How does a viscoelastic brake mechanism work for satellite deployment systems?

[Enthalpy]

Hello everybody!

I imagined and prototyped this brake long ago for the deployment mechanisms of a satellite I never built. In essence, it's clean, it can work in vacuum and in a wide temperature range, and it can wait for long without maintenance, so other uses must be possible, like semiconductor, food, optics processing...

It compresses a viscoelastic element and let's it roll to brake the relative movement of two parts, so that an alternate deformation results and dissipates energy. My viscoelastic element was a seal ring and the other parts were cylindrical - but feel free to make them flat or elliptic if it brings anything. Two gentle slopes help a lot inserting the ring. Repetitive operation would demand stoppers, whose shape is simplified on the sketch, as the rolling surface is better smooth right from the slope.

The braking resistance is difficult to predict unless you have a software for viscoelasticity and know the material's properties at varied frequencies, temperatures and stress... Even then, experimental verification remains necessary, so you can just skip the simulation step. The braking resistance increases with compression and scales as the dimensions squared.

The initial orientation of the ring is almost unperceivable, but after hours, the elastomer creeps and makes one position more stable. With the material I had (NBR?) this was important but didn't prevent movement under the expected force. Whether some material can damp at 10Hz but creep little over weeks is unclear.

Varied elastomers are used for seal rings, and all elastomers dampen except natural rubber. Silicone and Viton fit secondary vacuum and dampen a lot over varied temperatures. Creeping and limited friction are a drawback for Viton. The "viscoelastic properties" and "complex shear modulus" (search keywords) depend on the precise material including its molecular weight and cross-linking, which define the relation between temperature, frequency and the complex modulus, so "polybutadiene" or "polyurethane" is too imprecise.

The imaginary part of the shear modulus must increase with frequency over the whole desired temperature range, so the braking force increases with velocity and can stabilize the movement's speed. It's usually given at zero compression, alas.

My rod and tube were of POM-H. Examples of clean materials with great friction are nickel layers, anodization layers on aluminium, titanium, stainless steel, ceramic layers... but most bare metals and plastics seem to suffice.

Two viscoelastic elements in a brake help keep the moving elements parallel.

Marc Schaefer, aka Enthalpy

(Click on the thumbnail to view the sketch full-sized)

 

[Enthalpy]

A shock absorber can be built a similar way, especially if the compression of the viscoelastic element(s) increases over the stroke to compensate the decreasing speed.

The compression can be small in the waiting position, so the viscoelastic element creeps little - but beware the element must roll, not glide, soon at the stroke's beginning if not right from the beginning.

Clean, free of maintenance... like the brake.

Marc Schaefer, aka Enthalpy

 

[Danger]

Perhaps this is due to either fatigue or beer, both of which I have an abundance of, but I can't quite follow your plan. More detailed illustrations would be helpful.

 

[Enthalpy]

Here you are, a sketch of the shock absorber.

The slopes (which need not be uniform) squeeze the viscoelastic elements increasingly over the stroke, so the braking force doesn't have to drop as the speed does. More viscoelastic elements can also be brought in the narrow space as the piston travels.

As it now looks, using many viscoelastic elements can make this dry shock absorber as small as, or maybe smaller than, a hydraulic one. Heat is a remote limit (silicone rubber: 1500 J/kg/K, 1400 kg/m³), and the braking force relates indirectly with the squeezing pressure and with the larger wall's area instead of the head area as in a hydraulic brake.

Merry Christmas! - ¡Feliz Navidad! - Frohe Weihnachten! - Joyeux Noël! - Feliz Natal! - Buon Natale!
Marc Schaefer, aka Enthalpy

 

[Enthalpy]

The elastomer I used (NBR?) in my trial did not slip, but a perfluoroelastomer might more easily - or a fatty environment or high speed might lead to slip.

Two films wound around the viscoelastic element would prevent slipping shall this occur - with more turns than on the sketch below... Natural candidates are polyester or polyimide films, maybe a thin elastomer film containing fibres like a tyre does. The films don't need to wrap the whole width of the viscoelastic element nor cover the width in one single piece.

Threads are an alternative to films; instead of spirals, they can be wound as helices.

I give the idea, someone else shall find out how to assemble the thingy, especially with many viscoelastic elements... Split the shell in two parts, like a steam turbine? Hopefully brakes and dampers won't slip more than my prototype did. Corrugated running surfaces must help, high friction coefficients as well: aluminium bronze, nickel layer, ceramic layer...

Marc Schaefer, aka Enthalpy

 

[marcusl]

Your mechanism needs to provide the same damping at any temperature encountered in space. Will your materials do that? (I'm not a materials guy...)

 

[Enthalpy]

Your mechanism needs to provide the same damping at any temperature encountered in space. Will your materials do that? (I'm not a materials guy...)

Elastomer damping depends on temperature, absolutely right. Here some materials are better, especially silicone rubber, which is used as a damping material on spacecraft for that reason.

One may also combine several materials (in separate elements) whose highest damping lie at different temperatures, so that damping depends little on the temperature.

Some uses demand a constant damping, like the landing shock absorber of an airliner, other uses less so. Whether a satellite antenna takes 2s or 30s to deploy isn't important.

Marc Schaefer, aka Enthalpy

 

[Enthalpy]

I've sketched the viscoelastic elements as rings up to now, but any rotatable shape fits: ball, roller, truncated cone, gear... In the sketch below, balls are squeezed between two races to brake a rotation.

Usual construction variants of bearings apply: axial, angular contact, roller... Some would let adjust the braking effect through variable compression. Standard mechanical ingeniosity shall draw parts that can be assembled.

Such a rotating brake has some bearing capability which dampens vibrations. Combine with usual stiff bearings if preferred; one might even share the races among hard and viscoelastic rolling elements.

The path stability of rollers, needles and cones should be thought through or experimented. Maybe these can receive an internal hard axis to guide them in a cage. If using a cage with plain elastomers, it should slip well: consider perfluoropolymers, perhaps with a shape elastic enough to accommodate the elastomer.

The amount of squeezing can vary along the rotation angle if useful, for instance using oval races. Say, a dish antenna could hold on a satellite by soft damping bearings during flight, then open gently, more slowly at the end.

Braking a rotation over many turns has also uses, for instance at a differential, but the capacity of the viscoelastic elements to dissipate heat should be checked.

Marc Schaefer, aka Enthalpy

 

[Enthalpy]

The brake might be good for doors that close automatically (commonly by a spring). Especially if braking stronger at the end. The rotating variant looks simpler to integrate in a door.

More doors: a trunk door on a car. Presently they have a nitrogen spring that compensates the weight and brakes at the same time, but it tends to leak over the years.

Seen more similar doors, for instance at vacuum apparatus, that should not fall freely and have a brake for this reason.

Marc Schaefer, aka Enthalpy

 

[Enthalpy]

Some mechanisms use no spring but raise a door as it opens, letting the weight close it automatically.


This make it simpler to compress the viscoelastic elements more at the end so they brake more: build them as an axial bearing, or nearly as a usual axial bearing but conical...


If using a spring to close the door, a similar added movement at the viscoelastic elements can compress them at or near the end, using an oblique rod, a sliding slope...


The braking action can be releaved as the door is nearly shut, to get enough speed for the strike, and avoid a permanent compression of the viscoelastic elements.


Marc Schaefer, aka Enthalpy