700 replies to this topic

[ckh]

Mark,

 

You have a point. The spring constant k for a steel wire is pretty high, but it acts as spring with a mass on the end (the diagonal with it's support). 

Edited by ckh, 22 May 2019 - 09:46 AM.

[tommm]

Merely a mass (secondary assembly) suspended on a set of springs (spider wires). The frequencies of the fundamental resonances are easily calculated if the mass and the various spring constants are known...

I think there is more to it than first appears Bruce.  It is not just the 3 D harmonic oscillator where the motions in the 3 orthogonal dimensions are treated as independent.  There is not even a "spring" in the z direction (along the primary mirror optical axis).  The restoring force in that dimension depends on elongation of the springs in the x and y dimensions, which depends strongly on the wire angles within the x or y plane (whether pairs are crossed).  In addition, on an offset spider the "springs" are offset from each other further complicating motion, and motions in pitch, yaw, and roll must be considered.  The moment arm for yaw can be as large as the one for pitch depending on design.  Most holders with wire spiders I have seen have significant moments in both.  The vibrations are larger in amplitude than what you would estimate using the 3 D harmonic oscillator.  This is obvious in the case of z axis motion with uncrossed spider wires where the spring constant of the restoring force is far less than that for elongation of the spring along its axis.

Edited by tommm, 22 May 2019 - 01:01 PM.

[mark cowan]

Merely a mass (secondary assembly) suspended on a set of springs (spider wires). The frequencies of the fundamental resonances are easily calculated if the mass and the various spring constants are known. In principle one could perhaps use eddy current damping but the suspension system for the magnets used would need to be stiff and well damped. Various active damping schemes are also possible in principle but present similar issues with the supports for the required actuators (pieozoelectric, etc) and sensors.  . 

Yes the wires act as springs.  An advantage of the geometries I've shown before is that the tension required is very low and this reduces the spring constant of the system, so vibration is discouraged.

 

But I don't want to derail Jonathon's thread further.

Edited by mark cowan, 22 May 2019 - 04:13 PM.

[bcarter1234]

Sorry if this is the wrong place to ask but there seems to be a great gathering of spider knowledge here. Most things I do are seat of the pants. From a material science standpoint do I gain any benefit from the extra 10mm of width of these vanes or should I saw them down to 20mm to match the thickness of the ply hub? I don't claim this is "one spider to rule them all" but I hope it will "measure up". The diagonal is a 3.1" MA in a 13.1" Dob. Thanks in advance for any assistance.

 

Take care,

Brent

[Oberon]

I don't see any benefit to cutting the vanes down by 10mm, nothing that is worth the effort.

 

OTOH I suspect from the photos that those rulers are fairly thick, like say 1mm or so, which if true is much thicker than ideal. If so there is optical benefit to be gained by using a thinner material, ideally less than 0.5mm, despite the inconvenience of not using an off-the-shelf steel rule. That said I've seen steel rules as thin as 0.6mm.

 

The other key improvement to make is to relocate your wooden block so that the spider vanes are offset from centre. Currently your spider configuration is like the spider shown upper right; if you change it to match the spider shown on upper left it will perform far better, being many times more rigid.

Edited by Oberon, 23 May 2019 - 07:54 AM.

[bcarter1234]

Jonathan,

 

Thanks for the quick and helpful reply. I'll finish this one up as is since another hour in the shop should do it. I can use it while I build another with offset vanes as you recommend. The stainless rulers are .5mm thick by 20mm section. At $3 each they were hard to resist since it meant not having to thin aluminum into long and narrow strips on the table saw.

 

So back to my seat of the pants nature. I can intuitively see that the offset vanes will provide higher resistance to axial twist. I might upon reflection expect it to be more resistant to side to side tilt. What I can't grasp is how it could be as stiff in fore and aft tilt (in relation to the focuser) with the mount cantilevered as it is. I'm not arguing the veracity I just love learning new concepts. Knowledge being such a fine compliment to intuition. Please explain if you will.

 

Take care,

Brent

[Oberon]

OK, goodo, 0.5mm thin rulers is great, and thinner than they appear in the photos.  

The cantilever risk is nothing compared to the axial twist issue. And in any case your mirror will be cantilevered below. Also, with an offset spider it is possible to sneak the mirror COG up very close to the spider's structural center, within the vanes, and minimise the cantilever effect.

[bcarter1234]

Jonathan,

 

Doh! An excellent example of not seeing the forest for the trees. Had I considered this in context I might have realized the secondary in use moves mostly sideways through an arc such that the force of axial rotation experiences a significant (greater?) delta. Thanks for the clarification. 

 

Take care,

Brent

[bcarter1234]

The spider is installed. Of course leaving the rulers 30mm wide was unthinkable so they were cut down to 18mm to match the ply hub. I must have saved nearly 10 grams on the UTA. Obsession is a virtue no?

Just as Jonathan predicted it is dead stiff in every direction excepting axial twist, rather like a radial laced bicycle wheel in that regard. I'm sure it will serve. The next iteration will use offset vanes as recommened. Thanks for all the great information presented here.

 

Take care,

Brent

[bcarter1234]

Upgraded to the offset spider as suggested. Just as promised it is much stiffer in axial twist. I took the scope out for a few minutes tonight and had the best star images at 200X I've ever had. While using the autocollimator I could see the center spot reflections slip back into a close jumble as the altitude was changed. The movement seems to be coming from truss connections rather than the spider. Time to convert to a hexapod...

 

Take care,

Brent

[Oberon]

Good work!

Before going to a hexapod, which for your scope may be a fundamental change, suggest tweaking your current trusses to improve the triangles. That is, make the truss poles join on the ends, ideally sharing the some bolt/screw/pin. You will be amazed at the effect on rigidity. 

Suggestion; start a new thread to theme of improving rigidity on your scope, include plenty of photos of what you have now, and you'll attract a wide variety of good advice from experienced atm's without risking a derail of this one. 

 

[Oberon]

And at last...by Ed's request...behold the knockometer! A standard test for a spider's ability to resist vibration and settle.

 

Essentially a round wooden block hanging 1000mm from a hook fixed to the ceiling, when at rest it is only a few mm from the UTA...

 

...so with the help of a steel rule I give the UTA 3 standard hammer taps, letting it swing from 50mm, 100mm and 200mm.

 

The block with washer weighs 500g...

 

...and the entire swinging mass - an 1200mm M8 threaded rod, nuts, washers and wooden block weigh 948g.

 

Sorry, somebody else can work out what all that translates into force applied!

[Oberon]

So I ran the test on my heavy duty steel vane offset spider, which as expected performed exceptionally well. As the vibrations had settled within a single frame there isn't much to show, but here are the pics...

 

Standing still...

 

 

Then 50mm...

 

 

then 100mm...

 

 

...and finally 200mm...

 

 

As I said, the vibrations settled down so quickly I only captured a single frame at 30f/s, but I don't expect that to be the case with other spiders. In any case I'll work out a standard way of measuring the data and include the test with each spider we look at.

[CrazyPanda]

Sorry if this is the wrong place to ask but there seems to be a great gathering of spider knowledge here. Most things I do are seat of the pants. From a material science standpoint do I gain any benefit from the extra 10mm of width of these vanes or should I saw them down to 20mm to match the thickness of the ply hub? I don't claim this is "one spider to rule them all" but I hope it will "measure up". The diagonal is a 3.1" MA in a 13.1" Dob. Thanks in advance for any assistance.

 

 

 

Take care,

Brent

Depends. Do you plan on two connection points at the ends of the vanes, or one? If one, then there is no advantage to keeping them 20mm wide (though I agree with Jonathan that cutting them down in width is not worth the effort). If you have 2 connection points though, it will add rigidity with less tension required.

 

That being said, with steel rulers that thick, two connection points probably won't be a huge advantage. However, in the thin 0.008" aluminum flashing I'm using for my vanes, having two connection points makes them considerably more rigid under lower tension.

Edited by CrazyPanda, 31 May 2019 - 11:32 AM.

[mark cowan]

Nice job with the deathspider, er, knockometer.  The number of interest is just damping time, which is easy to see but perhaps difficult to measure looking at displacement...

[Ed Jones]

Wow,well done! 

[Oberon]

Nice job with the deathspider, er, knockometer.  The number of interest is just damping time, which is easy to see but perhaps difficult to measure looking at displacement...

Thats right. For the above I ran the camera (GoPro) at 120 frames per sec. The intention was to capture damping time accurately. The effect was merely 4 x identical frames, so only 30 frames per sec. I’ll do some research on that, so this isn’t the end of the story, but the flip side is that once I’d proven the system to work and in the right zone I pulled that spider off so that I could proceed with the real “like for like” comparisons featured a few posts ago. Only later did I discover that the 120 f/s hadn’t worked as intended.

 

In any case, for our purposes 30 f/s should be perfectly adequate, I’ve just got to settle on a practical and meaningful way of quantifying it.

Edited by Oberon, 31 May 2019 - 05:44 PM.

[Oberon]

Good news. It turns out that the GoPro has an auto low light setting that reverts to 30f/s under low light, thus quadrupling the frame to satisfy the 120f/s setting. Next time I'll turn it off.

[Bob4BVM]

So  Jonathan,

Reading thru all your work here to date, am I correct to assume that the standard sheet metal vaned spider beats out the wire spider ?

I am CRUSHED !!!

Bob

Edited by Bob4BVM, 03 June 2019 - 12:54 AM.

[Oberon]

Not so fast. I’ve yet to build and test an optimised wire spider. But that said, an optimised solid vane spider will always outperform a similar wire spider due to the added material. That a wire spider will have more flex is a given if all other parameters are the same. 

 

Merope’s wire spider is not optimised due to the shallow UTA and the high mass center support, and is barely adequate. I expect to do much better. However I don’t expect it is possible for any wire spider to outperform the solid spider I tested, except where it matters with respect to diffraction. But it doesn’t have to either.

[MitchAlsup]

Merope’s wire spider is not optimised due to the shallow UTA and the high mass center support, and is barely adequate. I expect to do much better. However I don’t expect it is possible for any wire spider to outperform the solid spider I tested, except where it matters with respect to diffraction. But it doesn’t have to either.

I fought with several wire spiders over a couple of years trying to get them to enable the telescope to move like binoculars. I failed at least 3 times.

 

Then I built a vane spider with 2.5" vane height, and never had another vibration problem from the secondary assembly. And the vanes are 0.007" thick lightly dusted with flat black paint, and have tolerated the abuse of storage, travel, setup and take down for 15 years.

 

This is why I like vane spiders. They are not as demanding on the engineering side and just plain work.

[Oberon]

I’ve not had vibration issues. All we observed from the test was increased flexure when compared to solid. Vibration issues suggest (to me) potential geometry issues, just as with vanes, but without more information I’m just guessing.

[Oberon]

And now the real testing begins.

Everything before was just preliminary set up.

[Oberon]

The first group of spiders compared are pretty standard straight vane spiders.

Each of them are lightweight but stiff construction from extruded aluminium section with 0.5mm thin aluminium vanes. The secondary is 75mm, and its central support ensures that the mirror mass has a 100mm moment from the center of the vanes. So far as possible each spider uses the same construction techniques. 

Two are centre mounted, two are offset.

Two have a 50mm central support, two have a 20mm support.

 

Essentially we are comparing the stability of offset and centre mounted vanes, and of a larger/smaller centre.

 

[Oberon]

#1

Full size image here.

https://www.cloudyni...-05-5050/large/