, the plane is too heavy.

Offline Dave Hull

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1902
Re: Titanium vs. Steel & Aluminum
« Reply #1 on: May 26, 2023, 06:14:40 PM »
Putting a threaded shank in bending is never recommended. However, threaded shank fittings are often used in full scale aircraft controls and cars under controlled circumstances. To avoid issues, there are some guidelines and things to watch out for:


--Make sure the end joints are free. That is, the ball joint has more angular travel available at the end position. Otherwise, you may be jamming the controls into the limit and putting bending stress into the threaded rod. That might result in low-cycle fatigue or simply breaking unexpectedly.

--Keep the mass of the pushrod assembly low. I have seen all kinds of different pushrods that have Goliath carbon tubes that are totally unnecessary. The end fitting mass counts, too.

--Keep the pushrod straight from pivot point to pivot point. In column loading (bending due to compressive stresses) it is the eccentricity or the amount of "offset from straight" that sets up the conditions for bending. This is the whole reason for the thin wall tubes: to get the area moment value up to better withstand bending. You can certainly do it with the appropriate diameter solid steel rod. The reason for the carbon is that per unit of area moment, it is less weight that solid (or tubular) steel. Thin wall aluminum tubing works fine as well.

--Related to the above, having the pushrod symmetric and balanced about its longitudinal axis essentially gets rid of the twisting under vibration mode. If you bend a 3/32" wire to poke down thru the bellcrank from one side, you are no longer mass balanced, and any lateral vibration (shaking force) will cause displacement and wear. (I replaced a rat race bellcrank with a very loose hole that resulted from this issue in a surprisingly short period of time.)

--If you use commercial screws, they are all rolled threads with a modified root radius. That means two things: that the material is essentially forged creating aligned crystalline structures and less potential for cracking while at the same time it is work-hardened at the surface. The modified root radius reduces the stress concentration at the "inside corner" of the thread. All good things.

--Corrosion is not your friend. If the threads are allowed to be attacked by corrosion, it eats away at the grain boundaries and much more easily initiates a crack.

--Heat treat matters. Some materials are much more susceptible to cracks at the wrong heat treat condition.

--In the other related discussion thread, there is a bogeyman term thrown around a bit: harmonics.  Better to say that you need to avoid a resonance in any of the structure and controls. Sitting on a resonant frequency means that the energy being pumped in excites the frequency that the structure wants to take off at. If the damping is limited, and the pumping continues, something will either chatter and wear out quickly, or it will fatigue and break. If we start off with the assumption that the engine rpm establishes the first energy mode, you do not want any structural resonances near that frequency. The engine will have higher frequency shaking forces as well. Just as the suspended pushrod assembly will have other resonance modes. Those structural modes are predictable, using sufficient analysis, which is what is done in aerospace. There is some art to doing this accurately, especially at higher order harmonics. If you can avoid the lowest frequency harmonics, you are generally fine since the higher frequencies will have less energy available. For fan blades the designers may have to agonize over harmonics up into the high-teens.

--The damping in a wood structure is generally quite good. The damping in a carbon structure is not. Foam in a carbon structure adds damping back in. A low friction ball joint won't have much damping.

--If your pushrod breaks because the structure was excited, you should be looking at the stiffness of the horn and bellcrank, not just the stiffness of the pushrod. Think of the arm on the horn like a diving board that has one moment-fixed end. Twang the free end and guesstimate the frequency the vibration. Don't let your engine run there, or be sure you have a lot of damping.

--In vibration control, there are situations that snubbers can play a role. If a component is subjected to 95 percentile type of inputs, it may be acceptable/advisable to use external means to control response.

--Don't ignore rattles in a plane. Figure out where they are coming from and decide whether they are innocuous or likely to be fatal to the airframe. I have a hand-me-down Fancy Pants that rattles. I haven't cut into the fuselage, but believe there's a good chance that it is the control system that is rattling....

--Specifically looking at the threaded aluminum cap design, where the screw exits the cap is still a stress concentration. If the cap was a shallow cone and threaded right to the tip, then you could take some credit for knocking down the concentration factor. There are tables for all kinds of concentration factors. Most are probably based on empirical testing of rotating beam specimens which means if you duplicate what was tested you are likely to get good correlation to their results. Using a blob of epoxy over the threads where they come out of the cap probably won't buy you much. The modulus of the steel screw and the aluminum hub are way higher. So the epoxy will yield in short order and you're not really better off than you were except that you added more mass. Initially, those deformations would add damping. But when that's gone it's like it was never there. A metal-filled adhesive (JB Weld) might have somewhat higher modulus. I don't have a number handy. But not something to count on if you're that close to failure. As you say, if the screw is torqued into the cap, I wouldn't worry about a bending failure under the actual screw head. All aerospace screws have a requirement for a radius under the head to help avoid failures there. But to be aware that you are going to need an absolute bare minimum of 3 full threads in the aluminum cap if you are to use normal torque values and expect it to remain solid over time. And, hopefully you are using some aerospace grade aluminum?

I had a design article I started on pushrods a year or two ago for our club newsletter. Maybe I'll go find it and see what it was missing. It was mostly simple graphics as I recall...?

Dave
« Last Edit: May 26, 2023, 06:43:49 PM by Dave Hull »


Advertise Here
Tags:
 


Advertise Here