Weight savings with Carbon Pushrods
Posted: Thu May 01, 2025 11:57 pm
The genesis of this project started with a random comment made on another list about using carbon for lighter pushrods and other parts. While there were many nay-sayers, carbon is actually ideal for pushrods because the most important parameter is stiffness – and carbon is really stiff.
While aluminum is by far the most common material used in constructing push rods for flight controls, switching to carbon can offer some significant weight savings. After some back of the envelope calculations looked promising, I prepared a spreadsheet to evaluate potential weight savings for different Carbon tube sizes.
There are two pushrods with a bell crank and a bob-weight in the center. The kit was supplied with 1.25” diameter, .062 wall aluminum tubing for these two tubes.
The limiting failure mode is buckling. The Euler critical buckling load is a simple equation. It carries with it some assumptions that need to be considered. For each assumption that is not met, a significant reduction in the critical buckling load can be expected. Assumptions include a purely axial load and a perfectly straight tube.
When a rod end bearing is attached to a bell crank, the load applied to the push rod contains a moment due to the friction in the rod end bearing. This moment becomes significant with large compressive forces.
The buckling load is also surprisingly sensitive to straightness. Long tubes may or may not be straight. Don’t be surprised if a 6’ tube is off by 1/16” in the center. This can cut its load carrying capability by a lot as the compressive load is also adding a bending moment within the tube.
Aligning the ‘bow’ in the same axis as the rotation, the road end bearing mentioned above can either have additive or opposing effects. That is, if the rotation of the rod end bearing opposes the bow in the rod, full buckling load may be achieved. On the other hand, if one flips the rod so that the bow and the rotation are additive one can quickly drop the buckling load by 50%.
Testing
I built a text fixture that allowed me to apply well in excess of 1,000 lbs of compressive load to full length push-rods. Generally, the FAA requires a pull of 150lb on the stick. This represents a panicked pilot hurtling towards the earth in a death spiral. If the pilot is able to break some part of the control system recovery is no longer possible.
The original kit supplied rods do meet the FAA 150lb pull criteria. This doesn't necessarily mean the rest of the system would also survive.
I reduced the criteria a bit. My tubes can handle a 63 lb pull on the stick. To put this into perspective, this represents a 6 G pull with a safety factor of 9x. It is worth noting that the rods don’t break at 63 lb. Instead, they start to bow. This means one cannot get any higher force out of them. Note that the forces in the tubes of a 320/360 are 5 times higher than the stick force applied, so a 63 lb stick force subjects the tubes to 300 lbs.
The end fittings are machined aluminum and are bonded with structural adhesive. While the bonded joint is not loaded at all from compressive forces, I did do a destructive pull test. Estimates placed the bond failure at 15,000 lbs, the threaded rod at about 5,000 lbs and the rod end bearing at 3,000 lbs. The bearing pulled apart at just over 3,000 lbs.
The final weight savings was almost exactly 50%, 848g vs 1,683g for both tubes together. The tubes themselves only weighed one third that of the aluminum but since I stuck with steel hardware, the total savings were diminished.
While aluminum is by far the most common material used in constructing push rods for flight controls, switching to carbon can offer some significant weight savings. After some back of the envelope calculations looked promising, I prepared a spreadsheet to evaluate potential weight savings for different Carbon tube sizes.
There are two pushrods with a bell crank and a bob-weight in the center. The kit was supplied with 1.25” diameter, .062 wall aluminum tubing for these two tubes.
The limiting failure mode is buckling. The Euler critical buckling load is a simple equation. It carries with it some assumptions that need to be considered. For each assumption that is not met, a significant reduction in the critical buckling load can be expected. Assumptions include a purely axial load and a perfectly straight tube.
When a rod end bearing is attached to a bell crank, the load applied to the push rod contains a moment due to the friction in the rod end bearing. This moment becomes significant with large compressive forces.
The buckling load is also surprisingly sensitive to straightness. Long tubes may or may not be straight. Don’t be surprised if a 6’ tube is off by 1/16” in the center. This can cut its load carrying capability by a lot as the compressive load is also adding a bending moment within the tube.
Aligning the ‘bow’ in the same axis as the rotation, the road end bearing mentioned above can either have additive or opposing effects. That is, if the rotation of the rod end bearing opposes the bow in the rod, full buckling load may be achieved. On the other hand, if one flips the rod so that the bow and the rotation are additive one can quickly drop the buckling load by 50%.
Testing
I built a text fixture that allowed me to apply well in excess of 1,000 lbs of compressive load to full length push-rods. Generally, the FAA requires a pull of 150lb on the stick. This represents a panicked pilot hurtling towards the earth in a death spiral. If the pilot is able to break some part of the control system recovery is no longer possible.
The original kit supplied rods do meet the FAA 150lb pull criteria. This doesn't necessarily mean the rest of the system would also survive.
I reduced the criteria a bit. My tubes can handle a 63 lb pull on the stick. To put this into perspective, this represents a 6 G pull with a safety factor of 9x. It is worth noting that the rods don’t break at 63 lb. Instead, they start to bow. This means one cannot get any higher force out of them. Note that the forces in the tubes of a 320/360 are 5 times higher than the stick force applied, so a 63 lb stick force subjects the tubes to 300 lbs.
The end fittings are machined aluminum and are bonded with structural adhesive. While the bonded joint is not loaded at all from compressive forces, I did do a destructive pull test. Estimates placed the bond failure at 15,000 lbs, the threaded rod at about 5,000 lbs and the rod end bearing at 3,000 lbs. The bearing pulled apart at just over 3,000 lbs.
The final weight savings was almost exactly 50%, 848g vs 1,683g for both tubes together. The tubes themselves only weighed one third that of the aluminum but since I stuck with steel hardware, the total savings were diminished.