JWST

[Screwdriver]

This is taking forever!

The tricky part (one of the many tricky parts) was getting the massive radius to intersect with the heaxpods. Has to be done in Sketchup (for me) and it has to be printable. This is going to be a model after all so I need to make parts that look about right but can still be printed as individual STL files.

It will also fold up like the real thing (that was also tricky to design in and not actually tested yet). I won't be making the sunscreen folding. That's way too complicated with tinfoil/mylar in the way.

40:1 scale so it will be pretty big!

(Oh, and I really want to stick a light in it, a little LED at the secondary mirror, just for fun...)

 

[Screwdriver]

This only took me all day...

Even at 40:1 which is a pretty big model (over half a metre!) the secondary mirror tripod is really spindly. I hate seeing models which are obviously too chunky. A bit chunky is ok but it should at least resemble the object.

I may need to incorporate a few metal rods...

(hinges are going to be a nightmare at this size).

 

[Screwdriver]

Some detail work on the mirror assembly and completed hinge mechanism.

Incredibly, it's tricky to find a good picture of the assembly as it will be deployed (without all the protective covers and gravity clamps). At least it no longer looks like a rabbit. No more boring Sketchups, next stop Ender3...

(then redesign the hinges I bet).

Bah. Couldn't resist giving the hinges a virtual test.

Perhaps unsurprisingly, after $10 billion, it folds up rather well. You can see this thing has been really thought through (and no I don't mean my crude attempt at copying it!).

 

[Screwdriver]

Lot of detail work on the hinges and telescoping arms.

I am dithering about with the folding nature because if there is a screen fitted, folding it will be a nightmare. The original JWST is using very thin kapton tape and frankly, that is also the thinnest material I could get away with using on this tiny model!

So I am considering how to design a fully deployed version which has a detachable sun screen which can be converted into a folding version with say, only one sheet for operability.

It does fold up though and you can see why some of the elements were designed as they are.

..I mean, it will fold up, once I get around to finalising it and printing it up.

 

[Screwdriver]

Finally!

...so, err, not much progress I grant but every step is a step closer. This is a test hinge, part of the primary mirror folding mech, shown here:

Four of them and they are separate from the main bodies because they are so fiddly and I'm not sure if I am happy with the hinge pin - largely because they're still on order (micro machine screws). At a later date I might even redesign them and print in place. They will either dovetail into the body or just superglue, not decided yet.

The other reason being, I don't have any more black PETG so I am using the tail end of a roll to try out the smaller parts. Plus I want to see how small I can go. These are a bit chunky but I have hidden them inside the body.

This next hinge part took ages to concoct. The JWST has angled rectangular section support arms and I wanted it to look right. Sketchup hates odd angles and the odd axes makes it super complex because the "push/pull" tool basically heads off on the axis which may or may not be along the body of the part if the same "group" has odd angles in it. You can separate groups but then you're in a world of pain when it comes to keeping things manifold for 3D printing.

That's the lower swingarm for the primary mirror boom and it's a bit like a giant wishbone. Might have just enough PETG to do a test print but based on the above, I think I may be easily over the threshold for it being strong enough. Might even be able to slim it down though it is almost on scale.

The first piece proper will be the fixed primary mirror (you can see a bit of it above. That's a 12 hour print all on its own. One of the reasons I am keeping the complex smaller parts separate. That piece will be sanded down, polished and get some gold foil so I don't want to keep remaking it!

 

[Dr.Al]

I'm very impressed that you managed to design that with Sketchup. Most times I've tried Sketchup I've ended up creating a whole new vocabulary of expletives.

 

[Screwdriver]

I'm very impressed that you managed to design that with Sketchup. Most times I've tried Sketchup I've ended up creating a whole new vocabulary of expletives.

It is tricky because there are many flaws (in the elderly version I use - 2017) and it is really a 3D sketching application, designed for architectural drawings. It's also difficult to export to .skp (I have to use a plugin, hence the elderly version) and intersections of curves is a nightmare.

Having said all that, there's nothing I have found that can translate a shape from your head into a 3D model with anything like the sheer speed of Sketchup. Means you can prototype a thought in seconds, see if it works and modify it until it does then, crucially, tweak it until it LOOKS right. That's the biggie for me. Form follows function but as often as not, if it looks right, it probably is.

The issue I have with mainstream CAD software is you can't quickly manipulate an object while holding the tool you want to use next. With Sketchup you can twist, rotate, spin, zoom, centre all from the mouse while at the same time selecting, pushing, pulling etc. You can grab individual objects and arrange them t your whim. Makes this JWST model actually pretty easy though I appreciate, it has too many foibles for some, especially for regular CAD.

 

[Screwdriver]

Look what my printer made...

I am seriously impressed with this Ender. Can hardly believe it will work at this scale. The tabs on those hinges are less than 1mm and the accuracy is perfectly acceptable for a model this size.

I made a little bump inside (0.4mm) so the interlocking cones stop the joint from splaying.

...and while I am quite happy with the result, this is just a test print before my micro screws turn up. This one is just from my random box of impossibly small screws.
Works though, straight as a nail, strong enough even at this scale and locks in position without going over (which would look horribly lame).

Incidentally, I also printed them in the folded position, one on top of the other. I had such a torrid time with supports being stronger than the part, I used the parts together to support themselves. Just needs a bit of planning in the design stage so you can get rid of supports altogether.


<edit> screen shot from Cura.

If you're careful with layer height and double check in Cura, you can get away with really tiny "gaps" that aren't even there. The cone shapes also help by putting the additional strength where it's needed when you twist them apart.

 

[Dr.Al]

Look what my printer made...

Wow, I wish my printer would print pound coins

 

[eLuSiVeMiTe]

It is tricky because there are many flaws (in the elderly version I use - 2017) and it is really a 3D sketching application, designed for architectural drawings. It's also difficult to export to .skp (I have to use a plugin, hence the elderly version) and intersections of curves is a nightmare.

Having said all that, there's nothing I have found that can translate a shape from your head into a 3D model with anything like the sheer speed of Sketchup. Means you can prototype a thought in seconds, see if it works and modify it until it does then, crucially, tweak it until it LOOKS right. That's the biggie for me. Form follows function but as often as not, if it looks right, it probably is.

The issue I have with mainstream CAD software is you can't quickly manipulate an object while holding the tool you want to use next. With Sketchup you can twist, rotate, spin, zoom, centre all from the mouse while at the same time selecting, pushing, pulling etc. You can grab individual objects and arrange them t your whim. Makes this JWST model actually pretty easy though I appreciate, it has too many foibles for some, especially for regular CAD.

Big fan of SketchUp. I concur with your thinking.
I've added a 3d mouse. A worthy investment. So much easier to get where you need to be

 

[Screwdriver]

While I was waiting for my PETG to turn up, I have been twiddling around with some of the smaller parts. I know, I said to myself, I'll get some gold leaf and gild the surfaces. Just like the real thing.

How difficult can it be??

Not acceptable! And took ages. That stuff is fantastically tricky to handle though that challenge in itself is quite good fun. I made up my own "size" from PVA and I think that was the biggest mistake. So I bought some proper size but I think this idea is a no-go. Not for a mirror because to go to all that effort, it would need to look a lot more like a mirror.

Here it is half way through a print (6 hours). I am printing the primary mirror in two halves sliced down the centre vertically because of the curve in the back.

My next test will be to glue some gold plastic foil to the sections though I will have one more crack at that secondary mirror once the correct "size" turns up. PVA is just too thick. Those foils are impossibly thin, about 0.1–0.125 μm or just a few hundred atoms!

I got it everywhere and it would stick to anything. Anything except the item I was intending...

 

[mtt.tr]

The issue I have with mainstream CAD software is you can't quickly manipulate an object while holding the tool you want to use next. With Sketchup you can twist, rotate, spin, zoom, centre all from the mouse while at the same time selecting, pushing, pulling etc.

This is also Thier strengths everything is based of a datum and parametrically aligned which in your use case is not something you'd find useful.

But being able to adjust a clearance on a part using a table makes adjustments so fast and it doesn't brick the model as everything is tied together.


You might find blend3r more what you are after

 

[Onoff]

Wow, I wish my printer would print pound coins

Why can't it?

Sketchfab

sketchfab.com

 

[ukracer]

Big fan of SketchUp. I concur with your thinking.
I've added a 3d mouse. A worthy investment. So much easier to get where you need to be

What 3D mouse did you end up getting?

 

[eLuSiVeMiTe]

What 3D mouse did you end up getting?

https://www.amazon.co.uk/3Dconnexion-3DX-700059-Spacemouse-Compact-Mouse/dp/B079Z3T2XC/ref=asc_df_B079Z3T2XC/?tag=googshopuk-21&linkCode=df0&hvadid=310552784321&hvpos=&hvnetw=g&hvrand=3022313139058374057&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=1006506&hvtargid=pla-489967759395&psc=1&th=1&psc=1

Think I paid 60

 

[Screwdriver]

This is also Thier strengths everything is based of a datum and parametrically aligned which in your use case is not something you'd find useful.

But being able to adjust a clearance on a part using a table makes adjustments so fast and it doesn't brick the model as everything is tied together.


You might find blend3r more what you are after

The parametric modelling feature is a big plus point of course, when you want to nudge something a bit, resize etc. So if you know beforehand what you want, then its great but I work at speed, roughing shapes out until they look right. I can only "see" that in 3D when I grab the object and spin it. The ability to manipulate an object or part of an object is so slow and cumbersome in a dedicated CAD software (I have tried them all), they are less useful to me for prototyping/one offs.

I did start teaching myself Fusion but then they took away another attractive option, generative design. That feature was not only a big motivation for me to learn it but also for Autodesk to unilaterally castrate the free feature set was a worry for the longer term. I am sure (relatively sure) I could get some speed out of Fusion after say a year or so but I don't want to spend a year teaching myself something that suddenly is no longer free.

 

[mtt.tr]

The parametric modelling feature is a big plus point of course, when you want to nudge something a bit, resize etc. So if you know beforehand what you want, then its great but I work at speed, roughing shapes out until they look right. I can only "see" that in 3D when I grab the object and spin it. The ability to manipulate an object or part of an object is so slow and cumbersome in a dedicated CAD software (I have tried them all), they are less useful to me for prototyping/one offs.

I did start teaching myself Fusion but then they took away another attractive option, generative design. That feature was not only a big motivation for me to learn it but also for Autodesk to unilaterally castrate the free feature set was a worry for the longer term. I am sure (relatively sure) I could get some speed out of Fusion after say a year or so but I don't want to spend a year teaching myself something that suddenly is no longer free.

Fusion killing off the hobby license is why I'm back to freecad, I'm actually enjoying it with the realthunder mod.

I'm lucky I guess that this style of modelling is Intuitive so I can break my thoughts into sketches.

 

[ukracer]

https://www.amazon.co.uk/3Dconnexion-3DX-700059-Spacemouse-Compact-Mouse/dp/B079Z3T2XC/ref=asc_df_B079Z3T2XC/?tag=googshopuk-21&linkCode=df0&hvadid=310552784321&hvpos=&hvnetw=g&hvrand=3022313139058374057&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=1006506&hvtargid=pla-489967759395&psc=1&th=1&psc=1

Think I paid 60

How did you get that for 60?

 

[eLuSiVeMiTe]

How did you get that for 60?

eBay a few years back

 

[ukracer]

https://www.amazon.co.uk/3Dconnexion-3DX-700059-Spacemouse-Compact-Mouse/dp/B079Z3T2XC/ref=asc_df_B079Z3T2XC/?tag=googshopuk-21&linkCode=df0&hvadid=310552784321&hvpos=&hvnetw=g&hvrand=3022313139058374057&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=1006506&hvtargid=pla-489967759395&psc=1&th=1&psc=1

Think I paid 60

I can get one of these locally for 80 quid.