Torsion bar for Jetpack

Background: Jetpacks need a torsion bar for stability linkage between the arms. Each movement of the pilots arms results in a steering change. The torsion bar returns the steering arm linkage to level and causes an increase in the force feedback for aggressive turns or corrections. It makes the ride stable.
Currently we build our torsion bar now out of a solid modified Nylon 28mm diameter, 680 mm long, CNC machined ends. Weight about 420 grams, just under a pound of Nylon. As in the photo below the black round bar:

This rendering shows the bar under deflection, showing twist and stretch with yellow lines. The main reason for our bearings in the shoulder joints is to handle the torsion bar forces and not lock up.

The bar is twisted torsionally up to 50 degrees and is stretched in length during normal operation. A lot of force is on this part. The bar has been very strong and no breakage. The issue is the optics, machine finish extruded round rod is just not that sexy. I am considering injection molding the part in order to get a better finish, and be able to change the profile.

My question is regarding the thickness of the part and injection molding suggestions. Thoughts on injection point to minimize warpage, in the center or at one end? Other issues to consider? We do not plan on molding in the mounting holes due to registration and stresses, those are planned to be drilled post molding.

Thanks for any suggestions.

What kind of finish are you looking for? Matte, gloss, semi-gloss?

I would think you would want to match the other black parts on your jetpack.

Are there any secondary decorative processes that would suit your needs? It seems like the extrusion is going to be better for part strength and costs.

Painting, media blasting, or possibly even just some kind of alternate material as a sleeve for the center section?

Strength superior in extruded due to material orientation? Additional cost in this part could be justified. An alternate material sleeve might be an answer to get a better surface finish.

Gloss or even a color that would add some visual highlight would be nice as a contrast. Black is the most technically convincing though for certain.

Disregarding CMF for a moment, the engineering implementation of the torsion bar seems less than great: The way it’s captured at each end with bolts driving right through the cross section of the bar, reducing the torsional stiffness at that point?

What you have here is the same as an anti-roll bar in vehicle suspension. Instead of capturing a straight bar at each end, the bar is implemented in a ‘U’ shape so that the two legs are anchored more easily without sacrificing torsional stiffness. Looking at the white parts of the pack, it looks like it might have been an easier implementation.

As for CMF, rather than applied finishes, or coatings, why not ‘truth to materials’?

The torsion bar rotates with the angle of the arms, and only comes into play when there is an angle differential. The space envelope is limited at each end of the radial travel. Move the arms down, they contact the frame, move them up, they contact the channel system. The part really has to be cylindrical to allow the required movement. U shaped torsion bars are pivotally anchored at two other points on the frame. Many different reasons for the anchoring method, machining of the alloy ends, alignment, assembly, user dis-assembly in case of bearing maintenance.

Splining the ends (as below) in nylon would not be strong enough to counteract the forces. The through bolt connection has not shown any signs of failure in testing or the field and aligns and assembles great.

Regarding cosmetics and truth to materials, in my subconscious, a torsion bar should be red. :slight_smile:

The composite material might be better suited to a large hex end like the old Chrysler products used rather than small splines, and a clamped style mount vs. through bolts? Just the first thing that popped in to my mind.

Since you’re strapping this to your back and flying up in the air, etc. while you’ve shown no signs of failure I’d still worry about cracks in the torsion bar eventually forming but being invisible since they’re covered. That said, what happens if the torsion bar fails? It at least sounds like a sudden, unexpected failure could be pretty unpleasant but this is purely armchair speculation (and I watched a Michael Bay film last night too). :slight_smile:

How about a more robust shape for the ends, like a hexagon, instead of a spline? It might be able to transfer your load better than a fine-pitch spline.

Another option might be to try using a bent steel rod. The shear modulus of steel is literally 20 times higher than nylon. The density of steel is only 8 times higher than nylon. You could make a steel part with equivalent stiffness (though with different ability to withstand torsional strain) that would be lighter and easier to shape. Maybe bend the ends of the rod to create features that arrest their rotation instead of a spline interface?

There is not only torsional twist, the is a longitudinal stretch as the ends rotate around a offset axis. Most of the solutions that we looked at for simply gripping the torsion failed to hold under the stretch force. The hex solution is great for steel in terms of forming and the shear strength. One of the interests in injection molding nylon would be the possibilities that open up for making a very large surface “key” on the ends.

Nylon in my experience (100’s of tons of material used in snowboard binding production) is fairly crack resistant. When it fails it fails completely. What happens if it were to fail on our jetpack during use is simply a sudden ease of turning. The tension is no longer there, everything works, just works faster, more jittery. No explosion :slight_smile:


Steel is an interesting approach. Any metal parts that we use have to be total seawater marine grade materials, 316 stainless in this case. The ends would certainly be more readily formed into a mount, the issue that we always saw was the stretch between the mounts. This could be dealt with by introducing spring type bends. A straight torsion bar would work great if we could put it through the center of the nozzle axis, which would put it into the stream of water requiring seals etc. The second factor is that due to its placement, the torsion bar always gets used as a handle, to lift the pack out of the water, carry, assist pilot. This requires a beam strength in addition to the torsion and stretch, in my estimation, the 8-10 mm torsion bar would likely get distorted.

Thanks for the input guys, nice to be able to bounce it around.

This is a fascinating thread.