Improved Experimental Pendulum

.

There ^^^ I suggest, is the essence of the problem.

In a more perfect world, your pendulum would be in solitary confinement within the structure … and all the ancillary gubbins would be outside.

The pendulum is light in weight, so stiffness is more important than strength.

[much too expensive, but] a filament-wound carbon fibre cone, bonded to a plate, might be mechanically ideal

… working-down from from that concept gets us to the tripod !

a proper one, not three parallel uprights

I’ll get something to eat now and leave you in peace

MichaelG.

Why not an A-frame design like my free pendulum used? You (mostly) only need rigidity in one direction anyway, and an A-frame is likely to be better than most vertical-only supports.

Could I just sound a note of caution on the issue of support compliance as I think it has come up here and in other threads? As a pendulum swings there is an equal and opposite restoring force on the suspension support, corresponding to the gravitational restoring force on the bob. In effect there is a "gravity spring" on which the bob moves, one end of which is attached to the "support spring", and the two are in series. The spring constants add as "reciprocal of sum of reciprocals". Thus the compliance of the support adds to the "gravity compliance" and changes the effective force of gravity. If the support compliance can change, for example with temperature, it can affect the period. Also, if the support compliance is damped, for example by using rubber feet, it can absorb energy from the pendulum.

I once came across an interesting design for a suspension for an electron microscope column that needed to be isolated from external vibration. To do this it was supported on vacuum-filled bellows, which have only a tiny natural compliance of the bellows material - the vacuum itself had zero compliance but could support the weight. Naturally some damping was needed, but as little as possible as the damping itself would transmit vibration across the support.

I think this means that the pendulum support needs to be as rigid and massive as possible, and have minimal damping. Or alternatively one avoids the problem by nulling out the pendulum force on its support by using a double pendulum.

Hi John

Could you explain a little more about your vacuum filled bellows. I can’t imagine how that works?

regards Martin

A subtlety with my suggestion might have been missed. I am not suggesting a closed tube, I am suggesting starting with a closed tube and then milling oval shaped access holes. These holes can be quite big - as long as large webs are left top and bottom, a lot of rigidity will remain while allowing good access to the mechanism.

Think of the steel roof trusses in your local supermarket or DIY store - commonly have large cut-away diamond or circular holes all along the centre web of the I beam, but retaining (most of) the structural strength, Now imagine that sort of thing rolled into a tube with the access slots around the side wall of the tube, blending into a full tube at the top and bottom, so it would give plenty of access but vastly improved rigidity.

I would draw what I mean but am busy battling with a broken central heating system - during the coldest part of winter - of course !

Edited By John Doe 2 on 07/03/2023 11:57:32

No need to draw it John, that's what I understood you to mean. I haven't dismissed the idea, because rigidity is important. The advantage of my tripod is it's easy to make, and provides lots of big gaps. The disadvantage is a tripod can't be as rigid as a tube, or cross-braced structure. Duncan suggested a Shukhov Tower in the other thread:

These minimise material whilst maximising strength and rigidity. My problem is designing one - it depends on hyperboloid curves, and then assembling it. And it leaves no space for fingers!

So I have to make compromises, and it's very possible I'll get them wrong. I hope someone else will have a go, because I'm getting more good ideas than I have time to try!

Though I've not commented on every suggestion made by forum friends, they've all got my attention, and none of them have been rejected entirely. Some are incorporated in the Mk4 clock, and I've no doubt it could be improved, especially if some of my restrictions are ignored, or met by some other clever arrangement!

Finding time is a problem, especially as I think and do practical stuff slowly, having to fix many silly mistakes as I go. So I'm trying to keep it simple, which is always hard work!

Dave

Think of an aneroid capsule, which is just a bellows with one or maybe 2 folds. Inside there is a vacuum or gas at low pressure. When you pull the ends apart air pressure exerts a force on each one against the pull. If there is actually a vacuum inside the amount of pull doesn't depend on the amount of movement. So except for the bellows material itself there compliance is infinite. Electron microscopes have plenty of vacuum available...

Hi John

Yes I can see that arrangement would produce a constant force tension spring. It was you saying that the vacuum supported the microscope so I was thinking in terms of compression.
Thanks Martin

Been out visiting and to buy food so the block is shaping up slowly. Had to go deeper than expected to get down to clean metal, but it will do.

Could stop here, except I don't often get the chance to square-off a hefty block and need the practice. HSS won't touch the scale and I think the quickest way to shift it is with an angle-grinder.

Decided, again, that I need a bigger milling machine! No chance, unless I buy another house first. As can be seen my biggest end-mill doesn't have the length needed to tidy the edges in one pass:

Tempting to order a carbide shell mill for this but I'll see how my fly-cutter gets on first.

The job is incredibly dirty, black dust everywhere, and the contents of my nose are extra disgusting. (Bogey photos available if anyone wants evidence!)

Probably spend this evening exploring frames with Solid Edge.

Dave

An unexpected feature of the Shukov tower is that all the diagonal bits are in fact straight, then you just have some simple hoops going circumferential, but it wasn't a serious suggestion. Would be a good talking point tho.

.

So it’s an impressively tall structure comprising a stack of the shapes we discussed before, but with the extra feature that the top hoop of each module is smaller than the bottom.

https://en.wikipedia.org/wiki/Hyperboloid#/media/File:Cylinder_-_hyperboloid_-_cone.gif

MichaelG.

.

Extending the metaphors … my proposed filament-wound carbon-fibre cone would be similar to the fuselage of the Wellington bomber.

https://www.barneswallisfoundation.co.uk/life-and-work/geodetic-aircraft-design/

Edited By Michael Gilligan on 08/03/2023 08:39:55

This is what I mean - of course the images have come out sideways, and I cannot find how to rotate them but I am sure you get the drift.

Moderator Note:  Unfortunately only moderators can rotate photos. It's a forum quirk and turning them is my main service to the community!     Dave

Edited By SillyOldDuffer on 08/03/2023 11:37:32

Played with some pillar layouts last night with 2D QCAD, which is easier to use for this job than SE:

I measured the space available for my fingers between pillars in three vertical configurations. Equalateral-tripod, square-quadripod, and rectangular-quadripod. Sorry about the blurred numbers - the tripod gives 72.3mm access, square 57.2mm, and the rectangle 70.77mm.

My hand: 4-fingers across = 90mm, 4 fingers + thumb = 110mm, and forefinger + thumb = 40mm.

The advantage of vertical pillars is space and simplicity of construction. A tilted tripod supporting a 240mm high platform would be angled at about 75°, requiring me to drill diagonally into the base and platform. Could be done, but it's harder to do and more error prone than going straight in with a milling machine. Perhaps in the Mk V!

The equilateral triangle both maximise stability, whilst the rectangle gives better access in exchange for less stability sideways (compared to a square.)

Spent this morning reading about columns with discouraging results. Apart from the maths boiling my brain, my book says things like "Great and apparently insuperable difficulties arise as soon as we attempt to obtain expressions which relate the strength of longer or more slender columns to their various dimensions."

It usefully explains why circular pillars are better than square or other shapes, the importance of fixing the ends, and would help if knowing the safe load was a concern. The safe load doesn't matter because my bob only weighs about 40g and even a spindly tripod will easily support it. And then the chapter on eccentric loads, which covers my case, doesn't cover dynamic loads (a moving pendulum), or say anything about rigidity, which is my chief concern!

I hope Solid Edge's Simulation tools will help, but frankly have to admit I barely understand the basics.

Going to spend an hour pondering what to do whilst de-scaling cast-iron in the snow. I guess cross-bracing a quadripod is the answer, but I'm not sure how to do it whilst allowing access. Before deciding, I need to model the chariot that carries the sensors and electromagnet. Either it fits through the gap or I have to emplace it first and then erect the scaffold around it. Hmmm.

Dave

Whatever you decide to do, Dave … please keep us informed

This is exciting ‘spectator sport’

… and I would be particularly interested if you can share what Solid Edge can tell about vibration modes.

MichaelG.

Slow progress for various reasons, but I'm getting there.

This is the base squared off and tapped to take the support columns:

And here I'm tapping the holes for the M12 adjustable feet:

Annoyingly the Aluminium tube needed to complete the four pillars has gone walk-about and I've decided to buy more.

The only remaining major operation on the base is cutting the 4mm wide circular slot needed to take a 4" diameter plastic soil pipe. Not sure whether to trepan the slot in the lathe or mill it. The problem is how to hold a block of that size. May open another thread on the subject - it's work-holding and cutting rather than clock related.

I'm dithering about where to put the vacuum outlet too. Under the block's centre of gravity for balance, but that puts it directly under pendulum BDC, where running the pump seems most likely to disturb the bob. Dithering about the rod as well, having gone off thin carbon-fibre, so probably another question coming in the Precision Pendulum thread.

Applying "lessons learnt" is proving painful!

Dave

Edited By SillyOldDuffer on 04/04/2023 12:31:39

Decided the base block was too big for my rotary table without making some sort of platform, so I'm going to trepan it. The block clamps nicely to my 240mm diameter faceplate using its levelling screws:

The milling table clamp on top is for balance, not holding. There are 3 of them, and I'll find out how effective they are when the faceplate is powered up. Only about 100rpm, but it will be shake, rattle and roll if the balance is off.

Now all I have to do is grind a trepanning tool, and I'm infamously cack-handed. What could possibly go wrong? Fingers crossed!

Dave

Worked out OK :

And the test section of pipe fits perfectly, phew:

Dave

Edited By SillyOldDuffer on 06/04/2023 19:48:27

Great progress, Dave

Presumably you will eventually be bedding the tube onto Silicone sealant or some-such

MichaelG.

Yes, that's how the Mark I was attempted. Didn't work well, slow leak, but seal finish was poor and there were 4 unhelpful through holes as well. (for the pillars) This time, I'm going to try an O-ring first. However, the construction depends too much on the weight of the pipe and air-pressure to squish the ring for my liking - it isn't a proper compression fitting. So silicon goo is the odds-on favourite.

Not too worried about the other end of the soil pipe. It came with a flange incorporating a neoprene seal - wrong way round - but I can glue the tight fitting top cap in as well.

The vacuum design has me uneasy for other reasons. Positioning the vent already mentioned, but also how to keep it and the electrical wiring air-tight. Ideally, the pump should take the pressure down once and thereafter the container never leaks. Fat chance! Other concerns: where to put the vacuum gauge; pipe size; will an ordinary water isolation valve be good enough; and making sure running the vacuum pump doesn't shake the whole shebang.

Slow progress - I'm wasting loads of time on domestics.

Dave

Hi Dave

A matrix of pins embedded in araldite in a flange of your choice makes for a good vacuum feedthrough although there are plenty of commercial versions available.

regards Martin