700 replies to this topic

[Oberon]

Late to the thread, and you certainly have done a lot of work, but what you have built in post 271 is not close to being structurally the same as in 266.  If it were your results would be significantly different.  Not trying to be critical but structurally the are far from the same.  Also believe 8 is supposed to be a 120 and a 60 degree not 45 degree.  6 I believe is also supposed to be 90s....   
 
I'm interested as I have spider issues on a 36 f3.5 as well as having to design a spider for a 76" scope shortly.
 
Again not trying to be critical...

There must be a misunderstanding. In fact post #271 and post #266 are geometrically identical; both are type #8 spiders (codename “Crayfish”), and both have several other structural features in common (same size secondary, same mass and moment etc). The key difference is that post #271 has vanes made from slightly thicker 1.2mm carbon fiber v 0.9mm Stainless Steel for #266. Also the load is not as centered as I would have liked it to be, which hypothetically put the CF spider at a disadvantage, but even so it proved stiffer than the SS spider in post #266, better able to resist roll (which simulates effect an off-center load). Please explain where you see the significant structural differences?

So far as angles are concerned, the ideal curved spider will subtend a total angle of 180 degrees to ensure diffraction is spread out evenly across 360 degrees. Thus a 3 vane spider vane will subtend 60 degrees per vane, whilst a 4 vane spider vane will subtend only 45 degrees per vane. Only a 2 vane spider will need 90 degree vanes. In short, the drawings below show the optimised angles, and exceeding these angles as you propose will only serve to weaken the structure, add mass, and increase diffraction.

Also, as noted in post #260, not all 4 vane spiders will spread the diffraction over 360 degrees. Spiders #6 and #7 both spread the diffraction energy only over 180 degrees, meaning half the field of view has double the diffractive energy and half has none. This is less than ideal. Consequently, of the 4 types illustrated, only #5 and #8 are optimised.

Sidenote:  #5 would benefit from a larger central support, and I have a much beefier CF version of #8 to test yet, built to support a 5” secondary. Seeing as #5 is an optimised and popular design I will revisit it also with thicker vanes to better assess practical design requirements.

Edited by Oberon, 01 September 2019 - 10:04 PM.

[JohnH]

I still think simple usually trumps complex when it comes to limiting diffraction spikes and stray vibrations.

 

 

One has to come to the conclusion about which they find more objectionable in a design, and work from there.

 

 

This is a Gary Wolansky design from Vancouver British Columbia, and while it doesn't eliminate the fractions spikes, it does have a very nice design to eliminate vibration and hides virtually all of the parts that are not the actual vanes in behind the secondary mirror.

 

 

https://www.cloudyni...olansky-spider/

Edited by JohnH, 02 September 2019 - 12:20 PM.

[gatorengineer]

Oberon,

 

266 and 271 show configurations that are not the same.  266 is correct.  What was made in 271 does not look like 266....  The curving 120 degree arc, should be on the opposite side of what you have made.  Now the entire weight of the spider is cantellivered.  If you flip it to the otherside it will make a big difference.  The other two arcs are also much shorter in 271 than in 266....  which will also improve rigidity.

 

Thanks for all of your hard work, and its a great thread......

[bobruben]

I used a Gary Wolanski spider on my 16 inch dob. The spider acts as invisible wires for the dew heater. I added clear plastic tubing and fiber washers to the secondary support bolts that keep the spider  "double vanes" in tension to also insulate the two sides from each other, and used those two conductors (and two aluminum UTA poles) for the dew heater.

Kudos for all of your tests on spiders, it is interesting, and nice to see data rather than lore.

Attached Thumbnails

• 

[Oberon]

Oberon,

 

266 and 271 show configurations that are not the same.  266 is correct.  What was made in 271 does not look like 266....  The curving 120 degree arc, should be on the opposite side of what you have made.  Now the entire weight of the spider is cantellivered.  If you flip it to the otherside it will make a big difference.  The other two arcs are also much shorter in 271 than in 266....  which will also improve rigidity.

Granted the load is slightly cantilevered, reason being that the carbon fibre spider was not constructed for this experiment but was a spare I had on the shelf (see the story of its construction in the Sterope thread #114 here to #116 here). 

Thinking about it, I'll redo the test by fixing the fiber vanes to an identical aluminium box centre-section I've been using on the other spiders. As the vanes are currently only screwed to the wooden block (not glued) then this is possible. Maybe next weekend then.

[Oberon]

#10

Single flat stalk (5x50mm aluminium plate)

Meanwhile I couldn't resist trialling a single stalk, seen below as #10. After all, my personal driver for these test was to compare options for my planned Bino-scope Sterope, and #12, #13 and #15 are the options under consideration...

 

Granted the stalk is offset, and this would have some effect, but I wasn't concerned about minor effects, I first wanted to get into the ball park of what would be required (and note that I will use a much heavier 120mm secondary, tilted at 30 degrees not 45). A single stalk could be masked by either of the curved masks shown on #12 and #15, so I wanted to see how a 5mm thick piece of aluminium plate would perform in that role before cutting steel, or carbon fibre, or anything complex, time consuming or expensive. 

So first some photos...

 

There is no structural reason why the stalk sits at 45 degrees here, it was just convenient to do so. It does effect the direction of flexure measured with the lasers during rotation of the experiment, but not the magnitude. It also affects the direction and potentially the magnitude of the axial stress test (roll) where I hang weights off the 100mm golden screw (right). However that test had a trivial impact (blue), so again its really not that significant. However the axial pitch test (yellow) was significant (weights hung of the screw closest to camera), showing that the simple single stalk is most vulnerable to twisting, which I had not predicted and was the most interesting result.

Full size image here.

 

Other photos

 

 

[jtsenghas]

Since you're getting creative on names for the options, might number 13 be the "water beetle"? 

[Pezdragon]

If you really wanted to maximize the stiffness of a single stalk the material is fired alumina. It’s 5 times stiffer than aluminum. I use it on my 10” sled focusing Dob at its just a rock.

 

Attached Thumbnails

• 

[Bob4BVM]

If you really wanted to maximize the stiffness of a single stalk the material is fired alumina. It’s 5 times stiffer than aluminum. I use it on my 10” sled focusing Dob at its just a rock.

 Pezdragon,

GOT my attention !

I am in process of a large binoscope which has no UTA cage or rings & is also sled focusing...

So I have been experimenting with various Alum & steel single-vane & V-vane sec supports. 

All still have more flex than I want since I am supporting a 2+# sec mirror.

So your  fired alumina is a new one on me, where do you find this magical material ?

Thx for posting !

Bob

[Diego]

#10

Single flat stalk (5x50mm aluminium plate)

Meanwhile I couldn't resist trialling a single stalk, seen below as #10. After all, my personal driver for these test was to compare options for my planned Bino-scope Sterope, and #12, #13 and #15 are the options under consideration...



Granted the stalk is offset, and this would have some effect, but I wasn't concerned about minor effects, I first wanted to get into the ball park of what would be required (and note that I will use a much heavier 120mm secondary, tilted at 30 degrees not 45). A single stalk could be masked by either of the curved masks shown on #12 and #15, so I wanted to see how a 5mm thick piece of aluminium plate would perform in that role before cutting steel, or carbon fibre, or anything complex, time consuming or expensive.

So first some photos...



There is no structural reason why the stalk sits at 45 degrees here, it was just convenient to do so. It does effect the direction of flexure measured with the lasers during rotation of the experiment, but not the magnitude. It also affects the direction and potentially the magnitude of the axial stress test (roll) where I hang weights off the 100mm golden screw (right). However that test had a trivial impact (blue), so again its really not that significant. However the axial pitch test (yellow) was significant (weights hung of the screw closest to camera), showing that the simple single stalk is most vulnerable to twisting, which I had not predicted and was the most interesting result.

Full size image here.



Other photos

Question: if the scope is going to be used in alt/az, wouldn't it make more sense to place the single stalk vertically at 90 degrees such like spider #15 as noted in the post above?

Maybe I'm missing something?

As always awesome work Oberon!!!

[Oberon]

You’re right, which is why #15 is there, it is the least gravitationally challenged, so less likely to flex. However #12 has the best optical performance in that is has the cleanest diffraction effects. If it can be made stiff, it will be preferred. 

 

see here 2nd row left side, also comment from Nils Olif Carlin here

[jtsenghas]

Jonathan, it's nice to see that I'm not the slowest one here on experimental builds in the interest of experiments and design optimization as much as anything else! Has it really been four years since that thread you referenced?  When did you buy those secondary mirrors? 

Edited by jtsenghas, 16 September 2019 - 09:37 AM.

[Pezdragon]

 Pezdragon,

GOT my attention !

I am in process of a large binoscope which has no UTA cage or rings & is also sled focusing...

So I have been experimenting with various Alum & steel single-vane & V-vane sec supports. 

All still have more flex than I want since I am supporting a 2+# sec mirror.

So your  fired alumina is a new one on me, where do you find this magical material ?

Thx for posting !

Bob

I was lucky to find a piece at where I worked that was available. This material Is extremely hard when fired so diamond tooling is needed for any shaping or holes. I designed my “spider” to need nothing more than a plain flat rectangular piece without holes. A quick internet search found this as a possible supplier.

http://valleydesign....na96-stock.html

[tommm]

There are several suppliers of Al2O3 to the semiconductor industry, most hot pressed.  There is also Macor which is cheaper.  Al2O3 is cheap relative to other ceramics such as AlN, SiN, Boron Carbide, and Y2O3. Several of these were first developed for armor for the (deep pockets) U.S. military.  They each have their advantages, e.g. AlN is a good dielectric but has higher thermal conductivity than aluminum, and high thermal shock.  Most of the suppliers will send you spec sheets for properties of their materials.

[Oberon]

I get serious and buy a roller...

[Oberon]

I then get very serious and build some heavy duty curved spiders, in 2mm aluminium sheet and in 2mm stainless steel sheet...

 

 

The aluminium spider weighs 400g...

 

While the SS spider weighs 800g...

Edited by Oberon, 06 December 2019 - 07:08 AM.

[Oberon]

Both spiders utilise curved vanes with a 75 degree arc (I'll discuss that choice separately) riveted onto a 100mm aluminium tube 4mm thick.



I print a template to ensure the curve is set right. It turns out that for a 400mm diameter mirror with a 100mm secondary support (to suit my 120mm secondary) a 250mm diameter curve is just right for 75 degrees.

 

The secondary support components are much heavier than those used previously, and weigh in at 650g. As this is roughly what I expect as a total when using my 120mm secondary, then I expect the experiments to produce realistic results.



And for completion, here is the aluminium spider mounted for testing...



 

[Starman1]

Didn't your earlier testing show the curved spiders to be insufficiently rigid?

2mm thick (0.079") seems really thick. 

Was that thickness to see if increased thickness of the vanes yielded a greater stiffness?

[Oberon]

Yes.

[MitchAlsup]

Didn't your earlier testing show the curved spiders to be insufficiently rigid?

2mm thick (0.079") seems really thick. 

Especially when one has anecdotal evidence that straight vanes of 0.008" thick are perfectly fine for 4" secondaries and 20" primaries. 20× thicker and probably still not as stiff.

 

Is someone forgetting that the area opaqued by the vanes is equal to the amount of light diffracted into spikes?

[Oberon]

See the OP

[Oberon]

#16

3 curved vanes - 2mm aluminium

 

Arc = 75 degrees

 

[Oberon]

#17

 

3 curved vanes - 2mm stainless steel

Arc = 75 degrees

 

[mark cowan]

Stainless steel more springy?  More weight I suppose.   Still those are pretty **** stiff, like you'd expect. 

[Oberon]

Observations:

 

Both spiders share identical build techniques and are identically dimensioned.

 

The only difference between them is the vane material, aluminium v stainless steel.

Both spiders return good performance, both meet the criteria for a Paracorr. In fact their performance is practically identical.

 

This is surprising because the two materials were chalk and cheese in the roller. For example, the aluminium could be bent to the correct curve in a single roll through the tool, whereas the stainless steel vanes took about 20 iterations and was hard work (I wasn't counting, but I did not enjoy the process). Also the stainless steel was much harder to work in every way whether drilling the rivet holes or bending the joints. Suffice to say I was expecting the SS spider to be much stiffer than the aluminium but if anything its performance was very slightly worse (although the differences are getting down into the noise territory and are likely an artefact of the reference laser).

Both spiders had good vibration performance recovering from the "knockometer" impact. Interestingly the aluminium spider was better damped, presumably because it is a softer metal with less "bounce" than steel. 

However the SS spider was better able to resist the "roll" test (blue line). This was the only test where the SS outperformed the aluminium, and it is consistent with the difficulties I had in rolling the curves in the first place.

Conclusion:

This test combined with the next and previous tests pretty conclusively shows that by far the most critical determinate for a curved vane spider is vane thickness. Not material, not geometry.

Edited by Oberon, 06 December 2019 - 08:46 PM.